Ⅰ. Introduction. National Laboratory of Biomacromelecules. Director Zihe Rao

Transcription

1 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 Ⅰ. Introduction Under the auspices of the National Committee of Planning and Development and the Chinese Academy of Sciences, the National Laboratory of Biomacromolecules was founded in 1989, and formally opened in Jan Over the past decade, the NLB successfully completed many fundamental research projects of national significance, accumulated talents, equipments, technologies, and experiences in the area of protein science. The NLB was repeatedly ranked exceptional in 1989, 1991, 1996, 2001 and 2004 in the evaluations of all National Laboratories in biological sciences organized by the Ministry of Science and Technology. In 1994, the 10 th anniversary of the inauguration of the National Key Laboratories, the NLB was awarded one of the outstanding groups and was selected as one of the first five labs to undertake further reform. In 1996, the lab was commended as the outstanding lab by the Bureau of Human resources, Chinese Academy of Sciences. Because of its excellent performance, the NLB was one of the first to enter the knowledge innovation program implemented by Chinese Academy of Sciences in Beijing in In 2004, when celebrating 20 th anniversary of inauguration of the National Laboratories, the NLB was again honored with a Golden Bull award as an outstanding group in National Key Laboratories has seen the change of leadership and Academic Committee of the NLB. Under the new leadership and academic committee, the NLB implemented a strategy Dynamic, Open, Collaborative, and Competitive, aiming to fulfill the important national needs and to reach frontlines of international science and technology. It is clear that the NLB will be continuously led by the Honorary Director, Prof. Chen-lu Tsou and other two senior academics Prof. Dong-cai Liang and Prof. Fu-yu Yang, built on the core of protein science, the NLB will actively pursue multidisciplinary study of protein 3-dimentional structure and function, structure and function of biological membrane and membrane proteins, function and folding principles of proteins, molecular basis of immunology and infectious diseases, molecular neurobiology, and systems biology. In the end of 2004, the NLB are carrying out 47 projects, with funds totaling 31.3 M, from various sources including the Ministry of Science and Technology, Chinese Academy of Sciences and other national funding agencies. The NLB adheres to the concept Human resource is the most precious resource, puts the highest priority in management of human resources, combines the action of attracting and training, stabilizing and metabolizing. The NLB has recruited a large number of outstanding young and mid-aged scientists to play critical roles, among them are receivers of Hundred Talents Plan of the Chinese Academy of Sciences, National Outstanding Young Researcher Award from the National Science Foundation China, and Cheng-gong professors of the Ministry of Education. A creative team based on graduate students and postdocs, led by first class scientist has emerged in the NLB. The NLB has integrated existing equipments, systematically constructed key infrastructure, combining with the platform of protein science in the hosting Institute of Biophysics. The NLB has fully prepared and has built the sound foundation for establishment of the National Lab for Protein Sciences, further facilitate the systematic and large scale development of the protein sciences of our country. National Laboratory of Biomacromelecules Director Zihe Rao 1

2 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 Ⅰ. 前言 生物大分子国家重点实验室于 1989 年经国家计划发展委员会和中国科学院批准成立,1991 年 1 月通过验收并正式对外开放运行 在十几年的成长历程中, 实验室承担了多项国家重大科研任务, 在蛋白质科学研究领域已形成了人才 装备 技术 经验的综合优势并取得了骄人业绩 在国家科学技术部对全国生物学科国家重点实验室和部门开放实验室的历次评估中, 生物大分子国家重点实验室分别于 1989 年 1991 年 1996 年 2001 年和 2004 年五次被评为优秀 ;1994 年我室被评为先进集体的同时被科技术部选为全国首批五个国家重点实验室试点单位之一 ;1996 年我室又被中国科学院人事局评为先进集体 ;1999 年作为优秀实验室首批进入了中国科学院在京区启动的生物科学的开放实验室知识创新试点 2004 年在国家重点实验室成立二十周年之际, 我室再度被国家科学技术部评为 国家重点实验室计划先进集体 称号, 荣获金牛奖 2003 年实验室进行了室领导及学术委员会的换届, 在新一届室领导和学术委员会的带领下本着 流动 开放 联合 竞争 的办室方针, 面向国家发展战略需求和国际科技发展前沿, 结合实验室多年来的工作积累, 明确地提出了实验室的定位 : 实验室将继续在我国生物物理和生物化学界前辈科学家邹承鲁 梁栋材 杨福愉三位先生的指导和带领下, 以蛋白质科学为核心, 发挥多学科交叉综合的优势, 结合人类健康相关的国家需求, 围绕生物膜与膜蛋白功能与结构 蛋白质功能与折叠原理 蛋白质三维结构与功能, 感染与免疫 系统生物学等重大科学问题开展原创性研究 截止 2004 年底实验室承担国家 科学院的在研任务 47 项, 从科技部和中科院得到的运行经费和各课题组在研研究经费总额达 3130 万元 生物大分子国家重点实验室坚持 人才是第一资源 的理念, 将人才工作作为各项工作的重中之重 实验室坚持引进与培养相结合 稳定与流动相统一的原则, 广纳优秀人才加盟, 引进了中国科学院 百人计划 入选者 国家杰出青年基金获得者和教育部 长江学者奖励计划 特聘教授等一大批优秀中青年科学工作者来室担当重任 形成了以一流领衔科学家为核心, 以一批有突出贡献的中年学术骨干和青年创新人才为主体, 以研究生 博士后为基础的金字塔形的创新人才队伍 生物大分子国家重点实验室在仪器设备方面整合了自有的仪器设备, 依靠中科院生物物理研究所的蛋白质科学研究平台, 通过学科调整和蛋白质科学关键仪器设备的系统性建设, 将为建设蛋白质科学国家实验室奠定良好基础, 并以此推进我国蛋白质科学研究走规模化 体系化的发展道路 生物大分子国家重点实验室 主任饶子和 2

3 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 Ⅱ. Organization 1. Honorary Directors Chenlu Tsou Dongcai Liang Fuyu Yang 2. Director and Vice Directors Director Vice Directors Zihe Rao Tao Xu Weimin Gong Xianming Pan Institute of Biophysics CAS Institute of Biophysics CAS Institute of Biophysics CAS Institute of Biophysics CAS Institute of Biophysics CAS Institute of Biophysics CAS Institute of Biophysics CAS 3 Academic Council: Chairman: Chihchen Wang Member of CAS Institute of Biophysics CAS Beijing Members: Jun Yu Professor Institute of Genomics CAS Beijing Zhixin Wang Member of CAS Institute of Biophysics CAS Beijing Zhaohui Ye Member of CAS Wuhan Institute of Physics and Mathematics CAS Wuhan Xianen Zhang Professor Ministry of Science and Technology Beijing Boliang Li Professor Shanghai Institute for Biological Sciences CAS Shanghai Jiayang Li Member of CAS Institute of Genetics and Developmental Biology CAS Beijing Qishui Lin Member of CAS Shanghai Institutes for Biological Sciences CAS Shanghai Yunyu Shi Member of CAS School of Life Science, University of Science and Hefei Technology of China Fuchu He Member of CAS Academy of Military Medical Science Beijing Zihe Rao Member of CAS Institute of Biophysics CAS Beijing Tao Xu Professor Institute of Biophysics CAS Beijing Aike Guo Member of CAS Institute of Biophysics CAS Beijing Guangxia Gao Professor Institute of Microbiology CAS Beijing Wenrui Chang Professor Institute of Biophysics CAS Beijing Xiyun Yan Professor Institute of Biophysics CAS Beijing Boqin Qiang Member of CAS Institute of Basic Medical Research CAMS Beijing Academic Secretary: Xiyun Yan 3

4 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 Ⅱ. 机构 1. 名誉主任 2. 主任 副主任主任 副主任 3. 学术委员会 邹承鲁梁栋材杨福愉饶子和徐涛龚为民潘宪明 中科院生物物理所中科院生物物理所中科院生物物理所中科院生物物理所中科院生物物理所中科院生物物理所中科院生物物理所 主任 : 王志珍 中国科学院院士 中科院生物物理所 北京 委员 : ( 以姓氏笔画为序 ) 于军 研究员 中科院北京基因组所 北京 王志新 中国科学院院士 中科院生物物理所 北京 叶朝辉 中国科学院院士 中科院武汉物理与数学研究所武汉 张先恩 研究员 科技部 北京 李伯良 研究员 上海生命科学研究院 上海 李家洋 中国科学院院士 中科院遗传发育所 北京 林其谁 中国科学院院士 上海生命科学研究院 上海 施蕴渝 中国科学院院士 中国科技大学生命科学院 合肥 贺福初 中国科学院院士 军事医学科学院 北京 饶子和 中国科学院院士 中科院生物物理所 北京 徐涛 研究员 中科院生物物理所 北京 郭爱克 中国科学院院士 中科院生物物理所 北京 高光侠 研究员 中科院微生物所 北京 常文瑞 研究员 中科院生物物理所 北京 阎锡蕴 研究员 中科院生物物理所 北京 强伯勤 中国科学院院士 中国医学科学院基础所 北京 学术秘书 : 阎锡蕴 ( 兼 ) 4

5 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES International Scientific Committee The International Scientific Committee is composed of eminent overseas scientists whose expertise is in fields closely related to the research of the Laboratory. The purpose of the International Committee is to advise on the priorities and direction for research, to maintain international scientific dialogue and exchange, and to allow appraisal of the Laboratory against an international standard. Chairman: Robert Huber (Nobel Prize Laureate) Max Plank Institute for Biochemistry (Germany) Vice Chairman: Erwin Neher (Nobel Prize Laureate) Max Plank Institute for Biophysical Chemistry (Germany) Members: Giuseppina Barsacchi Laboratori di Biologia Celluare e dello Sviluppo (Italy) Guy Dodson University of York (UK) Khalid Iqbal NYS Institute for Basic Research (USA) Neil Isaacs University of Glasgow (UK) Jack Johnson The Scripps Research Institute (USA) David Stuart University of Oxford (UK) Joel Sussman The Weizmann Institute of Science (Israel) Michael G. Rossmann Purdue University, (USA) Jack Johnson The Scripps Research Institute (USA) Bi-Cheng Wang University of Georgia, (USA) Brian Matthews University of Oregon (USA) Louise N. Johnson University of Oxford (UK) Neil Isaacs University of Glasgow (UK) Johannes Frederik Gerardus Vliegenthart Bijvoet Center for Biomolecular Research (Netherlands) 5 Research Faculty Chang Chen Hongyu Deng Zusen Fan Guangxia Gao Weimin Gong Haiying Hang Shigang He Guangju Ji Taijiao Jiang Tao Jiang Renjie Jiao Gang Jin Sarah Perrett Wei Liang Yingfang Liu Xianming Pan Zhihai Qin Zihe Rao Hong Tang Jie Tang Jinhui Wang Shengdian Wang Yi Wang Zhixin Wang Zhizhen Wang Zhongju Xiao Tao Xu Xiyun Yan Fuquan Yang Qinwei Yin Xianen Zhang Xujia Zhang 6 Senior Faculty Wenrui Chang Jianwen Chen Youguo Huang Guozhong Jing Dongcai Liang Jinfeng Wang Fuyu Yang Junmei Zhou Chenlu Tsou 5

6 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 国际科学委员会 为了与国际同行保持经常联系, 及时了解国际研究动向, 掌握和确定研究战略, 组织本实验室的国际同行评估, 本实验室成立了由国外科学家组成的 国际科学委员会 主席 : Robert Huber ( 诺贝尔奖获得者 ) Max Plank Institute for Biochemistry (Germany) 副主席 : Erwin Neher ( 诺贝尔奖获得者 ) Max Plank Institute for Biophysical Chemistry (Germany) 委员会成员 : Giuseppina Barsacchi Laboratori di Biologia Celluare e dello Sviluppo (Italy) Guy Dodson University of York (UK) Khalid Iqbal NYS Institute for Basic Research (USA) Neil Isaacs University of Glasgow (UK) Jack Johnson The Scripps Research Institute (USA) David Stuart University of Oxford (UK) Joel Sussman The Weizmann Institute of Science (Israel) Michael G. Rossmann Purdue University, (USA) Jack Johnson The Scripps Research Institute (USA) Bi-Cheng Wang University of Georgia, (USA) Brian Matthews University of Oregon (USA) Louise N. Johnson University of Oxford (UK) Neil Isaacs University of Glasgow (UK) Johannes Frederik Gerardus Vliegenthart Bijvoet Center for Biomolecular Research (Netherlands) 5 实验室固定成员 :( 以拼音为序 ) 陈畅 邓红雨 范祖森 高光侠 龚为民 杭海英 何士刚 姬广聚 蒋太交 江涛 焦仁杰 靳刚 柯莎 梁伟 刘迎芳 潘宪明 秦志海 饶子和 唐宏 唐捷 王晋辉 王盛典 王毅 王志新 王志珍 肖中举 徐涛 阎锡蕴 杨福全 殷勤伟 张先恩 张旭家 6 实验室资深人员 :( 以拼音为序 ) 常文瑞 陈建文 黄有国 静国忠 梁栋材 王金凤 杨福愉 周筠梅 邹承鲁 6

7 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 Ⅲ. Management 1 Management of the NLB The NLB is managed strictly according to the Temporary Regulations of Establishment and Management of National Key Laboratories. The director is responsible for running the NLB, and the academic committee is responsible for evaluation. The NLB is organized according to the research directions based on research unit. An academic conference is organized every year in Feb. to March. The director of the NLB presents a summary of the previous year and a proposal for running and expenditure of current year. 2. Academic committee The Annual Meeting convened by the Academic Council provides a forum to discuss and appraise the research of the Laboratory. At this meeting, ongoing research projects are assessed, new research proposals are considered for approval, the accounts are inspected, the research achievements of the Laboratory are examined and the priorities and direction for future research are set. 3. Research directions 1) Protein science and catalytic enzyme 2) Three dimensional structure and function of bio-macro molecules 3) Biology of membrane molecules 4) Molecular basis of immunology and infectious diseases 5) Molecular neuroscience 4. Management of projects Members of the NLB submits following materials to the secretary of the NLB: 1) All publications (a reprint and an electronic version) with the NLB in corresponding address, including papers, book chapters, and publications in national and international conferences 2) Photocopy of national and international awards received in past 12 months 3) Progress report (with a Chinese and an English version) 7

8 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 Ⅲ. 管理 1 实验室管理实验室依据 国家重点实验室建设与管理暂行办法 对实验室进行全面管理 本室实行实验室主任负责制 学术委员会评审制 ; 实验室采用以研究方向为导向, 以研究组为基本科研单元的运行模式, 每年 2-3 月间召开上年度学术年会 实验室主任就年度运行管理 经费使用提出可行方案, 并向学术委员会汇报上年度的实验室工作总结 2 学术委员会学术委员会结合本室学术年会每年召开全体会议一次, 评议实验室的工作, 内容包括确定实验室研究方向 制定及修改课题指南 审批课题申请 检查课题进展情况 监督经费使用 评审科研成果及审议学术活动计划等 3 研究方向 1) 蛋白质和分子酶学 2) 生物大分子三维结构和功能 3) 膜分子生物学 4) 感染与免疫的分子基础 5) 分子神经生物学 4 课题管理实验室成员每年应按时向实验室秘书提交以下材料 : 1) 当年发表的具有 生物大分子国家重点实验室 署名的全部著作目录 ( 包括专著 论文, 国际及全国性学术会议论文等 ), 并提交版面清楚平整的论文单印本一式一份及论文电子版 ; 2) 当年获得国际 国家或省部级科技奖励的证书复印件一份 ; 3) 年度工作报告 ( 中 英文各一份 ) 的电子文档 报告格式按实验室秘书提供的文档模版填写 8

9 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 Ⅳ. Summary of Research in 2004 Inspired by the mortem of the NLB rigorous, diligent, devoted, all members are working extremely hard and the NLB is making great progresses in Especially in the aspect of grant application, the the National Basic Research Program : Structure and Function of Bio-membrane and Membrane Proteins, led by Dr. Tao Xu has been approved by the MOST, with a total of 25M. In the the National Basic Research Program : Human Liver Structural Proteomics and New Methods and Techniques of Proteomics, a sub-project: Human Liver Structural Proteomics, led by Dr. Wei-min Gong, has been approved, with a total of 9M. A project Single Molecule Detection of Cell Membrane Protein CD146 and Molecular Recognition of Its Ligand and Signal Trunsduction in Living Cells led by Dr. Xi-yun Yan has been approved by the NSFC as a key research project. The NLB has been awarded the outstanding group and the Golden Bull award; the project led by Dr. Wen-rui Chang: Three Dimensional Structure and Function of Photon Catching Complex of Algae Photosynthesis has won the 1 st prize of Science and Technology in Beijing. This project has significant implication for the mechanism of photosynthesis and energy regeneration and usage. Dr. Zi-he Rao has been elected a member of Third World Academy and the outstanding personnel of 973. Dr. Wen-rui Chang has been awarded He Liang and He Li award for progress of science and technology, and the outstanding personnel of 973. The NLB has won research grants totaling 29M. The NLB currently has 8 postdoc, 58 Ph.D. students and 45 Master students. 1 Research Funds Running expenditure from the MOST and CAS: 2.3M. Total research grants: 29M. 2 Research Projects There are 47 items of scientific research programs which are taking on in the NLB till the end of the year of 2004, including the National Basic Research Program (973), 2 items (13 projects); the National High Technology Research and Development Program of China (863), 7 items; the Key Technologies R&D Program, 3 items; 3 items of the Major Program of the National Natural Science Foundation of China, 2 items of the Key Program, 1 item of the National Science Fund for Distinguished Young Scholars, 1 item of the State Key Laboratories Development Program, 9 items of the General Program; 13 items of the Knowledge Innovation Program of Chinese Academy of Sciences, and so on. 9

10 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 Ⅳ 年工作总结 生物大分子国家重点实验室在过去的一年里全室人员在严谨 勤奋 求精献身精神鼓舞下, 团结协作, 努力工作使得全室各项工作得以顺利开展, 完成情况良好 特别是今年在项目申请方面由徐涛研究员作为首席科学家申报的 国家重点基础研究发展规划 973 项目 : 生物膜和膜蛋白的结构与功能研究 获得科技部批准, 项目总经费 2500 万元 ; 在 973 项目 : 人类肝脏结构蛋白质组和蛋白质组新技术新方法研究 中, 龚为民研究员申报的子课题 : 人类肝脏结构蛋白质组学研究获得批准, 项目总经费 900 万元 ; 阎锡蕴研究员关于 活细胞单分子探测细胞膜蛋白 CD146 及其配体的分子识别和信号传导机制 研究项目也获得了基金委重点项目经费的支持 在成果取得方面, 生物大分子国家重点实验室被授予国家重点实验室计划先进集体荣誉称号并获金牛奖 ; 常文瑞课题组主持完成的 藻类光合作用捕光复合物的三维结构与功能研究 获北京市科学技术奖一等奖, 该项成果对于光合作用的机理研究和可再生能源的开发利用具有重要意义 ; 饶子和院士当选为第三世界科学院院士并获得 973 计划先进个人荣誉称号 ; 常文瑞研究员获得了 2004 年度何梁何利基金科学与技术进步奖,973 计划先进个人荣誉称号等 在经费争取方面, 我室今年在研科研项目经费总额达 2900 万元 在人才培养方面, 实验室设有博士后流动站, 在站博士后 8 人, 在读博士生 58 人, 硕士生 45 人 1 经费科技部和中科院支持的运行经费 :230 万元在研研究经费总额达 2900 万元 2 项目 2004 年底实验室承担的在研任务 47 项, 包括 973 计划主持项目 2 项, 子课题 13 项 ; 863 计划课题 7 项 ; 国家科技攻关计划课题 3 项 ; 科技部基础研究快速反应支持项目 1 项 ; 国家自然科学基金重大项目 3 项 重点项 2 项 国家杰出青年基金 1 项 优秀国家重点实验室基金 1 项 面上项目 9 项 ; 中科院创新工程重大项目课题 4 项 创新工程重要方向项目课题 8 项 生物科学与技术研究特别支持项目 2 项 中科院创新工程领域前沿项目 1 项, 院长基金特别支持项目 1 项, 院生物局特别支持项目 1 项 10

11 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES GRADUATE STUDENTS ⑴ Post-Doctoral Fellows Previous: Zhanfen Qin Xuesong Chen Current: Hongmei Zhang Xianzhi Dong Jianqiang Ni Changcheng Yin Chunling Zhang Xiaoxue Yan Tongbiao Zhao Yan Gan ⑵ Doctoral Students Students Conferred with the Ph.D Degree in 2004: Ling Yan Qiu Cui Lingyun Wang Chuanpeng Liu Chuanxi Cai Zhenfeng Liu Feng Wang Caihong Yun Yu Cao Yingang Feng Sun Huang Zhuo Li Dongsheng Liu Yongfang Zhao Yunzheng Zhao The National Laboratory of Biomacromolecules currently has 58 Doctoral students. ⑶ Master s Students Students Conferred with the Master Degree in 2004:Su Xu The National Laboratory of Biomacromolecules currently has 45 Master s students. ⑷ Student Awards 1) CAS Director s Scholarships were awarded to Yi Jiang and Hanchi Yan 2) Second Class Di-Ao Scholarships were awarded to Guoping Ren ⑸ Graduate Student Supervisor Awards 1)Prof. Chihchen Wang was awarded a Procter & Gamble Outstanding Graduate Student Supervisor Award in )Prof. Xiyun Yan was awarded a Procter & Gamble Outstanding Graduate Student Supervisor Award in

12 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 人才 ⑴ 博士后 出站博士后 : 秦占芬 在站博士后张红梅 闫小雪 陈雪松 董先智倪建强尹长城张春玲 赵同标甘燕 ⑵ 博士生 2004 年取得博士学位 15 人 : 闫玲崔球王凌云刘川鹏蔡传喜柳振峰王锋云彩红曹禹冯银刚黄隼李卓刘东升赵永芳赵云罡在读博士生 58 人 ⑶ 硕士生 2004 年取得硕士学位 1 人 : 徐苏硕士生 45 人 ⑷ 研究生获奖情况博士生江轶 严汉池获中国科学院院长优秀奖博士生任国平获中国科学院地奥奖学金二等奖 ⑸ 研究生导师获奖情况王志珍院士获中国科学院宝洁优秀研究生导师奖阎锡蕴研究员获中国科学院宝洁优秀研究生导师奖 12

13 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES Scientific Awards (1) Wenrui Chang 1 st Beijing municipal award of science and technology: Studies on 3D-structure of photosynthetic membrane proteins-crystal structure and function of spinach major light harvesting complex (2) Zihe Rao Elected as member of the third world academy of science (3) National Key Laboratory of Biomacromolecules Named Excellent Team of National Key Laboratory Program and awarded Golden Cattle Trophy (4) Wenrui Chang Won Liliang He Progress Award of Science and Technology of year 2004 (5) Wenrui Chang Named Excellent Scientist of National Loboratory by Ministry of Science and Technology of China (6) Wenrui Chang Won Award of Excellent Scientist of 973 project from Ministry of Science and Technology of China (7) Wenrui Chang Won Award of Excellent Scientist of 973 project from Ministry of Science and Technology of China 13

14 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 获奖情况 (1) 常文瑞等完成的 藻类光合作用捕光蛋白 - 色素复合物的三维结构与功能研究 成果获北京市科学技术奖一等奖 ( 图一 ) (2) 饶子和院士当选为第三世界科学院院士 (3) 生物大分子国家重点实验室获 国家重点实验室计划先进集体 称号并获金牛奖 (4) 常文瑞研究员获 2004 年度何梁何利基金科学与技术进步奖 ( 图二 ) (5) 常文瑞研究员获国家科技部颁发的 国家实验室先进个人 称号 ( 图三 ) (6) 饶子和院士获国家科技部颁发 973 计划先进个人 奖 ( 图四 ) (7) 常文瑞研究员获国家科技部颁发 973 计划先进个人 奖 ( 图五 ) 图一 图二 图三 图四 图五 14

15 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES International Exchance (1) International Metting Held 10th International Conference on Crystallography Beijing China June, 2004 (2) Participation in International Meetings Junmei Zhou Physical Aspects of Protein Folding Invited Speaker Time: 1/5-8/2004, Kyoto, Japan Topic: Folding and Chaperone Function of Escherichia Coli Trigger Factor Zihe Rao British Royal Society SARS Research Meeting Speaker Time: 1/12-15/2004, Landon, UK Topic: The The crystal structures of SARS virus main protease Mpro and its complex with an inhibitor Tao Xu Forty-eighth Research Meeting of Biophysics Society Speaker Time: 2/14-18/2004, Maryland, USA Topic: Regulation of Insulin Secretion by Extracellular ATP in Rat pancreatic beta cells Zihe Rao Sino-Japan SARS Research Meeting (organized by CAS) Speaker Time: 2/20-26/2004, Yokohama, Japan Topic: The The crystal structures of SARS virus main protease Mpro and its complex with an inhibitor Xiyun Yan The International Workshop on SARS Speaker Time: 2/23-24/2004, Japan Topic: Probing the structure of the SARS coronavirus using SEM Zihe Rao First Pacific Rim Conference on Protein Science Speaker Time: 4/14-20/2004, Yokohama, Japan Topic: The The crystal structures of SARS virus main protease Mpro and its complex with an inhibitor Sarah Perrett First Pacific Rim Conference on Protein Science Time: 4/14-18/2004, Japan Poster: Amyloid nucleation and hierarchical assembly of Ure2p fibrils Zihe Rao International Conference on SARS-one year after the (first) outbreak Speaker Time: 5/8-11/2004, Lubeck, German Topic: Structural Genomics Study of SARS coronavirus-the The crystal structures of SARS virus main protease Mpro and its complex with an inhibitor 15

16 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 国际交流 (1) 主办国际会议 第十届国际结晶学大会 2004 年 6 月中国北京 (2) 参加国际会议周筠梅 Physical Aspects of Protein Folding, Kyoto 时间 :2004 年 1 月 5-8 日, 日本邀请报告 :Folding and Chaperone Function of Escherichia coli Trigger Factor 饶子和英国皇家学会 SARS 研讨会大会报告时间 :2004 年 1 月 日, 英国伦敦题目 :Structural genomics study of the SARS coronavirus The Crystal Structures of SARS Virus Main Protease Mpro and Its Complex with an Inhibitor 徐涛第 48 届生物物理学会年会专题报告时间 :2004 年 2 月 14-18, 美国 (Maryland Baltimore) 题目为 :Regulation of Insulin Secretion by Extracellular ATP in Rat pancreatic beta cells 饶子和中 - 日 SARS 研讨会 ( 中科院组团 ) 大会报告时间 :2004 年 2 月 日, 日本题目 :Structural genomics study of the SARS coronavirus The Crystal Structures of SARS Virus Main Protease Mpro and Its Complex with an Inhibitor 阎锡蕴 The International Workshop on SARS,Tokyo, Japan 时间 :2004 年 2 月 日, 日本题目 :Probing the structure of the SARS coronavirus using SEM 饶子和 The 1st Pacific-Rim International Conference on ProteinScience 大会报告时间 :2004 年 4 月 日, 日本横滨题目 :Structural genomics study of the SARS coronavirus The Crystal Structures of SARS Virus Main Protease Mpro and Its Complex with an Inhibitor 柯莎 The1st Pacific Rim Conference on Protein Science 时间 :2004 年 4 月 14 日 -18 日, 日本墙报 : Amyloid nucleation and hierarchical assembly of Ure2p fibrils 饶子和 International Conference on SARS - one year after the (first)outbreak 大会报告时间 :2004 年 5 月 8-11 日德国 (Lubeck) 题目 :Structural genomics study of the SARS coronavirus The Crystal Structures of SARS Virus Main Protease Mpro and Its Complex with an Inhibitor 16

17 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 Qinwei Yin 2 nd RNAi International Conference Time: 5/10-13/2004, USA Poster: Effective suppression of proliferation of human melanoma cells by a combination of sirna molecules Chih-chen Wang 4 th International Workshop on the molecular biology of stress responses Invited Speaker Time: 5/13-16/2004, Wuhan and Yichang, China Topic: Dimerization by Domain Hybridization Bestows Chaperone and Isomerase Activities Xianen Zhang 8 th World Conference on Biosensors Speaker and Chair for enzyme sensors workshop Time: 5/24-26/2004, Spain Topic: Gene Technology: Opportunity in Biosensors Zihe Rao Tenth International Conference on Crystallography Speaker Time: 6/5-8/2004, Beijing, China Topic: A perspective on Structural Genomics efforts in China-a report, review and revision Wenrui Chang Tenth International Conference on Crystallography Invited Speaker Time: 6/5-8/2004, Beijing, China Topic: Crystal structure of spinach major light harvesting complex at 2.72 Å resolution Junmei Zhou FASEB Conference: Protein Misfolding, Amyloid and Conformational Disease Time: 6/12-17/2004, Colorado, USA Poster: Amyloid nucleation and hierarchical assembly of Ure2 fibrils: role of Asn/Gln repeat and non-repeat regions of the prion domain Sarah Perrett FASEB Conference: Protein folding in the cell Time: 7/31-8/5/2004, Vermont, USA Poster: Folding, misfolding and fibril formation of yeast prion protein Ure2 17

18 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 殷勤伟第二届 RNAi 国际会议时间 :2004 年 5 月 日, 美国墙报 :Effective suppression of proliferation of human melanoma cells by a combination of sirna molecules 王志珍 4 th International Workshop---On the molecular biology of stress responses, Wuhan and Yichang 时间 :2004 年 5 月 日邀请报告 : Dimerization by domain hybridization bestows chaperone and somerase activities. 张先恩第八届世界生物传感器大会分组口头报告并主持酶传感器分会会场时间 :2004 年 5 月 日, 西班牙题目 :Gene Technology: Opportunity in Biosensors 饶子和第 10 届国际晶体学大会 ICCBM10 大会主席分会场报告时间 :2004 年 6 月 5-8 日, 中国北京题目 :A perspective on Structural Genomics efforts in China a report, review and revision 常文瑞第十届国际结晶学大会大会特邀报告时间 :2004 年 6 月 5-8 日, 中国北京题目 :Crystal structure of spinach major light harvesting complex at 2.72Å resolution 周筠梅 FASEB Conference: Protein Misfolding, Amyloid and Conformational Disease, Colorado 时间 :2004 年 6 月 12 日 -17 日, 美国墙报 : Amyloid nucleation and hierarchical assembly of Ure2 fibrils: role of Asn/Gln repeat and non-repeat regions of the prion domain. 柯莎 FASEB Conference: Protein Folding in the Cell, Vermont 时间 :2004 年 7 月 31 日 -8 月 5 日, 美国墙报 :Folding, misfolding and fibril formation of the yeast prion protein Ure2 18

19 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 Xiyun Yan 2004 Asian-Pacific Rim Conference on Biology of Tumor Markers and 21th International Workshop on Tumor Markers Speaker Time: 8/21-25/2004, Xian, China Topic: A novel anti-cd146 antibody inhibits angiogenesis and tumor growth Wenrui Chang Thirteenth International Conference on Photosynthesis Invited Speaker Time: 8/29-9/3/2004, Canada Topic: Crystal structure of spinach major light harvesting complex at 2.72 Å resolution Zhihai Qin International Conference on Immune Escape of Tumors Speaker Time: 10/10-14/2004 Salzburg, Austria Topic: Diverse immunological mechanisms against transplanted and chemical carcinogen-induced tumors Zihe Rao Conference on Molecular Aspects and Prevention of SARS Speaker Time: 10/16-20/2004, Madrid, Spain Topic: Structural genomics study of the SARs-CoV. The crystal structure of the main protease and its inhibitors Xiyun Yan The first symposium between China and Japan on SARS Speaker Time: 10/17-18/2004, Beijing, China Xianen Zhang Topic: Generating SARS antibody library for SARS diagnosis and therapy 8 th China-Japan-South Korea Enzyme Engineering Conference Speaker Time: 10/24-27/2004, Hangzhou, China Topic: A Muts-based Protein Chip for Detection of DNA Mutations Chih-chen Wang HUPO3 rd Annual World Congress, Proteomics: decoding the genome Speaker Time: 11/2004, Beijing, China Topic: Thiol-Protein Oxidoreductases as Molecular Chaperones Xiyun Yan Immunotherapy for the New Century Speaker Time: 11/15-19/2004, Havana, Cuba Topic: A novel anti-cd146 antibody inhibits angiogenesis and tumor growth Zihe Rao Structrral Genomics & Proteomics EU Projects Meeting Speaker Time: 12/1-4/2004, Spain Topic: A perspective on Structural Genomics efforts in China 19

20 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 阎锡蕴 2004 亚太地区国际肿瘤生物学和医学暨第 21 届国际肿瘤标志物学 术会议 时间 : 2004 年 8 月 日, 中国西安 题目 :A novel anti-cd146 antibody inhibits angiogenesis and tumor growth 常文瑞第十三届国际光合作用大会特邀报告 时间 :2004 年 8 月 29 日 -9 月 3 日, 加拿大 题目 :Crystal structure of spinach major light-harvesting complex at 2.72Å resolution 秦志海 国际肿瘤免疫逃逸专题会议 上作专题报告 时间 :2004 年 10 月 10 日 -14 日, 奥地利 (Salzburg) 题目 :Diverse immunological mechanisms against transplanted and chemical carcinogen-induced tumors. 饶子和 Molecular Aspects and Prevention of SARS 大会报告时间 :2004 年 10 月 日西班牙马德里题目 :Structural genomics study of the SARS-CoV. The crystal structure of the main protease and its inhibitors 阎锡蕴 The first symposium between China and Japan on SARS,Beijng China, 时间 :2004 年 10 月 日, 北京题目 :Generating SARS antibody library for SARS diagnosis and therapy 张先恩第八届中日韩酶工程大会报告时间 :2004 年 10 月 日, 杭州题目 :A MutS-based Protein Chip for Detection of DNA Mutations 王志珍 HUPO3 rd Annual World Congress, Proteomics: decoding the genome 时间 :2004 年 11 月, 北京题目 : Thiol-Protein Oxidoreductases as Molecular Chaperones 阎锡蕴 Immunotherapy for the New Century, Havana, Cuba, 时间 :2004 年 11 月 日题目 :A novel anti-cd146 antibody inhibits angiogenesis and tumor growth 饶子和 Structural Genomics & Proteomics EU Projects Meeting 时间 :2004 年 12 月 1-4 日, 西班牙题目 :A perspective on Structural Genomics efforts in China 20

21 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 ⑶Visits Abroad Tao Xu NIH, USA, 2/2004 Wenrui Chang Wenrui Chang Invited speech at Verona University, Italy Invited speech at Sheffield University, UK Xiyun Yan Xiyun Yan Tour to National Institute of Nanotechnology and Universities in Canada, sponsored by Ministry of Science and Technology of China, 9/ /2004 Tour to Daikin Company in Japan for the arrangement of research cooperation, 10/2004 Chih-chen Wang Medical School, Hong Kong University, 10/15/2004 Topic: Thiol-Protein Oxidoreductases as Molecular Chaperones Chih-chen Wang Chinese University of Hong Kong, 10/18/2004 Topic: Thiol-Protein Oxidoreductases as Molecular Chaperones ⑷Visitiors from the Taiwan Area and Abroad David Westhead Jim Larrick Hoichi Yorinaka Tomomichi Ono Doctor, Leeds University, UK Time: 2/16-22/2004 Doctor, Panorama Research Institute, USA Time: 3/19-24/2004 Vice President of Xiongben University, Japan Time: 3/29/2004 Professor, Medical school, Xiongben University, Japan Time: 3/29/2004 T. Hoshino Rigaku Company, Japan David Stipp Time: 3/31/2004 Fortune magazine, USA Time: 5/26/2004 Orla M. Smith Valda Vinson Science journal, USA Time: 5/27/2004 Jack Johnson & Tianwei Lin Drs., The Scripps Research Institute Tom Blundell Michael G Rossmann Time: 6/3-4/2004 Chair of School of Biological Sciences, University of Cambridge Time: 6/7/2004 Doctor, Purdue University Time: 6/8/

22 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 ⑶ 对国外及港澳台地区的访问徐涛美国 National Institutes of Health, 常文瑞意大利 verona 大学邀请报告常文瑞英国 Sheffield 大学邀请报告阎锡蕴 2004 年 9 月 30 日至 10 月 13 日, 受加拿大国家研究院国家纳米技术研究所 (National Institute of NanoTechnology) 的邀请, 参加科技部组织的纳米科技代表团, 访问了加拿大的有关研究机构和大学 阎锡蕴 2004 年 10 月, 受日本大金公司邀请, 访问该公司并开展 SARS 防治方面的研究 王志珍 2004 年 10 月 15 日香港大学医学院并做题为 : Thiol-Protein Oxidoreductases as Molecular Chaperones 报告王志珍 2004 年 10 月 15 日香港中文大学并做题为 : Thiol-Protein Oxidoreductases as Molecular Chaperones 报告 ⑷ 来我室访问的外国和港澳台地区科学家 年 2 月 日英国 Leeds 大学 David WESTHEAD 博士来访 年 3 月 日美国 Panorama Research Institute 的 Jim Larrick 博士来访 年 3 月 29 日日本雄本大学 Hoichi Yorinaka 副学长与雄本大学药学院 Tomomichi Ono 教授等来访 年 3 月 31 日日本 T. Hoshino 日本理学公司来访 年 5 月 26 日美国 Fortune 杂志的 David Stipp 来访 年 5 月 27 日美国 Science 杂志的 Orla M. Smith Valda Vinson 来访 年 6 月 3-4 日美国 The Scripps Research Institute,Jack Johnson 林天伟博士来访 年 6 月 7 日英国 School of Biological Sciences, University of Cambridge 的主席 Tom Blundell 博士来访 年 6 月 8 日美国 Purdue University 的 Michael G Rossmann 博士来访 22

23 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 John W. Kappler Pjilippa Marrack Member of British Royal Society, Investigator of Howard Hughes Medical Institute, Professor of National Jewish Medical research Center, Denver, Colorado, USA Time: 6/9-10/2004 Investigator of Howard Hughes Medical Institute, Professor of National Jewish Medical research Center, Denver, Colorado, USA Time: 6/9-10/2004 I. Raska Doctor, Pharmaceutical Institute, Academy of Sciences, Czech Time: 6/17/2004 British Visiting Group of Structural Genomics: Kim Watson, University of Reading; Dame Louise Johnson, University of Oxford; David Gillham, Syngenta Inc; Drake Eggleston, GlaxoSmithKline Inc; Peter Collins, Rigaku/MSC Inc; Paul Loeffen, University of Oxford; Jennifer Moynihan, University of Reading Yangxin Fu Professor, University of Chicago, USA Time: 7/2004 Speech topic: Tumor Immune Therapy Wenqin Xu Doctor, University of Washington, USA Time: 7/27/2004 Rongguang Zhang Argone National Laboratory, USA Time: 7/27/2004 Rudi Balling Director, German Research Center of Biotechnology Time: 7/27/2004 Zongchao Jia Doctor, Department of Chemistry, Queens University Time: 7/30/2004 Jie Zheng Professor, Department of Physiology and Membrane Biology, School of Medicine, University of California at Davis, USA Time: 8/2004 Topic: Molecular Motion of Ion Channel Zhinan Yin Professor of Immunology of Yale University, USA Time: 8/2004 Speech topic: T cell Function in Tumor Immunity Hospital Universitario de Canarias, Sapin Time: 9/1/2004 Kyogo Itoh Professor, Department of Immunology, School of Medicine, Kurume University, Japan Time: 9/12/2004 Takehisa Matsumoto Riken Inc., Japan Time: 10/25/2004 Visiting group from Academy of Sciences of Uzbekistan: seven people Time: 10/27/

24 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 年 6 月 9-10 日美国 Howard Hughes Medical Institute National,Jewish Medical and Research Center 的英国皇家科学院院士 John W. Kappler 和 Philippa Marrack 教授来访 年 6 月 17 日捷克科学院药物所 I. Raska 博士来访 年英国结构基因组学使团来访 (Dr Kim Watson The University of Reading,Prof Dame Louise Johnson University of Oxford,Dr David GillhamSyngenta, Dr Drake Eggleston GlaxoSmithKline,Dr Peter Collins Rigaku/MSC Inc,Dr Paul Loeffen Oxford,Diffraction Limited,Dr Jennifer Moynihan The University of Reading) 年 7 月份邀请 University of Chicago 大学傅阳心教授访问我所并做关于肿瘤免疫治疗的学术报告 年 7 月 27 日美国 University of Washington 的许文清博士来访 年 7 月 27 日美国 Argone National Laboratory 的张荣光博士来访 年 7 月 27 日德国 German Research Centre fro Biotechnology 的所长 Rudi Balling 来访 年 7 月 30 日加拿大女王大学生化系的贾宗超博士来访 年 8 月 16 日日本东京大学的 Aikichi Iwamoto 博士来访 年 8 月 27 日美国华盛顿大学 Fred Hutchinson Cancer Research Center 的诺贝尔奖获得者 Leland H.Hartwell 校长来访 年 8 月, 加州大学戴维斯分校医学院生理及膜生物学系教授郑来访, 并做了题为 Molecular Motion of Ion Channel 的学术报告 年 8 月, 美国耶鲁大学免疫学教授尹芝南来我室访问并为全所做了题为 T cell Function in Tumor Immunity 的报告 年 9 月 1 日, Hospital Universitario de Canarias, Spain 年 9 月 12 日日本 Kurume 大学医学部免疫系 Kyogo Itoh 教授来访 年 10 月 25 日本 Takehisa Matsumoto 日本 Riken 公司来访 年 10 月 27 日乌兹别克斯坦乌兹别克斯坦科学院 7 人来访 24

25 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 Neil Isaascs, Richard Barron and Laurance Cogdell, Glasgow University Gerhard Mterlik Giovanni E. Mann Vladimir Torchilin Visiting group ⑸International Cooperation Time: 10/29/2004 Professor, CEO, Diamond Inc, UK Time: 11/3/2004 Professor, Guy s, King s & St Thomas School of Biomedical Sciences Time: 11/2004 Chair of Department of Pharmaceutical Sciences, Northeastern University, USA Time: 11/12/2004 Five people including Vice president of University of Tokyo Time: 11/22/2004 Zhushen Fan Zihe Rao Zihe Rao Zihe Rao Zihe Rao Xiyun Yan Xiyun Yan Worked on tumor immunotherapy project in Dr. Yangxin Fu s laboratory at University of Chicago Time: 2 nd half of 2004 with China-German Science Foundation Time: 4/2004-3/2007 Content: Structural Proteomics of SARS cornavirus Within FP6 Time: 5/1/2004 Content: Sino-European Project on SARS Diagnostics and Antivirals With Takehiko Sasazuki, President of International Medical Center of Japan Time: 12/7/2004 Content: Crystallographic analysis of important viral products, including structural analysis of the spike protein from SARS cornavirus as wellas several products from avian flu virus With Teruo Kirkae, Director, Department of Infectious Diseases Research Institue, International Medical Center of Japan Time: 12/16/2004 Content: Crystallographic analysis of important viral products, including structural analysis of the spike protein from SARS cornavirus as wellas several products from avian flu virus With Andrew Bradbury, Professor, Los Alamos National Laboratory, USA Content: Establishment of high number and variety of SARS antibody library using a new phage vector With Kyogo Ito, Professor, Kurume University, Japan Content: immune activities of antigen surface positions 25

26 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 年 10 月 29 日. 英国 Glasgow University 的 Neil Isaacs Richard Barron 和 Laurance Cogdell 年 11 月 3 日英国 Diamond 的 CEO Gerhard Materlik 教授来访 年 11 月英国 Guy s, King s & St Thomas School of Biomedical Sciences 的 Giovanni E. Mann 教授来访 年 11 月 12 日美国 Northeastern University Department of Pharmaceutical Sciences 的 Chairman Vladimir P. Torchilin 来访 年 11 月 22 日日本东京大学副校长等 5 人来访 ⑸ 国际合作研究 1) 范祖森 :2004 年下半年到 University of Chicago 大学傅阳心教授实验室开展肿瘤免疫治疗的合作研究 2) 饶子和 : 合作方中德科学基金, 时间 2004 年 4 月 年 3 月合作内容 Structural proteomics of SARS coronavirus 3) 饶子和 : 合作方第六欧盟框架计划, 时间 2004 年 5 月 1 日合作内容 Sino-European Project on SARS Diagnostics and Antivirals (SARS 诊断及治疗的中 欧合作计划项目 (SEPSDA)) 4) 饶子和 : 合作方 Takehiko Sasazuki, President, International MedicalCenter of Japan, 时间 2004 年 12 月 7 日合作内容 : 重要病毒的结构分析 (crystallographic analysis of important viral products, including structural analysis of the spike protein from SARS coronavirus as well as several products from avian flu virus) 5) 饶子和 : 合作方 Teruo Kirikae, Director, Department of Infectious Diseases Research Institute, International Medical Center of Japan, 时间 2004 年 12 月 16 日合作内容 : 重要病毒的结构分析 (crystallographic analysis of important viral products, including structural analysis of the spike protein from SARS coronavirus as well as several products from avian flu virus) 6) 阎锡蕴与美国 Alamos 国家重点实验室的 Andrew Bradbury 教授合作, 合作内容 : 利用新型噬粒载体构建大容量和多样性 SARS 抗体库 7) 阎锡蕴与日本 Kurume 大学 Kyogo Ito 教授合作, 合作内容 : 开展抗原表位免疫活性的研究 26

27 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 Xiyun Yan Junmei Zhou Tao Xu ⑹Domestic Meetings With Daikin Company, Japan Content: inhibitory role of novel PTAF against SARS virus With Hirosh Kihara, Professor, Japan Time: 4/2004 and 12/2004 Content: Experiments at the station of X-rays small angle (15A). with Dr. Nils Brose, Max Plank Institue Content: the role of Munc13 in insulin secetion Zihe Rao High Level Symposium on International Techniques of new medicine drug development and drug market Invited Speaker Time: 4/27/2004, Beijing Topic: Protein Science and Development of New Drugs Zihe Rao Zhongguancun Symposium Invited Speaker Time: 4/28/2004, Beijing Topic: Protein Science-from Aids to SARS Zihe Rao 7 th National Conference of Enzymology Invited Speaker Time: 5/16/2004, Kunmin Topic: Protein Science is basic science as well as spring for industry- from Aids to SARS Wenrui Chang 7 th National Conference of Enzymology Invited Speaker Time: 5/15-17/2004, Kunmin Topic: Studies on 3D-structure of photosynthetic membrane proteins-crystal structure of spinach major light harvesting complex Chih-chen Wang 7 th National Conference of Enzymology Invited Speaker Time: 5/15-17/2004, Kunmin Topic: Dimerization by Domain Hybridization Bestows Chaperone and Isomerase Activities Sarah Perrett 7 th National Conference of Enzymology Speaker Time: 5/15-17/2004, Kunmin Topic: Folding, misfolding and amyloid fibril formation of yeast prion protein Ure2 27

28 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES ) 阎锡蕴与日本大金公司合作, 合作内容 : 研究新型网膜 PTAF 对 SARS 病毒的抑制作用 9) 周筠梅与日本关西医科大学的 Hirosh Kihara 教授合作, 时间 :2004 年 4 月和 12 月两次, 合作内容 : 在 X- 射线小角散射站 (15 Å) 进行实验 本工作获日本筑波高能物理所光子工厂的资助 10) 徐涛与德国马普研究所 Dr. Nils Brose 合作, 内容 Munc13 在胰岛素分泌中的作用 ⑹ 国内学术会议饶子和国际新药开发技术与医药市场发展战略高层论坛特邀报告时间 :2004 年 4 月 27 日, 北京题目 : 蛋白质科学与创新药物饶子和中关村论坛特邀报告时间 :2004 年 4 月 28 日, 北京题目 : 蛋白质科学 From AIDS to SARS 饶子和第七届全国酶学会特邀报告时间 :2004 年 5 月 16 日, 昆明题目 : 蛋白质科学是基础研究和生物产业的源泉 从 AIDS 到 SARS 常文瑞第七届全国酶学会议大会特邀报告时间 :2004 年 5 月 15-17, 昆明题目 :Studies on 3D-structure of photosynthetic membrane proteins-crystal structure of spinach major light harvesting complex 王志珍第七届全国酶学学术讨论会, 昆明时间 :2004 年 5 月 15-17, 昆明题目 : Dimerization by domain hybridization bestows chaperone and isomerase activities. Oral presentation. 柯莎第七届全国酶学学术讨论会大会报告时间 :2004 年 5 月 15-17, 昆明题目 : 酵母 Prion 蛋白 Ure2 的折叠 错误折叠与淀粉样纤维的形成 28

29 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 Yi Shi 7 th National Conference of Enzymology Time: 5/16/2004, Kunmin Poster: The effect of hydrophobic fluorescent probe bis-ans on the chaperone function of trigger factor and its dimerization Ming Bo 7 th National Conference of Enzymology Time: 5/16/2004, Kunmin Poster: The glutathione peroxidase activity of yeast prion Ure2 Liling Zeng 7 th National Conference of Enzymology Time: 5/16/2004, Kunmin Poster: The effect of C-domain on the in vivo chaperone function of trigger factor Huiyong Lian 7 th National Conference of Enzymology Time: 5/16/2004, Kunmin Poster: Expression and purification of yeast chaperones Hsp104 and Ydj1 Yi Jiang 7 th National Conference of Enzymology Time: 5/16/2004, Kunmin Poster: Amyloid nucleation and hierarchical assembly of Ure2 fibrils: role of Asn/Gln repeat and non-repeat regions of the prion domain Zihe Rao 1 th Chinese International Workshop of Pharmacy Invited Speaker Time: 5/18/2004, Hangzhou Zihe Rao 1 th Sino-German Symposium of Young Scientists in Chemistry and Biology Invited Speaker Time: 5/25/2004, Beijing Wei Liang 1 th Chinese International Workshop of Pharmacy Time: 5/17-19/2004, Hangzhou Zihe Rao 2 nd Chinese Proteomics Research Meeting Invited Speaker Time: 8/10-12/2004, Dalian Wei Liang China s Involvement in Imaging and Therapeutics : Ceremony for starting publication and workshop Speaker Time: 9/3-6/2004, Beijing Topic: Guided Therapy of Tumors 29

30 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 石毅第七届全国酶学学术讨论会时间 :2004 年 5 月 16 日, 昆明墙报 : 疏水荧光探针 bis-ans 对 trigger factor 分子伴侣功能及二聚化的影响 柏鸣第七届全国酶学学术讨论会时间 :2004 年 5 月 16 日, 昆明墙报 : 酵母 prion 蛋白 Ure2 的谷胱甘肽过氧化物酶活性 曾丽玲第七届全国酶学学术讨论会时间 :2004 年 5 月 16 日, 昆明墙报 :C-domain 对 trigger factor 体内分子伴侣作用的影响 连惠勇第七届全国酶学学术讨论会时间 :2004 年 5 月 16 日, 昆明墙报 :Expression and purification of yeast chaperones Hsp104 and Ydj1. 江轶第七届全国酶学学术讨论会时间 :2004 年 5 月 16 日, 昆明墙报 :Amyloid nucleation and hierarchical assembly of Ure2 fibrils: role of Asn/Gln repeat and non-repeat regions of the prion domain. 饶子和国际药物制剂专题研讨会特邀报告时间 :2004 年 5 月 18 日, 杭州题目 : 基于 SARS 蛋白质结构的药物设计饶子和第一届中德双边化学生物学青年学术会议特邀报告时间 :2004 年 5 月 25 日, 北京题目 : 蛋白质科学是基础研究和生物产业的源泉梁伟第一届中国国际药剂学学术研讨会 时间 :2004 年 5 月 17-19, 杭州饶子和中国蛋白质组学第二届学术会议特邀报告时间 :2004 年 8 月 日, 大连题目 : 结构蛋白质组学研究进展梁伟 中国介入影像与治疗学 创刊暨学术研讨会时间 :2004 年 9 月 3-6, 北京题目 : 肿瘤的靶向介入治疗 的大会发言 30

31 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 Zihe Rao 3 rd National representative conference of Chinese Crystallography Society and Research Meeting Invited Speaker Time: 9/17-24/2004, Chengdu Zihe Rao 8 th Youth Scientific Symposium in Medicine and Health Invited Speaker Time: 10/10/2004, Beijing Zihe Rao 5 th Youth Symposium of Chinese Scientific Association Invited Speaker Time: 11/2/2004, Beijing Tao Xu 6 th Symposium of Biological Sciences Time: 11/1-6/2004, Chongqing Tao Xu Workshop of Biomembranes and Important Diseases Speaker Time: 12/23-26/2004, Sanya Topic: Intracellular membrane trafficking and regulation of blood glucose Zihe Rao Xiangshan Scientific Conference Invited Speaker Time: 12/28-29/2004, Beijing 31

32 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 饶子和中国晶体学会第三次全国代表会暨学术会议时间 :2004 年 9 月 日, 成都饶子和第八届医药卫生青年科技论坛特邀报告时间 :2004 年 10 月 10 日, 北京题目 : 蛋白质结构 : 从 AIDS 到 SARS 饶子和中国科协第五届青年学术年会特邀报告时间 :2004 年 11 月 2 日, 北京题目 : 蛋白质产业和蛋白质基础研究徐涛第六届生命科学学术研讨会时间 :2004 年 11 月 1-6 日, 重庆徐涛生物膜与重大疾病学术研讨会时间 :2004 年 12 月 日, 三亚题目 : 细胞内膜转运和血糖调控饶子和香山科学会议大会报告时间 :2004 年 12 月 日, 北京题目 :New progress in SARS basic research 32

33 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 Ⅴ. Research Projects and Progress 1. Proteins and Molecular Enzymology Name Chih-chen Wang Position Professor, Member of CAS Research Topic Chaperones in Protein Folding Summary of Research 1. Conformational change of dimeric DsbC molecule induced by GdnHCl, A study by intrinsic fluorescence. Unfolding-refolding of E. coli DsbC, a homodimeric molecule, induced by GdnHCl was studied by intrinsic fluorescence. It is shown that the sulfur atoms of Cys 141 and Cys 163 are far apart from the indole ring of the single Trp (Trp 140), and cannot quench its fluorescence, while the potential quenchers are Met 136 and His 170. It was revealed that only Tyr 171, Tyr 38 and Tyr 52 among the eight Tyr residues, contribute to the bulk fluorescence of the molecule. The character of intrinsic fluorescence intensity changes induced by GdnHCl (equilibrium and kinetic data) and its parametric representation and the existence of an isobestic point of fluorescence spectra at different GdnHCl concentrations suggest the one step character of DsbC denaturation and its reversibility. 2. N-Terminal fluorophore labeling combined with donor-donor energy migration for study of the unfolding of dimeric DsbC. We have developed a valuable method for the N-terminal specific fluorescence labeling by using transamination combined with donor-donor energy migration (DDEM) to study unfolding/folding of a dimeric protein. Transamination provides a general approach for the selective fluorophore attachment to the N-terminal amino acid residues, and the dimeric structure of DsbC allows the introduction of two identical fluorophores, therefore we can take advantage of the DDEM method to trace its unfolding behavior. This combination strategy is useful to investigate conformational changes of other dimeric proteins under variable conditions. Great progresses can be expected when the specific labeling method is combined with DDEM at the single-molecule level. Moreover, this labeling approach can also be applied to non-dimeric protein molecules, and therefore broadens the scope of application for fluorescence spectroscopy. 3. Dimerization, zinc-finger, chaperone and thiol-protein oxidoreductases of DnaJ. 4. Translocation of α-synuclein in E coli. 5. Protective action of protein disulfide isomerase on hypoxia in cells. 6. Suppression of chaperones on the amyloid formation of α-synuclein. Selected Publications 1. O. V.Stepanenko, I. M. Kuzenetsova, K. K. Turoverov *, C. J. Huang, and C. C. Wang *. Conformational change of dimeric DsbC molecule induced by GdnHCl: A study by intrinsic fluorescence. Biochemistry-USA 43, , (2004). 2. Xuejun Duan, Zhen Zhao, Jianping Ye, Huimin Ma *, Andong Xia *, Guoqing Yang, and Chih-chen Wang *. N-Terminal fluorephore labeling comined with donor-donor energy migration for study on unfolding of dimeric DsbC. Angew. Chem. Int. Ed. 43, , (2004). 33

34 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 Ⅴ. 研究课题及进展 1. 蛋白质和分子酶学 姓名 职称 研究方向 王志珍院士研究员副研究员长江学者百人计划国家杰出青年 分子伴侣和蛋白质折叠 本年度工作简介 1. 测定内源荧光研究盐酸胍诱导的分子伴侣 DsbC 同源二聚体的构象变化 分析了 DsbC 中色氨酸和酪氨酸残基的微环境 唯一的 Trp140 在两个单体之间没有能量传递, 其荧光不为 Cys98-Cys101 和 Cys141-Cys163 的硫原子淬灭 ; 但可能为 Met136 和 His170 淬灭 以及 52 位的酪氨酸残基是 DsbC 荧光的主要贡献者 ; 其它 5 个酪氨酸的能量传递给 Trp140 获在它们之间传递 不同浓度盐酸胍诱导下的 DsbC 去折叠过程的平衡态内源荧光的变化, 表明其去折叠过程为二态构象转变, 没有检测到去折叠的平衡态中间体 动力学研究表明变性稀释重折叠与去折叠过程相可逆 2. 用供体 - 供体能量转移 (DDEM) 技术研究 DsbC 同源二聚体的去折叠 用乙醛酸将 DsbC 二个 N 末端的氨基转移生成 α 羰醛, 利用 α 羰醛与肼的交联作用将荧光探针分子 BODIPY 特异地连接到 DsbC 的 N 末端, 是一种新的荧光标记策略 用 DDEM 技术计算 DsbC 不同程度去折叠后的能量转移效率探测分子到天然 DsbC 分子二个 N 末端间为 35Å( 晶体中 29Å); 在 1.5M 盐酸胍中, 二个 N 末端距离增加到 47Å, 表明分子部分去折叠但尚未解离 ;6M 盐酸胍使大大超过 BODIPY 的临界距离, 说明二体分子解离 N 末端特异标记结合 DDEM 技术可成为双体蛋白质折叠研究的通用方法 3. 分子伴侣 DnaJ 的二聚化, 锌指, 巯基 - 蛋白质氧化还原酶性质的研究 4.α-Synuclein 在大肠杆菌中的转运 5. 蛋白质二硫键异构酶对细胞缺氧损伤的保护作用 6. 分子伴侣对 α-synuclein 形成纤维和淀粉样沉淀的抑制作用 获奖 王志珍 2004 年度中国科学院宝洁优秀研究生导师 任国平 2004 年度中国科学院地奥二等奖 34

35 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 Name Jun-Mei ZHOU Position Professor Research Topic Protein structure, function and folding Summary of Research 1) Structure and Function of the Chaperone and Foldase Trigger Factor Trigger factor (TF) has both chaperone and cis-trans prolyl isomerase activities and is the first chaperone encountered by the nascent polypeptide in bacteria. While ribosome-bound TF is monomeric, the function of the dimeric form, and the reason for the excess concentrations of TF in the cytosol, are unknown. We have shown that it is specifically the dimeric form of TF that is able to bind folding intermediates of a substrate protein and hold them in a conformation that is competent to be refolded in an ATPase-dependent manner by the Hsp70 system, DnaK/DnaJ/GrpE. This may represent a previously unrecognised function of TF as a cytosolic binding chaperone. 2) Function, Folding and Amyloid Formation of the Yeast Prion Protein Ure2 We have carried out a detailed study of the mechanism of amyloid fibril formation for the yeast prion protein Ure2 and a series of N-terminal deletion mutants using a combination of techniques. The results demonstrate the importance of both Gln/Asn repeat and non-repeat regions of the N-terminal prion domain in the process of amyloid nucleation, consistent with the importance of these regions for manifestation of the prion phenotype in vivo. We have also investigated the enzymatic properties of the Ure2 protein and demonstrated that Ure2 has glutathione-dependent peroxidase activity in both native and fibrillar forms. This work represents important progress in elucidation of the role of Ure2 in vivo. Further, establishment of an in vitro activity assay for Ure2 provides a valuable tool for the study of structure-function relationships. Selected Publications 1. Zhu L, Qin ZJ, Zhou JM* & Kihara H. Unfolding kinetics of dimeric creatine kinase measured by stopped-flow small angle X-ray scattering. Biochimie 86: (2004). 2. Liu CP & Zhou JM.* Trigger factor-assisted folding of bovine carbonic anhydrase II. BBRC 313: (2004). 3. Jiang Y, Li H, Zhu L, Zhou JM* & Perrett S.* Amyloid nucleation and hierarchical assembly of Ure2p fibrils: Role of Asn/Gln repeat and non-repeat regions of the prion domain. J. Biol. Chem. 279: (2004). 4. Bai M, Zhou JM* & Perrett S* The yeast prion protein Ure2 shows glutathione peroxidase activity in both native and fibrillar forms. J. Biol. Chem. 279: (2004). 5. Liu CP, Perrett S & Zhou JM* (2005) Dimeric trigger factor stably binds folding-competent intermediates and cooperates with the DnaK-DnaJ-GrpE chaperone system to allow refolding. J. Biol. Chem. in press. 35

36 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 姓名 职称 研究方向 周筠梅院士研究员副研究员长江学者百人计划国家杰出青年 蛋白质的结构 功能与折叠 本年度工作简介 1,Trigger Factor 的结构与分子伴侣功能 Trigger factor(tf) 是大肠杆菌中新生肽链所遇到的第一个分子伴侣, 具有肽基脯氨酰基顺反异构酶 (PPIase) 活性, 在细胞内以三种状态存在 : 核糖体结合的单体 细胞质中游离的单体和二体 核糖体结合的 TF 帮助新生肽链折叠的机制已经研究的比较多, 而细胞质中大量游离的 TF, 特别是二体 TF 的功能尚不清楚 我们今年的研究结果表明只有二体 TF 能与蛋白质折叠的中间态紧密结合形成稳定的复合物, 并维持折叠中间态具有继续折叠的构象 ; 且只有那些被二体 TF 结合的折叠的中间体能够与 Hsp70 分子伴侣系统 (DnaK/DnaJ/GrpE) 相互作用, 在 ATP 存在下, 继续折叠为天然蛋白质 阐明了以前未意识到的,TF 可能作为细胞质内 结合分子伴侣 的作用 2, 酵母类 Prion 蛋白 Ure2 的功能 折叠和淀粉样纤维的形成我们应用 ThT 结合荧光和原子力显微镜技术, 详细研究了酵母类 Prion 蛋白 Ure2 及一系列 N- 端不同程度缺失突变体淀粉样纤维的形成 结果表明 Ure2 N- 端 Prion 结构域中的 Gln/Asn 重复序列和非重复序列对淀粉样纤维形成的成核过程都是十分重要的 此结果与体内研究发现的这些区域对其形成表型的影响一致 我们还进一步研究了 Ure2 的酶学性质, 发现天然 Ure2 和纤维形式的 Ure2 具有相同的谷胱甘肽过氧化物酶活性, 本工作是阐明 Ure2 体内功能的重要进展 并且, 在体外测定 Ure2 活性方法的建立, 为进一步研究 Ure2 的结构和功能之间的关系提供了有用的工具 本年度获奖情况 36

37 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 Name Position Research Topic Summary of Research Xian en Zhang Professor Molecular Recognition and Biosensors The mismatch repair (MMR) system plays an important role in maintaining the stability of the genome and defects in the mammalian pathway for mismatch repair are associated with a strong predisposition to tumor development. So far, the detailed mechanism of interaction between MMR proteins is not well understood. Our recent research interests include multi-mmr protein interactions and MMR protein-dna interactions, and exploring MMR molecules as biosensor recognition elements for effective detection of DNA mutations. The progress in the passed year is as follows: 1. The overexpression sytems of the proteins MutS and MutL have been established. 2. Four MutS fusion molecular systems have been constructed. Based on these fusions, two types of biochips have been built: protein chips and DNA chips, enabling the fast, specific and sensitive detection of DNA mutations, in combination of either fluorescence or enzymatic reaction. We have a number of publications on this aspect. 3. In the future we will focus on the structure-function relationships of protein-protein interactions related to DNA mismatch repair. Protein fusion technology and surface plasma resonance technology will be adopted to facilitate the research. Selected Publications 1. LJ Bi, YF Zhou, XE Zhang*, ZP Zhang, CG Zhang and Anthony E. G. CASS. High Expression of Gene Encoding for MutL Fusion Protein and Research on Its Chaperon Function. Chinese Journal of Biochemistry and Molecular Biology 20(2): , (2004). 2. Li-Jun BI, Ya-Feng ZHOU, Xian-En ZHANG *, Jiao-Yu DENG, Ji-Kai WEN, Zhi-Ping ZHANG. Construction and Characterization of Different MutS Fusion Proteins as Recognition Elements of DNA Chip for Detection of DNA Mutations, Biosensors and Bioelectronics, 2004 in press. 3. Sun L, Dong Y, Zhou Y, Yang M, Zhang C, Rao Z, Zhang XE*. Crystallization and preliminary X-ray studies of methyl parathion hydrolase from Pseudomonas sp. WBC-3. Acta Crystallogr. D Biol. Crystallogr. 60(5): 954-6, (2004). 37

38 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 姓名 职称 研究方向 张先恩院士研究员副研究员长江学者百人计划国家杰出青年 分子识别与生物传感 本年度工作简介 一 课题组科研进展情况 : DNA 错配修复 (Mismatch Repair,MMR) 是细胞复制后的一种修复机制, 具有维持 DNA 复制保真度, 控制基因变异的作用 大量研究表明,DNA 错配修复基因缺陷会导致在 HNPCC 和散发性结肠癌等的发生 目前,DNA 错配修复的分子机制作为当今研究的热点也取得了一些进展, 但修复过程中一些修复蛋白的具体作用, 蛋白质之间的协同作用以及作用机制还存在争论 我们主要的研究工作包括 DNA 错配修复系统介导的基因突变检测方法研究, 错配修复蛋白与 DNA 的作用及蛋白质之间协同作用研究 研究进展情况如下 : 1, 构建了 DNA 错配修复基因 muts 和 mutl 的高效表达系统 2, 利用 MutS 蛋白对错配碱基的特异识别 结合功能, 通过基因工程, 构建一系列 MutS 融合蛋白, 发展了基于 MutS 蛋白的基因突变检测蛋白质芯片和 DNA 芯片新方法 研究结果表明, 基于 MutS 蛋白的蛋白质芯片和 DNA 芯片都能成功检测含有错配或未配对碱基的寡核苷酸片段及含有单个碱基错配的不同长度结核杆菌 rpob 基因片段 3,DNA 错配修复蛋白结构与功能的关系以及修复蛋白的协同作用研究 目前已经完成了 DNA 错配修复蛋白 MutS 融合分子系统的构建以及 mutl 基因的克隆 表达和纯化工作, 为修复蛋白之间协同作用的动力学分析作准备 MutS MutL 和 MutH 的缺失突变和定点突变工作正在开展, 该工作是确定 MutS 与 MutL MutL 与 MutH 相互作用位点和关键氨基酸组成 相关研究成果已经分别发表在国内外学术刊物上 本年度获奖情况 发表在 Analytical Chemistry 上的文章 A MutS-based Protein Chip for Detection of DNA Mutations 获得 2004 年度 湖北省第十届自然科学优秀学术论文 特等奖 38

39 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 Name Wang Jinfeng Position Professor Research Topic NMR study of protein folding and solution structure Summary of Research We have accomplished the following studies during 2004: 1. Study of the folding pathway and mechanism of staphylococcal nuclease (SNase) (1) We studied the folding of the 1-79 residue fragment of SNase (SNase79). The residual α helix and β structures were determined. The β turn Y27-Q30 was identified as a folding initiation site for SNase79. The conclusion was drawn that the cis/trans heterogeneity of the X-prolyl bond Q30-P31 determines whether the sequence region T13-V39 forms a native-like β structure in SNase79. (2) We calculated the solution structure of [Pro - ]SNase which has lost enzyme activity. The internal motions of [Pro - ]SNase were analyzed. Analysis of the key factors that influence the enzyme function of proteins is in progress. (3) We determined the solution structure of the residue fragment of SNase (SNase140). The NMR experiments for backbone dynamics were completed. The plasmid for two relevant fragments was constructed. (4) The C-terminal residue fragment of SNase (SNaseα 3 ) which contains the α-helix sub-domain of SNase was expressed and purified. Interaction between SNaseα 3 and SNase121 was studied. A dramatic increase in enzyme activity of SNase121 and formation of native-like conformations of both fragments were observed. Study of the solution structure of the complex was started. 2. Structural genomics study (1) We calculated the solution structure of a 180-residue human translationally controlled tumor protein (TCTP). Determination of Ca 2+ binding site on TCTP is in progress. (2) We calculated the solution structure of human programmed cell death 5 protein (PDCD5). We have expressed and purified the N-terminal 22-residue fragment of PDCD and determined its solution structure as an intact α helix. (3) We obtained an initial solution structure of a protein from the human blood system which is a protein with unknown physiological function. (4) We calculated the solution structure of Ssh10b which is a member of Sac10b family from the hyperthermophilic archaeon Sulfolobus shibatae. Ssh10b was determined to be a dimer in aqueous solution. Selected Publications (1) Human programmed cell death 5 protein has a helical-core and two dissociated structural regions. Dongsheng Liu, Yingang Feng, Yuan Cheng, Jinfeng Wang*. BBRC 2004, 318: (2) Searching for folding initiation sites of staphylococcal nuclease: A study of N-terminal short fragments. Jixun Dai, Xu Wang, Yingang Feng, Guibao Fan, Jinfeng Wang*. Biopolymers 2004, 75:

40 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 姓名 职称 王金凤 院士研究员副研究员长江学者百人计划国家杰出青年 研究方向 本年度工作简介 2004 年内我组完成了如下方面的工作 : 一 金黄色葡萄球菌酶 (SNase) 体外折叠路径和机制研究 (1) 完成了酶蛋白 N- 末端 1-79 残基片段 (SNase79) 折叠研究 确定了 α- 螺旋和 β- 折叠的残余构象,β- 转角 Y27-Q30 是折叠的起始位点, 并阐明 β- 折叠是否在 SNase79 中形成取决于 Q30-P31 是处于 trans- 构象还是 cis- 构象 (2) 确定了因 P42 P47 和 P117 突变而失活的 SNase 的溶液三维结构 完成了其内运动特性的研究 正在从构象变化及内运动变化角度分析失活的决定因素, 以进一步探讨酶发挥功能机制 (3) 确定了 N- 末端 残基片段 SNase140 的溶液三维结构 完成了其内运动研究的实验 正在构建和表达两个相关片段, 从而可以进一步分析 C- 端几个氨基酸残基缺失对酶功能影响的决定因素 (4) 成功地表达了 C- 端 残基 ( 即包含 α3 亚结构域的残基 ) 片段 SNase α3 并已开展 SNase α3 和 SNase121(1-121 残基片段 ) 的相互作用研究 已观察到 SNase α3 和 SNase121 类天然构象的形成以及酶活的极大增加 正在解析复合体的三维结构 二 与结构基因组研究相关的工作 (1) 解析了 172 残基的人翻译控制的肿瘤蛋白质 TCTP 的溶液三维结构, 正在进一步构建突变体以确定 Ca 2+ 结合位点 (2) 解析了人白血病细胞凋亡相关蛋白质 PDCD5 的溶液三维结构 并表达了 N- 端 22 残基片段和确定了这一独立 α 结构域构象 (3) 初步解析了一个人血液系统未知结构功能蛋白质的三维结构 (4) 解析了嗜热菌蛋白质 Ssh10b 双体的溶液三维结构 本年度获奖情况 40

41 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 Name Sarah Perrett Position Associate Professor Research Topic Protein Misfolding and Disease Summary of Research The yeast prion protein Ure2 forms amyloid-like filaments in vivo and in vitro. This ability depends on the N-terminal prion domain, which contains Asn/Gln repeats, a motif thought to cause human disease by forming stable protein aggregates. We compared the time course of structural changes monitored by thioflavin T (ThT) binding fluorescence and atomic force microscopy (AFM) for Ure2 and a series of prion domain mutants. AFM height images at successive time points during a single growth experiment showed the sequential appearance of at least four fibril types that could be readily differentiated by height, morphology and/or time of appearance. N-terminal deletion mutants showed an increased lag time, demonstrating that the intact N-terminal domain is required for efficient amyloid nucleation, consistent with its importance in prion formation in vivo. Further, the results show that Ure2 amyloid formation is a multistep process via a series of fibrillar intermediates. The C-terminal domain of Ure2 has homology to glutathione transferases (GSTs), but lacks typical GST activity. A recent study found that deletion of the Ure2 gene causes increased sensitivity to oxidants, while prion strains show normal sensitivity. Using steady-state kinetic methods, we succeeded in demonstrating glutathione-dependent peroxidase activity for Ure2, using a variety of substrates. The mutant 90Ure2, which lacks the unstructured N-terminal prion domain, showed kinetic parameters identical to WT, indicating that the prion domain does not contribute to the enzyme activity of Ure2. Interestingly, fibrillar aggregates showed the same level of activity as soluble protein, consistent with normal sensitivity towards oxidants of prion yeast strains. Demonstration that protection against oxidant toxicity is an inherent property of both native and amyloid forms of the Ure2 protein represents important progress in elucidation of its role in vivo. Further, establishment of an in vitro activity assay provides a valuable tool for the study of structure-function relationships of the Ure2 protein as both a prion and an enzyme. Selected Publications 1. Jiang, Y., Li, H., Zhu, L., Zhou, J.M.* & Perrett, S.* Amyloid nucleation and hierarchical assembly of Ure2p fibrils: Role of Asn/Gln repeat and non-repeat regions of the prion domain. J. Biol. Chem. 279, , (2004) 2. Bai, M., Zhou, J.M.* & Perrett, S.* The yeast prion protein Ure2 shows glutathione peroxidase activity in both native and fibrillar forms. J. Biol. Chem. 279, , (2004) 3. Liu C.P., Perrett, S. & Zhou, J.M.* Dimeric trigger factor stably binds folding-competent intermediates and cooperates with the DnaK-DnaJ-GrpE chaperone system to allow refolding. J. Biol. Chem. (2005)in press. 41

42 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 姓名 职称 柯莎 院士研究员副研究员长江学者百人计划国家杰出青年 研究方向 蛋白质错误折叠与疾病 本年度工作简介在 2004 年度完成的工作主要有两方面 : 1) Ure2 淀粉样纤维的成核以及分步组装酵母 Ure2 蛋白在体内和体外均易形成淀粉样纤维, 其富含天冬酰胺和谷氨酰胺重复序列的 N 端 (1-89) 与 prion 性质密切相关 N 端的 位残基是夹在两段重复序列中的随机序列, 相对疏水, 在种间高度保守 我们以 ThT 结合荧光和原子力显微镜监测了 Ure2 及其 prion 结构域不同程度缺失的突变体在不同条件下淀粉样纤维形成过程 发现有原子力显微镜图象高度分别为 纳米 不同形态的纤维在 ThT 所追踪的平台期的不同时段出现 由原子力显微镜图象的高度还确认出 Ure2 的双体和寡聚体的存在 突变体 15Ure2 和 15 42Ure2 的纤维化也与野生型 Ure2 同样经由多个不同的纤维形态, 但是与野生型相比它们与 ThT 结合荧光的延迟期更长, 15 42Ure2 与 ThT 的结合较弱 我们的工作表明该保守区对于淀粉样纤维形成的成核过程以及与 ThT 的结合过程都有重要作用, 并且 Ure2 蛋白及其 N 端突变体的淀粉样化遵循经由多个纤维状中间体的分步组装机制 2) 酵母 prion 蛋白 Ure2 的谷胱甘肽过氧化物酶活性 Ure2 是酵母 prion [URE3] 的前体蛋白形式, 它与谷胱甘肽转硫酶 (GST) 具有同源性, 却并不表现 GST 活性 最近有研究表明, 缺失 Ure2 基因的酵母突变株表现出对重金属离子和氧化剂的敏感性增加, 并推测 Ure2 可能具有潜在的谷胱甘肽过氧化物酶 (GPx) 活性 为了证明这种抗氧化剂毒性的保护作用是 Ure2 及其 prion 形式的内在性质, 我们分别以 CHP,H 2 O 2,t-BH 为过氧化物底物, 采用稳态动力学方法对 Ure2 的 GPx 活性进行了测定 其最适 ph 和最适温度分别为 ph 8 和 40 ºC 在双倒数作图中表现为米氏酶特性, 其双底物反应动力学遵循顺序机制, 而不是乒乓机制 这一点与 Ure2 归属于 GST 家族是一致的 缺失 N 端无结构的 PrD 结构域的突变体 90Ure2 具有与野生型 Ure2 相同的动力学常数 纤维化的 Ure2 聚集体也表现出与可溶的 Ure2 同等水平的 GPx 活性 Ure2 的 GPx 活性的证实是阐明其体内功能的重要进展 此外, 体外活性测定方法的建立也为 Ure2 作为既是 prion 又是酶的蛋白质的结构与功能关系的研究提供了有价值的工具 本年度获奖情况 (1) 江轶获得 2004 年中科院院长优秀奖, (2) 柏鸣已具有获得 2004 年所长奖学金一等奖的资格 42

43 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES Three-dimensional Structure and Function of Biomacromolecules Name Dong-cai Liang Position Professor, Member of CAS Research Topic Structural Genomics Summary of Research A. Research projects and progress: 1. Structural genomics on proteins from human and T. tengcongensis 1) The expression of 26 genes were analyzed in detail;14 recombinant proteins were purified and 10 kinds of crystals (including microcrystals) were obtained. 2) X-ray diffraction data for 7 crystals were collected on our in-house Cu-anode X-ray source, and on the Beamline BL-6B Experimental Station (PF, Tsukuba, Japan). 3) The crystal structures of 5 proteins were solved using the methods of MAD, SAD/SIRAS and MR. The structures are listed below: a) KD93: The crystal structure of KD93, was determined using the MAD method. b) Calmodulin: Calmodulin plays a key role in transducing Ca2+ signals to different physiological effects. The crystal structure of a potato calmodulin (PCM6), the first three-dimensional structure of a plant calmodulin, was solved by the MR method. c) Insulin-like growth factor 1: The crystal structure of mini-igf-1(2) was solved by the SAD/SIRAS method using our in-house X-ray source. d) Other unpublished results: The crystal structures of U123 and the C domain of U268 have just been solved. 2. Structural genomics on proteins in a two-component regulatory system Some proteins in a two-component regulatory system were chosen carefully as the targets of the structural genomics and structural biology project. B. Faculty and graduate students: Our group currently has three research faculty, one post-doctoral fellow and six graduate students. Two students were conferred with the Ph. D degree in C. Funds: Our work was supported by National Natural Science Foundation of China; 973 Project; 863 Program and CAS Major Innovation Program. Selected Publications Three papers were published in They are: 1. Jun-Feng Liu, Xin-Quan Wang, Zhan-Xin Wang, Jian-Rong Chen, Tao Jiang, Xiao-Min An, Wen-Rui Chang, Dong-Cai Liang*. Crystal structure of KD93, a novel protein expressed in human hematopoietic stem/progenitor cells. Journal of Structural Biology 148, (2004). 2. Cai-Hong Yun, Yue-Hua Tang, You-Min Feng, Xiao-Min An, Wen-Rui Chang, Dong-Cai Liang* A crystal structure of mini-igf-1(2): an analysis of the disul.deisomerization property and receptor binding property of IGF-1 based on the three-dimensional structure. Biochemical and Biophysical Research Communications 326, (2004). 3. Yun, C.H., Bai, J., Sun, D.Y., Cui, D.F., Chang, W.R., and Liang, D.C.* Structure of potato calmodulin PCM6: the first report of the three-dimensional structure of a plant calmodulin. Acta Crystallogr. D. Biol. Crystallogr. 60, (2004). 43

44 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 生物大分子三维结构和功能 姓名 职称 研究方向 梁栋材院士研究员副研究员长江学者百人计划国家杰出青年 结构基因组学研究 本年度工作简介一 研究工作 1 人源及嗜热菌结构基因组方面 : (1) 重点分析了 26 个基因的重组表达, 纯化了 14 种重组蛋白, 获得了 10 种蛋白的晶体 ( 包括微晶 ) (2) 在本实验室铜靶光源,Mar-345 IP 记录仪及在日本筑波 PF 光子工厂的 BL6B 光源上收集了 7 种晶体的衍射数据 (3) 分别用多波长反常散射 (MAD) 单波长反常散射 SAD/SIRAS 方法和 MR 方法解析了 5 种蛋白质的晶体结构 解析的结构分别为 : (a)kd93: KD93 属于一个结构和功能未知的蛋白质家族 利用多波长反常散射技术 (MAD) 解析了这个蛋白质 1.9-Å 分辨率的晶体结构 ( J.S.B. 148 (2004) ) (b) 钙调素 : 钙调素直接或间接地参与了细胞内大多数重要的信号传导途径, 我们用分子置换法解析了土豆钙调素 PCM6 的晶体结构 ( Acta Crys. D.. 60 (2004), ) (c) 类胰岛素生长因子 -1: 我们用基于实验室光源的 SAD/SIRAS 方法解析了 mini-igf-1 (2) 的晶体结构 ( B. B. R. C. 326 (2004) ) (d) U123 和 U268 C- 结构域的结构解析工作刚刚完成 2 开展 双因子调控系统 的结构基因组研究 经过调研, 选择一批有重要功能的双因子调控基因进行结构基因组和结构生物学研究 二 人员情况及人才培养 : 在职人员 3 名, 博士后 1 名 2004 年毕业博士生 2 名, 在读研究生 6 名 三 参加下列研究项目并从中获得研究经费 : 国家自然科学基金委青年基金 ( 项目编号 ); 重点项目 ( 项目编号 ); 中科院创新工程重大项目 ( 项目编号 KSCX1-SW-17 ); 973 项目 ( 项目编号 2002CB ) 和 863 项目 ( 项目编号 2002BA711A13) 本年度获奖情况 < 藻类光合作用捕光蛋白 - 色素复合物的三维结构与功能研究 > 2003 年度北京市科学技术奖一等奖 ( 第四获奖人 ) 44

45 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 Name Zihe Rao Position Professor, Member of CAS Research Topic Human disease related protein structure, function and drug design Summary of Research In 2004, Prof. Rao s research group made good progress: we determined 26 three-dimensional protein structures; 26 papers were published in international SCI journals. Three patents were applied for. Our main achievements are as follows: 1. We analyzed two mutually overlapping fragments of FKBP52 (N (1-260) and C ( )) as well as the complex structure with the C-terminal pentapeptide from Hsp90. Based on the structures of the two overlapping fragments, the complete putative structure of FKBP52 can be defined. 2. Through analysis of the structure of the PTB domain of Dok1 and its complex with a phosphopeptide derived from RET receptor tyrosine kinase, we revealed the molecular basis for the specific recognition of RET by the Dok1 PTB domain. 3. We have determined the crystal structure of Pirin. The structure provides evidence that Pirin requires the participation of the metal ion for its interaction with Bcl-3 to co-regulate the NF-κB transcription pathway and the interaction with NFI in DNA replication. 4. We used the SARS M pro structure, which we successfully determined in 2003, to design a series of inhibitors that are effective against four kinds of coronavirus, and analyzed the structures of the SARS M pro and the porcine transmissible gastroenteritis virus (TGEV) M pro in complex with these inhibitors. We have also analyzed the crystal structures of the SARS-CoV and mouse hepatitis virus (MHV) spike (S) protein fusion cores and proposed a conserved mechanism of membrane fusion involving the spike protein. Last year we obtained a European Union Framework 6 grant to carry out the Sino-European Project on SARS Diagnostics and Antivirals (SEPSDA). 5. We are highly effective in the purification of membrane proteins and structure analysis, as well as in the exploration of new technical methods. Selected Publications 1. Wu B, Li P, Liu Y, Lou Z, Ding Y, Shu C, Ye S, Bartlam M, Shen B & Rao Z*. 3D structure of human FK506-binding protein 52: implications for the assembly of the glucocorticoid receptor/hsp90/immunophilin heterocomplex. Proc. Natl. Acad. Sci. USA, 101(22): , (2004). 2. Bartlam M, Wang G, Yang H, Gao R, Zhao X, Xie G, Cao S, Feng Y & Rao Z*. Crystal Structure of an Acylpeptide Hydrolase/Esterase from Aeropyrum pernix K1. Structure, 12(8): , (2004). 3. Pang H, Bartlam M, Zeng Q, Miyatake H, Hisano T, Miki K, Wong LL, Gao GF & Rao Z*. Crystal Structure of Human Pirin: AN IRON-BINDING NUCLEAR PROTEIN AND TRANSCRIPTION COFACTOR. J. Biol. Chem., 279(2): , (2004). 4. Shi N, Ye S, Bartlam M, Yang M, Wu J, Liu Y, Sun F, Han X, Peng X, Qiang B, Yuan J* & Rao Z*. Structural Basis for the Specific Recognition of RET by the Dok1 Phosphotyrosine Binding Domain. J. Biol. Chem., 279(6): , (2004). 5. Xu Y, Lou Z, Liu Y, Pang H, Tien P, Gao G.F & Rao Z*. Crystal structure of SARS-CoV spike protein fusion core. J. Biol. Chem., 279(47): , (2004). 45

46 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 姓名 职称 研究方向 饶子和院士研究员副研究员长江学者百人计划国家杰出青年 与人类重大疾病相关的蛋白质结构 功能及药物设计 本年度工作简介饶子和院士研究组包括了他在清华大学和中科院生物物理所的两个课题组, 2004 年科研进展顺利 : 共解析蛋白质三维结构 26 个, 提交给国际蛋白质数据库 (PDB) 数据 17 个蛋白质结构信息 ; 在国外发表 SGI 论文 26 篇, 其中包括 Proc. Natl. Acad. Sci. USA 一篇 J. Biol. Chem. 四篇 Structure 一篇 Biochemistry 两篇 Biophysical J. 一篇 申请专利 3 项 毕业博士研究生 2 人 ; 毕业硕士研究生 1 人 ; 博士后出站 1 人 主要研究成果如下 : 1 解析了 FKBP52 的两个相互交叠片段 (N(1-260) 和 C( )) 以及与 Hsp90 C 末端五肽复合物的晶体结构, 基于这两个相互交叠结构域的结构, 拼接得到 FKBP52 蛋白质全长的三维结构并推测了人 FKBP52 蛋白质的三维结构 糖皮质激素受体 Hsp90 免疫亲和素复合物结合模式 2 通过解析 Dok1 PTB 结构域以及它与源自 RET 受体酪氨酸激酶的磷酸化多肽复合物的晶体结构, 揭示出 Dok1 PTB 结构域对 RET 特异性识别的分子基础, 并发现 Dok1 不识别源自 TrkA 和 IL-4 的多肽序列, 可分别被 Shc 和 IRS1 识别 3 解析了人 Pirin 晶体结构, 这是一个可与 B 淋巴细胞癌基因编码的原癌蛋 (Bcl-3) 和核因子 I(NFI) 反应的核蛋白 功能研究显示,Pirin 需要金属离子的参与同 Bcl-3 反应来共同调控 NF-κB 转录通路和在 DNA 复制过程中与 NFI 相互作用, 因此由重金属的离子置换可建立起金属离子直接影响基因转录的一种新通路 4 继去年成功解析了 SARS 冠状病毒主要蛋白酶 (3CL pro ) 及其复合物的三维结构后, 利用 SARS 冠状病毒主蛋白酶的晶体结构设计了一系列针对四种冠状病毒都有效的抑制剂 ; 解析了 SARS 的主蛋白酶 猪的肠胃炎病毒主蛋白酶和上述抑制剂的晶体结构 还解析出 SARS-CoV S 蛋白融合核心的晶体结构, 并以融合蛋白 (S) 为研究对象, 解析出小鼠肝炎病毒 (MHV)S 蛋白融合核心的晶体结构, 提出 Spike 蛋白介导冠状病毒膜融合的分子机制 目前, 我们与国内及欧洲的科学家们合作, 共同获得了第六欧盟框架基金的支持, 开展 SARS 诊断及治疗的中 欧合作计划 (SEPSDA) 5 今年研究组最大的亮点是膜蛋白研究方面的进展, 我们在膜蛋白的高效 纯化 结构解析以及新技术方法的探索方面取得突破性的进展, 为 2005 年成果的诞生奠定了坚实的基础 本年度获奖情况饶子和院士被评选为 2004 年国家重点基础研究发展计划 (973 计划 ) 先进个人 46

47 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 Name Position Wenrui Chang Professor Research Topic 3D-Structure and function of photosynthesis membrane proteins Summary of Research 1. The first X-ray structure of LHC-II in icosahedral proteoliposome assembly at atomic detail was solved. One asymmetric unit of a large R32 unit cell contains ten LHC-II monomers. The 14 chlorophylls (Chl) in each monomer can be unambiguously distinguished as eight Chla and six Chlb molecules. Assignment of the orientation of the transition dipole moment of each chlorophyll has been achieved. All Chlb are located around the interface between adjacent monomers, and together with Chla they are the basis for efficient light harvesting. Four carotenoid-binding sites per monomer have been observed. The xanthophyll-cycle carotenoid at the monomer monomer interface may be involved in the non-radiative dissipation of excessive energy, one of the photoprotective strategies that have evolved in plants. 2. The structures of some enzymes have also been solved and the results were published: the ph-profile structure of C-terminal despentapeptide nitrite reductase; the structure of potato calmodulin PCM6; the crystal structure of KD93 and the crystal structure of EFEa and its complex. Selected Publications 1. Zhenfeng Liu, Hanchi Yan, Kebin Wang, Tingyun Kuang, Jiping Zhang, Lulu Gui, Xiaomin An & Wenrui Chang*. Crystal structure of spinach major light-harvesting complex at 2.72 Å resolution. Nature Vol 428, , (2004). 2. Wang F, Wang C, Li M, Gui LL, Zhang JP, Chang WR*. Crystallization and preliminary crystallographic analysis of earthworm fibrinolytic enzyme component B from Eisenia fetida. Acta Cryst. D 60, , (2004). 3. Hai-Tao Li, Chao Wang, Tschining Chang, Wen-Chang Chang, Ming-yih Liu, Jean Le Gall, Lu-Lu Gui, Ji-Ping Zhang, Xiao-Min An, Wen-Rui Chang*. ph-profile crystal structure studies of C-terminal despentapeptide nitrite reductase from Achromobacter cycloclastes. BBRC 316, , (2004). 4. Chao Wang, Feng Wang, Mei Li, Yong Tang, Ji-Ping Zhang, Lu-Lu Gui, Xiao-Min An, Wen-Rui Chang*. Structural basis for broad substrate specificity of earthworm brinolytic enzyme component A. BBRC 325, , (2004). 47

48 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 姓名 职称 研究方向 常文瑞院士研究员副研究员长江学者百人计划国家杰出青年 光合膜蛋白的三维结构与功能研究 本年度工作简介 1. 第一个以正二十面体蛋白脂质体形式堆积的高等植物主要捕光复合物 (LHC-II) 的原子水平的晶体结构被解析 在该晶体的超大 R32 晶胞的一个不对称单位中含有 10 个 LHC-II 单体, 每个单体内的 14 个叶绿素分子可以准确无误地确定为 8 个叶绿素 a 和 6 个叶绿素 b, 每个叶绿素分子的过渡偶极距方位均可确定 6 个叶绿素 b 分布在单体之间的界面上, 与叶绿素 a 一起形成高效吸收光能的基础 每个单体的 4 个类胡罗卜素分子也被确定, 一个参与叶黄素循环的类胡罗卜素分子位于单体之间的介面上, 可能参与了植物在髙光照条件下过多激发能的非幅射耗散从而实现光保护的机制 2. 一些重要功能酶的三维结构也被解析, 其结果也己在不同学术刊物上发表 : 如去羧端 5 肤的亚硝酸还原酶的不同 ph 条件的晶体结构, 土豆植物钙调素 PCM6 的晶体结构,KD93 的晶体结构以及蚓激酶 EFEa 的高分辫率修正及其复合物的晶体结构等 本年度获奖情况 1. 我国科学家破解膜蛋白晶体结构难题 入选 两院院士评选振邦杯 2004 年中国十大科技进展新闻 2. 国家重点基础研究发展计划 (973 计划 ) 先进个人 3. 国家重点实验室计划先进个人 4. 何梁何利基金 科学与技术进步奖 年度北京市科学技术奖一等奖 48

49 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 Name Weimin Gong Position Professor Research Topic Structural Biology Summary of Research Bisphosphoglycerate mutase is a trifunctional enzyme of which the main function is to synthesize 2,3-bisphosphoglycerate, the allosteric effector of hemoglobin. Protein crystals were obtained and diffract to 2.5 Å, producing the first crystal structure of bisphosphoglycerate mutase. The enzyme remains a dimer in the crystal. The conformational changes in the backbone and the side chains of some residues reveal the structural basis for the different activities between phosphoglycerate mutase and bisphosphoglycerate mutase. The bisphosphoglycerate mutase-specific residue Gly-14 may cause the most important conformational changes, which makes the side chain of Glu-13 orient toward the active site. The positions of Glu-13 and Phe-22 prevent 2,3-bisphosphoglycerate from binding in the way proposed previously. In addition, the side chain of Glu-13 would affect the Glu-89 protonation ability responsible for the low mutase activity. Other structural variations, which could be connected with functional differences, were also examined. eif3k, the smallest subunit of eukaryotic initiation factor 3 (eif3), interacts with several other subunits of eif3 and the 40 S ribosomal subunit. eif3k is conserved among high eukaryotes and may play a unique regulatory role in higher organisms. We reported the crystal structure of human eif3k, the first high-resolution structure of an eif3 component. This novel structure contains two distinct domains, a HEAT repeat-like domain and a winged-helix-like domain. Through structural comparison and sequence conservation analysis, we show that eif3k has three putative protein-binding surfaces and has potential RNA binding activity. Selected Publications 1. L Liu, Z Wei, Y Wang, M Wan, Z Cheng & W Gong*. Crystal Structure of Human Coactosin-like Protein. J. Mol. Biol. 344: , (2004). 2. Y Wang, Z Wei, Q Bian, Z Cheng, M Wan, L Liu & W Gong*. Crystal Structure of Human Bisphosphoglycerate Mutase. J. Biol. Chem. 279(37): , (2004) 3. Z Wei, P Zhang, Z Zhou, Z Cheng, M Wan, & W Gong*. Crystal Structure of Human eif3k, the first structure of eif3 subunits. J. Biol. Chem. 279(33): , (2004) 4. Z Zhou, X Song, Y Li & W Gong*. Unique Structural Characteristics of Peptide Deformylase from Pathogenic Bacterium Leptospira interrogans. J. Mol. Biol. 339(1): , (2004). 49

50 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 姓名 职称 研究方向 龚为民院士研究员副研究员长江学者百人计划国家杰出青年 结构生物学 本年度工作简介本年度我们实验室在人类重要蛋白质 抗菌药物设计靶蛋白和植物防御蛋白等研究方面取得重要进展, 根据这些蛋白的晶体结构和功能分析, 共发表研究论文 9 篇, 其中影响因子大于 5 的论文 4 篇, 总影响因子 35 其中主要的研究结果如下 二磷酸甘油酸变位酶 (BPGM) 是一种三功能酶, 其主要功能是合成血红蛋白的别构效应因子 2,3-BPG 从人脑 CDNA 文库中扩增出人二磷酸甘油酸变位酶基因并在大肠杆菌中表达, 并测得 BPGM 的第一个晶体结构 人 BPGM 在晶体结构中以二体形式存在 BPGM 一些残基的主链与侧链的构象变化是构成 BPGM 与 dpgm 不同活性的结构基础 BPGM 特异性残基 Gly14 是诱导构象发生变化的重要因素, 它使 Glu13 的侧链伸入活性中心, 影响 Glu89 质子化, 导致 BPGM 相对较低的变位酶活性 和 dpgm 相比,BPGM 中 Glu13 和 Phe22 的独特位置使催化残基 His11 附近的空间变小, 不利于变位酶活性所要求的 2,3-BPG 中间体的转动, 这可能是导致 BPGM 相对较低的变位酶活性的另一原因 eif3k 是真核翻译起始因子 3(eIF3) 最小的亚基, 它可以与 eif3 的多个其他的亚基以及 40S 核糖体亚基相互作用 eif3k 在高等真核生物 ( 包括哺乳动物 昆虫和植物 ) 中相当保守, 在人体各器官中也普遍存在 ; 但其却不存在于低等真核生物酵母的一些种类中 这说明 eif3k 可能在高等生物中扮演着一个独特的角色 在这篇文章中, 我们报道了 eif3 第一个高分辨率的亚基结构, 人类 eif3k 亚基的晶体结构 此新型结构由两个不同的结构域 (HAM 结构域和 WH 结构域 ) 组成 蛋白质结构比较和氨基酸序列同源性分析表明,eIF3k 有 3 个可能的蛋白质结合表面, 并且 eif3k 可能有 RNA 结合活性 对 eif3k 结构的解析和分析, 为进一步了解 eif3 复合物的结构和功能提供了关键信息 本年度获奖情况 1) 国务院政府特殊津贴 ; 2) 新世纪百千万人才国家级人选 50

51 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 Name Tao Jiang Position Professor Research Topic Crystal structure and functional study of biomacromolecules Summary of Research 1. To understand the processes involved in the catalytic mechanism of pyridoxal kinase (PLK), we determined the crystal structures of PLK.AMP-PCP-pyridoxamine, PLK.ADP.PLP, and PLK.ADP complexes. Comparison of these structures revealed that PLK exhibits different conformations during its catalytic process. After the binding of AMP-PCP (an analogue that replaces ATP) and pyridoxamine to PLK, this enzyme retains a conformation similar to that of the PLK.ATP complex. The distance between the reacting groups of the two substrates is 5.8 A, indicating that the position of ATP is not favorable for spontaneous transfer of its phosphate group. However, the structure of the PLK.ADP.PLP complex exhibited significant changes in both the conformation of the enzyme and the location of the ligands at the active site. Therefore, it appears that after binding of both substrates, the enzyme-substrate complex requires changes in the protein structure to enable the transfer of the phosphate group from ATP to vitamin B(6). Furthermore, a conformation of the enzyme-substrate complex before the transition state of the enzymatic reaction was also hypothesized. 2. (R)-roscovitine is a selective inhibitor of Cyclin-dependent kinases (CDKs). Recent biochemical investigations have shown that (R)-roscovitine interacts with PLK selectively. We determined the crystal structure of PLK in complex with (R)-roscovitine. Structural analysis revealed that roscovitines bind in the pyridoxal binding site, rather than in the anticipated ATP-binding site. This affords a better understanding of the catalysis mechanism of PLK and could aid in the design of roscovitine derivatives displaying strict selectivity for either PLK or CDKs. Selected Publications 1. Li MH, Kwok F, Chang WR, Liu SQ, Lo SC, Zhang JP, Jiang T*, Liang DC. Conformational changes in the reaction of pyridoxal kinase. J. Biol. Chem. 279 (17), (2004). 2. Liu JF, Wang XQ, Wang ZX, Chen JR, Jiang T, An XM, Chang WR, Liang DC*. Crystal structure of KD93, a novel protein expressed in human hematopoietic stem/progenitor cells. J. Struct. Biol. 148(3), (2004). 51

52 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 姓名 职称 研究方向 江涛院士研究员副研究员长江学者百人计划国家杰出青年 生物大分子晶体结构与功能研究 本年度工作简介 1. 为了阐明吡哆醛激酶的催化反应过程, 我们解析了 PLK -AMP-PCPpyridoxamine PLK-ADP-PLP 和 PLK-ADP 三种复合物的结构, 结构比较表明吡哆醛激酶在催化过程中有不同的构象 结合 AMP-PCP 和 Pyridoxamine 后, 酶的构象和复合物 PLK-ATP 的构象相似, 两种底物的反应基因之间的距离为 5.8Å, 表明 ATP 的位置不利于其磷酸基因的转移, 然而在复合物 PLK-ADP-PLP 中, 酶本身以及结合位点都发生很大的构象变化 因此, 在结合了两种底物后, 酶的构象改变以利于磷酸基团从 ATP 转移到 VB6 此外, 文中提出了转移反应之前酶和两种底物相互作用模型 (pre-reaction state model) 2. (R)-roscovitine 是细胞周期依赖激酶的选择性抑制剂, 最近的生物学研究表明, (R)-roscovitine 可以和 PLK 选择性的结合, 我们解析了 PLK 和 (R)-roscovitine 复合物的晶体结构, 结构分析表明 roscovitine 结合在 PLK 的吡哆醛结合位点, 而不是我们预先推测的 ATP 结合位点, 这有助于我们了解 PLK 的催化机理, 并且能够帮助我们设计对 PLK 或 CDK 更具有选择性的 roscovitine 的衍生物 本年度获奖情况 52

53 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 Name Liang Wei Position Associate Professor Research Topic Targeting delivery system of bio-macromolecular drug Summary of Research Permeation peptides, cell penetrating peptides (CCPs), or protein transduction domains (PTDs), including the Tat-peptide basic domain, have been intensely studied recently after early observations demonstrated rapid translocation properties, seemingly independent of receptor-mediated endocytosis, into various cell types. The ability of permeation peptides to cross cell membranes is specially promising for drug delivery, since one major barrier to successful therapies and imaging is the need for membrane translocation. A variety of multicationic oligomers, including arginine-rich peptides, as simple as 7-11 consecutive arginines are known to be more effective. Cargoes ranging in size from metal chelates, fluorescent dyes and proteins to iron oxide nanoparticles and liposomes can be delivered into cells, although the detailed mechanisms and subcellular localizations remain poorly understood and may differ, depending on cargo size, cell type, CPP sequence, and other experimental variables. We are attempting to develop a new strategy for sirna delivery into cells using arginine octamer peptides (R8) attached to the PEG-liposome surface (R8-liposomes). One assumes that liposomes formulated using a charge neutral ratio of DOTAP/siRNA, adding 3% mol PEG-DSPE, post-inserting 0.5-1% mol R8-PEG-PE on the outer membrane of liposomes will provide an effective transfection system in vitro or in vivo. In this case, sirna containing R8-liposomes should be nontoxic, and prevent aggregation and decrease adsorption of serum proteins and interaction with non-target cells. The presence of R8 increases the uptake capacity of sirna into cells and subsequent accumulation in the cytoplasm compartment and enhances sirna functions. We have examined cellular uptake and distribution of Rh-PE labeled liposomes with and without R8 attached by fluorescence microscopy. We have also compared transfection efficiency by using sirna containing R8-liposomes and lipofectamine 2000-siRNA complexes in the presence of serum proteins or in their absence. Finally, we are testing the toxicity of sirna containing R8-liposomes and sirna-free R8-liposomes and the biologic activity of sirna using RT-PCR and western blotting assay. The results provide new insights into the potential applications of R8-liposomes for delivery of sirna.(in preparation for publication.) Selected Publications 1. LIANG Wei*, DAVALIAN Dariush, TORCHILIN Vladimir P. The interaction of a novel peptoid enhancer oligomer of arginine with bovine submaxillary mucin. Acta Pharm Sinica. 2004, 39(12): LIANG Wei*, Wang Ya-Qin, DAVALIAN Dariush. A Novel HIV-1 Therapeutic Target: Tat transactivator Protein. Prog. Biochem. Biophys. 2004, 31(9): LIANG Wei*, Wang Ya-Qin, LIU Xing-Jun. Targeting intervention for tumor therapy. Chin J Inetrv Imaging Ther. 2004, 1(1):

54 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 姓名 职称 研究方向 梁伟院士研究员副研究员长江学者百人计划国家杰出青年 生物大分子药物的定向输送系统和药物作用靶标的筛选 本年度工作简介实验室组建已基本完成, 部分研究工作已开展 根据本实验室研究工作的需要及符合我所将来发展的要求, 我们建立了蛋白质与多肽的定位修饰技术 ; 适合生物大分子药物输送的自组装纳米制备技术 ; 检验和鉴定生物大分子药物功能的分子生物学和细胞生物学评价技术 已完成和正在进行的研究工作有 : 1 基于 HIV-1 病毒 TAT 蛋白碱性氨基酸区域的 TAT 的寡肽是一段核定位信号肽且具有不依赖内吞机制迅速穿透生物膜的功能,TAT 这一特点尤其适合作为核酸类物质的胞内输送的载体 在 TAT 寡肽的基础上, 我们筛选出了比其穿膜能力更强的由 8-12 个精氨酸组成的寡聚体 构建了八聚精氨酸联接的 PEG 化磷脂自组装胶束并实现了将其定量地插入到包载 hmd2 sirna 的长循环脂质体的脂双层的外层膜上, 三种不同肺癌细胞株体外转染试验显示其转染效率显著高于目前广泛使用的基因转染试剂 Lipofectamine 2000, 且毒性明显低于 Lipofectamine 2000, 对 Lipofectamine 2000 转染无效的肺鳞癌成功地实现了靶基因的沉默和相关基因 P53 的高表达 2 基于 PEG 化磷脂自组装胶束的前列腺素 E1 制剂已申请国家发明专利且通过初审 ( 申请号 : ), 其药动学和药效学试验正在进行中 3 检定出特异性识别肺肿瘤细胞的肽类配基的序列 ; 稳定的抗肿瘤药物高包载量的 PEG 化磷脂纳米胶束的筛选工作已完成 ; 细胞内药物含量的测定方法已建立 ; 原位肺癌动物模型的建立正在预试中 本年度获奖情况 54

55 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES Molecular Membrane Biology Name FuyuYang Position Professor, Member of CAS Research Topic A novel member from lysosome-chymotrysinogen B is involved in Apoptosis Summary of Research A protease which can convert Bid into its active form (t Bid) inducing cytochrome c release through mitochondrial outer membrane has been isolated and purified from rat liver lysosomes. Peptide mass fingerprinting and biochemical analysis identified this protease as Chymotrypsin B (Ctr b), a robust and stable serine endopeptidase previously known as a digestive enzyme with expression restricted to the pancreas. Our results clearly demonstrate that intracellular Ctr b is lysosomally localized and it is conceivable that it can be leaked into the cytosol as a pro-apoptotic molecule. Selected Publications 1 Yongfang Zhao,Xiaoxuan Fan, FuyuYang* and Xujia Zhang* Gangliosides modulate the activity of the plasma membrane Ca 2+ -ATPase from porcine brain synaptosomes. Arch. Biochem. Biophysics 427, , (2004). 2 Xiaoping Wang, Xuehai Han and Fuyu Yang* Critical segment of apocytochrome c for its insertion into membrane. Mol. Cell. Biochem. 262, 61-69, (2004). 55

56 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 膜分子生物学 姓名 职称 研究方向 杨福愉院士研究员副研究员长江学者百人计划国家杰出青年 生物膜的结构与功能 本年度工作简介一 与张旭家博士等合作研究神经节苷脂 脂筏 (lipid rafts) 对细胞质膜 肌浆网膜 Ca 2+ -ATP 酶构象与活性的影响 ( 详见张旭家的报告 ) 二 溶酶体 - 线粒体途径诱发细胞凋亡的研究在研究 Bid 诱发线粒体通过外膜释放细胞色素 c(cyt.c) 过程中发现溶酶体含有多种蛋白酶能酶介 Bid tbid 从而使 Cyt.c 从线粒体中释放并引起凋亡 我们经长期研究从大鼠肝溶酶体中分离纯化一种蛋白因子, 称之为 LBCP(Lysosomal Bid Cleavage Protease) 通过 LC-MS/MS 分析 酶切位点 生化特点等测定, 初步确定 LBCP 为 Chymotrypsinogen B 这一结果进一步结合 Bioinformatics 并用 RT-PCR 等方法得到了验证 Chymotrypsinogen 一般都认为仅分布于胰脏 我们从大鼠肝溶酶体中发现它的存在, 经活化后它能酶切 Bid-tBid 并能使线粒体外膜的透性发生改变释出 Cyt c. 诱发凋亡 迄今国内外未见类似报道 本年度获奖情况 56

57 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 Name Tao Xu Position Professor Research Topic The Molecular Mechanism of Membrane Trafficking Summary of Research We have made progress in the following four aspects: 1) A new pathway of exocytosis regulation. We have determined that PKC could increase the sensitivity of regulated exocytosis, which is the first evidence that Ca 2+ -sensing of secretion could be regulated by phosphorylation. By using kinetic model to analyze the effects of PKC on the exocytosis dynamics, we concluded that PKC increased the exocytosis sensitivity by reducing the Ca 2+ binding sites on the sensor protein without changing other kinetics parameters. The reduction of the Ca 2+ binding site reduced the threshold of the Ca 2+ need for secretion, and increased the fusion probability of vesicles far from the voltage gated calcium channel. 2) The investigation on the exocytosis of secretory lysosome The anti-infection, anti-virus, anti-tumor and anti-transplant effects of Natural killer cells (NK cells) depend on the exocytosis of secretory lysosome. We find that mass secretory lysosmes are produced in NK cells after target cell recognition. PKC but not Golgi apparatus is found to be involved during this process. 3) The development of detection method of single vesicle activities in live cells The study of vesicle transport and trafficking within cells is a hot topic. Previous researches were mainly focused on monitoring two-dimensional movement of vesicle within live cells. We have combined three-dimensional deconvolution fluorescence microscopy technique and single particle trafficking technique, which enable us to monitorl movements of single vesicle within PC-12 cells. 4) The dynamics of single GLUT4 vesicle in live cells GLUT4 is critically important in the blood sugar homeostasis. Evidences prove the defect of GLUT4 transport is linked to type II diabetes. The molecular mechanism by which GLUT4 storage vesicle (GSV) synthesizes and cycles is not clear. We have specifically labeled GSVs with green fluorescent protein. Using total internal reflection fluorescence microscopy, we have monitored the movements of GSV within the live cell and studied the characteristics of those movements. This result provides the basis for further investigation of the molecular mechanism of GSV movement.ther investigation of the molecular mechanism of GSV movement. Selected Publications [1]Hua Yang, Huisheng Liu, Zhitao Hu, Hongliang Zhu, and Tao Xu*. PKC-induced sensitization of Ca2+-dependent exocytosis is mediated by reducing the Ca2+-cooperativity in pituitary gonadotropes. Journal of General Physiology. (Accepted) [2]Dongfang Liu, Liang Xu, Fan Yang, Dongdong Li, Feili Gong, and Tao Xu*. Rapid biog enesis and sensitization of secretory lysosomes in NK cells mediated by target-cell recognition. PNAS. (In Press) [3]Qun-Fang Wan, Yongming Dong, Hua Yang, Xuelin Lou, Jiuping Ding, Tao Xu*. Protein Kinase Activation Increases Insulin Secretion by Sensitizing the Secretory Machinery to Ca2+. Journal of General Physiology. Vol.124:1-11. (2004) [4]Chen Hong LI, Li BAI, Dong Dong LI, Sheng XIA, Tao XU*. (2004) Dynamic tracking and Mobility analysis of single GLUT4 storage Vesicle in live 3T3-L1 Cells. Cell Research. (In Press) [5]Xia S, Xu L, Bai L, Xu ZQ, Xu T*. Labeling and dynamic imaging of synaptic vesicle-like microvesicles in PC12 cells using TIRFM. Brain Research. Vol.997(2): (2004) [6]DD Li, J Xiong, AL Qu, and T Xu*. Three-dimensional tracking of single secretory granules in live PC12 cells. Biophysical Journal. Vol.87(9): (2004) 57

58 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 姓名 职称 研究方向 徐涛院士研究员副研究员长江学者百人计划国家杰出青年 膜转运的分子机制 本年度工作简介我们主要在以下四个方向取得了具有代表性的成果 1) 发现了一条新的调控分泌的机制我们发现 PKC 可以显著增强分泌对钙离子的敏感性, 从而首次证明了分泌的钙离子传感可被蛋白磷酸化所调控 为了进一步揭示蛋白磷酸化调节钙离子传感器的内在机理, 我们利用动力学模型拟合, 分析 PKC 对钙触发分泌的动力学特性的影响 我们发现 PKC 减少钙离子传感器上的钙离子结合位点, 从而增强分泌对钙的敏感性 ; 但动力学参数并没有显著的变化 钙离子结合步骤数目的减少意味着分泌所需钙离子的阀值的降低 ; 同时距离钙离子通道的较远囊泡, 分泌的可能性也提高了 该文章已被 JGP 接受 2) 分泌型溶酶体分泌活动的调控机制研究自然杀伤细胞 (Natural killer cells, NK) 的抗感染, 抗病毒, 抗肿瘤以及移植排斥效应有赖于一种叫做 " 分泌型溶酶体 " 的分泌活动 我们发现 NK 细胞在杀伤靶细胞时需要大量重新生成分泌型溶酶体, 其快速产生的过程不涉及高尔基体但需要 PKC 的参与 此工作已被 PNAS 接受, 有关工作还在进一步研究和整理中 3) 发展了活细胞单个囊泡活动的检测方法随着显微成像技术和活体细胞荧光标记技术的发展, 直观观察和研究分泌囊泡的胞内转运过程已经成为目前细胞分泌研究的热点之一 但是国际上关于这方面的研究工作一直局限于分泌囊泡在细胞内二维平面内的运动, 我们发展了三维荧光反卷积显微成像技术和单微粒跟踪技术成功地跟踪了 PC-12 细胞中单个分泌囊泡在全细胞范围内的三维运动过程并揭示了其运动规律, 为以后深入阐明分泌囊泡转运过程中涉及的分子机制提供了技术基础 这一研究成果发表在 Biophys. J 上, 同时配发了特约评论文章, 对该工作给予好评 4) 活细胞中单个 GLUT4 囊泡动力学研究葡萄糖转运子 4(Glucose Transporter 4, GLUT4) 在人体整体糖的平衡中起到了重要的作用, 有越来越多的证据表明 II 型糖尿病与 GLUT4 的转运障碍有关 目前, 储存 GLUT4 的囊泡 (GLUT4 storage vesicles,gsv) 的生成和循环途径及其分子机制尚不清楚 我们采用 EGFP 特异地标记了 GSV 并运用全内反射荧光显微技术 (TIRFM) 研究了 GSV 在活细胞内的动态运动及其规律 此研究结果已发表在 Cell Research 杂志上, 为进一步揭示 GSV 运动的分子基础奠定了基础 本年度获奖情况 58

59 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 Name James Q. Yin Position Professor Research Topic RNAi and Functional Genomics Summary of Research SiRNA and mirna can efficiently induce mrna cleavage and/or translational repression at the posttranscriptional level in a sequence-specific manner. Recently, it has been demonstrated that these small RNAs guide genome modification in mammalian cells. However, their ability to direct cognate DNA methylation has been confirmed so far only in plants, and their function and mode of silencing are still elusive. We have reported that small RNAs derived from intron regions of the p53 gene (p53-srna) can target homologous DNA sequence in the promoter region of the heat shock protein 70.1 (hsp70.1) gene in both MCF7 and HeLa cells. Vector-based small RNA repressed expression of the hsp70.1 gene at the transcriptional level. Western blot analysis indicated that the expression of heat shock protein 70, a protein related to cell survival, was greatly reduced or completely inhibited owing to the promoter methylation of the hsp70.1 gene. These findings reveal that the expression of the p53 gene can regulate cell activities at both protein and RNA levels, suggesting that p53-srna may be a novel strategy for therapeutic gene silencing of tumor cells in human beings. Similarly, we also discovered several sirnas from an intron of BRCA1 and their target mrnas by using bioinformatics, and confirmed their existence in human cells with Northern blot analysis. Our results demonstrated that these sirnas could regulate gene expression by different modes of action. In addition, my research group conducted some experiments concerning the development of red blood cells. The primary findings showed that there was a separation of apoptotic nuclei from survival cytoplasm. DNA microarray data showed that enzymes associated with cell apoptosis and survival display differential synthesis in pro-erythroblast and late-stage erythroblast cells. Moreover, we are documenting data concerning the effects of SMAD3-siRNA on liver fibrosis, and the inhibitory roles of hdm2-sirna on lung cancer cells. Selected Publications 1. Tan FL& Yin JQ*. RNAi, a new therapeutic strategy against viral infection. Cell Res. 14(6), (2004). 2. Liu TG, Yin JQ*, Shang BY, Min Z, He HW, Jiang JM, Chen F, Zhen YS, Shao RG. Silencing of hdm2 oncogene by sirna inhibits p53-dependent human breast cancer. Cancer Gene Ther. 11(11), (2004). 3. Tan FL & Yin JQ*. RNAi, a new strategy for the treatment of cancers. Frontier Bioscience in press. 59

60 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 姓名 职称 殷勤伟 院士研究员副研究员长江学者百人计划国家杰出青年 研究方向 RNAi 技术和功能基因组学 本年度工作简介近年来, 从植物到人类的细胞中发现了一类内源性的 长度约 22 个核苷酸的非编码 RNA, 称为小 RNA 小 RNA 可由内源性基因编码, 位于编码基因间区或编码基因的内含子或外显子区, 通过 RNA 聚合酶 II 而被转录, 其前体含有冒状结构和多聚腺苷酸 它们在指导 mrna 分子的转译抑制 降解断裂或其它形式的调节通路中发挥重要作用 目前已经知道小 RNA 可以发挥多种功能, 如组织器官的定向发育 细胞生长分化的时空调节 信号通路的开启和关闭 细胞周期的监测与调控 学习与记忆 肿瘤的发生与凋亡 肥胖和衰老等 在 2004 年, 我研究组采用生物信息学技术在 p53 和 BRCA1 基因的内含子中发现了具有调节功能的小 RNA 分子, 经 RT-PCR 和 Western Blot 证明这些小 RNA 分子能够以序列特异的方式沉默相应的靶基因 来源于 p53 基因内含子的小 RNA 能够诱导 Hsp70-1 基因操纵子的甲基化从而抑制 Hsp70-1 基因的表达 尽管寻找小 RNA 的靶基因是件非常困难的任务, 但幸运的是 BRCA1 内含子的小 RNA 却有许多靶基因 来源于 BRCA1 基因内含子的小 RNA 能以不同的方式作用于靶基因从而调节那些基因的表达 这些小 RNA 分子亦能引起肿瘤细胞的行为和形态学的变化, 进一步的深入而系统的研究在进行中 总之, 根据其它实验室的研究结果和我们的发现, 可以提出如下假说 : 单一的小 RNA 能有多个不同的靶基因 同样, 多个不同的小 RNA 亦可与同一个 mrna 分子相结合并调控其生物活性 随着编码基因的表达, 从内含子和外显子剪切出的小 RNA 可以与其上下游的靶子 mrna 的不同结构区域结合, 从而指导转译抑制 切割降解或甲基化 小 RNA 的发现丰富了人们对蛋白质合成控制的认识, 补充了在 RNA 水平对靶 mrna 分子进行更迅速和有效的调节, 展现了细胞内基因表达调控全方位多层次的网络系统 此外, 我研究组还发现了红细胞核质凋亡分离的现象并在研究探讨可能的分子机制 我研究组也在研发 SMAD3-siRNA 对肝硬化的防治作用和 hdm-2-sirna 对肺癌细胞的生长抑制作用 本年度获奖情况 60

61 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 Name Huang Youguo Position Professor Research Topic Activity and Conformation of Membrane Proteins Summary of Research Main results on activity and conformation change of MRP1, a typical transmembrane protein, obtained in 2004, were as follows: 1. The relationship between MRP1 activities and its NBD conformational changes MIANS, a sulfhydryl-reactive fluorescence, was used to label the cysteines of MRP1 (multidrug resistance protein), and the results indicated that an increase in fluorescence intensity and a large emission blue shift took place after two Cys residues of MRP1 reacted with MIANS, which demonstrated that labeled Cys residues in MRP1 reside in a relatively hydrophobic environment. The experimental results obtained from fluorescence resonance energy transfer further uncover that two Cys residues of MRP1 modified by MIANS located in the vicinity of its NBDs, of which one lies close to NBD1, and the other near NBD2. ATP, ADP and anticancer drugs can all reduce the rate of reaction of MRP1 with MIANS. The collisional quenchers, acrylamide, l-, and Cs+ were used to assess local environments of MIANS bound to MRP1 and the results showed that the region around the MIANS-labeled cysteine is positively charged. Both MIANS and NEM, which are sulfhydryl-reactive reagents, inhibited MRP1 ATPase activity, whereas anticancer drugs activated it. These results demonstrated that all nucleotides and drugs could induce changes in conformation of the NBDs in MRP1. Nucleotides can bind directly to NBDs, but drugs may react first with TMDs, which in turn alters the accessibility of the two Cys residues bound by MIANS and affects MRP1 ATPase activity, which is coupled with the transport of its substrates. Taken together, the above experimental results provide direct evidence for further study on the coupling of translocation of the transported species to hydrolysis of ATP in MRP1. 2. Fluorescent modified phosphatidylcholine floppase activity of reconstituted multidrug resistance-associated protein MRP1 Multidrug resistance-associated protein (MRP1) may function as a floppase in human red blood cells to translocate phosphatidylserine and/or phosphatidylcholine from inner membrane leaflet to outer leaflet. Here we report that the purified and reconstituted MRP1 protein into asolectin proteoliposomes is mainly in an inside-out configuration and possesses the ability to flop a fluorescent labeled phosphatidylcholine (NBD-PC) from outer leaflet (protoplasmic) to inner leaflet (extracytoplasmic). The reconstituted MRP1 protein retains endogenous ATPase activity. ATP hydrolysis is required for the flopping since removal of ATP and/or Mg2+ inhibits the translocation of NBD-PC. Further evidence to support this conclusion is that the translocation of NBD-PC is inhibited by vanadate, which traps ATP hydrolysis product ADP in the nucleotide binding domains. In addition, the translocation of NBD-PC by proteoliposomes containing MRP1 protein is in a glutathione-dependent manner, similar to the process of translocating anticancer drugs such as daunorubicin. Verapamil, vincristine, vinblastine, doxorubicin and oxidized glutathione partially inhibited the translocation of NBD-PC, whereas MK 571, an inhibitor of MRP1 protein, inhibited the translocation almost completely. Taken together, the purified and reconstituted MRP1 protein possesses the ability to flop NBD-PC from outer to inner leaflet of the proteoliposomes. Selected Publications 1. Huang, Z. and Huang, Y.* The relationship between MRP1 activities and its NBD conformational changes. Sci. China. C Life Sci. 47(5): , Huang, Z., Chang, X., Riordan, J.R., Huang, Y.* Fluorescent modified phosphatidylcholine floppase activity of reconstituted multidrug resistance-associated protein MRP1. Biochim. Biophys. Acta. 1660(1-2): ,

62 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 姓名 职称 黄有国 院士研究员副研究员长江学者百人计划国家杰出青年 研究方向 膜蛋白的活性和构象 本年度工作简介本年度在多药耐药蛋白 MRP1 的活性和构象变化的关系方面取得的主要研究结果如下 : 1 MRP1 的 NBDs 结构域的构象变化与其转运药物的 ATP 酶活性相关用专一性标记蛋白分子中半胱氨酸 (Cys) 的荧光探针 MIANS 标记 MRP1(multidrug resistance protein) 的实验结果表明,MIANS 与 MRP1 中的 2 个 Cys 结合, 结合后不仅荧光强度增加, 而且发射波长蓝移, 表明所标记的位点处于相对疏水的环境, 荧光共振能量转移实验进一步表明,MIANS 标记的 Cys 与 MRP1 核苷酸结合结构域 (NBDs) 很接近, 其中 1 分子 MIANS 标记在 NBD1 附近, 另一分子标记在 NBD2 附近,ATP,ADP 以及化疗药物能阻止 MIANS 对 MRP1 的标记, 猝灭剂丙烯酰胺 Cs 和 I 对 MIANS-MRP1 荧光猝灭实验又表明,MIANS 标记的 Cys 位点处于一个带正点电荷的区域, 同时, 巯基结合试剂 MIANS 和 NEM 对 MRP1 ATP 酶的活性具有明显的抑制, 而化疗药物却有明显的激活作用, 上述实验结果提示, 核苷酸和化疗药物都能引起 MRP1 的 NBDs 的构象变化, 核苷酸可直接与 NBDs 结合, 而化疗药物可能通过改变 MRP1 的跨膜结构域 (TMDs) 的构象进而影响 NBDs 的构象, 从而调控 MRP1 ATP 酶的活性并影响对化疗药物的转运, 结果对于深入揭示 MRP1 的 ATP 的结合和水解与其转运化疗药物之间的偶联提供了较直接的实验证据 2 MRP1 的磷脂翻转酶活性可能与药物转运相关用纯化的 MRP1 蛋白的重建实验表明, 重建 MRP1 具有翻转 NBD-PC( 磷脂酰胆碱荧光类似物 ) 的活性 除去 ATP 和 Mg 2+ 可抑制其转运活性, 进一步的实验表明, 与 NBD 结合的 vanadate 亦抑制 NBD-PC 的转运 MRP1 的 NBD-PC 的转运与其它一些药物的转运具有相似性即依赖于 glutathione, 而且 MRP1 的专一性抑制剂 MK571 可完全抑制其转运活性 上述实验结果表明, 重建 MRP1 具有转运 NBD-PC 的能力, 这种转运亦可能与肿瘤细胞的多药耐药性有关 综上,MRP1 是 1992 年才确认的一种多药耐药蛋白, 越来越多的证据表明, 它如同 MDR 一样参与了肿瘤多药耐药性调控 本年度获奖情况 62

63 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 Name Chen Jianwen Position Professor Research Topic Caveolae and multi-drug resistance Summary of Research 1. Cholesterol is a key lipid in mediating the enzyme activity or signaling pathway of many proteins on the plasma membrane in mammalian cells. We demonstrated for the first time that after overexpressing caveolin-1, the plasma membrane cholesterol level was decreased by about 12% and 30% for doxorubicin-sensitive and doxorubicin-resistant Hs578T breast cancer cells, respectively. However, the total cholesterol level in both cell lines was increased by about 10%. By measuring fluorescence and flow cytometry using the fluorescence dyes 1,6-diphenyl-1,3,5-hexatriene and Merocyanine 540, we found that overexpressing caveolin-1 resulted in a similar increase in membrane fluidity and loosening of lipid packing density as cholesterol depletion by 1 mm methyl-b-cyclodextrin (MbCD) or 2-hydroxypropyl-b-cyclodextrin (HbCD). Moreover, we found that the transport activity of P-gp was significantly inhibited by 1 mm MbCD or HbCD, which is also similar to the inhibitory effect of caveolin-1 overexpression. Our data demonstrated for the first time that the reduction of the plasma membrane cholesterol level induced by overexpressing caveolin-1 may indirectly inhibit P-gp transport activity by increasing plasma membrane fluidity. 2. Caveolin-1, the principal component of caveolae, is a kda integral membrane protein. The interaction of the caveolin-1 scaffolding domain with signaling molecules can functionally inhibit the activity of these signaling proteins. Little is known about how caveolin-1 influences the expression of Pglycoprotein (P-gp), an ABC transporter encoded by multi-drug resistance (MDR1) gene. To elucidate the possible mechanism between caveolin-1 and P-gp expression, we overexpressed caveolin-1 in Hs578T/Dox breast adenocarcinoma cells, a multidrug resistant line, and then selected single clone cells with high levels of caveolin-1 expression. Both Western blot and confocal microscopy analyses showed that caveolin-1 was markedly overexpressed in the transfectants, while P-gp protein was almost abolished. Reverse transcription polymerase chain reaction also showed that the expression of P-gp mrna was significantly suppressed in the transfectants. This was confirmed further by Northern blot analysis. Moreover, through measuring the changes of drug resistance and P-gp transport activity in the transfectants, we found that overexpression of caveolin-1 reversed drug resistance of transfectants and lowered their P-gp transport activity to the level of Hs578T/S. Taken together, our results indicate that such suppression of P-gp in the transfectants overexpressing caveolin-1 may occur at the transcriptional level. Selected Publications 1. Yu-hong Pang and Jian-wen Chen *. Anisodamine Causes the Changes of Structure and Function in the Transmembrane Domain of the Ca2+-ATPase from Sarcoplasmic Reticulum. Biosci. Biotechol. Biochem 68 (1), , Chuanxi CAI and Jianwen CHEN*. Overexpression of Caveolin-1 Induces Alteration of Multidrug Resistance in Hs578T Breast Adenocarcinoma Cells. Int. J. Cancer 111, , Chuanxi Cai, Hua Zhu, and Jianwen Chen*. Overexpression of Caveolin-1 Increases Plasma Membrane Fluidity and Reduces P-glycoprotein Function in Hs578T/Dox Biochem.Biophys.Research.Communication 320, , Hua Zhu, Chuanxi Cai, Jianwen Chen*. Suppression of P-glycoprotein gene expression in Hs578T/Dox by the overexpression of caveolin-1. FEBS Letters 576, ,

64 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 姓名 职称 研究方向 陈建文院士研究员副研究员长江学者百人计划国家杰出青年 Caveolae 和多药耐药性 本年度工作简介 1. 胆固醇在介导哺乳动物细胞质膜上存在的许多蛋白的酶活和信号途径中是一个很关键的脂分子 我们的研究首次证明了在过表达 caveolin-1 后, 对阿霉素药物敏感和耐药的乳腺癌细胞 Hs578T 的质膜胆固醇水平分别降低了 12% 和 30% 但是, 两种细胞的总胆固醇水平都增大了约 10% 通过流式方法测定 DPH 和 MC 540 的荧光, 我们发现 caveolin-1 的过表达会引起膜流动性的增加和脂分子堆积密度的减低, 这与用 1 mm MbCD 或 HbCD 去除胆固醇的处理后得到的结果是相似的 并且, 我们发现 P-gp 的转运活性明显受到 1 mm MbCD 或 HbCD 的抑制, 这与过表达 caveolin-1 后的抑制程度也是相似的 我们的数据第一次表明, 由于过表达 caveolin-1 而导致的质膜胆固醇水平的降低可能通过增加膜流动性的方式间接抑制了 P-gp 的转运活性 2.Caveolin-1, 作为 caveolae 的主要组成成分, 是一个 KDa 的整合膜蛋白 Caveolin-1 的 scaffolding 结构域与信号分子的相互作用通常会抑制这些信号蛋白的活性 目前关于 caveolin-1 是如何影响 P 型糖蛋白 (P-gp), 一种多药耐药基因 (MDR1) 编码的 ABC 转运蛋白的表达, 了解的还不是很清楚 为了阐明 caveolin-1 与 P-gp 表达之间可能的机制, 我们在 Hs578T/Dox 这种多药耐药的乳腺癌细胞系中过表达了 caveolin-1 并筛选了高表达量的单克隆细胞 Western blot 和 confocal 实验的结果都显示 caveolin-1 大量过表达于转染细胞中, 但是 P-gp 蛋白基本已经消失 反转录聚合酶链反应实验表明在转染细胞中 P-gp 的 mrna 转录受到了明显的抑制 Northern blot 实验也进一步证明了这一结果 并且, 通过测定转染细胞中 P-gp 的耐药性和转运活性的变化, 我们发现 caveolin-1 的过表达逆转了转染细胞的耐药性以及使 P-gp 的转运活性降低到 Hs578T/S 的水平 我们的结果表明在过表达 caveolin-1 的转染细胞中, P-gp 的这种抑制可能发生在转录水平上 本年度获奖情况 64

65 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 Name Position Research Topic Zhang Xujia Professor Molecular Biology of Membrane Proteins Summary of Research 1. Electron microscopic study of V-ATPase from mung bean a. The vacuolar H + -ATPase from mung bean (Vigna radiate L.) was purified to homogeneity. The purified complex contained all the reported subunits from mung bean, but also showed a 40 kda subunit. The excised 40 kda band was also processed for HPLC/MS/MS peptide mapping. A peptide sequence obtained, YPPYQAIFSK, shows identity with the d subunit of Arabidopsis thaliana. This then strongly suggests that the band at 40 kda represents the d subunit of the V-ATPase from mung bean. b. Subcomplexes (V o ) with or without subunit d were purified and reconstituted into soybean liposomes. The V o subcomplex containing subunits a and c is a passive proton channel driven by K + /valinomycin mediated membrane potential and inhibited by bafilomycin A 1 or DCCD, while the V o subcomplex containing subunits a, c and d is not. This result suggested that the subunit d was just above the center of the c ring, like a cap to regulate the H + translocation across the membrane. 2. Sphingolipids regulate the plasma membrane Ca 2+ -ATPase from erythrocyte ghosts a. In contrast to the effect of sphingolipids on the PMCA from porcine brain synaptosomes, the sphingolipids GM1, GM2, GM3 and GD1b greatly stimulate the PMCA, but Asiola-GM2 has no apparent effect. This result suggests that the regulation of the PMCA by sphingolipids is isoform dependence, in view of the fact that porcine brain synaptosomes contain PMCA1 - PMCA4, but the erythrocyte ghosts contain mainly PMCA4. b. The mechanism by which the gangliosides regulate the PMCA was systemically studied. The activities of the PMCA in the presence of CaM, polypeptide and calpain were measured. The results demonstrated that gangliosides stimulate the PMCA via their direct interactions with the c-termini of the PMCA, subsequently releasing active sites. Interestingly, GM2 stimulated the enzyme at lower concentrations, but inhibited it at higher concentrations. This result suggests that GM2 interacts with two regions of the PMCA: one located at the c-termini, the other at the self-inhibition region of the PMCA. Selected Publications 1. Yongfang Zhao, Xiaoxuan Fan, Fuyu Yang,* and Xujia Zhang*. Gangliosides modulate the activity of the plasma membrane Ca 2+ -ATPase from porcine brain synaptosomes. Archives of Biochemistry and Biophysics 427, , (2004). 2. Zhuo Li, Xujia Zhang*. Electron-microscopic structure of the V-ATPase from mung bean. Planta 219, , (2004). 65

66 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 姓名 职称 研究方向 张旭家院士研究员副研究员长江学者百人计划国家杰出青年 膜蛋白的结构与功能 本年度工作简介一 绿豆 V-ATPase 电镜单颗粒的结构研究 V-ATPase 是一种多亚基复合物 在上一年获得绿豆 V-ATPase 电镜结构后, 开始研究每一种亚基的空间定位及其功能 1 首次发现绿豆 V-ATPase 包含一个与膜结合的 40kDa 的亚基, 生化实验表明该条多肽是 V o 的亚基 用串联质谱鉴定出它与植物 V-ATPase 的 d 亚基高度同源 表明该多肽是绿豆 V-ATPase 的 d 亚基, 我们称之谓 d mb 2 在不同条件下分离纯化包含 d 亚基和不含 d 亚基的两种 V o 亚复合物 分别将其重组到大豆磷脂制成的的脂质体, 发现 d 亚基的有或无直接对应 V o 亚复合物被动质子转运活性的关闭或开放状态, 提出 V o 被动质子转运通道可能位于 c 亚基环的中央孔,d 亚基则相当于一个 盖子 控制通道的开闭 考虑 d 亚基广泛存在于 V-ATPase 中, 该模型可能具有广泛意义 二 神经节苷脂对质膜 Ca 2+ -ATPase(PMCA) 的调控既上一年发现神经节苷脂对脑中 PMCA 具有调控作用的基础上, 开始系统地研究神经节苷脂与 PMCA 的作用机理, 包括与不同亚型 PMCA 的作用 1 神经节苷脂 GM1 GM2 GM3 和 GD1b 对从血中提取的 PMCA 都有激活效应, 而 Asialo-GM1 对 PMCA 没有作用 这一结果与我们以前从脑中提取的 PMCA 完全相反, 表明神经节苷脂对 PMCA 的调控与 PMCA 的亚型相关, 因为脑中 PMCA 包含 PMCA1~PMCA4, 而血中主要为 PMCA4 2 进一步研究了神经节苷脂对 PMCA 作用的机理 通过限制性酶切 钙调素竞争 短肽竞争等实验证明了神经节苷脂是通过与 PMCA 的 c 末端的直接相互作用, 从而使 PMCA 处于开放状态, 进而激活 PMCA 另外, 高浓度 GM2 抑制 PMCA 活性 机理研究表明,GM2 与 PMCA 的两个结构域相互作用 : 低浓度时作用于 PMCA 的 c 末端, 从而激活 PMCA; 而高浓度时, 作用于 PMCA 的自抑制位点, 进而抑制 PMCA 的活性 本年度获奖情况 66

67 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES Molecular Basis of Infection and Immunity Name Yan Xiyun Position Professor Research Topic Identification of novel targets using an antibody-based approach Summary of Research 1. Identification of novel tumor targets and their specific antibodies. After identifying CD146 as a novel target in tumor angiogenesis, we investigated new aspects of CD146 function: (1) CD146 plays a critical role in trophoblast invasion; (2) CD146 expression is reduced in the pregnancy disorder preeclampsia; (3) the function of CD146 in cell signal transduction and neuron development; (4) identification of CD146 ligands; (5) evaluation of soluble CD146 as a marker for diagnosis of diseases. 2. SARS epitope library and SARS antibody library: (1) Image of three-dimensional appearance of SARS-CoV virus particles using SEM; (2) SARS peptide and SARS antibody libraries were constructed, from which several SARS antibodies were selected and evaluated for anti- SARS therapy. 5. Imaging of single molecule behavior in living cells, especially visualization of the dynamic behaviour of the membrane protein CD146 and its ligands in living cells by means of nanotechnology. Selected Publications 1. Lin, Y., Yan, X.Y.*, Cao, W.H., Wang, C.Y., Xie, S.S. & Feng, J. Probing the Structure of the SARS Coronavirus Using Scanning Electron Microscopy. Antiviral Therapy 9: , (2004). 2. Liu, Q., Yan, X.Y.*, Li, Y., Zhang, Y., Zhao, X. & Shen, Y. Preeclampsia is Associated with the Failure of Melanoma Cell Adhesion Molecule (MCAM/CD146) Expression by Intermediate Trophoblast. Lab Invest84: (2004) 3. Lin,Y.,Yan, X.Y.* Progression and Direction of Humanized Antibody Research. Chinese Journal of Biotechnology 20: 1-4, (2004) 4. Qin Liu, Xingang Zhao, Ying Zhang, Yi Shen, Yixun Liu, Xiyun Yan *. Melanoma cell adhesion molecule (MCAM/CD146) is a critical molecule in trophoblast invasion. Prog. Biochem. Biophys. 31(4): , (2004) 5. HAN Wei, ZHANG Pan-he, CAO Wu-chun, YANG Dong-ling, Yoshio Okamoto, Shigeharu Taira, YAN Xi-Yun*. The inactivation effect of photocatalytic titanium apatite filter on SARS Virus. Prog. Biochem. Biophys. 31(11): , (2004) 67

68 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 感染与免疫的分子基础 姓名 职称 研究方向 阎锡蕴院士研究员副研究员百人计划国家杰出青年 疾病相关新型靶分子及其抗体的功能研究 本年度工作简介 1. 肿瘤靶分子及其抗体的新功能研究 继 2003 年发现肿瘤血管新靶标 CD146 之后, 我们深入研究了 CD146 的新功能, 发现 (1) 在胚胎植入过程中 CD146 分子是影响滋养层细胞侵入行为的关键分子 ;(2) 妊娠疾病先兆子痫胎盘 CD146 分子的表达明显下调 ;(3)CD146 分子参与细胞信号传导 ;(4) 寻找 CD146 配体 ;(5) 研究 CD146 分子在神经免疫系统中的分布及功能 ;(6) 可溶性 CD146 分子在疾病诊断中的意义 另外, 鉴定了一个新的肿瘤靶分子 T2-2 的分布及体内外功能 ; 完成了抗人肿瘤相关抗原 HerB2 的免疫毒素 ScFv-Trail 及 ScFv-TNF-α 的构建表达和功能研究 2.SARS 病毒抗原表位库和抗体库的研究 利用超高分辨率扫描电镜报道了 SARS 病毒的三维立体结构 ; 建立了 SARS 冠状病毒 BJ01 株抗原表位库和 SARS 抗体库, 获得 SARS 冠状病毒特异抗体并完成抗体的高效表达 纯化及中和活性实验 发现新型光触媒钛羟基磷灰石网膜 PTAF 对 SARS 病毒生长具有明显的抑制作用 3. 纳米材料与技术在细胞生物学研究的初探 2004 获得国家自然科学 纳米科技 基础研究领域重点项目支持, 利用纳米技术探测 CD146 分子在活细胞信号传递过程中的单分子行为 ; 并探讨纳米材料对基因扩增和基因转化的影响 专利 :2004 年申请发明专利 1 项 ; 获得专利授权 1 项 本年度获奖情况 2004 年培养的博士生荣获以下奖励 : 1)2004 年中国科学院院长优秀奖 2)2004 年宝洁奖学金 68

69 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 Name Position Research Topic Summary of Research Zhihai Qin Professor Interferon γ tumor stroma and tumor rejection As a new lab founded in the middle of 2004, we have now 3 PhD students, 3 visiting scientists, 1 research assistant and 1 technician. Three laboratories have been established for cell culture, molecular cloning and basic immunological experiments. The interest of our group is focused on the role of IFN-γ on tumor stroma cells during an immune response mediated tumor rejection. Research has begun on the following aspects: 1)Generation of conditional knockout mice (Jing Jiang, Yu Lu and Wei Yang). Mice with an IFN-γ receptor deficiency specifically on different tumor stoma cells, such as fibroblasts or endothelial cells will be generated in collaboration with the Experimental Animal Institute of the Chinese Academy of Medical Sciences. 2)Establishment of a three-dimensional co-culture system of mouse fibroblasts and tumor cells (Shuibai Liu and Chunhai Zhou). To analyze the interaction between tumor cells and fibroblasts in vitro, we are going to establish a 3D co-culture system. A series of mouse embryonic fibroblast and tumor infiltrating fibroblast cell lines from IFN-γR knockout mice have been established. We will reconstitute the expression of IFN-γR on these cells and investigate tumor/fibroblast interaction in the presence or absence of IFN-γ. 3)Improvement of in vivo cytokine detection assays (Bin Li). 4)Investigation of the role of inflammatory cytokines, such as IFN-γ and TNF, during the process of chemical carcinogenesis (Zhiguang Li and Xiangyue Zhang). Selected Publications 1. Qin, Z., and T. Blankenstein. A cancer immunosurveillance controversy. Nat Immunol 5:3, (2004). 2. Kim, H. J., T. Kammertoens, M. Janke, O. Schmetzer, Z. Qin, C. Berek, and T. Blankenstein. Establishment of early lymphoid organ infrastructure in transplanted tumors mediated by local production of lymphotoxin alpha and in the combined absence of functional B and T cells. J Immunol 172:4037, (2004) 3. Wu, T. H., C. N. Pabin, Z. Qin, T. Blankenstein, M. Philip, J. Dignam, K. Schreiber, and H. Schreiber. Long-Term Suppression of Tumor Growth by TNF Requires a Stat1- and IFN Regulatory Factor 1-Dependent IFN-gamma Pathway but Not IL-12 or IL-18. J Immunol 172:3243, (2004) 69

70 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 姓名 职称 秦志海 院士研究员副研究员长江学者百人计划国家杰出青年 研究方向 本年度工作简介 IFN-γ, 肿瘤间质细胞与肿瘤的免疫排斥 作为新建实验室, 本年度我们的工作重点是科研队伍的组建, 实验室建设和一些前期实验准备工作 本课题组现有博士研究生两名 硕转博士生一名 进修生三名 助研和实验员各一名 经过半年多时间的努力, 我们现已建立起细胞培养室 分子生物学和免疫学三个实验室, 并开始正常运行 围绕 IFN-γ, 肿瘤间质细胞与肿瘤免疫排斥的关系这一主攻方向, 我们从以下几个方面开展了工作 : 1) 构建肿瘤间质细胞特异性 IFN-γ 受体基因敲除小鼠 ( 蒋静 陆宇 杨薇 ) 为分析一些与肿瘤免疫排斥相关的效应因子, 如 IFN-γ 等对成纤维细胞和血管内皮细胞的作用, 我们拟构建转基因小鼠 现已完成了部分 IFN-γ 受体基因表达质粒的构建, 将与中国医学科学院动物所合作, 进一步构建转基因小鼠 2 建立体外肿瘤细胞与间质细胞的 3D 共培养体系, 研究其相互作用 ( 刘树柏 周春海 ) 我们已从 IFN-γR 基因敲除小鼠体内分离 建成多种胚胎成纤维细胞系列和浸润肿瘤组织的成纤维细胞系列 同时, 我们正在重建这些细胞对 IFN-γR 的表达 为近一步分析时间与剂量等因素对 IFN-γ 作用的影响, 我们利用 Tet-on/off 系统对该基因在培养体系内的表达进行了调控 3) 建立高敏感度的体内细胞因子检测方法 ( 李冰 ) 为分析免疫反应过程中, IFN-γ IL4 及 TNF 等细胞因子在体内的变化, 我们正在建立高敏感度的细胞因子检测方法 4) 分析细胞因子在肿瘤免疫监视过程中的作用机制 ( 李志广 张香月 ) 现已引进数种基因敲除小鼠, 如 :IFN-γR TNF-R1 TNF-R2 和 TNF-R1/R2 双基因敲除小鼠等, 拟分析 IFN-γ 和 TNF 在化学物如甲基胆蒽和二甲苯蒽等在小鼠体内致癌过程中的作用及免疫反应机制 本年度获奖情况 70

71 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 Name Haiying Hang Position Professor Research Topic Cell Cycle Checkpoint Protein Complex Summary of Research DNA damage arises during normal cellular metabolic processes or when cells are exposed to genotoxic agents. Cells respond to DNA damage by activating an intricate network of cell cycle checkpoint proteins and DNA repair factors. Mutations in these genes often lead to cancer development. One of the major tasks in this area is to identify key proteins in this network and discover their roles in cell cycle checkpoint control, DNA repair and tumor-prevention. Our current work focuses on a newly identified cell cycle checkpoint protein complex Rad9-Rad1-Hus1 (9-1-1 complex). We have successfully created rad1 and rad9 knockout cells and mice. Our studies with these cells have shown that rad1 and rad9 play essential roles in ensuring normal replication as well as S/M cell cycle checkpoint controls. We are also performing experiments to find out if deletion of rad1 or rad9 causes tumors in mice. Selected Publications 71

72 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 姓名 职称 杭海英 院士研究员副研究员长江学者百人计划国家杰出青年 研究方向 细胞周期检查点蛋白质的结构的结构与功能 本年度工作简介在正常的细胞代谢活动中或当细胞暴露于遗传毒性因子时, 会发生基因组 DNA 的损伤 细胞依赖由细胞周期调控蛋白和 DNA 修复蛋白组成的复杂系统对 DNA 的损伤进行控制和修复 当这些基因发生突变时, 常常会导致癌症的发生 Hus1 Rad1 和 Rad9 是细胞周期调控家族中的新成员, 是 DNA 损伤修复系统中必不可少的组成成分 据认为这三种蛋白通过形成三分子复合体 (9-1-1 复合体 ), 帮助细胞对抗遗传毒性压力, 包括对细胞周期进行调控和对 DNA 的损伤进行修复等 但是, 目前还不完全清楚 复合体在细胞周期调控和 DNA 修复中的作用机制, 更不了解它是否有抑制肿瘤发生的功能 我们建立了小鼠基因敲除细胞 (rad1 -/- 和 rad9 -/- ), 并发现 rad1 和 rad9 均是调控 DNA 正常复制和 S/M 细胞周期检查点所必需的 我们还建立了 rad1 和 rad9 基因敲除小鼠, 并正在利用这些小鼠检验 rad1 和 rad9 是否为抑癌基因 本年度获奖情况 72

73 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 Name Position Research Topic Summary of Research Jie Tang Professor Molecular Immunology 1. Functional studies of lymphocyte cell surface proteins CD72 is a regulatory receptor on B cells that suppresses B cell activation. Its function is closely related to a tyrosine phosphotase, SHP-1, which is associated with CD72 in nonactivated B cells. Upon CD72 ligand engagement, SHP-1 dissociates from CD72 so that its suppression of B cell receptor signaling is released. The function of CD72 can be regulated on the transcriptional level since multiple splicing variants of CD72 mrna exist in mouse B cells. The relationship between CD72 splicing variants and B cell function is our primary interest. We have cloned five novel CD72 splicing variants and expressed them in mammalian cells. Their association with SHP-1 and their functions in B cell activation will be studied in COS and WEHI cell lines. 2. Antibody therapy of immune system related disease Sepsis is one of the major immune system related diseases. To suppress the abnormal immune response is our main approach to treat this disease. We will develop monoclonal antibodies against HMGB-1, which is a cytokine that functions at a late stage in the immune response to infections. We have expressed HMGB-1 in E.coli and established a functional assay using purified recombinant protein. Animal immunization is underway. With the anti-hmgb-1 antibody, we will study the role of HMGB-1 in a murine sepsis model and validate this protein as a target for human sepsis. Selected Publications 73

74 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 姓名 职称 研究方向 唐捷院士研究员副研究员长江学者百人计划国家杰出青年 分子免疫学 本年度工作简介 ( 一 )B 细胞特异性受体信号转导调控的研究 CD72 是 B 细胞特异性受体, 在 B 细胞成熟与活化的各个阶段都有表达 在 CD72 的胞内区上有两个免疫受体酪氨酸抑制性基序 (ITIM) 每个 ITIM 内的酪氨酸被磷酸化后可与蛋白酪氨酸磷酸酯酶 SHP-1 的一个 SH2 结合性结构域相结合 在 B 细胞活化前,CD72 上的酪氨酸被磷酸化,SHP-1 与 CD72 结合, 控制 BCR 的酪氨酸磷酸化水平, 进而抑制 BCR 介导的信号转导 当 CD72 与其配体结合后, CD72 上的酪氨酸去磷酸化,SHP-1 与 CD72 分离, 从而解除了对 BCR 信号转导的抑制 我们的工作是在 pre-mrna 剪切水平上研究对 CD72 功能的调控 我们发现 CD72 在不同细胞和细胞活化的不同阶段剪切类型有变化, 有些剪切类型会导致 CD72 上的第二个 ITIM 缺失, 或者两个 ITIM 之间的距离增大, 还有一些类型由于缺失跨膜区而不能在膜上表达 这些类型对 CD72 功能的影响是我们进一步研究的课题 ( 二 ) 以 HMGB1 为靶点开发治疗败血症的单克隆抗体药物 在败血症的小鼠模型中,HMGB1 发挥后期调控因子的作用 在注射过 LPS 后 24 小时内加入抗 HMGB1 的多克隆抗体, 就能阻断 LPS 伤害作用, 明显增强生还比率 相比之下,HMGB1 较其它细胞因子在效应时间上明显靠后, 因此它给我们拓宽了治疗时间的范围, 相当于给我们开了一个全新的 高效的治疗窗口 拮抗 HMGB1 的药物开发, 将是利用这个新窗口的关键 但人源的 HMGB1 与鼠源的 HMGB1 有 98% 以上的同源性, 导致鼠对人的 HMGB1 具有免疫耐受性, 从而无法从正常小鼠或大鼠中得到单克隆抗体 由于鸡在进化中和哺乳动物差别比较大, 在其体内产生抗人源 HMGB1 抗体的几率要大得多, 已有文献证明确实能够免疫鸡得到具有中和活性的抗 HMGB1 抗体 我们掌握的鸡源性单克隆抗体筛选技术, 主要是利用分子生物学方法从免疫后的鸡脾中构建抗体文库, 并用酵母展示方法筛选特异性抗体 目前我们已完成了抗原的表达与纯化工作, 建立了 HMGB1 活性鉴定体系, 并开始对鸡进行免疫 酵母展示系统在本实验室也已建立起来, 在许多方面还有了进一步改进 本年度获奖情况 74

75 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 Name Fan Zusen Position Professor Research Topic Molecular Mechanisms of Tumorigenesis, Killing mechanisms and Cancer Immunotherapy Summary of Research I. Molecular mechanisms of tumorigenesis Cell cycle control and tumorigenesis are at the frontier of cancer research. pp32 was recently confirmed as a new tumor suppressor. We identified eight pp32-associated proteins using the yeast two-hybrid system and confirmed their real interactions with histone H3, TRAP1 and RYBP. pp32 inhibits cell growth through binding to histone H3 and RYBP. The results will be prepared for publication. The interaction between pp32 and TRAP1 is under investigation. II. CTL-mediated killing mechanism against cancers CTL and NK cells are important effector cells in immune responses against viruses, intracellular bacteria and tumors. They kill their target cells through granule contents, including perforin and granzymes. Human granzymes A, B, K, M and H and their inactive forms were expressed and purified. We are studying their molecular pathways to induce apoptosis and compare with known pathways of granzyme A and B, which will provide natural drugs for cancer immunotherapy. III. Cancer immunotherapy NK cells are innate effector cells that play an important role in the defense against virally infected and transformed cells. We demonstrated that the TNF superfamily member LIGHT is a critical ligand for activation of NK cells. HVEM is expressed on NK cells and its engagement with LIGHT mediates NK cell activation. The activated NK cells can trigger activation and maturation of tumor specific CD8 + T cells at its priming phase in vivo and in vitro. NK cells can directly prime CTL responses to bridge innate and adaptive immunity for breaking the tolerance of CD8 + cells inside tumors, which may provide new guidance for cancer immunotherapy. This work is being reviewed for Nature Medicine. Selected Publications Zusen Fan, Ping Yu, Yang Wang, May lynne Fu, Youjin Lee, Yugang Wang, Wenhua Liu, Yonglian Sun, Yang-Xin Fu*. Natural killer activation by LIGHT primes tumor specific CD8 + T cell immunity to reject established tumors. Nature Medicine (under review). 75

76 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 姓名 职称 研究方向 范祖森院士研究员副研究员长江学者百人计划国家杰出青年 肿瘤发生的分子基础 杀伤机理及免疫治疗 本年度工作简介 1. 肿瘤发生的分子基础原癌基因和抑癌基因如何调节细胞生长分化及癌变机理, 一直成为细胞生物学及肿瘤学研究的热点 pp32 被证实为新抑癌基因, 但 pp32 如何调节细胞生长及发挥抑癌作用的机制尚未阐明 我们发现 pp32 与 SET 等蛋白形成 SET 复合体, 参与颗粒酶 A 介导的细胞凋亡 以 pp32 为诱饵蛋白, 采用酵母双杂交系统筛选出 8 个与 pp32 相作用的蛋白, 其中组蛋白 H3, RYBP 和 TRAP1 与 pp32 的相互作用已被证实 并阐明了 pp32 与 H3 及 RYBP 结合, 抑制了基因转录, 进而阻止了细胞生长 部分结果将于近期内整理发表 正进一步研究 pp32 如何通过其作用蛋白 TRAP1 参与 TNFR 信号及 prb 调节途径调控细胞生长分化, 并从结构与功能的关系上, 以期阐明 pp32 抑制细胞生长的机制 2. 肿瘤的杀伤机理 CTL 和 NK 细胞诱导的杀伤作用是机体抗病毒感染和抗肿瘤的主要效应途径 介导细胞杀伤的颗粒内容物为颗粒酶和导致膜损伤的穿孔素与颗粒素, 各种颗粒酶如何协同诱导靶细胞凋亡以及穿孔素和颗粒素如何辅助颗粒酶进入细胞并如何发挥其特有的杀伤作用, 均尚未搞清 我们已表达纯化了穿孔素和颗粒素及各种颗粒酶 A B K M 和 H 及其突变体 正在探讨颗粒酶 K M 和 H 在 CTL 介导肿瘤杀伤中的作用, 并鉴定其作用底物, 分析其与颗粒酶 A B 杀伤途径的异同, 阐明其杀伤机制, 为抗肿瘤及抗病毒药物的研制提供理论基础和新思路 3. 肿瘤的免疫治疗 NK 细胞作为先天免疫的效应细胞在抗病毒及抗肿瘤中发挥重要作用 NK 细胞如何被活化和激发过继免疫反应尚不清楚 我们发现 TNF 超家族成员 LIGHT 是 NK 活化的重要配体,NK 表面表达 HVEM 受体并与其配体 LIGHT 结合介导 NK 细胞活化, 体外及动物实验研究表明激活的 NK 细胞可以通过 IFN-γ 直接活化肿瘤特异性 CTL 介导有效的肿瘤杀伤, 从而证明 NK 细胞是沟通先天免疫与过继免疫的桥梁 该研究论文正在 Nature Medicine 杂志审稿 在此基础上将深入探讨如何在肿瘤局部高效地表达 LIGHT 激发 NK 细胞的活性, 为临床有效地介导肿瘤的免疫治疗提供新的免疫药物和理论基础 本年度获奖情况 76

77 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 Ⅵ. 论文索引 List of Publications 1 LIU ZF, YAN HC, WANG KB et al. and CHANG WR*. Crystal structure of spinach major light-harvesting complex at 2.72 angstrom resolution. NATURE, 2004, 428: WU BL, LI PY, LIU YW et al. and RAO ZH*. 3D structure of human FK506-binding protein 52: Implications for the assembly of the glucocorticoid receptor/hsp90/immunophilin heterocomplex. P NATL ACAD SCI USA, 2004, 101: DUAN XJ, ZHAO Z, YE JP et al. and MA HM*, XIA AD*, WANG CC*. Donor Donor Energy-Migration Measurements of Dimeric DsbC Labeled at Its N-Terminal Amines with Fluorescent Probes: A Study of Protein Unfolding. ANGEW CHEM INT EDIT, 2004, 43: PENG H, BARTLAM MARK, ZENG QH et al. and RAO ZH*. Crystal structure of human pirin - An iron-binding nuclear protein and transcription cofactor. J BIOL CHEM, 2004, 279: SHI N, YE S, BARTLAM M et al. and RAO ZH*, YUAN JG*. Structural basis for the specific recognition of RET by the Dok1 phosphotyrosine binding domain. J BIOL CHEM, 2004, 279: JIANG Y, LI H, ZHU L et al. and ZHOU JM*, PERRETT S*. Amyloid nucleation and hierarchical assembly of Ure2p fibrils - Role of asparagine/glutamine repeat and nonrepeat regions of the prion domain. J BIOL CHEM, 2004, 279: WEI ZY, ZHANG P, ZHOU ZC et al. and GONG WM*. Crystal structure of human eif3k, the first structure of eif3 subunits. J BIOL CHEM, 2004, 279: LI MH, KWOK F, CHANG WR et al. and JIANG T*. Conformational changes in the reaction of pyridoxal kinase. J BIOL CHEM, 2004, 279: WANG YL, WEI ZY, BIAN Q et al. and GONG WM*. Crystal structure of human bisphosphoglycerate mutase. J BIOL CHEM, 2004, 279: XU YH, LIU YW, LOU ZY et al. and RAO ZH*. Structural basis for coronavirus-mediated membrane fusion: Crystal structure of MHV spike protein fusion core. J BIOL CHEM, 2004, 279: XU YH, LOU ZY, LIU YW et al. and RAO ZH*. Crystal structure of SARS-CoV spike protein fusion core. J BIOL CHEM, 2004, 279: BAI M, ZHOU JM*, PERRETT S* et al. The yeast prion protein Ure2 shows glutathione peroxidase activity in both native and fibrillar forms. J BIOL CHEM, 2004, 279: BARTIAM M, WANG GG, YANG HT et al. and RAO ZH*. Crystal Structure of an Acylpeptide Hydrolase/Esterase from Aeropyrum pernix K1. STRUCTURE, 2004, 77

78 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES : LIN Y, YAN XY*, CAO WC et al. Probing the structure of the SARS coronavirus using scanning electron microscopy. ANTIVIR THER, 2004, 9: ZHOU ZC, SONG XM, LI YK et al. and GONG WM*. Unique structural characteristics of peptide deformylase from pathogenic bacterium Leptospira interrogans. J MOL BIOL, 2004, 339: LIU L, WEI ZY, WANG YL et al. and GONG WM*. Crystal structure of human coactosin-like Protein. J MOL BIOL, 2004, 344: WAN QF, DONG YM, YANG H et al. and XU T*. Protein Kinase Activation Increases Insulin Secretion by Sensitizing the Secretory Machinery to Ca2+. J GEN PHYSIOL, 2004, 124: LIU BB, BARTLAM MARK, GAO RJ et al. and RAO ZH*. Crystal structure of the hyperthermophilic inorganic pyrophosphatase from the Archaeon Pyrococcus horikoshii. BIOPHYS J, 2004, 86: LI DD, XIONG J, QU AL et al. and XU T*. Three-dimensional tracking of single secretory granules in live PC12 cells. BIOPHYS J, 2004, 87: CAI CX, CHEN JW*. Overexpression of caveolin-1 induces alteration of multidrug resistance in Hs578T breast adenocarcinoma cells. INT J CANCER, 2004, 111: LIU JH, WANG ZX*. Kinetic analysis of ligand-induced autocatalytic reactions.. BIOCHEM J, 2004, 379: SU X, QING SB, PAN XM*. Thermal and conformational stability of Ssh10b protein from archaeon Sulfolobus shibattae. BIOCHEM J, 2004, 382: STEPANENKO OV, KUZNETSOVA IM, TUROVEROV KK* et al. andwang CC*. Conformational change of the dimeric DsbC molecule induced by GdnHCl. A study by intrinsic fluorescence. BIOCHEMISTRY-US, 2004, 43: XU YH, ZHU JQ, LIU YW et al. and RAO ZH*, GAO GF*. Characterization of the heptad repeat regions, HR1 and HR2, and design of a fusion core structure model of the spike protein from severe acute respiratory syndrome (SARS) coronavirus. BIOCHEMISTRY-US, 2004, 43: LI XM, LIU XQ, LOU ZY et al. and RAO ZH*, LIU YW*. Crystal Structure of Human Coactosin-like Protein at 1.9 A Resolution. PROTEIN SCI, 2004, 13: ZHU L, QIN ZJ, ZHOU JM*. Unfolding kinetics of dimeric creatine kinase measured by stopped-flow small angle X-ray scattering. BIOCHIMIE, 2004, 86:

79 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES FENG YM, HUANG S, ZHANG WZ et al. and JING GZ*. The effects of amino acid replacements of glycine 20 on conformational stability and catalysis of staphylococcal nuclease. BIOCHIMIE, 2004, 86: ZHU H, CAI CX, CHEN JW* et al. Suppression of P-glycoprotein gene expression in Hs578T/Dox by the overexpression of caveolin-1. FEBS LETT, 2004, 576: LIU JF, WANG XQ, WANG ZX et al. and LIANG DC*. Crystal sturcture of KD93, a novel protein expressed in human hematopoietic stem/progenitor cells. J STRUCT BIOL, 2004, 148: ZHOU L, ZHANG XJ*. Electron-microscopic structure of th V-ATPase from mung bean. PLANTA, 2004, 219: PANG H, LIU YG, HAN XQ et al. and RAO ZH*. Protective humoral responses to severe acute respiratory syndrome-associated coronavirus: implications for the design of an effective protein-based vaccine. J GEN VIROL, 2004, 85: WANG WN, PAN XM, WANG ZX*. Kinetic analysis of zymogen autoactiviation in the presence of a reversible inhibitor. EUR J BIOCHEM, 2004, 271: CAI CX, ZHU H, CHEN JW*. Overexpression of caveolin-1 increases plasma membrane fluidity and reduces P-glycoprotein function in Hs578T/Dox. BIOCHEM BIOPH RES CO, 2004, 320: LIU CP, ZHOU JM*. Trigger factor-assisted folding of bovine carbonic anhydrase II. BIOCHEM BIOPH RES CO, 2004, 313: LI HT, WANG C, CHANG TN et al. and CHANG WR*. ph-profile crystal structure studies of C-terminal despentapeptide nitrite reductase from Achromobacter cycloclastes. BIOCHEM BIOPH RES CO, 2004, 316: LIU DS, FENG YG, CHENG Y et al. and WANG JF*. Human programmed cell death 5 protein has a helical-core and two dissociated structural regions. BIOCHEM BIOPH RES CO, 2004, 318: BERNINI A, SPIGA O, CIUTTI A et al. and NICCOLAI N*. Prediction of quaternary assembly of SARS coronavirus peplomer. BIOCHEM BIOPH RES CO, 2004, 325: WANG C, WANG F, LI M et al. and CHANG WR*. Structural basis for broad substrate specificity of earthworm fibrinolytic enzyme component A. BIOCHEM BIOPH RES CO, 2004, 325: DAI JX, WANG X, FENG YG et al. and WANG JF*. Searching for folding initiation sites of staphylococcal nuclease: A study of N-terminal short fragments. BIOPOLYMERS, 2004, 75:

80 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES HUANG ZH, CHANG XB, RIORDAN JR et al. and HUANG YG*. Fluorescent modified phosphatidylcholine floppase activity of reconstituted multidrug resistance-associated protein MRP1. BBA-BIOMEMBRANES, 2004, 1660: XIA S, XU L, BAI L et al. and XU T*. Labeling and dynamic imaging of synaptic vesicle-like microvesicles in PC12 cells using TIRFM. BRAIN RES, 2004, 997: ZHAO YF, FAN XX, YANG FY et al. and YANG FY, ZHANG XJ*. Gangliosides modulate the activity of the plasma membrane Ca2+-ATPase from porcine brain synaptosomes. ARCH BIOCHEM BIOPHYS, 2004, 427: FENG J, WANG Q, WU YS et al. and ZHANG JP*. Triplet excitation transfer between carotenoids in the LH2 complex from photosynthetic bacterium Rhodopseudomonas palustris. PHOTOSYNTH RES, 2004, 82: ZHOU ZC, GONG WM*. Co-crystallization of Leptospira interrogans peptide deformylase with a potent inhibitor and moleculara-replacement schemes with eight subunits in an asymmetric unit. ACTA CRYSTALLOGR D, 2004, 60: CHANG SJ, SONG XM, YAN M et al. and GONG WM*. Purification, characterization and preliminary crystallographic studies of a cysteine protease from Pachyrrhizus erosus seeds. ACTA CRYSTALLOGR D, 2004, 60: WANG F, WANG C, LI M et al. and CHANG WR*. Crystallization and preliminary crystallographic analysis of earthworm fibrinolytic enzyme component B from Eisenia fetida. ACTA CRYSTALLOGR D, 2004, 60: YUAN CH, BAI J, SUN YD et al. and LIANG DC*. Structure of potato calmodulin PCM6: the first report of the three-dimensional structure of a plant calmodulin. ACTA CRYSTALLOGR D, 2004, 60: LIU L, WANG YL, ZHANG P et al. and GONG WM*. Expression, purification and preliminary crystallographic studies of human coactosin-like protein. ACTA CRYSTALLOGR D, 2004, 60: SUN L, DONG YJ, ZHOU YF et al. and ZHANG XE*. Crystallization and preliminary X-ray studies of methyl parathion hydrolase from Pseudomonas sp. WBC-3. ACTA CRYSTALLOGR D, 2004, 60: HU HY, WANG GG, YANG HT et al. and RAO ZH*, JIN C*. Crystallization and preliminary crystallographic analysis of a native chitinase from the fungal pathogen Aspergillus fumigatus YJ-407. ACTA CRYSTALLOGR D, 2004, 60: LIU BB, LI XM, GAO RJ et al. and RAO ZH*. Crystallization and preliminary X-ray analysis of inorganic pyrophosphatase from the hyperthermophilic archaeon Pyrococcus horikoshii OT3. ACTA CRYSTALLOGR D, 2004, 60: SHI N, LIU YW, NI MH et al. and RAO ZH*. Expression, crystallization and preliminary X-ray studies of the recombinant PTB domain of mouse dok1 protein. 80

81 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 ACTA CRYSTALLOGR D, 2004, 60: XU YH, SU N, QING L et al. and RAO ZH*. Crystallization and preliminary crystallographic analysis of heptad repeat complex of SARS coronavirus Spike protein. ACTA CRYSTALLOGR D, 2004, 60: XU YH, BAI ZH, QIN L et al. and RAO ZH*. Crystallization and Preliminary Crystallographic analysis of fusion Core of spike protein of the murine coronavirus mouse hepatitis virus (MHV). ACTA CRYSTALLOGR D, 2004, 60: WANG YL, CHENG ZJ, LIU L et al. and GONG WM*. Cloning, purification, crystallization and preliminary crystallographic analysis of human phosphoglycerate mutase. ACTA CRYSTALLOGR D, 2004, D60: SONG XM, ZHOU ZC, WANG J et al. and GONG WM*. Purification, characterization and preliminary crystallographic studies of a novel plant defensin from Pachyrrhizus erosus seeds. ACTA CRYSTALLOGR D, 2004, D60: LI X, LIU X, ZHAO Y et al. and RAO ZH*. Crystallization and preliminary crystallographic studies of human coactosin-like protein (CLP). ACTA CRYSTALLOGR D, 2004, 60: HAN XQ, BARTLAM M, JIN YH et al. and RAO ZH*. The expression of SARS-CoV M gene in P. Pastoris and the diagnostic utility of the expression product. J VIROL METHODS, 2004, 122: WANG XP, HAN XH, YANG FY*. Critical segment of apocytochrome c for its insertion into membrane. MOL CELL BIOCHEM, 2004, 262: PENG YH, CHEN JW*. Anisodamine causes th changes of structure and function in the transmembrane domain of th Ca2+-ATPase from sacroplasmic reticulum. BIOSCI BIOTECH BIOCH, 2004, 68: HUANG ZH, HUANG YG*. The relationship between MRP1 activities and its NBD conformational changes. SCI CHINA SER C, 2004, 47: ZHOU JM*. Prion Diseases and The "Protein only" Hypothesis. PROG BIOCHEM BIOPHYS, 2004, 31: LIU Q, ZHAO XG, ZHANG Y et al. and YAN XY*. Melanoma cell adhesion molecule (MCAM/CD146) is a critical molecule in trophoblast invasion. PROG BIOCHEM BIOPHYS, 2004, 31:

82 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 Ⅶ. 代表性论文选编 ( 影响因子大于 4) Selected Publications 1 LIU ZF, YAN HC, WANG KB et al. and CHANG WR*. Crystal structure of spinach major light-harvesting complex at 2.72 angstrom resolution. NATURE, 2004, 428: WU BL, LI PY, LIU YW et al. and RAO ZH*. 3D structure of human FK506-binding protein 52: Implications for the assembly of the glucocorticoid receptor/hsp90/immunophilin heterocomplex. P NATL ACAD SCI USA, 2004, 101: DUAN XJ, ZHAO Z, YE JP et al. and MA HM*, XIA AD*, WANG CC*. Donor Donor Energy-Migration Measurements of Dimeric DsbC Labeled at Its N-Terminal Amines with Fluorescent Probes: A Study of Protein Unfolding. ANGEW CHEM INT EDIT, 2004, 43: PENG H, BARTLAM MARK, ZENG QH et al. and RAO ZH*. Crystal structure of human pirin - An iron-binding nuclear protein and transcription cofactor. J BIOL CHEM, 2004, 279: SHI N, YE S, BARTLAM M et al. and RAO ZH*, YUAN JG*. Structural basis for the specific recognition of RET by the Dok1 phosphotyrosine binding domain. J BIOL CHEM, 2004, 279: JIANG Y, LI H, ZHU L et al. and ZHOU JM*, PERRETT S*. Amyloid nucleation and hierarchical assembly of Ure2p fibrils - Role of asparagine/glutamine repeat and nonrepeat regions of the prion domain. J BIOL CHEM, 2004, 279: WEI ZY, ZHANG P, ZHOU ZC et al. and GONG WM*. Crystal structure of human eif3k, the first structure of eif3 subunits. J BIOL CHEM, 2004, 279: LI MH, KWOK F, CHANG WR et al. and JIANG T*. Conformational changes in the reaction of pyridoxal kinase. J BIOL CHEM, 2004, 279: WANG YL, WEI ZY, BIAN Q et al. and GONG WM*. Crystal structure of human bisphosphoglycerate mutase. J BIOL CHEM, 2004, 279:

83 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES XU YH, LIU YW, LOU ZY et al. and RAO ZH*. Structural basis for coronavirus-mediated membrane fusion: Crystal structure of MHV spike protein fusion core. J BIOL CHEM, 2004, 279: XU YH, LOU ZY, LIU YW et al. and RAO ZH*. Crystal structure of SARS-CoV spike protein fusion core. J BIOL CHEM, 2004, 279: BAI M, ZHOU JM*, PERRETT S* et al. The yeast prion protein Ure2 shows glutathione peroxidase activity in both native and fibrillar forms. J BIOL CHEM, 2004, 279: BARTIAM M, WANG GG, YANG HT et al. and RAO ZH*. Crystal Structure of an Acylpeptide Hydrolase/Esterase from Aeropyrum pernix K1. STRUCTURE, 2004, 12: LIN Y, YAN XY*, CAO WC et al. Probing the structure of the SARS coronavirus using scanning electron microscopy. ANTIVIR THER, 2004, 9: ZHOU ZC, SONG XM, LI YK et al. and GONG WM*. Unique structural characteristics of peptide deformylase from pathogenic bacterium Leptospira interrogans. J MOL BIOL, 2004, 339: LIU L, WEI ZY, WANG YL et al. and GONG WM*. Crystal structure of human coactosin-like Protein. J MOL BIOL, 2004, 344: WAN QF, DONG YM, YANG H et al. and XU T*. Protein Kinase Activation Increases Insulin Secretion by Sensitizing the Secretory Machinery to Ca2+. J GEN PHYSIOL, 2004, 124: LIU BB, BARTLAM MARK, GAO RJ et al. and RAO ZH*. Crystal structure of the hyperthermophilic inorganic pyrophosphatase from the Archaeon Pyrococcus horikoshii. BIOPHYS J, 2004, 86: LI DD, XIONG J, QU AL et al. and XU T*. Three-dimensional tracking of single secretory granules in live PC12 cells. BIOPHYS J, 2004, 87: CAI CX, CHEN JW*. Overexpression of caveolin-1 induces alteration of multidrug resistance in Hs578T breast adenocarcinoma cells. INT J CANCER, 2004, 111:

84 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES LIU JH, WANG ZX*. Kinetic analysis of ligand-induced autocatalytic reactions.. BIOCHEM J, 2004, 379: SU X, QING SB, PAN XM*. Thermal and conformational stability of Ssh10b protein from archaeon Sulfolobus shibattae. BIOCHEM J, 2004, 382:

85 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 Ⅷ. Future Prospects With the support of the Ministry of Science and Technology and the Chinese Academy of Sciences, the NLB will continuously carry out. Taking full advantage of current resources and focusing incremental resources, to develop a 1 st class platform for protein sciences, including systems for high-throughput protein expression, for antibody research and development, for functional analysis of proteins, for structural genomics, for proteomics, for innovations of key technologies. Built on the core of protein science, the NLB will actively pursue multidisciplinary study of protein 3-dimentional structure and function, structure and function of biological membrane and membrane proteins, function and folding principles of proteins, molecular basis of immunology and infectious diseases, molecular neurobiology, nano-biology, computational biology, and systems biology, protein and multipeptide drugs, to generate a large number of breakthroughs to advance the fundamental understanding of basic principles of science and to meet the strategic need of the Nation. 85

86 ANNUAL REPORT NATIONAL LABORATORY OF BIOMACROMOLECULES 2004 Ⅷ. 今后发展方向 在科技部和中科院的支持下, 生物大分子国家重点实验室将一如既往地坚持 流动 开放 联合 竞争 的办室方针 ; 坚持定期评估 优胜劣汰 强强联合的管理机制 充分利用现有资源, 集中投入增量经费, 分阶段 有重点地建设包括高通量蛋白质表达与抗体研发系统 蛋白质功能分析研究系统 结构基因组学研究系统 蛋白质组学研究系统 关键技术自主创新系统的国际一流的蛋白质科学研究平台, 并在此基础上以蛋白质科学为核心, 发挥多学科交叉综合的优势, 积极筹备以蛋白质三维结构与功能研究 生物膜和膜蛋白功能与结构研究 蛋白质功能与折叠原理研究 感染与免疫的分子基础 分子神经生物学 纳米生物学与微纳仿生 计算生物学和系统生物学 蛋白质药物与多肽药物等领域为主要研究方向的蛋白质科学国家实验室 并取得一批重大原创性成果, 和一批与国家战略需求相关的重大成果 86

87 Crystal structure of spinach major lightharvesting complex at 2.72 Å resolution Zhenfeng Liu 1, Hanchi Yan 1, Kebin Wang 2, Tingyun Kuang 2, Jiping Zhang 1, Lulu Gui 1, Xiaomin An 1 & Wenrui Chang 1 articles 1 National Laboratory of Biomacromolecules, Institute of Biophysics, Chinese Academy of Sciences, 15 Datun Road, Chaoyang District, Beijing , People s Republic of China 2 Laboratory of Photosynthesis and Environmental Molecular Physiology, Institute of Botany, Chinese Academy of Sciences, 20 Nanxincun, Xiangshan, Beijing , People s Republic of China... The major light-harvesting complex of photosystem II (LHC-II) serves as the principal solar energy collector in the photosynthesis of green plants and presumably also functions in photoprotection under high-light conditions. Here we report the first X-ray structure of LHC-II in icosahedral proteoliposome assembly at atomic detail. One asymmetric unit of a large R32 unit cell contains ten LHC-II monomers. The 14 chlorophylls (Chl) in each monomer can be unambiguously distinguished as eight Chla and six Chlb molecules. Assignment of the orientation of the transition dipole moment of each chlorophyll has been achieved. All Chlb are located around the interface between adjacent monomers, and together with Chla they are the basis for efficient light harvesting. Four carotenoid-binding sites per monomer have been observed. The xanthophyll-cycle carotenoid at the monomer monomer interface may be involved in the non-radiative dissipation of excessive energy, one of the photoprotective strategies that have evolved in plants. Light harvesting is the primary process in photosynthesis. In green plants, the function of harvesting solar energy is fulfilled by a series of light-harvesting complexes in the thylakoid membrane of chloroplasts. LHC-II, the most abundant integral membrane protein in chloroplasts, exists as a trimer and binds half of the thylakoid chlorophyll molecules. Every monomeric LHC-II comprises a polypeptide of about 232 amino-acid residues, Chla and Chlb molecules 1,3 4carotenoids 2 and one tightly bound phospholipid 3. Besides the light-harvesting function, LHC-II has also been shown to function in the non-radiative dissipation of excess excitation energy formed under high-light conditions 4,5. It has a crucial role in minimizing the damaging effects of excess light by operating this photoprotective mechanism as light intensity becomes increasingly saturating. Moreover, LHC-II also takes part in regulating the distribution of excitation energy to photosystems II and I (ref. 6). The structure of LHC-II from pea has been determined by electron crystallography at 3.4 Å resolution parallel to the membrane plane, and at about 4.9 Å resolution perpendicular to this plane 7. This model revealed some basic structural features of LHC-II, including three transmembrane a-helices (helices A, B and C) and a short amphipathic helix (helix D), 12 chlorophyll tetrapyrroles with roughly determined locations and orientations, and two carotenoids. A more detailed structural picture of LHC-II, with an unambiguous determination of the identity of the chlorophylls (Chla or Chlb) and the orientation of their transition dipole moments, would be beneficial for a better understanding of the basic functional mechanism of LHC-II. We have obtained this information by solving the structure of LHC-II at higher resolution using X-ray crystallography. In our model of the X-ray structure of LHC-II at 2.72 Å resolution, we provide the basis for investigating quantitatively the underlying mechanism of the light-harvesting process and its adjustment in LHC-II. We also reveal for the first time an elegant arrangement of membrane proteins in the icosahedral proteoliposome assembly, and show that membrane proteins can be crystallized in a way that differs from those described in ref. 8. Structure determination Crystallization of trimeric LHC-II isolated from Spinacia oleracea is described briefly in the Methods. Structure determination of LHC-II organized in an icosahedral particle includes initial phasing by the single isomorphous replacement (SIR) method plus phase refining and extending by the real-space averaging method 9. Data collection, phasing and refinement statistics are listed in Table 1. The high-quality electron density map enabled us to trace 94% of Table 1 Data collection, phasing and refinement statistics Data set Native (1) Native (2) K 2 HgI 4 (derivative)... Resolution (Å) R merge * (0.112) (0.368) (0.131) Completeness (%) 97.9 (95.5) 90.8 (79.6) 80.5 (83.2),I/j. 6.1 (2.1) 14.4 (2.5) 6.4 (3.3) SIR phasing statistics No. of heavy atom sites 10 R cullis (centric/acentric) 0.66/0.69 Phasing power (centric/acentric) 1.13/1.67 Resolution (Å) FOM 0.36 Phase refinement and extension statistics Resolution (Å) FOM (0.801) Correlation coefficient (0.765) R factor (0.468) Structure refinement statistics Resolution Reflections (working set) 179,170 Reflections (test set) 9,326 R work /R free (%) 19.4/22.1 r.m.s.d. bond length (Å) r.m.s.d. bond angle (8) Coordinate error (Å) Luzzati 0.28 SigmaA 0.28 Number of non-hydrogen atoms Protein 16,619 Cofactors 11,720 Water Numbers in parentheses correspond to values in the highest resolution shell. *R merge ¼ S j S h ji j,h, I h. j/s j S h, I h., where h are unique reflection indices, I j,h are intensities of symmetry-related reflections and,i h. is the mean intensity. Reflections with I. 2.0 j I in native (1) and derivative data sets were used in the R merge calculation, whereas the 23.0 j I cutoff was applied in the native (2) data set. FOM (figure of merit), R cullis (centric) and phasing power determined by programs from the CCP4 suite (see Methods). Correlation coefficient¼ S hð, F o.2jf oj h Þð, F c.2jf cj h Þ=½S hð, F o.2jf oj h Þ 2 S hð, F c. 2jF cj h Þ 2 Š 1=2 ; R factor ¼ S hjf o 2 F cj=s hf o, where h are the unique reflection indices, F o are the observed structure factors and F c are the structure factors calculated from inversion of the noncrystallographic symmetry-averaged map. From Luzzati plot and SigmaA analysis, as determined with CNS (see Methods). 87 NATURE VOL MARCH Nature Publishing Group

88 articles the 232 amino acids and accurately locate the 14 chlorophylls and 4 carotenoids within one monomeric LHC-II. For the 14 chlorophylls, assignment of the orientation of the Q x and Q y transition dipolar moments was accomplished by proper positioning of the chlorophyll head groups. Ten chlorophylls were modelled with complete phytyl chains, but phytyl chains for the remaining four chlorophylls could be only partially modelled. With the help of 2F o F c and F o F c electron-density maps (Fig. 1a, b), all 14 chlorophylls were unambiguously characterized as eight Chla and six Chlb. The resulting Chla/b molar ratio of 1.33 is consistent with the value determined by earlier biochemical analyses 1,2. Three carotenoids were identified as two luteins and one neoxanthin, and the fourth member was interpreted as a mixed density involving the xanthophyll-cycle carotenoids. In addition, two lipids, one detergent molecule and about 70 water molecules per monomer have been positioned. Figure 1c, d shows two regions of the electron-density map calculated with the phases at 2.72 Å resolution. Icosahedral proteoliposome and crystal packing In the T ¼ 1 icosahedral particle (Fig. 2a), 20 LHC-II trimers are organized in a closed 532 point group symmetry, with their central C 3 axis serving as the icosahedral C 3 axis and oriented radially towards the sphere centre. One C 3 axis and two C 2 axes of the icosahedron superpose with the crystallographic axes. These trimers form a spherical shell with an outer diameter of about 261 Å and an inner diameter of about 160 Å. They are oriented in the shell with their flat lumenal surface facing the interior of the sphere and the less flat stromal surface facing outwards, taking part in the contacts with other particles in the crystal. The interactions between two adjacent trimers are mediated mainly by two digalactosyl diacylglycerol (DGDG) molecules and two pairs of chlorophylls through van der Waals contacts. They are all located near the icosahedral C 2 axis. The digalactosyl head group of each DGDG is simultaneously hydrogen bonded to the lumenal-surface amino acids from two adjacent trimers, functioning as a bridge. The hydrophobic fattyacid chains of DGDG extend into the membrane interior, interacting with hydrophobic residues and pigments of LHC-II. Other lipids that are expected to fill the gaps between LHC-II trimers and to form a spherical lipid-bilayer vesicle are mostly disordered. In the crystal lattice (Fig. 2b), LHC-IIs are assembled and packed in a manner different from those in Type I and Type II threedimensional (3D) crystals of membrane proteins as originally proposed in ref. 8. The LHC DGDG proteoliposomes assume the shape of closed spheres, presumably originating from curved, small patches of two-dimensional (2D) membrane protein crystals. Both the outer and inner surfaces of each proteoliposome are hydrophilic. The contacts between two proteoliposomes in the crystal lattice are polar interactions provided by the hydrophilic stromal surfaces of LHC-IIs. The hydrophobic intramembranous surfaces of LHC-II trimers are sheltered from crystal packing by the hydrophobic chains of lipids. We categorize this novel kind of 3D crystal as a Type III membrane-protein crystal. The apoprotein and LHC-II trimer The polypeptide main chain of each monomeric LHC-II was continuously traced from Ser 14 to Gly 231. The secondary structure model of spinach LHC-II reported here (Fig. 3a) is similar to the electron crystallographic model of LHC-II from pea 7. Between the primary structures of spinach and pea LHC-II, 89% of the 232 amino acids are conserved. However, deviations in the residue range, length, turns and orientation between helices in the two species were observed 7 (Fig. 3a). We also found a typical amphipathic short helix located in the BC loop region and named it helix E. Helix E has a length close to that of helix D and is related to helix D by the internal pseudo-c 2 axis. It is inclined with respect to the membrane plane by an angle of about 308. In the following EC loop, the polypeptide folds into two short antiparallel strands that are stabilized by an inter-strand ionic pair (Asp 111 His 120) and some hydrogen bonds. The basic structural and functional unit of LHC-II is the trimer. The whole trimerization region covers the amino-terminal domain, the carboxy terminus, the stromal end of helix B, several hydrophobic residues from helix C and also the pigments and lipid bound to these parts of the polypeptide chain (Fig. 3b). Chla 614, Chla 613, xanthophyll-cycle carotenoid, phosphatidylglycerol (PG), Chlb 601 and Chla 602 from one monomer together with Chlb 607, Chlb 609 and Chla 603 from the neighbouring monomer line up from the periphery of the trimer to the core region near the central C 3 axis at Figure 1 Electron-density map at 2.72 Å resolution. a, Chla and b, Chlb. Grey cage, 2F o F c density (1.5 j level); cyan cage, F o F c density (4.0 j level). No residual 2F o F c or F o F c density appears beside Chla C7-methyl, while strong 2F o F c and F o F c densities show up at the position of Chlb C7-formyl if it is omitted. c, N-terminal region including binding sites for a Chlb (cyan) and a phospholipid coordinated to a Chla (green). d, Two antiparallel polypeptide strands in the EC loop region with one Chlb bound. In c and d, 2F o F c densities (1.5 j level) are shown as a purple cage. 288 Figure 2 Organization and packing of the icosahedral particles. a, Schematic drawing of one-half of the LHC-II DGDG proteoliposome viewed along the c axis of the hexagonal cell. b, Packing diagram of Type III membrane-protein crystal, showing the contacts between icosahedral spherical particles in the hexagonal cell. Prosthetic groups are omitted for clarity. The N-terminal domain and AC loop region located at the stromal surface are involved in the crystal packing Nature Publishing Group NATURE VOL MARCH

89 the interface between monomers, forming extensive hydrophobic interactions. Six Chla (Chla 602 and Chla 603 from each monomer) constitute the core of the trimer. Our observations directly reveal the structural role of PG in stabilizing the LHC-II trimer and clearly indicate that hydrophobic interactions dominate the associations between monomers within a trimer. It was shown that removal of the first 49 or 51 amino-acid residues of the polypeptide by proteolytic cleavage led to loss of PG and complete dissociation of the trimer into monomers, and that hydrolysis of the PG by phospholipase A 2 has a similar effect in breaking down the LHC-II trimer 3. Chlorophyll-binding sites In a crystallographic asymmetric unit, the individual chlorophyllbinding sites in each LHC-II monomer are occupied by one type of chlorophyll (either Chla or Chlb). No mixed binding sites were observed. All central ligands of the 14 chlorophylls have been identified as side chains of seven amino-acid residues, two backbone carbonyls, four water molecules and the phosphodiester group of a PG (Supplementary Table 1; a comparison with a previous model 7 is also included). This coordination mode of Chla 611 to PG is the second case of its kind since its first discovery in photosystem I (ref. 10). On the other side, the phosphodiester group of PG forms a hydrogen-bonding and ionic interaction with the side chains of Tyr 44 and Lys 182 respectively (Fig. 1c). The polypeptide backbone NH and side chains also form hydrogen bonds with the C7-formyl groups (Chlb) and the C13 1 -keto groups of several chlorophylls (Supplementary Table 1). These interactions will not only strengthen the linkage between pigments and protein, but also influence the absorption characteristics of chlorophylls as shown previously 11. Except for Chlb 601, nearly all Chlb in the complex are selectively hydrogen-bonded to the polypeptide or to the coordinated water of Chlb 607 through their C7-formyls. The amide side chain of Gln 131 interacts with three Chlb molecules. One hydrogen bond is formed through the interaction of its C ¼ O with the coordinated water of Chlb 606, and two additional hydrogen bonds are formed by its NH 2 interacting with the C7-formyls of Chlb 607 and Chlb 609. Moreover, the C7-formyl of Chlb 606 is hydrogen-bonded to the coordinated water of Chlb 607. All these interactions bring three Chlb into close proximity, Figure 3 Secondary structure of monomeric LHC-II apoprotein and trimerization. View in parallel with the membrane plane. a, The vertical line indicates the approximate direction of the membrane normal and the position of the pseudo-c 2 axis. Helices are labelled A E. Helix E is newly defined, whereas others are labelled as before 7. The angle between the central axis of each helix and the membrane normal is shown in parentheses, with the residue range marked below each value. b, The interface between two adjacent monomers is shown. Colour code: yellow, amino-acid residues; green, Chla; cyan and blue, Chlb; magenta, xanthophyll-cycle carotenoids; pink, PG; red, water; maroon, C a traces of N-terminal (Ser 14 Asp 54) and C-terminal (Asp 215 Gly 231) polypeptide chain. The vertical line represents the local C 3 axis of an LHC-II trimer. articles resulting in the clustering of Chlb molecules in this region, which may facilitate the efficient energy transfer between these chlorophylls. It was suggested by functional investigations that Gln 131 is involved in the selective binding of Chlb molecules to LHC-II 12,13.As for the selective binding of Chla, we notice that the environment surrounding the C7-methyl groups of Chla molecules is mostly nonpolar. Hydrophobic repulsion or steric hindrance may be the factors affecting the binding affinity of Chlb to these Chla-binding sites. Chlorophyll arrangement for efficient light harvesting The chlorophylls in LHC-II are vertically distributed into two layers within the membrane, each layer lying close to the stromal or lumenal surface (Fig. 4a). Inside a monomer, the layer close to the stromal surface contains eight chlorophylls (five Chla and three Chlb), which surround the central helices A and B more or less evenly to form an elliptical ring (Fig. 4b). The average centre-tocentre distance between two neighbouring chlorophylls is about Å, with a maximum of Å and a minimum of 9.74 Å. Each chlorophyll inside this layer can find its symmetric mate related by the internal pseudo-c 2 axis. The remaining six chlorophylls (three Chla and three Chlb) are arranged in the layer close to the lumenal surface. They form two separate clusters comprising four chlorophylls (three Chlb and one Chla) and a Chla Chla dimer (Fig. 4c). Among them, Chlb 606 and Chla 604 are associated with the smallest centre-to-centre distance (8.05 Å) in LHC-II. The shortest distance between two chlorophyll layers is about Å (Chlb 609 to Chlb 606). Another interesting feature of this chlorophyll arrangement is the enrichment of Chlb molecules around helix C and at the interface between monomers (Fig. 4a). All six Chlb molecules are located in this region, with five of them belonging to one monomer and the remaining one (Chlb 601) from the neighbouring monomer. Chlb 601 (II) and Chlb 609 (I) (distance, Å) are the closest associated couple of chlorophylls between adjacent monomers within a trimeric LHC-II, indicating that this Chlb-rich region is of critical importance in energy equilibrating inside a functional trimer. In the trimeric LHC-II, all 24 chlorophylls from the stromal layer are organized into two irregular circular rings (Fig. 4d). The inner ring located in the core region of a trimer is composed of six Chla molecules that are thought to have an important role in intermonomeric energy transfer 14. The remaining nine Chla and nine Chlb (those covered by the yellow circular ring in Fig. 4d) form the outer ring and are arranged in a mosaic pattern, with three Chlb alternating with three Chla. This new pigment arrangement would favour the efficient absorption of incident light energy from all directions in a broad spectral region and the transfer of the excitation energy to the nearest exit, the putative terminal fluorescence emitter Chla 612 (Supplementary Table 2), in a few steps and at high rates. Energy transfer between two lumenal clusters are much less efficient than those within a stromal layer, as they are separated by larger distances (Fig. 4e). We infer that these lumenal chlorophyll clusters might serve as upstream energy collectors, absorbing energy and transmitting it to the stromal chlorophylls in a relatively independent way. The energy absorbed by the stromal chlorophylls is quickly focused on Chla 612/Chla 611 and is further transmitted to the neighbouring LHCs or reaction centres. Carotenoids as light-harvesting antennae The two central carotenoids with all-trans configurations are bound in the grooves on both sides of the supercoil (helices A and B) to form a cross-brace. They are assigned as lutein molecules (Fig. 4). Best fit with the electron density is achieved when the b-rings of both lutein molecules are oriented towards the lumenal surface and the e-rings point to the stromal surface. The polyene chains of lutein 620 and lutein 621 are inclined with respect to the membrane normal 89 NATURE VOL MARCH Nature Publishing Group

90 articles by angles of about 598 and 628, respectively. Both ring-shaped end groups of these two lutein molecules interact with four internal homologous segments of the polypeptide 15 located on both ends of helices A and B through van der Waals contacts and hydrogen bonds. Their polyene chains are firmly fixed in two elongated narrow hydrophobic cavities on both sides of the supercoil, providing strong and rigid linkage between helices A and B. They are indispensable for proper in vitro folding of LHC-II into stable complexes The third carotenoid, shaped like a bent-over hook, is located in the Chlb-rich region around helix C and is assigned as 9 0 -cis neoxanthin (Fig. 4). Its polyene chain forms an angle of about 588 with the membrane normal. A value of about 57 ^ 1.58 derived from linear dichroism spectra 20 confirms our assignment. The epoxycyclohexane ring of neoxanthin hangs over the chlorin ring of Chla 604 and is hydrogen-bonded to the hydroxyl of Tyr 112 via its C3 0 -hydroxyl. Side chains of Leu 134, Met 135, Val 138 from helix C and Trp 71 from helix B as well as chlorin rings and phytyl chains of Chlb 606 and Chlb 608 form a hydrophobic cleft that accommodates the hook-shaped polyene chain of neoxanthin. This binding site has been shown to be highly selective for neoxanthin 17,19. The cyclohexane ring of neoxanthin on the other end stretches into the exterior solvent region and exhibits weak electron density. The rate of singlet excitation energy transfer between carotenoids and chlorophylls is correlated with the mutual orientation between them, the centre-to-centre intermolecular distance and the closest distance between two conjugated parts (Supplementary Table 3). Six Chla are found to be in favourable orientations and distances with respect to two luteins for efficient singlet energy transfer from lutein to Chla. The data also show that efficient energy transfer from neoxanthin to Chlb 606 and Chlb 608 is highly possible. There is experimental evidence to suggest that singlet excitation energy of luteins is transferred exclusively to Chla molecules and not to Chlb 21. Neoxanthin was found to transfer its energy mostly towards Chlb 21,22. It can be concluded that lutein and neoxanthin found in LHC-II may function as effective accessory light-harvesting antennae, absorbing light in the blue green spectral region as a complement to Chla/b absorbing in the red region. This is in addition to their obvious structural role as well as their photoprotective role of quenching triplet chlorophylls and singlet oxygen 7. Structure-based non-photochemical quenching model The fourth carotenoid we discovered in LHC-II is located at the monomer monomer interface. The polyene chain of this carotenoid has an all-trans configuration and forms a small angle (348) with the membrane normal. As shown in Fig. 3b, a hydrophobic pocket is formed at the interface by several chlorophylls, hydrophobic residues from the polypeptide and the PG. Part of the polyene chain of this carotenoid together with one of its end groups is accommodated inside this pocket. The opposite end group sticks outside the binding pocket and faces the chlorin plane of Chlb 601 at the stromal side. The two ring-shaped end groups of this carotenoid exhibit distinct electron densities (one flat and the other bulgy between C-5 and C-6). This observation led to our original assignment of this carotenoid as an antheraxanthin, an intermediate in the xanthophyll cycle. However, later carotenoid composition analysis revealed that the major component of xanthophyll-cycle carotenoids in the LHC-II preparation used for crystallization is violaxanthin. We propose that the electron density may be interpreted by a mixed binding of different xanthophyll-cycle carotenoids at this site. The end group at the lumenal side points to the cavity formed Figure 4 Pigments in the LHC-II trimer and monomer. a, Stereo view showing the pigment arrangement pattern in the LHC-II trimer. View along the membrane normal from the stromal side. Monomers are labelled I III. For clarity, the chlorophyll phytyl chains and lipids are omitted. Green, Chla; blue, Chlb; yellow, lutein; orange, neoxanthin; magenta, xanthophyll-cycle carotenoids. b, c, Pigment pattern in a monomer at the stromal and lumenal sides, respectively. Colour designation the same as in a. d, e, Arrangement of chlorophylls within a LHC-II trimer at the stromal and lumenal sides, respectively. 290 Chlorophylls are represented by three atoms: the central magnesium atom and two nitrogen atoms. The connecting line between the two nitrogens defines the directions of the Q y transition dipole. Green, Chla nitrogen; blue, Chlb nitrogen; grey, magnesium; purple and blue ellipse, approximate monomer area. The magenta numerical note near the dark line connecting two chlorophylls indicates the centre-to-centre distance (Å) between them Nature Publishing Group NATURE VOL MARCH

91 around the local C 3 axis, suggesting that this cavity might be the docking site for violaxanthin de-epoxidase 23. The xanthophyll cycle was proposed to have a major role in adjusting the efficiency of light harvesting 4,24,25. It involves the conversion of violaxanthin into zeaxanthin through antheraxanthin. It was suggested that zeaxanthin molecules can act as direct quenchers of excess excitation by accepting singlet energy transferred from chlorophyll 26,27. We observed that at least three chlorophylls are close to the xanthophyll-cycle carotenoids and adopt favourable orientations for efficient singlet excitation transfer from chlorophylls to the xanthophyll-cycle carotenoids (Supplementary Table 3). In addition, we noticed that the distance between Chla 613 and Chla 614 is smaller than 10 Å, and their mutual orientation is close to an irregular distribution (Fig. 4e; Supplementary Table 2). These features agree well with the characteristics of statistical pair energy trap 28. We speculate that this pair of Chla molecules might also function as an excitation energy quencher, which would enhance the quenching effect of the xanthophyll-cycle carotenoids. Here we propose a structure-based non-photochemical quenching (NPQ) model concerning LHC-II (Fig. 5). Efficient nonphotochemical energy-transfer pathways are established upon aggregation of LHC-II trimers mediated by DGDG, so that the excitation energy is able to escape from one trimer to the adjacent trimer via these pathways. The following step is the migration of the excitation to the trapping site (the xanthophyll-cycle carotenoids and/or Chla 613 Chla 614 pair), where the non-radiative dissipation of excitation energy happens. This explains the NPQ observed upon incorporation of LHC-II into the liposomes containing DGDG 29. Conformational change induced by the protonation of photosystem II proteins including LHC-II was found to be necessary for NPQ 4,25. Among the seven lumen-exposed acidic residues in our structure, four of them (Glu 94, Asp 111, Glu 207 and Asp 211) form ion pairs with basic residues. The protonation of these acidic residues under low lumenal ph conditions may trigger the conformational changes of helix D and the BC loop. The linker chlorophylls (Chla 614 and Chlb 605) at the trimer trimer interface (Fig. 5) are coordinated to these regions of polypeptide chain. They may be moved and reoriented to promote the non-photochemical energy transfer and/or the quenching effect of the putative Figure 5 Structure-based non-photochemical chlorophyll fluorescence quenching model in oligomeric LHC-II. Top view along the icosahedral C 2 axis from the stromal side. DGDG is shown as a yellow transparent ball-and-stick model. Chlorophylls and xanthophyll-cycle carotenoids are represented as in previous figures. Black arrows represent the excitation energy-transfer pathways from one trimer to the neighbouring trimer and the orange arrow shows the possible transfer pathways from chlorophyll Q y to xanthophyll-cycle carotenoids S 1. The red stars indicate the putative quenching sites. For clarity, characters in one trimer are in black and those of the other are in grey. energy-trapping sites. As a consequence, the potential damaging effect of excess energy would be avoided. A Methods The LHC-II was isolated according to the protocol described previously 30. A single step of gel filtration chromatography with Hiload 16/60 Superdex 200 pg column (Pharmacia Biotech) was added to improve sample purification, ensuring crystallization reproducibility. The purified LHC-II was solubilized in a solution containing 0.8% n-nonyl-b-d-glucoside (BNG) (Anatrace) and 2 mg ml 21 DGDG (Lipid Products) to a final concentration of 4 mg ml 21 chlorophyll (about 8 mg ml 21 protein) and mixed with the crystallization solution containing 66.5 mm HEPES-NaOH ph 7.5, M citrate trisodium and 0.2% N,N-bis-(3-D-gluconamidopropyl)deoxycholamide (DBC) (Anatrace) in a ratio of 3.0:1.8 (v/v). The resulting drop was equilibrated against a well of 1 ml crystallization solution at Kusing the sitting-drop vapour-diffusion method. Green tabular crystals appeared a week later and grew to a maximum size of about mm after one month. Heavy-atom derivative was prepared by soaking the crystal for about 24 h in artificial mother liquor (50 mm HEPES-NaOH ph 7.5, 0.6% BNG, 0.1% DBC, 1.5 mg ml 21 DGDG, 1.09 M citrate trisodium) containing 0.5 mm K 2 HgI 4, followed by a backsoaking procedure for about 3 5 h in heavy-atom-free mother liquor. A cryoprotectant (50 mm HEPES-NaOH ph 7.5, 0.4% BNG, 0.15% DBC, 1.0 mg ml 21 DGDG, 1.15 M citrate trisodium, 11% saturated sucrose) was introduced to the crystal by soaking for a few minutes and then the crystal was flash-frozen for the X-ray diffraction experiment. The first native data set and the derivative data set were collected at PF (Tsukuba, Japan) beamline BL6B and the second native data set was collected at BSRF (Beijing, China) beamline 3W1A. A large crystal-to-detector distance ( mm) and a small oscillation (0.58) are necessary for reducing the overlap of reflections in the large diffraction angle region. Data were processed with Denzo and Scalepack 31. A typical icosahedral 532 point group symmetry was found by calculating self-rotation function with GLRF 32. We inferred that there is only one T ¼ 1 icosahedral particle residing in a primitive rhombohedral unit cell by analysing the crystal packing. The icosahedron is oriented with one of its 32 subgroups superposing with the crystallographic 32 point group. Each crystallographic asymmetric unit contains one-sixth of the icosahedron. Ten heavy-atom sites were located using SnB 33 in direct method mode at 5 Å. The arrangement of heavy atoms in the unit cell also abides with the icosahedral 532 symmetry, confirming our judgement based on self-rotation function and crystal packing analysis. The initial SIR phase was calculated with MLPHARE of the CCP4 suite 34 at 5 Å. Phase refinement and extension was performed according to the standard molecular replacement real-space averaging protocol 9. The NCS matrix was derived from the output of GLRF and improved by IMP 35. A molecular envelope enclosed by two icosahedral C 3 axes and one icosahedral C 5 axis covering an icosahedral asymmetric unit with a thickness of about 50 Å was generated by MAMA 35. Electron density was calculated using FFT (CCP4) and density averaging was performed with AVE 35. The SFALL program (CCP4) was used to calculate F c and a c from the averaged density map. The correlation coefficient and R factor were calculated using Rstats (CCP4). The a c were combined with the initial SIR phases a b using SigmaA (CCP4) and a new electron-density map was calculated with the F o and the combined phases. Phases were extended from 5.0 Å to 3.5 Å using a step size of 0.05 Å, and then to 2.72 Å with 0.02 Å step size. Phase refinement was performed for ten iterative cycles in each extension step. The final averaged electron-density map was of high quality, and the backbone was traced without much difficulty. Most of the pigments are clearly defined in the map. The electron-density map was interpreted using the program O 36. The polypeptide model was built with the help of a published primary sequence of Lhcb1 (ref. 37). The crystals mainly contain two highly homologous polypeptides, Lhcb1 and Lhcb2. Differences between the two polypeptides are confined to the N-terminal region, which may account for the weakness of the electron density in the region before Ser 14. Structure refinement was performed with CNS 38. At initial stages, two rounds of NCS-constrained simulated annealing (torsion angle dynamics protocol) were performed at 3.5 Å, followed by positional refinement. After the resolution was extended to 2.72 Å, the NCS constraint was switched to restraint mode and the restraint was gradually released at final stages. The individual B factors were refined using Å data. Peaks above 3.0 j in the F o F c electron density map were picked as candidates for water molecules, after the R factor was reduced to below 23%. Stereochemical restraints were introduced between chlorophylls and their central ligands during refinement. The final model was evaluated with PROCHECK 39 and all ten monomers have good geometry with only one Ramachandran outlier (Val 119) per monomer, the backbone carbonyl of which coordinates a chlorophyll. Secondary structures of polypeptides were analysed with STRIDE 40. Figures 1c, d and 2b was prepared with program O 36. All other figures were prepared with Molscript 41 and Raster3d 42. Received 29 October 2003; accepted 27 January 2004; doi: /nature articles 1. Peter, G. F. & Thornber, J. P. Biochemical composition and organization of higher plant photosystem II light-harvesting pigment-proteins. J. Biol. Chem. 266, (1991). 2. Ruban, A. V., Lee, P. J., Wentworth, M., Young, A. J. & Horton, P. Determination of the stoichiometry and strength of binding of xanthophylls to the photosystem II light harvesting complexes. J. Biol. Chem. 274, (1999). 3. Nußberger, S., Dörr, K., Wang, D. N. & Kühlbrandt, W. Lipid protein interactions in crystals of plant light-harvesting complex. J. Mol. Biol. 234, (1993). 4. Horton, P., Ruban, A. V. & Walters, R. G. Regulation of light harvesting in green plants. Annu. Rev. Plant Physiol. Plant Mol. Biol. 47, (1996). 5. Elrad, D., Niyogi, K. K. & Grossman, A. R. A major light-harvesting polypeptide of photosystem II functions in thermal dissipation. Plant Cell 14, (2002). 6. Nilsson, A. et al. Phosphorylation controls the three-dimensional structure of plant light harvesting 91 NATURE VOL MARCH Nature Publishing Group

92 articles complex II. J. Biol. Chem. 272, (1997). 7. Kühlbrandt, W., Wang, D. N. & Fujiyoshi, Y. Atomic model of plant light-harvesting complex by electron crystallography. Nature 367, (1994). 8. Michel, H. Crystallization of membrane proteins. Trends Biochem. Sci. 8, (1983). 9. Rossmann, M. G. in Methods in Macromolecular Crystallography (eds Turk, D. & Johnson, L.) (IOS Press, Amsterdam, 2001). 10. Jordan, P. et al. Three-dimensional structure of cyanobacterial photosystem I at 2.5 Å resolution. Nature 411, (2001). 11. McLuskey, K., Prince, S. M., Cogdell, R. J. & Isaacs, N. W. The crystallographic structure of the B LH3 light-harvesting complex from the purple bacteria Rhodopseudomonas acidophila strain Biochemistry 40, (2001). 12. Bassi, R., Croce, R., Cugini, D. & Sandonà, D. Mutational analysis of a higher plant antenna protein provides identification of chromophores bound into multiple sites. Proc. Natl Acad. Sci. USA 96, (1999). 13. Remelli, R., Varotto, C., Sandonà, D., Croce, R. & Bassi, R. Chlorophyll binding to monomeric lightharvesting complex. A mutation analysis of chromophore-binding residues. J. Biol. Chem. 274, (1999). 14. Gradinaru, C. C. et al. The flow of excitation energy in LHCII monomers: implications for the structural model of the major plant antenna. Biophys. J. 75, (1998). 15. Bassi, R., Sandonà, D. & Croce, R. Novel aspects of chlorophyll a/b-binding proteins. Physiol. Planta. 100, (1997). 16. Plumley, F. G. & Schmidt, G. W. Reconstitution of chlorophyll a/b light-harvesting complexes: Xanthophyll-dependent assembly and energy transfer. Proc. Natl Acad. Sci. USA 84, (1987). 17. Croce, R., Weiss, S. & Bassi, R. Carotenoid-binding sites of the major light-harvesting complex II of higher plants. J. Biol. Chem. 274, (1999). 18. Paulsen, H., Finkenzeller, B. & Kuhlein, N. Pigments induce folding of light-harvesting chlorophyll a/b-binding protein. Eur. J. Biochem. 215, (1993). 19. Hobe, S., Niemeier, H., Bender, A. & Paulsen, H. Carotenoid binding sites in LHCIIb. Relative affinities towards major xanthophylls of higher plants. Eur. J. Biochem. 267, (2000). 20. Croce, R., Remelli, R., Varotto, C., Breton, J. & Bassi, R. The neoxanthin binding site of the major light harvesting complex (LHCII) from higher plants. FEBS Lett. 456, 1 6 (1999). 21. Gradinaru, C. C., Stokkum, I. H. M. V., Pascal, A. A., Grondelle, R. V. & Amerongen, H. V. Identifying the pathways of energy transfer between carotenoids and chlorophylls in LHCII and CP29. A multicolor, femtosecond pump-probe study. J. Phys. Chem. B 104, (2000). 22. Croce, R., Müller, M. G., Bassi, R. & Holzwarth, A. R. Carotenoid-to-chlorophyll energy transfer in recombinant major light-harvesting complex (LHC-II) of higher plants. I. femtosecond transient absorption measurements. Biophys. J. 80, (2001). 23. Hieber, A. D., Bugos, R. C. & Yamamoto, H. Y. Plant lipocalins: violaxanthin de-epoxidase and zeaxanthin epoxidase. Biochim. Biophys. Acta 1482, (2000). 24. Demmig-Adams, B. & Adams, W. W. Photoprotection and other responses of plants to high light stress. Annu. Rev. Plant Physiol. Plant Mol. Biol. 43, (1992). 25. Gilmore, A. M. Mechanistic aspects of xanthophyll cycle-dependent photoprotection in higher plant chloroplasts and leaves. Physiol. Planta. 99, (1997). 26. Frank, H. A. et al. Photophysics of the carotenoids associated with the xanthophyll cycle in photosynthesis. Photosynth. Res. 41, (1994). 27. Ma, Y. Z., Holt, N. E., Li, X. P., Niyogi, K. K. & Fleming, G. R. Evidence for direct carotenoid involvement in the regulation of photosynthetic light harvesting. Proc. Natl Acad. Sci. USA 100, (2003). 28. Beddard, G. S., Carlin, S. E. & Porter, G. Concentration quenching of chlorophyll fluorescence in bilayer lipid vesicles and liposomes. Chem. Phys. Lett. 43, (1976). 29. Moya, I., Silvestri, M., Vallon, O., Cinque, G. & Bassi, R. Time-resolved fluorescence analysis of the photosystem II antenna proteins in detergent micelles and liposomes. Biochemistry 40, (2001). 30. Lou, S., Wang, K., Zhao, F., Xu, C. & Kuang, T. A comparative study on PS II light harvesting chlorophyll a/b protein complexes between spinach and cucumber. Acta Bot. Sin. 37, (1995). 31. Otwinowski, Z. & Minor, W. Processing of X-ray diffraction data collected in oscillation mode. Methods Enzymol. 276, (1997). 32. Tong, L. & Rossmann, M. G. The locked rotation function. Acta Crystallogr. A 46, (1990). 33. Weeks, C. M. & Miller, R. The design and implementation of SnB v2.0. J. Appl. Crystallogr. 32, (1999). 34. CCP4 Collaborative Computational Project, The CCP4 suite: programs for protein crystallography. Acta Crystallogr. D 50, (1994). 35. Jones, T. A. in Molecular Replacement (eds Dodson, E. J., Gover, S. & Wolf, W.) (SERC Daresbury Laboratory, Warrington, 1992). 36. Jones, T. A., Zou, J. Y., Cowan, S. W. & Kjeldgaard, M. Improved methods for building protein models in electron density maps and the location of errors in these models. Acta Crystallogr. A 47, (1991). 37. Mason, J. G. Nucleotide sequence of a cdna encoding the light-harvesting chlorophyll a/b binding protein from spinach. Nucleic Acids Res. 17, 5387 (1989). 38. Brünger, A. T. et al. Crystallography and NMR system: A new software suite for macromolecular structure determination. Acta Crystallogr. D 54, (1998). 39. Laskowski, R. A., MacArthur, M. W., Moss, D. S. & Thornton, J. M. PROCHECK: a program to check the stereochemical quality of protein structures. J. Appl. Crystallogr. 26, (1993). 40. Frishman, D. & Argos, P. Knowledge-based secondary structure assignment. Proteins Struct. Funct. Genet. 23, (1995). 41. Kraulis, P. J. MOLSCRIPT: a program to produce both detailed and schematic plots of protein structures. J. Appl. Crystallogr. 24, (1991). 42. Merritt, E. A. & Murphy, M. E. P. Raster3D version 2.0. A program for photorealistic molecular graphics. Acta Crystallogr. D 50, (1994). Supplementary Information accompanies the paper on Acknowledgements We thank D. C. Liang and the late P. S. Tang for their efforts in initiating this project; X. C. Gu for discussions; N. Sakabe and K. Sakabe at PF (Tsukuba, Japan) and the staff at BSRF (Beijing, China) for their support during data collection at the synchrotron facilities. This research was financially supported by the National Key Research Development Project of China, the National Natural Science Foundation of China, the National Key Special Research Program and the Knowledge Innovation Program of the Chinese Academy of Sciences. Competing interests statement The authors declare that they have no competing financial interests. Correspondence and requests for materials should be addressed to W.R.C. (wrchang@sun5.ibp.ac.cn). Coordinates and structure factors have been deposited in the Protein Data Bank under accession code 1RWT Nature Publishing Group NATURE VOL MARCH

93 3D structure of human FK506-binding protein 52: Implications for the assembly of the glucocorticoid receptor Hsp90 immunophilin heterocomplex Beili Wu*, Pengyun Li*, Yiwei Liu*, Zhiyong Lou*, Yi Ding*, Cuiling Shu, Sheng Ye*, Mark Bartlam*, Beifen Shen, and Zihe Rao* *Laboratory of Structural Biology, Tsinghua University, Beijing , China; National Laboratory of Biomacromolecules, Institute of Biophysics, Chinese Academy of Science, Beijing , China; and Beijing Institute of Basic Medical Science, Beijing , China Edited by Timothy A. Springer, Harvard Medical School, Boston, MA, and approved April 14, 2004 (received for review September 17, 2003) FK506-binding protein 52 (FKBP52), which binds FK506 and possesses peptidylprolyl isomerase activity, is an important immunophilin involved in the heterocomplex of steroid receptors with heat-shock protein 90. Here we report the crystal structures of two overlapped fragments [N(1 260) and C( )] of FKBP52 and the complex with a C-terminal pentapeptide from heat-shock protein 90. Based on the structures of these two overlapped fragments, the complete putative structure of FKBP52 can be defined. The structure of FKBP52 is composed of two consecutive FKBP domains, a tetratricopeptide repeat domain and a short helical domain beyond the final tetratricopeptide repeat motif. Key structural differences between FKBP52 and FKBP51, including the relative orientations of the four domains and some important residue substitutions, could account for the differential functions of FKBPs. Immunophilins are proteins possessing peptidylprolyl isomerase (PPIase) domains that bind immunosuppressant drugs. According to their binding affinity for different drugs, immunophilins have been divided into two families: FK506-binding proteins (FKBPs), which bind FK506 and rapamycin, and cyclophilins, which bind cyclosporin A (1, 2). FKBP52 is an immunophilin belonging to the FKBP family, with a molecular mass of 52 kda, and was first discovered as a component of an inactive steroid receptor heat-shock protein 90 (Hsp90) complex (3). The assembly pathway of steroid receptors involves multiple chaperone and cochaperone proteins such as Hsp90, Hsp40, Hsp70, Hsp70/90 organizing protein (Hop), and P23 (4 6). Matured steroid receptor complexes contained Hsp90 and at least one immunophilin: FKBP52, another FKBP (FKBP51), the cyclosporin A-binding protein cyclophilin 40 (Cyp40), or the protein phosphatase 5 (PP5). These cochaperones all possess tetratricopeptide repeat (TPR) domains, which form the binding sites for Hsp90 (7). The C-terminal MEEVD sequence motif of Hsp90 has been shown to be critical for the binding of Hsp90 to various TPR-containing proteins (8). Immunophilins are not required for steroid receptors to bind hormones in vitro (5, 9, 10), but they are known to influence steroid signaling pathways. Receptor-associated FKBP52 and PP5 are suggested to play a role in the shuttling of steroid receptors between cytoplasm and nuclear compartments (7, 11 13). Overexpression of FKBP51 may be responsible for the low hormone-binding affinity of glucocorticoid receptor in both human and squirrel monkey (14 16). Although FKBP52 and FKBP51 share 75% sequence similarity (Fig. 1a), they affect hormone binding by glucocorticoid receptor in opposing manners and have different Hsp90-binding characteristics. It is not clear what structural features are responsible for these functional differences between FKBP52 and FKBP51. The crystal structure of FKBP51 has been solved recently, as well as that of the PPIase domain of FKBP52 (16, 17). FKBP52 is difficult to crystallize because of its instability in solution. To overcome this problem, we have determined the structures of two overlapping fragments of FKBP52: N(1 260) [(1 260)a (space group P2 1 ) and N(1 260)b (space group P )], and C( ). Based on these two structures, we have defined the whole putative structure of FKBP52, which is architecturally similar to FKBP51 except that the orientations between different domains are diverse. The structural alterations give the clues for the differential effects of FKBP52 and FKBP51. We have also determined the structure of C( ) in complex with the C-terminal pentapeptide (MEEVD) from Hsp90. The complex structure maps out the binding pocket in the TPR domain and reveals several key residues involved in Hsp90 binding. Methods Protein Expression and Purification. Codons [N(1 260)] and [C( )] of human FKBP52 were cloned into the His-6-tag expression plasmid pet28a( ) (Novagen). N(1 260) was overexpressed in Escherichia coli strain BL21 (DE3), and the selenomethionine-labeled C( ) was produced by expression in methionine-deficient E. coli strain B834 (DE3). The soluble proteins were purified by using Ni 2 -nitrilotriacetic acid agarose (Qiagen, Valencia, CA) and chromatography on Superdex 75 and Resource Q (Pharmacia). Crystallization. The protein solution of N(1 260) used for crystallization contained 20 mm Tris (ph 8.0), 150 mm NaCl, and 20 mg/ml protein. Crystals were obtained by using the hangingdrop vapor-diffusion technique with reservoir solutions containing 28 31% polyethylene glycol 6000, 3 5% DMSO, and 100 mm Tris [ph 8.0 for N(1 260)a] and 200 mm calcium chloride, 20% polyethylene glycol 3350 [for N(1 260)b] in a drop formed by mixing 1 l of protein solution and 1 l of reservoir solution at 291 K. Crystals of C( ) were grown by using a reservoir solution of 100 mm Tris, ph M ammonium sulfate 2 4% (vol/vol) ethanol. Protein concentration was 10 mg/ml in 20 mm Tris, ph 8.0/100 mm NaCl/5 mm DTT. Crystals of the C( ) MEEVD complex (protein/peptide, 1:1.5) were grown by using a reservoir solution of 1.2 M sodium citrate (ph 6.7). This paper was submitted directly (Track II) to the PNAS office. Abbreviations: PPIase, peptidylprolyl isomerase; FKBP, FK506-binding protein; Hsp, heatshock protein; Hop, Hsp70 90 organizing protein; Cyp40, cyclophilin 40; PP5, protein phosphatase 5; TPR, tetratricopeptide repeat; FK1, first FKBP domain of FKBPs; FK2, second FKBP domain of FKBPs. Data deposition: The atomic coordinates and structure factors for N(1 260) (space group P2 1), C( ), and C( ) MEEVD have been deposited in the Protein Data Bank, (PDB ID codes 1Q1C, 1P5Q, and 1QZ2, respectively). B.W. and P.L. contributed equally to this work. To whom correspondence may be addressed. raozh@xtal.tsinghua.edu.cn or shenbf@mx.cei.gov.cn by The National Academy of Sciences of the USA PNAS June 1, 2004 vol. 101 no cgi doi pnas

94 Fig. 1. (a) Sequence alignment of human FKBP52 (hfkbp52) and human FKBP51 (hfkbp51). Amino acids with high consensus are shown in red. Human FKBP52 shares 60% amino acid sequence identity and 75% similarity with human FKBP51. Four domains are indicated by different underlines: single underline, FK1 domain; double underline, FK2 domain; thick underline, TPR domain; dashed underline, calmodulin-binding domain. (b) The final N(1 260) model contains residues (c) The final C( ) model contains residues (d) The overall structure of FKBP52 has been defined based on the superposition of overlapped regions of N(1 260) and C( ). (e) Stereo view of the structural comparison between FKBP51 (blue) and FKBP52 (yellow) shows their similar structural architectures but the different orientations of their corresponding domains. Data Collection and Processing. X-ray data were collected from a single C( ) crystal to 2.7 Å at 90 K. Multiwavelength anomalous diffraction data were collected on a charge-coupled device detector at the BL41XU beamline of SPring-8 (Hyogo, Japan). Data sets were indexed and scaled by using HKL2000 (18). X-ray data of N(1 260)a were collected at the 13th beamline at Beijing Synchrotron Radiation Facility (Beijing, China) to 1.9 Å. Data of N(1 260)b were collected on a Rigaku (Tokyo) RU2000 rotating Cu K anode source to 2.0 Å. Data sets were processed by using DENZO and SCALEPACK (18). The data for the C( ) MEEVD complex were collected to 3.0 Å at the Advanced Photon Source (Argonne, IL) and processed by using HKL2000. Structure Determination and Refinement. All 12 possible selenium sites of C( ) were found and refined at a 2.8-Å resolution by using SOLVE (19), which produced a mean figure of merit of After density modification with RESOLVE (20), 50% of all the residues of C( ) were automatically traced into the experimental density map; the remaining residues were traced manually by using O (21). CNS (22) was used for refinement and addition of solvent molecules. Structures of N(1 260)a and N(1 260)b were solved by molecular replacement using CNS. The structures of the FKBP52 N-terminal domain (residues ) (PDB ID code 1N1A) and the N-terminal fragment of C( ) (residues ) were used as starting models. This model was subjected to rigid-body refinement, and manual adjustments were made to the model in O. Water molecules were added by using CNS. The structure of the C( ) MEEVD complex was solved by molecular replacement with CNS, with C( ) as the search model. Data collection, processing, and refinement statistic are given in Table 1. Results and Discussion N(1 260), C( ), and Putative FKBP52 Structures. FKBP52 can be divided into four domains according to sequence analysis (17). Two overlapped segments of FKBP52 [designated N(1 260) and C( )] have been cloned, expressed, purified, and crystallized. N(1 260) includes the first two FKBP domains [named FK1 (first FKBP domain of FKBPs) and FK2 (second FKBP domain of FKBPs)], whereas C( ) is composed of the second FKBP domain, the TPR domain, and a calmodulinbinding domain. The final refined N(1 260) model contains residues of FKBP52 (Fig. 1b), whereas the C( ) model contains residues (Fig. 1c). Based on the superposition of overlapped regions of N(1 260) and C( ), the overall structure of FKBP52 has been clearly defined (Fig. 1d). The FKBP domains of FKBP52 both consist of a five- to six-stranded antiparallel -sheet, which is wrapped around a short -helix with a right-handed twist, and are similar to those of FKBP51, FKBP12, and macrophage infectivity potentiator protein (16, 23, 24) (Fig. 2a). The TPR domain of FKBP52 is all helical and consists of three units of a consensus 34-aa motif. Each single unit consists of two consecutive -helices containing residues (except 1 and 3, containing 21 and 23 residues, respectively) that cross at an angle of 20 to each other. The organizational pattern of the FKBP52 TPR domain is similar to those of FKBP51, Cyp40, PP5, and Hop (8, 16, 25, 26). There is an additional -helix ( 7) in the C terminus beyond the final TPR motif that contains the calmodulin-binding site (Fig. 2b). The overall structure of FKBP52 is very similar to that of FKBP51 except for their relative domain orientations (16) (Fig. 1e). The rms deviations BIOPHYSICS Wu et al. PNAS June 1, 2004 vol. 101 no

95 Table 1. Data collection and refinement statistics C( ) N(1 260)a N(1 260)b C( ) MEEVD Data collection Space group C222 1 P2 1 P C222 1 Unit cell a b c, Å , Wavelength, Å (peak) (edge) (remote) Resolution, Å Completeness, % (99.8) (99.9) (100.0) 99.2 (98.3) 99.9 (99.9) 98.7 (98.8) Reflections Total 291, , ,271 72,917 78, ,154 Unique 38,801 38,872 34,752 24,881 11,441 27,894 Redundancy R merge, %* 6.4 (32.9) 5.0 (34.4) 6.2 (42.2) 10.1 (40.1) 9.3 (31.4) 9.7 (58.9) I (I) 11.9 (4.2) 13.8 (4.7) 11.5 (4.0) 8.3 (2.7) 8.9 (5.4) 6.5 (1.9) Refinement statistics Resolution, Å R factor, % Working set Test set Rms deviation Bonds, Å Angles, Ramachandran plot, % Most favored Allowed Generously allowed Disallowed *R merge h l I ih I h h l I h, where I h is the mean of the observations I ih of reflection h. R work ( F obs F calc ) F obs ; R free is the R factor for a subset (10%) of reflections that was selected before refinement calculations and not included in the refinement. Ramachandran plots were generated by using PROCHECK (14). between corresponding FK1, FK2, and TPR domains are 0.6, 0.9, and 1.0 Å, respectively, for all C atoms. The FK1 and FK2 domains of FKBP52 are linked by a highly hydrophilic hinge (residues ), and the FK506-binding pocket of FK1 is oriented 180 from the putative FK506- binding pocket of FK2. Residues form an antiparallel -strand that interacts with residues in the 1 strand of FK2. This loop is stabilized by the formation of a hydrogen-bond network with the side chains of neighboring residues (Fig. 3a). The extensive interactions in the hinge region suggest that the flexibility of the hinge is restrained. The side chains of several residues in the interface (Met-48, Ile-49, Ile-154, and Arg-157) form a hydrophobic core, together with those of Asp-141, Leu-142, Ile-150, Ile-151, and Arg-210. Although a small difference is found in the orientation between FK1 and FK2 of N(1 260)a and N(1 260)b, which have different crystal packing, the conformations of the loop are mostly the same, as well as the hydrogen-bond interactions and the hydrophobic cores. These tight interactions and the rigid hinge stabilize the relative orientation of FK1 and FK2 and restrict the great change of this orientation. The FK1 domain is responsible for the PPIase activity of FKBP52, whereas the FK2 domain shows no PPIase activity (27, 28). Detailed structural information for FK1 has been reported (17). FK2 is architecturally similar to FK1 despite sharing only 32% sequence identity but lacks the large bulge-splitting strand 4 of FK1 (Fig. 2a). FK1 contains 14 residues directly related to substrate binding, yet only five residues are conserved in FK2. In FK1, Ile-87 and Tyr-113 form highly conserved hydrogen bonds with substrates, and Trp-90 provides a platform for substrate binding (17), whereas in FK2 the corresponding residues are substituted by Pro-201, Phe-232, and Leu-204, respectively. The insertion of one residue (Lys-234) in the loop and five residues (Gly-195, Glu-196, Met-197, Leu-198, and Asp-199) in the loop, which both flank the binding pocket of FK2, pushes the loops into the binding pocket. The large and extended side chains of Lys-232, Glu-233, and Phe-235 (corresponding to Ser-118, Pro-119, and Lys-121 of FK1, respectively) block the opening of the pocket and change the electrostatic state, in addition to the insertions mentioned above. We conclude that alterations in loops surrounding the binding pocket and substitution of residues important for the substrate binding account for the loss of the PPIase activity of FK2. Previously, a deletion mutant (deletion of Asp-195, His-196, and Asp-197) of FKBP51 has shown that the FK2 domain relates to progesterone receptor preference and target protein interaction (16). Presumably, FK2 domains of FKBP proteins may result from a duplication event. During evolution, FK2 domains lost their PPIase activity but act as organizers of different domains and provide interaction interfaces with other target proteins such as Hsp90 and steroid receptors. Structure of the C( ) MEEVD Complex. The overall structure of the peptide-bound C( ) is similar to that of the free protein. The peptide binds in a cavity formed by the 1, 3, and 5 helices of the TPR domain. Of the three molecules in the asymmetric unit, two (molecules A and B) bind the MEEVD peptide, whereas the third (molecule C) does not. The peptidebinding pockets in molecules A and B are composed of exactly the same residues, namely Lys-282, Asn-324, Met-327, Lys-354, and Arg-358 (Fig. 3 b and c). Some interactions are important to stabilize the binding of the peptide and are conserved in both molecules. In particular, there is a hydrogen bond formed between Lys-282 and Met-1, and two hydrogen bonds are cgi doi pnas Wu et al. 95

96 Fig. 2. Stereo view of the superposition of FK and TPR domains. (a) Two FKBP domains of FKBP51 and FKBP52 were superimposed onto FKBP12. FKBP12 (green), 51-FK1 (blue), and 52-FK1 (red) are similar. The structures of 51-FK2 (cyan) and 52-FK2 (yellow) are more closed than the others. (b) TPR domains are superimposed onto the TPR domains of FKBP52. FKBP52 is shown in yellow, FKBP51 is shown in cyan, Hop is shown in green, Cyp40 is shown in purple, and PP5 is shown in pink. The conformations of all the TPR domains are similar, containing six -helices ( 1 6). The orientations of the extra -helix ( 7) are different. (c) Superposition of the structures of TPR domains and the 7- helixes of FKBP51 (blue) and FKBP52 (yellow). Gln-333, Phe-335, and Ala-365 of FKBP52 are replaced by Arg-331, Tyr-333, and Leu-363 in FKBP51, which may be responsible for the differential binding pattern of FKBPs to Hsp90. The side chain of Ile-400 of FKBP52, corresponding to Ala-398 of FKBP51, will clash with Phe-369, and this may cause the different orientations of the 7-helix. formed between Lys-354, Arg-358, and Glu-2. Nevertheless, the orientations of the peptides in these two molecules are different. Additional amino acids N-terminal to this peptide sequence in Hsp90 also contribute to the binding and render specificity to the interaction between the TPR and Hsp90 (8). As a consequence, the binding between the MEEVD peptide and C( ) is not strong, which may account for the different observed binding modes. In the unit cell, the pocket of molecule A faces the pocket of molecule B, and there is enough space for the peptides to fill in, but the binding pocket of molecule C is blocked by the FK2 domain of another molecule C, which explains why this molecule does not bind the peptide. Previous mutation analysis and crystal structures of TPR domains have identified residues in Hop, Cyp40, and PP5 that are essential for Hsp90 binding (8, 25, 26, 29, 30). Comparing these proteins with FKBP52, the key residues for binding Hsp90 are conserved, which suggests a functional similarity between these proteins. Comparison with the Structure of FKBP51. Although the overall architectures of FKBP52 and FKBP51 are very similar, the relative orientations of their four domains are distinctly different (Fig. 1e). Superimposing the structures of FKBP52 and FKBP51 by using the FK2 domain as the reference, we observe that domain FK1 of FKBP51 is rotated at an angle ( 24, 33 ) relative to FK1 of FKBP52. The hinge linking FK1 and FK2 is stabilized by extensive hydrogen bonds and hydrophobic interactions in both FKBP proteins. The different hinge conformations of these two proteins and specific interactions in the interface between FK1 and FK2, together with a different hydrophobic core, may account for their different orientations. Thr-143 and Arg-206 of FKBP52 form four pairs of hydrogen bonds with Asp-141, Glu-144, and Glu-146. In FKBP51, the corresponding residues are Phe-143 and Lys-204, which do not form hydrogen bonds. Additionally, the side chains of Ile-154 and Arg-157 in FKBP52 (corresponding to Thr-152 and Lys-155 of FKBP51, respectively) provide tighter hydrophobic interactions. Available evidence suggests that the hinge may be involved in the interaction between FKBP52 and Hsp90. Thr-143 of FKBP52 has been identified as the major phosphorylation site for casein kinase II, and it is known that casein kinase II-phosphorylated FKBP52 does not bind to Hsp90 (31). Furthermore, Thr-143 is changed to Phe-143 in FKBP51, which may have a differential effect on the Hsp90 binding. The TPR domain orientation of FKBP52 is considerably different from that of FKBP51. There are fewer hydrogen bonds and hydrophobic interactions in the interface between FK2 and TPR domains, as well as the loop linking FK2 and TPR domains. Considering the flexibility of this area, it is likely that the different orientations of TPR domains are due to different crystal packing. The TPR domain chimera and truncation experiments have shown that the TPR domain of FKBP51 requires appropriate downstream sequences for Hsp90 binding, but the TPR domain of FKBP52 does not. Additionally, the C-terminal region of FKBP51, which functionally interacts with the TPR domain to permit Hsp90 binding, confers preferential association with progesterone receptor (32, 33). Comparing the TPR domain structures of FKBP52 and FKBP51 (Fig. 2), most residues in the binding interface between FKBP52 and Hsp90 are strictly conserved, with the exception of Gln-333, Phe-335, and Ala- 365 (corresponding to Arg-331, Tyr-333, and Leu-363 of FKBP51, respectively). The substitutions of Gln to Arg and Ala to Leu results in changes to the electronic state and steric relationship of the interaction interface, which would account for the differential binding pattern of FKBPs to Hsp90. Superimposing the TPR domains and the extra seventh -helix ( 7) of FKBPs, it can be observed that structural architectures of TPR domain are very similar. However, the 7-helix of FKBP52 shifts away by 30 from the Hsp90-binding interface relative to FKBP51 (Fig. 2). The different orientations of the 7-helix in FKBPs can be explained by the side chain of Ile-400 (corresponding to Ala-398 in FKBP51) in the 7-helix of FKBP52, which would otherwise clash with the phenyl group of Phe-369 in the FKBP51 conformation. The more compact orientation of the C terminus of FKBP51 may result in the BIOPHYSICS Wu et al. PNAS June 1, 2004 vol. 101 no

97 Fig. 3. (a) Stereo view of the hydrogen bonds between FK1 and FK2 of FKBP52. Hydrogen bonds at the interface of FK1 and FK2 form a complicated network, which stabilizes the conformation. Residues in FK1 are shown in red, residues in FK2 are shown in yellow, and residues in the loop are shown in white. (b and c) Stereo view of the MEEVD peptide bound to molecules A (b) andb(c) of C( ). The omit electron-density map is contoured at 0.7 above the mean. Residues of the peptide are shown in white, and residues of the TPR domain are shown in yellow. Residues involved in important interactions are shown in ball-and-stick representation. Hydrogen bonds are shown as dotted lines. proximity and direct interaction with Hsp90, which is consistent with previous studies. Conclusions A detailed structural analysis of FKBP52 and the complex with the MEEVD peptide has revealed the essential basis for the loss of PPIase activity of the FK2 domain and the key residues for Hsp90 binding. We observe a number of interesting structural variances between FKBP52 and FKBP51, including the relative orientations of corresponding domains and some important residue substitutions. The very real differences in orientation between the FK1 and FK2 domains in FKBP52 and FKBP51 could explain their differential effects on Hsp90 binding. We thank Xuemei Li, Yuna Sun, and Wei Zheng (Z.R. group) for technical assistance; Min Yao (Hokkaido University, Sapporo, Japan), Rongguang Zhang, Andrzej Joachimiak (Advanced Photon Source, Argonne, IL), Fei Sun, Feng Xu, and Feng Gao (Z.R. group) for assistance with data collection; and Dr. R. B. Sim for critical reading of the manuscript. The peptide MEEVD was a gift from Yuzhang Wu (Third Military Medical University, Chongging, China). This work was supported by Project 863 Grant 2002BA711A12, Project 973 Grant G , and National Natural Science Foundation of China Grant Dolinski, K., Muir, S., Cardenas, M. & Heitman, J. (1997) Proc. Natl. Acad. Sci. USA 94, Galat, A. (1993) Eur. J. Biochem. 216, Sanchez, E. R. (1990) J. Biol. Chem. 265, Pratt, W. B. & Toft, D. O. (1997) Endocr. Rev. 18, Cheung, J. & Smith, D. F. (2000) Mol. Endocrinol. 14, Morishima, Y., Kanelakis, K. C., Silverstein, A. M., Dittmar, K. D., Estrada, L. & Pratt, W. B. (2000) J. Biol. Chem. 275, Riggs, D. L., Roberts, P. J., Chirillo, S. C., Cheung-Flynn, J., Prapapanich, V., Ratajczak, T., Gaber, R., Picard, D. & Smith, D. F. (2003) EMBO J. 22, Scheufler, C., Brinker, A., Bourenkov, G., Pegoraro, S., Moroder, L., Bartunik, H., Hartl, F. U. & Moarefi, I. (2000) Cell 101, cgi doi pnas Wu et al. 97

98 9. Kosano, H., Stensgard, B., Charlesworth, M. C., McMahon, N. & Toft, D. (1998) J. Biol. Chem. 273, Dittmar, K. D. & Pratt, W. B. (1997) J. Biol. Chem. 272, Czar, M. J., Lyons, R. H., Welsh, M. J., Renoir, J. M. & Pratt, W. B. (1995) Mol. Endocrinol. 9, Galigniana, M. D., Radanyi, C., Renoir, J. M., Housley, P. R. & Pratt, W. B. (2001) J. Biol. Chem. 276, Dean, D. A., Urban, G., Aragon, I. V., Swingle, M., Miller, B., Rusconi, S., Bueno, M., Dean, N. M. & Honkanen, R. E. (2001) BMC Cell Biol. 2, Reynolds, P. D., Ruan, Y., Smith, D. F. & Scammell, J. G. (1999) J. Clin. Endocrinol. Metab. 84, Denny, W. B., Valentine, D. L., Reynolds, P. D., Smith, D. F. & Scammell, J. G. (2000) Endocrinology 141, Sinars, C. R., Cheung-Flynn, J., Rimerman, R. A., Scammell, J. G., Smith, D. F. & Clardy, J. (2003) Proc. Natl. Acad. Sci. USA 100, Li, P., Ding, Y., Wu, B., Shu, C., Shen, B. & Rao, Z. (2003) Acta Crystallogr. D 59, Otwinowski, Z. & Minor, W. (1997) Methods Enzymol. 276, Terwilliger, T. C. & Berendzen, J. (1999) Acta Crystallogr. D 55, Terwilliger, T. C. (2000) Acta Crystallogr. D 56, Brunger, A. T., Adams, P. D., Clore, G. M., DeLano, W. L., Gros, P., Grosse-Kunstleve, R. W., Jiang, J. S., Kuszewski, J., Nilges, M., Pannu, N. S., et al. (1998) Acta Crystallogr. D 54, Jones, T. A., Zou, J. Y., Cowan, S. W. & Kjeldgaard, M. (1991) Acta Crystallogr. A 47, Van Duyne, G. D., Standaert, R. F., Karplus, P. A., Schreiber, S. L. & Clardy, J. (1991) Science 252, Riboldi-Tunnicliffe, A., Konig, B., Jessen, S., Weiss, M. S., Rahfeld, J., Hacker, J., Fischer, G. & Hilgenfeld, R. (2001) Nat. Struct. Biol. 8, Das, A. K., Cohen, P. W. & Barford, D. (1998) EMBO J. 17, Taylor, P., Dornan, J., Carrello, A., Minchin, R. F., Ratajczak, T. & Walkinshaw, M. D. (2001) Structure (London) 9, Chambraud, B., Rouviere-Fourmy, N., Radanyi, C., Hsiao, K., Peattie, D. A., Livingston, D. J. & Baulieu, E. E. (1993) Biochem. Biophys. Res. Commun. 196, Pirkl, F., Fischer, E., Modrow, S. & Buchner, J. (2001) J. Biol. Chem. 276, Ward, B. K., Allan, R. K., Mok, D., Temple, S. E., Taylor, P., Dornan, J., Mark, P. J., Shaw, D. J., Kumar, P., Walkinshaw, M. D., et al. (2002) J. Biol. Chem. 277, Russell, L. C., Whitt, S. R., Chen, M. S. & Chinkers, M. (1999) J. Biol. Chem. 274, Miyata, Y., Chambraud, B., Radanyi, C., Leclerc, J., Lebeau, M. C., Renoir, J. M., Shirai, R., Catelli, M. G., Yahara, I. & Baulieu, E. E. (1997) Proc. Natl. Acad. Sci. USA 94, Barent, R. L., Nair, S. C., Carr, D. C., Ruan, Y., Rimerman, R. A., Fulton, J., Zhang, Y. & Smith, D. F. (1998) Mol. Endocrinol. 12, Cheung-Flynn, J., Roberts, P. J., Riggs, D. L. & Smith, D. F. (2003) J. Biol. Chem. 278, BIOPHYSICS Wu et al. PNAS June 1, 2004 vol. 101 no

99 Communications Fluorescence Spectroscopy Donor Donor Energy-Migration Measurements of Dimeric DsbC Labeled at Its N-Terminal Amines with Fluorescent Probes: A Study of Protein Unfolding** Xuejun Duan, Zhen Zhao, Jianping Ye, Huimin Ma,* Andong Xia,* Guoqiang Yang, and Chih-Chen Wang* Fluorescence resonance energy transfer (FRET) is a powerful technique for the determination of distances between two fluorophores. The overall geometry of protein structures [1 4] and the conformational changes of a molecule under different conditions can be studied by this method if appropriate sites of the molecule are labeled with fluorescence donor and acceptor probes. Nevertheless, it is rather difficult to specifically introduce two different fluorophore groups into one molecule, [2] especially into a homodimeric biomacromolecule that has two identical reactive sites. Different from the conventional FRET technique, donor donor energy migration (DDEM) takes advantage of certain fluorescence probes that display an overlap of their absorption and emission spectra and are therefore able to transfer energy between themselves. [2 4] Energy transfer in this case is a reversible process and can be measured through analysis of the timeresolved depolarization of the fluorescence emission (as donor donor energy migration results in additional depolarization). As only one type of probe is required, DDEM simplifies greatly not only the labeling operation but also the theoretical analysis and the time-resolved measurements and has been widely used to study the steady-state conformational changes of biomacromolecules. DsbC (1), a member of the Dsb family in the periplasm of Gram-negative bacteria, is a thiol-protein oxidoreductase that displays molecular chaperone activity. [5 7] The DsbC molecule is a V-shaped homodimer consisting of two 23.4-kDa subunits. [8] Each subunit is composed of a C-terminal thiore- [*] X. Duan, Dr. J. Ye, Prof. Dr. H. Ma, Prof. Dr. A. Xia, Prof. Dr. G. Yang Center for Molecular Sciences, Institute of Chemistry Chinese Academy of Sciences, Beijing (China) Fax: (+ 86) mahm@iccas.ac.cn andong@iccas.ac.cn Z. Zhao, Prof. C.-C. Wang National Laboratory of Biomacromolecules Institute of Biophysics, Chinese Academy of Sciences Beijing (China) Fax: (+ 86) chihwang@sun5.ibp.ac.cn [**] This work was supported by the National Natural Science Foundation of China (Grant no ), the Ministry of Science and Technology of China, and the Chinese Academy of Sciences. We thank Dr. Rudi Glockshuber (Eidgenössische Technische Hochschule, Hönggerberg, Switzerland) for the generous gift of pdsbc plasmid. Supporting information for this article is available on the WWW under or fromthe author Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim DOI: /anie Angew. Chem. Int. Ed. 2004, 43,

100 Angewandte Chemie doxin-like domain and a N-terminal domain, which is responsible for dimerization and is essential for the chaperone activity of the molecule. [8,9] The V-shaped structure of homodimeric DsbC led us to apply the DDEM method to explore its unfolding and dissociation behavior and to understand further its structure function relationship. In this context, the two N-terminal amino groups of DsbC are the sites of choice at which to link two identical probes; however, the labeling of other amino groups, such as e-amino groups of lysine residues, and the nonspecific modification of groups other than amino groups could also occur. Several new methods for the introduction of fluorescent probes into proteins have recently been developed to improve the specificity of labeling. [10 12] The most common approach is to engineer a pair of reactive cysteine residues to provide two thiol handles for conjugation. [2,13] Alternatively, a ketone handle, produced through the introduction of an unnatural keto-containing amino acid, can be labeled with hydrazide-functionalized fluorophores with no [1, 13] observed cross-reactivity. Herein, we describe a general method for the specific labeling of N-terminal groups through a transamination reaction, in which the N-terminal amino group of a protein molecule is converted into a reactive carbonyl group, which is then treated with a hydrazide-containing fluorophore. As the intermediate in transamination reactions involves the participation of an adjacent peptide bond, only the conversion of the terminal amino group occurs without modification of the internal amino groups on lysine residues. [14 17] Subsequently, the conformational changes of dimeric DsbC during unfolding (induced by guanidine hydrochloride (GuHCl)) were studied by DDEM. The fluorescent dye BODIPY FL (shown as the hydrazide derivative 2 in Figure 1) was employed as the probe owing to its high fluorescence quantum yield, its insensitivity to solvent polarity and ph, and its Förster radius of 57 Š. [2,13, 18] The N- terminal amino groups of 1 were modified as shown in Figure 1 by a) a transamination reaction in the presence of glyoxylate and CuSO 4, [14,17] b) coupling of the product 3 with BODIPY FL hydrazide (2), and c) reduction of the imine groups to the more stable amine form 4 of the labeled product. Sodium cyanoborohydride instead of borane-pyridine was used as the reducing agent owing to its better selectivity for imines [19] and the absence of quenching effects on the fluorescence from the BODIPY dye (data not shown). In a similar procedure, 1 was also labeled by following the transamination step carried out in the absence of glyoxylate. As shown in Figure 2, the absorption spectrum of the protein modified in the presence of glyoxylate exhibits a main peak at 280 nm characteristic of native protein and a less intense band at 505 nm for the BODIPY moiety, [13] whereas the absorption profile for the protein modified in the absence of glyoxylate shows only one band for native protein; this indicates that the BODIPY-labeled DsbC 4 can be prepared only through a transamination process carried out in the presence of glyoxylate. To confirm further that the DsbC molecule had been specifically labeled with BODIPY, 4 was also examined by MALDI TOF mass spectrometry. A peak at m/z , in Figure 1. Specific modification of the N-terminal amino groups of dimeric DsbC (1): a) transamination reaction with glyoxylate; b) coupling with 2; c) reduction with sodiumcyanoborohydride. The separation between the central B atomof the BODIPY FL dye and the terminal N atomof the hydrazide group of the linker armin 2 is 7.8 Š (calculated with CS Chem3D software). BODIPY FL = 4,4-difluoro-5,7- dimethyl-4-bora-3a,4a-diaza-s-indacene-3-propionic acid. Figure 2. Absorption spectra of DsbC labeled with BODIPY. DsbC (1) was modified in the presence (solid line, 4, 1.2 mm) and absence (dashed line, 1.0 mm) of glyoxylate. The inset shows the excitation (l em = 535 nm) and emission (l ex = 467 nm) spectra of 4 (4.1 mm); 5-nmexcitation and emission slits were used. agreement with the theoretical value of m/z expected for DsbC with two N-terminal BODIPY labels 4, was detected with a mass error < 0.3%. [20] Although the presence of a small amount of DsbC modified on only one N-terminal amino group cannot be ruled out, it should not affect the DDEM measurements, especially in dilute solution. The efficiency with which fluorophores are incorporated into DsbC is about 9%, which is ascribed to the limited accessibility of the N termini of the DsbC molecule. The fluorescence spectra of 4 display an excitation band at 505 nm and an emission band at 510 nm (see Figure 2 inset), Angew. Chem. Int. Ed. 2004, 43, Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

101 Communications which are almost the same as that for the free BODIPY dye [13] and indicate that the attachment of the fluorophore to DsbC does not alter its spectral properties. On the other hand, 4 shows the same circular dichroism spectrum as that of the native DsbC, [20] which suggests that the introduction of BODIPY does not affect the secondary structure of the protein. The native DsbC, the partially denatured DsbC formed in the presence of GuHCl (1.5 m), and the fully denatured DsbC formed in the presence of GuHCl (6m) and dithiothreitol (0.1m), all labeled with BODIPY, displayed near-identical fluorescence-decay profiles, [20] which were fitted to a single exponential function with a satisfactory low value of c 2 in the range of The fluorescence lifetime (t) in each of the three cases was about 6.7 ns (calculated based on I(t) = Ae ( t/t) ) and indicate that the fluorescence lifetime of BODIPY in BODIPY-labeled DsbC is unaffected by the extent of denaturing of the protein (Table 1).On the other hand, the decay rates of fluorescence anisotropy r(t) show a variation with different extents of denaturing of DsbC (Figure 3). The initial decay of r(t) of the fully denatured DsbC is much slower than that of the native DsbC. The fast decay of the fluorescence anisotropy from the native DsbC suggests that the observed emission is not from the originally excited BODIPY fluorophore. The other adjacent fluorophore in the same DsbC molecule could contribute to the observed emission by an energy-transfer mechanism and thereby lead to the fast depolarization. This is an experimental hallmark of donor donor energy-migration processes. [4] The interfluorophore distance R is defined as the distance between the centers of two fluorophores and can be estimated based on energy-transfer measurements. [2 4,21] The rate w of energy transfer between two interacting fluorophores is expressed by Equation (1) according to the Förster energytransfer [2,3, 21] mechanism: w ¼ 3 2 hk2 i 1 t 6 R0 R (t = fluorescence lifetime, R 0 = Förster radius (57 1Š for BODIPY), [2,18] and hk 2 i = orientation factor, for which an average value of 2 = 3 is usually taken; the parameter w obtained from DDEM measurements and the values of R estimated by Equation (1) are summarized in Table 1). Table 1: Results fromddem measurements [a] DsbC w [ns 1 ] t [ns] R [Š] R c [Š] Native Partially denatured Fully denatured 6.6 [a] The parameter w was obtained fromthe best-fit curves based on r(t) = Aexp( 2wt) + B; the value of c 2 for each fitting was in the range of The interfluorophore distance R in 4 denatured to different extents was calculated according to Equation (1). R c is the corrected value of R. The data quoted are the average of two independent experiments. The calculated interfluorophore distances in the native and partially denatured DsbC are 46 and 58 Š, respectively, and contain a contribution from the length of the linker group of the BODIPY dye (Figure 1). Moreover, it is reasonable to ð1þ Figure 3. a) c) Polarized fluorescence decays of I k (t) and I? (t) and d) f) anisotropy decays (along with the best-fit curves and the weighted residuals) of 4 (4.1 mm) at various extents of denaturing of the DsbC protein; a), d) native DsbC; b), e) denatured in GuHCl (1.5m); c), f) denatured in GuHCl (6m) with dithiothreitol (DTT, 0.1m). I k (t) and I? (t) represent the intensities of the fluorescence observed with the emission polarizer orientated parallel and perpendicular, respectively, to the excitation polarizer. All measurements were carried out at 273 K Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Angew. Chem. Int. Ed. 2004, 43,

102 Angewandte Chemie assume an averaged right-angled geometry between the two linker groups attached to the N termini of DsbC. [21] The corrected value R c for the native DsbC is thus 35 Š (Table 1), which is in agreement with the value of 29 Š measured from the crystal structure of the protein. [8] Similarly, the corrected distance between the two N termini in the partially denatured DsbC is 47 Š, which is markedly longer than 35 Š in the native protein. The much longer distance measured between the two N termini in a partially denatured molecule indicates that the unfolding of DsbC in the presence of GuHCl (1.5m) renders the molecule more loose and flexible but not dissociated (Table 1). The very slow decay of the fluorescence anisotropy of the fully denatured DsbC could only arise from the rotation of the probe molecule together with the fluctuation of the conformation of DsbC rather than from DDEM processes. The interfluorophore distance in this case, which is far longer than the critical distance R 0 of BODIPY and could not be determined by DDEM measurements, implies the dissociation of the dimeric molecule in the fully denatured protein. In summary, we have developed a valuable method, which consists of N-terminal-specific fluorescence labeling through a transamination reaction followed by DDEM measurements, to study the unfolding/folding processes of a dimeric protein. The transamination step provides a general approach for the selective attachment of a fluorophore to N-terminal amino acid residues, and the dimeric structure of DsbC allows the introduction of two identical fluorophores so that the DDEM method can be used to trace its unfolding behavior. This combined strategy is useful to investigate conformational changes of other dimeric proteins under variable conditions. An important development would be to combine the specific labeling method with DDEM measurement at the singlemolecule level. Furthermore, this labeling approach could also be extended to nondimeric protein molecules and would therefore broaden the scope of application of fluorescence spectroscopy. Experimental Section [7, 9] General: DsbC (1) was prepared as reported previously from plasmid pdsbc, which contains the full-length DsbC precursor gene. Glyoxylate was purchased from Acros. BODIPY FL hydrazide was purchased from Molecular Probes, Inc. MALDI TOF mass spectrometry was performed on a Bruker BIFLEX III instrument. 3: DsbC (1; 1 mg) was dissolved in an aqueous solution of sodium acetate (2 ml; 1m, ph 5.5) containing glyoxylate (0.1m) and CuSO 4 (5 mm) and was stirred for 1 h at 296 K. The reaction was quenched through the addition of ethylenediaminetetraacetic acid diammonium salt to a final concentration of 20 mm followed by dialysis against sodium phosphate buffer (0.1m, ph 7.4). 4: BODIPY FL hydrazide (2; 200 ml; 1.96 mm in MeOH) and concentrated HCl (to a final concentration of 0.5m) were consecutively added to 3, and the mixture was stirred for 1 h at 296 K in the dark. Sodium cyanoborohydride (5 equiv relative to the protein; Sigma) was then added and the solution was incubated overnight at 277 K. The mixture was applied onto a Sephadex G-25 column to remove any remaining free BODIPY dye and the excess reducing reagent. The protein fraction 4, which displays an absorbance at both 280 and 505 nm, was collected and then thoroughly dialyzed against phosphate buffer. The efficiency of labeling was calculated from the absorption spectra/molar absorptivities of the fluorescent probe 2 (e = 80000m 1 cm 1 at 505 nm) [2] and the dimeric protein 1 (e = 32340m 1 cm 1 at 280 nm). [22] As a control, the same procedure was performed with DsbC in the absence of glyoxylate. Received: March 22, Revised: May 10, 2004 [Z460072] Keywords: amines analytical methods energy transfer fluorescent probes protein folding [1] S. Weiss, Science 1999, 283, [2] J. Karolin, M. Fa, M. Wilczynska, T. Ny, L. B.-Š. Johansson, Biophys. J. 1998, 74, [3] T. Ikeda, B. Lee, S. Kurihara, S. Tazuke, S. Ito, M. Yamamoto, J. Am. Chem. Soc. 1988, 110, [4] H. Otto, T. Lamparter, B. Borucki, J. Hughes, M. P. Heyn, Biochemistry 2003, 42, [5] D. Missiakas, S. Raina, J. Bacteriol. 1997, 179, [6] D. Missiakas, C. Georgopoulos, S. Raina, EMBO J. 1994, 13, [7] J. Chen, J. L. Song, S. Zhang, Y. Wang, D. F. Cui, C. C. Wang, J. Biol. Chem. 1999, 274, [8] A. A. McCarthy, P. W. Haebel, A. Törrönen, V. Rybin, E. N. Baker, P. Metcalf, Nat. Struct. Biol. 2000, 7, [9] X. X. Sun, C. C. Wang, J. Biol. Chem. 2000, 275, [10] V. W. Cornish, D. R. Benson, C. A. Altenbach, K. Hideg, W. L. Hubbell, P. G. Schultz, Proc. Natl. Acad. Sci. USA 1994, 91, [11] R. L. Lundblad, Chemical Reagents for Protein Modification, 2nd ed., CRC Press, Boca Raton, [12] T. J. Tolbert, C. H. Wong, Angew. Chem. 2002, 114, ; Angew. Chem. Int. Ed. 2002, 41, [13] R. P. Haugland, Handbook of Fluorescent Probes and Research Products, 9th ed., Molecular Probes, Inc., Eugene, 2002, pp [14] P. Wu, L. Brand, Methods Enzymol. 1997, [15] R. Q. He, C. L. Tsou, Biochem. J. 1992, 287, [16] B. Hammack, S. Godbole, B. E. Bowler, J. Mol. Biol. 1998, 275, [17] H. B. F. Dixon, R. Fields, Methods Enzymol. 1972, 25, [18] J. Karolin, L. B.-Š. Johansson, L. Strandberg, T. Ny, J. Am. Chem. Soc. 1994, 116, [19] R. F. Borch, M. D. Bernstein, H. D. Durst, J. Am. Chem. Soc. 1971, 93, [20] See Supporting Information. [21] A. A. Deniz, T. A. Laurence, G. S. Beligere, M. Dahan, A. B. Martin, D. S. Chemla, P. E. Dawson, P. G. Schultz, S. Weiss, Proc. Natl. Acad. Sci. USA 2000, 97, [22] A. Zapun, D. Missiakas, S. Raina, T. E. Creighton, Biochemistry 1995, 34, Angew. Chem. Int. Ed. 2004, 43, Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

103 THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 279, No. 2, Issue of January 9, pp , by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A. Crystal Structure of Human Pirin AN IRON-BINDING NUCLEAR PROTEIN AND TRANSCRIPTION COFACTOR* Received for publication, September 9, 2003, and in revised form, October 13, 2003 Published, JBC Papers in Press, October 22, 2003, DOI /jbc.M Hai Pang, Mark Bartlam, Qinghong Zeng, Hideyuki Miyatake, Tamao Hisano, Kunio Miki, Luet-Lok Wong, George F. Gao **, and Zihe Rao From the Laboratory of Structural Biology, Tsinghua University and National Laboratory of Biomacromolecules, Institute of Biophysics, Chinese Academy of Science, School of Life Sciences and Engineering, Beijing, Beijing , China, the Theoretical Structural Biology Laboratory, RIKEN Harima Institute, SPring-8, Hyogo , Japan, the Department of Chemistry, Inorganic Chemistry Laboratory, South Parks Road, Oxford OX1 3QR, United Kingdom, and the **Nuffield Department of Clinical Medicine, University of Oxford, John Radcliffe Hospital, Headington, Oxford OX3 9DU, United Kingdom Pirin is a newly identified nuclear protein that interacts with the oncoprotein B-cell lymphoma 3-encoded (Bcl-3) and nuclear factor I (NFI). The crystal structure of human Pirin at 2.1-Å resolution shows it to be a member of the functionally diverse cupin superfamily. The structure comprises two -barrel domains, with an Fe(II) cofactor bound within the cavity of the N-terminal domain. These findings suggest an enzymatic role for Pirin, most likely in biological redox reactions involving oxygen, and provide compelling evidence that Pirin requires the participation of the metal ion for its interaction with Bcl-3 to co-regulate the NF- B transcription pathway and the interaction with NFI in DNA replication. Substitution of iron by heavy metals thus provides a novel pathway for these metals to directly influence gene transcription. The structure suggests an interesting new role of iron in biology and that Pirin may be involved in novel mechanisms of gene regulation. Pirin is a newly identified nuclear protein that is widely expressed in dot-like subnuclear structures in human tissues, in particular liver and heart (1). Pirins are highly conserved among mammals, plants, fungi, and even prokaryotic organisms and have been assigned as a sub-family of the cupin superfamily based on both structure and sequence homology (1, 2). The cupin superfamily is among the most functionally diverse of any protein family described to date, with both enzymatic and non-enzymatic members included (2). This cupin superfamily is grouped according to a conserved -barrel fold and two characteristic sequence motifs. Study of Pirin reveals two cupin domains from its primary sequence that are consistent with other members of the superfamily. The exact functions of Pirins are not yet known. No enzymatic activity has been described, but human Pirin has been found to bind to the nuclear factor I/CCAAT box transcription * This work was supported by Grants G (Project 973) and 2002BA711A12 (Ministry of Science and Technology, China) for the Human Structural Genomics Initiative. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. The atomic coordinates and structure factors (code 1J1L) have been deposited in the Protein Data Bank, Research Collaboratory for Structural Bioinformatics, Rutgers University, New Brunswick, NJ ( These authors contributed equally to this work. To whom correspondence should be addressed. Tel.: ; Fax: ; raozh@xtal.tsinghua.edu.cn. This paper is available on line at factor (NFI) 1 and to the oncoprotein B cell lymphoma 3-encoded (Bcl-3) in vivo (1, 3), suggesting that it is a transcription cofactor. NFI is known to stimulate DNA replication and RNA polymerase II-driven transcription (4). Bcl-3 is a distinctive member of the I B family, which inhibits the transcription factor NF- B by preventing NF- B nuclear translocation and DNA binding. However, there is also evidence that Bcl-3 preferentially binds to NF- B p50 or p52 homodimers to stimulate transcription (5). The functional nature of this difference between I B (inhibiting) and Bcl-3 (enhancing) is not known, but it is clear that they bind to different protein partners. Pirin is one of four binding partners of Bcl-3, together with Bard1, Tip60, and Jab1, and can be sequestered into quaternary complexes with Bcl-3, p50, and DNA (3). These four Bcl-3-interacting protein partners, which do not share any sequence homology, might play some crucial role in regulating the function of Bcl-3 and I B. Indeed, both Pirin and Bcl-3 are localized in the nucleus, and the potential roles of Pirin in NF- B-dependent transcriptional regulation are implicated in a number of experiments (6 10). Here we report the crystal structure of human Pirin to 2.1-Å resolution. Understanding of the Pirin structure is of critical importance in elucidating and understanding its function, in particular its interaction with the I B family of proteins in its role as a transcription cofactor. EXPERIMENTAL PROCEDURES Cloning, Expression, Purification, and Crystallization The complete gene fragment encoding human Pirin protein was subcloned into pet-28a expression vector from a human liver cdna library, and the human Pirin was highly expressed as a soluble protein in Escherichia coli strain BL21(DE3) with a 6-residue His tag attached to its N terminus. Purification of the Pirin protein was carried out through an affinity chromatography Co-NTA His Bind column (Qiagen) followed by size-exclusion chromatography on a Superdex 75 column (Amersham Biosciences). Crystals of Pirin were grown using the hanging dropvapor diffusion method from a solution containing 25 mg/ml Pirin in 14% polyethylene glycol 20000, 0.1 M MES, ph 6.5, at 16 C. For phase determination, a selenomethionyl derivative was produced and crystallized under similar conditions. The Pirin crystals belong to the space group P with unit cell parameters a 42.28, b 67.12, c Å, 90. Structure Determination Data sets were collected at three wavelengths from a single selenomethionine derivative crystal at 100 K on 1 The abbreviations used are: NFI, nuclear factor I; Bcl-3, B-cell lymphoma 3; MES, 4-morpholineethanesulfonic acid; AAS, atomic absorption spectrophotometry; r.m.s., root mean square; arac, 1- -D-arabinofuranosylcytosine; Ni-ARD, nickel-binding dioxygenase, acireductone dioxygenase; 2-OG, 2-oxoglutarate.

104 1492 Crystal Structure of Human Pirin TABLE I Data collection and refinement statistics The numbers in parentheses correspond to the highest resolution shell. beamline BL44B2 of SPring-8. Data were collected to 2.1-Å resolution using energies corresponding to the peak ( Å) and edge ( Å) of the experimentally determined selenium K-edge, and a low energy remote wavelength ( Å). All processing, scaling, and merging of datasets were performed using the HKL2000 package (11). Initial phases were calculated to 2.5-Å resolution with SOLVE (12) from seven heavy atom sites, and RESOLVE (13) was used for density modification and phase extension to 2.1 Å. The Pirin model was built using ARP/wARP (14) and O (15) and was subsequently refined with CNS (16) using the low energy remote data. An initial round of simulated annealing was followed by alternate cycles of manual rebuilding and minimization. The final model contains 288 residues, 188 water molecules, and an Fe 2 ion. The position of the metal ion was evident as a peak greater than 16 in the F o F c map. The Fe 2 ion was refined at full occupancy with a temperature factor of 17.5 Å 2. The coordinating atoms His 56 N 2, His 58 N 2, His 101 N 2, and Glu 103 O 2 have temperature factors of 10.8, 18.1, 12.9, and 13.8 Å 2, respectively. Identification of the Metal Ion The metal ion bound in the N-terminal domain of Pirin was confirmed to be iron by atomic absorption spectrophotometry (AAS). The AAS experiments were performed in the Tsinghua Analysis Center using a Carl Zeiss Technology Analytic Jena AAS 6 Vario instrument. The presence of the metal ions Mg 2,Ca 2, Mn 2,Fe 2,Co 2,Ni 2,Cu 2, and Zn 2 were analyzed. The results indicated that, after Co 2 -affinity chromatography, only Fe 2 remained in the protein solution after gel-filtration chromatography on a Superdex-75 column (Amersham Biosciences) and equilibration in 20 mm Tris-HCl (ph 8.0) and 150 mm NaCl. RESULTS AND DISCUSSION Structure Determination The structure of human Pirin was determined using the multiple-wavelength anomalous dispersion method from a single selenomethionine derivative crystal. Details of the data collection and structure refinement are given in Table I. The asymmetric unit of the Pirin crystal contains one monomer. The electron density map is of high quality such that 288 out of 290 residues could be built. No density was observed for the two N-terminal residues, but the remainder of the Pirin molecule, including the C-terminal end of the polypeptide chain, is well defined (Fig. 1a). Two cisproline residues could be identified in the structure (Pro 31 and Pro 50 ). The final model of the structure contains 288 residues, 188 water molecules, and 1 Fe 2 ion. The presence of iron was established by atomic absorption spectrophotometry, and quantitative measurements showed that there is 1 mol of Fe 2 per mol of protein. MAD peak MAD edge MAD remote Wavelength (Å) Resolution limit (Å) ( ) ( ) ( ) Total reflections 95,485 95,271 94,836 Unique reflections 30,975 30,927 30,894 Completeness 90.5 (81.5) 90.4 (81.4) 89.9 (79.8) a R merge 5.0 (27.5) 5.0 (28.8) 5.0 (29.3) I/ (I) 7.3 (1.9) 7.0 (1.8) 6.8 (1.8) Resolution range b R work 20.6 b R free 24.9 Number of protein atoms 2,244 Number of water molecules 187 r.m.s. deviation Bonds (Å) Angles ( ) 1.35 Average B factor (Å 2 ) Protein Water Ramachandran plot Favored (%) 88.6 Allowed (%) 11.4 a R merge h l I ih I h h I I h, where I h is the mean of the observations I ih of reflection h. b R work ( F p (obs) F p (calc) )/ F p (obs) ; R free R factor for a selected subset (5%) of the reflections that was not included in prior refinement calculations. 104 Structural Overview Pirin is composed of two structurally similar domains arranged face to face (Fig. 1, b and c). The core of each domain comprises two antiparallel -sheets, with eight strands forming a -sandwich. The fold of each domain is very similar, and the two domains can be superimposed with an r.m.s. difference of 1.3 Å for 64 equivalent residues. The N- terminal domain (residues 3 134) additionally features four -strands, and the C-terminal domain (residues ) also includes four additional -strands and a short -helix. The two domains are cross-linked, with 1 forming part of one sheet of the C-terminal domain, and strands 25 and 26 forming an extension of one sheet of the N-terminal domain. The C-terminal -helix packs against the outer surface of the N-terminal domain -barrel. The two domains are joined by a short linker of 10 amino acids (residues ) that contains a single turn of a 3 10 helix. Four additional 3 10 helices are located in the structure. Similar cavities are found in each domain. In the N-terminal domain, the cavity contains a metal binding site with a single Fe 2 ion located at about 6 Å from the protein surface (Fig. 1d). The C-terminal domain cavity is more compact than the N- terminal domain and is closed by the -strand formed by residues The C-terminal domain does not contain any metal binding site. Pirin Is a Member of the Cupin Superfamily It has been predicted that Pirin belongs to the cupin superfamily on the basis of primary sequence (2). From a Dali (17) search for structural similarity, the Pirin structure was confirmed to closely resemble members of the cupin superfamily (Fig. 2, a f), particularly the structures of quercetin 2,3-dioxygenase (18), glycinin g1 (19), and phosphomannose isomerase (20). As with Pirin, these three proteins are bicupins with two germin-like -barrel domains. Pirin also contains the two characteristic sequences of the cupin superfamily, namely PG-(X) 5 -HXH-(X) 4 -E-(X) 6 -G and G-(X) 5 -PXG- (X) 2 -H-(X) 3 -N separated by a variable stretch of amino acids (Fig. 2g). These motifs are best conserved in the N- terminal domain (residues and residues ) where the conserved histidine and glutamic acid residues correspond to the metal-coordinating residues. The C-terminal domain motifs (residues and residues ) lack the metal binding residues normally associated with the cupin fold.

105 Crystal Structure of Human Pirin 1493 FIG. 1. The Pirin structure. a, a stereo diagram of the Pirin structure shown asac trace. Every 20th residue is labeled. The top view (b) and side view (c) of the Pirin structure are shown. The structure is colored from N-terminal (blue) to C-terminal (red). The metal ion is shown as a large gray sphere, and the coordinating groups and water molecules are shown as ball-and-stick models. d, a stereo diagram showing the electron density map in the metal binding site. The metal ion is coordinated by three histidines, one glutamate, and two water molecules. The omit map is contoured at The cupin superfamily has possibly the widest range of biochemical functions of any superfamily identified to date (2). It is comprised of both enzymatic and non-enzymatic members, which have either one or two cupin domains. The cupin fold comprises a motif of six to eight antiparallel -strands located within a conserved -barrel structure (2). The variety of biochemical functions is reflected by the low sequence homology shared among the cupin superfamily members (Fig. 2g). A BLAST search for proteins homologous to human Pirin revealed significant conservation between mammals, plants, and prokaryotes, particularly within the N-terminal domain (1). Interestingly, sequence alignment of human Pirin and related proteins reveals two clusters that are highly conserved throughout all aligned sequences. These sequence clusters, corresponding to residues (cluster 1) and (cluster 2), include the four metal coordinating residues of human Pirin, which are strictly conserved among all aligned sequences. Pirin (Fig. 2a) and the bicupin metalloenzymes quercetin 2,3-dioxygenase (Fig. 2b) and phosphomannose isomerase (Fig. 2d) show similarities in the overall fold. Quercetin 2,3-dioxygenase is a copper-containing enzyme, and its N-terminal domain can be superimposed onto the N-terminal domain of Pirin with an r.m.s. difference of 1.5 Å for 84 equivalent residues. The Cu-binding site of quercetin 2,3-dioxygenase, formed by 3 histidines, a glutamic acid residue and a single water molecule, matches the metal site of Pirin. The three histidines of Pirin (His 56, His 58, and His 101 ) are structurally equivalent to those of quercetin 2,3-dioxygenase (His 66, His 68, and His 112 ). Similarly, the phosphomannose isomerase structure can be superimposed onto that of Pirin with an r.m.s. difference of 1.3 Å for 71 equivalent residues. Phosphomannose isomerase is a zinc-containing enzyme, and its metal binding site, comprising two histidines and two glutamic acid residues, also matches the metal site of Pirin. Regardless of the overall structural similarity between its two domains, the C-terminal domain of Pirin shows interesting differences from the N-terminal domain. Notably, the C-terminal domain of Pirin does not contain a metal binding site and its sequence does not contain the conserved metal-coordinating residues. Several members of the cupin superfamily do not contain the metal-coordinating residues, including the transcription factor arac from Escherichia coli (21). arac is a single

106 1494 Crystal Structure of Human Pirin FIG. 2. A comparison between Pirin and similar structures. a, Pirin structure; b, quercetin 2,3-dioxygenase structure (PDB code: 1JUH); c, glycinin g1 structure (PDB code: 1FXZ); d, phosphomannose isomerase structure (PDB code: 1PMI); e, oxalate oxidase (germin) structure (PDB code: 1FI2); f, arac dimer structure (PDB code: 2ARC). g, structure-based sequence alignment of Pirin with related structures. PIRIN-N, N-terminal domain of Pirin; PIRIN-C, C-terminal domain of Pirin; 2,3QD-N, N-terminal domain of quercetin 2,3-dioxygenase; 2,3QD-C, C-terminal 106

107 Crystal Structure of Human Pirin 1495 FIG. 3. The metal binding site. Metal binding sites of Pirin (a), quercetin 2,3-dioxygenase (PDB code: 1JUH, b), phosphomannose isomerase (PDB code: 1PMI, c), and oxalate oxidase (PDB code: 1FI2, d). domain member of the cupin family and binds arabinose for activation of transcription. There are notable similarities between the C-terminal domain of Pirin and arac. In arac, the arabinose binding site is located within the -barrel and is closed by an N-terminal domain arm. The -barrel of the C- terminal domain of Pirin is also closed by the N-terminal 1 strand. However, in the absence of arabinose, the N-terminal arm of arac becomes disordered, suggesting that it is important for ligand stability. A detailed analysis of the arabinose binding residues in the arac structure shows that the C-terminal domain of Pirin is unlikely to bind arabinose. The Metal Binding Site Surprisingly, a single Fe 2 ion is located in the Pirin N-terminal domain where it is coordinated by 3 histidine residues (His 56, His 58, and His 101 ) through their N 2 atoms, and one glutamic acid (Glu 103 ) through the O 2 atom (Fig. 3a). The metal binding site is exposed to the solvent, and the octahedral coordination environment is completed by two water molecules, with metal-ligand distances of 2.24 and 2.15 Å (Table II). A structure-based sequence alignment shows that the metal binding motif is highly conserved within a number of other cupin superfamily members (Fig. 2g). The metal binding domain of Pirin is most structurally similar to that in germin, a Mn 2 -containing oxalate oxidase in which the metal is also octahedrally coordinated by three histidines, a glutamic acid and two water molecules (22). As with quercetin 2,3-dioxygenase, the three histidines in oxalate oxidase (His 88, His 90, and His 137 ) are structurally equivalent to those in Pirin (His 56, His 58, and His 101 ). One notable difference between Pirin and related metalloproteins of the cupin superfamily is the location of the coordinating Glu residue. Quercetin 2,3-dioxygenase (Fig. 3b), phosphomannose isomerase (Fig. 3c), and oxalate oxidase (Fig. 3d) all contain a conserved Glu residue in the equivalent structural position to residue 63 of Pirin. However, this residue is not conserved in Pirin and the coordinating Glu residue is instead located in position 103 on -strand 10. The structure of a nickel-binding dioxygenase, acireductone domain of quercetin 2,3-dioxygenase; PROGLYCININ, glycinin g1; PMI-N, N-terminal domain of phosphomannose isomerase; PMI-C, C-terminal domain of phosphomannose isomerase; OXALATE, oxalate oxidase (germin); ARAC, arac. Only the regions adjacent to the two characteristic conserved cupin sequences are shown. The two conserved cupin sequence motifs are shaded in gray, and metal coordinating residues are shown in green boxes. The secondary structure elements relate to the Pirin N-terminal domain structure. 107

108 1496 Crystal Structure of Human Pirin TABLE II Metal-donor atom distances and temperature factors Distance to metal ion Temperature factor Å Å 2 Fe His 56 N His 58 N His 101 N Glu 103 O Water Water FIG. 4.A model of the Pirin Bcl-3 (p50) 2 complex (A and B). The ankyrin repeat domain of Bcl-3 is shown as a ribbon model colored red. The Pirin structure is shown as a ribbon model with the N-terminal domain colored blue and the C-terminal domain colored green. The (p50) 2 homodimer is shown as a ribbon model with one p50 molecule colored orange and the other colored yellow. dioxygenase (Ni-ARD) from Klebsiella pneumoniae, was recently determined by NMR and its active site modeled by comparative homology modeling (23). It was suggested that ARD might also be a member of the cupin superfamily. The Ni 2 ion in ARD was predicted to be coordinated by six ligands, namely three histidine residues, a glutamic acid, and two water molecules. The spatial arrangement of ligand groups in the Ni-ARD model is similar to that of Pirin and germin, with an average distance of 2.1 Å between the metal and donor atoms. Potential Functions of Pirin The bound metal ion may play an important role in Pirin function by stabilizing the N-terminal cupin domain structure and/or by imparting enzymatic activity to the protein. The iron found in the structure is the metal cofactor in numerous enzymes spanning a wide spectrum of activities. Metal-dependent Transcriptional Regulation Pirins are putative transcription cofactors. Heavy metals are known to play an important role in the transcriptional regulation of eukaryotic and prokaryotic genes (24 29). Pirin is newly identified in this study to be a metal-binding protein, and, interestingly, the metal-binding residues of Pirins are highly conserved across mammals, plants, fungi, and prokaryotic organisms. Pirin acts as a cofactor for the transcription factor NFI, the regulatory mechanism of which is generally believed to require the assistance of a metal ion (30). Our structural data support the hypothesis that the bound iron of Pirin may participate in this transcriptional regulation by enhancing and stabilizing the formation of the p50 Bcl-3 DNA complex. Metals have been implicated directly or indirectly in the NF- B family of transcription factors that control expression of a number of early response genes associated with inflammatory responses, cell growth, cell cycle progression, and neoplastic transformation (30). However, most metal-dependent transcription factors are DNA-binding proteins that bind to specific sequences when the metal binds to the protein. Pirin, on the other hand, appears to function differently and bind to the transcription factor DNA complex. The His 3 Glu ligand environment and octahedral geometry of the metal-binding site in Pirin are well suited for the binding of heavy metal ions. The Fe(II) cofactors in non-heme iron proteins such as Pirin are labile, and therefore substitution of iron in Pirin by other metal ions offers a novel mechanism by which heavy metals can directly interfere with gene transcription. Potential Enzymatic Activity of Pirins The Fe 2 binding site in Pirin closely resembles that found in germin, the archetypal cupin. Germin is a manganese-containing protein that has been shown to be an oxalate oxidase and superoxide dismutase (22, 31). We found that Pirin does not possess either superoxide dismutase or oxalate oxidase activity, but we cannot rule out oxidase activity with other substrates. Recently, a number of 2-oxoglutarate (2-OG)-dependent iron monooxygenases, including clavaminic acid synthase 1 (32), hypoxia-inducible factor (33, 34), and factor-inhibiting hypoxia-inducible factor, were suggested to be members of the cupin superfamily on the basis of sequence and structure homology (35). Pirin is structurally related to these enzymes, but 2-OG-dependent monooxygenases have a highly conserved HX(D/E)... H motif, with the two histidines and one aspartate or glutamate forming a facial triad, leaving three coordination sites on the octahedral Fe(II) center for the binding of 2-OG and oxygen. Because the metal binding site in Pirins has a highly conserved His 3 Glu set of ligands, with only two potentially vacant coordination sites on the Fe(II), Pirins are unlikely to be 2-OG-dependent monooxygenases. The other members of the cupin superfamily thought to contain iron are dioxygenases, and a phylogenetic analysis of cupins placed Pirins in the same clade as cysteine dioxygenases (2). Dioxygenases require electron transfer cofactor proteins, and the genes encoding the various proteins of a dioxygenase system are generally grouped together into an operon. Although no open reading frames encoding such cofactor proteins have been found on either the 5 or 3 sides of the human Pirin gene (1), this aspect warrants further investigation. Model of a Pirin Bcl-3 Complex The oncoprotein Bcl-3 is a distinctive member of the I B family of NF- B inhibitors and is located predominantly in the nucleus. It has the properties of a transcriptional co-activator and acts as a bridging factor between NF- B/Rel and nuclear co-regulators. Pirin is known to be one of several binding partners that can associate with Bcl-3 (3) to form part of a larger quaternary complex on NF- B DNA binding sites. Four binding partners of Bcl-3 have so far been identified, namely Pirin, Jab1, Bard1, and Tip60. There are no apparent shared sequence motifs in these four cofactors re-

109 quired for interaction with Bcl-3, and the only common properties between them is that they are nuclear proteins and associate with gene regulators. The structure of the Bcl-3 ankyrin repeat domain (ARD) is elongated and is comprised of seven ankyrin (ANK) repeats (36). Its central ankyrin repeats are similar to those of I B, with the key difference that Bcl-3 has a seventh ankyrin repeat at the C terminus in place of the PEST region of I B. Mutagenesis studies of Bcl-3 indicated that all seven ankyrin repeats are required for binary interaction with Pirin, whereas interactions with Jab1 or Bard1 require only five repeats (3). Pirin and Bcl-3 can also be sequestered into quaternary complexes with p50 and DNA, and Pirin is known to increase the DNA binding by Bcl-3 p50. As with Pirin, all seven ankyrin repeats are required for the Bcl-3 p50 interaction (3). The stoichiometry of the complex is not known, but Bcl-3 is able to recognize at least two proteins simultaneously (3). Because the structures of I B /NF- B (p65/p50 heterodimer) (37, 38) complex and NF- B p50 homodimer (39) are known, and to understand the interaction between Pirin, Bcl-3, and p50, we modeled the complex of these two proteins using protein-protein docking techniques (Fig. 4, A and B). A Bcl-3 (p50) 2 complex was constructed using the structures of the I B NF- B complex and the ankyrin repeat domain (ARD) of Bcl-3. The Bcl-3 ARD structure and the (p50) 2 homodimer were superimposed onto the I B NF- B complex. Pirin was then docked to the Bcl-3 (p50) 2 complex using the program FTDOCK (40). The surface of Pirin has a large acidic patch on the N-terminal domain surface formed by residues 77 82, , and We suggest that this acidic patch could interact with the large basic patch on ankyrin repeats 6 and 7 of Bcl-3. Additional contacts to ankyrin repeat 5 of Bcl-3 could be formed by residues of Pirin. Notably, the previous observation that Pirin binding to Bcl-3 requires all seven repeats of the ARD was based on C-terminal ARD deletion ( ARD6 7 or ARD5 7) mutants, therefore, ARD4 7 of Bcl-3 might be the only domain interacting with Pirin (3), which is consistent with our model. Further analysis involving N-terminal ARD deletion mutants should provide further insight into the Pirin-Bcl-3 recognition interactions. In our model of the complex, residues of the C- terminal arm of Pirin could make contacts with one of the p50 molecules. One possibility is that Bcl-3 binding to the (p50) 2 homodimer could induce conformational changes that lead to Pirin binding to p50, although gel retardation experiments showed that there is no direct binding between Pirin and p50 (3). Although we emphasize that this model is only hypothetical, the extensive contacts formed by Pirin with Bcl-3 and p50 may explain why Pirin can act as a transcription cofactor and enhance the association of Bcl-3 with p50 homodimers. Bcl-3 has a strong preference for p50 and p52 homodimers (3), and, under certain conditions, two molecules of Bcl-3 have been observed to bind to a single p50 or p52 homodimer (41). The addition of a second Bcl-3 molecule in our model, related to the first Bcl-3 molecule using the pseudodyad symmetry of the p50 homodimer, shows that the two Bcl-3 molecules have highly complementary hydrophobic contacts in the ankyrin repeats 1 and 2 (Fig. 4, A and B). I B, on the other hand, binds preferentially to p50-p65 heterodimers, and it is clear that the C-terminal of p65, containing the nuclear localization signal, would sterically hinder the binding of a second I B molecule. Our model gives some clues as to why Bcl-3 preferentially binds p50 or p52 homodimers and how p50 or p52 homodimers interact with Bcl-3. Biological Implications The structure of human Pirin represents the first crystal structure of the Pirin family and is an Crystal Structure of Human Pirin important step toward understanding the function of these proteins. Our work shows that human Pirin is a monomer with two cross-linked -barrel domains. Pirin is confirmed to be a member of the cupin superfamily on the basis of primary sequence and structural similarity. The presence of a metal binding site in the N-terminal -barrel of Pirin, which is highly conserved among Pirins, may be significant in its role in regulating NFI DNA replication and NF- B transcription factor activity. Fe(II) was surprisingly found to be bound to Pirin and could impart enzymatic activity to the protein. Substitution of iron by other heavy metals also offers a direct mechanism for heavy metals to influence gene transcription. This novel pathway may be relevant in the toxicity and other effects of these metals. Finally, the structure of Pirin strongly suggests that a new role of iron in biology, namely regulating DNA replication and gene transcription at the level of the DNA complexes, should be added to the list of the many vital functions of this metal in living organisms. Acknowledgments We are grateful to Feng Gao, Fei Sun, and Zhiyong Lou from the Rao laboratory for assistance, and to Rongguang Zhang and Andrzej Joachimiak from the Advanced Photon Source (Argonne National Laboratory) for help with data collection. We thank Zhi Xing of the Tsinghua Analysis Center for help with atomic absorption spectrophotometry. Critical discussions with Neil Isaacs (Glasgow University, UK) and David Stuart (Oxford University, UK) are greatly appreciated. REFERENCES 1. Wendler, W. M., Kremmer, E., Forster, R., and Winnacker, E. L. (1997) J. Biol. Chem. 272, Dunwell, J. M., Culham, A., Carter, C. E., Sosa-Aguirre, C. R., and Goodenough, P. W. (2001) Trends Biochem. Science 26, Dechend, R., Hirano, F., Lehmann, K., Heissmeyer, V., Ansieau, S., Wulczyn, F. G., Scheidereit, C., and Leutz, A. (1999) Oncogene 18, Santoro, C., Mermod, N., Andrews, P. C., and Tjian, R. (1998) Nature 334, Lenardo, M., and Siebenlist, U. (1994) Immunol. Today 15, Schwarz, E. M., Krimpenfort, P., Berns, A., and Verma, I. M. (1997) Genes Dev. 11, Richard, M., Louahed, J., Demoulin, J.-B., and Renauld, J. C. (1999) Blood 93, Bergman, A. C., Alaiya, A. A., Wendler, W., Binetruy, B., Shoshan, M., Sakaguchi, K., Bergman, T., Kronenwett, U., Auer, G., Appella, E., Jornvall, H., and Linder, S. (1999) Cell Mol. Life Sci. 55, Orzaez, D., de Jong, A. J., and Woltering, E. J. (2001) Plant Mol. Biol. 46, Mitchell, T. C., Hildeman, D., Kedl, R. M., Teague, T. K., Schaefer, B. C., White, J., Zhu, Y., Kappler, J., and Marrack, P. (2001) Nat. Immunol. 2, Otwinowski, Z., and Minor, W. (1997) Methods Enzymol. 276, Terwilliger, T. C., and Berendzen, J. (1999) Acta Crystallogr. Sec. D 55, Terwilliger, T. C. (2000) Acta Crystallogr. Sec. D 56, Perrakis, A., Morris, R., and Lamzin, V. S. (1999) Nat. Struct. Biol. 6, Jones, T. A., Zou, J. Y., Cowan, S. W., and Kjeldgaard, M. (1991) Acta Crystallogr. Sect. A 47, Brunger, A. T., Adams, P. D., Clore, G. M., DeLano, W. L., Gros, P., Grosse- Kunstleve, R. W., Jiang, J. S., Kuszewski, J., Nilges, M., Pannu, N. S., Read, R. J., Rice, L. M., Simonson, T., and Warren, G. L. (1998) Acta Crystallogr. Sec. D 54, Holm, L., and Sander, C. (1998) Nucleic Acids Res. 26, Fusetti, F., Schroeter, K. H., Steiner, R. A., Van Noort, P. I., Pijning, T., Rozeboom, H. J., Kalk, K. H., Egmond, M. R., and Dijkstra, B. W. (2002) Structure 10, Adachi, M., Takenaka, Y., Gidamis, A. B., Mikami, B., and Utsumi, S. (2000) J. Mol. Biol. 305, Cleasby, A., Wonacott, A., Skarzynski, T., Hubbard, R. E., Davies, G. J., Proudfoot, A. E., Bernard, A. R., Payton, M. A., and Wells, T. N. (1996) Nat. Struct. Biol. 3, Soisson, S. M., MacDougall-Shackleton, B., Schleif, R., and Wolberger, C. (1997) Science 276, Anand, R., Dorrestien, P. C., Kinsland, C., Begley, T. P., and Ealick, S. E. (2002) Biochemistry 41, Pochapsky, T. C., Pochapsky, S. S., Ju, T., Mo, H., Al-Mjeni, F., and Maroney, M. J. (2002) Nat. Struct. Biol. 9, DeMoor, J. M., and Koropatnick, D. J. (2000) Cell Mol. Biol. (Noisy-le-grand) 46, Makarova, K. S., Aravind, L., and Koonin, E. V. (2002) Trends Biochem. Sci. 27, Molkentin, J. D. (2000) J. Biol. Chem. 275, O Halloran, T. V. (1993) Science 261, Fujita, M., Yamada, C., Goto, H., Yokoyama, N., Kuzushima, K., Inagaki, M., and Tsurumi, T. (1999) J. Biol. Chem. 274, Liang, R., Igarashi, H., Tsuzuki, T., Nakabeppu, Y., Sekiguchi, M., Kasprzak,

110 1498 Crystal Structure of Human Pirin K. S., and Shiao, Y. H. (2001) Ann. Clin. Lab. Sci. 31, Chen, F., and Shi, X. (2002) Environ. Health Perspect. 110, Lane, B. G., Dunwell, J. M., Ray, J. A., Schmitt, M. R., and Cuming, A. C. (1993) J. Biol. Chem. 268, Zhang, Z., Ren, J., Stammers, D. K., Baldwin, J. E., Harlos, K., and Schofield, C. J. (2000) Nat. Struct. Biol. 7, Ivan, M., Kondo, K., Yang, H., Kim, W., Valiando, J., Ohh, M., Salic, A., Asara, J. M., Lane, W. S., and Kaelin, W. G., Jr. (2001) Science 292, Jaakkola, P., Mole, D. R., Tian, Y. M., Wilson, M. I., Gielbert, J., Gaskell, S. J., Kriegsheim, A., Hebestreit, H. F., Mukherji, M., Schofield, C. J., Maxwell, P. H., Pugh, C. W., and Ratcliffe, P. J. (2001) Science 292, Hewitson, K. S., McNeill, L. A., Riordan, M. V., Tian, Y. M., Bullock, A. N., Welford, R. W., Elkins, J. M., Oldham, N. J., Bhattacharya, S., Gleadle, J. M., Ratcliffe, P. J., Pugh, C. W., and Schofield, C. J. (2002) J. Biol. Chem. 277, Michel, F., Soler-Lopez, M., Petosa, C., Cramer, P., Siebenlist, U., and Muller, C. W. (2001) EMBO J. 20, Huxford, T., Huang, D. B., Malek, S., and Ghosh, G. (1998) Cell 95, Jacobs, M. D., and Harrison, S. C. (1998) Cell 95, Muller, C. W., Rey, F. A., Sodeoka, M., Verdine, G. L., and Harrison, S. C. (1995) Nature 373, Moont, G., Gabb, H. A., and Sternberg, M. J. E. (1999) Proteins 35, Bundy, D. L., and McKeithan, T. W. (1997) J. Biol. Chem. 272,

111 THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 279, No. 6, Issue of February 6, pp , by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A. Structural Basis for the Specific Recognition of RET by the Dok1 Phosphotyrosine Binding Domain* Received for publication, October 7, 2003, and in revised form, November 5, 2003 Published, JBC Papers in Press, November 7, 2003, DOI /jbc.M Ning Shi, Sheng Ye, Mark Bartlam, Maojun Yang, Jing Wu, Yiwei Liu, Fei Sun, Xueqing Han, Xiaozhong Peng, Boqing Qiang, Jiangang Yuan, and Zihe Rao From the Laboratory of Structural Biology, Tsinghua University and National Laboratory of Biomacromolecules, Institute of Biophysics, Chinese Academy of Science, Beijing and the National Laboratory of Medical Molecular Biology, Institute of Basic Medical Sciences, Chinese Academy of Medical Sciences, Peking Union Medical College, National Human Genome Center, Beijing , China Dok1 is a common substrate of activated protein-tyrosine kinases. It is rapidly tyrosine-phosphorylated in response to receptor tyrosine activation and interacts with ras GTPase-activating protein and Nck, leading to inhibition of ras signaling pathway activation and the c-jun N-terminal kinase (JNK) and c-jun activation, respectively. In chronic myelogenous leukemia cells, it has shown constitutive phosphorylation. The N-terminal phosphotyrosine binding (PTB) domain of Dok1 can recognize and bind specifically to phosphotyrosine-containing motifs of receptors. Here we report the crystal structure of the Dok1 PTB domain alone and in complex with a phosphopeptide derived from RET receptor tyrosine kinase. The structure consists of a -sandwich composed of two nearly orthogonal, 7-stranded, antiparallel -sheets, and it is capped at one side by a C-terminal -helix. The RET phosphopeptide binds to Dok1 via a surface groove formed between strand 5 and the C- terminal -helix of the PTB domain. The structures reveal the molecular basis for the specific recognition of RET by the Dok1 PTB domain. We also show that Dok1 does not recognize peptide sequences from TrkA and IL-4, which are recognized by Shc and IRS1, respectively. Protein-protein interactions play key roles in signal transduction. These interactions are often mediated by adapter proteins, which simultaneously associate with several kinases of a signaling pathway, forming an ordered module that permits sequential activation of each enzyme and by anchoring proteins, which are tethered to subcellular structures and localize their complement of enzymes close to their site of action. The docking protein Dok1 appears to function as an adapter. It has * This work was supported by National Sciences Foundation of China (Grants and ); the National Program for Key Basic Research Project ( 973 Grants G and G ); and the National High Technology Research and Development Program ( 863 Grant 2002BA711A12). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. The atomic coordinates and structure factors (code 1P5T) have been deposited in the Protein Data Bank, Research Collaboratory for Structural Bioinformatics, Rutgers University, New Brunswick, NJ ( To whom correspondence may be addressed. Tel.: ; Fax: ; yuanjiangang@pumc.edu.cn. To whom correspondence may be addressed: Laboratory of Structural Biology, School of Life Science and Engineering, Tsinghua University, Beijing , P. R. China. Tel.: ; Fax: ; raozh@xtal.tsinghua.edu.cn been identified as the highly phosphorylated 62-kDa protein that interacts with ras GTPase-activating protein in chronic myelogenous leukemia progenitor cells and v-abl-transformed preb cells (1, 2). The expression of v-abl or the chimeric protein p210bcr-abl in chronic myelogenous leukemia cells has been shown to lead to constitutive Dok1 phosphorylation (1, 2). Recent studies have shown that Dok1 is a common substrate of activated protein-tyrosine kinases such as v-abl (2), v-src (3), BCR (4), EphRs (5), RET (6), and integrin (7). It is rapidly tyrosine-phosphorylated in response to receptor tyrosine activation in various cell systems. Dok1 contains an N-terminal pleckstrin homology (PH) 1 domain followed by a central phosphotyrosine binding (PTB) domain and a proline- and tyrosine-rich C-terminal tail. The PH domain is known to bind to acidic phospholids and localize proteins to the plasma membrane, whereas the PTB domain is known to mediate protein-protein interactions by binding to phosphotyrosine-containing motifs (8). The C-terminal part of Dok1 contains multiple tyrosine phosphorylation sites. When phosphorylated, they become potential docking sites for Src homology 2-containing proteins such as ras GTPase-activating protein and Nck, leading to inhibition of ras signaling pathway activation and the c-jun N-terminal kinase (JNK) and c-jun activation, respectively (6). The many proteins that have been identified to contain PTB domains fall into two major groups. The first group contains PTB domains that have primary sequence similarity to the Shc PTB domain. The second group contains insulin receptor substrate (IRS)-like proteins such as IRS, Dok, and SNT/FRS2, which contain PTB domains with limited sequence similarity to the Shc PTB domain but similar binding characteristics (9). The Dok1 PTB domain belongs to the second group. It is 17% identical in sequence to the IRS PTB domain and was supposed to recognize sequences containing the NKLpY motif (10). To better understand the PTB domain specificity of Dok and the interaction between Dok1 and RET, we have determined the x-ray crystal structure of the murine Dok1 PTB domain alone and in complex with a phosphopeptide derived from RET. MATERIALS AND METHODS Peptide Synthesis and Binding Studies The following peptides were synthesized by Sigma: the Shc-specific TrkA Tyr(P)-490 (Ac- HIIENPQpYFSDAGGK-NH 2 ), the IRS1-specific IL-4R Tyr(P)-497 (Ac- LVIAGNPApYRSGGK-NH 2 ), RET Tyr(P)-1062 (Ac-STWIEN- 1 The abbreviations used are: PH, pleckstrin homology; PTB, phosphotyrosine binding; JNK, c-jun N-terminal kinase; IRS, insulin receptor substrate; MES, 4-morpholineethanesulfonic acid; PIPES, 1,4- piperazinediethanesulfonic acid; py, phosphotyrosine; MEN, multiple endocrine neoplasia. This paper is available on line at

112 TABLE I X-ray data collection, phasing, and refinement statistics Numbers in parentheses correspond to the highest resolution shell ( Å). MAD data Data set RET peptide complex Peak Edge Remote Wavelength (Å) Resolution (Å) Space group P P Unit Cell a/b/c (Å) 41.1/56.2/ /55.7/99.1 Reflections Total Unique 8330 (780) a 8423 (740) a 8238 (670) a 8410 (764) a Redundancy 7.2 (4.8) a 7.1 (3.9) a 7.3 (2.5) a 5.3 (4.2) a Completeness (%) 99.0 (94.5) a 97.6 (89.0) a 96.6 (80.1) a 91.0 (85.7) a R merge (0.348) a (0.361) a (0.460) a (0.349) a Mean I/ (I) 6.5 (1.7) a 5.6 (1.6) a 5.4 (1.3) a 13.4 (2.7) a Refinement statistics Structural Basis for the Recognition of RET by Dok Resolution range (Å) R work /R free (%) b 21.8/ /27.7 r.m.s.d. c from ideal values Bonds (Å) Angles ( ) Number of atoms Protein Water Ramachandran plot Most favored (%) Additionally allowed (%) a R merge h I I Ih I h / h I I h, where I h is the mean of the observations I ih of reflection h. b R work ( F p (obs) F p (calc) )/ F p (obs) ; R free R factor for a selected subset (5%) of the reflections that was not included in prior refinement calculations. c r.m.s.d., root mean square deviation. KLpYGMSDGGK-NH 2 ) and RET Tyr-1062 non-phosphopeptide (Ac- STWIENKLYGMSDGGK-NH 2 ). A C-terminal GGK extension was added to each of the peptides for coupling to the CM5 chip via the lysine side chain amino group. Binding analyses of the Dok1 PTB domain and the peptides were performed using a Biosensor BIAcore instrument (BIACORE 1000) (BIAcore) according to the manufacturer s instructions. CM5 research grade sensor chips (BIAcore) were used. All buffers were filtered before use. The peptide concentration of 200 g/ml and a contact time of 13.3 min at a flow rate of 3 l/min gave 200 resonance units. Three phosphopeptides were coupled to different flow cells of the CM5 chip, respectively. A reference surface was generated simultaneously under the same conditions but without peptide injection and used as a blank to correct for instrument and buffer artifacts. All measurements were conducted in HEPES-buffered saline buffer (10 mm HEPES, ph 7.4, containing 0.15 M NaCl, 3 mm EDTA, and 0.005% Tween 20) at a flow rate of 20 l/min at 25 C. After each measurement, the chip surface was regenerated with 5 l of6m guanidine-hcl (ph 7.0) buffer at a flow rate of 10 l/min at 25 C. The Dok1 PTB domain was injected at variable concentrations at 20 l/min flow rate, and binding to the peptides immobilized on the chip was monitored in real time. Response curves were prepared by subtracting the signal generated from the control flow cell. Kinetic parameters were determined using the software BIA evaluation 3.0. Protein Expression, Purification, and Crystallization The Histagged murine Dok1 PTB domain (residues ) was expressed and purified by Ni 2 -chelation chromatography. The N-terminal His tag was removed by thrombin digestion, and the protein was purified as described previously. 2 Se-Met-substituted Dok1 PTB domain was produced in the methionine auxotrophic Escherichia coli strain B834 (DE3) (Novagen). Crystals of Se-Met-derived Dok1 PTB domain were grown in a hanging drop by mixing 1 l of protein solution (7 mg/ml, 10 mm MES (ph 6.5), 50 mm NaCl, 10 mm DTT) and 1 l of reservoir solution containing 28% (v/v) polyethylene glycol 6000, 0.1 M MES (ph 6.0), 10 mm dithiothreitol (DTT). Crystals of Dok1 PTB domain in a complex with RET peptide were grown by the same method using 1 l of protein solution (10 mg/ml, 1 mm RET peptide, 10 mm MES (ph 6.5), 50 mm NaCl, 10 mm DTT) with 1 l of reservoir solution (30% (v/v) polyethylene glycol 6000, 0.1 M PIPES (ph 6.0), 10 mm dithiothreitol (DTT)). The resulting crystals grew after 1 week at 16 C. 2 N. Shi, submitted for publication. 112 Data Collection and Structure Determination Data were collected from a flash-frozen crystal after soaking the crystal in a reservoir solution containing 20% (v/v) glycerol. The MAD data were collected at the BL41XU beamline at SPring-8. Three different wavelengths were used to obtain the multiwavelength anomalous diffraction data: Å (peak), Å (edge), and Å (remote). Data were integrated, scaled, and merged using the HKL programs DENZO and SCALE- PACK (12). Crystals of the Dok1 PTB domain belong to the space group P , with unit cell parameters a 41.1 Å, b 56.2 Å, c 99.6 Å, 90, containing two molecules per asymmetric unit. Three selenium sites were located and refined at 2.5-Å resolution using SOLVE (13), which produced a mean figure of merit of After auto-modeling with RESOLVE (14), about 50% of all the residues were easily modeled into the experimental map. The remaining residues were traced manually with O (15). CNS (16) was used for refinement and the addition of solvent molecules. Data from the RET peptide complex crystals were collected at a wavelength of Å at the BL41XU beamline at SPring-8. The crystal also belonged to the space group P , containing two molecules per asymmetric unit, but with different unit cell parameters, a 45.5 Å, b 55.7 Å, c 99.1 Å, 90. The structure of the complex was phased by molecular replacement using CNS with the model of the free Dok1 PTB domain as the starting model. The RET phosphopeptides bound to the Dok1 PTB domain were located using an F o F c difference electron density map. Model building and fitting were carried out using O, and refinement and addition of water molecules were performed using CNS. Data collection, processing, and refinement statistics are given in Table I. The complex model consists of residues of mouse Dok1, the 10 residues of the RET phosphopeptide, and 17 water molecules. Model quality was checked with PROCHECK (17). Coordinates Coordinates and structure factors for the Dok1 PTB domain have been deposited in the Protein Data Bank (accession number 1P5T). Coordinates and structure factors for the Dok1 PTB domain and RET peptide complex have been deposited in the Protein Data Bank (accession number 1EUF). RESULTS AND DISCUSSION Specificity of Phosphopeptide Binding Affinity analysis was performed by means of surface plasmon resonance. The synthetic peptides derived from TrkA (residues ), IL-4R

113 4964 Structural Basis for the Recognition of RET by Dok1 FIG. 1. Biosensor analysis of the Dok1 PTB domain with immobilized phosphopeptides. Five different concentrations of Dok1 PTB domain were injected over three flow cells with different phosphopeptides and the reference flow cell. The sensorgram shows the relative response in resonance units (RU) after background subtraction versus time in seconds are recorded for the following peptide: RET (a), TrkA (b), and IL-4R (c). The concentrations of PTB domain are indicated by numbers in the corresponding graphs. (residues ), and RET (residues ) were coupled to the sensor chip, CM5 of BIAcore, and various concentrations of Dok1 PTB domain solutions were run over the chip. The dissociation constant (K d ) of binding of RET phosphopeptide to the Dok1 PTB domain was determined to be 3.2 M from the data in Fig. 1a. Measurements made in the presence of M of its non-phosphorylated counterpart were unchanged. However, no binding could be detected for immobilized Trka and IL-4 peptides (Fig. 1, b and c), indicating binding specificity of the Dok1 PTB domain to the receptor. Structural Overview The native structure of the Dok1 PTB domain was determined by MAD phasing to 2.5-Å resolution 113 with Se-Met derivative data. Final statistics for the structure are given in Table I. The electron density was of good quality and well defined for most of the structure. The final model consists of residues of mouse Dok1 in chain A, residues in chain B, and 16 water molecules. The PTB domain of Dok1 adopts a PH domain-like fold, with seven strands forming a -sandwich composed of two nearly orthogonal antiparallel -sheets (Fig. 2a). The -sandwich is capped at one end by a C-terminal -helix. Structure of the Dok1 PTB Domain-RET Peptide Complex To gain further insight into the molecular basis for the binding properties of the Dok1 PTB domain, we determined the

114 Structural Basis for the Recognition of RET by Dok FIG. 2. Overall structure of dok1 PTB domain. a, ribbon stereo diagram showing the fold of the Dok1 PTB domain (green) and the orientation of the bound RET phosphopeptide (white). The ribbon diagram was generated with the program BOBSCRIPT (11). b, structure-based sequence alignments of the nine Doks and hirs1 PTB domains. Sequences of mouse Dok1-( ), human dok1-( ), mouse Dok2-( ), human Dok2-( ), mouse Dok3-( ), mouse Dok4-( ), human Dok4-( ), mouse Dok5-( ), human Dok5-( ), and human IRS1-( ) were aligned. Numbers refer to mouse Dok1. The conserving residues were boxed in red and blue. Critical arginines for phosphotyrosine recognition are indicated by green dots. Alignment was generated using CLUSTAL X (1.8). structure of a 1:1 complex of the Dok1 PTB domain (residues ) with an 11-residue peptide derived from the C-terminal of RET (residues ). The structure of the complex was determined by molecular replacement using the native structure as a search model. The structure of the complex is displayed in Fig. 2a, and statistics for the structure determination are given in Table I. Clear density was observed for 114 all residues of the RET peptide, with the exception of Ser in the 8 position relative to the phosphotyrosine (py-8). The peptide-binding site on the Dok1 PTB domain is characterized by an L-shaped surface groove formed by residues from strand 5 and the C-terminal -helix, 2. The peptide forms a -turn to occupy the L-shaped binding site (Figs. 2a and 3). Phosphopeptide Recognition Although it is known that PTB

115 4966 Structural Basis for the Recognition of RET by Dok1 FIG. 3. Stereo view of the electron density map covering the RET peptide. A2 F o F c map is shown at 2.5-Å resolution using phases calculated from the final, refined model and contoured at 1.0 FIG. 4. The contacts between Dok1 PTB domain and RET peptide side chains that contribute specificity to the interaction. a, molecular surface representation of the Dok1 PTB domain structure calculated and shaded according to electrostatic potential using the program ViewerPro (Accelrys). As shown in b, Arg-208, Tyr-209, Gly-210, Ser-217, and Phe-218 of Dok1 PTB domain form a hydrophobic pocket, which may show a preference for large side chain hydrophobic residues such as Trp, Tyr, Phe, and Met in position py-6 of the peptide. As shown in c, large hydrophobic side chains are present at py-5 in the Dok1 PTB domain recognition motifs similar to Shc. As shown in d, Gln-252, Ile-249, and Thr-204 of Dok1 PTB domain form a hydrophobic pocket, which may prefer Leu or Ile in position py-1 of the peptide. The key residues are shown in ball-and-stick representation. domains mainly recognize NPXpY motifs, careful analysis of binding indicates that these domains have slightly different binding specificities (9). Asparagine in position 3 relative to phosphotyrosine (py-3) and the phosphotyrosine group are necessary for binding to most PTB domains. A hydrophobic residue at position 5 and a proline at 2 are crucial for the Shc PTB domain, but the amino acids from 6 to 8 residues N-terminal to the phosphotyrosine are important for IRS1 binding to the NPXpY motifs. The proline in the NPXpY motifs also appears to be more important for IRS1 PTB binding than for Shc PTB binding. In addition, IRS1 PTB favors a small hydrophobic amino acid such as alanine at the 1 position. Substituting this 115 alanine by a glutamate (such as insulin receptor) leads to a 30-fold loss of affinity for the IRS1 PTB domain. Studies with a combinatorial phosphopeptide library have indicated that the Dok1 PTB domain recognizes distinct sequences as compared with the IRS1 and Shc PTB domains. Leu at position 1 and hydrophobic amino acids Tyr, Met, and Phe at 6 were strongly selected for binding by the Dok1 PTB domain. Similar preferences for hydrophobic residues at position 5 to 8 have also been reported for other PTB domains. Our binding studies show that the Dok1 PTB domain can bind only with the RET peptide and not with the IL-4 receptor and TrkA peptides (Fig. 1). Previous experiments indicated

116 Structural Basis for the Recognition of RET by Dok FIG. 5.Stereo view of the interactions between residues at py-1 of the phosphopeptide, shown in brown, and Dok1 (a) or IRS1 (b) PTB domain. Residues involved in important interactions are shown in ball-and-stick representation. The residues interacting with py-1 are represented as green; the sulfur atom is represented in yellow. that IRS1 can bind with IL-4 and insulin receptor peptides and also with the RET peptide but not with the middle T, TrkA, Erb4, or epidermal growth factor receptor peptides that have hydrophobic residues at position 5 relative to Tyr(P) (9, 18). The Shc PTB domain can bind with mt, TrkA, Erb4, or epidermal growth factor receptor peptides and also with IL-4 and RET peptides (9, 19). The distinct specificities of these PTB domains correlate with and may account for some biological differences between these cytoplasmic substrates of tyrosine kinase-linked receptors. Interactions between RET Peptide and the Dok1 PTB Domain The RET peptide forms a -turn and fills an L-shaped groove on the surface of the PTB domain that is formed by residues from the 5 strand and the C-terminal helix. The estimated surface area of Dok1 PTB buried by the bound peptide is 761 Å 2. The recognition groove is composed of residues from the 5 strand, the C-terminal -helix, and the 3 10 turn connecting strands 4 and 5, including Tyr-203, Thr-204, Leu-205, Leu-206, Arg-207, Arg-208, Tyr-209, Arg-211, Ser- 217, Phe-218, Gly-221, Arg-222, Phe-242, Ile-249, Gln-252, Lys-253. These residues make extensive contacts with all 10 residues of the RET peptide, through both hydrogen bonds and hydrophobic interactions. The phosphotyrosine is coordinated by Arg-207 and Arg-222, which extend from the 5 and 6 strands, respectively, and which are conserved in all Dok family proteins (Fig. 2b). The py side chain lies in an open pocket created by the 3 10 turn and residues at the end of strands 5 and 6 (Fig. 2a). An extensive network of hydrogen bonds and ionic interactions coordinate the phosphate oxygens, consistent 116 with the observation that phosphorylation of the tyrosine is necessary for peptide binding. Replacing Arg-207 with alanine eliminates the ability of the Dok1 PTB domain to bind phosphopeptides (6). In addition, integrin 3 and 7 can bind to the Dok1 PTB domain with their tails containing the Dok1 PTB domain recognition motifs (7). The replacement with alanine of the Tyr-747 at py position of integrin 3 tails or the Tyr-778 at py position of integrin 7 tails also disrupted binding to the Dok1 PTB domain (7). The backbone of N-terminal residues of the RET peptide, including residues py-7 Thr, py-6 Trp, py-5 Ile, py-4 Glu, py-3 Asn, form a strand that hydrogen-bonds with strand 5 inan antiparallel orientation. In addition to backbone interactions, there are numerous contacts between the domain and peptide side chains that contribute specifically to the interaction. The indole ring of Trp-6 is bound in a pocket between 5 and 3 that is composed of Arg-208, Tyr-209, Gly-210, Ser-217, and Phe-218. This large pocket suggests that hydrophobic residues with large side chains might be selected here (Fig. 4b). Using a combinatorial peptide library approach, Songyang et al. (10) found that Tyr, Met, and Phe were strongly selected at this site. The side chain of Ile-5 shows numerous contacts with Phe-242 in the C-terminal -helix (Fig. 4c). Large hydrophobic side chains are present at py-5 in the Dok1 PTB domain recognition motifs. Integrin 3 and 7 can bind to the Dok1 PTB domain via their tails, which contain Dok1 PTB domain recognition motifs (7). Replacement of Asp-773 at the 5 position of integrin 7 tails with more hydrophobic Ala or Phe residues dramatically increased Dok1 PTB domain binding to 7 tails, and con-

117 4968 Structural Basis for the Recognition of RET by Dok1 versely, a substitution of Ala-742 to Asp at the 5 position in integrin 3 resulted in reduced binding to Dok1 PTB domain. The Asn at py-3 is similar to that in other PTB domain recognition motifs NPXpY and appears to play an important structural role in stabilizing the -turn of the peptides formed. The side chains of py-4 Glu and py-2 Lys extend away from the surface of the domain. In Dok1 PTB domain, Leu at the 1 position extends into a hydrophobic pocket composed of Gln-252, Ile-249, and Thr-204 and was exclusively selected (Fig. 4d). In addition, py 1 Gly forms a hydrogen bond with Thr-204, and the side chain of py 2 Met interacts with Lys-253 (Fig. 5a). Comparison with Other PTB Domains There is 17% sequence identity between the PTB domains of Dok1 and IRS1, whereas there is no significant sequence homology between the PTB domains of Dok1 and Shc. Despite the low sequence homology, the overall structure of the Dok1 PTB domain is similar to its IRS1 (20) and Shc (21) counterparts. Dok1 shares the FIG. 6. Stereo view of the superposition of Dok1 (red), IRS1 (blue), and Shc (green) PTB domains. Dok1, IRS1, and Shc share a common PH domain-like fold. For clarity, selected residues from the Shc PTB domain have been omitted. C atoms of core residues of the structures superimpose with an root mean square deviation of 1.0Å. PH domain-like fold of the PTB domain family (22) (Fig. 6) and a common mode of peptide binding, with the same -turn conformation and orientation of phosphopeptide observed in each of the PTB domains. There are further similarities between IRS1 and the Dok1 PTB domain. Arg-212 and Arg-227, which recognize the phosphotyrosine in IRS1, are equivalent to Arg-207 and Arg-222 in the Dok1 PTB domain. These two residues are also conserved throughout the IRS protein family (Fig. 2b). Interestingly, the Dok1 PTB domain has a different set of residues for recognizing the peptide. In IRS1, py-1 of the peptide interacts with a hydrophobic patch composed of Met-209, Met-260, Ser-261, and Met-257 (Fig. 5b) (20). Ala was selected in this position, and although py-1 can be substituted for Glu or Leu, they would result in an unfavorable interaction with this patch. When the py-1 Ala in IL-4R is substituted by a Glu, as in the case of the insulin receptor, the result is a 30-fold loss in binding to IRS1 (23). In Dok1, py-1 of the peptide interacts with a hydrophobic pocket composed of Gln-252, Ile-249, and Thr-204, and Leu was exclusively selected in this position (Fig. 5a). The different binding of TrkA and RET phosphopeptides to Dok1 may be due to the replacement of Gln by Leu at the py-1 position. It is demonstrated that py-1 Leu is very important to the Dok1 PTB domain binding motif. The proline in position py-2 is known to be crucial for high affinity binding for Shc and IRS1. Substitution of this residue reduces but does not abolish binding for Shc PTB domain. Meanwhile, substitution of the py-2 proline with alanine abolishes binding for IRS1 (9). The side chain of py-2 Lys extends away from the surface of the domain in Dok1 PTB, where it seems that a proline at position py-2 is not essential. Large hydrophobic side chains are present at py-5 in the Dok1 PTB domain recognition motifs, similar to Shc. However, there is insufficient space in the IRS1 domain complex to accommodate large, hydrophobic side chains at peptide position py-5. As with FIG. 7.The interaction between Dok1 PTB domain and three isoforms of RET. Arg-1064 in RET9 and Ala-1064 in RET 43 were modeled from our structure of the Dok1 PTB domain complexed with the RET51 phosphopeptide. Met-1064 in RET51 (a) and Arg-1064 in RET9 (b) both form an interaction with Lys-253 in the Dok1 PTB domain (3.53 and 3.73 Å, respectively), but Ala-1064 in RET 43 (c) does not. 117

118 Structural Basis for the Recognition of RET by Dok the Shc PTB domain, the Dok1 PTB domain can also recognize the motifs of growth factor receptors and transforming proteins that possess large hydrophobic side chains at py-5, whereas IRS1 does not bind to growth factor receptors. These differences indicate that Dok1 PTB recognizes distinct sequences (NXLpY) as compared with the Shc and IRS1 PTB domain (NPXpY). Dok1 PTB Domain Binding to RET Isoforms The RET proto-oncogene encodes a tyrosine kinase receptor that is essential for the development of the enteric nervous system and the kidney. Germline mutations of the RET proto-oncogene cause multiple endocrine neoplasia (MEN) 2A or 2B (24). RET has three isoforms, RET51, RET9, and RET43, formed by alternative splicing at a site just downstream of tyrosine 1062 (py) (25). These isoforms play different roles in tumor development. RET51-MEN2A and RET51-MEN2B mutant proteins have stronger transforming activity than RET9-MEN2A and RET9-MEN2B mutant proteins, respectively (26). The activity of RET43 is very low (27, 28). The Tyr-1604 (py 2) residue is different in each of these RET isoforms. The RET9 isoform has arginine in the py 2 position, whereas RET43 has alanine and RET51 has methionine in the equivalent position. In our model of the Dok1 PTB domain-ret peptide complex, the RET peptide is derived from RET51, and the Tyr-1604 (py 2) residue is a methionine that forms a hydrophobic interaction with residue Lys-253 that extends from the C-terminal -helix (the distance between C of Met 2toC of Lys-253 is 3.53Å) (Fig. 7a). When Met in the py 2 position is replaced by Arg, there is still an interaction between Arg 2 and Lys-253, but it is weakened (the distance between C of Arg 2 toc of Lys-253 is 3.73Å) (Fig. 7b). However, the substitution of Ala for Met at the py 2 position abolishes the hydrophobic interaction altogether (Fig. 7c). These findings are consistent with the relative transforming activities of the RET isoforms. Conclusions A detailed analysis of the structure of the Dok1 PTB domain and its complex with a RET phosphopeptide has revealed the basis for ligand recognition by the Dok1 PTB domain. We also show that the recognition of peptides by the Dok1 PTB domain is specific since Dok1 cannot bind IL-4 receptor and TrkA peptides that are recognized by Shc and IRS1 PTB domains, respectively. A structural comparison of the Dok1 PTB domain with other PTB domain structures explains their different peptide binding specificities. Furthermore, the distinct specificities of the PTB domains correlate with and should account for key biological differences between these cytoplasmic substrates of tyrosine kinase-linked receptors. Acknowledgments We thank Dr. Min Yao for assistance during data collection at beamline 41 XU at SPring-8, Hyogo, Japan. We also thank Dr. George F. Gao of Oxford University, Oxford, UK, for help with synthetic peptides. REFERENCES 1. Carpino, N., Wisniewski, D., Strife, A., Marshak, D., Kobayashi, R., Stillman, B., and Clarkson, B. (1997) Cell 88, Yamanashi, Y., and Baltimore, D. (1997) Cell 88, Shah, K., and Shokat, K. M. (2002) Chem. Biol. 9, Kato, I., Takai, T., and Kudo, A. (2002) J. Immunol. 168, Becker, E., Huynh-Do, U., Holland, S., Pawson, T., Daniel, T. O., and Skolnik, E. Y. (2000) Mol. Cell Biol. 20, Murakami, H., Yamamura, Y., Shimono, Y., Kawai, K., Kurokawa, K., and Takahashi, M. (2002) J. Biol. Chem. 277, Calderwood, D. A., Fujioka, Y., de Pereda, J. M., Garcia-Alvarez, B., Nakamoto, T., Margolis, B., McGlade, C. J., Liddington, R. C., and Ginsberg, M. H. (2003) Proc. Natl. Acad. Sci. U. S. A. 100, Dhe-Paganon, S., Ottinger, E. A., Nolte, R. T., Eck, M. J., and Shoelson, S. E. (1999) Proc. Natl. Acad. Sci. U. S. A. 96, Wolf, G., Trub, T., Ottinger, E., Groninga, L., Lynch, A., White, M. F., Miyazaki, M., Lee, J., and Shoelson, S. E. (1995) J. Biol. Chem. 270, Songyang, Z., Yamanashi, Y., Liu, D., and Baltimore, D. (2001) J. Biol. Chem. 276, Esnouf, R. M. (1997) J. Mol. Graphics 15, Otwinowski, Z., and Minor, W. (1997) in Macromolecular Crystallography, Part A (Carter, C. W., Jr., and Sweet, R. M., eds) Vol. 276, pp , Academic Press, Orlando, FL 13. Terwilliger, T. C., and Berendzen, J. (1999) Acta Crystallogr. Sect. D Biol. Crystallogr. 55, Terwilliger, T. C. (2000) Acta Crystallogr. Sect. D Biol. Crystallogr. 56, Jones, T. A., Zou, J. Y., Cowan, S. W., and Kjeldgaard, M. (1991) Acta Crystallogr. Sect. A 47, Brünger, A. T., Adams, P. D., Clore, G. M., DeLano, W. L., Gros, P., Grosse- Kunstleve, R.W., Jiang, J.-S., Kuszewski, J., Nilges, M., Pannu, N. S., Read, R. J., Rice, L. M., Simonson, T., and Warren, G. L. (1998) Acta Crystallogr. Sect. D Biol. Crystallogr. 54, Laskowski, R. A., MacArthur, M. W., Moss, D. S., and Thornton, J. M. (1993) J. Appl. Cryst. 26, Melillo, R. M., Carlomagno, F., De Vita, G., Formisano, P., Vecchio, G., Fusco, A., Billaud, M., and Santoro, M. (2000) Oncogene 20, Asai, N., Murakami, H., Iwashita, T., and Takahasi, M. (1996) J. Biol. Chem. 271, Zhou, M. M., Huang, B., Olejniczak, E. T., Meadows, R. P., Shuker, S. B., Miyazaki, M., Trub, T., Shoelson, S. E., and Fesik, S. W. (1996) Nat. Struct. Biol. 3, Zhou, M. M., Ravichandran, K. S., Olejniczak, E. F., Petros, A. M., Meadows, R. P., Sattler, M., Harlan, J. E., Wade, W. S., Burakoff, S. J., and Fesik, S. W. (1995) Nature 378, Murzin, A. G., Brenner, S. E., Hubbard, T., and Chothia, C. (1995) J. Mol. Biol. 247, He, W., O Neill, T. J., and Gustafson, T. A. (1995) J. Biol. Chem. 270, Watanabe, T., Ichihara, M., Hashimoto, M., Shimono, K., Shimoyama, Y., Nagasaka, T., Murakumo, Y., Murakami, H., Sugiura, H., Iwata, H., Ishiguro, N., and Takahasi, M. (2002) Am. J. Pathol. 2002, Ivanchuk, S. M., Myers, S. M., and Mulligan, L. M. (1998) Oncogene 16, Pasini, A., Geneste, O., Legrand, P., Schlumberger, M., Rossel, M., Fournier, L., Rudkin, B. B., Schuffenecker, I., Lenoir, G. M., and Billaud, M. (1997) Oncogene 15, Carter, M. T., Yome, J. L., Marcil, M. N., Martin, C. A., Vanhorne, J. B., and Mulligan, L. M. (2001) Cytogenet. Cell Genet. 95, Lee, D. C., Chan, K. W., and Chan, S. Y. (2002) Oncogene 21,

119 THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 279, No. 5, Issue of January 30, pp , by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A. Amyloid Nucleation and Hierarchical Assembly of Ure2p Fibrils ROLE OF ASPARAGINE/GLUTAMINE REPEAT AND NONREPEAT REGIONS OF THE PRION DOMAIN* Received for publication, September 22, 2003, and in revised form, November 4, 2003 Published, JBC Papers in Press, November 10, 2003, DOI /jbc.M Yi Jiang, Hui Li, Li Zhu, Jun-Mei Zhou, and Sarah Perrett From the National Laboratory of Biomacromolecules, Institute of Biophysics, Chinese Academy of Sciences, 15 Datun Road, Chaoyang District, Beijing , and the State Key Laboratory of Magnetism, Center for Condensed Matter Physics, Institute of Physics, Chinese Academy of Sciences, P. O. Box 603, Beijing , China The yeast prion protein Ure2 forms amyloid-like filaments in vivo and in vitro. This ability depends on the N-terminal prion domain, which contains Asn/Gln repeats, a motif thought to cause human disease by forming stable protein aggregates. The Asn/Gln region of the Ure2p prion domain extends to residue 89, but residues represent an island of normal random sequence, which is highly conserved in related species and is relatively hydrophobic. We compare the time course of structural changes monitored by thioflavin T (ThT) binding fluorescence and atomic force microscopy for Ure2 and a series of prion domain mutants under a range of conditions. Atomic force microscopy height images at successive time points during a single growth experiment showed the sequential appearance of at least four fibril types that could be readily differentiated by height (5, 8, 12, or 9 nm), morphology (twisted or smooth), and/or time of appearance (early or late in the plateau phase of ThT binding). The Ure2 dimer (h nm) and granular particles corresponding to higher order oligomers (h 4 12 nm) could also be detected. The mutants 15Ure2 and 15 42Ure2 showed the same time-dependent variation in fibril types but with an increased lag time detected by ThT binding compared with wild-type Ure2. In addition, 15 42Ure2 showed reduced binding to ThT. The results imply a role of the conserved region in both amyloid nucleation and formation of the binding surface recognized by ThT. Further, Ure2 amyloid formation is a multistep process via a series of fibrillar intermediates. Ure2p is the protein determinant of the epigenetic factor [URE3] of Saccharomyces cerevisiae, which has been demonstrated to represent a prion of yeast (1, 2). Analogous to the mammalian prion protein (3), Ure2p is aggregated, inactive, and protease-resistant in prion strains (1, 2). In addition, amyloid-like filaments are observed both in vivo (4) and in vitro (5, 6). Fibrils formed in vitro show characteristic green birefringence on Congo red binding (5, 7, 8) and also bind thioflavin T * This work was supported in part by Natural Science Foundation of China Grant , 973 Project of the Chinese Ministry of Science and Technology G , and State Key Development Plan Project The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. To whom correspondence may be addressed. Tel.: ; Fax: ; zhoujm@sun5.ibp.ac.cn or sarah.perrett@ iname.com. Supported by the Chinese Academy of Sciences, the Royal Commission for the Exhibition of 1851, and the Royal Society. This paper is available on line at (ThT) 1 (7 9), a dye considered highly specific for amyloid (10, 11). Amyloids are thought to form by conversion of the native protein structure to a generic cross- structure, identified by a characteristic x-ray diffraction pattern (12). However, an unusual property of Ure2 fibrils is their ability to maintain native-like ligand binding properties within the fibrillar arrays (8, 13). Consistent with this, Ure2 fibrils formed by standing at subambient temperatures show a Fourier transform infrared spectrum consistent with native-like helical content (8), and the cross- x-ray diffraction band becomes apparent only after incubation of the protein close to the T m (14). Atomic force microscopy (AFM) imaging of these native-like fibrils shows them to be homogeneous, with a height of around 12 nm and a periodicity of around 50 nm (8). Ure2 prion formation in vivo (2) and fibril formation in vitro (5, 15) are dependent on the presence of the N-terminal 90 amino acids. The N-terminal prion domain (PrD) is unstructured in the native state (6, 16) and is rich in Asn and Gln residues (see Fig. 1). Deletion of all or parts of the PrD has no discernible effect on the dimeric structure, thermodynamic stability, folding kinetics, or folding pathway of Ure2 under a wide range of conditions (9, 15, 16, 20 22). Expansion of Gln (or CAG) repeats is responsible for a number of human neurodegenerative diseases, including Huntington s disease (23). The ability of poly-gln or poly-asn regions to aggregate by forming a hydrogen bonded -sheet structure is thought to cause disease (24 27), either by direct toxicity (28) or by sequestration of other vital cellular proteins (29, 30). There is evidence for poly-gln and other amyloid diseases that the species most damaging to cells are protofibrils, or intermediates formed early in the aggregation process, rather than large fibrillar aggregates (31, 32). ThT binding provides a convenient method to measure the effect of different environmental factors on the kinetics of amyloid formation (9, 33). The time course of Ure2 amyloid formation monitored by ThT binding shows a sigmoidal curve, representing a lag time, an exponential growth phase, and a plateau region (7 9). The lag time can be circumvented by seeding with preformed amyloid-like fibrils (5 7) and is protein concentration-dependent (6, 9). This is consistent with a nucleation-dependent mechanism (34, 35), where the lag time reflects the kinetic barrier to association of a sufficient number of molecules to form a thermodynamically stable nucleus or seed. However, once this stable nucleus is formed, exponential growth from the fibril ends is then observed, until steady-state conditions are reached. Results from a number of laboratories indicate that conditions that produce this characteristic increase in dye binding for Ure2 1 The abbreviations used are: ThT, thioflavin T; AFM, atomic force spectroscopy; EM, electron microscopy; PrD, prion domain; WT, wild-type.

120 3362 Amyloid Nucleation and Hierarchical Assembly of Ure2 Fibrils also lead to the eventual appearance of well formed amyloidlike fibrils, visualized by EM or AFM (5 9). Oligomeric particles or rods of Ure2 have been observed by EM (6) and AFM (8). However, the relationship between the time course of changes in ThT binding and evolution of amyloid-like structure has not been investigated for Ure2. AFM offers particular advantages over EM, in that hydrated samples can be observed directly, in air, without requirement for staining. In addition, AFM can detect the presence of aggregated species as well as differences in fibril morphology which are not apparent in EM images (8, 36, 37). Here we compare the time course of structural changes monitored by ThT binding and AFM and examine the role of different parts of the Ure2 PrD sequence on the kinetics of fibril formation. EXPERIMENTAL PROCEDURES Materials ThT and Tris were from Sigma. Ure2 and N-terminal variants, 15Ure2, 15 42Ure2, 42Ure2, and 90Ure2 (see Fig. 1) were produced in Escherichia coli with a short His tag and purified as described previously (9, 16). 42Ure2 was constructed as described previously for the other mutants (16). Proteins were stored at 80 C and defrosted in a 25 C water bath immediately prior to use. Samples were prepared in 50 mm KH 2 PO 4 -Na 2 HPO 4 buffer containing 0.15 M NaCl, or in 50 mm Tris-HCl buffer containing 0.2 M NaCl as described (9) and centrifuged at 18,000 g for 30 min at 4 C to remove any aggregated protein. Amyloid Formation The kinetics of amyloid formation of Ure2 proteins was monitored using ThT binding fluorescence as described previously (7, 9, 33, 38). Incubation was at a constant temperature of 25 or 37 C with shaking, as described previously (9). Alternatively, proteins were incubated without shaking at 4 or 25 C. NaN 3 (0.02% w/v) was added to prevent bacterial growth. The ph range of buffers used was (at 25 C), and the protein concentration range was M for full-length Ure2 and M for Ure2 mutants. The ph values given in the text are correct at the temperature of incubation, allowing for the temperature dependence of the ph of Tris buffer. At regular time intervals, 10- l aliquots were removed from the reaction mixture and assayed for ThT binding, as described previously (9). Samples were incubated in parallel whenever possible. When comparing the time course of amyloid formation by ThT binding and AFM, samples were taken simultaneously from the same reaction vessel whenever possible. AFM A 10- l drop of the protein sample was deposited on freshly cleaved mica, allowed to stand for 10 min in air, then washed with three 200- l aliquots of distilled deionized water, before drying for 4 min in a stream of nitrogen. Tapping mode AFM was performed using a Nanoscope IIIa Multimode-AFM instrument (Digital Instruments) under ambient conditions. Super-sharp silicon tips (Silicon-MDT Ltd.) with resonance frequency of about 106 khz were used at a scan rate of 1 2 Hz. Once the tip was engaged, the set point value was adjusted to minimize the force exerted on the sample while maintaining the sharpness of the image. Height measurements of granular and fibrillar particles were performed manually using the software provided with the Nanoscope instrument. To compare the size distribution of granules over time, the maximal height of every granular particle detected within a representative scan area of fixed size was measured for each time point. To determine the distribution of heights of fibrils, the height profiles were measured perpendicular to the fibril over a region greater than the periodicity of height variations within an individual fibril. The average peak height for each fibril was recorded (maximal variation around 1 nm). Mean heights were obtained by fitting to a Gaussian curve. The errors shown are the standard deviation. Scanning different regions of the mica surface confirmed that the distribution of particles or fibrils was uniform and that the scan areas sampled were representative, with the exception that protofilaments were deposited preferentially at the edge of the grids, suggesting that they are more easily dislodged from the mica surface by the washing process than mature fibrils. This has important consequences for the interpretation of the data because it indicates that although protofilament formation clearly precedes fibril formation, protofilaments are present earlier and more abundantly than apparent from the AFM assay procedure. This may explain in part the increased tendency to observe protofilaments when fibrils are grown in situ on the mica (39). 120 FIG. 1.Primary structure of the Ure2p N-terminal PrD in the WT protein and in the prion domain deletion mutants. Repetitive regions detected by sequence analysis (16) are indicated by diagonal stripes or bold type. Hydrophobic regions of the PrD, identified by plotting the relative hydrophobicity (17, 18), are underlined. The C- terminal region has homology to the glutathione S-transferase protein family (19). RESULTS Ure2 Mutants 15Ure2 and 15 42Ure2 Show an Increase in the Lag Time Detected by ThT Binding ThT binding provides a convenient method to assay the effect of different factors on the kinetics of amyloid formation (9, 33). We compared the ThT-monitored kinetics of amyloid formation for full-length Ure2 and a series of PrD deletion mutants. The mutants examined were 15Ure2, which lacks the first stretch of repetitive sequence; 15 42Ure2, which retains all the Asn/Gln repeat regions, but lacks the island of normal random sequence within the PrD; 42Ure2, which lacks the first 41 residues of the PrD; and 90Ure2, in which the entire Asn/Gln repeat region has been deleted (Fig. 1). The WT and mutant Ure2 proteins were compared under a range of incubation conditions, including different temperatures (4, 25, and 37 C), buffer systems (sodium/potassium phosphate or Tris-HCl), and with or without agitation. Representative curves for a variety of conditions are shown in Fig. 2. The mutants 42Ure2 and 90Ure2 showed a negligible increase in ThT binding fluorescence even at maximal protein concentrations. This is consistent with the requirement of residues 1 65 for induction or maintenance of the prion state in vivo (2, 40). Interestingly, the mutant 15Ure2 consistently showed a sigmoidal time course monitored by ThT binding, similar to that for WT Ure2, but with a longer lag time (Fig. 2) Ure2 showed greatly reduced binding to ThT compared with 15Ure2 or WT. However, under certain conditions (particularly in Tris buffer at relatively high protein concentrations), the limited increase in ThT binding for 15 42Ure2 could be seen to be sigmoidal, revealing a lag time slower than for WT Ure2 and similar to 15Ure2 at the same protein concentration (Fig. 2B). This suggests a role of both Asn/Gln repeat (residues 1 14) and nonrepeat (residues 15 42) regions of the PrD in nucleation of amyloid structure. However, it also indicates that deletion of either one of these two regions does not ablate the ability to form an amyloid-like structure. Characterization of the Ure2 Dimer by AFM AFM height imaging provides a convenient method to characterize the morphology of biological macromolecules at submolecular resolution without requirement for staining (36, 37, 39, 41 43). Typ-

121 Amyloid Nucleation and Hierarchical Assembly of Ure2 Fibrils 3363 FIG. 2.Kinetics of formation of amyloid-like structure for Ure2 and PrD mutants monitored by ThT binding under a range of conditions. Values of ph are correct at the temperature of incubation. A, 40 M Ure2 ( ), 15Ure2 (Œ), 15 42Ure2 (f), and 42Ure2 ( ) incubated in Tris-HCl buffer, ph 7.5, 0.2 M NaCl at 25 C with shaking. B, 30 M Ure2 ( ); 40 M (Œ) and 80 M ( ) 15Ure2; 40 M (f) and 80 M ( ) 15 42Ure2; and 40 M ( ) and 80 M ( ) 42Ure2 incubated in Tris-HCl buffer, ph 7.2, 0.2 M NaCl at 37 C with shaking. C, 30 M Ure2 ( ), 40 M 15Ure2 (Œ), and 80 M 15 42Ure2 (f) incubated in sodium/potassium phosphate buffer, ph 8.0, 0.2 M NaCl at 37 C with shaking. D, 40 M Ure2 ( ), 15 42Ure2 (f), 42Ure2 ( ), and 90Ure2 ( ) incubated in Tris-HCl buffer, ph 9.0, 0.2 M NaCl at 4 C without shaking. ically, the sample is adsorbed onto mica, and the surface is scanned with an oscillating or tapping tip on a microcantilever. The variations in height can then be represented by a gray scale, ranging from white (high) to black (low), to form the image. A characteristic of AFM imaging is that the size of the scanning tip relative to the size of protein oligomers or fibrils results in an overestimation of the width, whereas the height of the sample above the mica surface can be measured with accuracy and reproducibility. Nevertheless, the height of biological molecules measured by AFM tends to be smaller than the diameter measured by other structural methods, which may reflect the absence of stain, the degree of hydration, and/or compression by the AFM tip. To calibrate the height of the Ure2 dimer by AFM, we prepared a series of dilutions of Ure2 and the mutants 15 42Ure2 and 90Ure2 in Tris-HCl buffer, ph 8.4, 0.2 M NaCl at 25 C. Under these conditions the protein shows a minimal tendency to aggregate and is expected to be predominantly dimeric at micromolar protein concentrations (16), although species with sedimentation values corresponding to monomers and tetramers are detected under similar buffer conditions at 15 C, depending on the protein concentration (6). After adsorption of a 1 M protein solution of WT or mutant Ure2 onto mica, AFM imaging reveals a homogenous population of spherical particles (Fig. 3). The predominant species has a height of nm (Fig. 3C, upper panel). The WT sample contained a minor population (less than 5%) of larger particles (Fig. 3A), which have heights in the range 4 8 nm (data not shown). At lower protein concentrations, an additional peak at nm appears (Fig. 3C, lower panel), consistent with population of the monomer. Identical peaks are observed for WT Ure2, 15 42Ure2, and 90Ure2, indicating that the PrD does not 121 contribute to the height of the Ure2 monomer or dimer as measured by AFM. Comparison of Time Course of Structural Changes Monitored by ThT Binding and AFM To ascertain the structural basis of the differences in kinetics detected by ThT binding for Ure2 and the PrD mutants described above, we measured the time course of structural changes in parallel by ThT binding and tapping mode AFM. In each case, aliquots were removed from the reaction vessel for assay. Fig. 4 shows a comparison of the time course measured by the two methods for Ure2 incubated in phosphate buffer, ph 8.0, at 37 C with shaking. Under these conditions, abundant well formed fibrils are formed within 24 h, as detected by EM, and the time course of ThT binding is highly reproducible and relatively rapid (9), so that a plateau is reached within 12 h (Fig. 4A). The initial Ure2 protein solution was found to contain a variety of sizes of granular particles (Fig. 4, B and E, 0h), ranging in height from 1 to 85 nm, although greater than 90% of the particles were less than 20 nm in height. The larger particles appear to be dispersed or precipitated after the onset of shaking and are concluded to represent amorphous aggregates. Occasional fibrillar structures are also detected in the initial sample (Fig. 4E, 0h). The presence of a low concentration of preformed fibrils, or nuclei for fibril formation, could account for the linear rather than exponential dependence of the lag time on initial Ure2 concentration, as discussed previously (9). Changes in the height distribution of granular particles over the time course of the ThT-monitored curve were examined by recording the heights of all granules within a representative 2- m square at hourly time points between the onset of incubation (0 h) and the onset of fibril formation (7 h). Within the lag phase region (0 2 h), there was a marked increase in the number of particles of height 4 6 nm, which was then followed (2 3 h) by an increase in particles of height 8 12 nm (Fig. 4C), coinciding with an increase in ThT binding fluorescence (Fig. 4A). Given the height of the Ure2 dimer of nm (Fig. 3), this then suggests that the 4 6-nm particles that appear during the lag phase are tetramers or hexamers, whereas the 8 12-nm particles are likely to represent larger oligomers. The earliest fibril-like structures appear late in the exponential phase of ThT binding (6 h) and have a height of around 5 nm, thus resembling the protofilaments observed for other amyloidogenic proteins (28, 31, 36, 37, 39, 42, 43). These are succeeded by 8-nm fibrils (7 h) and then 12-nm fibrils (8 h). Significant numbers of fibrillar structures are not detected until ThT binding has already reached a plateau (Fig. 4, A, D, and E). However, the number of protofilaments is underestimated in the AFM assay (see Experimental Procedures ). It is therefore likely that both protofilaments (5 nm; also 2.5 nm, see below) and larger granules (8 12 nm) contribute to the onset of ThT binding. Continued monitoring of the fibril heights and morphology reveals further structural changes over time, with little further variation in the ThT binding fluorescence (Fig. 4). Early in the plateau phase (8 h), the majority of fibrils have a height of around 12 nm and a relatively smooth appearance, although the height is observed to vary with a periodicity of nm, similar to the native-like fibrils produced after a week of standing in Tris buffer at 4 C (8). Over the course of the following hours and days of incubation at 37 C, there was a gradual decrease in the mean fibril height, until at 7 days the fibril heights resolved into two peaks, indicating the presence of two distinct fibril types (Fig. 4D). (Fitting the combined data to a two-peak Gaussian gives a height of nm for the fibril type that dominates early in the plateau phase, and a height of nm for the fibril type that dominates at

122 3364 Amyloid Nucleation and Hierarchical Assembly of Ure2 Fibrils FIG. 3.AFM characterization of the Ure2 dimer. AFM height images of Ure2 (A) and 15 42Ure2 (B) in Tris-HCl, ph 8.4, 0.2 M NaCl, room temperature, and 1 M protein concentration. The scale bar represents 200 nm, and the full range of the gray scale corresponding to height was 10 nm. C, height distribution of the particles for Ure2 (black) and 15 42Ure2 (gray) at1 M protein concentration (upper panel) or at approximately 0.1 M (lower panel). In each case the scan area was a 2- m square or less. extended incubation times.) The other change that is apparent over this time period is that most fibrils no longer appear smooth, but have a clearly distinguishable twisted or zigzag morphology (Fig. 5A). Careful inspection reveals that the small proportion of twisted fibrils present at time points when 12-nm fibrils dominate (twisted fibrils indicated by arrows in Fig. 5A, 8h) consistently have heights of around 8 nm. This then indicates that as well as protofilaments, at least two additional types of fibril are present, distinguishable by their height and morphology, the relative proportions of which change over time. The preferential loss of the thicker smooth fibrils over time suggests that the twisted fibrils may be derived from the smooth fibrils by conformational conversion or separation of the constituent strands. Comparison of Fibril Morphology for Ure2, 15Ure2, and 15 42Ure2 As described above, a parallel assay by ThT binding and AFM indicates that long fibrils of WT Ure2 are present by the end of the exponential growth phase (Fig. 5A, 6h), are abundant by early plateau phase (Fig. 5A, 8h), and persist even at very long incubation times (Fig. 5A, 4d). Variations in morphology over this time course include a change in the distribution of fibril heights, indicating the progressive appearance of 5-, 8-, 12-, and 9-nm fibrils (Figs. 4D and 5D). The ThT binding assay (Fig. 2) suggests that 15Ure2, and to a lesser extent 15 42Ure2, are also able to form amyloid-like structure, although with a longer nucleation-dependent lag time. A comparison by AFM of WT and mutant fibrils at equivalent stages of the ThT-monitored time course is shown in Fig. 5. Incubation of the proteins was under identical buffer conditions (phosphate buffer, ph 8.0, 37 C, with shaking), but with the protein concentration of the mutants increased to give a similar kinetic time course as WT. (At lower protein concentrations, similar short fibrils are observed for 15Ure2, but only amorphous structures could be detected for 15 42Ure2, data not shown.) The results indicate that the observed increase in ThT 122 binding for the mutant proteins (Fig. 2) coincides with formation of fibrillar structures. However, the mutant fibrils are shorter in length than WT and do not elongate further even at extended incubation times (Fig. 5, A C). This is particularly marked for 15 42Ure2, which shows only rod-like structures ( nm length). However, the WT and mutant fibrils or rods show a similar pattern of height changes over time, with progressive appearance of 5-, 8-, 12-, and 9-nm height protofilaments or fibrils (Fig. 5, D F). Reduced ThT Binding Ability of the 15 42Ure2 Mutant On first inspection, the low level of ThT binding for the 15 42Ure2 mutant, particularly in phosphate buffer, appears to correlate with the inability of the fibrils to extend beyond short rods (Fig. 5C). However, when investigating Ure2 and its mutants by AFM under nonaggregating conditions (Tris, ph 8.4, at 25 C), we made the unexpected discovery that although WT Ure2 aggregation is strongly reduced under these conditions (16), fibrils of 15 42Ure2 are readily formed on standing in this buffer when stored at 4 C (Fig. 6). WT Ure2 fibrils (Fig. 6, A and D, insets of E) formed at 4 C in this buffer, or at slightly lower ph as used to produce native-like Ure2 fibrils (8, 13), are morphologically indistinguishable from these 15 42Ure2 fibrils (Fig. 6, B, C, and E). Subpopulations of protofilaments with heights of around 2.5 and 5 nm can be clearly distinguished (Fig. 6, D and E), and bundling or branching of the filaments can occasionally be observed (Fig. 6C). In the high ph buffer, 15 42Ure2 fibrils are far more abundant than WT at all time points (Fig. 6E). In contrast to the phosphate buffer time course (Fig. 4), few granular particles greater than 4 nm were observed in Tris buffer, suggesting that larger granular particles are not obligate intermediates in fibril formation and that the ThT binding under these conditions is the result of the presence of protofilaments and fibrils. The contradiction between the extent of ThT binding (Fig. 2D) and the abundance of fibrils for Ure2 and 15 42Ure2

123 Amyloid Nucleation and Hierarchical Assembly of Ure2 Fibrils 3365 FIG. 4.Comparison of the time course of formation of amyloid-like structure monitored by ThT binding and AFM. 30 M WT Ure2 protein was incubated in sodium/potassium phosphate buffer, ph 8.0, 0.15 M NaCl, and 37 C with shaking. Structural changes were monitored by ThT binding (A) and AFM (B E). B, population distribution of heights of granular particles observed in the initial protein sample, derived from analysis of two representative scan areas each of 5 m square. C, population distribution of heights of granular particles observed over different time points, as indicated. All particles detected within a standard, representative scan area of 2 m square for each hourly time point were included. D, population distribution of heights of fibrils observed at different time points, as indicated. A representative 5- m square scan area from the center of the mica surface was analyzed in each case (but see Experimental Procedures and Fig. 6). E, AFM height images during the course of fibril formation. Samples were removed from a single reaction vessel at the time points indicated. The bars represent 1 m. The full range of the gray scale corresponding to height was 20, 30, or 40 nm for 0 4 h and 50 nm thereafter, except 2d, which was 100 nm. 123

124 3366 Amyloid Nucleation and Hierarchical Assembly of Ure2 Fibrils FIG. 5.Morphology and height distribution of amyloid-like structures monitored by AFM for WT and mutant Ure2. A and D, 30 M WT Ure2; B and E, 80 M 15Ure2; and C and F, 80 M 15 42Ure2. The incubation conditions were the same as in Fig. 4. For each protein, the samples were removed from a single reaction vessel at the time points indicated. A C, AFM height images. The bar represents 1 m. The full range of the gray scale corresponding to height was 30 nm (B, 5h) or 50 nm (all other panels, A C). D F, population distribution of heights of fibrils or rods at different incubation times as indicated, obtained by measuring the heights of all fibrils within a representative 5- m square scan area, which included the scan areas displayed in A C. (Fig. 6E) could be the result of a vastly lower affinity of WT Ure2 for the mica surface. However, the lack of WT Ure2 fibrils under these conditions is also apparent from EM (9, 15). The simplest explanation for the discrepancy between ThT binding 124 and the presence of fibrils for 15 42Ure2 is that the region is important for binding of the hydrophobic dye ThT. Plotting the relative hydrophobicity of the PrD sequence (17, 18) indicates that the region spanning residues is rela-

125 Amyloid Nucleation and Hierarchical Assembly of Ure2 Fibrils 3367 FIG. 6.Observation of protofilaments and fibrils of WT Ure2 and 15 42Ure2. Incubation was at 4 C in Tris-HCl buffer, 0.2 M NaCl, ph 8.6 (A and D) orph9.0(b, C, and E) for 3 or 7 days, as indicated. (Note: the same buffers have a ph of 8.0 or 8.4, respectively, at 25 C.) A C, the bar represents 200 nm, and the full range of the gray scale corresponding to height was 30 nm (A and C) or50nm(b). Fibrils of WT (A) or 15 42Ure2 (B and C) have a twisted or braided appearance with a height of 8 1 nm and a periodicity of nm. C, bundling or branching of 5-nm protofilaments into 8-nm fibrils. The height across the junction region (between arrows) is nm. D and E, population distribution of protofilaments and fibrils within representative 5- m scan areas at the center (black) or the edge (gray) of the mica surface for WT (D and insets of E) and 15 42Ure2 (E, main panels). tively hydrophilic. However, two hydrophobic regions consisting of residues 9 21 and flank this region (Fig. 1). This may then account for the reduced binding of ThT to the mutant 15 42Ure2, despite its ability to form fibrillar structures morphologically indistinguishable from those formed by WT Ure2. DISCUSSION The importance of the Ure2 N-terminal domain in prion induction and propagation has been established by genetic studies (2, 40), and the requirement of the PrD to form amyloidlike fibrils in vitro has also been demonstrated (5, 15). The mutants initially used in genetic studies to define the PrD were designed at the DNA level, on the basis of convenient endonuclease restriction sites (2, 40), whereas the mutants used in this study were designed at the protein level, based on interesting features of the of the PrD primary structure (Ref. 16 and see Fig. 1). Protein sequence analysis of Ure2 homologs in different species of yeast has recently identified a conserved region in the otherwise divergent PrDs, corresponding to residues (44). Within Saccharomyces species, the highly conserved region extends from residue 1 to 43. Overexpression of fragments spanning this region causes curing of the prion state in prion strains and inactivation of the Ure2 protein in non-prion strains (44). This suggests that the region spanning residues 1 43 is involved in intermolecular interactions important for formation of the prion state. Our finding that the Ure2 mutants 15Ure2 and 15 42Ure2, which have deletions within this crucial region, are nevertheless able to form amyloid-like fibrils is therefore somewhat surprising. Comparison with the Ure2 fragments that have been studied in vivo (Table I) would predict that these two mutants would be unable to induce or cure the prion state. Possible explanations for the curing properties of the conserved region when overproduced in vivo include a crystal poisoning mechanism caused by the interaction of heterogenous molecules with the growing prion seed; induction of chaperones, which then cure the prion state; or the requirement for a cellular cofactor to maintain the prion state, which is depleted by the Ure2 fragment (44, 45). Our in vitro results,

126 3368 Amyloid Nucleation and Hierarchical Assembly of Ure2 Fibrils TABLE I Properties of Ure2 mutants and fragments in vivo Experimental details are given in the references indicated., observed;, not observed (or very low rate of occurrence); ND, not determined; NA, not applicable. Protein Curing a Inactivation b Induction c Propagation d Spontaneous generation e Ure2 WT Ure ND ND Ure ND NA Ure ND NA Ure NA Ure ND ND NA Ure ND NA Ure ND NA Ure ND a Curing of the prion state when the gene is expressed in a WT background in an initially [URE3] prion strain (44, 45). b Inactivation of the Ure2 protein (but not necessarily induction of a stable prion [URE3] state) when the gene is expressed in vivo inawt background (44). c Induction of the [URE3] prion when the gene is expressed in a WT background in an initally non-prion strain (2). d Ability to propagate the prion state when the gene is expressed in vivo in a null (ure2 ) background (40). e Spontaneous generation of the [URE3] prion when the gene is expressed in vivo in a ure2 background (40). (This is not applicable to N-terminal fragments that lack the Ure2 functional region because they cannot show the heritable loss of Ure2 function which is the signature of the [URE3] prion.) in the absence of cellular cofactors, provide support for a crystal poisoning mechanism and demonstrate the importance of sequence complementarity during self-association to form amyloid-like structure. The increased lag time for nucleation-dependent formation of amyloid-like structure for the mutants examined in this study is consistent with a role of the conserved region of the Ure2 PrD in forming inter- or intramolecular interactions that help stabilize the prion seed. The low level of binding of the amyloidspecific dye ThT to 15 42Ure2, independent of fibril morphology, suggests that the island of normal random sequence within the PrD is involved in formation of the structural motif or binding surface recognized by ThT. This possibly reflects the relatively hydrophobic nature of this region (Fig. 1), which could also explain its involvement in curing and inactivation in vivo (Table I). The delayed fibril formation kinetics of the mutants compared with WT provides an explanation for their increased solubility in solution and in E. coli cell extracts (Table II). The difference in solubility of WT and mutant Ure2 in vitro is most marked in phosphate buffer (16), which correlates with destabilization of the native state relative to partially folded intermediates, likely to be precursors in amyloid formation and other aggregation processes (9). In phosphate buffer, some differences in morphology between WT and mutant Ure2 fibrils could be observed, but this was limited to fibril length. In Tris buffer, WT and mutant fibrils were indistinguishable in length and morphology. Further, under all conditions, the same series of fibrillar intermediates was observed for WT and the mutants. This then implies that the mechanism of assembly of WT and mutant Ure2 fibrils is the same. It will be interesting to test the properties of 15Ure2 and 15 42Ure2 in vivo, particularly to see whether overproduction of these mutants in a ure2 background could spontaneously generate a heritable prion-like state, as has been demonstrated for the WT protein (40). Comparison of the time course of structural changes monitored by ThT binding and by AFM for Ure2 indicates that increased binding of ThT correlates well with appearance of fibrillar structures. However, granular aggregates were also 126 TABLE II Properties of Ure2 mutants in vitro Exact experimental details are given in the references or figure legends indicated., observed;, not observed; ND, not determined. a Rate of Protein Solubility nucleation b ThT binding c Fibril length by AFM d WT Ure2 15Ure Ure2 42Ure2 ND 74Ure2 ND ND ND 90Ure2 a Solubility in E. coli cell extracts (16) or in pure solution (9) in phosphate buffer, ph 7.5. Note that all constructs can be solubilized using Tris-HCl buffer, ph 8.4 (16), and that amyloid-like fibrils of Ure2 PrD fusion proteins have been observed in bacterial cell extracts (7). b Relative rate of nucleation of amyloid-like structure, assayed by ThT binding (, shorter lag time;, longer lag time;, no increase in ThT binding observed; see Fig. 2). c Relative ThT binding in plateau phase (Fig. 2). d Relative lengths of fibrillar structures detected by AFM after incubation in phosphate buffer (, long fibrils;, short fibrils;, rods;, no fibrils; Figs. 4 and 5). Note that WT and mutant fibrils show the same time-dependent changes in fibril thickness (height) under all conditions and that no differences in fibril length were observed in Tris-HCl buffer (Figs. 5 and 6). observed under certain conditions and may also bind ThT. Of particular note in the AFM study presented here is the time dependent appearance of a series of Ure2 fibril types, including protofilaments (2 5 nm), intermediate fibrils (8 nm), and two types of mature fibrils (12 and 9 nm), implying a hierarchical mechanism of assembly. Consistent with this, some fibrils had a twisted appearance, and bundling or branching of fibrils could be observed. This is similar to the results obtained for other amyloidogenic proteins (28, 36, 37, 39, 42, 43), including polyglutamine peptides (46). The hierarchical assembly of fibrils, by twisting together of protofilaments or protofibrils to form a variety of fibril morphologies, is emerging as a common mechanism of fibril assembly for all amyloidogenic proteins (47, 48). The yeast prion Sup35, like Ure2, has an N-terminal PrD containing Asn/Gln repeats (49). Fibrils formed from polyglutamine peptides (46, 50, 51), Sup35 (49 52), or Ure2 (5, 6, 14) share many of the tinctorial, structural and mechanistic properties of typical amyloids. A striking and unusual property of Ure2, however, is the maintenance of native-like structural and functional properties within the amyloid-like fibrillar arrays (8, 13). Interestingly, fibrils formed by incubation of native, full-length Sup35 show globular appendages radiating from a central fibrillar core under transmission EM (52), highly reminiscent of the backbone model proposed for Ure2 fibril assembly (4, 5, 13, 53). Together, these results are consistent with a model for Ure2 fibril formation, in which the PrDs assemble hierarchically into the protofilaments and fibrils of typical amyloid, while still allowing the globular domains to be accommodated in a native-like state. CONCLUSIONS A pure solution of either of the mutants 15Ure2 or 15 42Ure2 is able to form amyloid-like fibrils morphologically indistinguishable from WT Ure2 but with an increased lag time, whereas 42Ure2 shows no detectable amyloid forming ability. This correlates with the importance of residues 1 43 of the Ure2 N-terminal PrD in mediating prion induction or curing in a WT background in vivo (2, 40, 44). The observation of a series of fibrillar intermediates during assembly of Ure2p fibrils supports the existence of a common, hierarchical mechanism for fibril assembly in all amyloidogenic proteins, even where native-like globular structure is accommodated within the fibrillar arrays.

127 Amyloid Nucleation and Hierarchical Assembly of Ure2 Fibrils 3369 Acknowledgments We thank Drs. L.S. Itzhaki, X. M. Pan, and L. Serrano for helpful discussions. REFERENCES 1. Wickner, R. B. (1994) Science 264, Masison, D. C., and Wickner, R. B. (1995) Science 270, Prusiner, S. B. (1998) Proc. Natl. Acad. Sci. U. S. A. 95, Speransky, V. V., Taylor, K. L., Edskes, H. K., Wickner, R. B., and Steven, A. C. (2001) J. Cell Biol. 153, Taylor, K. L., Cheng, N., Williams, R. W., Steven, A. C., and Wickner, R. B. (1999) Science 283, Thual, C., Komar, A. A., Bousset, L., Fernandez-Bellot, E., Cullin, C., and Melki, R. (1999) J. Biol. Chem. 274, Schlumpberger, M., Wille, H., Baldwin, M. A., Butler, D. A., Herskowitz, I., and Prusiner, S. B. (2000) Protein Sci. 9, Bousset, L., Thomson, N. H., Radford, S. E., and Melki, R. (2002) EMBO J. 21, Zhu, L., Zhang, X. J., Wang, L. Y., Zhou, J. M., and Perrett, S. (2003) J. Mol. Biol. 328, Levine, H. (1993) Protein Sci. 2, Khurana, R., Uversky, V. N., Nielsen, L., and Fink, A. L. (2001) J. Biol. Chem. 276, Sunde, M., Serpell, L. C., Bartlam, M., Fraser, P. E., Pepys, M. B., and Blake, C. C. (1997) J. Mol. Biol. 273, Baxa, U., Speransky, V., Steven, A. C., and Wickner, R. B. (2002) Proc. Natl. Acad. Sci. U. S. A. 99, Bousset, L., Briki, F., Doucet, J., and Melki, R. (2003) J. Struct. Biol. 141, Thual, C., Bousset, L., Komar, A., Walter, S., Buchner, J., Cullin, C., and Melki, R. (2001) Biochemistry 40, Perrett, S., Freeman, S. J., Butler, P. J. G., and Fersht, A. R. (1999) J. Mol. Biol. 290, Lesser, G. J., and Rose, G. D. (1990) Proteins 8, Li, X., and Pan, X. M. (2001) Proteins 42, Coshigano, P. W., and Magasanik, B. (1991) Mol. Cell. Biol. 11, Galani, D., Fersht, A. R., and Perrett, S. (2002) J. Mol. Biol. 315, Zhou, J. M., Zhu, L., Balny, C., and Perrett, S. (2001) Biochem. Biophys. Res. Commun. 287, Zhu, L., Kihara, H., Kojima, M., Zhou, J. M., and Perrett, S. (2003) Biochem. Biophys. Res. Commun. 311, Perutz, M. F. (1999) Trends Biol. Sci. 24, Perutz, M. F., Johnston, T., Suzuki, M., and Finch, J. T. (1994) Proc. Natl. Acad. Sci. U. S. A. 92, Perutz, M. F., and Windle, A. H. (2001) Nature 412, Perutz, M. F., Pope, B. J., Owen, D., Wanker, E. E., and Scherzinger, E. (2002) Proc. Natl. Acad. Sci. U. S. A. 99, Thakur, A. K., and Wetzel, R. (2002) Proc. Natl. Acad. Sci. U. S. A. 99, Bucciantini, M., Giannoni, E., Chiti, F., Baroni, F., Formigli, L. Zurdo, J., Taddei, N., Ramponi, G., Dobson, C. M., and Stefani, M. (2002) Nature 416, Nucifora, F., Sasaki, M., Peters, M. F., Huang, H., Cooper, J. K., Yamada, M., Takahashi, H., Tsuji, S., Troncoso, J., Dawson, V. L., Dawson, T. M., and Ross, C. A. (2001) Science 291, Chen, S., Berthelier, V., Yang, W., and Wetzel, R. (2001) J. Mol. Biol. 311, Haass, C., and Steiner, H. (2001) Nat. Neurosci. 4, Walsh, D. M., Klyubin, I., Fadeeva, J. V., Cullen, W. K., Anwyl, R., Wolfe, M. S., Rowan, M. J., and Selkoe, D. J. (2002) Nature 416, Uversky, V. N., Li, J., and Fink, A. L. (2001) J. Biol. Chem. 276, Jarrett, J. T., and Lansbury, P. T., Jr. (1993) Cell 73, Kelly, J. W. (2000) Nat. Struct. Biol. 7, Stine, W. B., Snyder, S. W., Ladror, U. S., Wade, W. S., Miller, M. F., Perun, T. J., Holzman, T. F., and Krafft, G. A. (1996) J. Protein Chem. 15, Ionescu-Zanetti, C., Khurana, R., Gillespie, J. R., Petrick, J. S., Trabachino, L. C., Minert, L. J., Carter, S. A., and Fink, A. L. (1999) Proc. Natl. Acad. Sci. U. S. A. 96, Naiki, H., Higuchi, K., Hosokawa, M., and Takeda, T. (1989) Anal. Biochem. 177, Goldsbury, C., Kistler, J., Aebi, U., Arvinte, T., and Cooper, G. J. S. (1999) J. Mol. Biol. 285, Masison, D. C., Maddelein, M. L., and Wickner, R. B. (1997) Proc. Natl. Acad. Sci. U. S. A. 94, Muller, D. J., Amrein, M., and Engel, A. (1997) J. Struct. Biol. 119, Harper, J. D., Wong, S. S., Lieber, C. M., and Lansbury P. T., Jr. (1999) Biochemistry 38, Chamberlain, A. K., MacPhee, C. E., Zurdo, J., Morozova-Roche, L. A., Hill, H. A. O., Dobson, C. M., and Davis, J. J. (2000) Biophys. J. 79, Edskes, H., and Wickner, R. B. (2002) Proc. Natl. Acad. Sci. U. S. A. 99, Edskes, H., Gray, V. T., and Wickner, R. B. (1999) Proc. Natl. Acad. Sci. U. S. A. 96, Chen, S., Berthelier, V., Hamilton, J. B., O Nuallain, B., and Wetzel, R. (2002) Biochemistry 41, Kad, N. M., Myers, S. L., Smith, D. P., Smith, D. A., Radford, S. E., and Thomson, N. H. (2003) J. Mol. Biol. 330, Khurana, R., Ionescu-Zanetti, C., Pope, M., Li, J., Nielson, L., Ramirez-Alvarado, M., Regan, L., Fink, A. L., and Carter, S. A. (2003) Biophys. J. 85, Serio, T. R., and Lindquist, S. L. (2001) Adv. Protein Chem. 59, Balbirnie, M., Grothe, R., and Eisenberg, D. S. (2001) Proc. Natl. Acad. Sci. U. S. A. 98, Perutz, M. F., Finch, J. T., Berriman, J., and Lesk, A. (2002) Proc. Natl. Acad. Sci. U. S. A. 99, Glover, J. R., Kowal, A. S., Schirmer, E. C., Patino, M. M., Liu, J. J., and Lindquist, S. (1997) Cell 89, Baxa, U., Taylor, K. L., Wall, J. S., Simon, M. N., Cheng, N., Wickner, R. B., and Steven, A. C. (2003) J. Biol. Chem. 278,

128 THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 279, No. 33, Issue of August 13, pp , by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A. Crystal Structure of Human eif3k, the First Structure of eif3 Subunits* Received for publication, May 10, 2004 Published, JBC Papers in Press, June 4, 2004, DOI /jbc.M Zhiyi Wei, Ping Zhang, Zhaocai Zhou, Zhongjun Cheng, Mao Wan, and Weimin Gong From the National Laboratory of Biomacromolecules, Institute of Biophysics, Chinese Academy of Sciences, Beijing , People s Republic of China and the School of Life Sciences, Key Laboratory of Structural Biology, University of Science and Technology of China, Hefei, Anhui , People s Republic of China eif3k, the smallest subunit of eukaryotic initiation factor 3 (eif3), interacts with several other subunits of eif3 and the 40 S ribosomal subunit. eif3k is conserved among high eukaryotes, including mammals, insects, and plants, and it is ubiquitously expressed in human tissues. Interestingly, eif3k does not exist in some species of yeast. Thus, eif3k may play a unique regulatory role in higher organisms. Here we report the crystal structure of human eif3k, the first high-resolution structure of an eif3 component. This novel structure contains two distinct domains, a HEAT (named for Huntington, elongation factor 3, A subunit of protein phosphatase 2A, target of rapamycin) repeat-like HAM (HEAT analogous motif) domain and a winged-helix-like WH domain. Through structural comparison and sequence conservation analysis, we show that eif3k has three putative protein-binding surfaces and has potential RNA binding activity. The structure provides key information for understanding the structure and function of the eif3 complex. Translation initiation is a sophisticated cellular process, especially in eukaryotes. In general, translation initiation in eukaryotic organisms involves three steps (1); first, the methionyl-initiator trna (Met-tRNA i Met ) binds to the 40 S ribosomal subunit to form a 43 S preinitiation complex; second, the preinitiation complex binds to mrna and scans to the AUG start codon in the mrna; and third, the 60 S ribosomal subunit joins the mrna-bound preinitiation complex to form an 80 S initiation complex, ready to commence translation. Each of these steps is stimulated by a number of proteins called eukaryotic initiation factors (eifs). 1 At least 11 eifs have been identified, * This work was supported by the Foundation for Authors of National Excellent Doctoral Dissertation of the People s Republic of China (Project No ), National Foundation of Talent Youth (Grant No ), the National High Technology Research and Development Program of China (Grant No. 2001AA233021), the 863 Special Program of China (Grant No. 2002BA711A13), the Key Important Project and other projects from the National Natural Science Foundation of China (Grant Nos , , , and ), and the Chinese Academy of Sciences (Grant No. KSCX1-SW-17). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. The atomic coordinates and structure factors (code 1RZ4) have been deposited in the Protein Data Bank, Research Collaboratory for Structural Bioinformatics, Rutgers University, New Brunswick, NJ ( These authors contributed equally to this work. To whom the correspondence should be addressed. wgong@sun5.ibp.ac.cn. 1 The abbreviations used are: eif, eukaryotic initiation factor; HTH, helix-turn-helix; HEAT, Huntington, elongation factor 3, A subunit of protein phosphatase 2A, target of rapamycin; HAM, HEAT analogous comprising over 25 polypeptides (2). In contrast, only three to five initiation factors are known in prokaryotes. This difference in protein complexity suggests that more protein-rna and protein-protein interactions rather than RNA-RNA interactions are required for efficient translation initiation in eukaryotic cells. In mammalian cells, eif3 is the largest initiation factor with an apparent molecular mass of about 600 kda. It plays a central role in steps 1 and 2 of the translation initiation process (1, 3). For instance, eif3 can bind to dissociated 40 S subunits and delay the reassociation with the 60 S ribosomal subunit for a long enough time to permit initiation. eif3 also stabilizes the binding of the Met-tRNA i Met eif2 GTP ternary complex to the 40 S subunits and promotes the formation of a 43 S preinitiation complex comprised of the 40 S subunit, the ternary complex, eif1, eif1a, and eif3. In addition, eif3 stimulates the binding of 5 -m 7 G-capped mrna by interaction with the mrna-associated factor eif4g. eif3 is a multisubunit protein complex. Various genes encoding eif3 subunits have been cloned from mammals, plants, and yeasts. Twelve different subunits (eif3a/p170, b/p116, c/p110, d/p66, e/p48, f/p47, g/p44, h/p40, i/p36, j/p45, k/p28, l/p69) have been identified in mammals, while in yeast only six subunits (eif3a/tif32, b/prt1, c/nip1, i/tif34, g/tif35, j/hcr1) have been found and are all homologous to the corresponding mammalian subunits. The five yeast subunits, eif3a, eif3b, eif3c, eif3g, and eif3i, can form a core complex (4). Mammalian and yeast eif3s differ not only in the number of subunits but also in the structure of some subunits. For example, mammalian eif3a contains a repeat region, but this region is absent in the yeast ortholog (2). These differences suggest that mammalian and plant eif3 have evolved to include additional subunits, which are likely to function as regulatory factors and the extra structural motifs provide the capacity to mediate extra protein-protein or protein-rna interactions required for the tighter regulation in higher eukaryotes. Some eif3 subunits (eif3a, eif3c, eif3e, eif3f, and eif3h) may serve as structural scaffolds or to provide docking sites for other proteins, since they contain PCI or MPN domains, which are found in components of large protein complexes and have been implicated in protein-protein interactions (2, 5). To understand the structure and function of the eif3 complex in eukaryotic translation initiation, we have begun a systematic structural study of the eif3 components. Here we describe the high-resolution crystal structure of human eif3k, the smallest non-core subunit of eif3. Mammalian eif3k has recently been characterized (6) to co-express with the five core subunits of eif3 and form a stable co-immunoprecipitatable complex with the core complex. eif3k also interacts directly motif; WH, winged-helix; PHAT, pseudo-heat analogous topology; PDB, Protein Data Bank. This paper is available on line at

129 34984 Crystal Structure of Human eif3k, the First Structure of eif3 Subunits TABLE I Summary of data collection and processing Numbers in parentheses represent the value for the highest resolution shell. MAD, multiwavelength anomalous dispersion. MAD data collection Edge Peak Remote Wavelength (Å) Resolution range (Å) ( ) ( ) ( ) No. of total reflections 116, , ,782 No. of unique reflections 12,780 (828) 12,597 (822) 9,690 (638) I/ 12.0 (3.7) 21.7 (9.5) 16.6 (7.3) Completeness (%) 98.7 (97.6) 100 (100) 100 (99.8) R merge (0.300) (0.211) (0.269) with eif3c, eif3g, and eif3j. The crystal structure of eif3k reveals a novel ear-like protein structure containing two domains. Structural comparisons and sequence conservation analysis suggest that eif3k is likely to act as a structural scaffold for protein-protein and protein-rna interactions. It has three putative protein-binding regions and has potential RNA binding activity associated with its HTH (helix-turn-helix) motif. EXPERIMENTAL PROCEDURES Expression and Purification of Human eif3k The complete cdna fragment encoding human eif3k protein was subcloned from a human brain cdna library into a pet-22b expression vector, and human eif3k was expressed highly as a soluble protein in Escherichia coli strain BL21 (DE3) with a 6-residue histidine tag fused to its C terminus. Purification of the eif3k protein was carried out through affinity chromatography with a Chelating Sepharose TM Fast Flow (Amersham Biosciences). For phase determination, the recombinant plasmid was transferred into Met-auxotrophic strain B834 to obtain the selenomethionyl derivative of human eif3k protein. Crystallization and X-ray Data Collection Crystals of seleno-methionine-substituted eif3k protein were grown using the hanging drop vapor diffusion method at 4 C. The initial crystallization conditions (2.0 M ammonium sulfate, 0.1 M Hepes at ph 7.5, 0.1 M NaCl) were obtained using Crystal Screen kits I and II from Hampton Research. By optimizing the concentration of ammonium sulfate to 1.6 M, crystals of larger size and better quality were obtained. A multiwavelength anomalous dispersion data set was collected from a single seleno-methioninesubstituted eif3k crystal at 100 K on beamline 3W1A of the Beijing Synchrotron Radiation Facility at the Institute of High Energy Physics, Chinese Academy of Sciences. The data were collected at three wavelengths ( edge Å, peak Å, and remote 0.9 Å). All data were processed and scaled with the DENZO and SCALEPACK (7), respectively. Data collection statistics are presented in Table I. Phasing, Model Building, and Refinement Six of seven expected selenium positions were found by SOLVE (8) using the three data sets, and the initial phase was calculated to 2.5 Å. RESOLVE (9, 10) was used for density modification and building of the initial model of human eif3k. The initial model containing about 75% of the residues was refined against the peak data set in the Å resolution range with maximum-likehood amplitude targets by using the Crystallography and NMR System (11). Subsequently, the refinement was extended to resolution bins of 2.3 Å with iterative manual adjustments and rebuilding of the model using the program O (12) and 2F o F c and F o F c electron density maps as references. A sulfate ion and water molecules were added to the model when the resolution extended to 2.1 Å, and the value of R free is about 30%. After that, individual atomic B factors were refined. Finally, the model was checked for errors with simulated annealing omit maps covering a 10-residue segment of the structure at a time. The stereochemical quality of the final model was checked by PROCHECK (13), and the final refinement statistics and geometry are excellent (Table II). Sequence and Structure Analysis Seven eif3k-related sequences were identified in the sequence data base using BLAST (14); multiple sequences alignment was done with T-coffee (15); the similarity score of residues were calculated using Blosum62 matrix by ESPript (16); structural similarity searches were performed with DALI (17); and electrostatic surface potentials were calculated with MOLMOL (18). Figs. 1, 3, 4, 5, and 7 were prepared using Ribbons (19); Fig. 2 was prepared with ESPript; and Fig. 6 was prepared using MOLMOL. 129 TABLE II Structural refinement statistics Numbers in parentheses represent the value for the highest resolution shell. Space group P Unit cell dimensions (Å) Resolution (Å) ( ) R work (%) a 18.7 (18.7) R free (%) b 22.2 (27.3) No. of reflections Working set 11,090 Test set 1,276 No. of atoms Protein atoms 1,714 (including 6 selenium atoms) Water molecules 149 Sulfate ions 5 in 1 sulfate ions r.m.s.d. from ideality Bond lengths (Å) Bond angles ( ) 1.1 Average B factor (Å 2 ) Main chain 19.1 Side chain 22.8 Water 29.0 Sulfate ion 43.6 Ramachandran plot Most favored regions (%) 94.9 Additionally allowed (%) 5.1 a R work F obs F calc / F obs, where F obs and F calc are observed and calculated structure factors. b R free T F obs F calc / T F obs, where T is a test data set of 10% of the total reflections randomly chosen and set aside prior to refinement. RESULTS Overall Structure The human eif3k structure was determined by multiwavelength anomalous dispersion with selenomethionine substituted protein. The model was refined to 2.1-Å resolution with a good agreement with diffraction data and has high stereochemical quality (Table II). The final model includes an eif3k monomer with 213 residues (residue and ), 6 selenium atoms, 1 sulfate ion, and 149 water molecules. The C-terminal 2 residues and the polyhistidine tag are disordered. No electron density for residues Gly 184 and Ser 185 and the side chains of Gln 119, Asp 182, Glu 183, and Ser 216 was observed. In addition, residues Phe 56, Glu 123, Trp 156, Ile 196, Lys 197, and Lys 204 have weak electron density. The overall structure of eif3k has an ear-like shape with dimensions of 35 Å 35 Å 60Å. It has 16 helices (14 helices and two 3 10 helices) and a three-strand -sheet (Fig. 1). The secondary structure elements are indicated in Fig. 2. Although sequence analysis did not suggest any known structural motifs, a three-dimensional structural comparison of the crystal structure of eif3k determined here using DALI (17) clearly showed that eif3k consists of two distinct domains. The N-terminal region of eif3k is a HAM domain, named for HEAT (20) analogous motifs, in a mostly right-handed superhelical arrangement, formed by a leading helix and three HEAT analogous repeats, H1 (consisting of 3 and 4), H2 ( 6 and 7), and H3

130 Crystal Structure of Human eif3k, the First Structure of eif3 Subunits FIG. 1. The overall structure of human eif3k. Human eif3k contains a HAM domain (green) and a WH domain (blue) followed by a C-terminal tail (yellow). The HAM domain consists of eight helices ( ) and two 3 10 helices (h); the WH domain consists of three helices and three strands ( ). The disordered region is indicated by a broken line connecting 2 and 3. The loop between h5 and 6 is overlapped by 16. The view in B is rotated 90 around a vertical axis from the view in A. ( 9 and 10). All of the repeats and the leading helix are followed by a short helix. The C-terminal half contains a WH (winged-helix) domain, which is followed by a long C-terminal tail flanked by an helix at both ends. The WH domain is a compact / structure containing three helices ( 12, 13, and 14) and three strands ( 1, 2 and 3). Approximately of solvent-accessible surface area is buried by the HAM-WH interdomain interactions. The combination of the two domains and C-terminal tail results in two distinct faces of the protein surface: a concave face and a convex face (Fig. 6). The HAM Domain; Comparison with HEAT-repeat-containing Proteins HEAT-repeat-containing proteins such as the PR65/A subunit of protein phosphatase 2A (21), nuclear transport protein karyopherin- 2 (22), and -adaptin C, a subunit of the AP2 complex (23), play important roles in assembling multiprotein complexes in various cellular life activities (24). The HEAT repeat motif is formed by two antiparallel helices (named A and B), and it is usually residues long. The motif occurs in blocks of at least 3 and up to 22 tandem repeats (20). HEAT repeat family members have similar patterns of hydrophobic residues, and a highly conserved proline is frequently found in certain A helices located in the middle of many HEAT repeats to facilitate the kink of the helices (20). The human eif3k HAM domain (residues ) is generally similar to that in HEAT-repeat-containing proteins, but it also shows some unusual features. Its three HEAT analogous motifs have 21 (H3) to 32 (H1 and H2) residues. Adjacent HEAT analogous motifs are connected by a short 3 10 helix (h5, between H1 and H2) or a short helix ( 8, between H2 and H3). Compared with the canonical HEAT-repeat-containing protein PR65/A, two and a half HEAT analogous repeats ( 3, 4, 6, 7, and 9) in the HAM domain resemble the HEAT repeats 13 and 14 and A of repeat 15 of PR65/A, and they can be superimposed with a r.m.s.d. (root mean square deviation) of 3.2 for 84 residues in the region (Fig. 3). An analysis of the protein sequences (data not shown) shows that several highly 130 conserved hydrophobic residues in the HEAT analogous motifs of the HAM domain occupy the same positions as the corresponding residues in HEAT repeats (20, 25). However, unlike most other HEAT repeats, no proline residues are found in the eif3k helices ( 3, 6, and 9) corresponding to the A of HEAT repeats. As a consequence, the helices in the HAM domain are not kinked. These differences between the HAM domain and other HEAT domains suggest that the HEAT analogous motifs in the HAM domain are related but distinct, and they may belong to the new subfamily of ADB (named for -adatin; Ref. 25). The absence of a consensus proline residue in A isa unique feature of the ADB repeat in the HEAT repeats family. This is exemplified by the high similarity (Z score 6.4 by DALI) between the HAM domain and the HEAT motifs (residues Ala 340 to Ala 434 )in -adaptin C, an ADB subfamily member. Nevertheless, it is possible that the HAM domain may belong to a novel class in HEAT repeats as H1, H2, and H3 are structurally different when these three HEAT analogous motifs are compared. The intra-repeat connecting helices in the HAM domain do not exist in the PR65/A HEAT repeats and most HEAT repeats of other proteins (Fig. 3). This unusual topology is also found in a few heat-repeat-containing proteins, such as the PHAT (pseudo-heat analogous topology) domain of Smaug (26) (Fig. 3). The two and a half HEAT analogous repeats of the HAM domain are more similar to the PHAT domain (r.m.s.d. 2.7 Å for 84 residues) than to the PR65/A HEAT repeats (Fig. 3). Another unusual conformation in the HAM domain is the packing between H3 and H2, which is almost side-by-side instead of face-to-face, as in normal HEAT repeats, due to its unusual left-handedness. The unusual H3 turn causes 6, 9, 10, and 12 to form a hydrophobic core (Fig. 4). A similar type of conformation also appears in the structure of the C-terminal region of the vesicular transport protein Sec17 (27) (Fig. 3). In addition, the eif3k HAM domain also shows some similarity to several other proteins containing antiparallel helical

131 34986 Crystal Structure of Human eif3k, the First Structure of eif3 Subunits FIG. 2. Sequence alignment of eif3k homologues. Protein sequences were from Homo sapiens (human), Mus musculus (house mouse), Drosophila melanogaster, Anopheles gambiae, Caenorhabditis elegans, Oryza sativa, and Arabidopsis thaliana. Sequence accession numbers of the Swiss-Prot data base are Q9UBQ5 (for human), Q9DBZ5 (for mouse), Q9W2D9 (for D. melanogaster), Q7QGK4 (for A. gambiae), Q9XUP3 (for C. elegans), Q94HF1 (for O. sativa), and Q9SZA3 (for A. thaiana). The sequence of A. gambiae is derived from an EMBL/GenBank TM /DDBJ whole genome shotgun entry, which should be considered preliminary data. The secondary structure of human eif3k, which is mainly defined by the analysis of the structure using DSSP program (47), is indicated above the alignment. Residues in the alignment that are identical are shown in red boxes; those that are similar are shown in yellow boxes. The sequences highlighted in a pink box correspond to the disordered regions (Ser 184, Gly 185, Ser 217, and Gln 218 ) missing from the human eif3k structure. The characters under the alignment denote the predicted function of the conserved residues. The elucidations for these characters are listed as follows: 1, 2, and 3, the residues on the binding surface I, II, and III, respectively (Fig. 6); h, intra-molecular hydrophobic interaction (h colored in red denotes residues that participate in forming the hydrophobic core between the HAM domain and the WH domain shown in Fig. 4); s, the two highly conserved residues that form a salt bridges (Fig. 7); t, the conserved glycines that make turns between the secondary structure elements; orange triangle, the residues that are special in C. elegans compared with other organisms (hydrophobic residues in other organisms were replaced by hydrophilic residues in C. elegans or in reverse). The sequence of C. elegans eif3k has distinct property in some position. For instance, the residues that are indicated by the orange triangle are very different between C. elegans and other organisms; and an additional sequence in the orange box between 10 and 11 that belong to the HAM domain and WH domain, respectively, exist in C. elegans only. 131

132 Crystal Structure of Human eif3k, the First Structure of eif3 Subunits FIG. 3.Comparison of the HAM domain with other HEAT or HEAT-like repeats. The HAM domain, PR65/A HEAT repeats, and the Sec17 C-terminal region all contain three antiparallel -helical repeats, while the PHAT domain of Smaug contains only two and a half antiparallel -helical repeats. A, a top view of four HEAT/HEAT analogous repeats; B, a lateral view. The red arrows denote the places of the locations of the helices connecting HEAT analogous motifs in the HAM domain and in the PHAT domain. The left-handed superhelix at the C termini of the HAM domain and the Sec17 is seen very clearly from the top view (A). FIG. 4. The hydrophobic core between the HAM and WH domains. Highly conserved hydrophobic residues in 6, 9, 10, and 12 form a main part of the core. The core-forming residues are showed in a ball-and-stick models; they include Leu 128, Ile 131, Leu 71, Phe 134, Leu 74, Val 138, Leu 77, and Phe 141 on the left side in an order from the top down and on the right side are Ile 102, Leu 121, Leu 105, Trp 118, Phe 117, Leu 109, and Phe 114. All of these residues are highly conserved (Fig. 2) except Leu 77, Leu 121, and Leu 128 (in 11), although the residues corresponding to these three are also hydrophobic among eif3k homologues in other species (Fig. 2). repeats, including tetratricopeptide repeats and armadillo repeats that are both involved in protein-protein interactions (24), and to part of the helix domain (residues ) of chondroitin AC lyase (28). The WH Domain; Conformation of the WH Domain The WH domain has the appearance of an earlobe in an ear-like structure (Fig. 1). This domain is stably packed against the HAM domain via the hydrophobic core formed mainly by the conserved hydrophobic residues in 6, 9, 10, and 12 (Fig. 4). The WH domain comprises three helices and three strands in the order of and 13 are antiparallel to each other, while 14 is almost perpendicular to 13 (Fig. 5). The latter two helices are connected by a short turn of three amino acids. This architecture belongs to a HTH motif 132 (29). The three short strands, each containing only 2 residues, form a twisted antiparallel -sheet, with 1 bonding to the hairpin formed between 2 and 3, which are bonded by exactly two hydrogen bonds. This arrangement of strands is a common feature of HTH motifs with a / topology (named winged-helix motifs). The hydrophobic residues in the helices, together with Trp 179 and Ile 189, interdigitate to form another hydrophobic core stabilizing the architecture of the WH domain. Canonical winged-helix motifs commonly have two large loops or wings (w1 and w2). Wing w1 connects 2 and 3, and wing w2 extends from strand 3 to the C terminus of the winged-helix domain (30). In the WH domain of eif3k, wing w1 is very short (residues ) (Fig. 5); and two residues, 191 and 192, are disordered in the structure. In addition, wing w2 is replace by the helix ( 15) of the C-terminal tail. Comparison with Other Winged-helix Motifs A structural similarity search in the PDB (Protein Data Bank) with DALI using the WH domain (residues ) resulted in a number of matches to proteins containing winged-helix motifs. The most similar proteins found by DALI are the selenocysteinespecific elongation factor SelB (PDB code: 1LVA; Z score 6.5 and r.m.s.d. 2.4 for 56 residues) (31), the purine operon repressor of Bacillus subtilis (PDB code: 1P41; Z score 6.1 and r.m.s.d. 1.9 for 54 residues) (32), Rap30 DNA-binding domain (PDB code: 2BBY; Z score 5.4 and r.m.s.d. 2.4 for 53 residues) (33), IclR transcriptional factor (PDB code: 1JMR; Z score 5.3 and r.m.s.d. 2.7 for 55 residues) (34), endonuclease FokI (PDB code: 2FOK; Z score 5.2 and r.m.s.d. 2.4 for 58 residues) (35), double-stranded RNA-specific adenosine deaminase (PDB code: 1QBJ; Z score 5.2 and r.m.s.d. 2.7 for 54 residues) (36), and the Esa1 histone acetyltransferase domain (PDB code: 1FY7; Z score 5.2 and r.m.s.d. 2.6 for 56 residues) (37). All of winged-helix motifs in these proteins, including the WH domain, differ in the length of the third helix (corresponding to 14 in the WH domain) (Fig. 5), which is associated with major-groove DNA binding and is called the recognition helix. The hairpin (the wing w1) is also involved in DNA binding via the DNA backbone and/or the minor groove.

133 34988 Crystal Structure of Human eif3k, the First Structure of eif3 Subunits FIG. 5.A comparison of winged-helix domains. The recognition helix in the WH domain is 14, and W1 indicates wing w1 in the WH domain. The two winged-helix domains on the upper left side (SelB-C and Esa1) do not have DNA/RNA binding activity, while the two winged-helix domains on upper rightside (FokI-D3 and Rap30) interact weakly with DNA and are also involved in protein-protein interactions. The lower three winged-helix domains (PurR, Z domain, and IclR) are all involved in DNA binding. The disordered region in the WH domain is indicated by a broken line. DISCUSSION The HAM Domain Can Supply Binding Surfaces for Protein- Protein Interactions HEAT repeats have been proposed to mediate protein-protein interactions (38, 39), and the internal repetition enlarges the available protein binding surface area. Another eukaryotic translation initiation factor, eif4gii, also contains a HEAT domain that interacts with eif4a (40) and eif1 and eif5 (41). The detailed binding regions in HEAT motifs have not yet been delineated, but the loops connecting A with B (AB loop) may be the binding surface based on sequence conservation, mutational analysis, and oncogenic mutation studies (21). The structure of the tetra-protein complex AP2 (23) provides a model for understanding how proteins bind the AB loop. In the HAM domain of human eif3k, the AB loops help the formation of the concave face that might be suitable for protein-protein interactions (see below). It is worth mentioning that in the protein PR65/A, the intra-repeat turns of repeats 13 15, which are similar to that in the HAM domain (Fig. 3), were found to be the protein-binding site (21, 42). Although the intra-repeat turns in the HEAT repeats are replaced by helices (h2, h5, and 8) in the HAM domain, the location of conserved hydrophobic residues on these helices (Fig. 2) suggests that they could still serve as the protein binding surface in eif3k. These differences might be helpful to enlarge the protein-binding surface of eif3k. The presence of HEAT analogous motifs in eif3k strongly suggests that the HAM domain is essential for interactions between eif3k and other proteins, especially other eif3 subunits that have been shown to interact with eif3k directly by glutathione S-transferase pull-down assays (6). Conserved Residues in eif3k Suggest Three Binding Surfaces To gain insights into the mechanisms by which eif3k 133 interacts with other proteins, we combined the structural knowledge with sequence conservation to identify potential protein-protein interaction sites on eif3k. We aligned the sequences of seven eif3k homologues from different organisms that are available in public protein databases (Fig. 2). And we found that many highly conserved residues are distributed throughout the whole protein. The conserved residues can be divided into three classes: hydrophilic amino acids (class I), hydrophobic amino acids involved in inside hydrophobic interactions (class II), and hydrophobic amino acids exposed on the surface (class III). The residues in class II are mainly located at the interface of helices that interdigitate to maintain the structure of eif3k. Most of these three classes of residues are marked in Fig. 2 with their predicted function labeled. Previous experiments indicated that eif3k could interact with several other eif3 subunits (eif3c, eif3g, and eif3j), as well as the 40 S ribosomal subunit (6). Thus, the aforementioned class I and class III residues can serve as potential binding sites for the other proteins. Fig. 6 shows the distribution of the conserved residues on the solvent-accessible protein surface. And a surface representation showing the surface electronic potential distribution is display side-by-side for comparison. It is interesting to note that the concave side of the protein could be a putative binding surface, which we call binding surface I, for other proteins. On the binding surface I, there are 14 conserved residues, including 6 electronegative residues, Asp 43, Glu 45, Asp 81, Asp 135, Asp 136, and Glu 203 ; 2 electropositive residues, Lys 197 and Lys 199 ; 3 class III residues, Tyr 42, Pro 78, and Leu 84 ; and some hydrophilic residues (Fig. 6, A and B). Thus, this concave face may bind a protein through hydrogen bonds (or salt bridges) and hydrophobic interactions. Considering this is the largest potential binding surface, we propose that this region be

134 Crystal Structure of Human eif3k, the First Structure of eif3 Subunits FIG. 6. Surface distributions of conserved residues and electrostatic potential of eif3k. A, conserved residues are mapped onto the concave face (colored light green). The conserved residues on the binding surface I are marked by red numbers. The binding surface I was circled in brief, and it appears to be capable of interacting with a rather large protein due to its large area. B, the electrostatic potential surface on the concave face of eif3k is colored red for electronegative residues and blue for electropositive residues. C, conserved residues are mapped onto the convex face. There are two putative binding surfaces (binding surface II and III) for other proteins on the convex face. D, the electrostatic potential surface on the convex face of eif3k. The charged residues are fewer in this face than in concave face. Obviously, there are more electronegative residues than electropositive residues on molecular surface. This is the reason why human eif3k is an acidic protein with pi 4.8 (6). All the conserved residues located on the surface could be found in Fig. 2 (labeled with numbers 1, 2, and 3). FIG. 7.The cleft in the WH domain. The cleft is composed mainly of 12 and 14, the first two helices of the WH domain. The salt bridge between the N of Arg 155 and a carboxyl oxygen atom of Asp 167 is on the left side of the cleft (shown as a dotted line) with an interatomic distance of 2.72 Å. There are two conserved residues, Asp 136 and Arg 139,at the bottom of the cleft and a sulfate ion at the hatch of the cleft. involved in the interaction between eif3k and eif3c, the biggest subunit among eif3c, eif3g, and eif3j. The other class III residues are located in a bipartite region of the convex face (Fig. 6C). The residues in or around h2 and h5 connecting 1 and H1, H1, and H2, respectively, are reasonably well conserved (Fig. 2). Among these residues, the hydrophobic residues Ile 18, Tyr 21, Pro 23, Phe 56 (at the C-terminal end of 4), and Pro 58 belong to class III. The class III residues Ile 205, Phe 207, and Val 210 are located in the C-terminal 134 tail of eif3k whose surface connects with the surface of h2 and h5 to form a rather large protein-binding surface (binding surface II, Fig. 6C). The distribution of these highly conserved residues on the protein surface is consistent with their potential involvement in protein-protein interactions mainly through hydrophobic interactions. The third binding surface (binding surface III) is in the WH domain comprising Thr149, Tyr 150, Gln 151, and Gln 192 (Fig. 6C). Binding surface III may be responsible for the protein

135 34990 Crystal Structure of Human eif3k, the First Structure of eif3 Subunits binding function of the WH domain. In the human eif3k structure, there is a salt bridge between the N of Arg 155 and a carboxyl oxygen atom of Asp 167 (Fig. 7), both of which are highly conserved with basic residues (arginine and lysine) and acidic residues (asparagine and glutamic acid) respectively. To our knowledge, this is unique to eif3k, as no salt bridges are found in the similar region of other wingedhelix motifs. The stabilization of this salt bridge may affect the shape of a cleft where a sulfate ion and two unidentified electron density masses are present. The two conserved residues, Asp 136 and Arg 139, are at the bottom of the cleft (Fig. 7). The significance of this conformation is still unknown. The Winged-helix in eif3k; RNA Binding or Protein Docking The winged-helix motif is a subfamily of the HTH family, which includes many DNA-binding proteins. However, the structure of the phylogenetically conserved core of the signal recognition particle revealed that a HTH motif could also bind RNA (43). The C-terminal fragment of the elongation factor SelB (SelB-C) provided another example of RNA binding by winged-helix domains (31). As pointed out earlier, eif3 is required for the binding of the 40 S ribosomal subunit at the 5 -end of mrna, and it stimulates the binding of the ternary complex, comprised of eif2, Met-tRNA, and GTP, to the 40 S subunit to form the preinitiation complex (44, 45). These properties of eif3 provide an intriguing possibility that the eif3k WH domain may be an RNA binding domain. Our surprising structural finding that eif3k contains a winged-helix domain points to a potentially interesting scenario that it may be involved in RNA binding. However, the RNA binding activity of eif3k has not yet been demonstrated experimentally. Alternative activities other than RNA binding involving the eif3k WH domain are also possible. A careful examination of the eif3k WH domain shows that it differs from canonical winged-helix motifs in details. For instance, in well characterized winged-helix motifs, the recognition helix interacts with DNA, and it contains conserved basic residues within the helix or regions immediately surrounding the helix. The corresponding eif3k helix does not contain conserved basic residues in or around 14 (Fig. 2). Furthermore, not all winged-helix motifs have documented nucleic acids binding activities. For instance, the winged-helix motif in the yeast Esa1 histone acetyltransferase domain is known to be involved in the binding of CoA (37), but its DNA or RNA binding activity has yet to be demonstrated; the D3 domain of FokI, also a winged-helix motif that is highly similar to the WH domain, was found to barely touch DNA, and its putative recognition helix did not interact with the major groove of DNA, thought to be important for protein-protein interactions (46); the Rap30 DNA-binding domain may also be involved in the assembly of the transcription preinitiation complex (33). Thus, the question of whether the recognition helix in eif3k interacts with RNA, or its WH domain is a proteindocking domain, remains to be clarified. CONCLUSIONS We have determined the crystal structure of human eif3k, which is the first eif3 subunit structure reported. It contains two distinct domains: a HAM domain and a WH domain. From the results of structure comparisons and residues conservation analysis, we postulate that eif3k might serve as a structural scaffold, with potential surfaces for binding to other eif3 subunits and to RNA. Our work provides a starting point for further systematic biochemical, genetic, and structural studies aimed at understanding the structure and function of eif3 complex and at elucidation of the mechanism of translation initiation. Acknowledgments We thank Dr. Peng Liu, Dr. Yuhui Dong, and Deqiang Yao from the Institute of High Energy Physics, Chinese Academy of Sciences for their kind help in diffraction data collection. We also thank Prof. R. Xu and Prof. J. W. B. Hershey for their constructive criticism and suggestions. REFERENCES 1. Pain, V. M. (1996) Eur. J. Biochem. 236, Hershey, J. W. B. & Merrick, W. C. (2000) in Translational Control of Gene Expression (Sonenberg N., Hershey, J. W. B. & Mathews, M. B., eds) pp , Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY 3. Hinnebusch, A. G. (2000) in Translational Control of Gene Expression (Sonenberg, N., Hershey, J. W. B. & Mathews, M. B., eds) pp , Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY 4. Phan L., Zhang X., Asano K., Anderson J., Vornlocher H.-P., Greenberg J. R., Goldfarb D. S., Qin J. & Hinnebusch A. G. (1998) Mol. Cell. Biol. 18, Kim T.-H., Hofmann K., Arnim A. G. & Chamovitz D. A. (2001) Trends Plant Sci. 6, Mayeur, G. L., Fraser, C. S., Peiretti, F., Block, K. L. & Hershey, J. W. B. (2003) Eur. J. Biochem. 270, Otwinowski, Z. & Minor, W. (1997) Methods Enzymol. 276, Terwilliger, T. C. & Berendzen, J. (1999) Acta Crystallogr. Sect. D Biol. Crystallogr. 55, Terwilliger, T. C. (2000) Acta Crystallogr. Sect. D Biol. Crystallogr. 56, Terwilliger, T. C. (2002) Acta Crystallogr. Sect. D Biol. Crystallogr. 59, Brunger, A. T., Adams, P. D., Clore, G. M., DeLano, W. L., Gros, P., Grosse- Kunstleve, R. W., Jiang, J. S., Kuszewski, J., Nilges, M., Pannu, N. S., Read, R. J., Rice, L. M., Simonson, T. & Warren, G. L. (1998) Acta Crystallogr. Sect. D Biol. Crystallogr. 54, Jones, T. A., Zou, J. Y., Cowan, S. W. & Kjeldgaard M. (1991) Acta Crystallogr. Sect. A 47, Laskowski, R. A., MacArthur, M. W., Moss D. S. & Thornton J. M. J. (1993) Appl. Crystallogr. 26, Altschul, S. F., Madden, T. L., Schäffer, A. A., Zhang, J., Zhang, Z., Miller, W. & Lipman, D. J. (1997) Nucleic Acids Res. 25, Notredame, C., Higgins, D. G. & Heringa, J. (2000) J. Mol. Biol. 302, Gouet, P., Courcelle, E., Stuart, D. I. & Metoz, F. (1999) Bioinformatics 15, Holm, L. & Sander, C. (1993) J. Mol. Biol. 233, Koradi, R. & Billeter, M. (1998) PDB Newslett. 84, Carson, M. (1997) Methods Enzymol. 277, Kobe, B., Gleichmann, T., Horne, J., Jennings, I. G., Scotney, P. D. & Teh, T. (1999) Structure (Lond.) 7, Groves, M. R., Hanlon, N., Turowski, P., Hemmings, B. A. & Barford, D. (1999) Proc. Natl. Acad. Sci. U. S. A. 86, Chook, Y. M. & Blobel, G. (1999) Nature 399, Collins, B. M., McCoy, A. J., Kent, H. M., Evans, P. R. & Owen, D. J. (2002) Cell 109, Peifer, M., Berg, S. & Reynolds, A. B. (1996) Cell 76, Andrade, M. A., Petosa, C., O Donoghue, S. I., Müller, C. W. & Bork, P. (2001) J. Mol. Biol. 309, Green, J. B., Gardner, C. D., Wharton, R. P. & Aggarwal, A. K. (2003) Mol. Cell 11, Rice, L. M. & Brunger, A. T. (1999) Mol. Cell 4, Féthière, J., Eggimann B. & Cygler, M. (1999) J. Mol. Biol. 288, Brennan, R. G. & Matthews, B. W. (1989) J. Biol. Chem. 264, Gajiwala, K. S. & Burley, S. K. (2000) Curr. Opin. Struct. Biol. 10, Selmer, M. & Su, X.-D. (2002) EMBO J. 21, Sinha, S. C., Krahn, J., Shin, B. S., Tomchick, D. R., Zalkin, H. & Smith J. L. (2003) J. Bacteriol. 185, Groft, C. M., Uljon, S. N., Wang, R. & Werner M. H. (1998) Proc. Natl. Acad. Sci. U. S. A. 95, Zhang, R.-G., Kim, Y., Skarina, T., Beasley, S., Laskowski, R., Arrowsmith, C., Edwards, A., Joachimiak, A. & Savchenko, A. (2002) J. Biol. Chem. 277, Wah, D. A., Bitinaite, J., Schildkraut, I. & Aggarwal, A. K. (1998) Proc. Natl. Acad. Sci. U. S. A. 95, Schwartz, T., Rould, M. A., Lowenhaupt, K., Herbert, A. & Rich, A. (1999) Science 284, Yan, Y., Barlev, N. A., Haley, R. H., Berger, S. L. & Marmorstein, R. (2000) Mol. Cell 6, Blatch, G. L. & Lässle, M. (1999) BioEssay 21, Andrade, M. A., Perez-Iratxeta, C. & Ponting, C. P. (2001) J. Struct. Biol. 134, Marcotrigiano, J., Lomakin, I. B., Sonenberg, N., Pestova, T. V., Hellen, C. U. T. & Burley, S. K. (2001) Mol. Cell 7, He, H., von der Haar, T., Singh, C. R., Ii, M., Li, B., Hinnebusch, A. G., McCarthy, J. E. G. & Asano, K. (2003) Mol. Cell. Biol. 23, Ruediger, R., Hentz, Fait, J., Mumby, M. & Gernot, W. (1994) J. Virol. 68, Batey, R. T., Rambo, R. P., Lucast, L., Rha, B. & Doudna, J. A. (2000) Science 287, Chaudhuri, J., Chowdhury, D. & Maitra, U. (1999) J. Biol. Chem. 274, Pestova, T. V., Borukhov, S. I. & Hellen, C. U. T. (1998) Nature 394, Wah, D. A., Hirsch, J. A., Dorner, L. F., Schildkraut, I. & Aggarwal, A. K. (1997) Nature 388, Kabsch, W. & Sander, C. (1983) Biopolymers 22,

136 THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 279, No. 17, Issue of April 23, pp , by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A. Conformational Changes in the Reaction of Pyridoxal Kinase* Received for publication, November 12, 2003, and in revised form, February 2, 2004 Published, JBC Papers in Press, February 5, 2004, DOI /jbc.M Ming-hui Li, Francis Kwok, Wen-rui Chang, Sheng-quan Liu, Samuel C. L. Lo, Ji-ping Zhang, Tao Jiang, and Dong-cai Liang From the National Laboratory of Biomacromolecules, Institute of Biophysics, Chinese Academy of Sciences, Beijing , China, Department of Applied Biology Chemical Technology, Hong Kong, China and Polytechnic University, Hung Hom, Kowloon, Hong Kong, China To understand the processes involved in the catalytic mechanism of pyridoxal kinase (PLK), 1 we determined the crystal structures of PLK AMP-PCP-pyridoxamine, PLK ADP PLP, and PLK ADP complexes. Comparisons of these structures have revealed that PLK exhibits different conformations during its catalytic process. After the binding of AMP-PCP (an analogue that replaced ATP) and pyridoxamine to PLK, this enzyme retains a conformation similar to that of the PLK ATP complex. The distance between the reacting groups of the two substrates is 5.8 Å apart, indicating that the position of ATP is not favorable to spontaneous transfer of its phosphate group. However, the structure of PLK ADP PLP complex exhibited significant changes in both the conformation of the enzyme and the location of the ligands at the active site. Therefore, it appears that after binding of both substrates, the enzyme-substrate complex requires changes in the protein structure to enable the transfer of the phosphate group from ATP to vitamin B 6. Furthermore, a conformation of the enzyme-substrate complex before the transition state of the enzymatic reaction was also hypothesized. Pyridoxal kinase (PLK) 1 catalyzes the phosphorylation of vitamin B6 (including pyridoxal, pyridoxine, and pyridoxamine) in the presence of ATP and Zn 2, which is an essential step in the synthesis of pyridoxal 5 -phosphate (PLP), an active form of the vitamin in mammals (1 3). PLK is expressed in all mammalian tissues because of the fact that PLP cannot cross cell membranes, and PLK is required for the activation process inside cells (4). Genes encoding PLK have been cloned from both mammalian and plant cells. PLK activity has also been detected in bacteria, because PLP can be synthesized through the PLP salvage pathway (5, 6). * This work was supported by the National Natural Science Foundation of China (Grant ), the Life Science Special Fund of Chinese Academy of Services (Grant STZ0017), and the National Key Research Development Project of China (Grant G ). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. The atomic coordinates and structure factors (code 1RFT for PLK AMP-PCP-pyridoxamine, 1RFU for PLK ADP PLP, and 1RFV for PLK ADP) have been deposited in the Protein Data Bank, Research Collaboratory for Structural Bioinformatics, Rutgers University, New Brunswick, NJ ( To whom correspondence should be addressed: National Laboratory of Biomacromolecules, Institute of Biophysics, Chinese Academy of Sciences, 15 Datun Rd., Chaoyang District, Beijing , China. Tel.: ; Fax: ; x-ray@sun5.ibp.ac.cn. 1 The abbreviations used are: PLK, pyridoxal kinase; PLP, pyridoxal 5 -phosphate; AMP-PCP, adenosine 5 -(, -methylenetriphosphate); r.m.s.d., root mean square deviation. This paper is available on line at Recently, the three-dimensional structures of PLK from sheep brain and its complex with ATP were determined (7). Although structural analyses have shown that PLK exhibits a folding pattern similar to the core structure of enzymes in the ribokinase superfamily (8 13), low sequence homology between the two types of enzymes has been found. Despite kinetic studies that have shown that ribokinase and adenosine kinase both follow an ordered substrate-binding mechanism, PLK binds ATP and pyridoxal randomly (8, 10, 14). During the binding of ATP, a flexible loop containing 12 amino acid residues in the active site of PLK was responsible for triggering a major conformational change of the protein structure by interacting with the bound ATP. It has been suggested that the purpose for the inability of ATP to interact with this loop before catalysis is to prevent the nucleotide from hydrolysis, which is an essential feature in the random substrate binding mechanism. Further interest in crystallographic studies of PLK in the presence of substrates has arisen from several considerations. First, the structure of PLK complexes in the presence of pyridoxal has never been revealed. Thus, the exact interactions between molecules in the active site of PLK are unknown. Secondly, in the PLK ATP complex, ATP -phosphate group is maintained in a position far away from the catalytic site of the enzyme. Although this could prevent the nucleotide from hydrolyzing, the enzyme would need to engineer a dramatic conformational change of this macromolecule for the transfer process. This contradicts the design of other kinases found in nature, indicating that cells may have some regulatory mechanism to control PLK activity. Finally, the catalytic process of PLK consists of several steps. The structure of PLK in the form of a complex with different substrates or its analog and products allows the systematic observation of reaction cycles. In this paper, the crystal structures of the PLK AMP-PCP-pyridoxamine complex, the PLK ADP PLP complex, and the PLK ADP complex were determined. These structures have provided detailed information regarding the interactions between PLK and its substrates or products. Comparison of these structures can lead to a thorough understanding of conformational changes that occur in the enzyme in the presence the ligands in this unique phosphorylation process catalyzed by PLK. MATERIALS AND METHODS Crystallization and Data Collection PLK was purified from sheep brain as described previously (1). Enzyme crystals were obtained by using the hanging drop vapor diffusion method at a constant temperature of 17 C. An initial attempt to diffuse AMP-PCP and pyridoxamine into the native orthorhombic crystals of the enzyme was not successful. However, co-crystallization of PLK in the presence of AMP-PCP and pyridoxamine resulted in the formation of crystals of the enzyme-substrate complex. A solution was prepared consisting of 10 mg/ml PLK, 1 mm

137 17460 Conformation Change of Pyridoxal Kinase TABLE I Structure determination and refinement PLK AMP-PCP-pyridoxamine PLK ADP PLP PLK ADP Data collection statistics Space group P P4 3 P Cell dimension a b 103.7, c 58.6 a b 109.1, c a 59.1, b 93.9, c Resolution (Å) Measured reflections 48, , ,511 Unique reflections 9,133 75,513 18,118 Completeness (%) 99.6 (100) 87.5 (95.2) 99.6 (99.9) a R merge (0.543) (0.465) (0.559) I / (I) 15.7 (3.4) 11.4 (2.3) 13.0 (3.0) Refinement statistics b R work c R free Number of protein atoms 2,413 19,512 4,798 Number of ligand atoms R.m.s.d. bonds (Å) R.m.s.d. angles ( ) Average B-factor (Å 2 ) Protein atoms Ligand atoms Solvent atoms a R merge h i ( I i (h) I(h) )/ h i I i (h), where I i (h) istheith integrated intensity of a given reflection and I(h) is the weighted mean of all of the measurements of I(h). b R work h F(h) o F(h) c / h F(h) o for the 90% of reflection data used in refinement. c R free h F(h) o F(h) o / h F(h) o for the 10% of reflection data excluded from refinement. pyridoxamine, 1 mm AMP-PCP, and 0.1 mm zinc acetate with a buffer of equal volume consisting of 100 mm KH 2 PO 4 -K 2 HPO 4 containing 1.4 M (NH 4 ) 2 SO 4, ph 8.2. The solution was then equilibrated against the buffer containing 1.4 M ammonium sulfate, ph 8.2, for approximately one month at 290 K. Similar conditions were used for the crystallization of a PLK AMP-PCP-pyridoxamine complex. When AMP-PCP was replaced by ADP and pyridoxamine was replaced by PLP, the PLK ADP PLP complex crystals were obtained. Interestingly, the PLK AMP-PCP-pyridoxamine complex crystals were trigonal, whereas the PLK ADP PLP complex crystals were tetragonal. PLK ADP complex crystals were prepared by soaking the orthorhombic crystal of native enzyme (15) in 75 mm KH 2 PO 4 -K 2 HPO 4 buffer containing 1 mm ADP, 1 mm ZnAc 2, and 30% polyethylene glycol 6000, ph 6.5. Data from PLK AMP-PCP-pyridoxamine and PLK ADP complex crystals were collected using the Mar345 image plate in the National Laboratory of Biomacromolecules (Beijing, China) at room temperature. Data from the PLK ADP PLP complex were collected on a Mar345 image plate in the Laboratory of Structural Biology, Tsinghua University (Beijing, China). Before data collection, the crystals were flashfrozen in liquid nitrogen after short soaking in a solution consisting of 17.5% glycerol, 1.4 M ammonium sulfate, and 100 mm potassium phosphate buffer at ph 8.2. All of the data were processed with the HKL suite of programs (16). Table I shows the data collection statistics. Structural Analysis and Refinement The structure of the PLK AMP-PCP-pyridoxamine complex was solved by molecular replacement using AMORE (17). The solution was obtained using the unliganded structure (with residues omitted) as a search model. Because diffraction intensities of this complex crystal were slightly dispersed and could affect the measurement of weak reflections, the refinement process was carried out using reflections only at above 2 cutoff. After rigid body refinement, the R-factor was dropped to and then pyridoxamine, AMP-PCP, zinc ions, and the missing water molecules were added. During model rebuilding, xyz and B-factor refinements were carried out alternately, resulting in an R-free of in the presence of 22 additional water molecules. The PLK ADP PLP complex crystal was initially indexed as P , but was later found to be nearly perfectly twinned with a twinning ratio of 0.48 (18). Therefore, data were reprocessed in P4 3. The complex structure containing eight monomers in an asymmetric unit was solved by molecular replacement with the MOLREP (19) program. Refinement was performed using the CNS (20) scripts for twinning data, and non-crystallographic symmetry restraint was applied. Initial simulated annealing (21) changed the R-factor to 0.264, and an electron density map was calculated according to the structure to which PLP and ADP molecules had been added. After several cycles of xyz refinement, B- factor refinement, and model rebuilding, a model with an R-factor of and an R-free of were obtained. The structure of the PLK ADP complex was determined by the difference Fourier method using the unliganded structure (Protein Data 137 Bank code 1LHP) as a model. Rigid body refinement reduced the R- factor from to The following cycles of xyz refinement, B-factor refinement, and model rebuilding including the addition of missing residues such as ADP molecules, Zn 2 ions, and water molecules resulted in an R-factor of and R-free of All of the model rebuilding were performed by program O (22), and the models were evaluated using PROCHECK (23). Computing programs including MOLSCRIPT (20), RASTER3D (25), and GRASP (26) were utilized to draw figures. RESULTS AND DISCUSSION Structures of Pyridoxal Kinase in Complexes with Substrates and Products In contrast with the orthorhombic crystals of native PLK and the PLK ATP complex, the PLK AMP-PCPpyridoxamine crystal belongs to the P space group with the complex having one monomer in the asymmetric unit. However, the two monomers related by the 2-fold crystallographic axis form a dimeric molecule in a fashion similar to the dimeric molecules found in the native PLK structure. The overall structure of the monomeric PLK AMP-PCP-pyridoxamine complex is almost identical to the structures of the enzyme in the absence of ligands and the PLK ATP complex. Of the 309 residues, the C deviations of 287 residues are 1 Å with an r.m.s.d. of 0.43 Å. In the active site of the enzyme, AMP-PCP and pyridoxamine were located in substrate binding sites, as revealed by the electron density map (Fig. 1a). In the pyridoxal binding site, three hydrogen bonds formed between pyridoxamine and its surrounding amino acid residues. One bond formed between the hydroxyl group in the side chain of Ser-12 and N1 of pyridoxamine. The second bond formed between the hydroxyl group of Thr-47 and O3, and the third bond formed between the carboxyl group of Asp-235 and O5. Tyr-84 and Val-19 also interacted with the two sides of the pyridine ring of pyridoxamine through hydrophobic interactions (Fig. 2). In contrast to the structure of the PLK ATP complex, which is missing the second substrate bound at the active site, our result showed that the side chain of Tyr-84 in the PLK AMP- PCP-pyridoxamine complex was slightly closer to the nucleotide because of the fact that the position of Tyr was stabilized by a hydrogen bond between itself and the guanidino group of Arg-86. As for the amino acid residue of Phe-43, its phenyl ring rotated at an angle of 80. In addition, the Thr-47 hydroxyl group moved toward Phe-43, resulting in a shorter hydrogen

138 Conformation Change of Pyridoxal Kinase FIG. 1. Structures of pyridoxal kinase in complexes with substrates or products. a, b, and c, the F o F c electron density map contoured at 3 using the reflection data for the PLK AMP-PCPpyridoxamine complex, for the PLK ADP PLP complex, and for the PLK ADP complex, respectively, showing clear density for the ligands bound to the enzyme. d, the superposition of the main chain of the three complexes. The overall structure of the PLK AMP-PCP-pyridoxamine complex (shown in yellow) is similar to that of the PLK ADP complex (shown in blue) and is also similar to the PLK ATP complex structure in which structure has already been solved. However, the PLK ADP PLP complex (shown in red) is significantly different from them, the peptide chain moves toward to the active center, and the overall structure becomes more compact. FIG. 2. Pyridoxal binding site. The molecule in the center is the pyridoxamine bound in the PLK AMP-PCP-pyridoxamine complex. The surrounding residues are shown in green, and the hydrogen bonds between the pyridoxamine and the residues are shown as purple dashes. The corresponding residues in the PLK ATP complex are in blue. A comparison of these structures reveals local conformational adjustments of the pyridoxal binding site when the substrates binds. bonding distance between Thr-47 and O3 of pyridoxamine. A conformational change was detected in Trp-52, which is maintained in a position located adjacent to Phe-43 and Thr-47 with its indole ring made at a complete rotation of 180 degrees. Therefore, the relative side chain orientation between Phe-43 and Trp-52 changes from edge-to-face to offset-stacked. At the nucleotide binding site, both the conformation of AMP-PCP and its interactions with surrounding amino acid residues were similar to those of ATP in the PLK ATP complex. Although the conditions for crystallization were similar to those of the PLK AMP-PCP-pyridoxamine complex, the crystals of the PLK ADP PLP complex were tetragonal. Monomers of the two different enzyme-substrate complexes had significant conformation differences. Of a total of 309 amino acids, 117 residues in the PLK ADP PLP complex had carbon atoms that deviated by 1 Å compared with the PLK AMP-PCPpyridoxamine complex. Such a dramatic variation in the conformation of the whole macromolecule could explain the generation of a new protein crystal form in the presence of different substrates. In the asymmetric unit of the PLK ADP PLP structure, eight monomers were found, forming four dimeric molecules in a manner similar to that of the native enzyme. An ADP molecule, a PLP molecule, as well as a Zn 2 ion bound at the active site of each monomer. The phosphate group of PLP and -phosphate of ADP bridged by a Zn 2 ion 138 were in close proximity to each other. The PLP phosphate group formed hydrogen bonds with the main chain nitrogen atoms of Gly-232, Thr-233, Gly-234, and Asp-235, as well as with the side chain of Asp-235 and Thr-127 (Fig. 3). Interactions between the other portions of the PLP molecule and the protein were similar to those in the PLK AMP-PCP-pyridoxamine complex. PLP bound tightly in an area of the protein, which was structurally more compact than the same domain in the PLK AMP-PCP-pyridoxamine complex (shown in Fig. 4b and discussed under Results and Discussion ). Remarkably, neither the pyridoxamine 4 -amino group in the enzyme-substrate complex nor the PLP 4 -aldehyde group in the enzyme-product complex were found to be covered by any of the amino acid chains of the protein (Fig. 4c). This finding is in agreement with previous research that reported that substrate variation within this group did not affect the catalytic activity of PLK (3). In other words, the fact that the 4 -substitution group of the vitamin is exposed to the solvent explains the broad substrate specificity of PLK. In the analysis of the structure of the PLK ADP complex, an ADP molecule and a Zn 2 ion were found in the active site of each monomer. Since this complex could be used as a model to represent the reaction state in which the phosphate transfer process had been completed, the product, PLP, was released from the enzyme. The overall conformation of the protein in

139 17462 Conformation Change of Pyridoxal Kinase FIG. 3. Pre-reaction state model based on the structure of the PLK ADP PLP complex. In the PLK ADP PLP complex, the PLP phosphate group is located in the catalyzing site and makes hydrogen bonds with both the side chain carboxyl group of Asp-235 and the main chain nitrogen atoms of residue These bonds created an anion hole. The distance between the PLP-phosphorus atom and the oxygen atom of the ADP -phosphate is only 2.5 Å. The pre-reaction model can be constructed by moving the PLP-phosphorus atom toward ADP by 0.8 Å (shown in black) and connecting it to the ADP -phosphate without a great adjustment of the other three oxygen atoms. this complex was almost identical to that of the PLK ATP complex (Fig. 1d) with an r.m.s.d. of only 0.25 Å for all of the residues with the exception of the loop Compared with the compact structure of the PLK ADP PLP complex, the PLK ADP complex has a relatively open conformation because the active site is exposed to the solvent region. In addition, the ADP molecule in this complex is significantly different from both ATP in the PLK ATP complex and ADP in the PLK ADP PLP complex, primarily at the position of the phosphate groups. Compared with the ADP molecule in the PLK ADP PLP complex, -phosphate of ADP in the PLK ADP complex could move a distance of 1.1 Å, leading to its inability to form hydrogen bonds with Thr-186 and Asn-150. Instead, a new hydrogen bond was formed with Thr-233. Another movement of 2.2 Å was observed to have taken place on the -phosphate, forming a new hydrogen bond with Thr-186 and decreasing its distance from Asp-118 (Fig. 5). Reaction State before the Phosphate Transfer In the PLK ATP complex, the -phosphate of ATP is far away from the catalytic site. This prevents ATP from being hydrolyzed before pyridoxal is bound (7). Therefore, it was correct to expect that the binding of pyridoxamine would induce some form of conformational changes on both ATP and PLK, causing the -phosphate of ATP to be close to the other substrate. In the other location of the catalytic site of PLK, pyridoxamine should have also been bound in a position that allowed its 5 -hydroxyl group to be able to start a nucleophilic attack on ATP. However, in structural studies, neither AMP-PCP nor the enzyme in the PLK AMP-PCP-pyridoxamine complex has exhibited any significant difference from the structure in the PLK ATP complex and the -phosphorus atom was unusually far away from the 5 -hydroxyl oxygen atom. At a distance of 5.8 Å, it would be impossible for a spontaneous reaction of phosphate transfer to take place. Based on these findings, it could be hypothesized that there is another state of protein conformation known as the pre-reaction state, which might take place just before the phosphate transfer so that the two substrates could be brought close enough together for the reaction to occur. Although this reaction state was not visible in the crystal structures of the enzyme-substrate complexes in this study, its existence could be proven based on the products formed in the reaction. Based on what is known regarding protein chemistry, we 139 believe that the conformation of the PLK ADP PLP complex is more similar to the conformation of the transition state (or the pre-reaction state) than to the conformation of the PLK AMP- PCP-pyridoxamine complex. There are several reasons for this. First, the PLP phosphate group, which corresponds to the -phosphate of ATP before the reaction, is located in the catalytic site of PLK. It forms a hydrogen bond with the side chain carboxyl group of Asp-235 as well as with the main chain nitrogens of residues , which are skeletons of a positively charged anion hole. It is strongly believed that this hole plays a functional role in the stabilization of the transition state of the phosphate group (Fig. 3). Furthermore, -phosphate of ADP is close to the PLP phosphate group that corresponds to the original positions of the - and -phosphate of ATP before the reaction. The distance between ADP O1B atom and PLP-phosphorus atom was 2.5 Å (Fig. 4a). In the enzymatic reaction during phosphate transfer, the bond between O1B of ATP and the -phosphorus atom was broken and a new bond was subsequently formed between the phosphorus atom of the -phosphate group and the oxygen atom of the 5 -hydroxyl group of pyridoxal. These three atoms in the PLK ADP PLP complex were almost collinear (Fig. 3). The arrangement of molecules in this format is consistent with the S N 2 mechanism. Therefore, a pre-reaction state model just before the formation of the transition state was constructed based on the structure of the PLK ADP PLP complex. The model was constructed by moving the phosphorus atom of PLP (in the structure of the PLK ADP PLP complex) by 0.8 Å along the line linking it to the O1B of ADP, and thus, a new bond between these two substrates was formed. At the same time, the bond between the phosphorus atom and oxygen atom within the PLP phosphate group was broken. The operation mentioned above would lead to the conversion of ADP and PLP to ATP and pyridoxal without requiring movement of other atoms in either ADP and PLP. Indeed, this process was the reverse of the phosphate transfer via the transition state in the reaction catalyzed by PLK. Conformational Changes before the Reaction Superposition of different ternary complexes of enzyme-substrate in this investigation revealed unusual changes in both the location and orientation of the two substrates. In contrast to the PLK AMP- PCP-pyridoxamine complex, it is assumed that the molecule of

140 Conformation Change of Pyridoxal Kinase FIG. 4.Structural changes before the reaction revealed by the comparison of the PLK AMP-PCP-pyridoxamine structure and the pre-reaction model. a, the substrates in the two structures. The two substrates in the PLK AMP-PCP-pyridoxamine complex are far away from each other. The -phosphate of AMP-PCP is not in the anion hole, and its phosphorus atom is 5.8 Å away from the O5 atom of pyridoxamine. However, in the pre-reaction model, the -phosphate of ATP forms hydrogen bonds with the catalyzing residues and the distance between its phosphorus atom and the O5 atom of pyridoxamine is only 2.5 Å. b, structural changes of residues around the active site. In contrast with the PLK AMP-PCP-pyridoxamine complex, the residues around the active site of the pre-reaction state model move toward the substrate and push them closer to each other. These residues bind the substrates tightly and restrict the reaction group in a suitable position for the phosphate transfer. c, surface presentation of the active site. The left side shows the PLK AMP-PCP-pyridoxamine complex in which the active site is open. The right side shows pre-reaction state model in which the conformational changes cause the active site to move closer each other and the two substrates are almost buried totally. However, in both of these two structures, the 4 -substituted group of pyridoxamine or pyridoxal are exposed, which allows variations within this group. pyridoxal in the pre-reaction state was translocated toward ATP at a distance of 1.4 Å. Similarly, some conformational adjustment should also occur on the pyridoxal 5 -hydroxyl group. Therefore, both the nitrogenous base and the ribose ring of ATP would have to maintain their positions in the active site, 140 whereas conformational changes of the enzyme would take place to affect the stability of the ATP phosphate groups. As a result, translocations of 1.5 Å were detected for the -phosphate group and translocations of 1.9 Å were detected for the -phosphate group. These movements would then enable the

141 17464 Conformation Change of Pyridoxal Kinase FIG. 5.The ADP molecule bound in the PLK ADP complex and the residues interacting with it. The hydrogen bonds between them are shown as blue dashes. The molecule shown as a thin black line is the ADP in the PLK ADP PLP complex. A significant conformational change happens between the two ADP molecules. two substrates to be placed in suitable positions for subsequent phosphate transfer. Despite the movement of substrates, significant conformational changes were also observed in the protein structure (Fig. 4b). Unlike the structure of the PLK AMP-PCP-pyridoxamine complex, over one-third of the C atoms in the PLK ADP PLP complex moved 1 Å or more. In addition, all of the atoms in helices 4, 5, and 6 were found to move 1.5 Å. This type of movement enabled peptide chains within the protein to move toward the active site, thus creating a compact structure in the enzyme (Fig. 1d). In the pyridoxal binding site, the loop connecting 2 and 2 moved approximately 2 Å toward the substrate pyridoxal. This movement enabled residues Val-41, Phe- 43, and Thr-47 on this loop to interact with pyridoxal directly and push pyridoxal toward the ATP. At the same time, Tyr-84, Tyr-127, His-46, Val-231, and Val-115 moved toward the pyridine ring of pyridoxal from two sides, causing immobilization of the substrate in the active site (Fig. 4b). Although small changes were detected at the ATP adenine ring, Leu-199, Lys- 225, and Phe-230 also moved closer to ATP for the purpose of immobilization. Amino residues interacted with the ATP phosphate group either directly or via cations including Thr-186, Ser-187, Asn-150, Glu-153, Asp-113, Asp-118, Tyr-127, and Thr-148, which are mainly located on one side of ATP. The peptide chain consisting of these amino acid residues tends to move against the ATP molecule, pushing the -phosphate group to the anion hole previously formed by the main chain nitrogens of residues at the N-terminal end of 7. Consequently, the phosphate was able to form new hydrogen bonds with these amino acid residues (Fig. 4b). Tyr-127 and Asp-118 in loop interacted with the dorsal portion of ATP. Dramatic conformational changes of this loop would eventually shift these two residues to new positions where these residues would provide the ATP molecule with increased stabilization. Mechanism of Pyridoxal Kinase Two major catalyzing mechanisms have been suggested for enzymes belonging to the ribokinase superfamily. An anion hole with the ability to stabilize the transition state of reactants is formed by the main chain nitrogen atoms of several continuous residues (8, 10, 13). In PLK, such an anion hole is created by residues at the N-terminal end of the 7 helix. A base-catalyzing group in the active site is needed to initiate the reaction through deprotonation of the hydroxyl group on the vitamin substrate so 141 FIG. 6. The entire process of pyridoxal kinase catalysis. a, the active site of PLK without any substrate bound to it (drawn according to the crystal structure of PLK). b, after the binding of ATP, a loop undergoes remarkable conformational changes and interacts with the ATP (drawn according the crystal structure of the PLK ATP complex). c, when ATP and pyridoxal are both bound in the enzyme, they are far away from each other and no significant changes occur on the overall structure of PLK (drawn according to the crystal structure of the PLK AMP-PCP-pyridoxamine complex with AMP-PCP replaced by ATP and pyridoxamine replaced by pyridoxal). d, the active site becomes more compact and binds the substrates more tightly. The two substrates move close to each other, and their reaction groups are located in the catalyzing position (drawn according to the proposed pre-reaction state model). e, phosphate transfer occurs, resulting in two substrates, ADP and PLP (drawn according to the crystal structure of the PLK ADP PLP complex). f, one of the two substrates, PLP, has been released from the enzyme. The active site is open again, and the conformation of ADP is different from the conformation of the PLK ADP PLP complex. The ADP leaves the enzyme after which new substrates will be bound (drawn according to the crystal structure of the PLK ADP complex). that the hydroxyl oxygen atom can spontaneously attack the -phosphate of ATP. This mechanism is carried out primarily by the side chain carboxyl group of Asp-235 in PLK. Another mechanism is carried out by other kinases in the ribokinase superfamily. For example, a positively charged residue, such as Lys-43 in ribokinase or Arg-136 in adenosine kinase, exerts a stabilizing effect on the -phosphate of ATP before transfer of the phosphate group. However, this positively charged residue was not found in the active site of PLK. Arg-120, the only possible residue located near the active site did not interact with ATP or ADP phosphate, suggesting that a positively charged amino acid is not present for the positioning or stabilization of the -phosphate. In contrast, the correct positioning of the reaction groups of substrates in PLK relies on extensive interactions between amino acid residues of the enzyme and

142 Conformation Change of Pyridoxal Kinase substrates. During the time when conformational changes occurred in the enzyme-substrate ternary complex, converting the complex from its initial state to the pre-reaction state, amino acid residues of PLK would move toward the active site and cause movement toward the substrates through several interactions described above. As a result, substrates bound to the protein became more rigid than before (Fig. 4b). A comparison of the structures of the PLK AMP-PCP-pyridoxamine complex and pre-reaction state model showed that the number of atom pairs with distances 4 Å between pyridoxal and protein increased from 29 to 46 and the number of atom pairs with short distances increased from 70 to 99. A decrease in the distance between atom pairs leads to a further decrease in the volume of an active site, leaving limited free space around the substrates inside the site. As a result, both substrates in this pre-reaction state could be completely buried (Fig. 4c) in the protein mass. By limiting the substrate in a restricted space, this phenomenon directed the reaction group, the -phosphate of ATP, and the 5 -hydroxyl group of pyridoxal to collide with each other, allowing the reaction to occur. The positioning and directing of substrates by cooperative conformational changes in the overall protein structure of enzymes is unique in the catalytic mechanism of PLK. CONCLUSION The Overall Catalyzing Process of PLK Crystallographic studies of PLK with respect to the binary and ternary complex of enzyme in the presence of substrates and products provide information for the elucidation of an integral mechanism in the catalytic process carried out by PLK. Initially, PLK exhibits an open conformation before the binding of any substrate. This exposes the ATP binding site to the solution of the reaction medium but does not expose the pyridoxal binding site (observed in the PLK structure and shown in Fig. 6a). After ATP is bound to the enzyme, the overall structure of PLK changes little at this stage. The loop over the pyridoxal binding site then swings onto the ATP binding site and interacts with the ATP phosphate. At this time, the ATP -phosphate is far away from the catalytic site of PLK to prevent hydrolysis (observed in the structure of PLK ATP complex and shown in Fig. 6b). These procedural incidents are consistent with the random substrate binding kinetics followed by PLK. When pyridoxal as a vitamin B 6 substrate binds to the enzyme, conformational changes of amino acid residues localized at the pyridoxal binding site cause the protein to become more compact than its normal state, enabling the substrate to be tightly bound. Interestingly, this conformational change does not extend to the ATP binding site maintaining the ATP -phosphate far away from the catalytic site (observed in the structure of PLK AMP-PCP-pyridoxamine complex and shown in Fig. 6c). The existence of such a stage in the catalytic mechanism of PLK may be related to the regulation of PLK through other proteins. Previous reports have shown that PLK could form complexes with some PLP-binding enzymes such as aspartate aminotransferase and pyridoxine-5-phosphate oxidase (24, 27) to restrict the release of free PLP into the cellular environment where phosphatases are present. It is possible that the binding of these proteins to PLK may accelerate further conformational changes in the enzyme and lead to spontaneous enzymatic reactions. However, investigation of this particular aspect is needed for further understanding of the mechanism behind protein-protein interactions between PLK and other binding proteins. Later on, further conformational changes cause the structure to become more compact and reach the pre-reaction state and both substrates are pushed by the surrounding amino acid residues to move to positions where their reaction groups are close to each other. This movement causes the ATP -phosphate to be stabilized by an anion hole in the active site, and the pyridoxal 5 -hydroxyl group forms a hydrogen bond with the side chain carboxyl group of Asp-235 (Fig. 6d). Electrons of the oxygen atom of the 5 -hydroxyl group starts a nucleophilic attack on the phosphorus atom -phosphate of ATP, forming a new bond. The phosphorus atom in the -position then moves a distance of 0.8 Å, causing the bond between it and the -phosphate group to break. Thus, the phosphate transfer is completed and the two products, ADP and PLP, are generated (observed in the structure of PLK ADP PLP complex and shown in Fig. 6e). To release reaction products, the overall conformation of the enzyme relaxes again. The active site is exposed to the reaction medium solution, which allows the products to be released as free molecules (observed in the structure of the PLK ADP complex and shown Fig. 6f). For the second catalytic cycle to start, PLK can accept new substrates again for the next cycle of catalysis. REFERENCES 1. Kerry, J. A., Rohde, M., and Kwok, F. (1986) Eur. J. Biochem. 158, McCormick, D. B., and Snell, E. E. (1959) Proc. Natl. Acad. Sci. U. S. A. 45, McCormick, D. B., and Snell, E. E. (1961) J. Biol. Chem. 263, Hanna, M. C., Turner, A. J., and Kirkness, E. F. (1997) J. Biol. Chem. 272, Yang, Y., Zhao, G., and Winkler, M. E. (1996) FEMS Microbiol Lett. 141, Yang, Y., Tsui, H. C., Man, T. K., and Winkler, M. E. (1998) J. Bacteriol. 180, Li, M.-H., Kwok, F., Chang, W.-R., Lau, C.-K., Zhang, J.-P., Lo, S. C. L., Jiang, T., and Liang, D.-C. (2002) J. Biol. Chem. 277, Sigrell, J. A., Cameron, A. D., Jones, T. A., and Mowbray, S. L. (1998) Structure 6, Mathews, I. I., Erion, M. D., and Ealick, S. E. (1998) Biochemistry 37, Schumacher, M. A., Scott, D. M., Mathews, I. I., Ealick, S. E., Roos, D. S., Ullman, B., and Brennan, R. G. (2000) J. Mol. Biol. 298, Campobasso, N., Mathews, I. I., Begley, T. P., and Ealick, S. E. (2000) Biochemistry 39, Ito, S., Fushinobu, S., Yoshioka, I., Koga, S., Matsuzawa, H., and Wakagi, T. (2001) Structure 9, Cheng, G., Bennett, E. M., Begley, T. P., and Ealick, S. E. (2002) Structure 10, Sigrell, J. A., Cameron, A. D., and Mowbray, S. L. (1999) J. Mol. Biol. 290, Li, M.-H., Kwok, F., An, X.-M., Chang, W.-R., Lau, C.-K., Zhang, J.-P., Liu, S.-Q., Leung, Y.-C., Jiang, T., and Liang, D.-C. (2002) Acta Crystallogr. D 58, Otwinowski, Z. (1993) in Proceedings of the CCP4 Study Weekend (Issacs, N., Bailey, S., and Sawyer, L.) pp , Daresbury Laboratories, Warrington, United Kingdom 17. Navaza, J., and Saludjian, P. (1997) Methods Enzymol. 276, Yeates, T. O. (1997) Methods Enzymol. 276, Vagin, A., and Teplyakov, A. (1997) J. Appl. Crystallogr. 30, Brunger, A. T., Adams, P. D., Clore, G. M., DeLano, W. L., Gros, P., Grosse- Kunstleve, R. W., Jiang, J. S., Kuszewski, J., Nilges, M., Pannu, N. S., Read, R. J., Rice, L. M., Simonson, T., and Warren, G. L. (1998) Acta Crystallogr. D 54, Brunger, A. T., Krukowski, A., and Erickson, J. W. (1991) Acta Crystallogr. A 46, Jones, T. A., Zuo, J. Y., Cowan, S. W., and Kjeldgaard, M. (1991) Acta Crystallogr. A 47, Laskowski, R. A., MacArthur, M. W., Moss, D. S., and Thornton, J. M. (1993) J. Appl. Crystallogr. 26, Kwok, F., and Churchich, J. E. (1980) J. Biol. Chem. 255, Kraulis, P. J. (1991) J. Appl. Crystallogr. 24, Merritt, E. A., and Bacon, D. J. (1997) Methods Enzymol. 277, Kim, Y. T., Kwok, F., and Churchich, J. E. (1988) J. Biol. Chem. 263,

143 THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 279, No. 37, Issue of September 10, pp , by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A. Crystal Structure of Human Bisphosphoglycerate Mutase* Received for publication, May 28, 2004, and in revised form, June 28, 2004 Published, JBC Papers in Press, July 16, 2004, DOI /jbc.M Yanli Wang, Zhiyi Wei, Qian Bian, Zhongjun Cheng, Mao Wan, Lin Liu, and Weimin Gong From the National Laboratory of Biomacromolecules, Institute of Biophysics, Chinese Academy of Sciences, Beijing , China and the School of Life Sciences, Key Laboratory of Structural Biology, University of Science and Technology of China, Hefei, Anhui , China Bisphosphoglycerate mutase is a trifunctional enzyme of which the main function is to synthesize 2,3-bisphosphoglycerate, the allosteric effector of hemoglobin. The gene coding for bisphosphoglycerate mutase from the human cdna library was cloned and expressed in Escherichia coli. The protein crystals were obtained and diffract to 2.5 Å and produced the first crystal structure of bisphosphoglycerate mutase. The model was refined to a crystallographic R-factor of and R free of with excellent stereochemistry. The enzyme remains a dimer in the crystal. The overall structure of the enzyme resembles that of the cofactor-dependent phosphoglycerate mutase except the regions of 13 21, , , and the C-terminal tail. The conformational changes in the backbone and the side chains of some residues reveal the structural basis for the different activities between phosphoglycerate mutase and bisphosphoglycerate mutase. The bisphosphoglycerate mutase-specific residue Gly-14 may cause the most important conformational changes, which makes the side chain of Glu-13 orient toward the active site. The positions of Glu-13 and Phe-22 prevent 2,3-bisphosphoglycerate from binding in the way proposed previously. In addition, the side chain of Glu-13 would affect the Glu-89 protonation ability responsible for the low mutase activity. Other structural variations, which could be connected with functional differences, are also discussed. Bisphosphoglycerate mutase (BPGM) 1 is an erythrocyte-specific trifunctional enzyme. The main activity is that of synthase (BPGM, EC ), catalyzing the formation of 2,3-bisphosphoglycerate (2,3-BPG) from 1,3-bisphosphoglycerate (1,3- * This work is supported by Project from the Foundation for Authors of National Excellent Doctoral Dissertation of China, Grant from the National Foundation of Talent Youth, Grant 2001AA from the National High Technology Research and Development Program of China, Grant 2002BA711A13 from the 863 Special Program of China, the Key Important Project and other projects from the National Natural Science Foundation of China (Grant No , No , No and No ), and Grant KSCX1-SW-17 from the Chinese Academy of Sciences. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. The atomic coordinates and structure factors (code 1T8P) have been deposited in the Protein Data Bank, Research Collaboratory for Structural Bioinformatics, Rutgers University, New Brunswick, NJ ( To whom correspondence should be addressed: Institute of Biophysics, 15 Datun Rd., Chaoyang District, Beijing , China. Tel.: ; Fax: ; wgong@sun5.ibp.ac.cn. 1 The abbreviations used are: BPGM, bisphosphoglycerate mutase; hbpgm, human BPGM; dpgm, cofactor-dependent phosphoglycerate mutase; BPG, bisphosphoglycerate BPG) (1 3) (see Fig. 1A). The second activity is that of a mutase (phosphoglycerate mutase, EC ) catalyzing the interconversion between 2- and 3-phosphoglycerate (Fig. 1B) (4). The third activity, as a phosphatase (S-succinylglutathione hydrolase, EC ), is to catalyze the hydrolysis of 2,3-BPG to 3- or 2-phosphoglycerate and a phosphate (Fig. 1C) (1, 2). The phosphatase reaction can be stimulated by a number of anions including chloride, phosphate, and particularly by 2-phosphoglycolate (5). These three enzymic activities have been found to occur at a unique active site with two different binding sites for the substrates, one for bisphosphoglycerate and another for monophosphoglycerate (6, 7). BPGM regulates the level of 2,3-BPG in human blood cells. In vivo, the concentration of 2,3-BPG is determined by the relative activities of the synthase and phosphatase reactions. In red blood cells, 2,3-BPG is the main allosteric effector of hemoglobin. It shifts the equilibrium between the oxy and deoxy conformations of hemoglobins by preferentially stabilizing the unliganded form. Sickle cell anemia is characterized by polymerization of deoxygenated hemoglobin mutants giving rise to deformed erythrocytes and vaso-occlusive complications. 2,3-BPG has been shown to facilitate this polymerization. The ability to modulate the 2,3-BPG level in vivo would have important implications in the treatment of ischemia and sick cell anemia. One therapeutic approach would be to decrease the intraerythrocytic level of 2,3-BPG by increasing the phosphatase activity of the BPGM. Based on its enzymatic properties and amino acid sequence homology, BPGM is closely related to the glycolytic housekeeping enzyme, cofactor (2,3-BPG)-dependent phosphoglycerate mutase (dpgm) (8, 9), involved in glycolytic and gluconeogenic pathways with sequence identities in the 40 50% range (6). In the SCOP data base, BPGM and dpgm are grouped into a superfamily along with fructose-2,6-bisphosphatases and acid phosphatase. Sequence and structural comparisons show that they share a set of conserved residues forming the catalytic core but otherwise exhibit many differences associated with substrate specificities and catalytic activities. For example, the synthase activity of BPGM is higher than dpgm, but the latter has stronger mutase activity. A modest resolution structure of dpgm from Saccharomyces cerevisiae was reported in 1974 (10). Recently, the high-resolution structures of dpgms and the complex with their inhibitors were reported (11 17). According to these structural results, the substrate binding sites and the mutase catalytic mechanisms have been discussed. However, no BPGM three-dimensional structure information is available, although BPGM has been crystallized (18), Here we report the crystal structure of human BPGM (hbpgm) for the first time. Structural comparison of hbpgm and dpgms show some significant differences between the two groups of enzymes. This paper is available on line at

144 Crystal Structure of Human Bisphosphoglycerate Mutase EXPERIMENTAL PROCEDURES Gene Cloning, Expression, and Protein Purification The hbpgm gene was amplified from human brain cdna library (Clontech) by a PCR using the 5 - and 3 -end special primers 5 CATATGTCCAAGTA- CAAACTTATTATG 3 and 5 CTCGAGTTTTTTAGCTTGTTTCAC 3 (Sangon) designed based on the mrna sequence of hbpgm (Gen- Bank TM accession number X04327). Those oligonucleotides were introduced with NdeI and XhoI restriction sites, respectively. Taq polymerase, T-vector, DNA ligase, and the relevant restriction enzymes were obtained from Takara. The polymerase chain reaction product ( 800 bases) was purified with the Gel Extraction mini kit (Watson BioTechnologies) and cloned into a T-vector. The positive clones were identified by restriction digestion. The DNA fragment was ligated into the NdeI/ XhoI-cleaved plasmid pet22b (Novagen) to give the pet22b-hbpgm construction that was amplified in E. coli BL21 (DE3). The recombinant protein contains eight non-native residues at the C terminus. The integrity of the gene was confirmed by DNA sequencing. Single colonies were cultured in Luria-Bertani broth medium with ampicillin (100 g ml 1 ) for expression. Cells were induced with 0.5 mm isopropyl- -Dthiogalactopyranoside after attaining an A 600 of 0.4 and grown for an additional 4 h at 310 K.The cells were harvested by centrifugation, resuspended in lysis buffer containing 0.5 M NaCl, 20 mm Tris-HCl (ph 7.5), and lysed by sonication on ice. After centrifugation, the supernatant was loaded onto a nickel-nitrilotriacetic acid column (Qiagen) and eluted using a step gradient of M imidazole. The fractions were tested for purity by SDS-PAGE. The purified human BPGM protein was desalted and concentrated to 30 mg ml 1, determined by a Bio-Rad Protein Assay. Sample purity and molecular weight ( 30 kda) were verified by SDS-PAGE and mass spectrometry, respectively. Crystallization and Data Collection Crystals were grown by hanging drop vapor diffuse method at 291 K from a protein solution containing hbpgm (30 mg ml 1 ), 20 mm Tris-HCl (ph 7.5) with 50 mm NaCl and an equal volume reservoir solution comprising 100 mm HEPES- NaOH (ph 7.0) and 20% (w/v) polyethylene glycol Crystals appeared after 3 days and kept growing for a week. The diffraction data were collected on a Mar-CCD detector at the Beijing Synchrotron Radiation Facility ( 0.9 Å) at the Institute of High Energy Physics, Chinese Academy of Science. A total of 200 frames of data to 2.5 Å were collected with oscillation range of 1 at 100 K. Data were processed with Denzo and Scalepack (19). Structure Determination and Refinement Molecular replacement was performed with program AmoRe (20) using data from 15 to 3.5-Å resolution with S. cerevisiae dpgm (PDB entry: 5PGM) as the searching model. Multiple cycles of refinement were performed with package of CNS (21). The non-crystallography symmetry was applied during all steps of refinement. Water molecules were included near the end of refinement, followed by manual modification in the graphics program O (22). The quality and stereochemistry of the final model was analyzed with PROCHECK (23). TABLE I Summary of data collection and structure refinement Numbers in parentheses represent the value for the highest resolution shell. Unit cell (Å ) a 38.6 b 61.8 c Resolution limit (Å) ( ) Completeness (%) 96.0 (97.2) Number of reflections 176,376 Number of independent reflections 18,907 R merge (0.183) I/ 9.94 (4.14) R-factor a (0.302) b R free (0.337) Number of non-h atoms 4,221 Protein atoms 3,989 Water molecules 232 Average B-factors (Å 2 ) Main chain Side chain Water Root mean square distance of bond lengths (Å) Root mean square distance of 1.34 bond angles ( ) a R-factor F obs F calc / F obs, where F obs and F calc are observed and calculated structure factors. b R free SUB F obs F calc / T F obs, where T is a test data set of 10% of the total reflections randomly chosen and set aside prior to refinement. RESULTS AND DISCUSSION Structure Determination Crystals grew with a rod-shaped morphology and belonged to space group P2 1 based on the systematic absences in the diffraction data. The unit cell dimensions are a 38.6 Å, b 61.8 Å, c Å, 95.0 with a solvent content of 48% corresponding to two monomers for each asymmetric unit. The final model refined at a 2.5-Å resolution contains residues from Ser-2 to Asp-250 of the total 267 residues (including the His tag) in each monomer with no definite electron density beyond residue 250. The data in Table I show a well refined structure with excellent stereochemistry. In the Ramachandran plot, 91.4% of the residues are located in the most favored areas, and 8.1% are located in the additional allowed regions. Only one residue, Ser-24, is located at generously allowed regions (discussed below). The model fits well with the corresponding electron density except for the C-terminal residues. Overall Structure The hbpgm monomer (Fig. 1)contains two domains including six -strands (named A F) and ten -helices (named 1 10). The / -fold of hbpgm resembles that of the dpgms from S. cerevisiae and E. coli (Fig. 2). The protein core consists of a 144 FIG. 1. Reactions catalyzed by BPGM. A, synthase activity; B, mutase activity; C, phosphatase activity. six -strand in the order A, B, C, D, and F in parallel and strand E in antiparallel conformations. The -sheet is flanked by six -helices. The C-terminal tail ends around the active site. Glu-249 and Asp-250 lie on the center of the active cleft mouth with their side chains oriented from the active cleft. Lys-246 is hydrogen bonded to OE-1 of Glu-249 via its NZ atom. The OE-1 atom of Glu-33 interacts with the NZ atom of Lys-247 by a hydrogen bond. It is noteworthy that there is always an acidic residue at

145 39134 Crystal Structure of Human Bisphosphoglycerate Mutase FIG. 2.Stereo view of the structural superposition of the hbpgm (green), E. coli dpgm (purple), and S. cerevisiae dpgm (blue). The regions (residue 13 21, , , and the C-terminal tail of hbpgm) whose conformations are different from PGM are colored in yellow. This figure was produced using the Lsqman (24) and Ribbons (25) programs. residue 33 in BPGMs. These interactions may make the C- terminal helix (helix 10) more stable and fix it to the active site pocket. The values of the B-factors indicate that the C-terminal region is more flexible compared with the other parts of the molecule. Similarly, in the C-terminal tail, 12 residues cannot be observed in S. cerevisiae dpgm and 9 residues are missing in E. coli dpgm in its unphosphorylated form. The C-terminal residues play an important role in the activities of the enzymes in this family. Removal of the 12 C-terminal residues from S. cerevisiae dpgm is associated with loss of mutase activity but no change in phosphatase activity (26). The deletion of the last seven residues completely abolished the three catalytic activities of the hbpgm (9). The C-terminal residues would be involved in transferring a phosphoryl group to the internal active site (27). Therefore, the disordering of the C-terminal tail appears to be a common feature in both of BPGMs and of dpgms and could accommodate the access of the substrate to the active site. FIG. 3.The dimer of hbpgm viewed along the non-crystallographic 2-fold axis. Dimer Association hbpgm forms a dimer in the crystal in agreement with those observed in solution (5). The dimer is formed between the surface of the C strands and 3 helices of the two monomers with a non-crystallographic 2-fold symmetry (Fig. 3) similar to E. coli dpgm and S. cerevisiae dpgm. The side chains of Ile-64, Trp-68, Leu-69, Leu-71, and Val-81 form a hydrophobic dimerization core. Specifically, the side chains of Trp-68 from both monomers stack with each other across the 2-fold axis. Those hydrophobic residues are conserved in all BPGMs and the dimerized dpgms but not in Schizosaccharomyces pombe dpgm, which is a monomer. In addition, the salt bridge between Lys-29 and Glu-72 and several hydrogen bonds formed by Glu-51, Phe-52, Asp-53, His-65, Glu-77, and Arg-140 strengthen the dimer interactions. Arg-140 forms two hydrogen bonds to the carbonyl oxygen of Phe-52 and to the OD-2 atom of Asp-53 via its NH 2 atom and forms a hydrogen bond to the OE-1 atom of Glu-51 via its NE atom. His-65 forms a hydrogen bond to the side chain of Glu-77 via its NE-2 atom. All of these residues discussed above are conserved in BPGMs. The two subunits of the dimer show few differences when overlaid. The 145 root mean square distance between their C atoms is 0.51 Å. The following discussion is based on the structure of monomer A. Comparison with dpgms Sequences Alignment and Backbone Comparison Similarities of the hbpgm structure to some dpgm structures (PDB codes 5PGM, 1E58, 1E59, 1QHF, and 1FTZ) were analyzed using the program DALI (28) and LSQMAN (24). A structurebased sequence alignment of BPGMs and dpgms is shown in Fig. 4. The sequence of hbpgm is 50% identical to dpgms. The catalytic site residues Arg-10, His-11, Arg-62, Glu-89, Arg- 90, Arg-116, Arg-117, and His-188 are conserved in all of them. However, some residues involved in the substrate binding in dpgms have been substituted in BPGMs (discussed below). The topology of the / domain of BPGM shows considerable similarity to dpgms. A superposition of hbpgm with the structure of S. cerevisiae dpgm and E. coli dpgm yields root mean square distances of 0.69 and 0.74 Å, respectively, for C atoms in 124 residues in three regions (hbpgm residues 4 11, 25 97, and ), indicating strong structural homology within

146 Crystal Structure of Human Bisphosphoglycerate Mutase FIG. 4. Structure-based sequence alignment of BPGMs from human (hbpgm), rabbit (rbbpgm), rat (rtbpgm), and mouse (mbpgm) and dpgms from human brain (hpgm-b), human muscle (hpgm-m), E. coli (ecpgm), S. cerevisiae (scpgm), and S. Pombe (sppgm). The secondary structure elements in the crystal structure of hbpgm are shown above the alignment in blue. Arrows indicate -strands and coils indicate helices. Strictly conserved and conservatively substituted residues are boxed and marked with red and yellow background, respectively. Residues conserved only for BPGMs are colored in purple. The residues in the C-terminal region that are different between BPGMs and dpgms are colored in red. Cys-23 is present in all BPGMs and human muscle dpgm, resulting in their sensitivity to Hg 2, and marked in green background. Glu-13 and Gly-14 are marked by black stars. Every 10 residues of the hbpgm sequence are marked with dots and the residue numbers in each sequence are showed in the front. The figure was generated using ESPript (29). those regions, consistent with the concept that BPGM and dpgm are structurally homologous enzymes. The most significant variations in the backbone occur in the regions 13 21, , , and the C-terminal tail (Fig. 2). With a four-residue insertion at residues 136, 137, 143, and 144, an additional -helix is formed from residue 133 to 138, resulting in the different conformations at the region of residues This region is far away from the active site, and the structural differences would not contribute to the catalytic activity. The fragments 13 21, , and the C-terminal fragment surround the active site. These conformational changes may have significant effect on the enzymatic activity. Residues Specific for BPGMs There are 24 residues (Fig. 4, pink) that are identical in BPGMs but differ from those in dpgms. Among them, Glu-77 is involved in the interactions 146 between the two subunits by forming a hydrogen bond with His-65. Ser-63 and Ser-186, in which the equivalent residues are both Ala in dpgms located near the active site, form two hydrogen bonds between their main-chain and side-chain oxygen atoms. The hydroxyl oxygen of Ser-192 forms three hydrogen bonds with the carbonyl oxygen atoms of His-188 and Gly-189, and the main-chain nitrogen atom of Arg-193. These additional hydrogen bonds may help stabilize the structure. Gly-14 is a critical BPGM-specific residue. A mutation of Gly-14 of hbpgm to Ser did not modify the synthase activity, whereas the mutase and phosphatase activities were 2-fold increased or decreased, respectively. However, replacing Gly-14 with Arg enhanced phosphatase activity by 28.6-fold, whereas synthase and mutase activities were decreased 10-fold (30). In dpgms, the equivalent residue Ser interacts with the

147 39136 Crystal Structure of Human Bisphosphoglycerate Mutase FIG. 5.Stereo drawing of the superposition of the active site pocket of hbpgm (green) and S. cerevisiae dpgm (gray). The critical residues in the active site pocket are labeled. substrate (12). In prostatic acid phosphatase, Arg-15, the equivalent residue of Gly-14 of BPGM, points toward the active site and is involved in the substrate binding too (31 33). Fig. 5 shows that the conformation around Gly-14 varies significantly. As a consequence, the side chain of Glu-13 in hbpgm points toward the active site, forming a hydrogen bond network with Arg-10, His-11, Glu-89, and Gly-189 via three water molecules, whereas the equivalent residue of dpgms orients away from the active site. Glu-89 is conserved in all BPGMs and dpgms. In mutase activity, Glu-89 acts as an acid or a base during enzyme phosphorylation and dephosphorylation (12). The phosphotransfer step of the mutase reaction requires a proton to be transferred to or from Glu-89. The hydrogen bond between Glu-13 and the side chain of Glu-89 would affect the protonation ability of Glu-89, so that the mutase activity is lower in BPGMs. In contrast, the first phosphotransfer of the synthase reaction involves the transfer of an acyl-phospho group and does not require a proton to be transferred (6). Ser-24 is reported to be involved in the binding of both monophosphoglycerates and 2-phosphoglycolate (7). This position is conserved as a Gly in all dpgms. It was proposed that its backbone nitrogen atom interacts with the substrate carboxylate group (15). Because its dihedral angle is located in a general allowed region in the Ramachandran plot, it was predicted that Ser-24 would be the major reason for the different enzymatic activities of BPGMs and dpgms by adopting a different dihedral values and changing the main-chain conformation (12, 15). In our hbpgm structure, no significant structural change was observed except the side chain of Ser-24 was added. Although the Ser-24 side chain fits the electron density well, its conformation is in an unfavorable region of the Ramachandran plot (Fig. 6). The unfavorable stereochemistry may be compensated by the formation of a hydrogen bond between its hydroxyl oxygen and the backbone nitrogen atom of Tyr-92. Its carbonyl oxygen maintains contact with the NH 2 atom of Arg-62 as in dpgms. So it is unlikely that Ser-24 plays an important role in distinguishing the activities of BPGMs and dpgms. Active Site and the Substrate Binding Pocket The active site pocket was found at the carboxyl end of the parallel -strands of the / -domain. His-11 and His-188 are located at the bottom of the active site pocket as well as Arg-10 and 147 Arg-62. The mouth of the cleft is defined by Lys-18, Asn-20, Arg-100, and Arg-116 on one side and Ile-208, Asn-209, and Thr-211 on the other side (Fig. 5). Residues lie on the surface of the active site pocket. The general shape of the active cleft looks similar to those of dpgms, but the details differ resulting from the conformational changes at regions of 13 21, , and the C termini. The catalytic site cleft contains many basic residues resulting in a highly positive electrostatic potential accommodating for the negative charged substrates as dpgms. His-11 and His-188 are conserved in BPGMs and dpgms. In the hbpgm structure, their side chains adopt very similar conformations to those in the structure of dpgm (Fig. 5). His-11 was proposed to accept a phosphate group during the reactions (6). His-188 was confirmed to be very important for catalysis, but its precise function is still speculative (34, 35). In the hbpgm structure, the NE-2 atom of His-11 forms a hydrogen bond to the ND-1 atom of His-188. The hydroxyl group of Ser-58 is hydrogen-bonded to atom NE-2 of His-188. This interaction is also conserved in dpgms in the rat prostatic acid phosphatase (between Thr-75 and His-257) and in fructose-2,6- bisphosphatase (between Ser-52 and His-141), suggesting they contribute to the correct orientation of the His-188 imidazole ring. Most of the residues involved in bisphosphoglycerate binding (Arg-10, Asn-17, Arg-62, Glu-89, Arg-90, Tyr-92, Arg-116, Arg- 117, and Asn-190) are conserved in BPGMs as in dpgms. Upon a superposition of the structures of hbpgm and dpgms, Arg- 10, Arg-62, Glu-89, Arg-90, and Asn-190 fit well; Asn-17, Arg- 116, and Arg-117 shift with the backbone movement. The side chains of Arg-116 and Arg-117 are flexible and cannot be observed in the electron density map. Thr-20 and Lys-97 (numbered as in dpgms), in which the side chains were proposed to be involved in 2,3-BPG binding in dpgms, are substituted by Cys-23, and Arg-100, respectively, in hbpgm. Cys-23 is conserved in BPGMs and in muscle dpgm leading to a susceptibility to inactivation by Hg 2 (6). In hbpgm, the guanidinium group of Arg-62, which is necessary for substrate binding, contacts the SH group of Cys-23 in van der Waals distance. If a positively charged Hg 2 bound to Cys-23, it should push away the same positively charged

148 Crystal Structure of Human Bisphosphoglycerate Mutase FIG. 6.The electron density map (2F o F c, 1.0 ) of Ser-24. The hydrogen bond between Ser-24 and Arg-90 is labeled. The and angles of this residue are 53.5 and 125.1, respectively. These two angles in the equivalent residue Gly in E. coli dpgm are 61.4 and 130.8, and in S. cerevisiae dpgm are 67.6 and 124.9, respectively. Arg-62 and prevent the substrate from binding. It also should be noted that in dpgms, the side chain of Lys-97 points toward the active site pocket. However, in hbpgm, the equivalent residue Arg-100 turns its side chain away from the active site 90, toward the surface of the active cleft. It was reported that the substrate binding is not the rate-limited step, and the different catalytic ability of BPGMs and dpgms is caused mainly by the phosphorylation rates of 1,3-BPG and 2,3-BPG (36). Whether the moving away of Arg-100 is necessary for the specificity of phosphorylation needs to be investigated further. Compared with dpgms, the space around the catalytic His-11 is more crowded in hbpgm. Several backbone and residue differences contribute to this. First, the backbone of residues shifts with a distance of 5.0 Å between Glu-19 and its equivalent residue in S. cerevisiae dpgm. Second, the side chain of Glu-13 points toward the active site, and the side chain of Phe-22 rotates closer to the catalytic His-11 leading to the small active site space. As the consequence, the 2,3-BPG binding model of E. coli dpgm (15) cannot be adopted in hb- PGM without the backbone or side-chain movements, because the side chain of Glu-13 leads to the steric clash with the phospho group near the catalytic site, and the side chains of Tyr-92 and Phe-22 lead to steric clash with the distal phospho group of 2,3-BPG. Moreover, the binding model of 3-phosphoglycerate to S. cerevisiae dpgm observed experimentally (37) cannot be adopted by hbpgm, in which Glu-13 occupies the position of the ligand, either. If the similar binding mode remains in hbpgm, large conformational changes should happen upon the substrate binding. It also should be noted that the binding of dpgms with 2,3-BPG is only a theoretical model derived from the complex of E. coli dpgm and a liner tetravanadate (15). We are trying to obtain the hbpgm-substrate complex crystals, which will provide deeper insight into the binding and catalytic mechanism. 148 The narrower active cleft and varieties of residues involved in substrate binding could be another reason for different phosphorylation rates of BPGMs and dpgms by 2,3- and 1,3-BPG, respectively, which leads to the different mutase and synthase activities. The mutase reaction requires that the 2,3-BPG stay as an intermediate and change its orientation in the active site during each round of catalysis (6). However, in hbpgm, the narrower active site pocket would limit the reorientation of 2,3-BPG. Although in the synthase reaction, 2,3-BPG should be released as a product. In conclusion, hbpgm is similar in the folding, dimerization pattern, and the active site architectures to dpgms as predicted from the sequence homology. The major structural variations occur at residues 13 21, , , and the C- terminal tail. Gly-14 may have caused the most important conformational changes where the inside flipping of Glu-13, together with the relocation of Phe-22, resulted in a more crowded substrate binding pocket and could prevent 2,3-BPG from binding in the way that was proposed in dpgms. In addition, the side chain of Glu-13 would affect Glu-89 protonation ability that is critical for the mutase activity. Acknowledgments We thank Prof. Peng Liu and Yuhui Dong for synchrotron data collection. REFERENCES 1. Rose, Z. B. (1968) J. Biol. Chem. 243, Rose, Z. B. (1970) J. Biol. Chem. 248, Rose, Z. B. (1980) Adv. Enzymol. Relat. Areas Mol. Biol. 51, Rosa, R., Gaillardon, J., and Rosa, J. (1973) Biochem. Biophys. Res. Commun. 51, Rose, Z. B., and Liebowitz, J. (1970) J. Biol. Chem. 245, Fothergill-Gilmore, L. A., and Watson, H. C. (1989) Adv. Enzymol. Relat. Areas Mol. Biol. 62, Ravel, P., Craescu, C. T., Arous, N., Rosa, J., and Carel, M. C. (1997) J. Biol. Chem. 272, Hass, L. F., Kappel, W. K., Miller, K. B., and Engle, R. L. (1978) J. Biol. Chem. 253, 77 81

149 39138 Crystal Structure of Human Bisphosphoglycerate Mutase 9. Garel, M.-C., Joulin, V., Le Boulch, P., Calvin, M.-C., Prehu, M.-O., Arous, N., Longin, R., Rosa, R., Rosa, J., and Cohen-Solal, M. (1989) J. Biol. Chem. 264, Campbell, J. W., Watson, H. C., and Hodgson, G. I. (1974) Nature 250, Rigden, D. J., Alexeev, D., Phillips, S. E. V., and Fothergill-Gilmore, L. A. (1998) J. Mol. Biol. 276, Rigden, D. J., Walter, R. A., Phillips, S. E. V., and Fothergill-Gilmore, L. A. (1999) J. Mol. Biol. 286, Bond, C. S., White, M. F., and Hunter, W. N. (2001), J. Biol. Chem. 276, Rigden, D. J., Bagyan, I., Lamani, E., Setlow, P., and Jedrzejas, M. J. (2001) Protein Sci. 10, Bond, C., White M. F., and Hunter, W. N. (2002) J. Mol. Biol. 316, Rigden, D. J., Mello, L. V., Setlow, P., and Jedrzejas, M. J. (2002) J. Mol. Biol. 315, Uhrinova, S., Uhrin, D., Nairn, J., Price, N. C., Fothergill-Gilmore, L. A., and Barlow, P. N. (2001) J. Mol. Biol. 306, Cherfils, J., Rosa, R., Garel, M. C., Calvin, M. C., Rosa, J., and Janin, J. (1991) J. Mol. Biol. 218, Otwinowski, Z., and Minor, W. (1997) Methods Enzymol. 276, Collaborative Computational Project 4 (1994) Acta Crystallogr. Sect. D Biol. Crystallogr. 50, Brunger, A. T., Adams, P. D., Clore, G. M., DeLano, W. L., Gros, P., Grosse- Kunstleve, R. W., Jiang, J.-S., Kuszewski, J., Nilges, M., Pannu, N. S., Read, R. J., Rice, L. M., Simonson, T., and Warren, G. L. (1998) Acta Crystallogr. Sect. D Biol. Crystallogr. 54, Jones, T. A., Zou, J. Y., Cowan, S. W., and Kjeldgaard, M. (1991) Acta Crystallogr. Sect. A 47, Laskowsky, R. A., MacArthur, M. W., Moss, D. S., and Thornton, J. M. (1993) J. Appl. Crystallogr. 26, Kleywegt, G. J. (1999) Acta Crystallogr. Sect. D Biol. Crystallogr. 55, Carson, M. (1987) J. Mol. Graphics 5, Walter, R. A., Nairn, J., Duncan, D., Price, N. C., Kelly, S. M., Rigden, D. J., and Fothergill-Gilmore, L. A. (1999) Biochem. J. 337, Winn, S. I., Watson, H. C., Harkins, R. N., and Fothergill, L. A. (1981) Philos. Trans. R. Soc. Lond. B. Biol. Sci. 293, Homl, L., and Sander, C. (1993) J. Mol. Biol. 233, Gouet, P., Courcelle, E., Stuart, D. I., and Metoz, F. (1999) Bioinformatics 15, Garel, M. C., Arous, N., Calvin, M. C., Craescu, C. T., Rosa, J., and Rosa, R. (1994) Proc. Natl. Acad. Sci. U. S. A. 91, Schneider, G., Lindqvist, Y., and Vihko, P. (1993) EMBO J. 12, LaCount, M. W., Handy, G., and Lebioda, L. (1998) J. Biol. Chem. 273, Ortlund, E., LaCount, M. W., and Lebioda, L. (2003) Biochem. 42, White, M. F., and Fothergill-Gilmore, L. A. (1992) Eur. J. Biochem. 207, White M. F., Fothergill-Gilmore, L. A., Kelly, S. M., and Price, N. C. (1993) Biochem. J. 291, Rose, Z. B., and Dube, B. (1976) J. Biol. Chem. 251, Glina, S. C., Andrew, R. D., Michail, N. I., John, W. C., and Jenifer, A. L. (1999) Acta Crystallogr. Sect. D Biol. Crystallogr. 55,

150 THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 279, No. 29, Issue of July 16, pp , by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A. Structural Basis for Coronavirus-mediated Membrane Fusion CRYSTAL STRUCTURE OF MOUSE HEPATITIS VIRUS SPIKE PROTEIN FUSION CORE* Received for publication, April 5, 2004, and in revised form, April 22, 2004 Published, JBC Papers in Press, April 27, 2004, DOI /jbc.M Yanhui Xu, Yiwei Liu, Zhiyong Lou, Lan Qin, Xu Li, Zhihong Bai, Hai Pang, Po Tien, George F. Gao **, and Zihe Rao From the Laboratory of Structural Biology, Tsinghua University, Beijing , China, the National Laboratory of Biomacromolecules, Institute of Biophysics, Beijing , China, the Institute of Microbiology, Chinese Academy of Sciences, Beijing , China, and the **Nuffield Department of Clinical Medicine, John Radcliffe Hospital, Oxford University, Oxford OX3 9DU, United Kingdom The surface transmembrane glycoprotein is responsible for mediating virion attachment to cell and subsequent virus-cell membrane fusion. However, the molecular mechanisms for the viral entry of coronaviruses remain poorly understood. The crystal structure of the fusion core of mouse hepatitis virus S protein, which represents the first fusion core structure of any coronavirus, reveals a central hydrophobic coiled coil trimer surrounded by three helices in an oblique, antiparallel manner. This structure shares significant similarity with both the low ph-induced conformation of influenza hemagglutinin and fusion core of HIV gp41, indicating that the structure represents a fusion-active state formed after several conformational changes. Our results also indicate that the mechanisms for the viral fusion of coronaviruses are similar to those of influenza virus and HIV. The coiled coil structure has unique features, which are different from other viral fusion cores. Highly conserved heptad repeat 1 (HR1) and HR2 regions in coronavirus spike proteins indicate a similar three-dimensional structure among these fusion cores and common mechanisms for the viral fusion. We have proposed the binding regions of HR1 and HR2 of other coronaviruses and a structure model of their fusion core based on our mouse hepatitis virus fusion core structure and sequence alignment. Drug discovery strategies aimed at inhibiting viral entry by blocking hairpin formation may be applied to the inhibition of a number of emerging infectious diseases, including severe acute respiratory syndrome. Coronaviruses are enveloped viruses with single-stranded, positive-sense genomic RNA that is kb in length (1, 2). * This work was supported in part by Projects 973 and 863 of the Ministry of Science and Technology of China Grants 200BA711A12, G , and 2003CB514116, the National Natural Science Foundation of China Grant , and the Key Project of the Knowledge Innovation Program of Chinese Academy of Sciences. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. The atomic coordinates and structure factors (codes 1WDF and 1WDG) have been deposited in the Protein Data Bank, Research Collaboratory for Structural Bioinformatics, Rutgers University, New Brunswick, NJ ( These authors contributed equally to this work. Supported by the Chunhui Project Scheme of Ministry of Education of China. To whom correspondence may be addressed. Tel.: ; Fax: ; george.gao@ndm.ox.ac.uk. To whom correspondence may be addressed: Laboratory of Structural Biology, Tsinghua University, School of Life Sciences and Engineering, Beijing , China. Tel.: ; Fax: ; raozh@xtal.tsinghua.edu.cn The Coronaviridae exhibit a broad host range, infecting many mammalian and avian species and causing upper respiratory, hepatic, gastrointestinal, and central nervous system diseases. Coronaviruses in humans and fowl primarily cause upper respiratory tract infections, whereas porcine and bovine coronaviruses establish enteric infections that result in severe economic loss (3). The coronaviruses also include mouse hepatitis virus (MHV), 1 infectious bronchitis virus, feline infectious peritonitis virus, and the newly emergent human severe acute respiratory syndrome-associated coronavirus (HcoV-SARS) (4). The surface glycoproteins of enveloped viruses play essential roles in viral entry into cells by mediating virion attachment to cells and the virus-cell membrane fusion, the initial events of the viral infections. The spike (S) protein is the sole viral enveloped glycoprotein responsible for cell entry in coronaviruses. It binds to the cell surface receptor and mediates subsequent fusion of the viral and cellular membranes (5). Under the electron microscope, the spike proteins can be clearly seen as 20-nm-long surface projections on the virion membrane (6). The spike proteins of coronaviruses share several features with other viral glycoproteins mediating viral entry, including the hemagglutinin (HA) protein of influenza virus, gp160 of human immunodeficiency virus (HIV) and simian immunodeficiency virus (SIV), GP of Ebola virus, and fusion protein of paramyxovirus (7, 8). These glycoproteins are all synthesized as single polypeptide precursors that oligomerize in the endoplasmic reticulum to form trimers. Most of the enveloped proteins with fusion activity contain two noncovalently associated subunits: S1 S2 in coronaviruses, HA1 HA2 in influenza viruses, gp120 gp41 in HIV/SIV, GP1 GP2 in Ebola virus, and F1 F2 in paramyxoviruses, all of which are generated by proteolytic cleavage. Nevertheless, some enveloped proteins have no cleavage in their precursors and yet still maintain fusion activity, such as the S proteins of some coronaviruses (9) and GP of Ebola virus (10). A hydrophobic region in the membrane-anchored subunit of enveloped proteins, termed the fusion peptide in class I fusion proteins, has been shown to insert into the cellular membranes during the fusion process (11, 12). The regions following the fusion peptide have a 4-3 heptad repeat (HR) of hydrophobic residues, a sequence feature characteristic of coiled coils. The first heptad repeat, termed HR1 (HRA or N peptide), is followed by a short spacer domain and a second heptad repeat, termed HR2 (HRB or C-peptide), followed by another short 1 The abbreviations used are: MHV, mouse hepatitis virus; HA, hemagglutinin; HIV, human immunodeficiency virus; HR, heptad repeat; MES, 4-morpholineethanesulfonic acid; MMLV, Moloney murine leukemia virus; S protein, spike protein; SARS, severe acute respiratory syndrome; SIV, simian immunodeficiency virus; TM, transmembrane. This paper is available on line at

151 spacer and the transmembrane (TM) anchor (13). Biochemical and structural analyses of HA 2 (14), HIV-1/SIV gp41 (15 21), Ebola virus GP2 (22, 23), and SV5F1 (24) indicate that these heptad repeat regions form six-helix bundles. The N-terminal heptad repeat forms a central coiled coil, which is surrounded by three HR2 helices in an oblique, antiparallel manner. Among all enveloped glycoproteins, the membrane fusion mechanism of the HA of influenza viruses has been studied in greater detail (7, 8, 25). HA is proteolytically cleaved to generate a receptor binding subunit (HA 1 ) and an anchored subunit (HA 2 ) containing the fusion peptide. Numerous evidences suggest that the HA of influenza undergoes a conformational change, from a native (nonfusogenic) to fusion-active (fusogenic) state during the viral fusion process. In the native HA, part of the heptad repeat region of HA 2 forms a nonhelical loop (26) but converts into a coiled coil when exposed to low ph (14). The later conformation is generally regarded to be a fusogenic state because the low ph also activates influenza membrane fusion. This conformation change is also the basis of the spring-loaded mechanism for activation of viral fusion (27). The spike protein of MHV A59 has been identified as a class I fusion protein and shares common features with other viral fusion proteins (28). However, there are several significant differences between the membrane-anchored subunits of coronavirus spike proteins and HA 2, gp41, GP2, and SV5F1. First, the ectodomain is much larger (550 residues versus residues) in size in S protein. Second, the HR1 region is predicted to be much larger (more than 100 residues versus residues) in size by the learn-coil VMF program (29) and also verified by proteinase K-resistant experiments (28). Third, the putative fusion peptides of all other viral fusion proteins are located at the N terminus of the membrane-proximal subunits, whereas the S protein features an internal fusion peptide. Highly conserved HR1 and HR2 regions in coronavirus spike proteins suggest that they share similar three-dimensional fusion core structures and a common mechanism for viral fusion. The binding region of the HR1 HR2 complex of other coronaviruses and likely structures of their fusion cores can be proposed based on MHV fusion core structure and sequence alignment. Analogous to HIV C-peptides, HR2 peptides of coronaviruses are likely to act in a dominant-negative manner to inhibit hairpin formation, thereby inhibiting viral entry. Thus, drug discovery strategies aimed at inhibiting viral entry by blocking hairpin formation may also be applied to the inhibition of emerging infectious diseases such as SARS. Although previous biochemical and electron microscopic analyses have shown that the HR1 HR2 complex in the S protein of MHV forms a fusion core (28), the exact binding region of the HR1 HR2 complex and the detailed structure of the fusion core remain unknown. Here we report the determination of the crystal structure of the fusion core of MHV A59 spike protein to 2.5 Å resolution by x-ray crystallography and discuss the implications of the structure for coronavirus membrane fusion. EXPERIMENTAL PROCEDURES Expression, Purification, and Crystallization The expression, purification, and preliminary crystallographic studies of the MHV 2-Helix protein have been described elsewhere. 2 The PCR-directed gene was inserted into pet22b (Novagen) vector, and the selenomethionine MHV 2-Helix derivative was expressed in M9 medium containing 60 mg liter 1 selenomethionine in Escherichia coli strain BL21 (DE3). The product was purified by nickel-nitrilotriacetic acid affinity chromatography followed by gel filtration chromatography. The purified MHV 2-Helix derivative was dialyzed against 10 mm Tris, ph 8.0, 10 mm NaCl 2 Y. Xu, Z. Bai, L. Qin, X. Li, G. F. Gao, and Z. Rao, submitted for publication. Crystal Structure of MHV Spike Protein Fusion Core and concentrated to 8 mg ml 1. Crystals with good diffracting quality could be obtained in 0.1 M MES, ph 6.5, 10% PEG 4000 (v/v), 8% dimethyl sulfoxide (v/v), 5 mm hexaminecobalt trichloride after 3 days. The expression, purification, and crystallization of nmhv is the same as for MHV 2-Helix. Crystals with good diffracting quality could be obtained in 0.1 M MES, ph 6.5, 13% PEG 4000 (v/v), 5% dimethyl sulfoxide (v/v). Data Collection and Processing The MHV 2-Helix crystal was mounted on nylon loops and flash-frozen in an Oxford Cryosystems cold nitrogen gas stream at 100 K. Multiple wavelength anomalous dispersion data were collected by a rotation method using a MarCCD detector with synchrotron radiation on beamline 3W1A of the Beijing Synchrotron Radiation Facility. Data were collected from a single selenomethionyl derivative crystal at peak ( Å), inflection ( Å), and remote ( Å) wavelengths to 2.5 Å. Data collection from the nmhv 2-Helix crystal was performed in-house on a Rigaku RU2000 rotating anode x-ray generator operated at 48 kv and 98 ma (Cu K ; Å) with a MAR 345 image plate detector. The crystal was mounted on nylon loops and flash-frozen at 100 K using an Oxford Cryosystems cold nitrogen gas stream. Data were indexed and scaled using DENZO and SCALEPACK programs (30). Data collection statistics are shown in Table I. Phase Determination and Model Refinement For MHV 2-Helix structural determination, initial multiple wavelength anomalous dispersion phasing steps were performed using SOLVE (31) and followed by density modification by RESOLVE (32). The program O (33) was used for viewing electron density maps and manual building. The initial structure was subsequently refined to a final R-value of 22.4% and free R-value of 29.1%. The quality of the structure was verified by PRO- CHECK (34). None of the main chain torsion angles is located in disallowed regions of the Ramachandran plot. The statistics for the structure determination and refinement are summarized in Table I. The figures were generated with the programs GRASP (35), SPDBView (36), and MOLSCRIPT (37). RESULTS AND DISCUSSION Structure Determination Two peptides (HR1 and HR2, Fig. 1, A and B) encompassing the N-terminal and C-terminal heptad repeats of the MHV spike protein assemble into a stable trimer of heterodimers (28). The HR1 and HR2 regions of the MHV S protein consist of residues and residues , respectively (Fig. 1, A and B). The fusion core of the MHV spike protein was prepared as a single chain by linking the HR1 and HR2 domains via an eight-amino acid linker (GGSGGSGG, single amino acid abbreviation used here). The constructs and the encoded proteins were also called MHV 2-Helix (Fig. 1A). The preparation and characterization of the 2-Helix proteins will be reported elsewhere. 2 The MHV 2-Helix forms crystals that have unit cell parameters a b 48.3 Å, c 199.6Å, 90, 120 and belong to the space group R3. The crystals contain two MHV 2-Helix molecules/ asymmetric unit, and the diffraction pattern extends to 2.5 Å resolution. 2 The crystal structure of the MHV 2-Helix was solved by multiple wavelength anomalous dispersion from a single selenomethionyl derivative crystal. Four selenium sites could be located in one asymmetric unit from Patterson maps calculated using the program CNS 1.0 (38). The experimental electron density map was easily interpretable in the helical regions. The model was improved further by cycles of manual building and refinement using the programs O (33) and CNS (38). The structure was subsequently refined to a final resolution of 2.5 Å with an R-value of 22.4% and free R-value of 29.1%. The final model statistics are summarized in Table I. Description of the Structure In the three-dimensional structure of the MHV 2-Helix, the fusion core has a rod-shaped structure 80 Å in length and a maximum diameter of 28 Å. The complex is a six-helix bundle comprising a trimer of MHV 2-Helix. The center of this bundle consists of a parallel trimeric coiled coil of three HR1 helices that were packed by three antiparallel HR2 helices (Fig. 2, A and B). The N terminus of HR1 and the C terminus of HR2 are located at the same end of

152 30516 Crystal Structure of MHV Spike Protein Fusion Core TABLE I Data collection and model refinement statistics Data set statistics MHV 2-Helix Peak Edge Remote nmhv 2-Helix native Space group R3 R3 R3 R3 Unit cell parameters (Å) Wavelength (Å) Resolution limit (Å) Observed reflections 49,717 50,167 49,352 83,487 Unique reflections 6,842 6,867 6,804 12,068 Completeness (%) 100(100) a 100(100) 100(100) 98.6(100) I/ (I) 14.3(4.1) 14.6(4.5) 13.7(3.8) 28.7(4.3) R b merge (%) 9.4(34.0) 8.4(31.0) 9.3(36.1) 9.4(29.9) Final refinement statistics MHV 2-Helix nmhv 2-Helix R c work (%) R d free (%) Resolution range (Å) Total reflections used 5,833 11,689 No. of reflections in working set 5,501 11,092 No. of reflections in test set r.m.s.d. e bonds(å) r.m.s.d. angles( ) a Numbers in parentheses correspond to the highest resolution shell. b R merge h l I ih I h / h I I h, where I h is the mean of the observations I ih of reflection h. c R work ( F obs F calc )/ F obs. d R free is the R-factor for a subset (5%) of reflections that was selected prior refinement calculations and not included in the refinement. e r.m.s.d., root mean square deviation from ideal geometry. the six-helix bundles, placing the fusion peptide and TM domains close together. A region of about 190 amino acids would be located at the other end of the six-helix bundle between HR1 and HR2. The linker and several terminal residues were disordered in both molecules. In one asymmetric unit, one molecule includes residues in HR1 and in HR2, and the other molecule includes residues in HR1 and in HR2. The two trimers are created by the same 3-fold axis of the crystallographic unit cells and are both parallel with the 3-fold axis of the crystallographic unit cells, and there is about 30 degrees rotation in the orientation parallel with the 3-fold axis between the two trimers. The root mean square deviation of the two molecules in one asymmetric unit is 0.36 Å. There is only one weak hydrogen bond between the two parallel trimers, from OH (Tyr 1233 ) to O (Glu 1254 ) with a distance of 3.34 Å. Residues of HR1 fold into a 15-turn -helix stretching the entire length of the coiled coil. As in other naturally occurring coiled coils of the fusion core, the residues in the a and d positions of HR1 are predominantly hydrophobic (Fig. 1B). A sequence alignment of MHV with other representative coronavirus spike proteins shows that the residues at these two heptad repeat positions are highly conserved (Fig. 1B). Residues of HR2 form a 5-turn amphipathic -helix, whereas residues and form two extended chains at the N and C terminus of HR2, respectively. Each HR2 fits into the long grooves formed by the interface of the three HR1 helices, and no interaction is observed between individual HR2 helices (Fig. 2, A and B). The C terminus of HR2 ends with Glu 1254, which is aligned with Gln 970 of HR1; Gln 970 is also the N terminus of the HR1 domain. The N terminus of HR2 starts with Asp 1216, which is aligned with Ile 1023 of HR1 (Fig. 2C). Linker between HR1 and HR2 and nmhv 2-Helix Structure To verify whether the linker (GGSGGSGG) between HR1 and HR2 affects the natural structure of the MHV 2-Helix, we made a new construct (termed nmhv 2-Helix) that includes a shorter HR1 and longer linker. In the nmhv 2-Helix construct, 152 the HR1 consists of residues , the HR2 consists of residues , and the new fusion core was prepared as a single chain by linking the HR1 and HR2 domains via a 22-amino acid linker (LVPRGSGGSGGSGGLEVLFQGP), which is flexible and long enough to allow the HR1 and HR2 to form a natural interaction (Fig. 1A). The nmhv 2-Helix forms crystals in space group R3 with lattice dimensions of a b 51.6 Å, c Å, 90, 120. The crystals contain two 2-Helix molecules/asymmetric unit and diffract x-rays to a resolution of at least 2.1 Å. The crystal structure of nmhv 2-Helix was determined by molecular replacement with the MHV 2-Helix structure as a search model. Rotation and translation function searches were performed in CNS (38). The model was improved further by cycles of manual building and refinement using the programs O (33) and CNS (38). The final R-value and free R-value for the refinement are 26.2 and 29.8%, respectively. The final model statistics are listed in Table I. The nmhv 2-Helix structure is largely similar to the MHV 2-Helix structure, with the exception of several residues at the N terminus of HR2 (Fig. 2C). The overall root mean square deviation between the two structures is 0.48 Å, which is calculated using the CCP4 program LSQKAB. The nmhv 2-Helix structure also contains two molecules/asymmetric unit. One molecule includes residues in HR1 and in HR2, whereas the other molecule includes residues in HR1 and in HR2. The linker is also disordered and cannot be traced in the structure. The N terminus of HR2 cannot be seen in the structure because the C terminus of HR1, which is important for binding the N terminus of HR2, has been discarded in the new construct. The linker is long and flexible enough, so we can conclude that both structures surely represent the natural structure of the complex of HR1 and HR2 because the choice of linker does not affect the real interaction between the two heptad repeat regions. We will focus our following structural analysis on MHV 2-Helix structure. Interactions between HR1 and HR2 Three HR2 helices pack obliquely against the outside of the HR1 coiled coil trimer in an antiparallel orientation. The HR2 helices interact with HR1

153 Crystal Structure of MHV Spike Protein Fusion Core FIG. 1. Structure determination of the MHV spike protein fusion core trimer. A, schematic representation of coronavirus MHV A59 spike protein and the MHV 2-Helix and nmhv 2-Helix constructs. S1 and S2 are formed after proteolytic cleavage (vertical arrow) and noncovalently linked. The enveloped protein has an N-terminal signal sequence (SS) and a TM domain adjacent to the C terminus. S2 contains two HR regions (hatched bars), termed HR1 and HR2 as indicated. FP (hatched bars) is a putative fusion peptide followed by HR1 region. For the MHV 2-Helix, two HR regions were linked to a single polypeptide with an 8-residue linker (GGSGGSGG). For the nmhv 2-Helix, HR2 and a shortened HR1 were linked with a 22-amino acid linker (LVPRGSGGSGGSGGLEVLFQGP). B, sequence alignment of coronavirus spike protein HR1 and HR2 regions. Letters above the sequence indicate the predicted hydrophobic HR a and d residues, which are highly conserved. C, helical wheel representation of HR1 and HR2. Three HR1 helices and one HR2 helix are represented as helical wheel projections. The view is from the top of the structure. The three central HR1 helices form a central hydrophobic core with the interaction of residues in the a and d positions. The three HR2 helices pack against these hydrophobic surface grooves through interactions with residues in the a and d positions in HR2 and e and g positions in HR1. These residues, mediating the interactions between HR1 and HR2, are always hydrophobic and conserved (see B). mainly through hydrophobic residues in three grooves on the surface of the central coiled coil trimer (see Fig. 4A). Sequence comparison between MHV and other coronavirus spike proteins shows that residues contributing to the HR1/HR2 interaction (e and g positions in HR1, a and d positions in HR2) are highly conserved (Fig. 1B). This pattern of sequence conservation can also be shown by a helical wheel representation of three HR1 helices and one HR2 helix (Fig. 1C). In this diagram, residues in the a and d positions of HR2 pack against residues in the e and g positions of HR1, mainly through hydrophobic interactions. Sequence 153 comparison between MHV and HcoV-OC43 spike proteins shows that no nonconservative changes exist at the e and g positions of HR1, and only three such changes (L to I, L to F and I to L) occur in HR2 at the a and d positions. In contrast, 7 of 26 nonconservative changes occur at the outside f, b, and c positions in HR1, and 5 of 12 nonconservative changes occur at positions other than a and d in the helical region ( ) of HR2 (Fig. 1B). Comparison with Other Fusion Protein Structures The structure of the MHV fusion core was compared with the known structures of other viral fusion proteins (Fig. 3A). Al-

154 30518 Crystal Structure of MHV Spike Protein Fusion Core FIG. 2.Overall views of the fusion core structure and superposition of nmhv (new construct for MHV fusion core) and MHV fusion core. A, top view of the MHV fusion core structure showing the 3-fold axis of the trimer. B, side view of the MHV fusion core structure showing the six-helix bundle. C, side view showing the superposition of nmhv fusion core (colored in blue) and MHV fusion core (colored in yellow). The columns at both sides of the map represent two HR1 and HR2 regions of nmhv and MHV fusion cores. The number at the end of these columns represents the end residues in the two structures. though there are significant differences in the sequences, sizes, and structural properties of these viral fusion proteins, the similarities in their overall structures suggest a common mechanism for membrane fusion. These fusion core proteins are all trimeric coiled coils with the putative fusion peptides located at the N-terminal end of HR1 and the TM domains located at the C-terminal end of HR2 (7). The three HR2 polypeptides that form the outer layer of the central core vary in conformation among these structures, but they always form -helices and always pack antiparallel to the interior coiled coil. In this pattern, which has been proposed to be a fusogenic conformation, all of these fusion proteins would have their fusion peptides and TM anchors aligned at the same end of the coiled coil. There are three major differences between the MHV and other viral fusion core structures and their relevant regions, indicating unique features of fusion core structure in coronavirus S proteins. First, the conformation of HR2 in the fusion core structure is different from those of all other fusion core HR2 regions (Fig. 3A). In the MHV fusion core structure, a major 5-turn -helix and a single-turn -helix could be observed, and the remaining parts are extended segments. We can divide the HR2 polypeptide into five parts (Fig. 4A, right), parts 1 5. Part 4 is a typical pattern and forms a 5-turn -helix of which the residues at positions a and d pack against residues at the e and g positions of the HR1. Part 2 is also an -helix that exhibits the 3-4 spacing pattern, although it has only four residues, FEKL. Of these four residues, the residues in the a position (Phe 1221 ) and d position (Leu 1224 ) also pack against hydrophobic grooves formed on the surface of the HR1 core. Parts 1, 3, and 5 should also be -helical based on prediction by the learn-coil VMF program (29). However, the structure shows 154 they are all extended to form a strand-like conformation. In these three parts, there is an interesting O-X-O motif where O represents hydrophobic residues, and X represents any residue but is generally hydrophilic. Part 1 contains the residues LSL, part 3 contains the residues VTL and LDL, and part 5 contains the residues INL. The two hydrophobic residues in these motifs pack against hydrophobic grooves on the surface of the HR1 core, either facing the central core or aligning with the hydrophobic groove. As a result, the O residues form hydrophobic interactions to stabilize the six-helix bundle, leaving the X residue directed into solvent. This pattern, which we think is a major reason why these three parts do not form -helices, is also observed in the structure of SV5F (24), HRSV F (39), MMLV TM (40) and Ebola GP2 (41) (Fig. 4B). In these structures, partial regions in HR2 or C-peptide are extended and strand-like rather than -helical. In the pattern in HR2 of these glycoproteins, the presence of an O-X-O motif would result in hydrophobic residues interacting with the hydrophobic grooves on the surface of HR1 core, thus destroying the typical -helix. In these three-dimensional structures, the distance between the two hydrophobic residues in the O-X-O motifs is about 5 Å, which is also the distance between the two adjacent helices. Thus, the two hydrophobic residues could exactly pack against the grooves of the central coiled coil formed by three HR1 helices. This compatibility of HR2 segments makes the fusion core more stable in solvent because most of the hydrophobic residues in HR2 are packed against the central core, leaving the hydrophilic residues exposed to solvent. This pattern also explains why not all residues in HR2 of fusion cores from many other viral proteins form -helices and why HR1 structures of these fusion cores are highly conserved, whereas HR2 regions always differ in their three-dimensional conformations (Fig. 3A). Second, a proteinase K-resistant fusion core of MHV A59 spike protein has been reported by Bosch et al. (28). After digestion by proteinase K, the fusion core comprising residues in HR1 and in HR2 remains intact. In this fusion core, the HR1 region is about 30 residues longer than the HR1 region of the fusion core we constructed. Although the fusion core structure we determined here is only part of the proteinase K-resistant fusion core, we propose that residues and in HR1 would also form coiled coils in the natural fusion core on the basis of their resistant capacity and other biochemical analyses (28). In the proteinase K-resistant fusion core of MHV spike protein, the central coiled coil HR1 region has about 80 amino acids, which is considerably longer than HR1 segments in other fusion cores such as HIV gp41 and SV5F1. This length is comparable with that of the fusion core in influenza HA, whose mechanisms for the membrane fusion have been studied in extensive detail (25). In addition, the HR1 region of the MHV spike protein is predicted to contain more than 100 amino acids by the learncoil VMF program (29). This long helical coiled coil might be consistent with the long sequence of S protein, which is more than 1200 residues, to form the central skeleton of spikes on the surface of coronavirus. Third, although the fusion peptide is not part of the fusion core, it is also very important for investigating viral fusion mechanism. The putative fusion peptide of the MHV A59 spike protein is located at the N-terminal end of the HR1 region, and the cleavage site of the spike protein is about 250 amino acids away from the fusion peptide (28). The cleavage sites of other class I viral fusion proteins are all typically located adjacent to the fusion peptides (13). In the latter pattern, the likely role of the six-helix bundle structure is to facilitate juxtaposition of the viral and cellular membranes by bringing the fusion pep-

155 Crystal Structure of MHV Spike Protein Fusion Core FIG. 3. Viral fusion proteins and models for membrane fusion. A, comparison of MHV fusion core with other viral fusion protein structures. The proteins under comparison include SV5F, Ebola GP2, HIV gp41, MMLV Env-TM, and low ph-induced influenza virus HA, tbha2. Top and side views are shown for the six fusion core structures. B, model for coronavirus-mediated membrane fusion. The first state is the native conformation of coronavirus spike protein on the surface of viral membrane. It has been reported that the spike protein is trimeric in this conformation and about 200 Å in length (6), but the exact structure of the full-length protein remains unknown. The second state is the prehairpin state of the S2 subunit. After several conformational changes, the fusion peptide inserts into the cellular membrane with the aid of other regions of S protein and possibly including the receptor. Although the internal fusion peptide is not exposed at the N-terminal of S2, it could insert into part of the target membrane by means of some hydrophobic residues. This insertion would be stable enough to drive the membrane motion with the conformational changes of HR1 region, which is adjacent to the fusion peptide. The third state is conformational change and juxtaposition of the target and viral membranes. With the help of other regions of S protein, the HR1 and HR2 regions move together and facilitate juxtaposition of the cellular and viral membrane. The last state is the postfusion conformation. The coiled coil will reorient with its long axis parallel to the membrane surface. The fused cellular and viral membranes make it possible for subsequent viral infections. tide, which inserts into the cellular membrane, close to the transmembrane segment, which is anchored in the viral membrane (7). In the case of the MHV spike protein, the question of why the HR1 region and fusion peptide are so far away from the cleavage site remains unknown. Nevertheless, some viruses such as coronavirus (9) and Ebola virus (10) do not have cleavage sites in fusion proteins but still retain their fusion activity. This indicates that the location of the fusion peptides, whether exposed at the N terminus of the membrane-anchored polypeptide or not, is not an essential requirement for the viral fusion. We will give a possible mechanism in the further discussion. Evidence for the Conformational Change Structural studies of the influenza virus HA and HIV gp41 have established a paradigm for understanding the mechanisms of viral and cellular membrane fusion (7). For coronaviruses, direct evidence for the conformational change in spike protein is lacking, although the crystal structure of the MHV fusion core bears similarity to these fusion-active state molecules. The structure of the MHV 2-Helix could correspond to the fusion core of MHV spike protein in either the fusogenic or the native form of the envelope glycoprotein, or both. Several considerations provide good evidence that the fusion core in the crystal structure presented here is the final, stable form of the protein, which is a fusion-active state following one or more conformational changes. First, the fusion core of MHV spike protein is exceedingly 155 stable to both thermal denaturation and proteinase K digestion (28). The six-helix bundle has a melting temperature of about 85 C and could not be separated in general SDS-PAGE unless boiled at 100 C with a loading buffer containing a high concentration of SDS (28). These properties indicate that the complex is very stable and could not be dissociated by any biologically relevant interaction. This form of fusion core must be present at the later stage of conformational changes for viral fusion, although it is not known whether the complex maintains this conformation throughout the entire process. Second, virus-cell entry inhibition and cell-cell fusion inhibition experiments (28) also provide strong evidence that the fusion core is formed after one or more conformational changes. HR2 of the MHV 2-Helix could block viral entry and cell-cell fusion in a concentration-dependent manner and appears to be a potent inhibitor (28). In HIV gp41, the C-peptide and its derivatives have been shown to act as dominant-negative inhibitors by binding to the endogenous N-peptide coiled coil trimer within viral gp41 (19, 42, 43). A reasonable interpretation of the data for the MHV fusion core is that HR2 functions in a dominant-negative manner similar to the C-peptide in HIV gp41 by binding to the transiently exposed coiled coil in the prehairpin intermediate and thus preventing the conformational changes required for viral fusion. Third, mutations in the MHV spike protein that abolish viral-cell fusion activity often map to the HR1 or HR2 residues, which are expected to stabilize the fusion core structure re-

156 30520 Crystal Structure of MHV Spike Protein Fusion Core FIG. 4.O-X-O motifs in HR2 regions of MHV and the comparison with those of other fusion proteins. A. Left and center, surface map showing the hydrophobic grooves on the surface of three central HR1 helices. Three HR2 helices pack against the hydrophobic groove in an antiparallel manner. The helical regions in HR2 extended regions could be observed clearly. The helical region of HR2 just packs against the deep groove, and the extended region packs against the shallow groove. Right, detailed structure of O-X-O motifs in MHV HR2 region. One HR2 helix is divided into five parts based on its secondary structure. The helical regions (parts 2 and 4) HR2 are colored in red, and extended regions (parts 1, 3, and 5) are colored in blue. The essential residues of the three extended regions and O-X-O motifs in these regions are shown; residues colored in green represent the hydrophobic residues in O-X-O motifs. The three panels on the left show the enlarged images of parts 1, 3, and 5. The hydrophobic residues in these motifs are all packed against the hydrophobic grooves on the surface of three HR1 helices. B, detailed structures of O-X-O motifs in other fusion proteins including SV5F, HRSV F, MMLV Env-TM, and Ebola GP2. They all contain similar motifs in HR2 regions. The regions in which O-X-O motifs are located form extended regions but not -helices, in a way similar to the MHV 2-Helix. ported here. These studies show that mutations in some essential positions in HR regions abolish infectivity and membrane fusion (44, 45). The L981K and F977K mutations are particularly noteworthy because cells expressing mutant spike proteins with one of these mutations are almost completely defective in membrane fusion, although the surface expression level of spike proteins remains the same as for the wild type (44). Residue Leu 981 is in the a position and Phe 977 in the d position of the HR1 peptide, and thus both are essential for the formation of the central hydrophobic coiled coil (Fig. 1C). The L981K and F977K mutations in HR1 region substitute hydrophobic residues with hydrophilic residues, destroying the hydrophobic interaction and the formation of the six-helix bundles and subsequently abolishing the membrane fusion. In contrast, 156 replacement of the same residues with other hydrophobic amino acids (F977L, L981I) does not reduce fusion activity (44). L1224A/L1231A, L1224A/I1238A, L1224A/L1245A, and I1231A/L1245A mutations in HR2 also abolish fusion activity greatly (45). In these double substitution mutants, the Leu and Ile residues are both in d positions in HR2 and are also very important for the formation of the fusion core (Fig. 1C). These double substitution mutants could not maintain a stable coiled coil structure even though Leu or Ile was changed for Ala, a hydrophobic residue. The locations of these particular mutations indicate that the interactions between HR1 and HR2 are critical for membrane fusion. Lastly, the structural similarities between the MHV 2-Helix complex, the fusion core of low ph-induced conformation of

157 influenza HA2 (14), and the structure of HIV gp41 (19), each of which has been proposed to have fusion-active conformations, indicate that the MHV 2-Helix structure studied here represents the core of the fusogenic conformation of spike protein after conformational change (Fig. 3A). In all three structures, the putative hydrophobic fusion peptides are located close to the N-terminal end of the HR1 region, which forms a central coiled coil. Three strands of HR2 which pack against the coiled coil trimer in an antiparallel manner stabilize this hydrophobic coiled coil. These common features suggest that the MHV spike protein also possesses a conformational change mechanism similar to influenza HA and HIV gp41 (7). Implications for Models of Membrane Fusion Mechanisms Although we have no structures of full-length MHV spike protein either in the prefusion or postfusion state, we can propose a model for MHV membrane fusion mechanism based on the fusion core structure studied here and previous analysis of spike proteins. Current models for the class I viral fusion mechanisms suggest that the exposed fusion peptide located at the N terminus of the membrane-anchored subunit may be important for the juxtaposition of two membranes prior to fusion (7). For coronaviruses, the S2 fragment contains a putative internal fusion peptide that is not exposed at the N terminus (44). This pattern of internal fusion peptide in the MHV spike protein is reminiscent of fusion loops in class II viral fusion proteins, which are internally located in the fusion protein (46, 47). Recent structural studies of the dengue virus envelope protein show that the highly conserved internal fusion loop penetrates only 6 Å into the hydrocarbon layer of the cellular membrane (46). In this structure, an aromatic anchor formed by Trp 101 and Phe 108 inserts into the cellular membrane. Studies of a 20-residue influenza virus A fusion peptide with a detergent micelle suggests a kinked -helix, with the N and C termini embedded in the outer leaflet and the kink on the surface (48). Both the fusion loop from dengue virus E protein and the fusion peptide from influenza virus indicate that the anchoring into the lipid bilayer may not require complete insertion of the fusion peptide into the membrane. The exposure of the fusion peptide at the N terminus of the fusion protein is also not essential for the viral fusion. In this case, which also applies to several coronavirus spike proteins, the fusion peptides may insert into the membrane via several hydrophobic resides or a kinked hydrophobic loop (Fig. 3B). In our model shown in Fig. 3B, the internal fusion peptide of spike proteins forms a small hydrophobic core, which would insert into the membrane after several conformational changes induced by receptors or ph changes. HR2 and HR1 would then form a coiled coil to facilitate juxtaposition of the cellular and viral membranes, followed by virus-cell membrane fusion. In coronavirus spike proteins, S1 is mainly responsible for binding to receptor and S2 for fusion (5). The flexibility between HR1 and HR2 is not sufficient enough to allow for complete conformational changes to occur, although the two heptad repeat regions have a great tendency to pack closely together. Instead, other intervening regions of the spike protein should promote the formation of the fusion core. Further experiments should reveal how the exact conformational change occurs and how best to inhibit the viral fusion. Potential Inhibitors of Coronaviruses Entry into Cells The sequence alignment of HR1 and HR2 regions in spike proteins among coronaviruses shows significant similarity in their heptad repeat regions (Fig. 1B). For example, between MHV and SARS, identity of HR1 is 60% and positive is 91%, identity of HR2 is 35 and positive is 85%. Important residues for six-helix bundle formation, located in the a, d, e, and g positions of HR1 Crystal Structure of MHV Spike Protein Fusion Core and a and d positions of HR2, are all highly conserved. We can conclude that coronavirus spike proteins share a similar binding region of the HR1 HR2 complex and three-dimensional fusion core structure. Analogous to the HIV C-peptides, the HR2 region of coronavirus spike proteins most likely functions in a dominant-negative manner by binding to the transiently exposed hydrophobic grooves in the prehairpin intermediate and consequently blocking the formation of the fusion-active hairpin structures (43). These strategies have been used successfully in fusion inhibitors design for HIV (7, 42, 49 51). A similar approach may be applied to identify inhibitors of coronavirus infection. HR1 and HR2 regions and their derivatives are all potential inhibitors. Cavities and grooves on the surface of the central coiled coil are strong potential binding sites for small molecule inhibitors. In conclusion, the crystal structure of MHV fusion core shows the first fusion core structure of any coronavirus. Although the structure shares common features with those of other viral fusion proteins, it has unique characteristics that distinguish it from other fusion core structures. Sequence alignment of HR regions among coronaviruses indicates a similar structure among coronavirus spike proteins and suggests a common mechanism for viral fusion. This structure will also open an avenue toward the structure-based fusion inhibitor design of peptides, or peptide analogs, e.g. small molecules, targeted against emerging infectious diseases, such as SARS. Acknowledgment We thank Dr. Mark Bartlam for comments and critical reading. REFERENCES 1. Spaan, W., Cavanagh, D., and Horzinek, M. C. (1988) J. Gen. Virol. 69, Lee, H. J., Shieh, C. K., Gorbalenya, A. E., Koonin, E. V., La Monica, N., Tuler, J., Bagdzhadzhyan, A., and Lai, M. M. (1991) Virology 180, Siddell, S., Wege, H., and Ter Meulen, V. (1983) J. Gen. Virol. 64, Rota, P. A., Oberste, M. S., Monroe, S. S., Nix, W. A., Campagnoli, R., Icenogle, J. P., Penaranda, S., Bankamp, B., Maher, K., Chen, M. H., Tong, S., Tamin, A., Lowe, L., Frace, M., DeRisi, J. L., Chen, Q., Wang, D., Erdman, D. D., Peret, T. C., Burns, C., Ksiazek, T. G., Rollin, P. E., Sanchez, A., Liffick, S., Holloway, B., Limor, J., McCaustland, K., Olsen-Rasmussen, M., Fouchier, R., Gunther, S., Osterhaus, A. D., Drosten, C., Pallansch, M. A., Anderson, L. J., and Bellini, W. J. (2003) Science 300, Gallagher, T. M., and Buchmeier, M. J. (2001) Virology 279, Davies, H. A., and Macnaughton, M. R. (1979) Arch. Virol. 59, Eckert, D. M., and Kim, P. S. (2001) Annu. Rev. Biochem. 70, Hernandez, L. D., Hoffman, L. R., Wolfsberg, T. G., and White, J. M. (1996) Annu. Rev. Cell Dev. Biol. 12, Yamada, Y. K., Takimoto, K., Yabe, M., and Taguchi, F. (1997) Virology 227, Wool-Lewis, R. J., and Bates, P. (1999) J. Virol. 73, Hernandez, L. D., Peters, R. J., Delos, S. E., Young, J. A., Agard, D. A., and White, J. M. (1997) J. Cell Biol. 139, Damico, R. L., Crane, J., and Bates, P. (1998) Proc. Natl. Acad. Sci. U. S. A. 95, Chambers, P., Pringle, C. R., and Easton, A. J. (1990) J. Gen. Virol. 71, Bullough, P. A., Hughson, F. M., Skehel, J. J., and Wiley, D. C. (1994) Nature 371, Lu, M., Blacklow, S. C., and Kim, P. S. (1995) Nat. Struct. Biol. 2, Weissenhorn, W., Dessen, A., Harrison, S. C., Skehel, J. J., and Wiley, D. C. (1997) Nature 387, Blacklow, S. C., Lu, M., and Kim, P. S. (1995) Biochemistry 34, Weissenhorn, W., Wharton, S. A., Calder, L. J., Earl, P. L., Moss, B., Aliprandis, E., Skehel, J. J., and Wiley, D. C. (1996) EMBO J. 15, Chan, D. C., Fass, D., Berger, J. M., and Kim, P. S. (1997) Cell 89, Tan, K., Liu, J., Wang, J., Shen, S., and Lu, M. (1997) Proc. Natl. Acad. Sci. U. S. A. 94, Caffrey, M., Cai, M., Kaufman, J., Stahl, S. J., Wingfield, P. T., Covell, D. G., Gronenborn, A. M., and Clore, G. M. (1998) EMBO J. 17, Weissenhorn, W., Carfi, A., Lee, K. H., Skehel, J. J., and Wiley, D. C. (1998) Mol. Cell 2, Weissenhorn, W., Calder, L. J., Wharton, S. A., Skehel, J. J., and Wiley, D. C. (1998) Proc. Natl. Acad. Sci. U. S. A. 95, Baker, K. A., Dutch, R. E., Lamb, R. A., and Jardetzky, T. S. (1999) Mol. Cell 3, Wiley, D. C., and Skehel, J. J. (1987) Annu. Rev. Biochem. 56, Wilson, I. A., Skehel, J. J., and Wiley, D. C. (1981) Nature 289, Carr, C. M., and Kim, P. S. (1993) Cell 73, Bosch, B. J., van der Zee, R., de Haan, C. A., and Rottier, P. J. (2003) J. Virol. 77, Singh, M., Berger, B., and Kim, P. S. (1999) J. Mol. Biol. 290,

158 30522 Crystal Structure of MHV Spike Protein Fusion Core 30. Otwinowski, Z., and Minor, W. (1997) Methods Enzymol. 276, Terwilliger, T. C., and Berendzen, J. (1999) Acta Crystallogr. Sect. D Biol. Crystallogr. 55, Terwilliger, T. C. (2001) Acta Crystallogr. Sect. D Biol. Crystallogr. 57, Jones, T. A., Zou, J. Y., Cowan, S. W., and Kjeldgaard (1991) Acta Crystallogr. Sect. A 47, Laskowski, R. A., MacArthur, M. W., Moss, D. S., and Thornton, J. M. (1993) J. Appl. Crystallogr. 26, Nicholls, A., Sharp, K. A., and Honig, B. (1991) Proteins 11, Guex, N., and Peitsch, M. C. (1997) Electrophoresis 18, Kraulis, P. J. (1991) J. Appl. Crystallogr. 24, Brunger, A. T., Adams, P. D., Clore, G. M., DeLano, W. L., Gros, P., Grosse- Kunstleve, R. W., Jiang, J. S., Kuszewski, J., Nilges, M., Pannu, N. S., Read, R. J., Rice, L. M., Simonson, T., and Warren, G. L. (1998) Acta Crystallogr. Sect. D Biol. Crystallogr. 54, Zhao, X., Singh, M., Malashkevich, V. N., and Kim, P. S. (2000) Proc. Natl. Acad. Sci. U. S. A. 97, Fass, D., and Kim, P. S. (1995) Curr. Biol. 5, Malashkevich, V. N., Schneider, B. J., McNally, M. L., Milhollen, M. A., Pang, J. X., and Kim, P. S. (1999) Proc. Natl. Acad. Sci. U. S. A. 96, Wild, C. T., Shugars, D. C., Greenwell, T. K., McDanal, C. B., and Matthews, T. J. (1994) Proc. Natl. Acad. Sci. U. S. A. 91, Chan, D. C., and Kim, P. S. (1998) Cell 93, Luo, Z., and Weiss, S. R. (1998) Virology 244, Luo, Z., Matthews, A. M., and Weiss, S. R. (1999) J. Virol. 73, Modis, Y., Ogata, S., Clements, D., and Harrison, S. C. (2004) Nature 427, Gibbons, D. L., Vaney, M. C., Roussel, A., Vigouroux, A., Reilly, B., Lepault, J., Kielian, M., and Rey, F. A. (2004) Nature 427, Han, X., Bushweller, J. H., Cafiso, D. S., and Tamm, L. K. (2001) Nat. Struct. Biol. 8, Eckert, D. M., and Kim, P. S. (2001) Proc. Natl. Acad. Sci. U. S. A. 98, Eckert, D. M., Malashkevich, V. N., Hong, L. H., Carr, P. A., and Kim, P. S. (1999) Cell 99, Root, M. J., Kay, M. S., and Kim, P. S. (2001) Science 291,

159 THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 279, No. 47, Issue of November 19, pp , by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A. Crystal Structure of Severe Acute Respiratory Syndrome Coronavirus Spike Protein Fusion Core* Received for publication, August 2, 2004, and in revised form, August 25, 2004 Published, JBC Papers in Press, September 1, 2004, DOI /jbc.M Yanhui Xu, Zhiyong Lou, Yiwei Liu, Hai Pang, Po Tien, George F. Gao, and Zihe Rao From the Laboratory of Structural Biology, Tsinghua University, Beijing and National Laboratory of Bio-Macromolecules, Institute of Biophysics, Beijing , China, Nuffield Dept of Clinical Medicine, John Radcliffe Hospital, Oxford University, Oxford OX3 9DU, United Kingdom, and Institute of Microbiology, Chinese Academy of Sciences, Beijing , China Severe acute respiratory syndrome coronavirus is a newly emergent virus responsible for a recent outbreak of an atypical pneumonia. The coronavirus spike protein, an enveloped glycoprotein essential for viral entry, belongs to the class I fusion proteins and is characterized by the presence of two heptad repeat (HR) regions, HR1 and HR2. These two regions are understood to form a fusion-active conformation similar to those of other typical viral fusion proteins. This hairpin structure likely juxtaposes the viral and cellular membranes, thus facilitating membrane fusion and subsequent viral entry. The fusion core protein of severe acute respiratory syndrome coronavirus spike protein was crystallized, and the structure was determined at 2.8 Å of resolution. The fusion core is a six-helix bundle with three HR2 helices packed against the hydrophobic grooves on the surface of central coiled coil formed by three parallel HR1 helices in an oblique antiparallel manner. This structure shares significant similarity with the fusion core structure of mouse hepatitis virus spike protein and other viral fusion proteins, suggesting a conserved mechanism of membrane fusion. Drug discovery strategies aimed at inhibiting viral entry by blocking hairpin formation, which have been successfully used in human immunodeficiency virus 1 inhibitor development, may be applicable to the inhibition of severe acute respiratory syndrome coronavirus on the basis of structural information provided here. The relatively deep grooves on the surface of the central coiled coil will be a good target site for the design of viral fusion inhibitors. Severe acute respiratory syndrome (SARS) 1 is a new lifethreatening form of atypical pneumonia (1, 2) caused by a novel coronavirus, SARS-CoV (3 10). Phylogenetic analysis of SARS- CoV shows that it is not closely related to any of the previously * This work was supported by Projects 973 and 863 of the Ministry of Science and Technology of China (Grants 200BA711A12, G , GZ236(202/9), and 2003CB The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. The atomic coordinates and structure factors (code 1WNC ) have been deposited in the Protein Data Bank, Research Collaboratory for Structural Bioinformatics, Rutgers University, New Brunswick, NJ ( To whom correspondence and reprint requests should be addressed. Tel.: ; Fax: ; raozh@xtal. tsinghua.edu.cn. 1 The abbreviations used are: SARS, severe acute respiratory syndrome; HIV, human immunodeficiency virus; HR, heptad repeat; SARS-CoV, severe acute respiratory syndrome coronavirus; MHV, mouse hepatitis virus characterized coronaviruses isolated from either humans or animals, and it has therefore been assigned to a new, distinct group within the genus (4, 5, 9, 10). Coronaviruses are enveloped, positive-strand RNA viruses with the largest genomes of any RNA virus and are characterized by 3 4 envelope proteins embedded on the surface (11, 12). Both the receptor binding and the subsequent membrane fusion process of coronaviruses are mediated by the spike membrane glycoprotein (S protein) (13). Recent studies show that murine coronavirus (mouse hepatitis virus (MHV)) uses a spike-mediated membrane fusion mechanism similar to that of so-called class I virus fusion proteins (14, 15). Class I viral fusion proteins, including the hemagglutinin protein of influenza virus, gp160 of human immunodeficiency virus (HIV-1), glycoprotein of Ebola virus, and fusion protein (F protein) of paramyxovirus (16, 17), are all type I transmembrane glycoproteins that are displayed on the surface of viral membrane as oligomers. Most of these glycoproteins are synthesized as single chain precursors containing a protease cleavage site, and these precursors are cleaved into two noncovalently associated subunits: S1 and S2 in coronavirus, hemagglutinin 1 and 2 in influenza virus, gp120 gp41 in HIV/simian immunodeficiency virus, glycoprotein-1 and -2 in Ebola virus, and F1 and -2 in paramyxovirus. Class I viral fusion proteins also contain a fusion peptide and at least two heptad repeat regions, termed HR1 and HR2. After binding to the receptor or induced by low ph, the fusion proteins undergo a series of conformational changes to mediate membrane fusion. The first step involves exposure of the fusion peptide, a hydrophobic region in the membrane-anchored subunit, which then inserts into the cellular lipid bilayer. Subsequently, HR1 and HR2 peptides form a trimer-ofhairpins-like structure via a transient pre-hairpin intermediate to facilitate juxtaposition of the viral and cellular membranes followed by virus-cell membrane fusion and viral entry. Biochemical and structural analysis of these fusion cores from class I viral fusion proteins shows that these complexes of two heptad repeat regions form a stable six-helix bundle, which is designated as a fusion core in which three HR1 helices form a central coiled coil surrounded by three HR2 helices in an oblique, antiparallel manner (18 26). The coronavirus spike protein shares many features with other class I viral fusion proteins. It is a type I membrane protein that associates into trimers on the surface of coronavirus membrane (27). The distal subunit (S1) of the spike protein contains the receptor binding domain (28, 29), and the membrane-anchored subunit (S2) contains a putative internal fusion peptide and two heptad repeat regions (HR1 and HR2) (14, 15). Agents that prevent conformational changes in the fusion protein by stabilizing the intermediate state are expected to prevent fusion activation and, thus, inhibit viral entry. In the This paper is available on line at

160 Crystal Structure of SARS Spike Protein Fusion Core FIG. 1. Structure determination of the MHV spike protein fusion core trimer. A, schematic diagram of SARS-CoV spike protein indicating the location of structurally significant domains. S1 and S2 are formed after proteolytic cleavage (vertical arrow) and noncovalently linked. The enveloped protein has an N-terminal signal sequence (SS) and a transmembrane domain (TM) adjacent to the C terminus. S2 contains two HR (heptad repeat) regions (hatched bars), HR1 and HR2 as indicated. The HR1 ( ) and HR2 ( ) used in this study were derived from the LearnCoil-VMF prediction program (37). The 2-Helix protein construct consists of HR2 and part of HR1, which is the major region binding HR2, connected by a 22-amino acid linker (LVPRGSGGSGGSGGLEVLFQGP) as indicated. B, sequence alignment of coronavirus spike protein HR1 and HR2 regions. Letters above the sequence indicate the predicted hydrophobic residues at the a and d positions in two heptad repeat regions, which are highly conserved. FP, feline panleukopenia; IBV, infectious bronchitis virus; FIPV, feline infectious peritonitis virus. case of HIV-1 gp41, peptides derived from C peptides can effectively inhibit infection in a dominant-negative manner by binding to the transiently exposed hydrophobic grooves of central coiled coil in the intermediate state and consequently blocking the formation of the fusion-active six-helix bundle structure (30). We have recently determined the three-dimensional structure of murine coronavirus MHV spike protein fusion core and proposed a model for coronavirus-mediated membrane fusion in which the S protein undergoes a series of conformational changes similar to those of influenza virus and HIV-1 (14). Recent studies show that HR1 and HR2 of SARS-CoV spike protein form a stable six-helix bundle, and synthesized peptides corresponding to the HR2 region have inhibitory activity for viral fusion (31 36). 2 These studies also propose the binding regions between HR1 and HR2 and identify the inhibition efficiency of peptides derived from the HR2 region. However, the detailed three-dimensional structure of the HR1/HR2 complex remains unknown. To verify that the SARS-CoV spike protein indeed forms a trimer-of-hairpins structure and to provide a structural basis for the design of viral fusion inhibitors, we characterized the binding of the two HR regions of the SARS-CoV spike protein and solved the crystal structure of this fusion core complex to 2.8 Å of resolution. The structure shows a similar conformation to other class I viral fusion proteins, especially MHV spike protein fusion 2 Y. Xu, J. Zhu, Y. Liu, Z. Lou, F. Yuan, Y. Liu, D. K. Cole, L. Ni, N. Su, L. Qin, X. Li, Z. Bai, J. I. Bell, H. Pang, P. Tien, Z. Rao, and G. F. Gao, submitted for publication. 160 core, suggesting that a similar approach might be used in identifying inhibitors of SARS-CoV infection effectively. This structure also provides important detailed structural information and target site for structure based drug design. EXPERIMENTAL PROCEDURES Construction, Expression, and Purification The SARS spike gene was cloned from SARS coronavirus GZ02 (GenBank TM accession number AY390556). Two peptides (HR1 and HR2 regions of SARS-CoV spike protein) interact, and their binding regions were characterized by recent biochemical studies and sequence alignment with MHV fusion core complex. 2 The HR1 and HR2 regions of SARS-CoV spike protein consist of residues and residues , respectively, corresponding to the HR1 and HR2 regions in new MHV structure (14). The fusion core of SARS-CoV spike protein was prepared as a single chain (termed SARS 2-Helix) by linking the HR1 and HR2 domains via a 22-amino acid linker (LVPRGSGGSGGSGGLEVLFQGP), which is flexible and long enough to facilitate a natural interaction between the HR1 and HR2 peptides and allows for easy expression and purification of the fusion core complex (Fig. 1A). The PCR-directed gene was inserted into pet22b (Novagen) vector, and the SARS 2-Helix protein was expressed in LB culture medium in Escherichia coli strain BL21(DE3). The product was purified by nickel nitrilotriacetic acid affinity chromatography and was further purified by gel filtration chromatography (Superdex 75, Amersham Biosciences). Crystallization and Structure Determination The purified protein was dialyzed against 10 mm Tris-HCl, ph 8.0, 10 mm NaCl and then concentrated to 15 mg ml 1. Crystals with good diffracting quality could be obtained after 2 weeks by using the hanging drop method by equilibrating a 2- l drop (protein solution mixed 1:1 with reservoir solution) against a reservoir containing 0.1 M citric acid, ph 2.5, 10 15% polyethylene glycol 4000, 0.02 M spermidine tetra-hcl. The SARS 2-Helix crystal was mounted on nylon loops and flash-frozen in cold nitrogengas stream at 100 K using an Oxford Cryosystems coldstream with 0.1

161 49416 Crystal Structure of SARS Spike Protein Fusion Core TABLE I Data collection and final refinement statistics Numbers in parentheses correspond to the highest resolution shell. R merge h I Iih Ih / h I Ih, where Ih is the mean of the observations Iih of reflection h. R work ( F obs F calc )/ F obs ; R free is the R factor for a subset (5%) of reflections that was selected before refinement calculations and not included in the refinement. r.m.s.d., root mean square deviation from ideal geometry. Data collection statistics Space group C2 Unit cell parameters a Å, b 66.3 Å, c 70.0 Å, Wavelength (Å) Resolution limit (Å) 2.8 Observed reflections 27,312 Unique reflections 11,974 Completeness (%) 91.4 (80.8) I/ (I) 5.7 (1.2) R merge (%) 13.9 (51.1) Final refinement statistics R work (%) 23.3 R free (%) 27.3 Resolution range (Å) Total reflections used 10,718 Number of reflections in 10,166 working set Number of reflections in 552 test set Average B (Å 2 ) 42.7 r.m.s.d. bonds (Å) r.m.s.d. angles( ) 1.8 M citric acid, ph 2.5, 25% polyethylene glycol 4000 as the cryoprotectant. The crystals have unit-cell parameters a Å, b 66.3 Å, c 70.0 Å, 90, and belong to space group C2. The crystals contain 6 SARS 2-Helix molecules in one asymmetric unit, and the diffraction pattern extends to 2.8 Å of resolution. Data collection was performed in-house on a Rigaku RU2000 rotating copper-anode x-ray generator operated at 48 kv and 98 ma (Cu K ; Å) with a MAR 345 image-plate detector. Data were indexed, integrated, and scaled using DENZO and SCALEPACK programs (38). The structure of SARS 2-Helix was determined by molecular replacement with the MHV 2-Helix structure (PDB code 1WDF) as a search model. Rotation and translation function searches were performed in the program CNS (39). The model was improved further by cycles of manual building and refinement using the programs O (40) and CNS (39). The quality of coordinates was examined by PROCHECK (41). The figures were generated with the programs GRASP (42), SPDBView (43), and MOLSCRIPT (44). RESULTS AND DISCUSSION Structure Determination The crystal structure of SARS 2-Helix was solved by molecular replacement using the program CNS (39) with the MHV 2-Helix structure (PDB code 1WDF) employed as a search model. Six molecules (two trimers of SARS 2-Helix) per asymmetric unit were located from crossrotation and translation function searches. The model was improved further by cycles of manual building and refinement using the programs O (40) and CNS (39). The structure was subsequently refined to a final resolution of 2.8 Å with an R value of 23.3% and R free value of 27.3%. No residue was in disallowed regions of the Ramachandran plot. The statistics for the data collection, structure determination, and refinement are summarized in Table I. Description of the Structure One asymmetric unit contains six SARS 2-Helix molecules, including residues in HR1 and in HR2 (A molecule), residues in HR1 and in HR2 (B molecule), residues in HR1 and in HR2 (C molecule), residues in HR1 and in HR2 (D molecule), residues in HR1 and in HR2 (E molecule), and residues in HR1 and in HR2 (F molecule), respectively. The linker and several terminal residues could not be traced in the electron density map due to their disordered nature. 161 FIG. 2. Overall views of the fusion core structure. A, top view of the SARS-CoV spike protein fusion core structure showing the 3-fold axis of the trimer. B, side view of the SARS-CoV spike protein fusion core structure showing the six-helix bundle. Here, we chose A, B, and C molecules as the SARS spike protein fusion core for the following structure description. In the SARS 2-Helix three-dimensional structure, the fusion core has a rod-shaped structure with a length of 70 Å and a diameter of 28 Å. Similar to MHV 2-Helix and other class I viral fusion proteins, the SARS spike protein fusion core is a six-helix bundle comprising a trimer of 2-Helix molecules. The center of the fusion core consists of a parallel trimeric coiled coil of three HR1 helices surrounded by three HR2 helices in an oblique, antiparallel manner (Fig. 2, A and B). Residues in HR1 fold into a 12-turn -helix stretching the entire length of the fusion core. As in MHV 2-Helix and other class I viral fusion proteins, the residues in positions a and d of HR1 are predominantly hydrophobic (Fig. 1B). Residues in HR2 form a 5-turn -helix, whereas residues at the N terminus and residues at the C terminus of HR2 form two extended conformations, respectively. Three HR2 helices pack against the grooves formed by the interface of the central three HR1 helices, and no interaction was observed between individual HR2 regions. The N terminus of HR2 starts with Ile 1150, which is aligned with Gln 947 of HR1. The C ter-

162 Crystal Structure of SARS Spike Protein Fusion Core FIG. 3. Detailed structure of the SARS-CoV spike protein fusion core and OXO motifs in HR2 regions. A, surface map showing the hydrophobic grooves on the surface of the central coiled coil (right side). Three HR2 helices pack against the hydrophobic grooves in an oblique antiparallel manner (left side). The helical regions and extended regions in HR2 helices could be observed clearly, and the boundaries of these regions are marked. B, OXO motifs in HR2 regions of SARS-CoV spike protein fusion core structure. The enlarged images show two regions containing OXO motifs. The hydrophobic residues in these motifs all pack against the hydrophobic grooves on the surface of three HR1 helices. 162 minus of HR2 ends with Leu 1184, which is aligned with Gln 902 of HR1 (see Fig. 4). Interactions between HR1 and HR2 Three HR2 helices interact with HR1 helices mainly through hydrophobic interaction between hydrophobic residues in HR2 regions and the grooves on the surface of the central coiled coil. Similar to those in MHV 2-Helix, HR2 helices in SARS 2-Helix also contain OXO motifs, in which O represents a hydrophobic residue, and X represents any residue but is generally hydrophobic IN- ASVVNI 1160 and 1178 LIDL 1181 in HR2 regions are both composed of OXO motifs; the side chains of the O residues inset into or align with the hydrophobic grooves of the central coiled coil, whereas the side chains of the X residues are directed into solvent (Fig. 3B). As described in our previous paper (14), the OXO motifs are responsible for the partially extended conformation of HR2, and this pattern also makes the fusion core stable in solvent, as most of the hydrophobic residues in HR2 helices are packed against the central coiled coil, leaving the hydrophilic residues exposed to solvent. Comparison with MHV 2-Helix Structure Among coronavirus spike proteins, only the structure of the MHV spike protein fusion core has been determined (14). In general, the fusion core from SARS-CoV and MHV adopt a similar fold, consistent with their high sequence identity and similarity between the two proteins. The structure of SARS spike protein fusion core was compared with that of MHV spike protein (Fig. 4). These structures can be superimposed with a root mean square difference of 0.91 Å for all C atoms. Alignments of the peptides derived from HR1 and HR2 regions of SARS-CoV spike protein and those of other coronaviruses reveal high sequence identity and similarity, suggesting that structures of spike protein fusion cores from other coronavirus might share significant similarity with those of MHV and SARS-CoV (Fig. 1B). Although the main chains of the MHV 2-Helix and SARS- CoV 2-Helix structures can be superposed closely, the two complexes have significant differences in their hydrophobic grooves on the surface of the central coiled coil (Fig. 4B). As discussed in our previous paper detailing our SARS-CoV fusion core model, the central coiled coil had relatively deep grooves and relatively shallow grooves. 2 The deep grooves, consisting of three hydrophobic deep pockets or cavities, were clearly deeper than the corresponding grooves on the surface of MHV 2-Helix control coiled coil. From the structure presented here, the helical regions of SARS-CoV HR2 segment (residues ) pack exactly against the relatively deep grooves of the central coiled coil and the extended regions (residues ) pack against the relatively shallow grooves. Based on our previous biochemical analysis and the crystal structure, we propose that the deep groove is an important target site for the design of viral fusion inhibitors (16, 30, 34, 45 48). Conformational Change and Membrane Fusion Mechanisms Our previous structural study of the MHV spike protein fusion core led us to propose a model for coronavirusmediated membrane fusion mechanism (14). The remarkable similarity between SARS-CoV and MHV spike protein fusion core structures as well as similar HR2 peptide inhibition phenomena and remarkable stability to both thermal denaturation and proteinase K digestion (31 36) suggest a conserved mechanism of membrane fusion mediated by the spike protein. Similar to the MHV spike protein and HIV gp41, the SARS-CoV spike protein likely undergoes a series of conformational changes to become fusion-active. The fusion loop but not fusion peptide, which will insert into the cellular membrane, and distinct conformational states proposed for the MHV spike protein fusion core, including the native state, the pre-hairpin intermediate, and fusion-active hairpin state, may also apply to SARS-CoV spike protein. The existence of the pre-hairpin intermediate conformation of the SARS-CoV spike protein is strongly supported by the viral inhibition assay, in which the peptides corresponding to the HR2 regions can inhibit viral fusion in a dominant-negative manner (32, 34). A reasonable interpretation of these phenomena is that the HR2 peptide functions by binding to the transiently exposed hydrophobic groove on the surface of central coiled coil, thus blocking the conformational transition to the fusion active form and subsequent membrane fusion and viral entry. Binding Regions of SARS-CoV Spike Protein Fusion Core Recent studies on the fusion-active complex of SARS-CoV have confirmed that HR1 and HR2 associate into an antiparallel six-helix bundle, with structural features of other typical class I viral fusion proteins by means of CD, native PAGE, proteolysis protection analysis, and size-exclusion chromatography. In their biochemical analysis, Ingallinella et al. (31) mapped the specific boundaries of the key region of interaction between HR1 and HR2 peptides as residues 914 and 949 in HR1 region and residues 1148 and 1185 in HR2 region (31). The exact

163 49418 Crystal Structure of SARS Spike Protein Fusion Core FIG. 4. Comparison between fusion core structure of SARS-CoV and MHV. A, side view showing a structural comparison between SARS-CoV spike protein fusion core (colored in green) and MHV spike protein fusion core (colored in purple). The columns at both sides of the map represent two HR1 and HR2 regions of MHV and SARS fusion cores. The numbers at the end of these columns represent the specific boundaries of the HR1-HR2 interaction region in the two structures. B, surface map showing the comparison between hydrophobic grooves on the surface of three central HR1 regions of MHV (left side) and SARS-CoV (right side). The figure on the right side shows the deep and relatively shallow grooves on the surface of central HR1 coiled coil of SARS-CoV. Three numbers, 1 3, in circles represent three deep cavities, composing the deep grooves. The residues represent the boundaries of the grooves and cavities. TABLE II Amino acid sequences and EC 50 values of inhibitory peptides Peptide EC 50 Sequence M NP-1 ( ) a Marginal GVTQNVLYENQKQIANQFNKAISQIQESLTTTSTALGKLQ CP-1 ( ) a 19 GINASVVNTQKEIDRLNEVAKNLNESLIDLQELGKYE HR1 ( ) b 3.68 NGIGVTQNVLYENQKQIANQFNKAISQIQESLTTTSTA HR2 ( ) b 5.22 IQKEIDRLNEVAKNLNESLIDLQELGK HR2-1 ( ) c ELDSPKEELDKYFKNHTSPDVDLGDISGINASVVNIQKEIDRLNEVAKNLNESLIDLQELGKYE HR2-2 ( ) c PKHELDKYFKNHTSPDVDLGDISGINASVVNIQKEIDRLNEVAKNLNESLIDLQELGKYE HR2-8 ( ) c ELDSFKEELDKYFKNHTSPDVDLGDISGINASVVNIQKEIDRLNEVAKNLNESLIDLQELGKYEQYIK HR2-9 ( ) c ELDSFKEELDKYFKNHTSPDVDLGDISGINASVVNIQKEIDRLNEVAKNLNESLIDLQEL HR2-6 ( ) c 50 DISGINASVVNIQKEIDRLNEVAKNLNESLIDLQELGKYE a Data are from the paper by Liu et al. (35). b Data are from the paper by Yuan et al. (33). c Data are from the paper by Bosch et al. (32). boundaries, residues 902 and 947 in HR1 region and residues 1150 and 1184 in HR2 region, can be observed from the crystal structure of SARS-CoV spike protein fusion core with a slight difference from previous results. The N terminus of the HR1 region (Gln 902 ) aligns with the C terminus of the HR2 region (Leu 1184 ), and the C terminus of the HR1 region (Gln 947 ) aligns with the N terminus of the HR2 region (Ile 1150 ). Although the residue numbers of the HR1 and HR2 regions are different (46 in HR1 and 35 in HR2), the two peptides are equivalent in length in the three-dimensional structure since HR1 forms a typical -helix, whereas HR2 forms a partial helical conformation (Fig. 3A). This pattern of HR2 helices is also strongly supported by proteolysis protection experiments (31). The real boundaries of the fusion core might extend beyond those found in the structure studied here, whereas the major binding regions and the interactions between HR1 and HR2 peptides can be identified clearly from the SARS-CoV spike protein fusion core structure. Inhibitory Molecules for SARS-CoV Infection If small, bioavailable molecules that prevent hairpin formation can be 163 identified, they may serve as useful drugs against SARS-CoV infection. In the case of HIV-1, several strategies to block hairpin formation have been successfully developed to identify viral entry inhibitors that bind to the hydrophobic pocket and grooves on the surface of the central coiled coil consisting of HIV-1 gp41 N peptides. These useful viral entry inhibitors include D peptides, 5-Helix, and synthetic peptides derived from N or C peptides (45 47). Successful viral entry inhibitors have also been identified for other viruses, such as T20 for HIV-1 (49, 50) and GP610 for Ebola virus (51). These strategies could also be used for the design of SARS fusion inhibitors. The well defined hydrophobic grooves on the surface of the central coiled coil of the SARS-CoV spike protein fusion core identified here may be a significant target for drug design. Several peptides derived from the HR1 and HR2 regions of SARS-CoV spike proteins have been found to have inhibitory activity in recent studies (32, 33, 35) (Table II). Analogous to the HIV-1 C peptides and MHV HR2 peptides, the HR2 peptides of SARS-CoV spike protein likely function in a dominantnegative manner by binding to the transiently exposed hydro-

164 Crystal Structure of SARS Spike Protein Fusion Core phobic grooves in the pre-hairpin intermediate, thus, blocking viral entry. The efficacy of HR2 peptides of SARS-CoV spike protein is, however, significantly lower than corresponding HR2 peptides of murine coronavirus mouse hepatitis virus in inhibiting MHV infection (32). Synthetic peptides with the highest inhibitory efficacy encompass residues , derived from the HR2 region (33). This peptide is just the corresponding region that binds to the relatively deep grooves on the surface of central coiled coil. It is not clear why HR2 peptides of SARS-CoV have lower inhibitory efficacy. However, the structural information provided here will be useful for the design of antiviral compounds such as D peptides, 5-Helix, and some peptides (or mutants) derived from HR1 or HR2 peptides based on the crystal structure of SARS-CoV spike protein fusion core. The peptides encompassing residues will be good targets for mutagenesis in the search for peptides with higher inhibitory efficacy. The exposed hydrophobic grooves in the intermediate state and, particularly, the relatively deep grooves on the surface of central coiled coil will be good targets for the discovery of viral entry inhibitors (Fig. 3A and 4B).. Acknowledgment We thank Dr. Mark Bartlam for comments and critical reading. REFERENCES 1. Tsang, K. W., Mok, T. Y., Wong, P. C., and Ooi, G. C. (2003) Respirology 8, Poutanen, S. M., Low, D. E., Henry, B., Finkelstein, S., Rose, D., Green, K., Tellier, R., Draker, R., Adachi, D., Ayers, M., Chan, A. K., Skowronski, D. M., Salit, I., Simor, A. E., Slutsky, A. S., Doyle, P. W., Krajden, M., Petric, M., Brunham, R. C., and McGeer, A. J. (2003) N. Engl. J. Med. 348, Drosten, C., Gunther, S., Preiser, W., van der Werf, S., Brodt, H. R., Becker, S., Rabenau, H., Panning, M., Kolesnikova, L., Fouchier, R. A., Berger, A., Burguiere, A. M., Cinatl, J., Eickmann, M., Escriou, N., Grywna, K., Kramme, S., Manuguerra, J. C., Muller, S., Rickerts, V., Sturmer, M., Vieth, S., Klenk, H. D., Osterhaus, A. D., Schmitz, H., and Doerr, H. W. (2003) N. Engl. J. Med. 348, Eickmann, M., Becker, S., Klenk, H. D., Doerr, H. W., Stadler, K., Censini, S., Guidotti, S., Masignani, V., Scarselli, M., Mora, M., Donati, C., Han, J. H., Song, H. C., Abrignani, S., Covacci, A., and Rappuoli, R. (2003) Science 302, Stadler, K., Masignani, V., Eickmann, M., Becker, S., Abrignani, S., Klenk, H. D., and Rappuoli, R. (2003) Nat. Rev. Microbiol. 1, Ksiazek, T. G., Erdman, D., Goldsmith, C. S., Zaki, S. R., Peret, T., Emery, S., Tong, S., Urbani, C., Comer, J. A., Lim, W., Rollin, P. E., Dowell, S. F., Ling, A. E., Humphrey, C. D., Shieh, W. J., Guarner, J., Paddock, C. D., Rota, P., Fields, B., DeRisi, J., Yang, J. Y., Cox, N., Hughes, J. M., LeDuc, J. W., Bellini, W. J., and Anderson, L. J. (2003) N. Engl. J. Med. 348, Kuiken, T., Fouchier, R. A., Schutten, M., Rimmelzwaan, G. F., van Amerongen, G., van Riel, D., Laman, J. D., de Jong, T., van Doornum, G., Lim, W., Ling, A. E., Chan, P. K., Tam, J. S., Zambon, M. C., Gopal, R., Drosten, C., van der Werf, S., Escriou, N., Manuguerra, J. C., Stohr, K., Peiris, J. S., and Osterhaus, A. D. (2003) Lancet 362, Peiris, J. S., Lai, S. T., Poon, L. L., Guan, Y., Yam, L. Y., Lim, W., Nicholls, J., Yee, W. K., Yan, W. W., Cheung, M. T., Cheng, V. C., Chan, K. H., Tsang, D. N., Yung, R. W., Ng, T. K., and Yuen, K. Y. (2003) Lancet 361, Marra, M. A., Jones, S. J., Astell, C. R., Holt, R. A., Brooks-Wilson, A., Butterfield, Y. S., Khattra, J., Asano, J. K., Barber, S. A., Chan, S. Y., Cloutier, A., Coughlin, S. M., Freeman, D., Girn, N., Griffith, O. L., Leach, S. R., Mayo, M., McDonald, H., Montgomery, S. B., Pandoh, P. K., Petrescu, A. S., Robertson, A. G., Schein, J. E., Siddiqui, A., Smailus, D. E., Stott, J. M., Yang, G. S., Plummer, F., Andonov, A., Artsob, H., Bastien, N., Bernard, K., Booth, T. F., Bowness, D., Czub, M., Drebot, M., Fernando, L., Flick, R., Garbutt, M., Gray, M., Grolla, A., Jones, S., Feldmann, H., Meyers, A., Kabani, A., Li, Y., Normand, S., Stroher, U., Tipples, G. A., Tyler, S., Vogrig, R., Ward, D., Watson, B., Brunham, R. C., Krajden, M., Petric, M., Skowronski, D. M., Upton, C., and Roper, R. L. (2003) Science 300, Rota, P. A., Oberste, M. S., Monroe, S. S., Nix, W. A., Campagnoli, R., Icenogle, J. P., Penaranda, S., Bankamp, B., Maher, K., Chen, M. H., Tong, S., Tamin, A., Lowe, L., Frace, M., DeRisi, J. L., Chen, Q., Wang, D., Erdman, D. D., Peret, T. C., Burns, C., Ksiazek, T. G., Rollin, P. E., Sanchez, A., Liffick, S., Holloway, B., Limor, J., McCaustland, K., Olsen-Rasmussen, M., Fouchier, R., Gunther, S., Osterhaus, A. D., Drosten, C., Pallansch, M. A., Anderson, L. J., and Bellini, W. J. (2003) Science 300, Siddell, S., Wege, H., and Ter Meulen, V. (1983) J. Gen. Virol. 64, Cavanagh, D. (1983) J. Gen. Virol. 64, Gallagher, T. M., and Buchmeier, M. J. (2001) Virology 279, Xu, Y., Liu, Y., Lou, Z., Qin, L., Li, X., Bai, Z., Pang, H., Tien, P., Gao, G. F., and Rao, Z. (2004) J. Biol. Chem. 279, Bosch, B. J., van der Zee, R., de Haan, C. A., and Rottier, P. J. (2003) J. Virol. 77, Eckert, D. M., and Kim, P. S. (2001) Annu. Rev. Biochem. 70, Hernandez, L. D., Hoffman, L. R., Wolfsberg, T. G., and White, J. M. (1996) Annu. Rev. Cell Dev. Biol. 12, Bullough, P. A., Hughson, F. M., Skehel, J. J., and Wiley, D. C. (1994) Nature 371, Lu, M., Blacklow, S. C., and Kim, P. S. (1995) Nat. Struct. Biol. 2, Weissenhorn, W., Dessen, A., Harrison, S. C., Skehel, J. J., and Wiley, D. C. (1997) Nature 387, Chan, D. C., Fass, D., Berger, J. M., and Kim, P. S. (1997) Cell 89, Tan, K., Liu, J., Wang, J., Shen, S., and Lu, M. (1997) Proc. Natl. Acad. Sci. U. S. A. 94, Caffrey, M., Cai, M., Kaufman, J., Stahl, S. J., Wingfield, P. T., Covell, D. G., Gronenborn, A. M., and Clore, G. M. (1998) EMBO J. 17, Weissenhorn, W., Carfi, A., Lee, K. H., Skehel, J. J., and Wiley, D. C. (1998) Mol. Cell 2, Weissenhorn, W., Calder, L. J., Wharton, S. A., Skehel, J. J., and Wiley, D. C. (1998) Proc. Natl. Acad. Sci. U. S. A. 95, Baker, K. A., Dutch, R. E., Lamb, R. A., and Jardetzky, T. S. (1999) Mol. Cell 3, Delmas, B., and Laude, H. (1990) J. Virol. 64, Bonavia, A., Zelus, B. D., Wentworth, D. E., Talbot, P. J., and Holmes, K. V. (2003) J. Virol. 77, Taguchi, F. (1995) J. Virol. 69, Chan, D. C., and Kim, P. S. (1998) Cell 93, Ingallinella, P., Bianchi, E., Finotto, M., Cantoni, G., Eckert, D. M., Supekar, V. M., Bruckmann, C., Carfi, A., and Pessi, A. (2004) Proc. Natl. Acad. Sci. U. S. A. 101, Bosch, B. J., Martina, B. E., Van Der Zee, R., Lepault, J., Haijema, B. J., Versluis, C., Heck, A. J., De Groot, R., Osterhaus, A. D., and Rottier, P. J. (2004) Proc. Natl. Acad. Sci. U. S. A. 101, Yuan, K., Yi, L., Chen, J., Qu, X., Qing, T., Rao, X., Jiang, P., Hu, J., Xiong, Z., Nie, Y., Shi, X., Wang, W., Ling, C., Yin, X., Fan, K., Lai, L., Ding, M., and Deng, H. (2004) Biochem. Biophys. Res. Commun. 319, Zhu, J., Xiao, G., Xu, Y., Yuan, F., Zheng, C., Liu, Y., Yan, H., Cole, D. K., Bell, J. I., Rao, Z., Tien, P., and Gao, G. F. (2004) Biochem. Biophys. Res. Commun. 319, Liu, S., Xiao, G., Chen, Y., He, Y., Niu, J., Escalante, C. R., Xiong, H., Farmar, J., Debnath, A. K., Tien, P., and Jiang, S. (2004) Lancet 363, Tripet, B., Howard, M. W., Jobling, M., Holmes, R. K., Holmes, K. V., and Hodges, R. S. (2004) J. Biol. Chem. 279, Singh, M., Berger, B., and Kim, P. S. (1999) J. Mol. Biol. 290, Otwinowski, Z., and Minor, W. (1997) Methods Enzymol. 276, Brunger, A. T., Adams, P. D., Clore, G. M., DeLano, W. L., Gros, P., Grosse- Kunstleve, R. W., Jiang, J. S., Kuszewski, J., Nilges, M., Pannu, N. S., Read, R. J., Rice, L. M., Simonson, T., and Warren, G. L. (1998) Acta Crystallogr. D Biol. Crystallogr. 54, Jones, T. A., Zou, J. Y., Cowan, S. W., and Kjeldgaard, M. (1991) Acta Crystallogr. A. 47, Laskowski, R. A., MacArthur, M. W., Moss, D. S., and Thornton, J. M. (1993) J. Appl. Crystallogr. 26, Nicholls, A., Sharp, K. A., and Honig, B. (1991) Proteins 11, Guex, N., and Peitsch, M. C. (1997) Electrophoresis 18, Kraulis, P. J. (1991) J. Appl. Crystallogr. 24, Eckert, D. M., and Kim, P. S. (2001) Proc. Natl. Acad. Sci. U. S. A. 98, Root, M. J., Kay, M. S., and Kim, P. S. (2001) Science 291, Eckert, D. M., Malashkevich, V. N., Hong, L. H., Carr, P. A., and Kim, P. S. (1999) Cell 99, Wild, C. T., Shugars, D. C., Greenwell, T. K., McDanal, C. B., and Matthews, T. J. (1994) Proc. Natl. Acad. Sci. U. S. A. 91, Kilby, J. M., Hopkins, S., Venetta, T. M., DiMassimo, B., Cloud, G. A., Lee, J. Y., Alldredge, L., Hunter, E., Lambert, D., Bolognesi, D., Matthews, T., Johnson, M. R., Nowak, M. A., Shaw, G. M., and Saag, M. S. (1998) Nat. Med. 4, Imai, M., Okada, N., and Okada, H. (2000) Microbiol. Immunol. 44, Watanabe, S., Takada, A., Watanabe, T., Ito, H., Kida, H., and Kawaoka, Y. (2000) J. Virol. 74,

165 THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 279, No. 48, Issue of November 26, pp , by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A. The Yeast Prion Protein Ure2 Shows Glutathione Peroxidase Activity in Both Native and Fibrillar Forms* Received for publication, June 14, 2004, and in revised form, September 10, 2004 Published, JBC Papers in Press, September 15, 2004, DOI /jbc.M Ming Bai, Jun-Mei Zhou, and Sarah Perrett From the National Laboratory of Biomacromolecules, Institute of Biophysics, Chinese Academy of Sciences, 15 Datun Road, Chaoyang District, Beijing , China and Graduate School of the Chinese Academy of Sciences, 19 Yuquan Road, Fengtai District, Beijing , China Ure2p is the precursor protein of the Saccharomyces cerevisiae prion [URE3]. Ure2p shows homology to glutathione transferases but lacks typical glutathione transferase activity. A recent study found that deletion of the Ure2 gene causes increased sensitivity to heavy metal ions and oxidants, whereas prion strains show normal sensitivity. To demonstrate that protection against oxidant toxicity is an inherent property of native and prion Ure2p requires biochemical characterization of the purified protein. Here we use steady-state kinetic methods to characterize the multisubstrate peroxidase activity of Ure2p using GSH with cumene hydroperoxide, hydrogen peroxide, or tert-butyl hydroperoxide as substrates. Glutathione-dependent peroxidase activity was proportional to the Ure2p concentration and showed optima at ph 8 and 40 C. Michaelis-Menten behavior with convergent straight lines in double reciprocal plots was observed. This excludes a ping-pong mechanism and implies either a rapid-equilibrium random or a steady-state ordered sequential mechanism for Ure2p, consistent with its classification as a glutathione transferase. The mutant 90Ure2, which lacks the unstructured N-terminal prion domain, showed kinetic parameters identical to wild type. Fibrillar aggregates showed the same level of activity as native protein. Demonstration of peroxidase activity for Ure2 represents important progress in elucidation of its role in vivo. Further, establishment of an in vitro activity assay provides a valuable tool for the study of structure-function relationships of the Ure2 protein as both a prion and an enzyme. The glutathione S-transferases (GSTs) 1 are a multifunctional family of enzymes broadly distributed in nature that play a critical role in the cellular detoxification process (reviewed in Refs. 1 6). GSTs have the general function of conjugating GSH to electrophilic substances to reduce their toxicity. As a consequence, GSTs are involved in development of resistance toward drugs, insecticides, and herbicides and have a * This work was supported by the Natural Science Foundation of China ( , ), the 973 Project of the Chinese Ministry of Science and Technology (G ), and the Chinese Academy of Sciences Knowledge Innovation Project (KSCX2-SW214-3). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Recipient of a special grant from the Chinese Academy of Sciences and research fellowships from the Royal Commission for the Exhibition of 1851 and the Royal Society. To whom correspondence may be addressed. Tel.: ; Fax: ; sarah. perrett@iname.com or zhoujm@sun5.ibp.ac.cn. 1 The abbreviations used are: GST, glutathione S-transferase; CDNB, 1-chloro-2,4-dinitrobenzene; GPx, glutathione peroxidase; CHP, cumene hydroperoxide; t-bh, tert-butyl hydroperoxide; ThT, thioflavin T. This paper is available on line at protective role against a range of diseases including cancer. GSTs are dimeric proteins with a relatively conserved N-terminal thioredoxin-like domain and a more variable C-domain. GST activity is typically tested using the universal GST substrate, 1-chloro-2,4-dinitrobenzene (CDNB). However, a number of GSTs identified by structural criteria have failed to show activity toward CDNB (5). Some GSTs have been shown to have overlapping functions with other glutathione-binding enzymes such as glutathione peroxidases (GPxs) and glutaredoxins (1 3). These enzyme families share the GSH-binding thioredoxin domain but are otherwise structurally and mechanistically dissimilar (7 10). Characterization of the GSTs, glutaredoxins, and phospholipid hydroperoxide glutathione peroxidases present in Saccharomyces cerevisiae reveals some redundancy of function between the different classes of enzyme (11 13). The versatility of the glutathione-binding enzymes and their tendency to show overlapping functions may contribute significantly to the ability of the host organism to adapt to change. Ure2p is the protein determinant of the S. cerevisiae prion [URE3] (14). Analogous to the mammalian prion (15), the heritable [URE3] prion phenotype is conveyed by a structural change in Ure2 to an aggregated form (16). It has recently been demonstrated that drugs isolated using a yeast prion cellscreening assay are also active against mammalian prions, suggesting that there may be features of the cellular mechanism of prion formation and/or maintenance that are common to yeast and mammalian prions despite the diversity of the proteins involved (17). Conversion of Ure2 to the prion form depends on the N-terminal 90 amino acids (18). This N- terminal prion domain also directs the formation of amyloidlike fibrils in vitro (19, 20) and is predominantly unstructured in the native dimeric state (21, 22). Ure2 is involved in the regulation of nitrogen metabolism in vivo. This function is carried out by the C-terminal region of the protein and is lost on conversion to the prion form (16, 23). The crystal structure of the C-terminal region has been solved in both apo (24, 25) and glutathione-bound (26) forms, confirming the classification of Ure2 as a glutathione transferase (6, 23). However, attempts to demonstrate typical GST activity for Ure2, such as using CDNB, have so far proved unsuccessful (11, 21, 23, 27). A Ure2 homologue in Aspergillus nidulans was found to lack the nitrogen metabolite repression activity of Ure2 but contributed to heavy metal and xenobiotic resistance, including resistance to oxidative stress (28). Further, it was recently found that the deletion of the S. cerevisiae Ure2 gene increases the sensitivity of the cell to metals and cellular oxidants such as hydrogen peroxide (27, 29), raising the possibility that Ure2 possesses GPx activity. This has implications for the role of Ure2 in vivo. In addition, it implies the potential to establish an in vitro activity assay for Ure2, which would serve as an invaluable tool in structure-function analysis.

166 50026 The Yeast Prion Ure2p Has Glutathione Peroxidase Activity In this study, we tested the activity of Ure2 toward a number of oxidant substrates in vitro to establish whether the peroxidase activity related to the Ure2 gene is an inherent property of the Ure2 protein. Having detected GPx activity for Ure2, we employed steady-state kinetic techniques to characterize the kinetic parameters for the substrates cumene hydroperoxide (CHP), tert-butyl hydroperoxide (t-bh), and H 2 O 2 and to investigate the reaction mechanism. Further, we compared the activity of wild type Ure2 with that of a Ure2 deletion mutant, 90Ure2, which lacks the prion domain. Finally, we compared the activity of the native dimeric protein with that of fibrillar aggregates. EXPERIMENTAL PROCEDURES Materials GSH, -NADPH, CHP, t-bh, and glutathione reductase were from Sigma. Ure2 and 90Ure2 were produced in Escherichia coli with an N-terminal His 6 tag and purified by nickel affinity chromatography or produced without a tag and purified by a series of ionic exchange and gel filtration chromatographic steps, as described previously (21). Proteins were stored at 80 C in 50 mm Tris-HCl buffer, ph 8.4, containing 0.2 M NaCl and defrosted in a 25 C water bath immediately prior to use. The protein concentration in terms of monomers was measured by absorbance at 280 nm using an extinction coefficient of 40,700 M 1 cm 1 (21) unless otherwise stated. Enzyme Assays and Steady-state Kinetic Analysis The GPx activity of Ure2 was determined using GSH with one of the hydroperoxides, CHP, H 2 O 2,ort-BH, as substrates using a coupled spectrophotometric assay as described previously (30) with slight modifications. Unless otherwise indicated, the assay was carried out at 25 C in a 1-ml reaction volume containing 100 mm sodium phosphate buffer, ph 7.5, 4 mm sodium azide, mm GSH, 0.15 mm -NADPH, 0.24 units of glutathione reductase, and M Ure2. The reaction mixture was preincubated at 25 C for 6 min, after which the reaction was started by the addition of the hydroperoxide substrate to a final concentration of mm to both cuvettes. The progress of reactions was monitored continuously by following the decrease in NADPH absorbance at 340 nm on a Shimadzu UV2501PC14 spectrophotometer. Initial rates were determined from the linear slope of progress curves obtained with an extinction coefficient for NADPH of 6220 M 1 cm 1 after subtracting the non-enzymatic velocities due to the auto-oxidation of GSH by the hydroperoxide determined from the corresponding blank. When bovine serum albumin was used in place of Ure2, no increase over the base-line rate was observed. The presence or absence of 4 mm sodium azide had no affect on the Ure2 activity. When GSH was omitted from the reaction mixture, no Ure2 activity was observed. Steady-state kinetic analysis was carried out by obtaining sets of initial velocities over a wide range of concentrations of one substrate while the concentration of the other substrate was kept constant (31, 32). The data were fitted to the Michaelis-Menten equation or the Lineweaver-Burk equation. The values obtained from these plots, and Eadie-Hofstee plots were the same within error. Single or global fitting was carried out using the regression wizard of SigmaPlot. The errors shown are the S.E. of the fit, or the mean S.E. obtained from independent measurements, as appropriate. Determination of true kinetic parameters and investigation of the reaction mechanism were performed by obtaining Michaelis-Menten curves at a series of concentrations of the second substrate and then fitting the data globally to the Michaelis-Menten model describing two-substrate sequential binding (32 34). The values obtained by linear extrapolation using secondary plots or by global fitting were the same within error. Assay of Ure2 GPx Activity during the Time Course of Amyloid-like Fibril Formation The initial sample was centrifuged at 18,000 g for 30 min at 4 C to remove any preexisting aggregates, and 300 l ofthe supernatant was transferred into each of a series of tubes, one for each time point. The reaction mixture contained 30 M full-length Ure2 in 50 mm sodium phosphate buffer, ph 7.5, containing 0.2 M NaCl. The samples were incubated in parallel at a constant temperature of 37 C with shaking as described previously (35, 36). Under these conditions, the increase in fluorescence due to ThT binding correlates directly with the appearance of fibrillar aggregates of Ure2 (36). At each time point, one of the samples was placed on ice. A 50- l aliquot of the complete reaction mixture was removed and assayed for GPx activity using 1 mm GSH and 1.2 mm CHP as substrates, as described above. A further 10- l aliquot of the reaction mixture was removed to assay for ThT binding, as described previously (35 37). After centrifugation of the remaining 166 FIG. 1.Glutathione-dependent reduction of CHP catalyzed by Ure2p. The initial velocity of the Ure2-catalyzed reaction is plotted as a function of the Ure2 concentration. The reaction conditions were 100 mm sodium phosphate buffer, ph 7.5, 1 mm GSH, and 1.2 mm CHP at 25 C. Other details are as described under Experimental Procedures. Inset, the non-enzymatic rate (solid line) was measured and subtracted from the reaction rate measured in the presence of Ure2 (1 M (dasheddotted line); 2.5 M (dashed line)) in each case. In the absence of GSH, no reaction was observed. 240 l of sample, a 50- l aliquot of the resulting supernatant was assayed for GPx activity. A further 10- l aliquot of supernatant was used for protein concentration determination using the method of Bradford (38). The precipitate was resuspended in 240 l of the same buffer, and then a 50- l aliquot of the resuspended mixture was assayed for GPx activity. Thus, the final protein concentration in the GPx assays was 1.5 M for the total reaction mixture and a maximum of 1.5 M in either the supernatant or the pellet fraction, depending on the relative distribution of protein between the fractions during the course of fibril formation. The pattern of change observed was highly reproducible in independent experiments. RESULTS Ure2 Shows Glutathione Peroxidase Activity The ability of Ure2 to reduce hydroperoxides was tested in vitro using purified Ure2 with the oxidant substrate CHP and the reducing agent GSH. Reactions were followed by the oxidation of NADPH, which is coupled to the reduction of GSSG to GSH by glutathione reductase. The rate of non-enzymatic oxidation of NADPH in the absence of Ure2 was subtracted in each case. When Ure2 was added in the presence of all of the other components of the assay, a significant increase in the oxidation rate of NADPH was observed (Fig. 1, inset) and the initial velocity of the Ure2-catalyzed reaction was found to be proportional to the Ure2 concentration (Fig. 1, main panel). As is characteristic of an enzyme-catalyzed reaction (32), the enzymatic activity of Ure2 toward CHP showed ph and temperature optima, in this case at around ph 8.0 and 40 C (Fig. 2). In contrast, the uncatalyzed rate was observed to increase steeply above ph 8 or above 30 C (data not shown). Therefore, we adopted ph 7.5 and 25 C as the standard conditions for the Ure2 GPx activity assay. A control using bovine serum albumin in place of Ure2 over the same protein concentration range showed no detectable GPx activity (data not shown). The presence or absence of 4 mm sodium azide had no effect on the Ure2 GPx activity, ruling out the possibility that the observed activity is due to contamination with a heme-containing peroxidase such as catalase or myeloperoxidase. Ure2 showed no peroxidase activity in the absence of GSH. This then demonstrates that Ure2 has GSH-dependent peroxidase activity. Comparison of Different Hydroperoxide Substrates To further characterize the GPx activity of Ure2, we employed steady-state methods to obtain the apparent kinetic parameters for the enzymatic reaction. The results of steady-state kinetic analysis of Ure2 GPx activity toward the substrates

167 The Yeast Prion Ure2p Has Glutathione Peroxidase Activity FIG. 2.Optima of Ure2 catalyzed glutathione peroxidase activity toward CHP. A, ph dependence of activity measured at 25 C. B, temperature dependence of activity measured at ph 7.5. The Ure2 concentration was 1.2 M. Other details are as in Fig. 1. CHP, hydrogen peroxide, and t-bh are shown in Fig. 3 and Table I. When the concentration of one substrate was fixed and the concentration of the other substrate was varied, the GPx activity was hyperbolic with respect to substrate concentration (Fig. 3) and double-reciprocal Lineweaver-Burk plots were linear (Fig. 4), as is typical of adherence to Michaelis-Menten kinetics (31). Ure2 showed activity toward all three substrates with an apparent preference in the order H 2 O 2 CHP t-bh. This indicates that Ure2 has GPx activity toward hydrogen peroxide as well as typical organic hydroperoxide substrates. The Prion Domain Does Not Contribute to Peroxidase Activity The apparent kinetic parameters obtained for wild type Ure2 (with and without a His 6 tag) and for the prion domain deletion mutant 90Ure2 are shown in Table II. The results show that not only does the presence of a His 6 tag have no effect on the enzymatic activity of Ure2 but also that the detected activity cannot be attributed to contamination with a similar enzyme, given the radically different purification methods for tagged and non-tagged protein (see Experimental Procedures ). The parameters obtained for 90Ure2 are the same within error as those obtained for wild type Ure2, indicating that the prion domain does not contribute to the GPx activity. This result is consistent with the finding that the Ure2 prion domain is essentially unstructured in the native dimer (21, 22) and has no affect on the stability or folding of Ure2 (21, 39). Investigation of the Reaction Mechanism and Determination of True Kinetic Parameters To further investigate the mechanism of Ure2 GPx activity, we obtained a data set of initial velocities over a wide range of GSH and CHP concentrations. The exact steady-state solutions of the mechanisms for twosubstrate reactions are extremely complicated. However, in practice, only a limited number of mechanisms are observed 167 FIG. 3. GPx activity of Ure2p with different hydroperoxide substrates measured under steady-state conditions. The fit to the Michaelis-Menten equation is shown. The Ure2 concentration was M. A, varying concentrations of GSH with a fixed hydroperoxide substrate concentration of 1.2 mm for CHP ( )orh 2 O 2 (E)and5mM for t-bh ( ). B, varying concentrations of hydroperoxide substrates with a fixed GSH concentration of 1 mm. Symbols are as in A. and, under certain conditions, the rate equations can be reduced to simple forms (31 34). Fig. 4 shows double-reciprocal plots of the initial velocity versus one substrate concentration, obtained for a range of concentrations of the second substrate. The slope of the lines was observed to decrease with increasing concentration of the second substrate, and the lines intersected at a common point. This rules out a ping-pong mechanism (which is characterized by parallel double-reciprocal plots). The pattern observed is consistent with a sequential mechanism, and the data can be fitted to Equation 1 (32 34), V max [A][B] v K ia K mb K mb [A] K ma [B] [A][B] (Eq. 1) where K ma and K mb are the Michaelis constants for substrates A and B, respectively, and K ia is the inhibition constant for substrate A (which under certain circumstances is equal to the dissociation constant for A binding to the enzyme). It has been demonstrated there are two situations where the form of the equation simplifies in this way (32, 33), namely, a rapid-equilibrium random mechanism or a compulsorily ordered mechanism (Fig. 5). Because of the equivalent form of the rate equations, it cannot be distinguished from the current data whether binding is random or ordered. However, if binding is ordered, given that GSH binds to Ure2 in the absence of a second substrate (26), GSH must bind first. The true kinetic parameters obtained by fitting of the data shown in Fig. 4 to Equation 1 are shown in Table III. Ure2 Fibrillar Aggregates Show Peroxidase Activity Ure2 was incubated under conditions that have been thoroughly characterized by electron microscopy and atomic force microscopy and that promote rapid and abundant fibril formation (35,

168 50028 The Yeast Prion Ure2p Has Glutathione Peroxidase Activity TABLE I Apparent steady-state kinetic constants for Ure2 activity toward different hydroperoxide substrates The apparent kinetic constants were determined from Michaelis-Menten plots of initial velocities versus varying concentrations of one substrate with a fixed concentration of the other substrate under standard assay conditions as described under Experimental Procedures and shown in Fig. 3. The Ure2 concentration, E 0, was M. The values shown are the mean S.E. of repeated measurements. Fixed hydroperoxide substrate a Fixed GSH at1mm Substrate K m(gsh)(app) V max(app) / E 0 K m(app) V max(app) / E 0 mm s 1 mm s 1 CHP H 2 O t-bh a Fixed CHP or H 2 O 2 at 1.2 mm; fixed t-bh at5mm. FIG. 4. Double-reciprocal plots of Ure2p activity at a fixed concentration of one substrate versus varying concentrations of the second substrate for GSH and CHP. The concentration of Ure2p in the reactions was 0.6 M. The observed pattern of intersecting straight lines excludes a ping-pong reaction mechanism and is consistent with either a steady-state ordered sequential mechanism or a rapidequilibrium random sequential mechanism. The parameters obtained by global fitting of the data to Equation 1 are shown in Table III. 36). Under these conditions, it was found that the increase in the binding fluorescence of the amyloid-specific dye ThT correlates directly with the time dependent appearance of fibrillar aggregates of Ure2, providing a convenient method to quantify the extent of fibril formation (36). Aliquots were removed at regular time intervals for analysis (see Experimental Procedures ). The course of fibril formation was monitored by assaying the ThT binding of the incubation mixture and by measuring the decrease in the protein concentration in the supernatant fraction subsequent to sedimentation of the aggregates (Fig. 6A). In parallel, the GPx activity was assayed for the complete reaction mixture, and for the supernatant and pellet fractions (Fig. 6B). Concomitant with the formation of fibrillar aggregates and the loss of protein from the supernatant fraction, the GPx activity of the solution was lost from the supernatant fraction and instead was found in the pellet fraction. The level of activity of the complete incubation mixture, or of 168 the sum of the pellet and supernatant fractions, remained almost constant throughout the course of the experiment (Fig. 6B). This then indicates that Ure2 GPx activity is maintained within ordered aggregates and suggests that the level of activity is essentially unaffected by fibril formation, at least under the conditions used here. DISCUSSION The bovine spongiform encephalopathy epidemic (40) and the subsequent emergence of a new variant of the equivalent human disease (41) has prompted a massive worldwide effort to understand the prion phenomenon (15). The finding that prions also exist in fungi (14) has contributed significantly to establishing the viability of the prion concept (42). To understand the molecular mechanism of prion formation requires characterization of the structural and folding properties of the prion proteins. The natural tendency of prion proteins to aggregate makes this a difficult task. Nevertheless, significant progress has been made in recent years. High resolution structures are available for the mammalian prion protein, PrP (43, 44), and for Ure2 (24 26). In addition, the stability and kinetics of folding have been studied by a number of spectroscopic methods for PrP (45 47) and Ure2 (20, 21, 35, 39, 48, 49). The disadvantage of purely spectroscopic methods for folding studies is that it is often difficult to separate the native-structure signal from those of native-like or partially folded states. Therefore, the availability of an assay for native activity is an extremely important tool in structure-function analysis (31). In the case of Ure2, it was found that the native state could be distinguished from a spectroscopically identical misfolded native-like state by the difference in their unfolding kinetics (39). However, this unfolding assay requires the addition of high concentrations of chemical denaturant and is not readily applicable to fibrillar aggregates. Thus, the establishment of an in vitro activity assay for Ure2, as described here, not only addresses questions regarding the physiological structure and function of Ure2 but also provides an important tool for further mechanistic analysis of Ure2 as both an enzyme and a prion. Ure2 showed glutathione-dependent peroxidase activity toward both hydrogen peroxide and standard organic hydroperoxide substrates (Fig. 3 and Table I). This finding indicates that Ure2, while lacking typical GST activity (23), nevertheless belongs to the subset of GST proteins that are active against oxidant substrates (1, 5). Most GPxs contain a selenocysteine, which reacts covalently with GSH, generally via a ping-pong enzyme reaction mechanism (7). In contrast, GSTs use a conserved tyrosine, serine, or cysteine residue to interact with the thiol group of GSH, thus increasing the reactivity of GSH, typically via a sequential mechanism (8 10). Thus, the observation of a sequential mechanism for Ure2 peroxidase activity (Figs. 4 and 5) is in agreement with the designation of Ure2 as a GST. The residue Asn 124 has been suggested as a candidate for the catalytically essential residue in the Ure2 GSH-binding domain (26), although this remains in question, particularly

169 The Yeast Prion Ure2p Has Glutathione Peroxidase Activity Details are as for Table I. TABLE II Apparent steady-state kinetic constants for Ure2 and 90Ure2 Fixed CHP at 1.2 mm Fixed GSH at1mm Protein K m(gsh)(app) V max(app) / E 0 K m(chp)(app) V max(app) / E 0 mm s 1 mm s 1 Ure2 (no tag) Ure Ure FIG. 5. Minimal scheme for a sequential enzyme reaction mechanism. The enzyme, E, binds to two substrates, A and B, to give products, P and Q. In a rapid-equilibrium situation, the binding of the substrates is rapid compared with the rate of the reaction. The reaction may be ordered with a particular substrate always binding first, or the sequence of substrate binding may be random. TABLE III True kinetic constants for Ure2 derived from steady-state kinetic analysis The kinetic parameters were measured using the coupled enzyme assay with Ure2 0.6 M, GSH mm, and CHP mm to obtain a series of sets of data as shown in Fig. 4. The parameters were obtained by global fitting of the data to Equation 1. The errors shown are the S.E. of the fit. Kinetic parameter Value Method of determination k cat s 1 V max / E 0 K m(gsh) mm Measured K m(chp) mm Measured K I(GSH) mm Measured K m(chp) mm K I(GSH) K m(chp) /K m(gsh) k cat /K m(gsh) M 1 s 1 k cat /K m(gsh) k cat /K m(chp) M 1 s 1 k cat /K m(chp) because enzymatic activity had not been demonstrated until now (25). The establishment of an assay for Ure2 activity paves the way for mutagenesis studies to define the residues required for Ure2 catalytic activity. The kinetic parameters measured for Ure2, namely K m values in the millimolar range and k cat /K m values in the range (Table III), are consistent with the values observed for related enzymes. For example, the S. cerevisiae GSTs (which react with CDNB but not with hydroperoxides) also showed apparent K m(gsh) values in the millimolar range (11). The S. cerevisiae glutaredoxins showed apparent k cat /K m values of around 10 3 for CDNB and 10 4 for CHP with apparent substrate K m values in the millimolar range (12). The apparent specificity constant for the bacterial GST from Proteus mirabilis reacting with CDNB is around 10 3 M 1 s 1 with an apparent K m(gsh) of 0.34 mm and an apparent K m(cdnb) of 2.5 mm (10). The observation of oxidant sensitivity for Ure2 mutants (27), combined with demonstration here that peroxidase activity is an inherent property of the Ure2 protein, indicates that Ure2 is functional in S. cerevisiae cells as a peroxidase. A particularly interesting and controversial aspect of the Ure2 prion protein is its ability to assemble into amyloid-like fibrils (19, 22, 36) while still retaining native-like structural properties (50, 51). We observed the same level of GPx activity in fibrillar aggregates of Ure2 as for the same concentration of 169 FIG. 6.Relationship between peroxidase activity and the time course of fibril formation. Incubation was in 50 mm sodium phosphate buffer, ph 7.5, 0.2 M NaCl at 37 C with shaking, conditions that strongly favor fibril formation (35, 36). A, fibril formation was monitored by assaying changes in ThT binding ( ) and by measuring the protein concentration in the supernatant fraction after centrifugation (E). B, in parallel, the GPx activity in the resuspended pellet fraction ( ), supernatant fraction (E), and total reaction mixture ( ) was assayed. The initial velocities are shown for a final protein concentration in the GPx assay of 1.5 M for the total reaction mixture and a maximum of 1.5 M in either the pellet or the supernatant fraction, depending on the distribution of the protein between the fractions over time. fully dispersed soluble protein (Fig. 6). This finding is consistent with the observation that fibrillar aggregates of Ure2 maintain the ability to bind GSH (50) and fibrils formed from other enzymes linked to the Ure2 prion domain can still react with their specific substrates, provided that the substrate is small enough to diffuse into the fibrillar arrays (51). In principle, the kinetics of an immobilized enzyme may be different from the kinetics of the enzyme free in solution because of one or more of the following factors: 1) a change in conformation; 2) a change in environment; 3) a change in the effective concentration of the substrate; or 4) diffusional effects (52). Diffusional effects will be negligible if the substrate can easily reach the enzyme, or if catalysis is slow with respect to diffusion (i.e. k cat /K m is low) as is the case for Ure2. The results presented here provide direct support for the suggestion that formation of fibrillar aggregates of Ure2 does not involve significant structural change within the C-terminal globular region but rather that the native-like structure and activity are preserved. This finding then supports the hypothesis that a loss of nitrogen metabolite repression in the [URE3] prion state is due to a steric blocking mechanism, rather than to a loss of native-like Ure2 structure (51). Furthermore, these results are consistent with the observation that prion strains show normal sensitivity to oxidant stress (27), indicating that the Ure2 fibril formation assay is an

170 50030 The Yeast Prion Ure2p Has Glutathione Peroxidase Activity excellent model for the structural changes accompanied by prion formation in yeast cells. The availability of a convenient and relevant in vitro assay system to conduct structure-function analysis will allow further investigation of the interplay of the various roles of Ure2: as a regulator of nitrogen metabolism, a detoxification enzyme, and a prion. Acknowledgments We thank Prof. Z. X. Wang, Prof. X. M. Pan, and Dr. H. Wu of this institute and Dr. L. Itzhaki (University of Cambridge) for advice and helpful discussions. REFERENCES 1. Sheehan, D., Meade, G., Foley, V. M., and Dowd, C. A. (2001) Biochem. J. 360, Armstrong, R. N. (1997) Chem. Res. Toxicol. 10, Mannervik, B., Cameron, A. D., Fernandez, E., Gustafsson, A., Hansson, L. O., Jemth, P., Jiang F., Alwyn Jones, T., Larsson, A. K., Nilsson, L. O., Olin, B., Pettersson, P. L., Ridderström, M., Stenberg, G., and Widersten, M. (1998) Chem. Biol. Interact , Board, P. G., Cogga, M., Chelvanayagam, G., Easteal, S., Jermiin, L. S., Schulte, G. K., Danley, D. E., Hoth, L. R., Griffor, M. C., Kamath, A. V., Rosner, M. H., Chrunyk, B. A., Perregaux, D. E., Gabel, C. A., Geoghegan, K. F., and Pandit, J. (2000) J. Biol. Chem. 275, Vuilleumier, S. (1997) J. Bacteriol. 179, Rossjohn, J., Board, P. G., Parker, M. W., and Wilce, M. C. J. (1996) Protein Eng. 9, Saito, Y., Hayashi, T., Tanaka, A., Watanabe, Y., Suzuki, M., Saito, E., and Takahashi, K. (1999) J. Biol. Chem. 274, Nay, B., Fournier, D., Baudras, A., and Baudras, B. (1999) Insect Biochem. Mol. Biol. 29, Labrou, N. E., Mello, L. V., and Clonis, Y. D. (2001) Biochem. J. 358, Caccuri, A. M., Antonini, G., Allocati, N., Di Ilio, C., De Maria, F., Innocenti, F., Parker, M. W., Masulli, M., Lo Bello, M., Turella, P., Federici, G., and Ricci, G. (2002) J. Biol. Chem. 277, Choi, J. H., Lou, W., and Vancura, A. (1998) J. Biol. Chem. 273, Collinson, E. J., and Grant, C. M. (2003) J. Biol. Chem. 278, Avery, A. M., and Avery, S. V. (2001) J. Biol. Chem. 276, Wickner, R. B. (1994) Science 264, Prusiner, S. B. (1998) Proc. Natl. Acad. Sci. U. S. A. 95, Masison, D. C., and Wickner, R. B. (1995) Science 270, Bach, S., Talarek, N., Andrieu, T., Vierfond, J. M., Mettey, Y., Galo, H., Dormont, D., Meijer, L., Cullin, C., and Blondel, M. (2003) Nat. Biotechnol. 21, Masison, D. C., Maddelein, M. L., and Wickner, R. B. (1997) Proc. Natl. Acad. Sci. U. S. A. 94, Taylor, K. L., Cheng, N., Williams, R. W., Steven, A. C., and Wickner, R. B. (1999) Science 283, Thual, C., Bousset, L., Komar, A. A., Walter, S., Buchner, J., Cullin, C., and Melki, R. (2001) Biochemistry 40, Perrett, S., Freeman, S, J., Butler, P. J. G., and Fersht, A. R. (1999) J. Mol. Biol. 290, Thual, C., Komar, A. A., Bousset, L., Fernandez-Bellot, E., Cullin, C., and Melki, R. (1999) J. Biol. Chem. 274, Coshigano, P. W., and Magasanik, B. (1991) Mol. Cell. Biol. 11, Bousset, L., Berlhali, H., Janin, J., Melki, R., and Morera, S. (2001) Structure 9, Umlaud, T. C., Taylor, K. L., Rhee, S., Wickner, R. B., and Davis, D. R. (2001) Proc. Natl. Acad. Sci. U. S. A. 98, Bousset, L., Berlhali, H., Melki, R., and Morera, S. (2001) Biochemistry 40, Rai, R., Tate, J. J., and Cooper, T. G. (2003) J. Biol. Chem. 278, Fraser, J. A., Davis, M. A., and Hynes, M. J. (2002) Appl. Environ. Microbiol. 68, Basu, U., Southron, J. L., Stephens, J. L., and Taylor, G. J. (2004) Mol. Gen. Genomics 271, Flohe, L., and Gunzler, W. A. (1984) Methods Enzymol. 104, Fersht, A. R. (1998) Structure and Mechanism in Protein Science, W. H. Freeman, San Francisco 32. Laidler, K. J., and Bunting, P. S. (1973) The Chemical Kinetics of Enzyme Action, Oxford University Press, Oxford 33. Seubert, P. A., Renosto, F., Knudson, P., and Segel, I. H. (1985) Arch. Biochem. Biophys. 240, Wu, H., Zheng, Y., and Wang, Z. X. (2003) Biochemistry 42, Zhu, L., Zhang, X. J., Wang, L. Y., Zhou, J. M., and Perrett, S. (2003) J. Mol. Biol. 328, Jiang, Y., Li, H., Zhu, L., Zhou, J. M., and Perrett, S. (2004) J. Biol. Chem. 279, Uversky, V. N., Li, J., and Fink, A. L. (2001) J. Biol. Chem. 276, Bradford, M. M. (1976) Anal. Biochem. 72, Galani, D., Fersht, A. R., and Perrett, S. (2002) J. Mol. Biol. 315, Hope, J., Reekie, L. J., Hunter, N., Multhaup, G., Beyreuther, K., White, H., Scott, A. C., Stack, M. J., Dawson, M., and Wells, G. A. (1988) Nature 336, Will, R. G., Ironside, J. W., Zeidler, M., Cousens, S. N., Estibeiro, K., Alperovitch, A., Poser, S., Pocchiari, M., Hofman, A., and Smith, P. G. (1996) Lancet 347, Tuite, M. F., and Cox, B. S. (2003) Nature Rev. Mol. Cell. Biol. 4, Riek, R., Hornemann, S., Wider, G., Billeter, M., Glockshuber, R., and Wuthrich, K. (1996) Nature 382, Zahn R., Liu, A., Luhrs, T. Riek, R., von Schroetter, C., Lopez Garcia, F., Billeter, M., Calzolai, L., Wider, G., and Wuthrich, K. (2000) Proc. Natl. Acad. Sci. U. S. A. 97, Torrent, J., Alvarez-Martinez, M. T., Heitz, F., Liautard, J. P., Balny, C., and Lange, R. (2003) Biochemistry 42, Apetri, A. C., Surewicz, K., and Surewicz, W. K. (2004) J. Biol. Chem. 279, Hosszu, L. L., Jackson, G. S., Trevitt, C. R., Jones, S., Batchelor, M., Bhelt, D., Prodromidou, K., Clarke, A. R., Waltho, J. P., and Collinge, J. (2004) J. Biol. Chem. 279, Zhou, J. M., Zhu, L., Balny, C., and Perrett, S. (2001) Biochem. Biophys. Res. Commun. 287, Zhu L., Kihara, H., Kojima, M., Zhou, J. M., and Perrett, S. (2003) Biochem. Biophys. Res. Commun. 311, Bousset, L., Thomson, N. H., Radford, S. E., and Meiki, R. (2002) EMBO J. 21, Baxa, U., Speransky, V., Steven, A. C., and Wickner, R. B. (2002) Proc. Natl. Acad. Sci. U. S. A. 99, Laidler, K. J., and Bunting, P. S. (1980) Methods Enzymol. 64,

171 Structure, Vol. 12, , August, 2004, 2004 Elsevier Ltd. All rights reserved. DOI /j.str Crystal Structure of an Acylpeptide Hydrolase/Esterase from Aeropyrum pernix K1 Mark Bartlam, 1,3 Ganggang Wang, 1,3 Haitao Yang, 1,3 Renjun Gao, 2 Xiaodong Zhao, 1 Guiqiu Xie, 2 Shuigui Cao, 2 Yan Feng, 2 and Zihe Rao 1, * 1 Laboratory of Structural Biology -melanocyte-stimulating hormone, suggesting a possible biological role in controlling the concentration of these factors (Jones and Manning, 1988). Sequence analysis indicates that acylpeptide hydrolases share a Tsinghua University and strong resemblance in their C-terminal domains, which National Laboratory of Biomacromolecules is thought to reflect their peptidase activity (Barrett and Institute of Biophysics Rawlings, 1992; Rawlings et al., 1991). Chinese Academy of Science To date, human, porcine, and rat acylpeptide hy- Beijing drolases from various tissues have been characterized China (Kobayashi et al., 1989; Mitta et al., 1996; Miyagi et al., 2 Key Laboratory for Molecular Enzymology 1995). All are 732 amino acids in length, share more than and Engineering of Ministry of Education 90% sequence identity with each other, and are reported Jilin University to form homotetramers. The catalytic residues in human Changchun and porcine enzymes have been identified by chemical China modification and site-directed mutagenesis experiments (Mitta et al., 1998; Scaloni et al., 1992). A deficiency in the expression of human APH has been linked Summary with small cell lung carcinomas and renal carcinomas (Erlandsson et al., 1991; Naylor et al., 1989), while APH Acylpeptide hydrolases (APH; also known as acylam- in porcine brain has been identified as a sensitive site ino acid releasing enzyme) catalyze the removal of an for reaction with organophosphorus compounds and is N-acylated amino acid from blocked peptides. The a potential target for cognitive enhancing drugs (Richcrystal structure of an APH from the thermophilic arch- ards et al., 2000). Further evidence suggests that APH aeon Aeropyrum pernix K1 to 2.1 Å resolution confirms may be a more sensitive target for cognitive-enhancing it to be a member of the prolyl oligopeptidase family organophosphorus compounds than acetylcholinesterof serine proteases. The structure of apaph is a sym- ase (AChE) (Duysen et al., 2001; Richards et al., 2000). metric homodimer with each subunit comprised of two As such, the acylpeptide hydrolases have a high potendomains. The N-terminal domain is a regular seven- tial for drug discovery. An acylpeptide hydrolase from bladed -propeller, while the C-terminal domain has the thermophilic archaeon Pyrococcus horikoshii has a canonical / hydrolase fold and includes the active also been characterized, but is 100 residues shorter than site and a conserved Ser445-Asp524-His556 catalytic its mammalian counterparts and forms a homodimer triad. The complex structure of apaph with an organo- (Ishikawa et al., 1998). phosphorus substrate, p-nitrophenyl phosphate, has Despite advances in understanding the biological also been determined. The complex structure unamfunctions of acylpeptide hydrolases, little is known of biguously maps out the substrate binding pocket and the structural basis for the sequential deacetylation of provides a basis for substrate recognition by apaph. N-terminally acetylated proteins. Human APH has been A conserved mechanism for protein degradation from crystallized, but the structure is still unavailable today archaea to mammals is suggested by the structural (Freese et al., 1993). Recently, suitable crystals of an features of apaph. APH from the thermophilic archaeon Aeropyrum pernix Introduction K1 (apaph) were obtained for structure determination (Wang et al., 2002, 2003). Aeropyrum pernix K1 is an aerobic strain classified as crenarchaeota in archaeon Acylpeptide hydrolases (APH; also known as acylamino (Sako et al., 1996); the complete genome of A. pernix acid releasing enzyme or acylaminoacyl peptidase K1 has been sequenced, and four genes (Ape1547, [EC ]) catalyze the removal of an N-acylated amino acid from blocked peptides (Tsunasawa et al., Ape1832, Ape2290, and Ape2441) have been desig- 1975). Peptide substrates of various sizes and with difdescribes the structure of apaph, the gene product of nated as encoding acylpeptide hydrolases. This paper ferent acyl groups at the N terminus (acetyl, formyl, and chloroacetyl) can be hydrolyzed by APH to generate an Ape1547, to 2.1 Å resolution. Although apaph shares acyl amino acid and a peptide with a free N terminus only 27% sequence identity with human acylpeptide that is shortened by one amino acid. Acetylation occurs hydrolase, there is a surprising conservation of second- during or following the biosynthesis of the polypeptide ary structure between eukaryotic APH proteins and chains, suggesting that this process protects the intraapaph in complex with an organophosphorus (OP) sub- apaph. We have also determined the structure of cellular proteins from proteolysis (Hershko et al., 1984). Acylpeptide hydrolases are also active on small acetshare strate, p-nitrophenyl phosphate (pnp). Serine proteases ylated bioactive peptides, such as -endorphin and common inhibitors and have the potential to react with OP compounds, and so the complex of apaph with *Correspondence: raozh@xtal.tsinghua.edu.cn pnp should provide a structural basis for the design of 3 These authors contributed equally to this work. specific inhibitors for acylpeptide hydrolases. 171

172 Structure 1482 Table 1. Data Collection and Refinement Statistics Data Collection Statistics MAD Peak MAD Edge MAD Remote apaph-pnp Complex Wavelength (Å) Space group P P Unit cell (Å, ) a Å, b Å, a Å,b Å,c c Å, Å, Resolution limit (Å) ( ) ( ) ( ) ( ) Total reflections 427, , , ,959 Unique reflections 67,958 59,504 59,604 31,768 Completeness 99.2 (97.5) 99.5 (98.2) 99.6 (98.5) 100 (100) a R merge 7.9 (21.2) 6.3 (17.7) 12.4 (37.9) 10.0 (37.7) I/ (I) 16.6 (6.1) 18.4 (7.5) 9.7 (3.8) 6.9 (2.0) Redundancy 7.4 (7.2) 7.4 (7.2) 7.5 (7.0) 6.8 (7.0) Refinement Statistics apaph Resolution range (Å) b R work b R free Rms deviation bonds (Å) angles ( ) Average B factors (Å 2 ) Protein (chain A/B) 19.4/ /34.4 Water Ramachandran plot Favored (%) Allowed (%) Generously allowed (%) Disallowed (%) apaph-pnp Complex Numbers in parentheses correspond to the highest resolution shell. a R merge h i I ih I h / h i I h, where I h is the mean of the observations I ih of reflection h. b R work ( F p (obs) F p (calc) )/ F p (obs) ;R free R factor for a selected subset (5%) of the reflections that was not included in prior refinement calculations. Results and Discussion and 2A). The N-terminal domain (residues ) is a -propeller with seven blades; each blade consists of Structure Determination a four-stranded antiparallel sheet. The C-terminal do- The structure of apaph was determined by MAD phasfold, main (residues ) has a canonical / hydrolase ing from a selenomethionyl derivative. Data sets were with a central eight-strand mixed sheet flanked collected at peak, edge, and remote wavelengths on by five helices on one side and six helices on the other. beamline BL41XU of SPring-8 (Hyogo, Japan). The crysthe This central sheet is all parallel with the exception of tal contained two molecules in the asymmetric unit. The second strand. A short helix at the N-terminal quality of the experimental electron density map was (residues 8 23) extends from the -propeller domain good such that residues of chain A and residues and forms part of the hydrolase domain. Between the of chain B could be traced continuously. No denwhich domains is a large cavity approximately 45 Å in width, sity was observed for residues 1 7 and 582. The final is accessible via a tunnel in the -propeller do- model consists of 1148 residues, 580 water molecules, main. Three prolines Pro312, Pro359, and Pro370 in 2 -octyl glucoside molecules, and 2 glycerol molecules. each protomer are in the cis conformation. Data collection and refinement statistics are summa- Interestingly, the structure includes a detergent molerized in Table 1. cule, -octyl glucoside, bound within the central cavity The structure of the apaph-pnp complex was deter- of each subunit. The detergent molecule, added during mined from a single crystal by molecular replacement, crystallization, does not bind to the active Ser445 but using data collected in-house to 2.7 Å resolution. Clear instead forms a single hydrogen bond with the O atom electron density for the substrate was observed from of Ser26. Further hydrophobic contacts are made with an F o F c difference map in the binding pocket of residues Phe41, Val46, Phe381, Ile558, Ala564, and each subunit. Continuous electron density was ob- Leu568. A glycerol molecule is also found within the served for residues of chain A and of chain central channel of the -propeller in each subunit. Coor- B. The final model consists of 1146 residues, 181 water dination of the glycerol occurs via hydrogen bonds with molecules, and 2 pnp molecules. Ser157 and Leu251 and hydrophobic contacts with Ser199, Ala200 and Trp250. Structural Overview A DALI (Holm and Sander, 1998) search for structural The structure of apaph is a symmetric homodimer, and similarity confirms that the overall architecture of apaph each subunit is comprised of two domains (Figures 1 most resembles prolyl oligopeptidase (POP; PDB ID 172

173 Crystal Structure of apaph 1483 Hydrophobic interactions are mediated by residues positioned on helices 1, 11, and 12. In contrast to acylpeptide hydrolases from A. pernix K1 and P. horikoshii, which form homodimers, their mammalian counterparts have been reported to form homotetramers. A multiple sequence alignment shows that the residues in the dimer interface are conserved in mammalian enzymes, suggesting a common mode of dimer formation. Unlike the archaeal enzymes, however, the mammalian enzymes all possess a large hydrophobic insertion near the N-terminal, which may be involved in higher oligomer formation similar to DPP-IV (Engel et al., 2003). Lack of this N-terminal insertion may explain why apaph forms a dimer rather than a tetramer. 1QFM) (Fulop et al., 1998) and dipeptidyl peptidase IV/ CD26 (DPP-IV; PDB ID 1ORV) (Engel et al., 2003), both of which are members of the prolyl oligopeptidase family of serine proteases. As with apaph, both structures possess an N-terminal -propeller domain and a C-ter- minal / hydrolase domain. The -propeller domain of POP features seven blades, whereas the corresponding domain of DPP-IV has eight blades. Unsurprisingly, similarities were also found between the C-terminal domain of apaph and other esterase structures sharing the same hydrolase fold, including bacterial cocaine ester- ase (Larsen et al., 2002). Dimerization The structure of apaph is a symmetric dimer in which the two subunits are related by a 2-fold rotation axis (Figure 2B). Protomers A and B are essentially similar, with an rmsd between them of 0.4 Å for all C atoms. The dimer interface is located exclusively in the C-terminal hydrolase domain, and the total surface area buried by the dimer is 2060 Å 2. The subunits are arranged such that the central hydrolase sheet of one subunit forms an extension of the central hydrolase sheet of the second subunit. Hydrogen bonds are formed between strand 37 of each subunit by the residues Lys544, Thr545, Phe546, Ala548, His549, Ile550, and Asp553. Additional hydrogen bonds are formed between residues located on the N-terminal 1 helix (Ser10, Glu17) of one subunit and helix 11 (Gln522) of its partner subunit. Figure 1. Topology of the apaph Structure A topology diagram showing the domain structure of apaph. The helices are shown as yellow cylinders, strands in the N-terminal domain are shown in blue, and strands in the C-terminal domain are shown in red. Two cysteine residues (Cys416 and Cys453) located in the C-terminal domain form a disulfide bridge between helices 6 and 7 and are shown as green circles. -Propeller Domain The N-terminal domain (residues ) is a regular -propeller consisting of seven blades, each of which is made up of four antiparallel strands (Figure 2C). Blade III is the single exception since it has an additional fifth strand due to crosslinking from blade II. The sheets are twisted and radially arranged around a pseudo 7-fold axis such that they pack face to face. The central tunnel of the -propeller is lined with hydrogen donors and acceptors, which are water solvated. The -propeller is connected to the catalytic domain via two polypeptide main chains. The two domains are stabilized by 29 hydrogen bonds and salt bridges, with additional stability provided by hydrophobic forces. Internal structural stability is provided predominantly by hydrophobic interactions. There is also a high fre- quency of charged residues (Arg and Glu) located on the end of strands in the -propeller domain, resulting in the formation of 12 ion pairs between neighboring blades. These ion pairs may be a contributory factor to the high thermostability of apaph. It is worth noting that five of the seven blades have an aspartate residue (Asp52, Asp119, Asp140, Asp224, and Asp274) located on the end of the third strand. A similar motif is found in the -propeller domain of POP (Fulop et al., 1998). The significance of this aspartate motif is not clear, but in both structures, the aspartates are directed into solvent. While the primary sequence similarity between -propeller domains is generally low, their three-dimen- sional structures can be closely superimposed. The apaph -propeller domain is more regular and compact than the POP propeller, with an rmsd between them of 2.7 Å. It more closely resembles the seven-bladed propellers of Integrin V 3 (PDB ID 1JV2) or G protein (PDB ID 1TBG), with respective rmsds of 1.9 Å and 2.1 Å. A number of -propeller domain structures have evolved ways to close the circle between the first and last blades, including covalent bonds between the first and last blades, or by strand exchange between the first and last blades. The apaph -propeller is not stabilized in this way, and only hydrophobic interactions exist between the first and last blades. Catalytic Domain The catalytic domain has a canonical / -hydrolase fold and spans residues 8 23 and (Figure 2D). The central eight-stranded sheet is all parallel, with the exception of the second strand, and is twisted by more 173

174 Structure 1484 Figure 2. The Structure of apaph (A) A ribbon diagram showing the structure of apaph. The coloring is from blue at the N terminus to red at the C terminus. The N- and C termini are labeled. (B) The apaph dimer structure. Each subunit is colored from blue at the N terminus to red at the C terminus. (C) The -propeller domain viewed down the pseudo 7-fold axis. (D) A view of the / -hydrolase domain showing the twisted central sheet. than 90 in line with other serine proteases (Figure 2D). The central sheet is flanked by six helices on one side and five helices on the other. The primary sequence of apaph contains only two cysteines (Cys416 and Cys453), which are located in the catalytic domain and form a disulfide bond linking helices 6 and 7. Active Site The serine proteases are known to possess a conserved Ser-Asp-His catalytic triad. The three-dimensional arrangement of Ser445, Asp524, and His556 in apaph matches with other hydrolase structures. This triad is located in the C-terminal hydrolase domain where it is covered by the tunnel formed by the N-terminal -propeller domain. Ser445 is located on the turn between 34 and 7; Asp524 is located on the loop between 36 and 11; and His556 is located on the loop between 36 and 12. Characteristic of / hydrolases, the active serine is located on a sharp turn known as a nucleophile elbow. The sequence surrounding the active serine is Gly-Tyr- Ser-Tyr-Gly, which is consistent with the Gly-X-Ser-X- Gly consensus sequence observed in the / hydrolase folds of the lipase, esterase, and serine protease super- family. The main chain conformation of Ser445 is strained, with (φ, ) (61.3, ). This is an energetically unfavorable conformation also observed in other / hydrolases and is believed to provide an energy reservoir for catalysis. The location of several glycine residues (Gly443, Gly447, and Gly448) in very close proximity to the catalytic Ser445 allows the avoidance of any steric hindrance in the sharp turn of the nucleophile elbow. The net result is that Ser445 is well exposed and readily accessible to both the catalytic His556 imidazole group and the substrate. The main entrance to the active site is via a tunnel in 174

175 Crystal Structure of apaph 1485 the -propeller domain, whose diameter is approximately 7 Å; this is large enough for small peptides to enter but would clearly prevent larger peptides and proteins from accessing the central cavity and undergoing accidental hydrolysis. A second, smaller side opening also provides access to the active site and is located between blades 1 and 2. The 6Åwide opening is lined by the residues Asn65, Arg81, Asp82, Glu88, Asp553, Ala557, Ile558, Asn559, and Asn563. As with the main propeller entrance, the side entrance is also water solvated. Assuming the reaction proceeds via a general serine protease mechanism, this side opening may provide an exit for the reaction product following nucleophilic attack and formation of an acyl-enzyme intermediate. Substrate Recognition and Catalytic Mechanism Acylpeptide hydrolases are unique among the prolyl oligopeptidase family for their substrate preference, which is a short peptide blocked at the N terminus. In order to understand more about the substrate specificity of apaph, we determined the structure of a complex with p-nitrophenyl phosphate, a small organophosphorus compound known to be a nonspecific inhibitor of esterases. As with many serine proteases, apaph is a bifunctional enzyme, possessing both acylpeptide hydrolase and esterase activity (Y.F., unpublished data). This similarity between esterases and acylpeptide hydrolases means that they share common inhibitors (Scaloni et al., 1994). The complex structure unambiguously maps out the S1 substrate binding pocket located in close proximity to the active site (Figures 3A and 3B). The pocket provides a hydrophobic environment for the substrate and is lined by the residues Met477, Phe485, Phe488, Ile489, Leu492, Trp474, Tyr446, and Val471. Of these, only Met477 is conserved in human, porcine, and rat APH. It is surprising to note that the phosphate group of pnp is not covalently attached to the catalytic serine. Instead, in protomer B the O3P atom of the phosphate group is hydrogen bonded to the O atom of Ser445 with a distance of 2.8 Å (Figure 3C). The O2P atom forms a hydrogen bond with the main chain amide of Gly369 (2.9 Å), indicating the location of the oxyanion binding site. Finally, a water molecule is hydrogen bonded to the O4P atom (2.4 Å). The phenyl ring is stabilized by hydrophobic interactions with two phenylalanines, Phe485 and Phe488, as well as with Thr527. A similar orientation of the substrate is observed in the pocket of protomer A, but with the lack of the water molecule hydrogen bonded to O4P. Consequently, the hydrogen bond distance between O3P and Ser445 O is reduced to 2.3 Å, and the hydrogen bond distance between O2P and the Gly369 main chain amide is also reduced to 2.3 Å. Previous studies of apaph have indicated that the substrate recognition is not specific. Of a series of Acamino acid-pnas tested, apaph shows the highest activity for Ac-Phe and Ac-Leu substrates, while the lowest activity is for Ac-Ala and Ac-Lys (Y.F., unpublished data). Conversely, the human and rat forms of APH show highest activity for Ac-Ala, Ac-Met, and Ac-Ser. In addition, apaph also has high activity for the dipeptides Ala-Phe and Ala-Asp. It is likely that Phe485 and Phe488, which serve as anchors for the pnp substrate, influence the recognition of Phe and Leu in the P1 position. Further study of the substrate specificity using p-nitrophenyl alkanoate esters (C2-C18 acyl groups) shows that apaph has optimal activity for substrates with an acyl chain length of C8 (Y.F., unpublished data). The S2 pocket is not mapped by the pnp substrate, but its location can be inferred through a detailed comparison with POP and DPP-IV. The putative S2 pocket is also a hydrophobic environment and is particularly rich in phenylalanines (Phe153, Phe155, Phe163, and Phe371). The S2 pocket of DPP-IV features a dual Glu- Glu recognition motif (Glu205-Glu206), which binds the free amino terminus of the P2 residue and is essential for enzyme activity. While apaph does not have this Glu-Glu motif, it does have an equivalent binding site formed by Phe153 and Phe155. The S2 site is also lined by Arg526, which is structurally equivalent to Arg125 in DPP-IV and Arg643 in POP. This arginine has been confirmed in both DPP-IV and POP to stabilize and acti- vate the P2 residue carbonyl oxygen. Arg526 is also found to be conserved in all other APH sequences. Fur- ther work is required to confirm the specific role of this residue. The oxyanion binding site is an essential feature for serine protease catalysis. The negatively charged oxy- anion is generated from the carbonyl of the scissile bond and stabilized by two hydrogen bonds. In apaph, a hydrogen bond is made by the main chain nitrogen of Gly369 to the O2P atom of the phosphate group of pnp. The second hydrogen bond is most likely provided by the main chain amide of Tyr446, which is located 4 Å from the O2P atom of the phosphate group of pnp. This arrangement, in which one of the bonds is formed by the main chain amide group adjacent to the catalytic serine, is typical of the / hydrolase fold family. A different hydrogen bonding pattern is observed in POP, wherein the equivalent to the Gly369 main chain hydro- gen bond is provided by the hydroxyl group of Tyr473 and not by the main chain amide group. However, super- position of their active sites shows that the OH group of Tyr473 in POP is in a structurally equivalent position to the amide nitrogen of Gly369 in apaph. Conclusions In summary, we have successfully determined the struc- ture of an acylpeptide hydrolase from the thermophilic archaeon Aeropyrum pernix K1. To the best of our knowledge, this is the first structure of an acylpeptide hydrolase to be determined. The structure confirms that acylpeptide hydrolases are members of the prolyl oligopeptidase family of serine proteases. The apaph struc- ture shares the catalytic / hydrolase domain of other serine proteases, as well as the -propeller domain found in prolyl oligopeptidase family structures for the specific recognition of small peptides. The apaph struc- ture also includes many features known to be important for serine protease catalysis, such as the Ser-Asp-His catalytic triad, Gly-X-Ser-X-Gly sequence motif, and oxyanion binding site. Despite the relatively low se- quence similarity among acylpeptide hydrolases, the 175

176 Structure 1486 Figure 3. The Active Site of apaph (A) Cross-sections through the surface of apaph showing the central cavity and the bound pnp molecule near the active site. The pnp molecule is shown in cyan, and the location of the active Ser445 is marked in yellow on the molecular surface. The apaph structure is shown in ribbon form with the same coloring as in Figure 2. (B) A stereo diagram showing the active site and S1 pocket of apaph in the complex structure. Active site residues are shown in green, and the bound pnp substrate is shown in yellow. An omit map contoured at 1 is shown covering the pnp substrate. (C) A schematic showing the pnp substrate and active site residues. Ser445, Asp524, and His556 form the catalytic triad, while Gly369 and Tyr446 form the oxyanion binding site. The gray circle represents a water molecule. Hydrogen bonds are shown as dotted lines, and hydrogen bond distances are given. Residues Phe485, Phe488, and Thr527 form hydrophobic contacts with pnp. structural features of apaph suggest a general serine protease mechanism for protein degradation from archaea to mammals. We have also determined the structure of a complex with a small OP substrate, p-nitrophenyl phosphate. The complex structure reveals the basis for substrate recognition by apaph and maps out the active site residues important for catalysis, including the oxyanion binding site. The design of specific inhibitors for acylpeptide hydrolases will be important for further study of this important family of enzymes. Experimental Procedures Cloning, Expression, Purification, and Crystallization apaph The cloning, expression, purification, and crystallization of apaph have been described previously (Chen et al., 2002; Wang et al., 176

177 Crystal Structure of apaph ). Briefly, the purified protein was concentrated to 10 mg/ml References and transferred into a buffer containing 20 mmol/l Tris-HCl (ph 8.0). For preparation of the Se-Met protein derivative, 5 mm DTT was Barrett, A.J., and Rawlings, N.D. (1992). Oligopeptidases, and the used for antioxidation during the purification procedure. Crystallization emergence of the prolyl oligopeptidase family. Biol. Chem. Hoppe trials were conducted at 291 K in 16-well plates using the hang- Seyler 373, ing-drop vapor-diffusion method. The best crystals were obtained Brunger, A.T., Adams, P.D., Clore, G.M., DeLano, W.L., Gros, P., from reservoir of 6% PEG4000, 50 mm/l NaAC (ph 4.6), 15 mm/l Grosse-Kunstleve, R.W., Jiang, J.S., Kuszewski, J., Nilges, M., DTT, 0.2 mm/l EDTA. Pannu, N.S., et al. (1998). Crystallography & NMR system: a new pnp inhibitor complex software suite for macromolecular structure determination. Acta apaph crystals were soaked with 20 mm p-nitrophenyl phosphate Crystallogr. D Biol. Crystallogr. 54, dissolved in the buffer containing 6% PEG 4000, 25 mm NaAC (ph Chen, Y., Zhang, X., Liu, H., Wang, Y., and Xia, X. (2002). Study 4.6). A 5 l aliquot of such solution was added to the drop, and the on Pseudomonas sp. WBC-3 capable of complete degradation of crystals were soaked overnight. All of the crystals were cryopromethylparathion. Wei Sheng Wu Xue Bao 42, tected in 30% glycerol prior to freezing in a cryostream. Duysen, E.G., Li, B., Xie, W., Schopfer, L.M., Anderson, R.S., Broomfield, C.A., and Lockridge, O. (2001). Evidence for nonacetylcholines- Data Collection terase targets of organophosphorus nerve agent: supersensitivity of Three data sets were collected at peak ( Å), edge ( acetylcholinesterase knockout mouse to VX lethality. J. Pharmacol Å), and remote ( Å) wavelengths from a single Exp. Ther. 299, selenomethionyl derivative crystal on beamline BL14XU of SPring-8 Engel, M., Hoffmann, T., Wagner, L., Wermann, M., Heiser, U., Kiefer- (Hyogo, Japan). Data were processed and scaled using HKL 2000 sauer, R., Huber, R., Bode, W., Demuth, H.U., and Brandstetter, H. (Otwinowski and Minor, 1997) to a maximum resolution of 2.1 Å. (2003). The crystal structure of dipeptidyl peptidase IV (CD26) re- Data from a single crystal of apaph in complex with p-nitrophenyl veals its functional regulation and enzymatic mechanism. Proc. Natl. phosphate were collected in-house on a Rigaku RAXIS-IV detec- Acad. Sci. USA 100, tor at wavelength Å and 100 K. All data were processed with MOSFLM (Leslie, 1992) to 2.7 Å resolution, and scaled and merged Erlandsson, R., Boldog, F., Persson, B., Zabarovsky, E.R., Allikmets, with SCALA (Evans, 1997). R.L., Sumegi, J., Klein, G., and Jornvall, H. (1991). The gene from the short arm of chromosome 3, at D3F15S2, frequently deleted in renal cell carcinoma, encodes acylpeptide hydrolase. Oncogene 6, Structure Determination and Refinement The structure of apaph was determined to 2.1 Å resolution by the multiwavelength anomalous dispersion (MAD) method. CNS Evans, P.R. (1997). SCALA. In Joint CCP4 ESF-EACBM Newslet- (Brunger et al., 1998) was used to locate 20 selenium atoms and ter, pp calculate initial phases to 3.0 Å. The initial phases and heavy-atom Freese, M., Scaloni, A., Jones, W.M., Manning, J.M., and Remington, sites were then input into RESOLVE (Terwilliger, 2000) for phase S.J. (1993). Crystallization and preliminary X-ray studies of human extension and density modification. Noncrystallographic symmetry erythrocyte acylpeptide hydrolase. J. Mol. Biol. 233, (NCS) consistent with point group (rotation) symmetry was found Fulop, V., Bocskei, Z., and Polgar, L. (1998). Prolyl oligopeptidase: among 18 of the 20 heavy-atom sites, and NCS averaging was an unusual beta-propeller domain regulates proteolysis. Cell 94, consequently used by RESOLVE to improve the quality of the elec tron density map. A total of 980 residues were traced automatically Hershko, A., Heller, H., Eytan, E., Kaklij, G., and Rose, I.A. (1984). Role in two molecules, and the remainder of the structure was built into of the alpha-amino group of protein in ubiquitin-mediated protein the experimental electron density map using O (Jones et al., 1991) breakdown. Proc. Natl. Acad. Sci. USA 81, and ARP/wARP (Perrakis et al., 1999). Refinement of the model was Holm, L., and Sander, C. (1998). Protein folds and families: sequence performed using CNS with alternate cycles of manual rebuilding in and structure alignments. Nucleic Acids Res. 26, O. NCS restraints were applied in the early stages of refinement and were released in later stages. The current model has working Ishikawa, K., Ishida, H., Koyama, Y., Kawarabaysi, Y., Kawahara, J., and free R factors of 19.3% and 23.0%, respectively. The model Matsui, E., and Matsui, I. (1998). Acylamino acid-releasing enzyme has good stereochemistry, with 90.1% of residues in the most fa- from the thermophilic archaeon Pyrococcus horikoshii. J. Biol. vored region of the Ramachandran plot generated by PROCHECK Chem. 273, (Laskowski et al., 1993) and none in disallowed regions. Jones, W.M., and Manning, J.M. (1988). Substrate specificity of an The structure of the apaph-pnp complex was refined to 2.7 Å acylaminopeptidase that catalyzes the cleavage of the blocked resolution using the native wild-type structure as a starting point. amino termini of peptides. Biochim. Biophys. Acta 953, Manual adjustments were made to the model in O, and refinement Jones, T.A., Zou, J.Y., Cowan, S.W., and Kjeldgaard, M. (1991). was performed in CNS. From an F o F c difference map, a Improved methods for building protein models in electron density p-nitrophenyl phosphate inhibitor molecule was located near the maps and the location of errors in these models. Acta Crystallogr. active site of each subunit. The current model has working and free A 47, R factors of 20.8% and 26.9%, respectively. Stereochemistry of the Kobayashi, K., Lin, L.W., and Yeadon, J.E. (1989). Cloning and sestructure is good, with 85.0% of residues in the most favored region quence analysis of a rat liver cdna encoding acyl-peptide hydrolase. of the Ramachandran plot generated by PROCHECK. J. Biol. Chem. 264, Acknowledgments Larsen, N.A., Turner, J.M., and Stevens, J. (2002). Crystal structure of a bacterial cocaine esterase. Nat. Struct. Biol. 9, We thank Min Yao for assistance during data collection at SPring-8, Laskowski, R.A., MacArthur, M.W., Moss, D.S., and Thornton, J.M. Japan. We are also grateful to Kazuhiko Ishikawa, Hiroyasu Ishida, (1993). PROCHECK: a program to check the stereochemical quality Susumu Ando, and Yoshitsugu Kosugi. Z.R. was supported by of protein structures. J. Appl. Crystallogr. 26, grants from Projects 863 (no. 2002BA711A12) and 973 (no. Leslie, A.G.W. (1992). Recent changes to the MOSFLM package for G ) of the Ministry of Science and Technology, China. processing film and image plate data. In Joint CCP4 ESF-EAMCB Y.F. was supported by the EYTP (the Excellent Young Teachers Newsletter on Protein Crystallography 26. Program of MOE, China). Mitta, M., Ohnogi, H., and Mizutani, S. (1996). The nucleotide sequence of human acylamino acid-releasing enzyme. DNA Res. 3, Received: April 1, Revised: May 16, 2004 Mitta, M., Miyagi, M., and Kato, I. (1998). Identification of the catalytic Accepted: May 25, 2004 triad residues of porcine liver acylamino acid-releasing enzyme. J. Published: August 10, 2004 Biochem. (Tokyo) 123,

178 Structure 1488 Miyagi, M., Sakiyama, F., and Kato, I. (1995). Complete covalent structure of porcine liver acylamino acid-releasing enzyme and identification of its active serine residue. J. Biochem. (Tokyo) 118, Naylor, S.L., Marshall, A., Hensel, C., Martinez, P.F., Holley, B., and Sakaguchi, A.Y. (1989). The DNF15S2 locus at 3p21 is transcribed in normal lung and small cell lung cancer. Genomics 4, Otwinowski, Z., and Minor, W. (1997). Processing of X-ray diffraction data collected in oscillation mode. In Macromolecular Crystallography, Part A, C.W. Carter, Jr., and R.M. Sweet, eds. (New York: Academic Press), pp Perrakis, A., Morris, R., and Lamzin, V.S. (1999). Automated protein model building combined with iterative structure refinement. Nat. Struct. Biol. 6, Rawlings, N.D., Polgar, L., and Barrett, A.J. (1991). A new family of serine-type peptidases related to prolyl oligopeptidase. Biochem. J. 279, Richards, P.G., Johnson, M.K., and Ray, D.E. (2000). Identification of acylpeptide hydrolase as a sensitive site for reaction with organophosphorus compounds and a potential target for cognitive enhancing drugs. Mol. Pharmacol. 58, Sako, Y., Nomura, N., and Uchida, A. (1996). A novel aerobic hyperthermophilic archaeon growing at temperature up to 100 degrees C. Int. J. Syst. Bacteriol. 46, Scaloni, A., Jones, W.M., and Barra, D. (1992). Acylpeptide hydrolase: inhibitors and some active residues of the human enzyme. J. Biol. Chem. 267, Scaloni, A., Barra, D., Jones, W.M., and Manning, J.M. (1994). Human acylpeptide hydrolase: studies on its thiol groups and mechanism of action. J. Biol. Chem. 269, Terwilliger, T.C. (2000). Maximum-likelihood density modification. Acta Crystallogr. D Biol. Crystallogr. 56, Tsunasawa, S., Narita, K., and Ogata, K. (1975). Purification and properties of acylamino acid-releasing enzyme from rat liver. J. Biochem. (Tokyo) 77, Wang, G.G., Gao, R.J., Ding, Y., Yang, H., Cao, S., Feng, Y., and Rao, Z. (2002). Crystallization and preliminary crystallographic analysis of acylamino acid releasing enzyme from hyperthermophilic archaeon Aeropyrum pernix. Acta Crystallogr. D Biol. Crystallogr. 58, Wang, G.G., Gao, R.J., Yang, H.T., Cao, S.G., Feng, Y., and Rao, Z. (2003). Archaeal acylamino acid releasing enzyme/lipase: crystallization and preliminary crystallographic analysis in a new crystal form. Chin. Sci. Bull. 48, Accession Numbers The coordinates and structure factors for apaph have been deposited in the Protein Data Bank with PDB accession code 1VE6. The coordinates and structure factors for the apaph-pnp complex have been deposited in the Protein Data Bank with PDB accession code 1VE7. 178

179 Short communication Antiviral Therapy 9:x-xx Probing the structure of the SARS coronavirus using scanning electron microscopy Yun Lin 1,2, Xiyun Yan 1*, Wuchun Cao 3, Chaoying Wang 4, Jing Feng 1,2, Jinzhu Duan 1,2 and Sishen Xie 4 1 National Laboratory of Biomacromolecules, Institute of Biophysics, Chinese Academy of Sciences (CAS), Beijing, China 2 State Key Laboratory of Microbial Resources, Institute of Microbiology, CAS, Beijing, China 3 Department of Epidemiology, Institute of Microbiology and Epidemiology, Beijing, China 4 Center for Condensed Matter Physics, Institute of Physics, CAS, Beijing, China *Corresponding author: Tel: ; Fax: ; yanxy@sun5.ibp.ac.cn A novel coronavirus, SARS-CoV, has been confirmed to be the aetiological agent of SARS. Transmission electron microscope (TEM) images played an important role in initial identification of the pathogen. In order to obtain greater morphological detail of SARS-CoV than could be obtained by TEM, we used ultra-high resolution scanning electron microscopy (SEM) to image the virus particles. We show here the three-dimensional appearance of SARS-CoV. Enhanced detail of the ultrastructure reveals the trimeric structure of the nm spikes on the virion surface. These results contribute to characterization of the SARS agent and development of new antiviral strategies. Introduction It has been confirmed by several research groups [1 4] that the worldwide outbreak of severe acute respiratory syndrome (SARS) is caused by a novel coronavirus, named SARS coronavirus (SARS-CoV). The ultrastructural features of coronaviruses, including SARS-CoV, have been established by transmission electron microscope (TEM) imaging [2]. However, little is known about the appearance of a complete virus particle and the ultrastructure of the virus surface. Here we provide the first observation of SARS-CoV particles by means of a dedicated ultra-high resolution scanning electron microscope (SEM) (HITACHI S- 5200, Japan). These data contribute to a more comprehensive understanding of SARS-CoV. Materials and methods Preparation of purified SARS virus Given the serious nature of SARS, all clinical specimens were handled in a biosafety level 3 laboratory. Serum specimens were inactivated by β-lactone before outside serological testing. SARS virus strain BJ01 [5] was obtained from a patient with SARS and established in the Vero E6 cell line. Culture supernatant was collected and centrifuged at low-speed to remove cells fragments. After sucrose-gradient ultracentrifugation, the purified virus pellet was resuspended in TNE (0.25 M NaCl, 0.02 M Tris-Cl ph 7.5, M EDTA) buffer. Enzyme-linked immunosorbent assay Inactivated virions were coated onto 96-well plates and incubated with either normal sera or convalescent sera from SARS patients, and then detected by HRP conjugated anti-human IgG. Positive and negative controls were from a diagnostic kit of antibodies for SARS virus (BGI-GBI Biotechnology Ltd, Beijing, China) and were independent of the sera we used. The cut off value was determined by the equation: cut off value=0.13+value negative control (OD450 nm). If the colour reaction represented a higher value than the cut off value, it was judged as positive. Sample preparation for scanning electron microscopy The purified viruses were submitted to the fixation and sputter coating procedure before viewing by SEM. Briefly, 3 5 mm cover slides or 200 mesh copper grids were coated with 1% poly-l-lysine (87 000) and then rinse with phosphate-buffered saline (PBS, ph 7.2) to remove excessive poly-lysine. Viruses with different dilution in TNE were dripped on cover slides or copper grids and let adhere for about 30 min. Excessive samples were sucked up and rinsed with PBS. After fixation in 2.5% glutaraldehyde for 30 min and a further 30 min in 1% osmium tetroxide, samples were dehydrated through an ethanol series in buffer: 50% 70% 90% 100% 100% for 5 min each and critically point dried from ethanol. Samples were mounted 2004 International Medical Press /02/$

180 Y Lin et al. Figure 1. Specific binding of the purified SARS-CoV BJ01 to convalescent sera from SARS patients OD450 nm on specimen stubs with conductive paint and coated to a thickness of 10 nm with Au in a sputter coater. Samples were viewed in lens on the Hitachi S-5200 SEM with slow scan mode, 10 kv accelerating voltage and 0 0.3mm working distance. Results 1:10 1:100 1:1000 Sample dilution Positive control Negative control In order to identify the biological characteristics of the purified virus, we first tested the reactivity of the virus with convalescent sera from SARS patients using an enzyme-linked immunosorbent assay (ELISA). Comparing with positive and negative controls, the purified SARS-CoV showed very strong binding to the convalescent sera but not to normal sera (Figure 1). This specific reaction indicated that the virus we purified from the culture supernatant of Vero E6 cells was SARS-CoV. To better understand the morphological detail of SARS-CoV, we used ultra-high resolution scanning electron microscopy to image the virus particles. Under the SEM, we observed large numbers of virus particles with predominantly nm in diameter either isolated or in aggregates (Figure 2). These SARS-CoV particles with a typical crown structure were further confirmed by using TEM (data not shown). Incidentally, we found a few virus particles as large as 400 nm in diameter, which had identical shape and surface substructure as the 200 nm particles and presented in a single field (Figure 2C). The virus particles appear round and full, with numerous tiny surface projections. Some are tightly adhered with their projections sticking into each other, forming a mosaic patch and leading to the compression of the virions. These flower-shaped projections corresponding to the spike, the main constituent of which was S protein, have a size ranging from 10 to 20 nm, in agreement with the reported diameter of the SARS virus spike. With further magnification of this part (Figure 2D, arrowheads), we observed the projections with a regular structure composed of three subunits, like a flower with three petals. Conclusion In this study, we provide the first three-dimensional structure of the causative agent in SARS infection and the detailed information about the ultrastructural surface of SARS-CoV by using SEM, which offers advantages over TEM for the study of biological molecules such as viruses, nucleic acid and proteins, due to its higher resolution and less disruptive nature of the technique. We observed all the surfaces of SARS virions bear flower-shaped projections that contain three S protein subunits. The S glycoprotein is known as a major immunogenic determinant of pathogenesis, binds to specific receptors and then undergo a temperature-dependent, receptor- mediated conformational change that leads to fusion of the viral envelope with host membranes to initiate infection. Delmas et al. demonstrated by in vitro experiment that coronavirus S protein tends to oligomerize into a trimer and the trimerization was a rate-limiting step after polypeptide synthesis [6]. Together with our structural observations of trimeric S protein on the surface of virion, we hypothesize that the identification of the receptor binding sites and the conformational change after S protein activation, as well as molecular interpretation of trimerization, may give rise to new strategies for development of novel antiviral drugs. It has been accepted that the size of coronaviruses are about nm observed by TEM. However, we found in the SEM images that the size of the SARS CoV BJ01 varied, most of them were about nm, a few up to 400 nm in diameter, which is significantly larger than the known nm size of SARS-CoV, even if the thickness of Au coating (20 nm) was added. We noticed that the larger virions had the same flower-shaped projections as those with normal size, and both particles in different sizes were seen in the same view. Although we are not able to prove these variations by further experiments at the moment because of the lack of SARS-CoV material, it remains an interesting question to be addressed in future. Taken together, our data vividly revealed the threedimensional appearance of SARS-CoV as well as their ultrastructural surfaces with three subunits-based projections. These results would enrich the morphological studies of the SARS-CoV and contribute to better understanding of SARS-CoV, with application to both basic virology and clinical practice International Medical Press 180

181 SARS-CoV SEM Figure 2. SARS-CoV observed under scanning electron microscope A B C D (A) Virions with diameter of 200 nm. (B) Virions with sizes of 100 and 200 nm. (C) Virions of 400 nm in diameter. (D) The ultrastructure of the surface projections. Two typical spikes are magnified to show the trimer structure (insets). Acknowledgements We acknowledge Jinzhu Zhang for advice on virus identification, Lixin Liu and Guangxia Gao for suggestions on virus purification, and Sarah Perrett for editing the manuscript. This work is supported by the National 973 grant specific for SARS prevention and therapy. References 1. Nicholls JM, Poon LL, Lee KC, Ng WF, Lai ST, Leung CY, Chu CM, Hui PK, Mak KL, Lim W, Yan KW, Chan KH, Tsang NC, Guan Y, Yuen KY & Peiris JS. Lung pathology of fatal severe acute respiratory syndrome. Lancet 2003; 361: Ksiazek TG, Erdman D & Goldsmith CSA. Novel coronavirus associated with severe acute respiratory syndrome. New England Journal of Medicine 2003; 348: Rota PA, Oberste MS, Monroe SS, Nix WA, Campagnoli R, Icenogle JP, Penaranda S, Bankamp B, Maher K, Chen MH, Tong S, Tamin A, Lowe L, Frace M, DeRisi JL, Chen Q, Wang D, Erdman DD, Peret TC, Burns C, Ksiazek TG, Rollin PE, Sanchez A, Liffick S, Holloway B, Limor J, McCaustland K, Olsen-Rasmussen M, Fouchier R, Gunther S, Osterhaus AD, Drosten C, Pallansch MA, Anderson LJ & Bellini WJ. Characterization of a novel coronavirus associated with severe acute respiratory syndrome. Science 2003; 300: Drosten C, Gunther S, Preiser W, van der Werf S, Brodt HR, Becker S, Rabenau H, Panning M, Kolesnikova L, Fouchier RA, Berger A, Burguiere AM, Cinatl J, Eickmann M, Escriou N, Grywna K, Kramme S, Manuguerra JC, Muller S, Rickerts V, Sturmer M, Vieth S, Klenk HD, Osterhaus AD, Schmitz H & Doerr HW. Identification of a novel coronavirus in patients with severe acute respiratory syndrome. New England Journal of Medicine 2003; 348: Qin Ede, Zhu Qingyu, Yu Man et al. A complete sequence and comparative analysis of a SARS-associated virus (Isolate BJ01). Chinese Science Bulletin. 2003; 48: Delmas B & Laude H. Assembly of coronavirus spike protein into trimers and its role in epitope expression. Journal of Virology 1990; 64: Received 14 August 2003, accepted 25 November 2003 Antiviral Therapy 9:

182 doi: /j.jmb J. Mol. Biol. (2004) 339, Unique Structural Characteristics of Peptide Deformylase from Pathogenic Bacterium Leptospira interrogans Zhaocai Zhou 1,2, Xiaomin Song 1,2, Yikun Li 1,2 and Weimin Gong 1,2 * 1 National Laboratory of Biomacromolecules, Institute of Biophysics, Chinese Academy of Sciences, Beijing, People s Republic of China 2 School of Life Sciences, Key Laboratory of Structural Biology, University of Science and Technology of China, Hefei Anhui, , People s Republic of China *Corresponding author Peptide deformylase (PDF), which is essential for normal growth of bacteria but not for higher organisms, is explored as an attractive target for developing novel antibiotics. Here, we present the crystal structure of Leptospira interrogans PDF (Li PDF) at 2.2 Å resolution. To our knowledge, this is the first crystal structure of PDF associating in a stable dimer. The key loop (named the CD-loop: amino acid residues 66 76) near the active-site pocket adopts closed or open conformations in the two monomers forming the dimer. In the closed subunit, the CD-loop and residue Arg109 block the entry of the substrate-binding pocket, while the active-site pocket of the open subunit is occupied by the C-terminal tail from the neighbouring molecule. Moreover, a formate group, as one product of deformylisation, is observed bound with the active-site zinc ion. Li PDF displays significant structural differences in the C-terminal region compared to both type-i and type-ii PDFs, suggesting a new family of PDFs. q 2004 Elsevier Ltd. All rights reserved. Keywords: peptide deformylase; dimerization; active-site pocket; conformational change; formate group Introduction Bacterial protein synthesis begins with a formylated methionine residue. 1,2 Following translation initiation, peptide deformylase (PDF) catalyzes the hydrolytic removal of the N-terminal fmet residue of most nascent polypeptides in eubacteria and the organelles of certain eukaryotes. 3 9 Although N-formylation is not essential for the survival of all bacterial species, it can stimulate protein synthesis by facilitating the participation of MettRNA fmet in translation initiation and by preventing its recognition by the elongation apparatus. 10,11 However, the first step in N-terminal processing by PDF is essential for the bacterial survival, because mature proteins do not retain N-formylmethionine and all known N-terminal peptidases cannot utilize formylated peptides as substrates. This formylation/deformylisation cycle is apparently unique to eubacteria and is not utilized in eukaryotic cytosolic protein Abbreviations used: PDF, peptide deformylase; Li, Leptospira interrogans; Ec, Escherichia coli. address of the corresponding author: wgong@sun5.ibp.ac.cn biosynthesis. 2 Thus, inhibition of PDF would halt bacterial growth and spare host cell function. In fact, as an attractive target for the development of novel antibiotics, the peptide deformylase is currently studied extensively, not only regarding catalytic properties, but also from a three-dimensional structure perspective. It was long believed that Escherichia coli PDF was a zinc enzyme until it was characterized as a ferrous native enzyme with extraordinary lability caused by oxygen-mediated inactivation The ferrous can be replaced by a nickel ion without significant loss of catalytic efficiency, whereas the zinc form, prepared from the apo-enzyme or by displacement of ferrous or nickel ion, proved to be virtually inactive. However, an eukaryotic PDF from Arabidopsis thaliana seedlings (AtPDF1A) was recently characterized by Serero et al. 27 to be a zinc enzyme with full activity. To date, attempts to explain the enormous activity difference between ferrous PDF and zinc PDF from a structure analysis of a metal centre have not been successful. 23 Phylogenetic tree analysis and systematic sequence alignment classify available PDF structures into type-i and type- II. 25,28 Most of type-i PDFs, represented by E. coli deformylase, belong to Gram-negative organisms /$ - see front matter q 2004 Elsevier Ltd. All rights reserved. 182

183 208 Unique Structural Characteristics of Zinc LiPDF and have, 30% sequence identity with each other; whereas type-ii PDFs, represented by Staphylococcus aureus deformylase, come from a distinct group of Gram-positive organisms. Compared to type-i PDFs, type-ii PDFs are characterized by three sequence insertions and a hydrophobic C terminus. 28 As a ubiquitous environmental bacterium, Leptospira can cause strong leptospirosis infection of animal and human by entering the host through mucosa and broken skin. The resultant bacteria is frequently found as the complication after operation. Most of the related pathogens show drug tolerance to conventional antibiotics, making it an urgent need to develop effective and novel drugs. Here we analysed the crystal structure of peptide deformylase from Gram-negative bacterium pathogen Leptospira interrogans (Li PDF) in the zinc bound form. Interestingly, Li PDF was observed to be a dimer in crystals although Kumar et al. 26 have seen complex oligomers in crystals of Pseudomonas falciparum PDF, 26 consistent with its stable biological association in solution. 29 Such a dimer has not been reported in other PDF structures. Unexpectedly, unique closed and open substrate-binding pockets were formed, respectively, in the two monomers. More importantly, the active-site zinc ion was observed to coordinate with one of its catalytic reaction products, a formate group, instead of a water molecule as in other PDF structures. Besides, the key loop around the active cleft (CD-loop) and the C terminus exhibit great differences compared with other PDFs, suggesting that Li PDF belongs to a new PDF family. Results and Discussion Overall structure and dimer association of Li PDF Similar to other PDF structures, the core structure of Li PDF contains two a-helices, seven b- strands and three 3 10 helices (Figure 1). As highly conserved structural elements, these three 3 10 helices comprise the bottom of the active-site hydrophobic pocket and help to build the internal microenvironment required for catalysis. The first conserved 3 10 helix, which plays a role in sealing up the back of the active-site pocket, is embedded in the middle of the coil preceding the N-terminal aa helix. Located on the BC loop, that is flanked with the first and the second strands of the N-terminal three-stranded antiparallel b-sheet, is the second 3 10 helix, which provides Gln54 as a key residue from the first conserved box for the indirect coordination of the zinc ion (Figure 1). The central ai helix forms the core of the PDF structure and provides two residues, His144 and His148, for the coordination of the active-site metal ion. Residue Leu103, acting as a Lewis acid for the polarisation of the formyl carbonyl group in the process of substrate cleavage, together with three other key residues Gly49, Gln54 and Glu145 within the active centre remains conserved. Consistent with the result of size exclusion column chromatography experiments, 29 the asymmetric unit of Li PDF crystals consist of two subunits (named A and B) forming a dimer, related by a non-crystallographic 2-fold axis. To our knowledge, this is the first crystal structure of a dimerized PDF (although P. falciparum PDF exhibited a tendency to form dimers; Kumar et al. 26 ). Each of the dimers is part of a tetramer sitting on a crystallographic 2-fold axis. Detailed examinations of the four monomers (monomer A, B, A 0 and B 0 ; A 0 and B 0 are symmetry related molecules of A and B, respectively) in the tetramer definitely revealed that electrostatic interactions predominate on the interface between the crystallographic 2-fold related subunits (A A 0 or B B 0, buried surface,800 Å 2 ). Contrarily, quite a few hydrophobic residues contribute to the interactions between the non-crystallographic symmetry related subunits (A B or A 0 B 0 ) and they are most likely to compose the biological dimer (Figure 2). The dimeric association is primarily mediated by the hydrophobic interactions of strands be, bh of one subunit and their counterparts of the other subunit, mostly contributed by Phe164, Phe166, Met108 and Met9. The solvent-accessible surface area buried between the subunits is approximately 10% of the surface area of one subunit. Whether such dimerisation is of any significance to the catalysis mechanism remains to be elucidated. The unique C terminus With five X-ray structures of PDFs from different origins currently available, 23,25,26,28,30,31 we have performed a detailed comparison of Li PDF with type-i and type-ii deformylases. Some important structural differences have been revealed ranging from topology to the substrate-binding pocket. Although the overall structure of Li PDF is similar to other PDFs, sequence alignments (Figure 1) and structural comparison clearly distinguished Li PDF from either type-i or type-ii PDFs. Li PDF apparently belongs to the type-i family based on its N-terminal character. However, the C-terminal part shows little sequence or structural homology to either type-i or type-ii PDFs and is even much more hydrophobic than type-ii PDFs. Type-I PDFs possess a long C-terminal a-helix, which is parallel to bf. 16,18,19 In contrast, the C terminus of type-ii PDF turns back to form a mixed b-sheet with be and bf. 14 However, the C terminus of Li PDF, consisting of a short b-strand (bj) and an a-helix (ak), heads to the direction between type-i and type-ii PDFs (Figure 3). bj joins the C-terminal antiparallel b-sheet, interacting with bf, instead of be as in type-ii PDFs. Both molecules in the asymmetric unit have the same C-terminal conformation. But their last four tail residues go in different directions mostly due to different packing environments. PDF was 183

184 Unique Structural Characteristics of Zinc LiPDF 209 Figure 1. Structure based sequence alignment of Li PDF with type-i and type-ii PDFs. This alignment was basically performed with program CLUSTALW 39 and redressed according to the result of structure overlapping by LSQMAN. The sequence labels are shown on the left in each block. Type-I (E. coli, P. aeruginosa, P. falciparum) and type-ii (S. aureus, Bacillus stearothermophilus) peptide deformylases are indicated in green and orange, respectively, while Li PDF is highlighted in blue. Three identified motifs have been boxed with strictly conserved residues in dark blue. Striking sequence variations are observed in a loop region near the active pocket (namely CD-loop in Li PDF). The secondary structural elements of Li PDF determined by PROMOTIF are drawn above the sequences and labelled as for Figure 4 The lines connecting to the small ball in purple indicate those residues coordinating with the zinc ion in Li PDF. Note: (*) identity; (:) strongly similar; (.) weakly similar. Three conserved 3 10 helices are presented in pentagons. thought to act while bound to the ribosome 3,32,33 and its C-terminal part was proposed to serve to anchor the enzyme to the ribonucleoproteic complex for the proximity of substrate. 34 This is consistent with the variable conformations observed in PDF s C terminus. Kleywegt, G. J.; LSQMAN v Department of Cell and Molecular Biology, Uppsala University, Biomedical Centre, Box 596, SE Uppsala, Sweden. Hutchinson, E. G. & Thornton, J. M.; PROMOTIF v2.0; Biomolecular Structure and Modelling Unit, Department of Biochemistry and Molecular Biology, University College, Gower Street, London WC1E 6BT, UK. Two conformations of the CD loop Besides a unique C terminus, a striking conformational change of the CD-loop (from Ser66 to Pro76 between bc and bd) adjacent to the active centre was observed. Compared to type-i and type-ii PDFs, Li PDF has the longest CD-loop (Figure 1). The structures of the two subunits are essentially identical with the r.m.s. deviation between the corresponding C a atoms to be 0.34 Å except for the CD-loop and the last four C-terminal residues. The CD-loops of type-i and type-ii PDFs, although adopting different conformations due to different lengths, are both positioned to make the active-site pocket open for substrate binding (Figure 4(a) and (b)). Nevertheless, utterly different 184

185 210 Unique Structural Characteristics of Zinc LiPDF Figure 2. Dimerization of Li PDF. Interactions between NCS related molecules in the asymmetric unit that form a stable dimer. Different conformations (open and closed) of the substrate-binding pocket were observed in these two subunits, which result in the non-symmetric appearance of the dimer. The C terminus is also related with a different packing environment and is in keeping with the case of the substrate-binding pocket. Residues forming the hydrophobic core (Phe164, Phe166, Met108 and Met9) are rendered in ball-and-stick format. The active-site zinc ion is highlighted with a purple ball. This Figure and Figures 3, 4, 5 and 6 were generated by MOLSCRIPT 40 and rendered by Raster3D. 41 from them, the corresponding CD-loop of Li PDF was observed to have two distinct conformations in the two subunits. In one subunit, the CD-loop is located with the active-site pocket open similar to other PDFs (this subunit is named as the open form; Figure 4(c)). In the other subunit, surprisingly, the CD-loop markedly turns towards the active-site crevice as a closing cover (this subunit is therefore named as the closed form; Figure 4(d)). The side-chains of Arg71 and Tyr72 (on the tip of CD-loop), which are hydrogen bonded with each other, protrude deeply into the active-site cleft Figure 4. Comparison of CD-loop conformations. Green, CD-loop; brown, C-terminal tail; white, the additional part. EcPDF and SaPDF represent type-i and type-ii peptide deformylase, respectively. a, Ec PDF; b, Sa PDF; c, open form of Li PDF; d, closed form of Li PDF. (Figure 5(a)). The side-chain of Arg109 shifts inside, forming a hydrogen bridge with Arg71(O), modifying the electrostatic and hydrophobic nature of the substrate-binding cleft. Thus, the CD-loop, together with Arg109, blocks the activesite entrance, which is critical for substrate/ inhibitor binding. If the natural inhibitor of PDF, actinonin, is modeled into this closed form, serious collisions are observed between the Tyr72 sidechain and the side-chains at P2 0 and P3 0 of the inhibitor, suggesting that the CD-loop lid must be opened before substrate molecules access the catalytic site. No structural equivalents of Arg71 and Tyr72 were observed in other PDFs structures. The fact that the CD-loop in the closed subunit is completely solvent-exposed and not involved in crystal packing, indicates that the closed form would exist naturally, whereas, in the open form, a neighbouring molecule related by a crystallographic 2-fold axis protrudes its C-terminal tail into the open active pocket (Figure 5(b)). Metal binding site Figure 3. Overlaping of C a atoms of Li PDF with type-i (E. coli) and type-ii (S. aureus) PDFs. White, E. coli PDF (Ec PDF); green, S. aureus PDF (Sa PDF); yellow, Li PDF. This Figure was calculated by Lsqman and checked by program O. The sequence alignment in Figure 1 is referred to as such a kind of structure overlapping. Li PDF, which was purified and crystallized with no extra metal ions added, was determined by atomic absorption spectroscopy to contain one zinc atom per monomer. The activity of Li PDF 29 is much higher than that of E. coli zinc-pdf 24 but much lower than that of E. coli Fe(II)-PDF. 20 It is still unclear whether such differences are related 185

186 Unique Structural Characteristics of Zinc LiPDF 211 Figure 5. Stereo view of closed and open subunits of Li PDF. The protein skeleton is shown using a coil model with the active-site pocket in green and the CD-loop in black. The small purple sphere represents the active-site zinc ion. (a) The closed subunit: CD-loop residues Arg71 and Tyr72 protrude into the pocket. These two residues, together with Arg109, block the entry of the pocket. The glycine residue in the S 0 pocket is represented in a bond model. (b) The open subunit: the CD-loop and Arg109 withdraw from the closed position. The symmetry related neighbouring tail is highlighted in pink. Asterisks represent residues on this neighbouring tail. The simulated product MAS was modelled by superimposing E. coli PDF (PDB entry ID 1BS8) to Li PDF. to the special dimer association of Li PDF. Comparison of the metal coordination with available structures reveals a perfectly conserved catalytic core, supporting the view that all PDFs adopt the same mechanism. Three residues, His144, His148, Cys102, are absolutely conserved and involved in the coordination with the metal ion. A water molecule is generally present as the fourth ligand. However, the metal-coordinating water molecule universally observed in other PDF structures is replaced with a well-defined formate group (from crystallization solution) in the structure of Li PDF, leading to the five-coordination of the zinc ion. This formate group, one of the products of the deformylisation reaction, provides two formate oxygen atoms as coordinating ligands with one of them binding to the zinc ion weakly (2.8 Å) (Figure 6). Besides, one oxygen atom of the formate group also hydrogen bonds to the proton shuttle residue Glu145 and the other oxygen atom to the 186

187 212 Unique Structural Characteristics of Zinc LiPDF Figure 6. Electron map of the metal binding site. (a) If one water molecule was filled; (b) if two water molecules were filled; (c) if none was filled, the formate group is present here for the purpose of comparison; (d) a formate group was filled. Notes: green: 2Fo 2 Fc map; red: Fo 2 Fc map with a 3 sigma cut-off. The electron density of this local area is essentially identical for the two subunits in the asymmetric unit. main-chain nitrogen atom of Leu103 and Gln54 NE2. Consistent with previous reports,30,31 the side-chain of Gln54 is not only hydrogen bonded with the ligand water molecule at ground state, but also involved in binding the formyl group of the substrate during the catalytic process. At the opposite side of the formate group, a highly conserved three-molecule water chain is observed, contributing to the stabilization of the active centre. The first water molecule (nearest to the active-site zinc ion) interacts with Gln154 with a hydrogen bond. The middle water molecule hydrogen bonds to the carbonyl oxygen atom of His144. The third water molecule hydrogen bonds to the side-chains of Arg114 and Asp147. These conserved residues and water molecules, together with Glu100 and Arg157, create an extensive hydrogen bond network, which maybe partially contributes to the stability of catalytic intermediates. Concerning the catalytic cycle, Becker et al.23 proposed a five-step mechanism in which the cleaved formate group from the substrate binds to the active-site metal ion, forming an intermediate complex during the last two steps. Our structure essentially supports such a catalytic model. Superimposing of the Li PDF structure with E. coli deformylase (Ec PDF) complexed with its natural inhibitor actinonin (PDB entry: 1LRU) reveals that the formate group of Li PDF assumes almost the same spatial position as the pseudo-formyl end of the actinonin (r.m.s.d.: 1.2 A ). In fact, our observation corresponds to the enzyme formate complex, providing the first direct evidence for the states in the last two steps. Although one coordinating water molecule (W1) is surely necessary for the ground state of PDF, detailed checking of models available revealed some spatial variations of this water molecule. It appears that the second water molecule (W2) is not always present. Substrate binding pocket Residues Ile86 and Leu125 in Ec PDF are replaced with Phe98 and Tyr137 in Li PDF, respectively, with their bigger side-chains contributing to 187

188 Unique Structural Characteristics of Zinc LiPDF 213 a narrower substrate entrance. The resultant convergence effect to the pocket has serious consequences for substrate/inhibitor binding, which is a possible factor for the relative low catalytic efficiency of Li PDF compared with nickel-containing Ec PDF. 17 However, Li PDF exhibits much higher activity compared with zinc bound Ec PDF. 29 No direct explanation for the fact that PDF activity is depending on the bound metal has been obtained from the present structure. In addition, a glycine molecule from the crystallization buffer solution is observed in the bottom of the S1 0 pocket of Li PDF, along the direction of P1 0 side-chain. This indicates that the S1 0 pocket of Li PDF is, at least, not strictly hydrophobic, in line with the fact that Tyr137 projects its side-chain hydroxyl group to form the bottom of S1 0. In the crystal packing of Li PDF, the C-terminal tail of one neighbouring molecule is docked snugly into the active-site pocket of the open form, reminding us that an active site is occupied by the simulated product Met-Ala-Ser in the complex structure (PDB entry: 1BS8) (Figure 5(b)). The second rearmost residue Leu177* (the asterisks represent residues on this neighbouring C-terminal tail) extends its side-chain into the S1 0 hydrophobic pocket with the side-chain of Asp178* directing to the position of P3 0. It seems that such a C-terminal tail mimics a real substrate peptide normally much longer than the simulated three amino acid residue product. Interestingly, the part of the C-terminal tail outside the pocket is braced by a CD-loop (through interactions between Leu163* and Pro73, His174* and Tyr72 in Li PDF), apparently facilitating the proper anchoring of the substrate to the active-site. To date, no description on the full-length substrate binding has been reported and our observations provide a possible model for the naturally occurred enzyme substrate interactions. To conclude, this study shows the first structure of a dimerized peptide deformylase and reveals some unique characters of Li PDF distinct from type-i and type-ii PDFs, suggesting a novel family of PDFs. Instead of the well-accepted open substrate-binding cleft, a striking closed pocket is observed. Besides, important differences are revealed regarding the catalytic site. For the first time, a formate group as one of the reaction products is directly observed to coordinate with the metal ion. Moreover, the C-terminal tail of a neighbouring molecule is observed to occupy the active-site pocket of the open form subunit. Materials and Methods Purification and crystallization Isolation and purification were carried out basically following the protocol described. 29 Additional steps were mainly towards the application of FPLC S-100 and Mono-Q columns, which greatly improved the purity of the crystallizing sample. Crystals were grown by vapour-diffusion at 277 K by the hanging-drop method. Good crystals were obtained within four days under the condition of 4 M sodium formate, Gly HCl buffer (ph 3.0). Interestingly, the optimized crystal is very small but possesses excellent diffraction quality. Data collection and processing Data were collected on a MAR Research image-plate system with a local X-ray source at 100 K. The wavelength was Å and the exposure time was five minutes per image. The oscillation step for each image was 18 The crystal-to-detector distance was set to 150 mm. The data were processed with DENZO and SCALE- PACK. 27 The space group was determined to be P with which the correct molecular replacement solution was obtained. Details of data collection and processing statistics are shown in Table 1. Structure determination and refinement Phases were obtained by molecular replacement with the AMoRe package. 35 The structure of P. aeruginosa PDF (PDB code: 1LRY) was taken as an initial search model over the resolution range Å. A weak solution with two molecules in one asymmetric unit was obtained in space group P Remarkably, the C-terminal peptides of the two subunits in this solution were found to seriously involve each other when checked with program O. 36 Then the C terminus was cut and nonconserved amino acid residues except glycine were set to alanine. The subsequent refinement was performed with CNS, 37 using standard protocols and solvent correction. After rigid body and energy minimization refinements, manual rebuilding of the model including the C-terminal peptide was carried out with program O. Twofold NCS redundancy was exploited with about 90% atomic coordinates being restrained to obey noncrystallographic symmetry. After several cycles of local rebuilding and refinement the electron density appeared to be quite good. The zinc atom and the coordinating formate group, as well as the glycine molecule in the S1 0 pocket were incorporated into the refined model by direct examination of F o 2 F c map. The R-factor of the final model is ðr free ¼ 0:236Þ (Table 1). Statistics Table 1. Statistics of data collection and refinement Space group P Unit cell parameters a ¼ b ¼ Å, c ¼ Å Resolution range (Å) (last shell) ( ) No. total reflections 178,754 No. independent reflections 22,512 Completness (%) 98.1(99.6) R merge 0.103(0.514) I/s (I) 14.49(3.0) Percentage of reflections used for 8 R-free No. protein atoms 2807 (occupancy. 0) No. ligand atoms 6 No. metal atoms 2 No. solvent molecules 521 R-value (last shell) 0.187(0.288) R-free (last shell) 0.236(0.417) r.m.s.d. of bond length (Å) r.m.s.d. of bond angle (deg.)

189 214 Unique Structural Characteristics of Zinc LiPDF generated by PROCHECK 38 showed that the model quality is beyond the normal standard. Protein Data Bank accession code The atomic coordinates of the refined model of Li PDF have been deposited in RCSB Protein Data Bank under the accession number 1RN5. Acknowledgements This work is supported by the Foundation for Authors of National Excellent Doctoral Dissertation of P.R. China (project number ), National Foundation of Talent Youth (grant number ), the National High Technology Research and Development Program of China (grant number 2001AA233021), the 863 Special Program of China (grant number 2002BA711A13), the Key Important Project and other projects from the National Natural Science Foundation of China (grant numbers , , and ) and Chinese Academy of Sciences (grant number KSCX1-SW-17). References 1. Meinnel, T., Mechulam, Y. & Blanquet, S. (1993). Methionine as translation start signal: a review of the enzymes of the pathway in Escherichia coli. Biochimie, 75, Kozak, M. (1983). Comparison of initiation of protein synthesis in procaryotes, eucaryotes, and organelles. Microbiol. Rev. 47, Adams, J. M. (1968). On the release of the formyl group from nascent protein. J. Mol. Biol. 33, Livingston, D. M. & Leder, P. (1969). Deformylation and protein synthesis. Biochemistry, 8, Takeda, M. & Webster, R. E. (1968). Protein chain initiation and deformylation in B. subtilis homogenates. Proc. Natl Acad. Sci. USA, 60, Adams, J. M. & Capecchi, M. R. (1966). N-formylmethionyl-sRNA as the initiator of protein synthesis. Proc. Natl Acad. Sci. USA, 55, Bradshaw, R. A., Brichey, W. W. & Walker, K. W. (1998). N-terminal processing: the methionine aminopeptidase and N alpha-acetyl transferase families. Trends Biochem. Sci. 23, Mazel, D., Pocher, S. & Marliere, P. (1994). Genetic characterization of polypeptide deformylase, a distinctive enzyme of eubacterial translation. EMBO J. 13, Giglione, C., Pierre, M. & Meinnel, T. (2000). Peptide deformylase as a target for new generation, broad spectrum antimicrobial agents. Mol. Microbiol. 36, Guillon, J. M., Heiss, S., Soutourina, J., Mechulam, Y., Laalami, S., Grunberg-Manago, M. & Blanquet, S. (1996). Interplay of methionine trnas with translation elongation factor Tu and translation initiation factor 2 in Escherichia coli. J. Biol. Chem. 271, Meinnel, T. & Blanquet, S. (1995). Enzymatic properties of Escherichia coli peptide deformylase. J. Bacteriol. 177, Schmitt, E., Guillon, J. M., Meinnel, T., Mechulam, Y., Dardel, F. & Blanquet, S. (1996). Molecular recognition governing the initiation of translation in Escherichia coli. Biochimie (Paris), 78, Hirel, P. H., Schmitter, M. J., Dessen, P., Fayat, G. & Blanquet, S. (1989). Extent of N-terminal methionine excision from Escherichia coli proteins is governed by the side-chain length of the penultimate amino acid. Proc. Natl Acad. Sci. USA, 86, Dalboge, H., Bayne, S. & Pedersen, J. (1990). In vivo processing of N-terminal methionine in E. coli. FEBS Letters, 266, Varshavsky, A. (1997). The N-end rule pathway of protein degradation. Genes Cells, 2, Solbiati, J., Chapman-Smith, A., Miller, J. L., Miller, C. G. & Cronan, J. E. (1999). Processing of the N termini of nascent polypeptide chains requires deformylation prior to methionine removal. J. Mol. Biol. 290, Tso, J. Y., Hermodson, M. A. & Zalkin, H. (1982). Glutamine phosphoriboxylpyrophosphate amidotransferase from cloned Escherichia coli purf. NH2- terminal amino acid sequence, identification of the glutamine site, and trace metal analysis. J. Biol. Chem. 257, Persson, B., Flinta, C., von Heijne, G. & Jornvall, H. (1985). Structures of N-terminally acetylated proteins. Eur. J. Biochem. 152, Arai, K., Clark, B. F., Duffy, L., Jones, M. D., Kaziro, Y., Laursen, R. A. et al. (1980). Primary structure of elongation factor Tu from Escherichia coli. Proc. Natl Acad. Sci. USA, 77, Groche, D. et al. (1998). Isolation and crystallization of functionally competent Escherichia coli peptide deformylase forms containing either iron or nickel in the active site. Biochem. Biophys. Res. Commun. 246, Rajagopalan, P. T. et al. (1997). Peptide deformylase: a new type of mononuclear iron protein. J. Am. Chem. Soc. 119, Rajagopalan, P. T. & Pei, D. (1998). Oxygen-mediated inactivation of peptide deformylase. J. Biol. Chem. 273, Becker, A., Schlichting, I., Kabsch, W., Groche, D., Schultz, S. & Volker Wagner, A. F. (1998). Iron center, substrate recognition and mechanism of peptide deformylase. Nature Struct. Biol. 5, Ragusa, S., Blanquet, S. & Meinnel, T. (1998). Control of peptide deformylase activity by metal cations. J. Mol. Biol. 280, Baldwin, E. T., Harris, M. S., Yem, A. W., Wolfe, C. L., Vosters, A. F. et al. (2002). Crystal structure of type II peptide deformylase from Staphylococcus aureus. J. Biol. Chem. 277, Kumar, A., Nguyen, K. T., Srivathsan, S., Ornstein, B., Turley, S., Hirsh, I. et al. (2002). Crystals of peptide deformylase from Plasmodium falciparum reveal critical characteristics of the active site for drug design. Structure, 10, Serero, A., Giglione, C. & Meinnel, T. (2001). Distinctive features of the two classes of eukaryotic peptide deformylases. J. Mol. Biol. 314, Guilloteau, J. P., Mathieu, M., Giglione, C., Blanc, V., Dupuy, A., Chevrier, M. et al. (2002). The crystal structures of four peptide deformylases bound to the antibiotic actinonin reveal two distinct types: a 189

190 Unique Structural Characteristics of Zinc LiPDF 215 platform for the structure-based design of antibacterial agents. J. Mol. Biol. 320, Li, Y., Chen, Z. & Gong, W. (2002). Enzymatic properties of a new peptide deformylase from pathogenic bacterium Leptospira interrogans. Biochem. Biophys. Res. Commun. 295, Chan, M. K., Gong, W., Rajagopalan, P. T., Hao, B., Tsai, C. M. & Pei, D. (1997). Crystal structure of the Escherichia coli peptide deformylase. Biochemistry, 36, Hao, B., Gong, W., Rajagopalan, P. T., Zhou, Y., Pei, D. & Chan, M. K. (1999). Structural basis for the design of antibiotics targeting peptide deformylase. Biochemistry, 38, Livingston, D. M. & Leder, P. (1969). Deformylation and protein biosynthesis. Biochemistry, 8, Pine, M. J. (1969). Kinetics of maturation of the amino termini of the cell proteins of Escherichia coli. Biochim. Biophys. Acta, 174, Schmitt, E., Guillon, J. M., Meinnel, T., Mechulam, Y., Dardel, F. & Blanquet, S. (1996). Molecular recognition governing the initiation of translation in Escherichia coli. Rev. Biochimie. 78, Navaza, J. (1994). AMORE: an automated procedure for molecular replacement. Acta Crystallog. sect. A, 50, Jones, T., Zou, J., Cowan, S. & Kjeldgaard, M. (1991). Improved methods for building protein models in electron density maps and the location of errors in these models. Acta Crystallog. sect. A, 47, Brunger, A. T., Adams, P. D., Clore, G. M., DeLano, W. L., Gros, P., Grosse-Kunstleve, R. W. et al. (1998). Crystallography & NMR system: A new software suite for macromolecular structure determination. Acta Crystallog. sect. D, 54, Laskowski, R. A., MacArthur, M. W., Moss, D. S. & Thornton, J. M. (1993). PROCHECK: a program to check the stereochemical quality of protein structures. J. Appl. Crystallog. 26, Thompson, J. D., Higgins, D. G. & Gibson, T. J. (1994). CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucl. Acids Res. 22, Kraulis, P. (1991). MolScript, a program to produce both detailed and schematic plots of protein structures. J. Appl. Crystallog. 24, Merritt, E. A. & Bacon, D. J. (1997). Raster3D-Photorealistic molecular graphics. Methods Enzymol. 277, Edited by I. Wilson (Received 29 December 2003; received in revised form 16 March 2004; accepted 16 March 2004) 190

191 doi: /j.jmb J. Mol. Biol. (2004) 344, COMMUNICATION Crystal Structure of Human Coactosin-like Protein Lin Liu 1,2, Zhiyi Wei 1,2, Yanli Wang 1,2, Mao Wan 2, Zhongjun Cheng 1,2 and Weimin Gong 1,2 * 1 National Laboratory of Biomacromolecules, Institute of Biophysics, Chinese Academy of Sciences, Beijing People s Republic of China 2 School of Life Sciences, Key Laboratory of Structural Biology, University of Science and Technology of China Hefei, Anhui , People s Republic of China *Corresponding author Human coactosin-like protein is an actin filament binding protein but does not bind to globular actin. It associates with 5-Lipoxygenase both in vivo and in vitro, playing important roles in modulating the activities of actin and 5-Lipoxygenase. Coactosin counteracts the capping activity of capping protein which inhibits the actin polymerization. We determined the crystal structures of human coactosin-like protein by multi-wavelength anomalous dispersion method. The structure showed a high level of similarity to ADF-H domain, although their amino acid sequences share low degree of homology. A few conserved hydrophobic residues that may contribute to the folding were identified. This structure suggests coactosin-like protein bind to F-actin in a different way from ADF/Cofilin family. Combined with the information from previous mutagenesis studies, the binding sites for F-actin and 5-Lipoxygenase were analyzed, respectively. These two sites are quite close, which might prevent F-actin and 5-Lipoxygenase from binding to coactosin simultaneously. q 2004 Elsevier Ltd. All rights reserved. Keywords: human coactosin-like protein; crystal structure; F-actin; 5-Lipoxygenase; capping protein Actin plays important roles in cell architecture, motility, phagocytosis, endocytosis and cytoplasmic streaming. 1 Its functions are modulated by a large number of actin-binding proteins (ABPs). The structures of actin and many of its binding proteins have been determined to high resolution using X-ray crystallography and NMR spectroscopy, whereas electron microscopy and image processing have established the interaction sites on F-actin for myosin and a range of actin-binding proteins. 2 Coactosin is a 17 kda actin-binding protein originally isolated from Dictyostelium discoideum. 3 It is revealed that coactosin is able to counteract the activity of capping proteins that retard actin polymerization, while coactosin itself has no effect on actin polymerization. 4 The human version of coactosin named coactosin-like protein (CLP) shows a significant homology to coactosin with 33.3% identity and 75% homology in amino acid sequence. Human CLP nucleotide L.L. & Z.W. contributed equally to this work. Abbreviations used: ABPs, actin-binding proteins; CLP, coactosin-like protein; SMS, Smith Magenis syndrome; MAD, multi-wavelength anomalous dispersion; PDB, Protein Data Bank; rmsd, root-mean-square deviation. address of the corresponding author: wgong@sun5.ibp.ac.cn sequence was initially found as a sequence flanking a deletion on chromosome 17 characterizing the Smith Magenis syndrome (SMS). The SMS critical region overlaps with a breakpoint cluster region associated with primitive neuro-ectodermal tumors, suggesting that the CLP gene is involved in DNA rearrangements of somatic cells. 5 CLP is also reported as a human pancreatic cancer antigen by SEREX method. 6 CLP binds directly to filamentous-actin (F-actin) but does not form a stable complex with globular actin (G-actin). CLP binds to actin filaments with a stoichiometry of 1 : 2 (CLP: actin subunits), but could be cross-linked to only one subunit of actin. 7 Human CLP was first obtained in a yeast twohybrid screen using 5-Lipoxygenase (5LO), also an actin-binding protein, as a bait. 8 5LO is a 78 kda enzyme, the first enzyme in cellular leukotriene biosynthesis. 5LO catalyzes two-step conversion of arachidonic acids to leukotrienes, which are potent mediator of inflammation of allergy disorders including arthritis, asthma, and allergic reactions. In resting cells, 5LO is localized in soluble compartments, in the cytosol and/or within the nucleus. Upon activation, 5LO becomes associated with the nuclear membrane. The migration of 5LO is probably of the most importance for regulation of the cellular 5LO activity. 9 Modulation of /$ - see front matter q 2004 Elsevier Ltd. All rights reserved. 191

192 318 Crystal Structure of Human Coactosin-like Protein translocation and activation of 5LO may involve interactions with other proteins. Structure overview The crystal structure of human CLP is the first structure of coactosin/clp family. The structure was solved by multi-wavelength anomalous dispersion (MAD) method using a seleno-methionine substituted protein crystal, and refined at 2.8 Å resolution (Table 1). There is only one methionine (Met70) in the recombinant human CLP protein. Mass spectroscopy experiment proved that the first Met was removed when expressed in Escherichia coli (data not shown). To our knowledge, this is one of the structures solved using MAD method with the most number of residues (148 residues tagged with six His residues in the C-terminal end) per selenium atom. The final models contain two monomers in the asymmetric unit. The C-terminal region seems very disordered, with no electron density fitting for residues in chain A, and residues in chain B, respectively. The monomer structure consists of a six-stranded mixed b-sheet in which the four central strands (b2-b5) are anti-parallel and the two edge strands (b1 and b6) run parallel with the neighboring strands. The sheet is surrounded by two a-helices on each side (Figure 1). Such a structure is a typical ADF-H domain, 10 which is the core structure of many actin-binding proteins. The two monomers are almost structurally identical with an root-meansquare deviation (rmsd) of 0.65 Å for 130 residues (Figure 2(B)), related by a non-crystallographic Table 1. Data collection, MAD phasing and refinement statistics Data collection Space group C2 Unit cell parameters az124.0 Å, bz37.0 Å, cz60.3 Å, bz Data sets Edge ( Å) Peak ( Å) Remote ( Å) Resolution range (Å) ( ) ( ) ( ) No. of total reflections No. of unique reflections 5856 (387) 5756 (377) 6039 (403) I/s 17.7 (5.9) 22.6 (9.2) 25.0 (9.8) Completeness (%) 88.8 (88.6) 87.4 (85.7) 91.5 (91.6) R merge (%) a 5.6 (13.5) 5.0 (11.5) 4.9 (13.1) MAD phasing (20 3Å) Mean FOM b 0.44 Mean FOM b after density modification 0.70 Structure refinement Resolution (Å) ( ) R cryst /R free (%) c 19.8 (29.8)/26.7 (37.5) No. of reflections Working set 5382 Test set 657 Average B-factor (Å 2 ) Main-chain 34.4/37.3 d Side-chain 33.5/37.2 d Water 31.0 No. of atoms Protein atoms 2051 Water molecules 33 Selenium atoms 2 rmsd Bond distances (Å) Bond angles (deg.) 2.1 The recombined human CLP was expressed in E. coli bacterial strain BL21(DE3) using expression vector pet22b(c) (Novagen), and was purified using affinity chromatography as a standard procedure. For phase determination, the recombinant plasmid was transferred into Met-auxotrophic strain B834(DE3) to obtain the seleno-methionyl derivative of human CLP protein. Crystals of CLP were obtained at 277 K by the hanging-drop vapor diffusion method from the protein sample (20 mg/ml protein in 50 mm Tris HCl (ph 8.0), 5 mm NaCl, 2 mm imidazole) combined in a 1 : 1 ratio with a reservoir solution consisting of 30% (w/v) PEG2K, and 0.1 M Hepes (ph 7.0). A MAD data set was collected from a single seleno-methionine substituted CLP crystal at 100 K on Beijing Synchrotron Radiation Facility (BSRF) beamline 3W1A at the Institute of High Energy Physics, Chinese Academy of Sciences. All data were processed and scaled with the DENZO and SCALEPACK. 21 Two expected selenium positions were found in an asymmetric unit by SOLVE. 22 RESOLVE 23,24 also built a fragmented model of modest quality containing backbone atoms for about 50% of all residues. The program O 25 was used to rebuild and connect the fragments manually for the initial model. The model was refined against the highenergy remote data set in Å resolution range by using CNS. 26 NCS restraints were applied through all stages of refinement aside from last cycle. The stereochemical quality of the final model was checked by PROCHECK. 27 a R merge ZSjI i KI m j=si i, where I i is the intensity of the measured reflection and I m is the mean intensity of all symmetry-related reflections. b Mean FOM ðfigure of meritþz!jspðaþe ia a=spðaþj:, where a is the phase and P(a) is the phase probability distribution. Numbers in parentheses represent the value for the highest resolution shell. c R cryst ZSjjF obs jkjf calc jj=sjf obs j, where F obs and F calc are observed and calculated structure factors. R free ZS T jjf obs jkjf calc jj=s T jf obs j, where T is a test data set of about 10% of the total reflections randomly chosen and set aside prior to refinement. d Values for the two different monomers (A and B) respectively in an asymmetric unit. Numbers in parentheses represent the value for the highest resolution shell. 192

193 Crystal Structure of Human Coactosin-like Protein 319 Figure 1. The stereo view of ribbon diagram of human CLP. The six-stranded mixed b-sheet is in purple; the helices are in lightgreen; and the connecting loops are in gray. Structural elements are labeled in the left diagram. Figures 1 3(B) were prepared using Ribbons fold screw axis with a 32 Å translation along the axis (Figure 2(A)). Recently, the secondary structure of CLP determined by NMR has been reported. 11 NMR structures of human CLP (PDB code: 1WNJ) and mouse CLP (PDB code: 1UDM) were released in PDB. Superposition of these two solution structures and our crystal structure in backbone (Figure 2(B)) shows that the loop between b4 and b5 are obviously flexible, while the other regions keep highly similar. It has been reported that human CLP could exist both as a monomer and a dimer. 7 During the purification, we also observed human CLP always showed two bands with the molecular masses corresponding to the monomer and the dimer on SDS-PAGE, but gave a sharp single peak in mass spectroscopy with the monomer mass. The interactions between the two monomers in the crystal are only two hydrogen bonds, (carbonyl oxygen of Gly67-A to nitrogen of Gly33-B, and N 3 of Lys126-A to carbonyl oxygen of Asp32-B,) the biological significance of this packing needs further investigation. Comparison with ADF-H proteins A structural similarity search in the Protein Data Bank (PDB) with program DALI 12 indicates that human CLP shares highly homology with ADF/ cofilin family (A/Cs) in secondary structures arrangement and peptide folding, with the conformational variations most occurring in the loops (Figure 3(B)), although the amino acid sequence identity between human CLP and A/Cs is as low as less than 15%. Another group of actin-binding proteins structurally similar to human CLP is gelsolin/villin family, which contains a common b-sheet and a long helix as CLP (Figure 3(B)). Sequence alignment of human CLP with other CLP or coactosin, A/C family and gelsolin/villin family shows some conserved hydrophobic residues throughout the amino acid sequence (Figure 3(A)). These conserved hydrophobic residues interdigitate to form two hydrophobic cores (Phe29, Tyr31, Phe59, Phe61, Val101 and Val105 forming the first hydrophobic core, and Cys53, Trp81 and Leu120 forming the second hydrophobic core, Figure 3(B)), which are essential for stabilizing the similar folds of CLP, cofilin and gelsolin. This result provides key information to explain why CLP and ADF-H domains have low similarity in sequence but are highly conserved in three-dimensional structure. Although CLP is similar to ADF-H domains in folding, the actin-binding models revealed by the putative cofilin-actin complex UNC-60B 13 could not be simply applied to CLP. Cofilin binds to both G-actin and F-actin, while CLP binds to F-actin only. Systematic mutagenesis studies suggest that the residues (except Lys75) whose counterparts involved in cofilin actin binding have no effects on CLP F-actin interactions. Our CLP structure may provide explanation on why CLP does not bind to G-actin. Compared with yeast cofilin, the delegate of AC family, the N terminus 1 MSRSG 5 of which Ser4 and Gly5 are highly conserved for both G-actin and F-actin binding is not at all conserved in human CLP with the N-terminal sequence 1 MATKI. 5 Another important residue for G-actin binding to cofilin, Arg96 located in the kinked helix, is replaced with Leu89 in human CLP. The kinked a-helix (a3 in CLP), which is considered to be the F-actin-binding region in both cofilin and gelsolin, is also kinked in CLP as in cofilin. Although residue G95 is in the middle of a3 in CLP, it would not be the fact causing the bend because it is not conserved in coactosin and AC family, and its dihedral angles are in the most favored regions in Ramachandran plot. In order to characterize the geometry of the kinked a-helices 193

194 Figure 2. The two monomers of human CLP. (A) The ribbons diagram of two human CLP monomers packing in an asymmetric unit. The two monomers are related by a noncrystallographic 2-fold screw axis. (B) The stereo view of the superposition of the two monomers (green and red), NMR structure of human CLP (blue), and NMR structure of mouse CLP (yellow). The structural difference of these four structures, the loop connecting b4 and b5, are shadowed. Actin-binding site Lys75 and 5LO binding site Lys131 are showed in stick and ball model (purple). The each model of the two NMR structures used for superposition is the first model of 20 models. The N termini (residues before initial Met) and the C termini (residues after Ala132) of the selected models are cut before superposition, because the N termini do not exist in nature structure and the N- and C termini are highly disordered in the two NMR structures. 194

195 Crystal Structure of Human Coactosin-like Protein 321 Figure 3. (A) Sequence alignment of human CLP and the proteins in ADF-H and gelsolin/villin families. The sequences of CLP from Homo sapiens (Human), CLP from Mus musculus (House Mouse), and coactosin from D. discoideum were aligned using ClustalW. The sequence of human CLP is aligned with AC family members: yeast cofilin (PDB code: 1COF), Acanthamoeba actophorin (PDB code: 1CNU), Arabidopsis thaliana ADF1 (PDB code: 1F7S), human destrin (PDB code: 1AK6), mouse ADF-H domain (PDB code: 1M4J), and gesolin family members: human gelsolin (PDB code: 1H1V) domain G4, human Cap-G (PDB code: 1J72) domain1, chicken villin (PDB code: 2VIK) mainly based on the structural element alignment gave by DALI. The secondary structure of human CLP, which is defined by the analysis of the structure using DSSP program, is indicated above the alignment. The conserved hydrophobic residues forming the two hydrophobic cores are boxed in blue, and red respectively. The essential actin and 5LO-binding sites are 195

196 322 Crystal Structure of Human Coactosin-like Protein in human CLP and AC family members, program HELANAL 16 was used to calculate the bending angle. The results show that the maximum bending angle of the kinked a-helix in human CLP (21.38) is smaller than those in AC family members (w40 508). A critical role of maize ADF residues Tyr67 (strictly conserved in AC family) in proper protein folding has been demonstrated by mutating Tyr67 (equivalent to yeast cofilin Tyr64) to phenylalanine, which implicates that the strong hydrogen bond between the hydroxyl of Tyr64 (in yeast cofilin) and the carbonyl oxygen of Tyr101 (in yeast cofilin) is necessary for F-actin binding. The strong hydrogen bond would be helpful for stabilizing the kinked a-helix of AC family members. However, the corresponding position to yeast cofilin Tyr64 is Phe59 in human CLP (Figure 3(A)), which indicates that the smaller bending angle of human CLP is properly caused by missing the strong hydrogen bond conserved in AC family. In addition, the kinked a-helix of human CLP is shorter (14 residues) than that of structures of AC family and gelsolin/villin family (18 19 residues). Arg96 and Lys98 of yeast cofilin in this helix are involved in cofilin actin interactions, 14 but the corresponding residues in human CLP are Leu89 and Arg91. There is no evidence indicating they are related to actin binding. These structural variations imply the actin-binding mode of human CLP is different from that in AC family. F-actin and 5LO binding sites in CLP Polar residues cover most area of the CLP surface, with few hydrophobic residues exposed, suggesting that CLP binds to other proteins with hydrogen bonds and/or salt bridges. The critical binding residues for F-actin and 5LO are Lys75 and Lys131, respectively. 7,8 Mutation of Arg73 also affects CLP binding to b-actin. 8 In the current structure, Lys75 is located in the bottom of a cleft formed by b5 and the C-terminal helix (a4) (Figure 2(B)). The C-terminal residues of chain A, which could not be seen in chain B, cover the cleft and bury Lys75. If Lys75 interacts with F-actin directly, the relocation of the C-terminal peptide should be necessary. Besides Lys75 in the bottom of the cleft, Arg73 and Lys130 are on each side of the cleft. Based on this basic cleft, F-actin should interact with CLP with an acidic protrusion. By searching the Holmes Lorenz model of F-actin, 19,20 two potential CLP binding regions, with their negative charged surface and protruding shape, were found in the F-actin molecule. The first is the N terminus of actin subdomain 1, which is a stretched peptide with the sequence of 1 DEDE. 4 The second is the region of subdomain 1, which is consisted of 361 EYED 364. To be consistent, the biochemical studies on Lys75 also suggested the N-terminal peptide of actin subdomain 1 could act as a fishing rod to attracting the positive charged surfaces of actin-binding proteins. 7 Lys131 lies on the surface of helix 4. The distance between the NZ atom of Lys75 to the C a atom of Lys131 is 7.0 Å only. Since F-actin is an elongated structure and 5LO is a 78kD large protein, Lys75 and Lys131 are so spatially close that the steric hindrance could make the CLP 5LO F-actin ternary complex impossible. The current structure gives support to the experimental result that the 5LO CLP and CLP F-actin interactions are mutually exclusive, suggesting a modulation in actin dynamics. This is also consistent with the fact that no F-actin CLP 5LO ternary complex has ever been observed experimentally. Interestingly, these two critical binding residues are both close to the C-terminal region in structure and the C-terminal structure is quite disordered, suggesting the flexibility of the C-terminal region of CLP is important for actin and 5LO binding. In addition, actin-binding proteins compose a complex system of different kinds of proteins, regulating in different ways and competing with each other in binding with monomeric or polymeric actin. Coactosin interferes with the activity of capping proteins in this complicated system. The hydrophobic region of subdomains 1 and 3 are supposed to be the binding surfaces for capping proteins. 15 According to the F-actin CLP models suggested by us, these surfaces would not be exposed when human CLP binds to F-actin, so that CLP could efficiently counteract capping proteins association with F-actin. But because human CLP only binds to the outer side of F-actin, it would not have any effect on actin polymerization. Coordinates The atomic coordinates for human CLP have been deposited with the Protein Data Bank (PDB accession code: 1T2L). Acknowledgements This work is supported by the Foundation for Authors of National Excellent Doctoral Dissertation of People s Republic of China (Project No ), National Foundation of Talent Youth (Grant No ), the National High Technology Research boxed in cyan and yellow, which conserved in CLPs and coactosin. This figure was prepared using ESPript. 18 (B) The ribbon representation of human CLP, AC family (rendering with yeast cofilin) and gelsolin/villin family (rendering with human gelsolin) structures. Residues showed with ball-stick correspond to the ones explained above with the same color. The C terminus of human CLP are colored in purple, which buries the actin-binding site Lys75. The black arrow denotes the long helix that interacts with G-actin directly in gelsolin domain G

197 Crystal Structure of Human Coactosin-like Protein 323 and Development Program of China (Grant No. 2001AA233021), the 863 Special Program of China (Grant No. 2002BA711A13), the Key Important Project and other projects from the National Natural Science Foundation of China (Grant Nos , , and ) and Chinese Academy of Sciences (Grant No. KSCX1-SW-17). We thank Prof. Peng Liu and Yuhui Dong for diffraction data collection. References 1. Alberts, B., Bray, D., Lewis, J., Raff, M., Roberts, K. & Watson, J. D. (1994). The Molecular Biology of the Cell (3rd edit.). Garland Press, New York. 2. McGough, A. (1998). F-actin binding proteins. Curr. Opin. Struct. Biol. 8, de Hostos, E. L., Bradtke, B., Lottspeich, F. & Gerisch, G. (1993). Coactosin, a 17 kda F-actin binding protein from Dictyostelium discoideum. Cell Motil. Cytoskeleton, 26, Rohrig, U., Gerisch, G., Morozova, L., Schleicher, M. & Wegner, A. (1995). Coactosin interferes with the capping of actin filaments. FEBS Letters, 374, Chen, K. S., Manian, P., Koeuth, T., Potocki, L., Zhao, Q., Chinault, A. C. et al. (1997). Homologous recombination of a flanking repeat gene cluster is a mechanism for a common contiguous gene deletion syndrome. Nature Genet. 17, Nakatsura, T., Senju, S., Ito, M., Nishimura, Y. & Itoh, K. (2002). Cellular and humoral immune responses to a human pancreatic cancer antigen, coactosin-like protein, originally defined by the SEREX method. Eur J Immunol. 32, Provost, P., Doucet, J., Stock, A., Gerisch, G., Samuelsson, B. & Radmark, O. (2001). Coactosin-like protein, a human F-actin-bining protein: critical role of lysine-75. Biochem. J. 359, Provost, P., Doucet, J., Hammarberg, T., Gerisch, G., Samuelsson, B. & Radmark, O. (2001). 5-Lipoxygenase interacts with coactosin-like protein. J. Biol. Chem. 276, Peters-Golden, M. & Brock, T. G. (2000). Intracellular compartmentalization of leukotriene biosynthesis. Am. J. Respir. Crit. Care Med. 161, S36 S Lappalainen, P., Kessels, M. M., Cope, M. J. & Drubin, D. G. (1998). The ADF homology (ADF-H) domain: a highly exploited actin-binding module. Mol. Biol. Cell, 9, Dai, H., Wu, J., Xu, Y., Tang, Y., Ding, H. & Shi, Y. (2004). 1 H, 13 C and 15 N resonance assignments and the secondary structures of human coactosin like protein (hclp) D123N. J. Biomol. NMR, 29, Holm, L. & Sander, C. (1993). Protein structure comparison by alignment of distance matrices. J. Mol. Biol. 233, Ono, S., McGough, A., Pope, B. J., Tolbert, V. T., Bui, A., Pohl, J. et al. (2001). The C-terminal tail of UNC- 60B (actin depolymerizing factor/cofilin) is critical for maintaining its stable association with F-actin and is implicated in the second actin-binding site. J. Biol. Chem. 276, McLaughlin, P. J., Gooch, J. T., Mannherz, H. G. & Weeds, A. G. (1993). Structure of gelsolin segment 1-actin complex and the mechanism of filament severing. Nature, 364, Lappalainen, P., Fedorov, E. V., Fedorov, A. A., Almo, S. C. & Drubin, D. G. (1997). Essential functions and actin-binding surfaces of yeast cofilin revealed by systematic mutagenesis. EMBO J. 16, Bansal, M., Kumar, S. & Velavan, R. (2000). HELANAL: a program to characterize helix geometry in proteins. J. Biomol. Struct. Dynam. 17, Carson, M. (1997). Ribbons. Methods Enzymol. 277, Gouet, P., Courcelle, E., Stuart, D. I. & Metoz, F. (1999). ESPript: multiple sequence alignments in PostScript. Bioinformatics, 15, Lorenz, M., Popp, D. & Holmes, K. C. (1993). Refinement of the F-actin model against X-ray fiber diffraction data by the use of a directed mutation algorithm. J. Mol. Biol. 234, Holmes, K. C., Popp, D., Gebhard, W. & Kabsch, W. (1990). Atomic model of the actin filament. Nature, 347, Otwinowski, Z. & Minor, W. (1997). Processing of X-ray diffraction data collected in oscillation mode. In Macromolecular Crystallography (Carter, C.W.a.S.R.M., ed) Methods in Enzymology, vol Terwilliger, T. C. & Berendzen, J. (1999). Automated MAD and MIR structure solution. Acta Crystallog. sect. D, 55, Terwilliger, T. C. (2000). Maximum likelihood density modification. Acta Crystallog. sect. D, 56, Terwilliger, T. C. (2002). Automated main-chain model-building by template-matching and iterative fragment extension. Acta Crystallog. sect. D, 59, Jones, T. A., Zou, J. Y., Cowan, S. W. & Kjeldgaard, M. (1991). Improved methods for building protein models in electron density maps and the location of error in these models. Acta Crystallog. sect. A, 47, Brunger, A. T., Adams, P. D., Clore, G. M., DeLano, W. L., Gros, P., Grosse-Kunstleve, R. W. et al. (1998). Crystallography & NMR system (CNS): a new software suite for macromolecular structure determination. Acta Crystallog. sect. D, 54, Laskowski, R. A., MacArthur, M. W., Moss, D. S. & Thornton, J. M. (1993). PROCHECK: a program to check the stereochemical quality of protein structures. J. Appl. Crystallog. 26, Edited by R. Huber (Received 25 June 2004; received in revised form 14 September 2004; accepted 16 September 2004) 197

198 Protein Kinase Activation Increases Insulin Secretion by Sensitizing the Secretory Machinery to Ca 2 Qun-Fang Wan, 1 Yongming Dong, 1 Hua Yang, 1 Xuelin Lou, 1 Jiuping Ding, 1 and Tao Xu 1,2 1 Institute of Biophysics and Biochemistry, School of Life Science, Huazhong University of Science and Technology, Wuhan , China 2 National Laboratory of Biomacromolecules, Institute of Biophysics, Chinese Academy of Sciences, Beijing , China The Journal of General Physiology abstract Glucose and other secretagogues are thought to activate a variety of protein kinases. This study was designed to unravel the sites of action of protein kinase A (PKA) and protein kinase C (PKC) in modulating insulin secretion. By using high time resolution measurements of membrane capacitance and flash photolysis of caged Ca 2, we characterize three kinetically different pools of vesicles in rat pancreatic -cells, namely, a highly calciumsensitive pool (HCSP), a readily releasable pool (RRP), and a reserve pool. The size of the HCSP is 20 ff under resting conditions, but is dramatically increased by application of either phorbol esters or forskolin. Phorbol esters and forskolin also increase the size of RRP to a lesser extent. The augmenting effect of phorbol esters or forskolin is blocked by various PKC or PKA inhibitors, indicating the involvement of these kinases. The effects of PKC and PKA on the size of the HCSP are not additive, suggesting a convergent mechanism. Using a protocol where membrane depolarization is combined with photorelease of Ca 2, we find that the HCSP is a distinct population of vesicles from those colocalized with Ca 2 channels. We propose that PKA and PKC promote insulin secretion by increasing the number of vesicles that are highly sensitive to Ca 2. key words: INTRODUCTION exocytosis insulin calcium sensitivity PKA PKC Insulin secretion is subject to precise regulation by nutrient and nonnutrient secretagogues. Despite the well-known depolarization secretion coupling initiated by metabolism of glucose and other nutrient secretagogues, nutrients also activate intracellular signaling pathways that lead to the activation of protein kinases such as PKC and PKA (Nesher et al., 2002). On the other hand, nutrient-induced insulin responses can be radically modified by nonnutrient secretagogues, including a wide variety of hormones and neurotransmitters, through the same intracellular regulators as nutrient secretagogues. For example, cholinergic muscarinic agonists generate diacylgycerol (DAG) and subsequently activate PKC (Verspohl and Wienecke, 1998), glucagon and glucose-dependent insulinotropic polypeptide elevate camp with subsequent activation of PKA, whereas somatostatin, galanin, or the 2-adrenoreceptor agonists inhibit adenylate cyclase and reduce intracellular camp (Sharp, 1996). Although protein kinases have been implicated in the control of insulin secretion, precisely how they participate in producing a controlled insulin response is not well understood. In addition to modulating cell excitability, calcium influx, and gene expression (Bozem et al., 1987; Liu and Heckman, 1998), recent evidence also suggests that PKA or PKC act directly on secretory machinery. In pancreatic -cells, PKA activation potentiates insulin secretion by increasing the total number of vesicles that are available for release (Renstrom et al., 1997; Rorsman et al., 2000). PKC activation has also been linked to priming of Ca 2 -mediated insulin secretion as well as enhancement of non- Ca 2 mediated exocytosis (Eliasson et al., 1996; Efanov et al., 1997). Direct interactions of PKA or PKC with the secretory machinery has also been suggested in other cell types, such as chromaffin cells and hippocampal neurons, where the size of the RRP and its rate of replenishment is increased (Smith et al., 1998; Stevens and Sullivan, 1998). Recently, a direct modulation of the Ca 2 sensitivity of fusion by PKC has been demonstrated in chromaffin cells and gonadotropes (Yang et al., 2002; Zhu et al., 2002). This study investigates the mechanisms whereby insulin secretion is regulated by PKA and PKC in rat pancreatic -cells. We focused on the secretory response distal to Ca 2 signaling by using whole cell membrane capacitance (C m ) measurements and direct and spatially uniform manipulation of [Ca 2 ] i with caged Ca 2. We Downloaded from on November 30, 2004 Address correspondence to Tao Xu, Institute of Biophysics and Biochemistry, School of Life Science and Technology, Huazhong University of Science and Technology, Wuhan , China. Fax: ; txu@mail.hust.edu.cn Abbreviations used in this paper: DAG, diacylgycerol; HCSP, highly calcium-sensitive pool; IRP, immediately releasable pool; RRP, readily releasable pool. 653 J. Gen. Physiol. The Rockefeller University Press /2004/12/653/10 $8.00 Volume 124 December

199 have identified a small, highly calcium-sensitive pool (HCSP) in addition to the previously reported RRP and the reserve pool. The size of the HCSP dramatically increased after treatment with PMA or forskolin, which also increased the size of the RRP, albeit to a lesser extent. To better characterize the specific isoforms of PKC involved in the modulation of secretory vesicles, we tested the effects of various PKC inhibitors on secretory responses. Furthermore, we evaluated whether the actions of PKC and PKA converge to influence the secretory process. MATERIALS AND METHODS Cell Culture and Solutions All experiments were performed on isolated rat pancreatic -cells at C. Pancreatic -cells from adult male Wistar rats were prepared as described previously (Ashcroft et al., 1984). In brief, rats were killed by cervical dislocation, and the islets were collected from the pancreas after collagenase digestion. The islets were further digested by dispase II to dissociate single -cells. Dispersed cells were kept in DMEM supplemented with 25 mm HEPES, 2 g/l NaHCO 3, 100 IU/ml penicillin, 100 g/ml streptomycin, and 10% FCS. Cells cultured for 3 5 d were used in the experiments. Standard bath solution for the experiments contained (in mm) 138 NaCl, 5.6 KCl, 1.2 MgCl 2, 2.6 CaCl 2, 3 glucose, 5 HEPES (ph 7.2, 310 mosm). For pipette solutions, we generally prepared 2 concentrated buffers, which contained 250 mm Csglutamate and 80 mm HEPES (ph 7.2). We added to the 2 buffer CaCl 2, ATP, GTP, caged Ca 2, and Ca 2 indicators. Standard internal solution consisted of (in mm) 110 Cs-glutamate, 2 MgATP, 0.3 GTP, and 35 HEPES with different loaded caged Ca 2 of either NP-EGTA or DM-nitrophen, and 0.2 mm various Ca 2 indicators, such as fura-6f or furaptra. The basal [Ca 2 ] i was measured to be around 200 nm by fura-2. For experiments with depolarization, the bath solution contained 10 mm CaCl 2.The pipette solution was adjusted to ph 7.2 with either HCl or CsOH. The osmolarity was adjusted to around 300 mosm. Stock solutions of forskolin, PMA, Gö6976, and Gö6983 were prepared in DMSO. Stock solutions of PKC19-31 were made in 5% acetic acid. The final concentration of DMSO or acetic acid in diluted solutions was 0.02%. DMEM, Dispase-II, FBS, and BSA were from GIBCO BRL; PKC19-31, Gö6976, and Gö6983 were purchased from Calbiochem; NP-EGTA, DM-nitrophen, fura-2, fura- 6F, and furaptra were from Molecular Probes; forskolin, Rp-cAMP, PMA, and all other chemicals were purchased from Sigma-Aldrich. Membrane Capacitance (C m ) Measurement We selected cells with diameters 11 m for study, so that 90% of the cells were expected to be -cells (Rorsman and Trube, 1986). Conventional whole-cell recordings were conducted using Sylgard-coated pipettes with series resistance ranging from 4 to 12 M. An EPC-9 patch-clamp amplifier was used together with PULSE LOCK-IN software (Heka Elektronics). A 1042-Hz, 20- mv peak-to-peak sinusoidal voltage stimulus was superimposed on a holding potential of 70 mv. Currents were filtered at 2.9 khz and sampled at 15 khz. The capacitance traces were imported to IGOR Pro (WaveMetrics) for further analysis. Flash Photolysis Flashes of ultraviolet light and fluorescence excitation light were generated as described previously (Xu et al., 1997). In the flash experiments, exocytosis was elicited by photorelease of caged Ca 2 preloaded into the cell via the patch pipette. Flashes of UV light were generated by a flash lamp (Rapp Optoelektronik). [Ca 2 ] i was measured with the Ca 2 indicator dyes fura-2, fura-6f, or furaptra. The dyes were excited with light alternating between 340 and 385 nm from a monochromator-based system (TILL photonics). The resulting fluorescence signal was measured by a photomultiplier. [Ca 2 ] i was determined from the ratio (R) of the fluorescence signals excited at the two wavelengths, following the equation [Ca 2 ] i Keff * (R Rmin)/(Rmax R), where Keff, Rmin, and Rmax are constants obtained from intracellular calibration as previously described (Xu et al., 1997). In brief, four solutions with [Ca 2 ] i of nominal zero (10 mm EGTA with no added Ca 2 ), 20 M (20 mm DPTA with 4 mm CaCl 2 ), 80 M (20 mm DPTA with 10 mm CaCl 2 ), and 10 mm (10 mm CaCl 2 with no added buffer) were dialyzed against the cytosol in the whole-cell patch clamp recording. Three to five recordings were made for each calibration solution to estimate the calibration constants. Data Analysis Data analysis was performed using IGOR Pro software (Wavemetrics), and results were presented as mean SEM with the indicated number of experiments. Statistical significance was evaluated using Student s t test. P 0.05 was considered to be statistically significant. RESULTS Characteristics of Secretory Response in Rat Pancreatic -cells We first characterized the secretory response to different [Ca 2 ] i levels in single rat pancreatic -cells. Exocytosis was elicited by flash photorelease of Ca 2 in the whole-cell patch-clamp configuration. Following a flash, [Ca 2 ] i was uniformly elevated to the M range within few milliseconds. Thus, the sizes of distinct vesicle pools and their secretory kinetics at a given [Ca 2 ] i could be studied directly without the complications of [Ca 2 ] i microdomains or modulation of Ca 2 influx. Fig. 1 A displays a typical C m response to a step-like [Ca 2 ] i elevation. The C m trace clearly displayed multiple kinetic components of exocytosis, indicating the presence of different vesicle pools as has been suggested for other cell types (Neher, 1993; Heinemann et al., 1994; Xu et al., 1998; Voets, 2000). Each exponential component of the C m trace is usually interpreted as release of a discrete vesicle pool, whereas the sustained linear increase is thought to reflect refilling from a reserve pool of vesicles (Sorensen et al., 2002). When we looked into the detailed kinetics of the initial exocytotic burst at an expanded time scale (Fig. 1 B), we observed a small but very fast component of exocytosis at low M [Ca 2 ] i. The amplitude of this fast component reflects release of 6 12 vesicles ( 20 ff) if we assume that one insulin-containing granule contributes ff of membrane as determined in pancreatic -cells (Ammala et al., 1993; Braun et al., 2004). This component was readily identifiable in 50% of the cells (n 93) studied. This variability likely results from the rela- Downloaded from on November 30, Sensitization of Insulin Secretion 199

200 tively small size of this pool and considerable cell-to-cell variation of secretory competence. The most rapid component of exocytosis had a relatively fast time constant of 20 ms at a [Ca 2 ] i level of 3.1 M. In contrast, recent studies in -cells report exocytosis from a readily releasable pool (RRP) with an amplitude of 200 ff and time constants of 1 s or longer upon photoelevation of [Ca 2 ] i to 3 M (Takahashi et al., 1997; Barg et al., 2001). Thus, this small, fast component in the exocytotic burst at low [Ca 2 ] i is kinetically distinguishable from the previously described RRP but is similar to what has recently been described as a highly Ca 2 -sensitive pool (HCSP) in chromaffin cells (Yang et al., 2002) and rat insulinoma INS-1 cells (Yang and Gillis, 2004). As we elevated [Ca 2 ] i to higher values, a slower but larger phase of exocytosis became dominant (Fig. 1 C). This slower phase had an amplitude ( 200 ff) and kinetics comparable to the previously reported RRP (Takahashi et al., 1997; Barg et al., 2001; Olofsson et al., 2002). PMA Significantly Increased the Size of the HCSP Next, we tested the effect of PMA, a PKC activator, on exocytosis in pancreatic -cells. A prominent HCSP component with much larger size than in control cells is evident after PMA treatment (Fig. 2, A and B). When the averaged time courses of C m increase in response to 655 Wan et al. 200 Figure 1. The kinetics of exocytosis in pancreatic -cells at different [Ca 2 ] i levels. (A) Example trace of a secretory response following flash photolysis that exhibited two distinct phases. Superimposed is the double exponential fit (solid line) with the rate constants and amplitudes indicated. The exocytotic burst is expanded in the inset. (B and C) Expanded exocytotic bursts in response to different post-flash [Ca 2 ] i levels with superimposed double exponential fits (solid lines). (B) At lower [Ca 2 ] i levels, a small-amplitude exponential component with a relatively fast rate constant was followed by a slower, but larger amplitude second exponential component. (C) At higher [Ca 2 ] i levels, the slower but larger phase of exocytosis dominated the exocytotic burst. similar post-flash [Ca 2 ] i levels were compared, PMA clearly increased the burst component of exocytosis as well as the sustained component (Fig. 2 C). To investigate the Ca 2 dependence of secretion, the exocytotic bursts were further fitted by a double exponential to obtain the rate constants and amplitudes of the HCSP and RRP. Fig. 3 compared the sizes and the rate constants of the HCSP and RRP from control (n 93) and PMA-treated (n 117) cells between [Ca 2 ] i levels of 0.7 and 30 M. Despite the comparable rate constants, the amplitudes of the HCSP were much greater in PMA-treated cells, as shown in the upper part of Fig. 3. We did not observe a clear dependence of the amplitude on [Ca 2 ] i. The rate constant of release from the HCSP was dependent on [Ca 2 ] i between 0.8 and 3 M and saturated at [Ca 2 ] i above 3 M. We fitted the HCSP rate constant to the equation Rate R max /(1 (K d / [Ca 2 ] i ) n ), and obtained an estimated K d of M. The best-fit Hill coefficient (n) was 1.9, suggesting a less cooperativity in the Ca 2 -dependent fusion of HCSP compared with the RRP. A shallow dependence of the rate of exocytosis of the HCSP on [Ca 2 ] i is also suggested in pituitary gonadotropes (Zhu et al., 2002), chromaffin cells (Yang et al., 2002), and in INS-1 cells (Yang and Gillis, 2004). The results reveal a novel mechanism to increase the apparent Ca 2 dependence Downloaded from on November 30, 2004

201 of insulin release by recruiting more vesicles into a highly calcium-sensitive state. Specificity of PMA Action in Regulating Insulin Secretion PMA activates PKC due to its structural similarity to the endogenous activator, DAG, by binding to the C1 domain of PKC. The specificity of PMA activation of PKC has been questioned because PMA also activates protein kinase D, RasGRPs, and diacylglycerol kinase with equal potency (Kazanietz, 2002). Munc13, a family of proteins that prime exocytosis, also possess a PMA-binding (C1) domain and are translocated to the plasma membrane upon bath application of PMA (Rhee et al., 2002). Moreover, a recent report demonstrates that Munc13-1 functions in regulating insulin secretion (Sheu et al., 2003). Thus, we applied a number of specific PKC inhibitors to determine whether PKC is involved in the effect of PMA, and which PKC isoform might be involved in regulating secretion. Fig. 4 summarizes the averaged amplitudes of the HCSP and the RRP from control cells and cells treated with PMA and various PKC inhibitors. A specific pharmacological tool to test the involvement of PKC is the inhibitory peptide, PKC19-31, a pseudosubstrate sequence that interacts with the PKC substrate binding site in the C4 region of the catalytic domain. We included 1 M PKC19-31 (IC nm) in the pipette solution and waited Figure 2. PMA increases the size of the HCSP. (A and B) Examples of the kinetics of exocytosis from PMA-treated (100 nm for 2 3 min) cells at two different post-flash [Ca 2 ] i levels. Superimposed solid curves are exponential fits. (C) Averaged [Ca 2 ] i (upper traces) and C m responses (lower traces) from control (circles, n 14) and PMAtreated (triangles, n 16) cells. for 3 min after establishing the whole-cell configuration before flash. As shown in Fig. 4 A, PKC19-31 abolished the stimulatory effect of PMA on the HCSP. At least five PKC isoenzymes (,,,, and ) have been found in rat pancreatic -cells (Kaneto et al., 2002). PMA activates the Ca 2 -dependent isoforms PKC and PKC II, and the Ca 2 -independent isoforms PKC and PKC (Csukai and Mochly-Rosen, 1999). Here, we tried to better define the relevant PKC isoforms that are involved in regulating insulin secretion by comparing the differential effects of Gö6976 and Gö6983. Gö6976 selectively inhibits Ca 2 -dependent PKC (IC nm) and PKC I (IC nm), whereas it does not affect the kinase activity of the Ca 2 -independent PKC isozymes even in the micromolar range (Gschwendt et al., 1996). Gö6983 selectively inhibits several PKC isozymes but does not discriminate between them. As shown in Fig. 4 A, Gö6976 (100 nm) and Gö6983 (100 nm) were equally potent in blocking the stimulatory effect of PMA. PMA enhanced the size of the RRP and the sustained component to a much lesser extent than its effect on the HCSP. This effect was also blocked by various PKC inhibitors (Fig. 4 A). Furthermore, we tested whether the PKC inhibitors used in this study could exert any effect on exocytosis in the absence of PMA. Fig. 4 B reveals that PKC19-31, Gö6976, and Gö6983 did not in- Downloaded from on November 30, Sensitization of Insulin Secretion 201

202 Figure 3. The rate constants and amplitudes of exponential fits to C m responses with similar post-flash [Ca 2 ] i levels are averaged and plotted versus [Ca 2 ] i. Circles and triangles represent the fast and slow components of double exponential fits. For [Ca 2 ] i 10 M, it is hard to distinguish the HCSP from RRP, therefore we only fit a single exponential to the C m traces and the rate constants are denoted as squares. The open and filled symbols represent data from control and PMA-treated cells, respectively. The Ca 2 dependence of the rate of exocytosis from the HCSP was fitted by the equation (dashed line) Rate R max /(1 (K d /[Ca 2 ] i ) n ), where R max, K d, and n were s 1, M, and , respectively. The dotted lines in the upper panel mark the averaged HCSP sizes for control and PMA-treated cells. fluence the different components of secretion significantly, suggesting little tonic activity of PKC. The Effect of PKA on Different Exocytotic Components in -cells To investigate the role of PKA in insulin secretion, we studied the effects of forskolin, an activator of adenylate cyclase, on the different secretory components in pancreatic -cells. After application of 10 M forskolin for 2 3 min, the amplitude of the HCSP was greatly enhanced (Fig. 5, A and B, see Fig. 1 for comparison). The averaged C m response to similar [Ca 2 ] i levels showed a pronounced increase in amplitude over that from control cells (Fig. 5 A). As with PMA, forskolin did not change the kinetics of the secretory response, whereas it increased the size of the HCSP (Fig. 6 A). The Ca 2 dependence of the release rate from the HCSP and RRP was fitted to the same equation as in Fig. 3. As summarized in Fig. 6 (B and C), forskolin dramatically increased the size of HCSP as well as RRP. The stimulatory effect of forskolin was blocked by a PKA antagonist, Rp-cAMP (10 M), demonstrating the involvement of PKA activation. Control experiments showed that Rp-cAMP had no effect on exocytosis in the absence of forskolin. Figure 4. The effect of PMA is mediated by PKC activation. (A) Averaged amplitudes of different kinetic components for control cells and cells with various treatments. The C m response was fitted by triple exponentials and the amplitudes of the three components were taken as the size of the HCSP, the RRP, and the sustained component. PKC19-31 (1 M, n 19), Gö6976 (100 nm, n 22), or Gö6983 (100 nm, n 14) were included in the pipette solution during perfusion with PMA in the external solution. (B) The effect of PKC19-31 (n 21), Gö6976 (n 12), or Gö6983 (n 19) alone on the three kinetic components of exocytosis. Values represent the mean SEM. Asterisks denote significant differences (t-test, *P 0.05, **P 0.01). The sizes of the HCSP, RRP, and sustained components were normalized to their control values from paired experiments, respectively, to guard against day-to-day variation. Convergent Effects of PKA and PKC on Exocytosis To test whether PKC and PKA act on the same targets in enhancing exocytosis, we challenged pancreatic -cells with combined application of 100 nm PMA and 10 M forskolin in the bath solution and compared their effect with that of PMA or forskolin applied alone. As shown in Fig. 7, the cocktail of PMA plus forskolin exerted no greater effect than each compound alone. Thus, the effects of PKC and PKA are not additive, suggesting that activation of either one may converge on the same secretory pathway in the regulation of insulin secretion. The Relationship between the HCSP and Rapid Depolarization-evoked Exocytosis A subset of vesicles has been postulated to colocalize with Ca 2 channels and is often termed the immediately releasable pool (IRP) of vesicles (Horrigan and Bookman, 1994). The size of the IRP in pancreatic -cells has been estimated to be around 30 ff (Barg et al., Downloaded from on November 30, Wan et al. 202

203 2001), which is similar to the basal size of the HCSP in -cells. To clarify the relationship between the HCSP and the IRP, we have designed a cross-depletion experiment by combining brief membrane depolarization and flash photolysis of caged Ca 2. We employed a dual-pulse protocol (two successive 30 ms depolarizing pulses separated by 100 ms at holding potential) to estimate the size of the IRP as previously described (Gillis et al., 1996), then a subsequent flash uncaging to elicit release from the HCSP. One example response in the presence of forskolin is depicted in Fig. 8 A, where elevation [Ca 2 ] i to 4.3 M is still capable of eliciting robust exocytosis from the HCSP after the depletion of IRP. Next, we repeated the cross-depletion protocol in the presence or absence of forskolin to determine whether the IRP and HCSP are differentially regulated. Fig. 8 B summarizes the results from 35 cells. The size of the IRP has not been significantly enhanced upon treatment of forskolin; however, the size of the HCSP following the depletion of the IRP was increased from to ff. This result further suggests that granules in the HCSP are distinct from those in the IRP. DISCUSSION We have characterized in detail the secretory response in single rat pancreatic -cells to different [Ca 2 ] i levels. We find a highly Ca 2 -sensitive phase of exocytosis Figure 5. Forskolin significantly increases the size of the HCSP. (A and B) Examples of the kinetics of exocytosis from forskolin-treated cells (10 M for 2 3 min) at two different post-flash [Ca 2 ] i levels. Superimposed solid curves are exponential fits. (C) Averaged [Ca 2 ] i (upper traces) and C m responses (lower traces) from control and forskolin-treated cells, from experiments that had similar post-flash [Ca 2 ] i values (n 14 for control, circles; n 13 for forskolin, triangles). in mouse -cells in addition to the previously described RRP and reserve pool of vesicles. We have shown that application of either PMA or forskolin dramatically increases the size of the HCSP by a factor of approximately four, an effect involving activation of PKC or PKA as testified by the effectiveness of specific protein kinase inhibitors. The effects of PKC and PKA were not additive, suggesting that a convergent mechanism is used to modulate the secretory machinery in pancreatic -cells. [Ca 2 ] i Dependence of Exocytosis in Pancreatic -cells The Ca 2 dependence of insulin secretion is controversial. It has been reported that little exocytosis is evoked in pancreatic -cells upon photolysis of caged Ca 2 to levels 3 M (Takahashi et al., 1997; Barg et al., 2001), yet we detect robust exocytosis at [Ca 2 ] i 1 M. It is quite possible that the HCSP we report here has been overlooked in previous studies on -cells because of its small and variable amplitude, yet we also see robust release from the larger RRP at 1 M. Our results are consistent with a number of studies that have demonstrated insulin secretion at [Ca 2 ] i levels of 1 M (e.g., Bergsten, 1995; Bokvist et al., 1995; Proks et al., 1996; Lang et al., 1997). One possible reason for the discrepancy between our results and other caged Ca 2 studies are differences in the [Ca 2 ] i level before the flash and/or the presence of ATP in the pipette so- Downloaded from on November 30, Sensitization of Insulin Secretion 203

204 Figure 7. The effects of PMA and forskolin on exocytosis are not additive. Application of forskolin and PMA produced a fourfold increase in the size of the HCSP. PMA and forskolin also increased the size of RRP and the sustained component; however, the augmentation was not as dramatic as that of the HCSP. The sizes of the HCSP, RRP, and sustained components were normalized to their control values from paired experiments. Figure 6. Summary of the effects of forskolin on different components of exocytosis. (A) The rate constants and amplitudes of exponential fits to C m responses plotted versus [Ca 2 ] i. Circles and triangles represent the fast and slow components of double exponential fits. The open and filled symbols represent data from control and forskolin-treated (10 M) cells, respectively. Squares denote the rate constants of single exponential fit to the C m traces for [Ca 2 ] i 10 M. The dashed line is reproduced from Fig. 3. The lines in the upper panel mark the averaged HCSP size for control and forskolin-treated cells. (B) Comparison of the amplitude of the HCSP between control cells and those treated with forskolin, Rp-cAMP (10 M) alone, or Rp-cAMP plus forskolin. (C) Application of forskolin significantly increased the amplitude of the RRP. Data are displayed as mean SEM. **P lution. For example, in the study by Takahashi et al. (1999), the pipette solution contained 10 mm DMnitrophen together with mm CaCl 2 and no ATP. The basal [Ca 2 ] i was estimated to be 5 nm. In the present study we have determined the basal [Ca 2 ] i to be 200 nm and we included 2 mm ATP in the pipette solution. It should be noted that both basal Ca 2 and 659 Wan et al. 204 ATP are important for the priming and maintenance of the docked RRP (von Ruden and Neher, 1993; Parsons et al., 1995; Eliasson et al., 1997; Xu et al., 1998; Smith et al., 1998; Takahashi et al., 1999). Thus, our results supported the notion that pancreatic -cells possess a high affinity Ca 2 sensor for exocytosis. PKC Isoforms Involved in Insulin Regulation PMA has been employed to investigate the role of PKC in the regulation of insulin secretion. However, because of the ubiquitous expression of many PKC isoforms and the large number of PKC regulators and substrates, the precise role of PKC activity and the identity of the relevant PKC isoforms have often remained elusive. Recently, Munc13s have been suggested as alternative DAG and Ca 2 receptors that function in regulating vesicle priming (Rhee et al., 2002). Thus, we used various specific PKC inhibitors including PKC19-31, Gö6976, and Gö6983 to demonstrate that PKC was indeed involved in the enhancement of exocytosis in -cells. Among multiple PKC isoforms, only classical PKCs (, I,, and ) and atypical PKCs (, /, and ) can be activated by PMA. BIS and the BIS-derived PKC inhibitor, Gö6983, inhibit several PKC isoforms (,,,, and ) without discriminating between them. Gö6976 selectively inhibits Ca 2 -dependent PKC (IC nm) and PKC I (IC nm), whereas it does not affect the kinase activity of the Ca 2 -independent PKC isoforms (,, and ) even in the micromolar range (Gschwendt et al., 1996). The fact that Gö6976 and Gö6983 were equally effective in blocking the stimulatory effect of PMA on the HCSP suggested the involvement of classical PKCs. Among all PKC isoforms, PKC,,,,, and are reported to be expressed in rat pancreatic islets. This narrows the PKC isoforms to PKC and PKC. PKC has been proposed in the priming of synaptic vesicles in the Calyx of Held (Wu Downloaded from on November 30, 2004

205 Figure 8. The HCSP is distinct from the IRP. (A) Sample response to two successive brief depolarizing pulses (30 ms in duration) to 10 mv followed by flash photolysis of caged Ca 2. Note that the C m response to the second pulse is substantially smaller than the response to the first pulse, demonstrating the depletion of the IRP. At the time indicated by the arrow, a UV flash was given to elevate [Ca 2 ] i to 4.3 M. The dashed line is single-exponential fit with the fastest time constant indicated. The experiment is performed in the presence of 10 M forskolin. (B) Summary of the estimated size of IRP and HCSP in the absence (open bars, n 8) and presence of forskolin (filled bars, n 27). The size of the HCSP is significantly increased by forskolin application (t test, P 0.05), whereas the size of the IRP remains unchanged. and Wu, 2001), whereas PKC is thought to induce c-myc expression and suppress insulin gene transcription (Kaneto et al., 2002). Recent experiments also have suggested that PMA-stimulated insulin secretion involved activation of PKC but not PKC (Carpenter et al., 2004). Thus, we suspect that PKC is involved in the enhancement of the HCSP in rat pancreatic -cells. Substrates of Protein Kinases Identification of kinase substrates and of their cellular functions is crucial to a full understanding of the regulatory roles of protein kinases in insulin secretion. A number of unidentified kinase substrates have been localized to -cell secretory vesicles or membrane fractions, which might be involved in vesicle trafficking, priming, and fusion. In this study, we have restricted the kinase substrates to those that are important for regulating vesicle priming and fusion by using patchclamped -cells and intracellular Ca 2 photorelease techniques. We found that activation of either PKA or PKC might act on a common site for changing the Ca 2 sensitivity of the primed vesicles. Despite a common effect on the HCSP, stimulation of PKA by 10 M forskolin gives a greater enhancement of the RRP than stimulation of PKC by 100 nm PMA in our hand. Considering synaptotagmin and the soluble N-ethylmaleimide sensitive fusion attachment protein receptor (SNARE) complex as an integrated calcium sensor for fusion (Xu et al., 1998), one could envision that phosphorylation of synaptotagmin or SNARE proteins, or other SNARE-interacting proteins, might finely adjust the Ca 2 sensing of exocytosis. The refilling and priming of vesicles involves multiple possible downstream effectors of protein kinases including SNARE proteins, SNAP, and Rab proteins, etc. Overexpression of phosphomimetic and phosphorylation-defective mutant variants of SNAP-25 has recently dissected the role of PKC-mediated phosphorylation of SNAP-25 during vesicle recruitment in chromaffin cells (Nagy et al., 2002). Similar methods can be employed to investigate the substrates of protein kinases in -cells. Alternatively, a more systematic proteomic approach will help to elucidate the kinase targets that are important in the regulation of insulin secretion. Model of Insulin Secretion Control by Protein Kinases Our study has revealed a novel, small HCSP with high Ca 2 sensitivity in rat pancreatic -cells. The origin of this HCSP remains to be revealed since small GABAcontaining synaptic-like vesicles have been suggested in pancreatic -cells in addition to dense-core insulin-containing granules (Thomas-Reetz and De Camilli, 1994; Takahashi et al., 1997). However, recently the release of GABA-containing synaptic-like vesicles has been estimated to contribute only 1% of the capacitance signal in -cells (Braun et al., 2004). In an accompanying report, Yang and Gillis (2004) have shown that quantal 5-HT release correlates well with the exocytosis HCSP in insulin-secreting INS-1 cell line. Simultaneous measurement of insulin and 5-HT release with modified carbon fiber electrodes demonstrates that 5-HT is released exclusively from insulin-containing granules (Aspinwall et al., 1999). Thus, it is possible that the HCSP is composed of the same type of insulin-containing granules as the conventional RRP. We have shown that PKC and PKA activation can dramatically increase the size of the HCSP. Similar effects have also been reported in the rat insulinoma INS-1 cells (Yang and Gillis, 2004), albeit augmentation of the fraction of granules in the HCSP was less prominent in this study. The preferential augmentation of the HCSP has been also reported in chromaffin cells (Yang et al., 2002) and possibly in INS-1 cells when Downloaded from on November 30, Sensitization of Insulin Secretion 205

206 stimulated with glucose (Yang and Gillis, 2004). We have found that the bulk of the HCSP is not released in response to brief depolarization sufficient to deplete the IRP, suggesting that most of the granules in the HCSP do not colocalize with Ca 2 channels. The same conclusion has been drawn from studies in pituitary gonadotropes (Zhu et al., 2002), chromaffin cells (Yang et al., 2002), and INS-1 cells (Yang and Gillis, 2004). Thus, the HCSP is likely to respond to global elevation of [Ca 2 ] i rather than localized Ca 2 microdomains. A preferential enhancement of the HCSP would mean that insulin secretion can be potentiated at substimulatory [Ca 2 ] i values upon activation of intracellular protein kinases (Jones et al., 1985, 1986) This mechanism may also explain how PMA can give rise to a slowly developing component of insulin secretion even at a subthreshold glucose concentration (Bozem et al., 1987). It is plausible that both nutrient and nonnutrient secretagogues might modulate insulin secretion by recruiting more granules into the HCSP through activation of PKA and PKC. We would like to thank Dr. Kevin Gillis for helpful suggestions to the manuscript and Mrs. X.P. Xu for cell preparation. This work was supported by National Science Foundation of China grants ( , , and ), and 973 Program of China (G and 2004CB720000) to T. Xu. We are also grateful for support from the Li Foundation. The laboratory of T. Xu is supported by the Partner Group Scheme of the Max Planck Institute for Biophysical Chemistry, Göttingen, and the Sinogerman Scientific Center. Olaf S. Andersen served as editor. Submitted: 27 April 2004 Accepted: 15 October Wan et al. 206 REFERENCES Ammala, C., F.M. Ashcroft, and P. Rorsman Calcium-independent potentiation of insulin release by cyclic AMP in single -cells. Nature. 363: Ashcroft, F.M., D.E. Harrison, and S.J. Ashcroft Glucose induces closure of single potassium channels in isolated rat pancreatic -cells. Nature. 312: Aspinwall, C.A., L. Huang, J.R. Lakey, and R.T. Kennedy Comparison of amperometric methods for detection of exocytosis from single pancreatic -cells of different species. Anal. Chem. 71: Barg, S., X. Ma, L. Eliasson, J. Galvanovskis, S.O. Gopel, S. Obermuller, J. Platzer, E. Renstrom, M. Trus, D. Atlas, et al Fast exocytosis with few Ca 2 channels in insulin-secreting mouse pancreatic -cells. Biophys. J. 81:3308. Bergsten, P Slow and fast oscillations of cytoplasmic Ca 2 in pancreatic islets correspond to pulsatile insulin release. Am. J. Physiol. 268:E282 E287. Bokvist, K., L. Eliasson, C. Ammala, E. Renstrom, and P. Rorsman Co-localization of L-type Ca 2 channels and insulin-containing secretory granules and its significance for the initiation of exocytosis in mouse pancreatic -cells. EMBO J. 14: Bozem, M., M. Nenquin, and J.C. Henquin The ionic, electrical, and secretory effects of protein kinase C activation in mouse pancreatic -cells: studies with a phorbol ester. Endocrinology. 121: Braun, M., A. Wendt, B. Birnir, J. Broman, L. Eliasson, J. Galvanovskis, J. Gromada, H. Mulder, and P. Rorsman Regulated exocytosis of GABA-containing synaptic-like microvesicles in pancreatic -cells. J. Gen. Physiol. 123: Carpenter, L., C.J. Mitchell, Z.Z. Xu, P. Poronnik, G.W. Both, and T.J. Biden PKC is activated but not required during glucose-induced insulin secretion from rat pancreatic islets. Diabetes. 53: Csukai, M., and D. Mochly-Rosen Pharmacologic modulation of protein kinase C isozymes: the role of RACKs and subcellular localisation. Pharmacol. Res. 39: Efanov, A.M., S.V. Zaitsev, and P.O. Berggren Inositol hexakisphosphate stimulates non-ca 2 -mediated and primes Ca 2 - mediated exocytosis of insulin by activation of protein kinase C. Proc. Natl. Acad. Sci. USA. 94: Eliasson, L., E. Renstrom, C. Ammala, P.O. Berggren, A.M. Bertorello, K. Bokvist, A. Chibalin, J.T. Deeney, P.R. Flatt, J. Gabel, et al PKC-dependent stimulation of exocytosis by sulfonylureas in pancreatic cells. Science. 271: Eliasson, L., E. Renstrom, W.G. Ding, P. Proks, and P. Rorsman Rapid ATP-dependent priming of secretory granules precedes Ca 2 -induced exocytosis in mouse pancreatic -cells. J. Physiol. 503(Pt 2): Gillis, K.D., R. Mossner, and E. Neher Protein kinase C enhances exocytosis from chromaffin cells by increasing the size of the readily releasable pool of secretory granules. Neuron. 16: Gschwendt, M., S. Dieterich, J. Rennecke, W. Kittstein, H.J. Mueller, and F.J. Johannes Inhibition of protein kinase C mu by various inhibitors. Differentiation from protein kinase C isoenzymes. FEBS Lett. 392: Heinemann, C., R.H. Chow, E. Neher, and R.S. Zucker Kinetics of the secretory response in bovine chromaffin cells following flash photolysis of caged Ca 2. Biophys. J. 67:2546. Horrigan, F.T., and R.J. Bookman Releasable pools and the kinetics of exocytosis in adrenal chromaffin cells. Neuron. 13: Jones, P.M., J.M. Fyles, and S.L. Howell Regulation of insulin secretion by camp in rat islets of Langerhans permeabilised by high-voltage discharge. FEBS Lett. 205: Jones, P.M., J. Stutchfield, and S.L. Howell Effects of Ca 2 and a phorbol ester on insulin secretion from islets of Langerhans permeabilised by high-voltage discharge. FEBS Lett. 191: Kaneto, H., K. Suzuma, A. Sharma, S. Bonner-Weir, G.L. King, and G.C. Weir Involvement of protein kinase C 2 in c-myc induction by high glucose in pancreatic -cells. J. Biol. Chem. 277: Kazanietz, M.G Novel nonkinase phorbol ester receptors: the C1 domain connection. Mol. Pharmacol. 61: Lang, J., M. Fukuda, H. Zhang, K. Mikoshiba, and C.B. Wollheim The first C2 domain of synaptotagmin is required for exocytosis of insulin from pancreatic -cells: action of synaptotagmin at low micromolar calcium. EMBO J. 16: Liu, W.S., and C.A. Heckman The sevenfold way of PKC regulation. Cell. Signal. 10: Nagy, G., U. Matti, R.B. Nehring, T. Binz, J. Rettig, E. Neher, and J.B. Sorensen Protein kinase C-dependent phosphorylation of synaptosome-associated protein of 25 kda at Ser187 potentiates vesicle recruitment. J. Neurosci. 22: Neher, E Cell physiology. Secretion without full fusion. Nature. 363: Nesher, R., E. Anteby, M. Yedovizky, N. Warwar, N. Kaiser, and E. Cerasi Cell protein kinases and the dynamics of the insu- Downloaded from on November 30, 2004

207 lin response to glucose. Diabetes. 51(Suppl 1):S68 S73. Olofsson, C.S., S.O. Gopel, S. Barg, J. Galvanovskis, X. Ma, A. Salehi, P. Rorsman, and L. Eliasson Fast insulin secretion reflects exocytosis of docked granules in mouse pancreatic -cells. Pflugers Arch. 444: Parsons, T.D., J.R. Coorssen, H. Horstmann, and W. Almers Docked granules, the exocytic burst, and the need for ATP hydrolysis in endocrine cells. Neuron. 15: Proks, P., L. Eliasson, C. Ammala, P. Rorsman, and F.M. Ashcroft Ca 2 - and GTP-dependent exocytosis in mouse pancreatic -cells involves both common and distinct steps. J. Physiol. 496(Pt 1): Renstrom, E., L. Eliasson, and P. Rorsman Protein kinase A-dependent and -independent stimulation of exocytosis by camp in mouse pancreatic -cells. J. Physiol. 502(Pt 1): Rhee, J.S., A. Betz, S. Pyott, K. Reim, F. Varoqueaux, I. Augustin, D. Hesse, T.C. Sudhof, M. Takahashi, C. Rosenmund, and N. Brose phorbol ester- and diacylglycerol-induced augmentation of transmitter release is mediated by Munc13s and not by PKCs. Cell. 108: Rorsman, P., and G. Trube Calcium and delayed potassium currents in mouse pancreatic -cells under voltage-clamp conditions. J. Physiol. 374: Rorsman, P., L. Eliasson, E. Renstrom, J. Gromada, S. Barg, and S. Gopel The cell physiology of biphasic insulin secretion. News Physiol. Sci. 15:72. Sharp, G.W Mechanisms of inhibition of insulin release. Am. J. Physiol. 271:C1781 C1799. Sheu, L., E.A. Pasyk, J. Ji, X. Huang, X. Gao, F. Varoqueaux, N. Brose, and H.Y. Gaisano Regulation of insulin exocytosis by Munc13-1. J. Biol. Chem. 278: Smith, C., T. Moser, T. Xu, and E. Neher Cytosolic Ca 2 acts by two separate pathways to modulate the supply of release-competent vesicles in chromaffin cells. Neuron. 20: Sorensen, J.B., U. Matti, S.H. Wei, R.B. Nehring, T. Voets, U. Ashery, T. Binz, E. Neher, and J. Rettig The SNARE protein SNAP-25 is linked to fast calcium triggering of exocytosis. Proc. Natl. Acad. Sci. USA. 99:1627. Stevens, C.F., and J.M. Sullivan Regulation of the readily releasable vesicle pool by protein kinase C. Neuron. 21: Takahashi, N., T. Kadowaki, Y. Yazaki, G.C. Ellis-Davies, Y. Miyashita, and H. Kasai Post-priming actions of ATP on Ca 2 -dependent exocytosis in pancreatic cells. Proc. Natl. Acad. Sci. USA. 96: Takahashi, N., T. Kadowaki, Y. Yazaki, Y. Miyashita, and H. Kasai Multiple exocytotic pathways in pancreatic cells. J. Cell Biol. 138: Thomas-Reetz, A.C., and P. De Camilli A role for synaptic vesicles in non-neuronal cells: clues from pancreatic cells and from chromaffin cells. FASEB J. 8: Verspohl, E.J., and A. Wienecke The role of protein kinase C in the desensitization of rat pancreatic islets to cholinergic stimulation. J. Endocrinol. 159: Voets, T Dissection of three Ca 2 -dependent steps leading to secretion in chromaffin cells from mouse adrenal slices. Neuron. 28: von Ruden, L., and E. Neher A Ca-dependent early step in the release of catecholamines from adrenal chromaffin cells. Science. 262: Wu, X.S., and L.G. Wu Protein kinase c increases the apparent affinity of the release machinery to Ca 2 by enhancing the release machinery downstream of the Ca 2 sensor. J. Neurosci. 21: Xu, T., T. Binz, H. Niemann, and E. Neher Multiple kinetic components of exocytosis distinguished by neurotoxin sensitivity. Nat. Neurosci. 1: Xu, T., M. Naraghi, H. Kang, and E. Neher Kinetic studies of Ca 2 binding and Ca 2 clearance in the cytosol of adrenal chromaffin cells. Biophys. J. 73:532. Yang, Y., and K.D. Gillis A highly Ca-sensitive pool of granules is regulated by glucose, PKC, and camp in insulin-secreting INS-1 cells. J. Gen. Physiol. 124: Yang, Y., S. Udayasankar, J. Dunning, P. Chen, and K.D. Gillis A highly Ca 2 -sensitive pool of vesicles is regulated by protein kinase C in adrenal chromaffin cells. Proc. Natl. Acad. Sci. USA. 99: Zhu, H., B. Hille, and T. Xu From the cover: sensitization of regulated exocytosis by protein kinase C. Proc. Natl. Acad. Sci. USA. 99: Downloaded from on November 30, Sensitization of Insulin Secretion 207

208 420 Biophysical Journal Volume 86 January Crystal Structure of the Hyperthermophilic Inorganic Pyrophosphatase from the Archaeon Pyrococcus horikoshii Binbin Liu,* z Mark Bartlam,* z Renjun Gao, y Weihong Zhou,* Hai Pang,* Yiwei Liu,* Yan Feng, y and Zihe Rao* z *Laboratory of Structural Biology, Tsinghua University and National Laboratory of Biomacromolecules, Institute of Biophysics, Chinese Academy of Sciences, Beijing, China; y Key Laboratory for Molecular Enzymology and Engineering of Ministry of Education, Jilin University, Changchun , China; and z National Laboratory of Macromolecules, Institute of Biophysics, Chinese Academy of Science, Beijing , China ABSTRACT A homolog to the eubacteria inorganic pyrophosphatase (PPase, EC ) was found in the genome of the hyperthermophilic archaeon Pyrococcus horikoshii. This inorganic pyrophosphatase (Pho-PPase) grows optimally at 888C. To understand the structural basis for the thermostability of Pho-PPase, we have determined the crystal structure to 2.66 Å resolution. The crystallographic asymmetric unit contains three monomers related by approximate threefold symmetry, and a hexamer is built up by twofold crystallographic symmetry. The main-chain fold of Pho-PPase is almost identical to that of the known crystal structure of the model from Sulfolobus acidocaldarius. A detailed comparison of the crystal structure of Pho- PPase with related structures from S. acidocaldarius, Thermus thermophilus, and Escherichia coli shows significant differences that may account for the difference in their thermostabilities. A reduction in thermolabile residues, additional aromatic residues, and more intimate association between subunits all contribute to the larger thermophilicity of Pho-PPase. In particular, deletions in two loops surrounding the active site help to stabilize its conformation, while ion-pair networks unique to Pho-PPase are located in the active site and near the C-terminus. The identification of structural features that make PPases more adaptable to extreme temperature should prove helpful for future biotechnology applications. Submitted June 12, 2003, and accepted for publication August 27, Binbin Liu and Mark Bartlam contributed equally to this work. Address reprint requests to Zihe Rao, Laboratory of Structural Biology, Dept. of Biological Science and Biotechnology, Tsinghua University, Beijing , P. R. China. Tel.: ; Fax: ; raozh@xtal.tsinghua.edu.cn. Ó 2004 by the Biophysical Society /04/01/420/08 $2.00 INTRODUCTION Inorganic pyrophosphatase (PPase, EC ) is an essential enzyme that specifically catalyzes the hydrolysis of the phosphoanhydride bond in inorganic phosphate (PPi) (Chen et al., 1990; Lahti, 1983). PPi is a central phosphorus metabolite and is a by-product of various reversible nucleoside 58-triphosphate dependent reactions including trna charging and DNA and protein synthesis that utilize ATP in vivo. PPases from a wide variety of sources have been studied, but those from Saccharomyces cerevisiae and Escherichia coli are the most highly characterized, both biochemically and structurally. E. coli pyrophosphatase is a homohexamer with 175 amino acids per monomer (Avaeva et al., 1997), whereas S. cerevisiae pyrophosphatase is a homodimer with 286 amino acids per monomer (Heikinheimo et al., 2001). PPases from archaebacterium exhibit different structural and catalytic properties (Hansen et al., 1999; Richter and Schafer, 1992). The archaeal PPases so far reported are relatively thermostable, especially in the presence of divalent metal cations (Ichiba et al., 1998). Understanding the structural basis for the enhanced stability of proteins from hyperthermophilic organisms relative to their mesophilic and thermophilic counterparts is a highly relevant but complex and challenging problem. Previous comparisons of highresolution crystal structures of enzymes with the same fold and function in mesophiles, thermophiles and hyperthermophiles have revealed a number of potentially stabilizing features. A proper understanding of the molecular basis of thermal stability in proteins could have important consequences for their application in a range of biotechnological processes. For example, thermostable pyrophosphatases have common uses in cycle sequencing methods using thermostable DNA polymerases (Vander Horn et al., 1997). The crystal structures of PPases from thermophilic bacterium Thermus thermophilus (T-PPase; PDB ID 2PRD) (Teplyakov et al., 1994), thermophilic archaebacterium Sulfolobus acidocaldarius (S-PPase; PDB ID 1QEZ) (Leppanen et al., 1999), and mesophile E. coli (E-PPase; PDB ID 1JFD) (Avaeva et al., 1997) have provided basic clues for the PPase catalytic mechanism and thermostability, but many important aspects remain to be resolved. Microorganisms can be classified according to their optimal growth temperature, T opt, into four groups: psychrophilic (0 \ T opt \ 208C), mesophilic (20 \ T opt \ 508C), thermophilic (50 \ T opt \ 808C), and hyperthermophilic (80 \ T opt \ 1208C). Considerable efforts have been made during recent years to analyze the structural features that determine the extraordinary thermal stability of proteins from hyperthermophiles. Here we have isolated an inorganic pyrophosphatase (Pho-PPase) from the hyperthermophilic archaeon Pyrococcus horikoshii OT3, whose optimum growth temperature (958C) is significantly higher than those of S. acidocaldarius (75 808C), T. thermophilus (75 808C) and E. coli (378C). Pho-PPase showed a higher optimal activity at 888C and an alkaline optimal ph of 10.3 (at 888C). The enzyme has extreme thermostability and does not lose 208

209 Crystal Structure of Pho-PPase 421 activity at 1008C. In addition, Pho-PPase is stable against various denaturants. All of these properties are different from those of other archaeal PPases: full details of the characterization of inorganic Pho-PPase will be reported elsewhere (Feng et al., unpublished results). To gain a more penetrating insight into its function, here we describe the structure determination of Pho-PPase and a comparison of the structure with its mesophilic and thermophilic counterparts in an attempt to understand the structural basis for thermal stability. MATERIALS AND METHODS Crystallization and x-ray data collection The preparation and preliminary characterization of Pho-PPase crystals have been described elsewhere (B. Liu X. Li, R. Gao, W. Zhou, G. Xie, M. Bartlam, H. Pang, Y. Feng, and Z. Rao, submitted). Briefly, the gene encoding inorganic pyrophosphatase from the archaea P. horikoshii was cloned into pet15b (Novagen) and expressed in E. coli strain BL21. After purification, the purified protein was concentrated to 20 mg/ml. Crystallization trials were set up using the hanging drop/vapor diffusion method with Crystal Screen reagent kits I and II (Hampton Research). Crystals suitable for diffraction were obtained after two weeks from the condition 3.8% PEG 4000, 0.1 M Na acetate, ph , 0.02 M MgCl 2. A set of data at 2.66 Å resolution were collected in house. All data were integrated using DENZO/ HKL and scaled and merged with SCALEPACK (Otwinowski and Minor, 1997). Crystal lattice properties and data-collection statistics are listed in Table 1. Structure determination and refinement The structure of Pho-PPase was solved by the molecular-replacement (MR) method using the protein S-PPase (PDB ID 1QEZ) (Leppanen et al., 1999) as the starting model. The initial R value for the MR solution obtained from the cross-rotation and translation search in CNS (Brunger et al., 1998) was 48.4%, using reflections in the Å resolution range. This R value is ;10% lower than those of the other possible solutions. After rigidbody refinement, the R value decreased to 46.8, and side-chain atoms were fitted into the 2 Fo - Fc electron-density map. The structure was further refined to 26.5% (using reflections in the resolution range Å) following cycles of simulated-annealing refinement using CNS and manual rebuilding in O (Jones et al., 1991). After the placement and refinement of 93 water molecules and individual B factor refinement, the R factor was reduced to 23.2%. A Ramachandran plot generated by PROCHECK (Laskowski et al., 1993) shows that the structure has reasonable stereochemistry with no residues in disallowed regions. Refinement statistics are summarized in Table 1. Coordinates for this structure have been deposited in the Protein Data Bank with PDB ID 1UDE. RESULTS AND DISCUSSION Monomeric structure The current model of P. horikoshii inorganic pyrophosphatase includes residues in chain A, in chain B, in chain C, and 93 water molecules. Pho-PPase is arranged in a globular form and belongs to the a1b class of protein folds. The Pho-PPase monomer structure is composed of nine b-strands and two a-helices arranged in a b1-b2-b3-b4-b5-b6-b7-b8-a1-b9-a2 topology (Fig. 1). TABLE 1 Data collection, refinement, and model statistics Data collection Space group P Unit cell parameters a ¼ 71.8, b ¼ 86.7, c ¼ 92.8 Å a ¼ b ¼ g ¼ 908 Matthews coefficient (Å 3 Da ÿ1 ) 1.93 Solvent content (%) Resolution (Å) Total observations Unique reflections Redundancy 6.93 Average I/s(I) 12.1 (4.3) R merge (%) 7.5 (13.9) Completeness (%) 99.8 (98.0) Refinement Reflections (observed) (test) 1687 Resolution range (Å) Protein atoms 4148 Solvent atoms 93 R work (%) R free (%) Average B value (Å 2 ) RMSD bonds (Å) RMSD angles (8) 1.6 Ramachandran plot (%) Most favored regions 76.9% Additionally allowed regions 19.6% Generously allowed regions 3.3% Disallowed regions 0.2% Pho-PPase shares 47% sequence identity with S-PPase (Fig. 2 A), and the two proteins have a similar core structure. Indeed, superposition of the Pho-PPase structure with S-PPase, T-PPase, and E-PPase shows that the four PPases are spatially homologous (Fig. 2 B). The RMSD between C a atoms of the four PPases range from 0.83 to 1.14 Å. The mutual positions of the central b-barrel structure and a-helices are similar. However, the structure-based sequence alignment (Fig. 2 A) shows that Pho-PPase contains a single residue deletion in the loop formed by residues 25 30, two residues deleted in the loop formed by residues , and three residues deleted in the loop formed by residues Oligomeric structure The Pho-PPase structure consists of three molecules in the asymmetric unit. The first subunit contains residues out of a total of 178 residues; the second subunit contains residues 4 170; and the third subunit contains residues with a Y170A substitution. Subunits 2 and 3 can be superimposed onto the first subunit with RMSDs of 0.61 Å and 0.65 Å, respectively. The three Pho-PPase monomers are related by a noncrystallographic threefold axis to form a tight trimer with an extensive subunit interface. As can been seen in Fig. 1 B, the main contact region concerns Biophysical Journal 86(1)

210 422 Liu et al. FIGURE 1 (A) A ribbon representation of the Pho-PPase monomer structure. The structure is colored from blue at the N-terminal to red at the C-terminus. Secondary structure elements have been labeled. (B) Top and side views of the Pho-PPase hexamer. Three subunits are related by a noncrystallographic threefold axis to form a trimer. Two trimers are related by a crystallographic twofold axis to form a tightly packed hexamer. strands b2 andb3 of the b-barrel of one subunit and a b-hairpin (residues 78 85) of another subunit. There are many hydrophobic and hydrophilic interactions between subunits to enhance the enzyme stability, which will be discussed in further detail below. From the crystallographic symmetry, the trimers are packed such as to form a tightly packed hexamer with twofold crystallographic symmetry. The intertrimer interactions in Pho-PPase are listed in Table 2. There are no largescale distortions in monomeric structure, and the increased thermostability of Pho-PPase likely results from an increase in both hydrophilic and hydrophobic interactions. Intertrimer interactions are concentrated in strand b3, helix a1, and the loop between b8 and a1. The overall surface area buried in the hexamer is 2342 Å 2, which is comparable to the 2430 Å 2 buried by the T-PPase hexamer and higher than the 2090 Å 2 buried by the E-PPase hexamer (Salminen et al., 1996). Active center The active site cavity of Pho-PPase is formed between the b-barrel and helix a1. The overall shape and size of the active site is very similar to those of E. coli and yeast PPases (Avaeva et al., 1997). It has been previously suggested that 17 residues might be involved in Mg 21 and PPi binding, 15 of which appear to be conserved in all sequences of soluble PPases known to date. Accordingly, all 15 of these conserved residues (E23, K31, E33, R44, Y52, Y56, D66, D68, D71, D98, D103, K105, Y140, K141, K146) also reside in the active-site cavity of Pho-PPase (Fig. 2). This implies that Pho-PPase will share the general catalytic mechanism of the PPase family in vivo, although the maximal enzyme activity of Pho-PPase is gained under a strong alkaline environment in vitro. Although Pho-PPase includes the conserved residues involved in Mg 21 binding, no Mg ion was observed in the Pho-PPase structure despite the addition of MgCl 2 during crystallization. Many crystals soaked in MgCl 2 were found to crack, which suggests that conformational changes may occur to prevent Mg 21 binding. A similar observation was made in the structure of E-PPase (Kankare et al., 1994). The crystal structure of Pho-PPase reveals that K102, equivalent to residue E101 in E-PPase, is important for the active-site cavity and is located in the highly conserved region including the essential catalytic residues (numbered E97, D102, and K104 in E-PPase). Mutagenesis studies of E-PPase have shown that the enzyme activity of mutants E99D, D102E, and K104E almost disappeared or decreased rapidly (Hyytia et al., 2001). However, when E101 is replaced by a more negative aspartic acid residue, the enzyme activity of the mutant was increased by 10%. Alignment with other PPases shows that the residue located in the site is conserved as either an acidic or neutral residue. We propose that K102 is related to the alkaline optimal ph (10.3) of Pho-PPase, since only in such an alkaline environment would K102 not be positive. Further site-directed mutagenesis studies are in progress to investigate the role of K102. Structural basis of thermostability in Pho-PPase The free energy of stabilization of globular proteins is rather small. It lies in the range from 30 to 65 kj/mol (Pfeil et al., 1986), which is equivalent to the energy contributed by a few hydrogen bonds, ion pairs, or hydrophobic interactions. The increase in free energy of stabilization observed for thermophilic proteins is of the same order of magnitude (Harris et al., 1980; Nojima et al., 1978). Recent structural studies also have identified several factors which are more often observed among thermophilic proteins and may account for their stability. These include an increased number of salt bridges or hydrogen bonds; optimized stability of helices, loops, and N- and C-termini; decreased solvent-exposed surface area; stronger interactions between the subunits in oligomers; and even an increased number of buried solvent molecules in hydrophilic cavities. We compare the present 2.66 Å resolution structure with three Biophysical Journal 86(1)

211 Crystal Structure of Pho-PPase 423 FIGURE 2 (A) A structure-based sequence alignment between Pho-PPase and three related PPase structures from T. thermophilus (T-PPase; PDB ID 2PRD) (Teplyakov et al., 1994), S. acidocaldarius (S-PPase; PDB ID 1QEZ) (Leppanen et al., 1999) and E. coli (E-PPase; PDB ID 1JFD) (Avaeva et al., 1997). Secondary structure elements are shown for Pho-PPase. (B) A stereodiagram showing the superposition of the four PPase structures. Structures are represented as C a backbone traces. The coloring is as follows: red, Pho- PPase; blue, S-PPase; green, T-PPase; magenta, E-PPase. RMSD between C a atoms of the four PPases range from 0.83 to 1.14 Å. other PPases, to identify which factors may be important in stabilizing Pho-PPase. Amino acid composition and thermostability The amino acid composition of a protein has long been thought to be correlated with its thermostability. Compared with related PPase structures, the unusual amino acid composition of Pho-PPase would account for its extreme thermostability. The structure-based sequence alignment (Fig. 2) shows the amino acid sequences of the four PPases compared in this study. More charged residues were found in Pho-PPase, which is consistent with the idea that the number of ion pairs is an important determinant of protein thermostability. For instance, Pho-PPase contains a total number of eight arginine residues, which have a tendency to form multiple ion-pairs and H-bonds. Pho-PPase also contains 14 proline residues, which is the largest proportion in the four PPases studied here. Since proline residues affect local mobility of the chain by decreasing the conformational entropy of the unfolded state, the increased rigidity of the structure would be expected to increase the overall thermostability. Such stabilization by the introduction of proline residues into loop regions is a well-documented phenomenon (Matthews et al., 1987). A recent study of inorganic pyrophosphatase from thermophilic bacterium PS-3 was carried out in which proline residues were systematically replaced by alanines (Masuda et al., 2002). The authors found that most of the proline residues in PS-3 PPase play very important roles, and many of them are critical for the structural integrity of the protein. They also concluded that the thermostability of PS-3 PPase is profoundly related with its subunit structure. The total number of hydrophobic residues (Gly, Ala, Val, Leu, Ile, Met, Phe, Trp, and Pro) is also highest in Pho- PPase. Compositional differences between the PPases are more pronounced among exposed sites. Pho-PPase has an increased number of exposed hydrophobic residues, which are presumably involved in oligomer formation and stability. Interestingly, previous site-directed mutagenesis of E-PPase has shown that aromatic residues play a very important role for the thermostability (Hyytia et al., 2001). There are an increased number of aromatic residues found in Pho-PPase compared to three other PPases. In particular, Biophysical Journal 86(1)

212 424 Liu et al. TABLE 2 Oligomeric interactions in E-PPase, T-PPase, S-PPase, and Pho-PPase E-PPase a distance T-PPase a distance S-PPase b distance Pho-PPase distance From To (Å) From To (Å) From To (Å) From To (Å) Intratrimeric A-B hydrophilic contacts S1 S N27 Y N16 K R28 E Y30 Q I39 V Y30 Y N29 E L39 V L41 V I39 V N29 E V41 V V44 E V41 V N29 P F44 L L45 R D42 T L40 I L45 R V44 T K41 E Q115 L Y46 K L42 I R44 E V45 G Intertrimeric A-D hydrophilic contacts S46 Q A48 F S48 N R28 E A48 F Q132 K S48 Og N T48 H H136 D H136 T H136 H P49 H H136 E H135 E R139 E R138 E Intertrimeric A-E hydrophilic contacts N24 Y R116 E R77 E R116 E R116 E Q115 Q Intratrimeric A-B hydrophobic contacts L2 A P6 L L5 E F4 L D26 Y Y30 V L5 V R28 E P27 Y I39 V I28 Y N29 E I28 Y L41 V I28 L L39 L Y30 S R43 E Y30 Y L40 T L39 S V44 Y K40 V K41 E F40 V V44 V V41 L L42 M V41 L V44 D V41 T L42 I V41 L L80 P V44 Y K44 E F44 T Y46 P V45 M F44 P Y80 P V45 D F44 S Y47 P P82 Y Intertrimeric A-D hydrophobic contacts P27 F A48 F T47 K R28 E S46 L A48 P S48 P T48 H T47 H A48 E S48 K T48 F A48 F Q49 H M49 M P49 F M49 M F50 F H136 L P49 E H136 L Q132 L H140 H P49 P H136 D T140 T F50 F H140 H H51 H H135 L R139 R Intertrimeric A-E hydrophobic contacts R116 E N76 N P114 A H119 P P115 A P114 A Y115 P Y115 A Biophysical Journal 86(1)

213 Crystal Structure of Pho-PPase 425 TABLE 3 Ionic interactions within a monomer E-PPase distance T-PPase distance S-PPase distance Pho-PPase distance From To (Å) From To (Å) From To (Å) From To (Å) E13 K K10 E D13 K E15 K D14 R E13 R K29 E K30 D D14 H K29 E D33 K K41 E K29 D D42 R E36 K R44 E R43 E R43 E R43 E R44 K D70 K E64 R R43 D D71 K K94 E D70 K E63 R R77 D D97 K E98 K E63 R K95 E D102 K E98 K E64 R K95 E H140 E D102 K D70 K D98 K H140 D D114 R E86 K D103 K E159 K D122 R R88 D K117 D E98 K E145 K D118 K E98 K E163 R K127 D ÿ4.09 D102 K E132 H H140 E R139 E E145 K R139 K E144 K E157 K E157 K E162 R E169 K there is an increased frequency of phenylalanine and tyrosine residues in Pho-PPase, which are liable to form hydrophobic and aromatic interactions. Many of these aromatic residues are observed to form a cluster located at the bottom of the active site, and it is possible that stacking interactions involving the aromatic residues may contribute to enhanced thermostability of Pho-PPase. The frequency of Asn (2) and Gln (0), which can be classed as thermolabile due to their tendency to undergo deamidation at high temperatures and therefore may be naturally discriminated against in thermostable proteins, is substantially reduced in Pho-PPase. Cysteine was also completely absent in Pho-PPase, which is easy to interpret since cysteine is highly sensitive to oxidation at high temperature. The frequency of glycine was not decreased, but changed in location. Interestingly, residue L83 of Pho- PPase is strictly conserved as a glycine in other PPases and is located in the short loop connecting b-strands 5 and 6, thus increasing the rigidity of the Pho-PPase structure. Ionic interactions and thermostability The importance of ion-pairs as determinants of protein thermostability was first highlighted by Perutz and Raidt (Perutz and Raidt, 1975) while comparing ferredoxin and hemoglobin structures, and ion-pairs were subsequently proposed to be important for the stability of a number of other thermostable proteins (Korndorfer et al., 1995; Walker et al., 1980). Indeed, several reports of high resolution structures of hyperthermophilic proteins show the number of ion-pairs in most of the hyperthermophilic proteins is higher than in their mesophilic counterparts (DeDecker et al., 1996; Hennig et al., 1995; Yip et al., 1995). A total of 12 ionic interactions are formed per monomer in the E-PPase short c-axis crystal form, which is equivalent to the number in T-PPase and lower than S-PPase (17 ionic pairs). In contrast, there are 28 ionic interactions scattered throughout the Pho-PPase monomer (Table 3). It appears that Pho-PPase is more stabilized by ion-pairs than T-PPase, E-PPase, and S-PPase. Although the number of ion-pairs is increased, the multicenter ion interactions are decreased. Interestingly, two long ion-pair networks are observed in Pho-PPase. The first ion-pair network is located in the C-terminus a2 helix, which may stabilize the C-terminus against thermal denaturation. The second ion network is located in the active center and involves residues and There is also an increase in the number of intrasubunit ion-pairs in Pho- PPase, and their involvement in complex networks mirrors that observed in the structure of the hyperthermophilic Pyrococcus furiosus glutamate dehydrogenase (Yip et al., 1995). The presence of ion-pair networks has also been observed in Sulfolobus solfataricus indole-3-glycerol phosphate synthase (Hennig et al., 1995) and the TATA-box binding protein from P. furiosus (DeDecker et al., 1996). The presence of ion-pair networks may be energetically favorable due to the shared entropic cost upon ion-pair formation (Nakamura, 1996). The positioning of ion-pairs is also crucial (Daggett and Levitt, 1993). Loop or random coil regions of a protein tend to be the most flexible areas and are the most likely to deform at high temperatures. Compared with the three other PPases used in this study, Pho-PPase has three shorter loops, of Biophysical Journal 86(1)

214 426 Liu et al. which two lie in the active site and the third sits at the molecular surface. In contrast, the loops found in E-PPase are longest of the four PPase structures. This observation, that the higher the growth temperature of the organism the shorter the protein loops, is consistent with high temperature molecular dynamic simulations on bovine pancreatic trypsin inhibitor, which revealed that loop and turn regions are likely to be the parts of the structure that unfold first during thermal denaturation. The structure of thermostable endocellulase (Sakon et al., 1996) also showed a reduced size of loop regions, and a similar strengthening of loop regions was also observed in S. solfataricus indole-3-glycerol phosphate synthase (Hennig et al., 1995). Oligomerization and thermostability In the PPase family, the oligomerization state stabilizes the conformation of the enzyme that binds substrate and vice versa (Baykov et al., 1995). The oligomeric packing appears to provide a general strategy for enhancing the thermostability of the PPase family. Pho-PPase seems to be a more tightly packed hexamer. The total number of oligomeric hydrophilic contacts is listed in Table 2. There are notably more intermonomer hydrophilic interactions in Pho-PPase compared with the three other PPases, resulting in tighter twofold A-E and threefold A-B interfaces. This effect can be measured by the accessible surface area (ASA) buried per monomer on oligomerization, which is 13.6% higher than for S-PPase and 12.1% higher than for E-PPase. These increased hydrophilic interactions may provide the extra energy necessary for stabilization. Similar observations are found in several thermophilic protein structures. For example, thermostability was attributed to improved subunit interfaces in L-lactate dehydrogenase from Bacillus stearothermophilus (Kallwass et al., 1992), malate dehydrogenase from Thermus flavus (Kelly et al., 1993), and ornithine carbamoyltransferase from P. furiosus (Villeret et al., 1998). Other studies also indicate that multimer formation and subunit interactions are critical for thermal stability of, for example, hemocyanin from the ancient tarantula Eurypelma californicum (Sterner et al., 1995), phosphoribosyl anthranilate isomerase from the hyperthermophile Thermotoga maritima (Hennig et al., 1997), GluDH from the hyperthermophile P. furiosus (Vetriani et al., 1998), and chorismate mutase from the thermophilic archaeon Methanococcus jannaschii (MacBeath et al., 1998). Helices and thermostability It has been previously reported that the helical conformation is stabilized by oppositely charged ion-pair interactions (i.e., Glu-Lys, Glu-Arg, Asp-Lys, Asp-Arg) in the positions (i,i14) or (i,i13) (Scholtz et al., 1993). The a1 helix contains a single ion-pair between the nonconserved residues K127-D131. Examination of the C-terminus a2 helix sequence indicates that Pho-PPase contains a significant increase in the number of charged residues compared with the other PPases (Fig. 2 A). These residues E157, R161, E162, R165, E168, K171 are ideally distributed in the sequence to allow the formation of intrahelix ion-pairs. The structure of Pho-PPase shows three intrahelical ion-pairs formed between residues E157-R161, E162-R165, and E169-K172 (Table 3), which are not conserved in other PPase structures. These additional ion-pairs may be responsible for the increased thermostability of Pho-PPase by stabilizing the C-terminus and increasing its resistance to denaturation. CONCLUSIONS Comparison of the structure of Pho-PPase with T-PPase, S-PPase, and E-PPase has identified a number of determining factors for the thermostability of PPase enzymes. Pho-PPase is the most thermostable of the four structures, and this is reflected by the following factors. First, shorter and more convergent loops in the active site imply that the two loops are important for enzyme-substrate or ion binding under high temperature, and this is consistent with the idea that shorter loops are more helpful for thermostable PPases to stabilize the conformation of the enzyme-substrate complex. Second, the basis of thermostability in PPases is generally believed to be related to the oligomer structure; Pho-PPase has an increase in the number of both intermonomer and intertrimer interactions, and this results in tighter packing of the hexamer. Third, an increase in the number of ion-pairs in particular, two ion-pair networks helps to stabilize the structure. One ion-pair network in the loop connecting strand b8 and helix a1 may help to stabilize the conformation of the active site, while another network in helix a2 may stabilize the C-terminus and increase its resistance to thermal denaturation. The structure of Pho-PPase and the comparison between PPase structures should stimulate further kinetic and structural studies of enzymes adapted to extreme temperature and prove helpful for future biotechnology applications. We are grateful to Feng Gao for help with data collection. This research was supported by the following grants: Project 973 G ; Ministry of Science and Technology (MOST) 2002BA711A12. REFERENCES Avaeva, S., S. Kurilova, T. Nazarova, E. Rodina, N. Vorobyeva, V. Sklyankina, O. Grigorjeva, E. Harutyunyan, V. Oganessyan, K. Wilson, M. Dauter, R. Huber, and T. Mather Crystal structure of Escherichia coli inorganic pyrophosphatase complexed with SO4(2-). Ligand-induced molecular asymmetry. FEBS Lett. 410: Baykov, A. A., V. Y. Dudarenkov, J. Kapyla, T. Salminen, T. Hyytia, V. N. Kasho, S. Husgafvel, B. S. Cooperman, A. Goldman, and R. Lahti Dissociation of hexameric Escherichia coli inorganic pyrophosphatase into trimers on His-136!Gln or His-140!Gln substitution and its effect on enzyme catalytic properties. J Biol Chem. 270: Biophysical Journal 86(1)

215 Crystal Structure of Pho-PPase 427 Brunger, A. T., P. D. Adams, G. M. Clore, W. L. DeLano, P. Gros, R. W. Grosse-Kunstleve, J. S. Jiang, J. Kuszewski, M. Nilges, N. S. Pannu, R. J. Read, L. M. Rice, T. Simonson, and G. L. Warren Crystallography & NMR system: a new software suite for macromolecular structure determination. Acta Crystallogr. D. 54: Chen, J., A. Brevet, M. Fromant, F. Leveque, J. M. Schmitter, S. Blanquet, and P. Plateau Pyrophosphatase is essential for growth of Escherichia coli. J. Bacteriol. 172: Daggett, V., and M. Levitt Protein unfolding pathways explored through molecular dynamics simulations. J. Mol. Biol. 232: DeDecker, B. S., R. O Brien, P. J. Fleming, J. H. Geiger, S. P. Jackson, and P. B. Sigler The crystal structure of a hyperthermophilic archaeal TATA-box binding protein. J. Mol. Biol. 264: Hansen, T., C. Urbanke, V. M. Leppanen, A. Goldman, K. Brandenburg, and G. Schafer The extreme thermostable pyrophosphatase from Sulfolobus acidocaldarius: enzymatic and comparative biophysical characterization. Arch. Biochem. Biophys. 363: Harris, J. I., J. D. Hocking, M. J. Runswick, K. Suzuki, and J. E. Walker D-glyceraldehyde-3-phosphate dehydrogenase. The purification and characterisation of the enzyme from the thermophiles Bacillus stearothermophilus and Thermus aquaticus. Eur. J. Biochem. 108: Heikinheimo, P., V. Tuominen, A. K. Ahonen, A. Teplyakov, B. S. Cooperman, A. A. Baykov, R. Lahti, and A. Goldman Toward a quantum-mechanical description of metal-assisted phosphoryl transfer in pyrophosphatase. Proc. Natl. Acad. Sci. USA. 98: Hennig, M., B. Darimont, R. Sterner, K. Kirschner, and J. N. Jansonius A structure of indole-3-glycerol phosphate synthase from the hyperthermophile Sulfolobus solfataricus: possible determinants of protein stability. Structure. 3: Hennig, M., R. Sterner, K. Kirschner, and J. N. Jansonius Crystal structure at 2.0 A resolution of phosphoribosyl anthranilate isomerase from the hyperthermophile Thermotoga maritima: possible determinants of protein stability. Biochemistry. 36: Hyytia, T., P. Halonen, A. Salminen, A. Goldman, R. Lahti, and B. S. Cooperman Ligand binding sites in Escherichia coli inorganic pyrophosphatase: effects of active site mutations. Biochemistry. 40: Ichiba, T., T. Shibasaki, E. Iizuka, A. Hachimori, and T. Samejima Cation-induced thermostability of yeast and Escherichia coli pyrophosphatases. Biochem. Cell Biol. 66: Jones, T. A., J. Y. Zou, S. W. Cowan, and M. Kjeldgaard Improved methods for building protein models in electron density maps and the location of errors in these models. Acta Crystallogr. A. 47: Kallwass, H. K., W. K. Surewicz, W. Parris, E. L. Macfarlane, M. A. Luyten, C. M. Kay, M. Gold, and J. B. Jones Single amino acid substitutions can further increase the stability of a thermophilic L-lactate dehydrogenase. Protein Eng. 5: Kankare, J., G. S. Neal, T. Salminen, T. Glumhoff, B. S. Cooperman, R. Lahti, and A. Goldman The structure of E. coli soluble inorganic pyrophosphatase at 2.7 A resolution. Protein Eng. 7: Kelly, C. A., M. Nishiyama, Y. Ohnishi, T. Beppu, and J. J. Birktoft Determinants of protein thermostability observed in the 1.9-A crystal structure of malate dehydrogenase from the thermophilic bacterium Thermus flavus. Biochemistry. 32: Korndorfer, I., B. Steipe, R. Huber, A. Tomschy, and R. Jaenicke The crystal structure of holo-glyceraldehyde-3-phosphate dehydrogenase from the thermophilic bacterium Thermotoga aquaticus DNA polymerase at 2.5 A resolution. J. Mol. Biol. 246: Lahti, R Microbial inorganic pyrophosphatases. Microbiol. Rev. 47: Laskowski, R. A., M. W. MacArthur, D. S. Moss, and J. M. Thornton PROCHECK: a program to check the stereochemical quality of protein structures. J. Appl. Crystallogr. 26: Leppanen, V. M., H. Nummelin, T. Hansen, and R. Lahti Sulfolobus acidocaldarius inorganic pyrophosphatase: structure, thermostability, and effect of metal ion in an archael pyrophosphatase. Protein Sci. 8: MacBeath, G., P. Kast, and D. Hilvert A small, thermostable, and monofunctional chorismate mutase from the archaeon Methanococcus jannaschii. Biochemistry. 37: Masuda, H., T. Uchiumi, M. Wada, T. Ichiba, and A. Hachimori Effects of replacement of prolines with alanines on the catalytic activity and thermostability of inorganic pyrophosphatase from thermophilic bacterium PS-3. J. Biochem. (Tokyo). 131: Matthews, B. W., H. Nicholson, and W. J. Becktel Enhanced protein thermostability from site-directed mutations that decrease the entropy of unfolding. Proc. Natl. Acad. Sci. USA. 84: Nakamura, H Roles of electrostatic interaction in proteins. Q. Rev. Biophys. 29:1 90. Nojima, H., K. Hon-Nami, T. Oshima, and H. Noda Reversible thermal unfolding of thermostable cytochrome c-552. J. Mol. Biol. 122: Otwinowski, Z., and W. Minor Processing of x-ray diffraction data collected in oscillation mode. In: Macromolecular Crystallography, part A. C. W. Carter Jr. and R. M. Sweet, editors. Academic Press Perutz, M., and H. Raidt Stereochemical basis of heat thermostability in bacterial ferredoxins and in hemoglobin A2. Nature. 255: Pfeil, W., V. E. Bychkova, and O. B. Ptitsyn Physical nature of the phase transition in globular proteins. Calorimetric study of human alphalactalbumin. FEBS Lett. 198: Richter, O. M., and G. Schafer Purification and enzymic characterization of the cytoplasmic pyrophosphatase from the thermoacidophilic archaebacterium Thermoplasma acidophilum. Eur. J. Biochem. 209: Sakon, J., W. S. Adney, M. E. Himmel, S. R. Thomas, and P. A. Karplus Crystal structure of thermostable family 5 endocellulase E1 from Acidothermus cellulolyticus in complex with cellotetraose. Biochemistry. 35: Salminen, T., A. Teplyakov, J. Kankare, B. S. Cooperman, R. Lahti, and A. Goldman An unusual route to thermostability disclosed by the comparison of Thermus thermophilus and Escherichia coli inorganic pyrophosphatases. Protein Sci. 5: Scholtz, J. M., H. Qian, V. H. Robbins, and R. L. Baldwin The energetics of ion-pair and hydrogen-bonding interactions in a helical peptide. Biochemistry. 32: Sterner, R., T. Vogl, H. J. Hinz, F. Penz, R. Hoff, R. Foll, and H. Decker Extreme thermostability of tarantula hemocyanin. FEBS Lett. 364:9 12. Teplyakov, A., G. Obmolova, K. S. Wilson, K. Ishii, H. Kaji, T. Samejima, and I. Kuranova Crystal structure of inorganic pyrophosphatase from Thermus thermophilus. Protein Sci. 3: Vander Horn, P. B., M. C. Davis, J. J. Cunniff, C. Ruan, B. F. McArdle, S. B. Samols, J. Szasz, G. Hu, K. M. Hujer, S. T. Domke, S. R. Brummet, R. B. Moffet, and C. W. Fuller Thermo Sequenase DNA polymerase and T. acidophilum pyrophosphatase: new thermostable enzymes for DNA sequencing. Biotechniques 22(4): , Vetriani, C., D. L. Maeder, N. Tolliday, K. S. Yip, T. J. Stillman, K. L. Britton, D. W. Rice, H. H. Klump, and F. T. Robb Protein thermostability above 100 degreesc: a key role for ionic interactions. Proc. Natl. Acad. Sci. USA. 95: Villeret, V., B. Clantin, C. Tricot, C. Legrain, M. Roovers, V. Stalon, N. Glansdorff, and J. Van Beeumen The crystal structure of Pyrococcus furiosus ornithine carbamoyltransferase reveals a key role for oligomerization in enzyme stability at extremely high temperatures. Proc. Natl. Acad. Sci. USA. 95: Walker, J. E., A. J. Wonacott, and J. I. Harris Heat stability of a tetrameric enzyme, D-glyceraldehyde-3-phosphate dehydrogenase. Eur. J. Biochem. 108: Yip, K. S., T. J. Stillman, K. L. Britton, P. J. Artymiuk, P. J. Baker, S. E. Sedelnikova, P. C. Engel, A. Pasquo, R. Chiaraluce, and V. Consalvi The structure of Pyrococcus furiosus glutamate dehydrogenase reveals a key role for ion-pair networks in maintaining enzyme stability at extreme temperatures. Structure. 3: Biophysical Journal 86(1)

216 Biophysical Journal Volume 87 September Three-Dimensional Tracking of Single Secretory Granules in Live PC12 Cells Dongdong Li,* Jun Xiong,* Anlian Qu,* and Tao Xu* y *Institute of Biophysics and Biochemistry, School of Life Science and Technology, Huazhong University of Science and Technology, Wuhan , People s Republic of China; and y National Laboratory of Biomacromolecules, Institute of Biophysics, Chinese Academy of Sciences, Beijing , People s Republic of China ABSTRACT Deconvolution wide-field fluorescence microscopy and single-particle tracking were used to study the threedimensional mobility of single secretory granules in live PC12 cells. Acridine orange-labeled granules were found to travel primarily in random and caged diffusion, whereas only a small fraction of granules traveled in directed fashion. High K 1 stimulation increased significantly the percentage of granules traveling in directed fashion. By dividing granules into the nearmembrane group (within 1 mm from the plasma membrane) and cytosolic group, we have revealed significant differences between these two groups of granules in their mobility. The mobility of these two groups of granules is also differentially affected by disruption of F-actin, suggesting different mechanisms are involved in the motion of the two groups of granules. Our results demonstrate that combined deconvolution and single-particle tracking may find its application in three-dimensional tracking of long-term motion of granules and elucidating the underlying mechanisms. INTRODUCTION Vesicular trafficking constitutes a major means for intracellular and transmembrane cargo transportation. For example, hormones and neurotransmitters are stored in vesicles and are released upon fusion with the plasma membrane. Before a vesicle can fuse with the plasma membrane, it has to undergo a series of steps such as budding, translocation, tethering, docking, and priming (Pfeffer, 1999; Toonen and Verhage, 2003; Jahn et al., 2003). To better understand the mechanisms of different steps in vesicular trafficking, it is desirable to directly visualize the movement of vesicles inside living cells. Two-dimensional (2-D) confocal microscopy has been used to explore granule movement in single optical section (Pouli et al., 1998; Levitan, 1998). The contribution from the missing z direction is ignored by assuming that observed granules move within the observation plane during the experimental time. Three-dimensional (3-D) single-particle tracking (SPT) in living cells has been difficult for confocal fluorescence microscopy. This probably is due to low time resolution inherited with the laser scanning, high photobleaching, and toxicity at the expense of low light collection efficiency. Recently developed total internal reflection fluorescence microscope (TIRFM) has proved very successful in studying the movement of single granules. The advantages of low background, high temporal resolution, minimal photobleaching, and phototoxicity have made TIRFM very popular in tracking the docking and fusion of granules (Steyer and Almers, 1999; Oheim and Stühmer, 2000; Ohara-Imaizumi et al., 2002). However, as the penetration depth of the evanescent field is limited to several hundreds of nanometers, TIRFM can only be employed to study granules right underneath the plasma membrane close to the coverglass-solution interface, whereas granules residing deeper inside the cytosol are invisible under TIRFM. It is unclear whether the unphysiological adhesion of cell membrane to the coverglass will have an impact on the mobility and function of secretory granules. On the other hand, although TIRFM is generally used for 2-D tracking, z-mobility can be indirectly inferred from the changes in the intensity of fluorescence (Ölveczky et al., 1997; Johns et al., 2001). However, since TIRFM is extremely sensitive to the objects vertical movement, fluctuation in fluorescence unrelated to changes in axial position (such as variation in excitation power, quenching, or dequenching of fluorescence due to ph change, etc.) will result in misinterpretation of the z-position. To avoid the restriction of TIRFM in observing only the cell-glass interface and to expand our knowledge on 3-D movement of granules throughout living cells, we have constructed a system to combine deconvolution wide-field fluorescence microscopy (WFFM) and 3-D single-particle tracking technique. We have employed this method to study the 3-D mobility of single granules in living PC12 cells. Indeed, 3-D mobility of granules distant from the plasma membrane has yet to be demonstrated. We have compared the mobility of granules that are either close to the plasma membrane (GP) or located deep inside the cytosol (GC). Our data illustrated that the mobility of internal granules is greater than those near the plasma membrane. Whereas the majority of granules wander in random and caged fashion, stimulation increases significantly the percentage of granules that travel in a directed fashion. We also found that the mobility of GP is more prone to the disruption of F-actin, Submitted March 19, 2004, and accepted for publication May 25, Address reprint requests to Tao Xu, txu@mail.hust.edu.cn; or Anlian Qu, alqu@mail.hust.edu.cn. Ó 2004 by the Biophysical Society /04/09/1991/11 $2.00 doi: /biophysj

217 1992 Li et al. whereas that of GC is less affected. Our results demonstrate that combining deconvolution and 3-D SPT is useful in following the long-term 3-D movement of single granules within living cells. MATERIALS AND METHODS Cells, labeling, and solutions PC12 cells were cultured as described (Lang et al., 1997). Cells were used shortly after being transferred to homemade coverslip chambers. Granules were labeled by incubating the cells in a buffer containing 3 mm acridine orange (AO) (Molecular Probes, Eugene, OR) for 15 min at 21 C. Then the cells were washed twice in dye-free buffer, which contained (in mm): 150 NaCl, 5.4 KCl, 2 MgCl 2, 1.8 CaCl 2, and 10 HEPES (ph 7.4). Cells were viewed ;60 min later. To verify our determination of the membrane contour, cells were first incubated in normal external solution containing 4 mm FM1-43 (Molecular Probes, Eugene, OR) for 5 10 min until we saw a stable fluorescence staining that was uniformly distributed along the plasma membrane. After locating the plasma membrane, FM1-43 was washed out from the membrane in dye-free buffer, which resulted in a rapid disappearing of the staining. Then, for the same cell we labeled its granules with AO as described above. To avoid unnecessary photobleaching, fluorescence excitation was turned on only during image acquisition. [K 1 ] o was raised by local perfusion with a multi-channel perfusion system (MPS-1, YiBo Life Science Instrument, Wuhan, China). The stimulation buffer contained (in mm): 90.4 NaCl, 65 KCl, 2 MgCl 2, 1.8 CaCl 2, and 10 HEPES (ph 7.4). Sometimes, dense-core granules were labeled by transient transfection with human pro-neuropeptide Y (NPY) (plasmid kindly provided by W. Almers, Vollum Institute, Oregon Health and Science University, Portland, OR) fused to the N-terminal of DsRed, which would be used to compare with AO-labeled granules in size, but not used for successive 3-D imaging due to the longer exposure time required for imaging DsRed labeled granules. Actin cytoskeletons were disrupted by incubating the cells with 5 mm latrunculin B (ICN Biomedicals, Aurora, Ohio) for 2 min. Image collection Cells were grown on high refractive-index glass coverslips and viewed under an inverted microscope (IX70; Olympus America, Melville, NY) with a 1.65 numerical-aperture (NA) objective (APO 3100 O HR, Olympus). Excitation light from a fiber optical-coupled monochromator (Polychrome IV; TILL Photonics, Bayern, Germany) was passed through a shutter that opened only during camera exposure. The wavelength selection and shutter were controlled by the image acquiring software (TILL vision 4.0, Till Photonics). Images were acquired with a cooled charge-coupled device (PCO SensiCam, Kelheim, Germany) with pixel size of mm at the specimen plane. Appropriate dichroic mirror (505 DCLP from Chroma Technology, Brattleboro, VT) and emission filter (535LP from Chroma) were used for imaging. A series of 2-D images were collected via moving the focal plane through a cell with the piezoelectric z axis controller (E-662. LR, Physik Instrumente, Karlsruhe, Germany). The focal plane advanced ;75% of the distance moved by the objective due to the discrepancy in refractive index between the immersion oil and cytosol. Accordingly, we reconstructed 3-D images of cells with those stacks of 2-D sections. For obtaining successive 3-D images of a cell, we selected the plane where the outmost edge of the cell appeared sharpest as the reference plane. Starting from the reference plane, we sampled sixteen 2-D sections with 0.2 mm stepping size to generate one 3-D image. An illustration of the imaged region is shown in Fig. 1 D between the dashed lines. As to the same imaged region, we successively recorded its 3-D images every 5 s for 70 s to generate a stack including fifteen 3-D images. In our imaging protocol, exposure time for each optical section was 5 ms, and the waiting interval for the next section was 170 ms. Occasionally, more sections were sampled to include the whole cell. Point spread function and deconvolution Knowing the point spread function (PSF) of the imaging system is the prerequirement for deconvolution. For our 1.65-NA objective, we determined the 3-D PSF through either experimental or theoretical method. In experimental measurement, we dried subresolution fluorescent beads (PS- Speck; diameter, mm; Molecular Probes) on the coverslip (refractive FIGURE 1 Point spread function and deconvolution. (A) Comparison of jotf (u, v)j distribution of corresponding equal-defocus (0.4 mm) optical sections between experimental and theoretical PSF. Experimental PSF was averaged from eight measurements, which gave quite similar OTF as that of theoretical PSF (P ¼ 0.98, KS test). (B) OTF distribution of the point light source from different defocus sections on the focal plane. (C) Evaluation of deconvolution result using different iterations and high-pass filtering; 15-mm beads with a spherical shell of fluorescence ; mm thick were used for the evaluation. Equatorial sections were used to measure the width of the shell (n ¼ 7 beads). Deconvolution removes the out-offocus blur and gives better estimates of the width. DC stands for deconvolution, and HP for high-pass filtering at a spatial frequency of 1/mm. (D) Selected deconvolved 3-D images of a PC-12 cell from stacks of images, wherein secretory granules were labeled by AO. Left panel, lateral sections with a distance of 1 mminz direction. Right panel, axial sections with a distance of 3 mminy direction. Dashed lines depict the 3-D region we sampled for analysis at a z-step interval of 0.2 mm. Scale bars, 1 mm. Biophysical Journal 87(3)

218 Three-Dimensional Mobility of Granules 1993 index, 1.78; thickness, 0.15 mm; Olympus) and then collected its 3-D images at different focal distances. According to the parameters of our imaging system, the theoretical PSF was derived from the mathematic model described in Gibson and Lanni, The iterative expectation-maximum algorithm based on a maximumlikelihood approach (Conchello and McNally, 1996) was used to deconvolve every recorded 3-D fluorescence image. The deconvolution algorithm used in our study was regularized with intensity regularization that can avoid the appearance of some artificial bright spots in deconvolved images (Conchello and McNally, 1996; Markham and Conchello, 1997). To verify the result of deconvolution, we employed a well-defined 3-D specimen, FocalCheck Fluorescent Microsphere from Molecular Probes. The bead is a spherical shell of fluorescence with a thickness of mm as provided by Molecular Probes. Evaluation of 3-D SPT with simulation 2-D SPT has been widely applied to monitor subpixel displacements of individual fluorescence particles between successive images, and some evaluation frameworks have been performed (Gosh and Webb, 1994; Kues et al., 2001; Cheezum et al., 2001; Thompson et al., 2002). The rationale of subpixel displacement detection is that fluorescence of single particles spreads over more than one pixel in recorded images, so the subpixel displacement can be tracked by weighting the fluorescence distribution from multiple pixels between successive time-lapse images. Here, we calculated centroids of successive images to estimate subpixel displacements in three dimensions. The centroid of a single axis is m C x ¼ + i¼1 n + j¼1 t. + ðx i 3 I ijk Þ k¼1 m + i¼1 n + j¼1 t + I ijk ; (1) k¼1 in which x i is the coordinates of a pixel on the x axis, and I ijk denotes the fluorescence intensity of the corresponding pixel. A threshold was defined as the fraction of the maximum fluorescence intensity, and those normalized pixels below the threshold were appointed zero. We applied computergenerated granule trajectories and movement, based on the measured parameters, to assess the influences of different signal/noise ratios (SNRs) and thresholds on SPT performance. The simulated granule, with similar size with actual granules (120 nm in diameter, Tooze et al., 1991), was initially created in one high-resolution matrix whose cell size was 0.01 mm mm mm(x, y, z), which was then convolved with the 3-D PSF mentioned above. The simulated CCD image was acquired by covering the same physical extent of the PSF-convolved high-resolution matrix with a low-resolution matrix whose cell size was set to 0.07 mm mm mm (x, y, z), which approximates the real sampling size used in our experiments. Subsequently, shot noise after a Poisson distribution was added for every pixel. For a particular SNR, several stacks, each of which contained sequential 3-D images of the simulated granule, were generated with computer. Subpixel displacements of the simulated granule between successive 3-D images were achieved through randomly moving granule along three dimensions in high-resolution matrix. The mean-square displacement (MSD) is of critical importance for assessing the parameters of granule s 3-D mobility, so we evaluated the performance of our SPT algorithm for estimating the MSD of the simulated granule at different SNR levels. The resemblance between SPT-estimated MSD and the actual one was assessed by the P-values given in the Kolmogorov-Smirnov (KS) test. For every given SNR, the optimized threshold for each simulated granule tracking was determined by maximizing the P-values (see Fig. 2). In addition, the bias and standard deviation (STD) was calculated and used as another indication of tracking accuracy for the selection of appropriate threshold: bias ¼ Æa ÿ âæ STD ¼ Æða ÿ ÆaæÞ 2 æ 1=2 : (2) FIGURE 2 SNR and threshold selection are critical for the performance of SPT. (A) Simulated granule (with SNR ¼ 9) trajectories based on measured parameters were used to evaluate the influence of different threshold levels (labeled in figure) on the estimation of MSD. The resemblance between SPT-estimated MSD and the actual one was assessed by the P-values given in the Kolmogorov-Smirnov (KS) test, with a maximum P-value indicating the best resemblance. (B) Assessing mean P-values and bias between estimated and actual MSD under different threshold levels for SNR ¼ 9 (n ¼ 7 simulated trajectories). Dashed line indicates a P-value of 0.1. The maximum P-value is attained at a threshold of (C) Threshold selection is dependent on the SNR of each granule. For every given SNR, the optimized threshold for each simulated granule tracking was determined by maximizing the P-values. Averaged maximal P-values and the corresponding biases are displayed against different SNRs. Larger SNRs give better SPT performance with larger P-values and less biases. Error for the bias was given with STD for a given threshold. Both SPT and simulation program were written in MATLAB 6.1. Mobility analysis of granules residing in different subcellular regions For comparing the mobility of granules residing close to the plasma membrane and the ones that are deep inside the cytosol, we divided the total population of granules into two groups according to their relative distances from the plasma membrane. One group included granules localized,1 mm from the contour of the plasma membrane, termed GP. The rest of granules inside the cytosol were termed GC. The edge-detecting algorithm (first described by Canny, 1986) implemented in MATLAB 6.1 software was employed to detect the membrane contour from the deconvolved image of AO-loaded cells (see Results section for more details). The mobility of each granule can be analyzed by determining its MSD as a function of time interval ndt. MSD in three dimensions can be calculated as follows: MSDðnDtÞ ¼ 1 Nÿn N ÿ n + f½xð jdt 1 ndtþÿxðjdtþš 2 j¼1 1 ½yð jdt 1 ndtþÿyðjdtþš 2 1 ½zð jdt 1 ndtþÿzðjdtþš 2 g; (3) Biophysical Journal 87(3)

219 1994 Li et al. where n ¼ 1,2...(N ÿ 1), N is the total number of images in the recorded sequence, Dt is the time interval between two successive 3-D images in a stack, x(jdt), y(jdt) and z(jdt) are the coordinates of the granule at time jdt, and x(jdt 1 ndt), y(jdt 1 ndt), and z(jdt 1 ndt) are the coordinates of the granule in another image taken ndt later. For random diffusion, granules move with a single 3-D diffusion coefficient (D (3) ): MSDðnDtÞ ¼6D ð3þ ðndtþ: (4) In directed diffusion, a drift velocity V 2 ¼ v 2 x 1 v2 y 1 v2 z is superimposed on random diffusion: MSDðnDtÞ ¼6D ð3þ ðndtþ 1 V 2 ðndtþ 2 : (5) The feature of diffusion limited in a cage, termed caged diffusion, can be characterized by the following approximate equation (Saxton and Jacobson, 1997): MSDðnDtÞ ¼R 2 ½1 ÿ A 1 expðÿ6a 2 D ð3þ ndt=r 2 ÞŠ; (6) where A 1 ¼ 0.99 and A 2 ¼ 0.85, and R is the radius of the spherical cage in which the particle is free to diffuse with D (3). To distinguish the proportion of granules diffusing with different fashions, we analyzed granules by fitting their MSDs with Eqs. 4 6 and then selected the best fits by KS test, and sometimes with the assistant of x 2 value. The scope of movement of a granule is defined as the mean value of its displacements in three dimensions during the observation period. Statistics For normally distributed data, population averages were expressed as mean 6 SE unless otherwise stated, and statistic significance was assessed by the Student s t-test. STD was used in Fig. 3 C. Skewed distribution was confirmed by Fisher equation (Becherer et al., 2003). The median and median standard error (MSE) was used to describe skewed distributed data. Statistical significance of the difference between two skewed distributions was assessed with KS test, x 2 value, or both. P, 0.05 and P, 0.01 were denoted as * and **, respectively. RESULTS Deconvolution reduces out-of-focus fluorescence in WFFM images We first determined the experimental and theoretical PSF of our 1.65-NA objective as described in the Materials and Methods section. For quantitative comparison of the experimental and theoretical PSF, we calculated and compared the optical transfer function (OTF) of each section planes based on the respective PSFs. We found that the jotfj of experimental PSF at given section planes agreed quite well with that of theoretical PSF. As shown in Fig. 1 A, the jotfjs at defocus of 0.4 mm for the two PSFs were quite similar (P ¼ 0.98, KS test). Similar analyses were done for sections with defocus,1 mm. We have further confirmed that MSD plots calculated from images processed with either experimental or theoretical PSF are quite similar (data not shown). Thus, for simplicity, we employed theoretical PSF for subsequent deconvolution calculation. Next, we evaluate how point light sources at different axial planes influence the image of focal plane. As depicted in Fig. 1 B, when a point light source is relatively far from the focal plane (open circle), it contributes mainly to the low spatial frequency signals at the focal plane; whereas high spatial frequency fluorescence comes mainly from light sources at adjacent FIGURE 3 Selection of fluorescence spots. (A) Left, examples of AO-labeled particles after different image processing. Right, axial and lateral line profile across the centroid of selected fluorescence spot (marked by the white arrow in the left panel ) is plotted against corresponding distance. Each line profile was fitted with single Gaussian distribution. The spot size was determined as the FWHM of the Gaussian fit (marked with the dashed lines). Scale bars, 1 mm. (B) Size distribution of 188 AO-labeled fluorescent spots from images processed by the deconvolution 1 high-pass filtering method. Only fluorescent spots with SNR $ 6 were selected for analysis. (C) Comparison of the size of AO-labeled spots and NPY-DsRed labeled large densecore granules. NY-DsRed transfected cells were imaged with 100 ms exposure time, which gave equivalent SNR to that of AO-labeled particles with only 5 ms exposure time (right panel). Error bars represent STD. Biophysical Journal 87(3)

220 Three-Dimensional Mobility of Granules 1995 optical sections. Thus, applying high-pass filtering to the images could help to remove out-of-focus fluorescence. Before applying deconvolution to live cell images, we assessed the algorithm using well-defined 15-mm diameter beads with a thin fluorescent shell of mm in width. As would be expected, sections in unprocessed 3-D images of the bead were seriously blurred by out-of-focus light (inset in Fig. 1 C), and the width of shell was much larger than its actual size. Deconvolution computationally reduced the outof-focus blurring, and gradually gave better estimates of the width of the shell with increasing iteration times (Fig. 1 C). Since low spatial frequency signal (,1/mm) contributes little to the analysis of the mobility of granules that are generally,1 mm in size (Steyer and Almers, 1999; Oheim and Stühmer, 2000), we first deconvolved time-lapse images with a small number of iterations (100 for this study), and then high-pass filtered them at a spatial frequency of 1/mm. As depicted in Fig. 1 C, deconvolution with 100 iterations followed by high-pass filtering at a spatial frequency of 1/mm gave similar estimate of the width of the shell as that of simple deconvolution with 1000 iterations. Thus, for the rest of the image processing, we routinely employed the 100- iteration deconvolution together with high-pass filtering at a spatial frequency of 1/mm. The recorded and processed images from an AO-loaded PC12 cell are displayed in Fig. 1 D for comparison. Accuracy of SPT depends on the SNR and threshold level Both the SNR of the object and the threshold level applied to the image are critical for the performance of centroiddependent SPT. In this study, simulated granule trajectories and movement reconstructions, based on measured parameters, were used to evaluate the performance of SPT. We have generated sequential stacks of 3-D images containing single simulated granules at different SNR. Then, we used different SNRs as a parameter and identified the optimized threshold for each granule tracking according to its SNR. For a particular SNR, different threshold selection exerts significant impact on the SPT-estimated MSD as shown in Fig. 2 A. The reason is that as lower threshold is applied, too many background fluorescence remains around tracked granules, making the centroid-based SPT insensitive to granule movement and underestimate MSD throughout; whereas high threshold excludes too many pixels critical for the calculation of centroid and induces unexpected variability in the tracking as well as overestimation of MSD. Hence, we have examined the influence of different thresholds on the significance of difference (assayed by the P-value in KS test) and the bias between the real MSD and the one estimated from SPT (Fig. 2 B). Thus, appropriate threshold should be chosen to maximize the P-value and minimize the bias at given SNR. The relationship of optimized thresholds, corresponding P-values, and biases versus SNRs is revealed in Fig. 2 C. We notice that higher SNRs give better accuracy in SPT. This emphasizes the necessity to choose fluorescence spots with SNR equal to or above 6 for mobility tracking. Based on the SNR of each granule, we selected corresponding thresholds according to Fig. 2 C. For intermediate SNRs not simulated, thresholds were chosen by interpolation between the neighboring integer SNRs. Identification of fluorescence spots Most near-membrane fluorescence spots in PC12 cells labeled with AO have been identified as single large dense-core granules (Avery et al., 2000). However, as AO tends to accumulate in acidic compartments, it is important to verify that the cytosolic fluorescence spots that we selected for analysis are actually granules. After deconvolving and highpass filtering the images, we identified fluorescence spots with an SNR equal to or.6 (marked with arrows in Fig. 3 A) for further analysis. The lateral and axial fluorescence profiles of each selected AO spot were fitted with Gaussian functions, and the full width at half-maximum (FWHM) of the fitted Gaussian function was taken as a measure of the size of particles (one example is shown in Fig. 3 A, right). Fig. 3 B displays the size distributions of 188 fluorescent spots with the peak lateral and axial size at mm and mm (mean 6 STD), respectively. The larger axial size is caused by the relatively lower axial resolution inherited in WFFM. We then compared the averaged size of these fluorescence spots with that of NPY-DsRed labeled large dense-core secretory granules (Lang et al., 2000; Holroyd et al., 2002) with similar SNR and found their sizes were nearly equivalent (Fig. 3 C). In fact, the lateral size of our selected fluorescent spots is close to the TIRFM observation of dense-core granules in PC12 cells (Lang et al., 2000). Thus, we verified that the selection criteria seem to restrict most of the remaining fluorescence spots to large dense-core granules. We further selected those fluorescence spots with both lateral and axial size within the mean 6 2s of the Gaussian size distribution for further mobility analysis. Compared with NPY-DsRed labeled granules, larger SNR could be obtained for AO-loaded granules during much shorter exposure time (Fig. 3 C), so the employment of AO would provide us the ease of labeling, rapid sampling, and less photodamage for imaging live cells. 3-D tracking of single granules in resting cells Previous studies have suggested that granules diminished in their mobility as they approached the plasma membrane (Steyer et al., 1997; Johns et al., 2001). Also, the dense cortical actin network underneath the plasma membrane might influence the mobility of granules (Oheim and Stühmer, 2000; Lang et al., 2000). To confirm the usefulness Biophysical Journal 87(3)

221 1996 Li et al. of our 3-D SPT in studying the mobility of granules at different locations inside the cell, we have separated total granules into two groups, GPs and GCs, according to their relative distances from the detected contour of plasma membrane. The membrane contour was detected using the edge-detecting algorithm of Canny (1986). Briefly, it calculates the intensity gradient of images with the derivative of a Gaussian filter and locates the edge by looking for local maximum of the gradient. This method is robust to noise and likely to detect even weak edges. We compared the contour detected from deconvolved AO-stained images with that obtained from FM1-43 staining in the same cell. FM1-43 is not fluorescent in solution, and becomes fluorescent when incorporated into the membrane lipids (Leung et al., 2002; Angleson et al., 1999). As shown in Fig. 4 A, the membrane contours recognized from AO- and FM1-43-stained images were quantitatively comparable. The mean absolute difference between the two contours was mm (estimated from four cells) with a maximum and minimum value of 0.42 mm and 0.1 mm, respectively. The same section as in Fig. 4 A was high-pass filtered and displayed in high magnification in Fig. 4 B. Granules that meet our selection criteria are marked with circles. One granule identified as GP is indicated with an arrow in Fig. 4 B. The time-lapse images of this GP in the x, y and x, z plane are displayed in Fig. 4 C.As shown in Fig. 4 C, the granule seems quite consistent in its size and shape. However, sometimes the rotation of the nonspherical granule may change its projection in the 2-D section. Thus, as further evidence that the deconvolution did not introduce severe artifacts in our study, we have calculated the averaged 3-D volume of granules and found little variation between successive images, as demonstrated in Fig. 4 D. Different motion types of granules Next we studied the mobility of single granules by 3-D SPT. Only AO-loaded granules present for the entire imaging time were chosen for tracking analysis. Occasionally, we observed disappearance of granules in the middle of the image sampling, which might be due to either the explosions of AO-loaded granules or exocytosis when the granule is close to the plasma membrane. These granules were not included in subsequent analysis. We found that fluorescence decayed with an averaged time constant of s due to photobleaching. Thus the images were corrected for photobleaching before analysis. By fitting the MSD plot with appropriate equations, we have identified three types of motion among 246 total granules from 17 cells. A straight line fit to the MSD is indicative of random diffusion, which was generally observed for both GPs and GCs at similar frequency (Fig. 5 A). The major portion of the granules seems to move in a confined compartment and exhibit a negative curvature in the MSD plots (Fig. 5 B), which manifests a caged diffusion as if the granule is confined in a cage constructed by some neighboring obstruction. The averaged MSD of GC was much larger than that of GP, reflecting that cytosolic and nearmembrane granules are probably restricted by different structures. Functional studies have proposed that granules might be transported along some cytoskeletal tracks powered by motor proteins (Kamal and Goldstein, 2000; Rogers and FIGURE 4 3-D tracking of nearmembrane granules. (A) Membrane position was defined by the edgedetecting algorithm explained in the Materials and Methods section. Membrane contours, detected under AO staining and FM1-43 staining from the same cell, are displayed for comparison. The mean absolute difference between them is mm (data analyzed from four cells). (B) The same section as in panel A was high-pass filtered and displayed in high magnification. Granules meeting our selection criteria were marked with circles, which were subsequently included in the SPT analysis. (C) Sequential images of a granule marked by an arrow in B. The radial distance between the centroid of the granule and the membrane is 0.53 mm. Thus, the granule should be considered as a GP according to our definition. Scale bars for panels A, B, and C, 1 mm. (D) The averaged volumes from 25 granules calculated from the pixels that exceed the halfmaximum fluorescence are displayed as a function of time. Biophysical Journal 87(3)

222 Three-Dimensional Mobility of Granules 1997 FIGURE 5 Three types of motion exist for both GPs and GCs. Example trajectories (left) and averaged MSDs (right) of granules that travel in random diffusion (A), caged diffusion (B), or directed diffusion (C) are depicted, respectively. The numbers of granules analyzed are indicated in the graph. For comparison, averaged MSD of six immobilized 175-nm diameter beads imaged under SNR ¼ 6 condition is also displayed. Open circle, solid circle, and open triangle symbolize the GC, GP, and immobilized beads, respectively. * and ** mark the points from which on significant difference exists between the two MSD plots. Gelfand, 2000). In our experiments, we have also observed a small portion of both GPs and GCs (;10%) moving along a fixed direction as if transported by some already constructed tracks (Fig. 5 C). Their MSD plots resulted in a positive curvature because the V 2 term of Eq. 5 would dominate the MSD for longer time (Fig. 5 C). Interestingly, the directed velocity of granule transportation for GC ( ÿ3 mm/s, median 6 MSE, n ¼ 15) is also larger than that of GP ( ÿ3 mm/s, median 6 MSE, n ¼ 13) (P, 0.05, KS test). For comparison, we recorded sequential 3-D images of immobilized mm diameter beads under various imaging conditions. As an example, the averaged MSD plot from six beads under SNR ¼ 6 is displayed in Fig. 5 A, wherein a D (3) of ÿ4 mm 2 /s is calculated. This probably reflects our lower limit of detection under this imaging condition. Comparing the mobility between GP and GC In the preceding analysis, we have identified various motion types for granules according to MSD analysis, whereas detailed comparison of the 3-D mobility between GP and GC has not been made. In Fig. 6 A, D (3) distributions of 127 GPs and 119 GCs from 17 cells were compared. Because of the markedly skewed distribution, we have plotted the cumulative histogram of D (3) in the inset and checked the significance FIGURE 6 Comparison of the mobility between GP and GC. (A) D (3) histograms of GP (n ¼ 119) and GC (n ¼ 127) display a markedly skewed distribution. Inset, cumulative histogram reveals significant difference (P, 0.01, KS test). (B) D (3) distribution in semilogarithmic presentation. (C) Left, distribution of the radius of cage for GPs (n ¼ 69) and GCs (n ¼ 61) that move in caged diffusion. Right, cumulative histograms reveal significant difference (P, 0.05, KS test). (D) Scope of movement also displays significant difference (P, 0.05, KS test) between GP and GC. For this figure, wave symbols are defined in A. of difference with KS test. It is noticed that GC moves with a significant higher D (3) than that of GP (P, 0.01). Under resting conditions, 71% of GPs moved with a D (3), ÿ4 mm 2 /s, whereas only 23.7% of GCs traveled within that D (3) scope. The median of D (3) for GPs was ÿ4 mm 2 /s (n ¼ 127), which approximates the value measured for membrane-proximal granules by TIRFM ( ÿ4 mm 2 /s, Steyer and Almers, 1999), but much lower than that of GCs ( ÿ4 mm 2 /s, n ¼ 119). The D (3) histograms of GP and GC differ significantly from a simple broad Gaussian distribution, instead featuring a markedly skewed distribution and large variation among granules. To check whether there exist distinct groups for GP and GC, we calculated the logarithmic D (3) distributions and displayed them in semilogarithmic presentation (Fig. 6 B). Two separate peaks are obvious for GP and GC, suggesting they are indeed segregated into two distinct pools based on their mobility, whereas no distinct subgroups are observed within either GP or GC (Fig. 6 B). Next, we compared the histograms of the radius of cage for caged diffusion between GP and GC (Fig. 6 C). The histogram of the radius of cage for GC extends broader to the right, and the difference in cumulative histogram comparison is significant by KS test (P, 0.05). GCs moved in significant larger cages with a median of mm(n ¼ 61) in comparison with GP with a median of mm(n ¼ 69). Finally, we found that GCs traveled a larger mean distance (scope of movement) ( mm, median 6 MSE, n ¼ 127) in three dimensions than GPs over the same observation period ( mm, median 6 MSE, n ¼ 119) (P, 0.05), as depicted in Fig. 6 D. Biophysical Journal 87(3)

223 1998 Li et al. Stimulation increases the percentage of granules traveling in directed fashion Most granules wandered around their residing positions in random or caged diffusion under resting condition, and only a small fraction traveled in directed fashion. We found that high K 1 (HK) stimulation significantly increased the number of granules that traveled in directed fashion for both GP and GC (Fig. 7 A), suggesting an up-regulated active transportation of granules. HK stimulation leads to the influx of Ca 21. Ca 21 has been suggested to play an important role in accelerating the recruitment of secretory granules (von Rüden and Neher, 1993; Heinemann et al., 1993; Smith et al., 1998). Presumably, active transportation is involved in the recruitment of granules. Under stimulation, the median of D (3) was ÿ4 mm 2 /s (n ¼ 113) and ÿ4 mm 2 /s (n ¼ 97) for GP and GC, respectively, which were slightly higher than those of resting cells, but the differences were not statistically significant (P. 0.35, KS test). In contrast, the scopes of movement of GP and GC were significantly increased by HK stimulation (P, 0.05, Fig. 7 C). The influence of actin cytoskeleton on the mobility of granules Cortical actin cytoskeleton is localized subjacent to the plasma membrane with a thickness of ;0.4 mm (Lang et al., 2000). The mobility of granules going through the cortical FIGURE 7 Effects of high K 1 stimulation and latrunculin-b on granule mobility. (A) Percentages of granules that fall into different types of motion for GP and GC. C is abbreviation for control condition, HK for high K 1 stimulation, LB for latrunculin B treatment, and LB 1 HK for pretreatment with LB followed by high K 1 stimulation (n ¼ 8 11 cells for each condition mentioned above). (B) LB decreases the D (3) of GP significantly (P, 0.05, KS test) but does not change the D (3) of GC. (C) Comparison of the scope of movement for GP and GC under different conditions. For B and C, data are presented as median 6 MSE, and the numbers of granules selected for analysis under various conditions are from 77 to 127 granules. actin cytoskeleton has been explored with TIRFM, whereas whether granules located inside the cell are affected by actin cytoskeleton remains less studied. Here we have employed latrunculin B (LB) to disrupt F-actin formation in PC12 cells, and performed 3-D SPT to evaluate the mobility of granules. As shown in Fig. 7 B, LB significantly decreased the D (3) of GP (P, 0.05, KS test) but left that of GC unaltered, suggesting actin cortex facilitates the movement of nearmembrane granules rather than blocks it, which is consistent with the previous observation made by TIRFM (Lang et al., 2000). Although LB itself did not change the distribution of GPs and GCs in different motion types, the stimulatory effect of HK on the number of directed diffusion for GP was blocked by LB, an effect not seen for GC (Fig. 7 A). Similarly, LB blocked HK-induced increase in the scope of movement for GP but not for GC (Fig. 7 C). Our results suggest that although actin cytoskeleton participates actively in the movement of near-membrane granules, the trafficking of granules deep inside the cytosol is less affected by actin cytoskeleton. DISCUSSION Three-dimensional granule tracking To characterize the 3-D mobility of secretory granules throughout a whole cell, we have combined deconvolution WFFM and centroid-based 3-D SPT method to study granule trafficking in live PC12 cells. Although confocal microscopy has proved effective in removing the out-of-focus fluorescence, which improves the resolution for SPT, the inherited low time resolution limits its application in time-resolved 3-D particle tracking. Besides, confocal microscopy suffers from higher photobleaching and photodamage. In contrast, WFFM has the merit of less photobleaching and photodamage due to relatively higher light collection efficiency, especially with the highest NA (1.65) objective used in this study. The out-of-focus fluorescence in 3-D images recorded by WFFM can be computationally removed using deconvolution. In fact, WFFM combined with deconvolution has demonstrated the advantage over confocal microscopy for quantitatively studies of weakly fluorescent organelles in living cells (Swedlow and Platani, 2002). In classical 2-D time-resolved particle tracking, the contribution from the missing axial direction is usually ignored. TIRFM-based tracking attempts to resolve the z-mobility from the fluctuation in granule s fluorescence that is exponentially related to the change in axial position. However, small changes in the axial position will result in large changes in fluorescence measured with TIRFM. Also, fluctuation in fluorescence unrelated to axial position will result in misinterpretation of the z-position. Moreover, TIRFM can only be employed to study vesicles right underneath the plasma membrane, whereas vesicles residing deeper inside the cytosol are invisible under TIRFM. Despite the recent Biophysical Journal 87(3)

224 Three-Dimensional Mobility of Granules 1999 development in 3-D tracking of single particles (Kao and Verkman, 1994; Speidel et al., 2003; Levi et al., 2003), the 3-D mobility of granules within living cells has not been demonstrated. In this study, we have employed a centroidbased 3-D SPT from time-resolved stacks of 3-D images, and the displacements of single granules in three dimensions could be directly obtained. We found that granules can be assumed to move the same way in z direction as they behave in lateral direction, suggesting the missing z direction information in 2-D tracking probably will not impose significant distortion in assessing the lateral mobility of granules. However, 3-D tracking does offer the advantage of following granules in a larger 3-D space, not restricting to a single optical plane. We have demonstrated that appropriate threshold selection is very critical for the precision of granule tracking. Specific threshold should be selected for each given SNR to assure the best result in 3-D tracking. We have found that higher SNRs give better accuracy in SPT, thus improving SNR will help to get better precision in granule tracking. With our NA 1.65 objective, a theoretical resolution of 0.18 mm in x, y plane and 0.45 mm in z direction is expected. The evident discrepancy between lateral and axial resolution caused by diffraction limit leads to nonisotropic resolution and precision in 3-D particle tracking. In fact, we found that the threshold optimized (optimizing procedure, see Fig. 2) for lateral tracking performance was different from that optimized for axial tracking (data not shown) at a given SNR. Increases in resolutions along x, y directions and z direction, especially breaking the distortion occurred in z direction, will balance the tracking performance in different directions. The recently developed stimulated emission depletion technique (Klar et al., 2000) and beam-scanning multifocal multiphoton 4Pi-confocal microscopy (Egner et al., 2002) have finally broken the diffraction barrier and achieved a nearly spherical resolution of ;100 nm. These techniques will find more application in the demand for more precise and isotropic 3-D particle tracking. However, the drawbacks of these techniques inherited with confocal microscopy, i.e., low time resolution, high photobleaching, and photodamage, will limit their use when long-term 3-D tracking in living cell is demanded. 3-D mobility of secretory granules in PC12 cells AO accumulates within granules as well as nonvesicular compartments like lyso- or endosome. By choosing fluorescence spots with SNR $ 6 and by applying high-pass filtering, the selected fluorescence spots has single Gaussian profiles in fluorescence with lateral and axial FWHMs indistinguishable from those of NPY-DsRed labeled densecore granules, suggesting we were mainly analyzing AOlabeled large dense-core granules in this study. AO-loaded granules are much brighter than NPY-DsRed labeled ones, as 5 ms exposure of AO-loaded granules gave similar SNR to that of NPY-DsRed labeled granules with 100 ms exposure time (Fig. 3 C). Thus, the advantage of AO is not only its ease of use, but also the less exposure time and hence less photodamage. PC12 cells are densely packed with secretory granules. Our selection criteria might exclude most of the weakly stained spots (i.e., small synaptic like vesicles) and large spots for analysis; however, ;65% of the observed granules remained for analysis after selection. Moreover, although the density of our selected AO-labeled granules ( /mm 2 ) is lower than that of NPY-labeled granules observed under TIRFM (0.358/mm 2, Taraska et al., 2003) in PC12 cells, it is comparable with the density of cytosolic granules in chromaffin cells observed using electron microscopy (0.11/mm 2, Steyer et al., 1997). The characterization of particle motion imposes demands on the temporal and spatial resolution required for the measurement. When there is only simple free diffusion, as indicated by the linear dependence of MSD on time, the time resolution of the measurement does not influence the determination of the diffusion coefficient from the slop of the plot of MSD(nDt). In contrast, when diffusion is constrained by a cage or tether, a characteristic time of observation is required together with a corresponding requirement for sufficient temporal resolution to resolve the dynamic motion (Qian et al., 1991). Previous study has suggested that granules wander rapidly with a diffusion coefficient of ÿ4 mm 2 /s within a cage that leaves ;70 nm space around granules, whereas the cage itself diffuses in random 10-fold more slowly over longer distances (Steyer and Almers, 1999). The slow effective sampling rate (0.2 Hz) used in this study due to 3-D stack generation will miss the fine movement of granules within the cage. However, it will accurately track the long range diffusion of granules, which presumably reflects the movement of the cage. Imaging under TIRFM is restricted to a thin layer underneath the plasma membrane, which adheres unphysiologically to the coverglass. Whether the adhesion of plasma membrane to coverglass will affect the mobility of neighboring granules and their fusion remains to be examined. Interestingly, we found that GPs residing far away from the basal plasma membrane exhibit similar diffusion coefficient as that measured by TIRFM (Steyer and Almers, 1999, Becherer et al., 2003). GP in this study also traveled in a similar cage as that observed under TIRFM. Along with the fact that diffusion of fluorescent dyes could be observed as a cloud after fusion, it is likely the adhesion of basal membrane to coverglass does not exert significant effect on the surrounding granules. Secretory granules have long been assumed to belong to distinct functional pools (Bratanova-Tochkova et al., 2002; Duncan et al., 2003). By simultaneously tracking nearmembrane and cytosolic granules, we have found significant differences between these two groups of granules in their D (3), radius of cage, and scope of movement. The mobility of these two groups of granules is also differentially affected by Biophysical Journal 87(3)

225 2000 Li et al. disruption of F-actin. These results suggest that the motion of the two groups of granules is mediated by different mechanisms. However, we failed to distinguish subgroups within either GPs or GCs. The distributions of diffusion coefficient for either GP or GC did not exhibit multiple distinct peaks, suggesting granules are smoothly organized into one population, rather than into distinct pools, and most of them travel with lower mobility. This result is consistent with a recent study in the neurites of differentiated PC12 cells under TIRFM by Ng et al. (2003), who reported a broad and asymmetric distribution in diffusion coefficient without a separation of a distinct pool of granules. With the advantage of our method to study the mobility of granules deep inside the cytosol, we now extend this feature to cytosolic granules. Correlation of granules with cortical actin cytoskeleton Although some granules wandered with random diffusion, granule trafficking is likely to be mediated by different cytoskeleton systems along with a number of motor proteins (Hirokawa, 1998; Kamal and Goldstein, 2002). Thus, the motion of single granules might be influenced by expected or unexpected organelles on their way. The caged diffusion of granules much adjacent to the plasma membrane has been previously elucidated by TIRFM (Steyer and Almers, 1999; Oheim and Stühmer, 2000; Johns et al., 2001), which is seen as the entrapment of granules inside the cortical actin cytoskeleton meshwork. In this study, we have found that besides near-membrane granules, a majority of cytosolic granules traveled in a caged fashion as well. Interestingly, lots of GCs are likely trapped in a larger cage than GPs, suggesting a relatively looser meshwork inside cytosol. In addition to its role in restricting the movement of granules, cytoskeleton meshwork has also been thought to actively participate in granule motion, as well as to supply directional tracks for granule transportation (Rogers and Gelfand, 2000; Lang et al., 2000). In this study, we did find a small fraction of granules traveling in a directed fashion. Stimulation with HK solution increased the number of granules in directed traveling as well as the velocity of directed traveling (data not shown), whereas the diffusion coefficients remained unchanged, indicating an increased impelling of granules in an active manner. The augmentation in directed traveling was significantly inhibited by disruption of actin cytoskeleton for GP, but not for GC (Fig. 7 A). Moreover, the D (3) and the scope of movement of GP were also more prone to the treatment of LB than those of GC (Fig. 7, B and C). Taking together, we propose that the movement of nearmembrane granules are likely mediated as well as constrained by cortical actin network, whereas cytoskeleton other than cortical actin participates in the movement of cytosolic granules. Further experiments are demanded to identify the correlations between granule mobility and the underlying molecular mechanisms employing similar techniques used in this study. We thank J. Yao and T. Liang for assistance in the cell preparation, L. Xu and L. Bai for help in the adjustment of the imaging system, and J. Ding for helpful comments on the manuscript. This work was supported by National Science Foundation of China Grants No to A. Qu, Nos , , and to T. Xu, and National Basic Research Program of China (973) Grant G and 2001CCA04100 to T. Xu. We are grateful for the support from the Li Foundation and the Sinogerman Scientific Center. The laboratory of T. Xu belongs to a Partner Group Scheme of the Max Planck Institute for Biophysical Chemistry, Göttingen, Germany. REFERENCES Angleson, J. K., A. J. Cochilla, G. Kilic, I. Nussinovitch, and W. J. Betz Regulation of dense core release from neuroendocrine cells revealed by imaging single exocytic events. Nat. Neurosci. 2: Avery, J., D. J. Ellis, T. Lang, P. Holroyd, D. Riedel, R. M. Henderson, J. M. Edwardson, and R. Jahn A cell-free system for regulated exocytosis in PC12 cells. J. Cell Biol. 148: Becherer, U., T. Moser, W. Stuhmer, and M. Oheim Calcium regulates exocytosis at the level of single vesicles. Nat. Neurosci. 6: Bratanova-Tochkova, T. K., H. Cheng, S. Daniel, S. Gunawardana, Y. Liu, J. Mulvaney-Musa, T. Schermerhorn, S. G. Straub, H. Yajima, and G. W. G. Sharp Triggering and augmentation mechanisms, granule pools, and biphasic insulin secretion. Diabetes. 51:s83 s90. Canny, J A computational approach to edge detection. IEEE Trans. Pattern Anal. Mach. Intell. 8: Cheezum, M. K., W. F. Walker, and W. H. Guilford Quantitative comparison of algorithms for tracking single fluorescent particles. Biophys. J. 81: Conchello, J. A., and J. G. McNally Fast regularization technique for expectation maximization algorithm for computational optical sectioning microscopy. In Three-Dimensional Microscopy: Image Acquisition and Processing. Proc IS & T/SPIE Symposium on Electronic Imaging: Science and Technology. C. J. Cogswel, G. Kino, and T. Wilson editors. SPIE The International Society for Optical Engineering, SPIE code No , Bellingham, WA. Duncan, R. R., J. Greaves, U. K. Wiegand, L. Matskevich, G. Bodammer, D. K. Apps, M. J. Shipston, and R. H. Chow Functional and spatial segregation of secretory vesicle pools according to vesicle age. Nature. 422: Egner, A., S. Jakobs, and S. W. Hell Fast 100-nm resolution threedimensional microscope reveals structural plasticity of mitochondria in live yeast. Proc. Natl. Acad. Sci. USA. 99: Gibson, S. F., and F. Lanni Experimental test of an analytical model of aberration in an oil-immersion objective lens used in threedimensional light microscopy. J. Opt. Soc. Am. A. Opt. Image Sci. Vis. 9: Gosh, R. N., and W. W. Webb Automated detection and tracking of individual and clustered cell surface low density lipoprotein receptor molecules. Biophys. J. 66: Heinemann, C., L. von Rüden, R. H. Chow, and E. Neher A twostep model of secretion control in neuroendocrine cells. Pflugers Arch. 424: Hirokawa, N Kinesin and dynein superfamily proteins and the mechanism of organelle transport. Science. 279: Holroyd, P., T. Lang, D. Wenzel, P. D. Camilli, and R. Jahn Imaging direct, dynamin-dependent recapture of fusing secretory granules on plasma membrane lawns from PC12 cells. Proc. Natl. Acad. Sci. USA. 99: Biophysical Journal 87(3)

226 Three-Dimensional Mobility of Granules 2001 Jahn, R., T. Lang, and T. C. Südhof Membrane Fusion. Cell. 112: Johns, L. M., E. S. Levitan, E. A. Shelden, R. W. Holz, and D. Axelrod Restriction of secretory granule motion near the plasma membrane of chromaffin cells. J. Cell Biol. 153: Kamal, A., and L. S. Goldstein Connecting vesicle transport to the cytoskeleton. Curr. Opin. Cell Biol. 12: Kamal, A., and L. S. Goldstein Principles of cargo attachment to cytoplasmic motor proteins. Curr. Opin. Cell Biol. 14: Kao, H. P., and A. S. Verkman Tracking of single fluorescent particles in three dimensions: use of cylindrical optics to encode particle position. Biophys. J. 67: Klar, T. A., S. Jakobs, M. Dyba, A. Egner, and S. W. Hell Fluorescence microscopy with diffraction resolution barrier broken by stimulated emission. Proc. Natl. Acad. Sci. USA. 97: Kues, T., A. Dickmanns, R. Luhrmann, R. Peters, and U. Kubitscheck High intranuclear mobility and dynamic clustering of the splicing factor U1 snrnp observed by single particle tracking. Proc. Natl. Acad. Sci. USA. 98: Lang, T., I. Wacker, J. Steyer, C. Kaether, I. Wunderlich, T. Soldati, H.-H. Gerdes, and W. Almers Ca21-triggered peptide secretion in single cells imaged with green fluorescent protein and evanescent-wave microscopy. Neuron. 18: Lang, T., I. Wacker, I. Wunderlich, A. Rohrbach, G. Giese, T. Soldati, and W. Almers Role of actin cortex in the subplasmalemmal transport of secretory granules in PC-12 cells. Biophys. J. 78: Leung, Y. M., L. Sheu, E. Kwan, G. Wang, R. Tsushima, and H. Gaisano Visualization of sequential exocytosis in rat pancreatic islet b cells. Biochem. Biophys. Res. Commun. 292: Levi, V., Q. Ruan, K. Kis-Petikova, and E. Gratton Scanning FCS, a novel method for three-dimensional particle tracking. Biochem. Soc. Trans. 31: Levitan, E. S Studying neuronal peptide release and secretory granule dynamics with green fluorescent protein. Methods. 16: Markham, J., and J. A. Conchello Tradeoffs in regulated maximumlikelihood image restoration. In 3D Microscopy: Image Acquisition and Processing IV. Proc. of the 1997 SPIE s Biomedical Optics Symposium, BiOS97. C. J Cogswel., J. A. Conchello, and T. Wilson, editors. SPIE The International Society for Optical Engineering, SPIE code No , Bellingham, WA. Ng, Y.-K., X. Lu, A. Gulacsi, W. Han, M. J. Saxton, and E. S. Levitan Unexpected mobility variation among individual secretory vesicles produces an apparent refractory neuropeptide pool. Biophys. J. 84: Ohara-Imaizumi, M., Y. Nakamichi, T. Tanaka, H. Ishida, and S. Nagamatsu Imaging exocytosis of single insulin secretory granules with evanescent wave microscopy. J. Biol. Chem. 277: Oheim, M., and W. Stühmer Tracking chromaffin granules on their way through the actin cortex. Eur. Biophys. J. 29: Ölveczky, B. P., N. Periasamy, and A. S. Verkman Mapping fluorophore distributions in three dimensions by quantitative multiple angle-total internal reflection fluorescence microscopy. Biophys. J. 73: Pfeffer, S. R Transport-vesicle targeting: tethers before SNAREs. Nat. Cell Biol. 1:E17 E22. Pouli, A. E., E. Emmanouilidou, C. Zhao, C. Wasmeier, J. C. Hutton, and G. A. Rutter Secretory-granule dynamics visualized in vivo with a phogrin-green fluorescent protein chimaera. Biochem. J. 333: Qian, H., M. P. Sheetz, and E. L. Elson Single particle tracking. Analysis of diffusion and flow in two-dimensional systems. Biophys. J. 60: Rogers, S. L., and V. I. Gelfand Membrane trafficking, organelle transport, and the cytoskeleton. Curr. Opin. Cell Biol. 12: Saxton, M. J., and K. Jacobson Single-particle tracking: applications to membrane dynamics. Annu. Rev. Biophys. Biomol. Struct. 26: Smith, C., T. Moser, T. Xu, and E. Neher Cytosolic Ca 21 acts by two separate pathways to modulate the supply of releasable-competent vesicles in chromaffin cells. Neuron. 20: Speidel, M., A. Jonas, and E. Florin Three-dimensional tracking of fluorescent nanoparticles with subnanometer precision by use of offfocus imaging. Opt. Lett. 28: Steyer, J. A., and W. Almers Tracking single secretory granules in live chromaffin cells by evanescent-field fluorescence microscopy. Biophys. J. 76: Steyer, J. A., H. Horstmann, and W. Almers Transport, docking and exocytosis of single secretory granules in live chromaffin cells. Nature. 388: Swedlow, J. R., and M. Platani Live cell imaging using wide-field microscopy and deconvolution. Cell Struct. Funct. 27: Taraska, J. W., D. Perrais, M. Ohara-Imaizumi, S. Nagamatsu, and W. Almers Secretory granules are recaptured largely intact after stimulated exocytosis in cultured endocrine cells. Proc. Natl. Acad. Sci. USA. 100: Thompson, R. E., D. R. Larson, and W. W. Webb Precise nanometer localization analysis for individual fluorescent probes. Biophys. J. 82: Toonen, R. F. G., and M. Verhage Vesicle trafficking: pleasure and pain from SM genes. Trends Cell Biol. 13: Tooze, S. A., T. Flatmark, J. Tooze, and W. B. Huttner Characterization of the mature secretory granule, an intermediate in granule biogenesis. J. Cell Biol. 115: von Rüden, L., and E. Neher A Ca-dependent early step in the release of catecholamines from adrenal chromaffin cells. Science. 262: Biophysical Journal 87(3)

227 Int. J. Cancer: 111, (2004) 2004 Wiley-Liss, Inc. Publication of the International Union Against Cancer OVEREXPRESSION OF CAVEOLIN-1 INDUCES ALTERATION OF MULTIDRUG RESISTANCE IN Hs578T BREAST ADENOCARCINOMA CELLS Chuanxi CAI and Jianwen CHEN* National Laboratory of Biomacromolecules, Institute of Biophysics, Chinese Academy of Sciences, Beijing, China Caveolin-1 is a major caveolae-coat protein involved in a variety of cell signaling processes. Some studies have suggested that the level of caveolin-1 expression positively correlates with multi-drug resistance in cancer cells. We demonstrated for the first time that Hs578T doxorubicin resistant cells (Hs578T/Doxo), which contain low levels of endogenous caveolin-1 and high levels of P-glycoprotein, are rendered drug-sensitive by overexpression of exogenous caveolin-1. MTT assays showed that after overexpressing caveolin-1, the drug resistance of Hs578T/Doxo cells to doxorubicin and cisplatin was reduced from and g/ml to and g/ml, respectively (i.e. reduced by 97% and 64%, respectively). Furthermore, using rhodamine-123 efflux assays, we observed a significant decrease in P-glycoprotein activity in caveolin-1 overexpressing cells, similar to that observed with 5 M cyclosporine A or 10 M verapamil, 2 inhibitors of P-glycoprotein activity. Using confocal microscopy, subcellular fractionation and co-immunoprecipitation assays, a possible physical interaction between caveolin-1 and P-glycoprotein in the caveolae membrane was observed in Hs578T/Doxo cells overexpressing caveolin-1. These results suggest that overexpression of caveolin-1 changes the state of the cells from drug-resistant to drug-sensitive by inhibiting P-glycoprotein transport activity Wiley-Liss, Inc. Key words: caveolin-1; Hs578T breast adenocarcinoma cells; MDR; P-glycoprotein A major pitfall with chemotherapy in cancer patients is the development of tumor cells that are resistant to a broad range of therapeutic drugs commonly used in clinical treatment. 1 Although the precise mechanism of multi-drug resistance (MDR) remains largely unclear, there is abundant evidence that many cases of MDR are due to overexpression of a plasma membrane ATPase called P-glycoprotein (P-gp). 2 As an energy-dependent drug efflux transporter, P-gp decreases the effective concentration of active drugs in the cytosol and thereby reduces their cytotoxic efficacy. This is a very important mechanism of MDR in cancer cells. Caveolae are invaginations of the plasma membrane that are enriched in cholesterol and sphingolipids, which impart unique physical properties to these membrane sub domains. Caveolae have been shown to play an important role in many signal transduction processes, by acting as a compartmentalization center for signaling molecules. 3 9 Caveolin-1 is a major caveolar coat protein 10 that has the ability to engage in complex interactions with other protein components in the caveolae and to affect their functions. 11 Some studies have shown that the expression level of caveolin-1 closely correlates with the development of MDR in cancer cells. High expression levels of caveolin-1 and high surface density of caveolae have been identified in a number of MDR cancer cell lines, including adriamycin-resistant MCF-7 breast adenocarcinoma cells, colchicine-resistant HT-29 colon carcinoma cells, 12 vinblastine-resistant SKVLB1 ovarian carcinoma cells and Taxol-resistant A549-T24 lung carcinoma cells. 13 Thus, it has been proposed that the acquisition of resistance to different drugs by various cell types might be associated with high expression of caveolin-1. A more recent clinical study, however, showed that caveolin-1 expression in ovarian carcinoma showed no correlation with MDR. 14 Although several studies have suggested colocalization of caveolin and P-gp in caveolae, 12,15 the functional effect of caveolin-1 interaction with P-gp and its overall impact on the development and progression of MDR in cancer cells is largely unknown. To identify the possible effects of caveolin-1 on multidrug resistance and P-gp transport activity, we overexpressed caveolin-1 in Hs578T breast adenocarcinoma cells, and examined the changes of drug resistance and P-gp transport activity in these cells. In addition, we examined colocalization, co-fractionation and co-immunoprecipitation of caveolin-1 and P-gp. Our results suggest that overexpression of caveolin-1 in Hs578T breast cancer cells renders the cells drug-sensitive by inhibiting P-gp drugtransport activity. MATERIAL AND METHODS Materials Monoclonal antibody (mab) and polyclonal antibody (pab) against caveolin-1 were purchased from Transduction Laboratories (Lexington, KY) and Santa Cruz Biotechnology (Santa Cruz, CA), respectively. Anti-human P-gp mab (C219) was obtained from Signet Laboratories (Dedham, MA). FITC-labeled goat anti-mouse IgG and Texas Red-labeled goat anti-rabbit IgG were purchased from Molecular Probe Inc. Cell culture supplies (RPMI 1640 medium, geneticin, FBS, L-glutamine, trypsin-edta, Lipofectamine 2000, penicillin/streptomycin and Opti-Prep) were all from Life Technologies, Inc. (Carlsbad, CA). Percoll was purchased from Amersham Pharmacia Biotech. BCIP/NBT reagents were purchased from Promega. The pci-neo and pci-neo-cav-1 plasmids were generous gifts from Dr. Eric J. Smart (Kentucky Medical School, KY). 16 Protein-G Sepharose, verapamil, cyclosporine A and Rhodamine-123 were purchased from Sigma (St. Louis, MO). All other reagents were obtained from regular commercial sources with highest purity. Cell lines, plasmids and transfections The parental human breast cancer cell line Hs578T/S and its doxorubicin-resistant subclone Hs578T/Doxo were the generous gifts of Dr. Kjell Grankvist (Umea University, Sweden). 17 The cells were all cultured at 37 C, 5% CO 2 in RPMI 1640 medium Abbreviations: Cav-1, caveolin-1; CsA, cyclosporine A; DDP or cis, cisplatin; DMSO, dimethyl sulphoxide Doxo, doxorubicin; mab, monoclonal antibody; MDR, multidrug resistance; pab, polyclonal antibody; P-gp, P-glycoprotein; Rh-123, rhodamine-123; S, sensitive; Vm, verapamil. Grant sponsor: National Laboratory of Biomacromolecules; Grant number: ; Grant sponsor: Natural Science Foundation of China; Grant number: , ). *Correspondence to: National Laboratory of Macromolecules, Institute of Biophysics, Chinese Academy of Sciences, Datun Road 15, Chaoyang District, Beijing , China. Fax: chenmaci@sun5.ibp.ac.cn Received 20 August 2003; Revised 17 December 2003, 12 February 2004; Accepted 26 February 2004 DOI /ijc Published online 4 May 2004 in Wiley InterScience ( wiley.com). 227

228 CAVEOLIN-1 ALTERATION OF MULTIDRUG RESISTANCE 523 supplemented with 10% FBS, 2 mm glutamine, 100 U/ml penicillin, 100 g/ml streptomycin. The entire cdna sequence of caveolin-1 was cloned into the EcoRI site of pci-neo, to generate pci-neo-cav-1. PCI-neo-cav-1 and its empty vector were transfected into Hs578T cell lines as described. 16 Briefly, 5 g of pci-neo-cav-1 plasmid was transfected into 10 6 cells in a 100-mm dish, using Lipofectamine reagent according to the manufacturer s instruction. Twenty-four hours after transfection, the culture medium was changed to a selection medium containing 1.5 mg/ml geneticin for two days. Stable clones of cells overexpressing caveolin-1 were subsequently maintained in a culture medium containing 500 g/ml geneticin. As controls, the cells were transfected with the mock pci-neo vector, and subjected with same selection treatment with geneticin. Isolation of caveolae-enriched membrane fraction Detergent-free caveolae fractions were prepared using the modified method of Smart et al. 18 Briefly, cells were homogenized with a tight Dounce homogenizer. The post-nuclear supernatant fraction, obtained after centrifugation at 1,000g for 10 min, was layered on top of a 30% Percoll solution and centrifuged at 84,000g for 30 min in a Beckman Ti60 rotor. The plasma membrane (PM) faction was collected from the top of the gradient, and then sonicated with a Vibra Cell sonicator (Sonics & Materials, Danbury, CT). The sonicate was mixed with 50% Opti Prep to make a 23% Opti Prep solution. This was placed on the bottom of a Beckman SW41 centrifuge tube, and a linear 20 10% Opti-Prep gradient was layered on the top. The sample was then centrifuged at 52,000g for 90 min in a SW41 Beckman swinging bucket rotor. A sample of the bottom fraction (fractions 12 and 13) was collected and designated non-caveolae membrane (NCM). The top 5 ml of the gradient (fractions 1 6) was collected and mixed with 50% Opti-Prep. This was overlaid with 2 ml of 5% Opti-Prep and centrifuged at 52,000g for 90 min. An opaque band located just above the 5% interface was collected and designated caveolae fraction (CM). Western blot and immunoprecipitation Proteins were separated by SDS-PAGE with a standard reducing condition protocol. Samples were resuspended in sample buffer and loaded on 7.5% acrylamide-bisacrylamide gels for P-gp or 12.5% acrylamide-bisacrylamide gels for caveolin-1. After electrophoresis, proteins were electroblotted on to a nitrocellulose membrane. Blots were blocked by 5% nonfat dry milk, 0.05% Tween 20 in Tris-buffered saline (10 mm Tris, ph 8.0, 135 mm NaCl). Immunoblotting was carried out with designated antibodies and detection was carried out with the BCIP/NBT reagents according to the manufacturer s instructions. A standard protocol was used for co-immunoprecipitation studies of P-gp and caveolin-1. In brief, confluent Hs578T/Doxo pci-cav-1 cells were harvested by trypsin/edta solution and washed two times with ice cold PBS containing 0.1 mm PMSF. The cells were pelleted and resuspended in 0.5 ml modified RIPA buffer (150 mm NaCl, 50 mm Tris-Cl, ph 8.0, 1 mm EGTA, 1% Triton X-100, 0.1% SDS, 1% sodium deoxy cholate) containing protease inhibitors (0.1 mm PMSF, 1 M pepstatin A, 1 mm Benzamidine, 10 M Leupeptin, 1 g/ml Aprotinin). The whole cell lysate (100 g) was pre-washed with protein G-Sepharose beads and then incubated overnight with 1 g of monoclonal anti-p-gp antibodies (C219) or 1 g of anti-caveolin-1 pab (N- 20). As a negative control, another 0.5 ml of RIPA buffer without cell lysate was processed in the same way as the whole cell lysate. The immune complexes were collected on protein G-Sepharose beads by incubating for 2 hr and then washed extensively (4 times) with RIPA buffer. The beads were re-suspended in 40 l of 2-fold concentrated Laemmli electrophoresis buffer, separated by SDS- PAGE and electroblotted onto PVDF membranes for the detection of caveolin-1 or P-gp. 15 Cell survival MTT assay The 3-(4,5-dimethylthiazol-2-yl)-2,5-di-phenyltetrazolium bromide (MTT) colorimetric assay was used to assess the sensitivity of the cells to agents in vitro as described. 19 Cells were trypsinized and harvested during exponential growth, and equal numbers of cells (5,000 for Hs578T/S, Hs578T/S pci-neo and Hs578T/S pci-cav-1 or 10,000 for Hs578T/Doxo, Hs578T/Doxo pci-neo and Hs578T/ Doxo pci-cav-1) were inoculated into each well with 150 l of culture medium. After overnight incubation, 100 l of culture medium containing various concentrations of drugs were added, and the cells were incubated for 48 hr. Thereafter, 20 l of MTT (2 mg/ml in PBS) were added to each well and incubated for a further 4 hr. The resulting formazan product was dissolved in 100 l of dimethyl sulfoxide after aspiration of the culture medium. Plates were placed on a plate shaker for 5 min and read immediately at 550 nm using a microplate reader. The survival percent was calculated by the following equation: cell survival percent (mean OD of one grade)/(mean OD of the blank control) 100%. The IC 50 for doxorubicin or cisplatin was defined as the doxorubicin or cisplatin concentration at which the survival rate was 50%. Efflux of Rhodamine-123 To examine the transport activity of P-glycoprotein, efflux of Rhodamine-123 (Rh-123) was measured as described. 20,21 Briefly, the cells were harvested and adjusted to 10 6 /ml cells, then incubated with 2 M of Rh-123 for 30 min at 37 C. After 3 times of washing to remove the extracellular free dye, the cells were incubated in dye-free media or media containing 10 M verapamil or 5 M cyclosporine A (dissolved in DMSO) at 37 C. The efflux of Rh-123 was analyzed at successive time points by flow cytometry using a Becton-Dickinson FACS-420 instrument. Transmission electron microscopy Samples were fixed with glutaraldehyde, postfixed with osmium tetroxide, and stained with uranyl acetate and lead citrate, as described. 12 Samples were examined under a Philips TECAI 20 TEM. Confocal microscopy imaging Cells were cultured in 35-mm glass-bottom dishes or on a cover glass and the following protocol was used: 4 fixation was in PBS containing 4% paraformaldehyde at room temperature for 20 min. Fixed cells were rinsed with PBS and treated with 25 mm NH 4 Cl in PBS for 10 min to quench free aldehyde groups. The cells were permeabilized by incubation for 15 min in freshly prepared 0.1% Triton X-100/ PBS. The cells were preincubated for 1 hr in PBS containing 5% normal goat serum, and then incubated for 2 hr in diluted P-gp (1:25) or caveolin-1 (1:100) antibody in PBS containing 2.5% goat serum. Immunostaining was carried out by incubation for 1 hr in 1:200 diluted FITC-labeled goat anti-mouse IgG or Taxes Red-labeled goat anti-rabbit IgG. After extensive PBS washes, immunostained cells were examined with a Bio-Rad Radiant-2100 confocal microscope. More than 100 cells were inspected per experiment, and the results displayed are typical. Data analysis All data points represented the mean value of at least 3 independent experiments, with 5 duplicates for each. Values are represented as the mean SE. Statistical significance was determined by Student s t-test with p RESULTS Resistance to cisplatin and doxorubicin in Hs578T cells Through subclonal selections, we have obtained the following stable clones of cancer cells: parental human breast cancer cell line Hs578T (Hs578T/S) and its doxorubicin-resistant derivative cell line Hs578T/Doxo. The drug sensitivity or resistance of these cells was assayed using the MTT method. As shown in Table I, the 228

229 524 CAI AND CHEN Hs578T/Doxo cells were around 2-fold and 24-fold more resistant to cisplatin and doxorubicin than the parental Hs578T/S cells, respectively, suggesting that doxorubicin is a more specific substrate of P-gp than cisplatin. Overexpression of caveolin-1 and increase of caveolae organelles in Hs578T/S and Hs578T/Doxo cells Hs578T cells were transfected with the pci-neo-cav-1 plasmid, or with the pci-neo vector as a control. After culture in geneticin selection medium for two weeks, the expression of caveolin-1 and P-gp were detected by Western-blot assay. As shown in Figure 1a (lane 1,3), the expression levels of endogenous caveolin-1 were similar in Hs578T/S and Hs578/Doxo cell lines transfected with pci-neo vectors, but the expression level of P-gp was significantly higher in the drug-resistant cells (Hs578/Doxo vector) than in the drug-sensitive cells (Hs578T/S vector). TABLE I RESISTANCE TO CISPLATIN AND DOXORUBICIN IN Hs578T CELLS Reagent Hs578T/S IC 50 ( g/ml) 1 Hs578T/Doxo Cisplatin (1.9 ) 3 Doxorubicin (24 ) 1 Cell survival was determined by MTT assay. 2 Data are displayed as the mean SD of 3 determinations, each obtained from 5 duplicate cultures. 3 The values in parentheses represent the relative resistance, determined by dividing the IC 50 value of the resistant cells by that of the sensitive cells. The different expression patterns of P-gp and caveolin-1 in these cells raise an intriguing possibility that the mutual interaction between caveolin-1 and P-gp may impact the development of MDR in these cells. To test this possibility, we introduced exogenous caveolin-1 into these cells and therefore altered the ratio of caveoline-1 to P-gp, The subsequent response to different drugs was then studied. The densitometry results confirmed that the expression level of caveolin-1 in Hs578T/S pci-cav-1 or Hs578T/Doxo pci-cav-1 cells was fold higher than in Hs578T/S pci-neo or Hs578T/ Doxo pci-neo cells. Under the same conditions, the expression level of P-gp in Hs578T/Doxo pci-neo and Hs578T/ Doxo pci-cav-1 cells was about 20-fold higher than in Hs578T/S pci-neo and Hs578T/S pci-cav-1 cells, as detected by densitometry, irrespective of whether these cells were transfected with pci-cav-1 or empty vector (Fig. 1a, lane 1 vs. lane 2, lane 3 vs. lane 4). To further test the impact of overexpressing caveolin-1 on the formation of caveolae in Hs578T cells, we studied the changes of caveolae organelles using electron microscopy. As shown in Figure 1b, the number of caveolae increased greatly in Hs578T cells overexpressing caveolin-1. These caveolae appeared as attached omega-shaped flasks and bunches like clusters of grapes. Interestingly, a significant amount of caveolae-like intracellular vesicles could be detected in the cells overexpressing caveolin-1. Studies from Lavie et al. have shown that the caveolae-like vesicular compartments contain both caveolin-1 and P-gp. 12,15 The changes in membrane ultrastructure could in principle alter the MDR property of the Hs578T cells. FIGURE 1 Stable overexpression of caveolin-1 in Hs578T cells and the changes in caveolae organelles. (a) Drug sensitive (Hs578T/S) and drug resistant (Hs578T/Doxo) cells were stably transfected with caveolin-1, orvector as control. The expression levels of caveolin-1 in Hs578T/S and Hs578T/Doxo were about 2.5- or 4-fold higher than their control cell lines, respectively. The expression level of P-gp in Hs578T/Doxo is about 20-fold higher than in Hs578T/S cells, regardless of the caveolin-1 expression level. Each lane contained 30 g of total protein from whole cell lysates. The results shown are representative of three separate experiments. (b) Cells were fixed with osmium tetroxide, and stained with uranyl acetate and lead citrate. Representative electron micrographs display views of the plasma membrane, showing a large increase in the number of caveolae organelles in cells overexpressing caveolin

230 CAVEOLIN-1 ALTERATION OF MULTIDRUG RESISTANCE 525 FIGURE 2 Reduced multi-drug resistance in Hs578T cells overexpressing caveolin-1. The cell survival percentages of (a) Hs578T/S and (b) Hs578T/Doxo cells stably transfected with caveolin-1 (solid circle) or vector only (open square) were analyzed by MTT uptake assay with various doses of either cisplatin (left panels) or doxorubicin (right panels). (c) The relative drug resistances (%) were determined by dividing the IC 50 value of the drug for each individual cell line by that of their control cell line. The decrease in IC 50 of both cisplatin and doxorubicin in caveolin-1-transfected Hs578T/Doxo cells was significant (*p 0.01). The data shown are the mean ( SE) of 3 independent experiments, each with 5 duplicates. Reduced multi-drug resistance in Hs578T cells overexpressing caveolin-1 We examined the effect of overexpressed caveolin-1 on the drug resistance in Hs578T cells. We selected a specific substrate of P-gp, doxorubicin, and a non-specific substrate, cisplatin, to study the changes of multidrug resistance in these cells. As shown in Figure 2a, the IC 50 values for cisplatin in Hs578T/S pci-neo and Hs578T/S pci-cav-1 cells were about and g/ml, respectively (p 0.05) (left panel), and the IC 50 values for doxorubicin in Hs578T/S pci-neo and Hs578T/S pci-cav-1 cells were about and g/ml, respectively (p 0.05) (right panel). This indicated that although the Hs578T/S cells overexpressing caveolin-1 seem to be more sensitive to these drugs than the Hs578T/S cells transfected with the pcl-neo vector, the difference is not very large. In contrast, the difference in drug sensitivity could be observed more clearly in Hs578T/Doxo cells. The IC 50 values for cisplatin and doxorubicin in Hs578T/Doxo pci-neo were g/ml and g/ml, respectively, whereas those in Hs578T/Doxo pci- Cav-1 cells were reduced to and g/ml, respectively (p 0.01) (Fig. 2b). This equates to a decrease in drug resistance in terms of the IC 50 values of 64% for cisplatin and 97% for doxorubicin (Fig. 2c). These data demonstrate that overexpression of caveolin-1 markedly reduces the drug resistance in Hs578T/Doxo cells, particularly for doxorubicin, a specific substrate of P-gp. This suggests that overexpressed caveolin-1 may inhibit P-gp transport activity in Hs578T/Doxo cells. This would also be consistent with the high expression of P-gp in these cells compared to the parent strain, after selection in the presence of doxorubicin. Reduced P-gp transport activity in Hs578T cells overexpressing caveolin-1 Because the overexpression of caveolin-1 did not change the protein levels of P-gp (Fig. 1a, lane 1 vs. lane 2, lane 3 vs. lane 4), the reduced drug resistance in cells overexpressing caveolin-1 is likely to be due to the inhibition of P-gp transport activity. To verify this hypothesis, we examined the activity of P-gp by mea- 230

231 526 CAI AND CHEN FIGURE

232 CAVEOLIN-1 ALTERATION OF MULTIDRUG RESISTANCE 527 suring the efflux of rhodamine-123, a P-gp specific substrate, in Hs578T cells. Cells were loaded with Rhodamine-123, and then washed with dye-free medium to remove the extracellular dye. The fluorescence of remaining Rh-123 in the cells was monitored by flow cytometry. The fluorescence intensity change of Rh-123 reflects the change of P-gp transport activity in the membrane. The higher the fluorescence intensity of remaining Rh-123 in the cells, the lower the P-gp transport activity. As shown in Figure 3a, the Rh-123 fluorescence intensities at time 0 min were different in the 4 cell lines: F Hs578T/S pci-cav-1 F Hs578T/S pci-neo F Hs578T/ Doxo pci-cav-1 F Hs578T/Doxo pci-neo. This indicates that P-gp transport activity is lower in Hs578T/S cells than in Hs578T/Doxo cells, consistent with the relative expression levels of P-gp. It also indicates, however, that P-gp transport activity is lower when the expression level of caveolin-1 is increased. The Rh-123 fluorescence intensity at time 0 is related to the equilibrium of dye uptake and efflux, and the passive dye uptake rate is expected to be identical in the 4 cell lines. Therefore, the reduced intensity of Rh-123 fluorescence in drug-resistant cells and in those overexpressing caveolin-1 implies a reduction in efflux rate, which may be mediated by the function of P-gp. This was investigated further by measuring the rate of decrease of Rh-123 fluorescence intensity in the 4 cell lines (Fig. 3b,c). As shown in Figure 3b (right vs. left), Hs578T/S pci-cav-1 cells and Hs578T/S pci-neo cells had similar rates of decrease of Rh-123 fluorescence. This suggests that overexpression of caveolin-1 has little effect on the efflux of Rhodamine-123 when P-gp expression is low. Compared to drug resistant cells transfected with only the pci-neo vector, however, Hs578T/Doxo pci-cav-1 cells had a much lower rate of decrease of Rh-123 fluorescence (Fig. 3c, right vs. left). The rate of fluorescence decrease in Hs578T/Doxo pci-neo cells was significantly faster than in Hs578T/S pci-neo cells, indicating that the rate of efflux correlates with the level of P-gp expression. To quantify these changes, we took the mean fluorescence intensity of each sample at time 0 as 100% and plotted the relative fluorescence intensity at various time points as a percentage of this (Fig. 3d). The data demonstrate that Hs578T/Doxo Cav-1 cells have a similar fluorescence decline rate to Hs578T/S cells. This suggests that over-expression of caveolin-1 reduces the efflux rate of Rhodamine-123 in Hs578T/Doxo cells to the level observed in drug sensitive cells. To compare the inhibitory effect of caveolin-1 on the Rh-123 efflux from Hs578T cells with that of P-gp inhibitors, we used the drugs, verapamil and cyclosporine A, 2 known inhibitors of P-gp transport activity. 21 The results are shown in Figure 3e. We found that 5 M cyclosporine A or 10 M verapamil were able to reduce the efflux rate of Rh-123 from Hs578T/Doxo cells to the level of FIGURE 3 Inhibition of P-gp transport activity in Hs578T cells overexpressing caveolin-1. P-gp transport activity was measured by the changes in Rhodamine-123 fluorescence intensity as described in Experimental Procedures. (a) The Rhodamine-123 fluorescence intensity distribution was different in the 4 cell lines after 30 min of dye loading (red, Hs578/Doxo pci-neo; blue, Hs578T/Doxo pci-cav-1; orange, Hs578T/S pci-neo; green, Hs578T/S pci-cav-1). The left shift of fluorescent intensity in (b) Hs578T/S or (c) Hs578T/Doxo cells at different time points (0, 5, 10, 20, 30, 50 min as indicated) was due to the efflux of dye by P-gp-mediated transport. The efflux rate in Hs578T/Doxo cells overexpressing caveolin-1 (c, right panel) has been markedly reduced compared to control cells containing only vector (left panel). (d) The dynamic changes in Rhodamin-123 fluorescence intensity were re-plotted as relative intensity compared to time 0 (mean SE, n 3). Compared to drug sensitive Hs578T cells (circles), Hs578T/Doxo pci-neo cells (open triangles) showed a sharper decline, which is consistent with higher P-gp transport activity in drug-resistant cells. The decreased decline rate in Hs578T/Doxo pcl-cav-1 cells (solid triangles) indicates that overexpression of caveolin-1 rendered the drug-resistant cells drug-sensitive. (e) P-gp transport activity in Hs578T/Doxo cells (open triangles) was inhibited by verapamil (solid circles) and cyclosporine A (solid triangles) to a level similar to Hs578/S cells (open circles). The traces represent the combine results of three independent experiments. Hs578T/S cells. This not only suggests that the drug resistance of Hs578T/Doxo cells can be attributed to augmented P-gp levels, but also confirms that overexpressing caveolin-1 in Hs578T/Doxo cells has a similar effect on the activity of P-gp as known inhibitors of P-gp transport activity (Fig. 3d,e). Localization of P-gp in caveolae membrane and co-immunoprecipitation of P-gp and caveolin-1 Our data show that overexpression of caveolin-1 can functionally inhibit the transport activity of P-gp in Hs578T cells. This inhibition effect could be due to direct interaction, which would require physical contact between caveolin-1 and P-gp, presumably in the caveolae membrane region. To test this possibility, we used true co-localization confocal microscopy experiments of intact cells to detect P-gp and caveolin-1 in the caveolae domain of the plasma membrane. As shown in Figure 4a, we found that P-gp colocalizes with caveolin-1. We also found that P-gp and caveolin-1 co-fractionate in the caveolae fraction of the plasma membrane. As shown in Figure 4b, we employed a detergent-free method to isolate the caveolae fraction from Hs578T/Doxo pci-cav-1 cells. The data show that caveolin-1 and P-gp could be detected in the whole plasma membrane fraction (PM). The strong bands of caveolin-1 and P-gp immunoreactive proteins in the caveolae membrane fraction (CM) indicates, however, that the majority of caveolin-1 and P-gp is localized in the caveolar membrane sub domain. Caveolin-1 and P-gp proteins were undetectable in the noncaveolae membrane fraction (NCM). We also carried out coimmunoprecipitation in Hs578T/Doxo pci-cav-1 whole cell lysates using protein G-Sepharose beads and antibodies against caveolin-1 and P-gp. As shown in Figure 4c, caveolin-1 protein could be pulled down by anti-p-gp antibody and P-gp could also be co-immunoprecipitated with caveolin-1 by anti-caveolin-1 antibody. Taken together, these results suggest that caveolin-1 and P-gp may be able to interact directly within the caveolae membrane compartment. DISCUSSION Controlling the overexpression of particular proteins in the cell is increasingly employed as a tool to investigate their function and physiological significance. 24,25 We used this method in Hs578T carcinoma cells and their MDR derivatives to investigate the role of caveolin-1 in the reduction of the transport activity of P-gp, and hence the ability of caveolin-1 overexpression to cause multidrug resistant Hs578T/Doxo cells to become drug sensitive. We also observed that P-gp and caveolin-1 co-localize and co-fractionate in the caveolar membrane fraction, meaning that a direct physical interaction between these proteins is possible. It is generally believed that development of MDR in cells is due to the increased expression of P-gp. Our data, however, suggest that changes in the expression of caveolin-1 also play an important role in regulating the resistant nature of cells in response to multi-drug treatment. Conflicting results have been reported in the literature on the role of caveolins in the development of MDR. On the one hand, the caveolin-1 expression level is upregulated in MDR phenotypes of a number of human cell lines, including colon, ovarian, breast and lung carcinoma On the other hand, expression of caveolin-1 and caveolin-2 was not detected in several MDR cell lines that express high levels of P-glycoprotein, such as J7.V1-1 and J7.T3-1.6 derived from murine macrophage, and Caco-V100 derived from human colon carcinoma cells. 13 Unlike epothilone B-resistant A549 cells, epothilone B-resistant MCF-7 cells do not express any caveolins. It has also been shown that caveolin-1 expression has no association with that of P-gp protein or MDR1 mrna in human ovarian carcinoma. 14 In our study, we observed comparable amounts of caveolin-1 in Hs578T cells, irrespective of whether they were drug-sensitive or drug-resistant (Fig. 1a, lane 1,3). Thus, it is possible that the expression level of caveolin-1 in MDR cells 232

233 528 CAI AND CHEN FIGURE 4 Co-localization and co-immunoprecipitation of caveolin-1 and P-gp. (a) Co-localization of P-gp with caveolin-1 in Hs578T/ Doxo cells. Cells were double stained and examined by confocal microscopy. For each experiment, at least 100 cells were examined, and the figures shown represent the typical staining pattern for the majority of cells. The yellow fluorescent points (right hand panel) indicate colocalization of P-gp and caveolin-1 in the same domain of the plasma membrane. (b) Hs578T/ Doxo cells stably transfected with pci-neo-cav-1 were processed to isolate plasma membrane (PM), non-caveolae membrane (NCM) and caveolae membrane (CM) using a detergent-free method. Most of caveolin-1 and P-gp were localized in the caveolae membrane fraction, and there was no detectable caveolin-1 or P-gp in the noncaveolae membrane fraction. (c) P-gp and caveolin-1 could be co-immunoprecipitated (co-ip) from cell lysates using either anti-p-gp or anticaveolin-1 antibody (lane 2). Lane 1 was the negative control lacking cell lysate. The experiments were carried out at least 3 times with the same results. varies in a tissue-dependent, cell-dependent or drug-dependent manner. Most importantly, our data show that overexpression of caveolin-1 reduces P-gp transport activity and renders multi-drug resistant cancer cells sensitive to anticancer drugs. The inhibitory effect of overexpressing caveolin-1 on P-gp transport activity was found to be similar to the effect of verapamil or cyclosporine A, which are known inhibitors of P-gp activity (Fig. 3e). These functional studies not only provide some clues as to the role of caveolin-1 in multi-drug resistance of cancer cells, but also indicate that caveolin-1 may function as an important regulator of P-gp transport activity. Co-fractionation and co-immunoprecipitation assays in Hs578T/Doxo-Cav-1 cells (Fig. 4) indicate that P-gp and caveolin-1 are predominantly found in the caveolae membrane fraction. Confocal microscopy confirms that caveolin-1 and P-gp colocalize in intact cells, meaning that direct association of these proteins in the caveolae membrane may occur. This is consistent with previous findings that caveolin-1 and P-gp colocalize in the low density detergent-insoluble membrane fraction and that there is a physical interaction between these proteins in resistant CH R C5 cells and brain capillaries. 15 This indicates that co-localization of caveolin-1 and P-gp as well as their physical association is not depend on the cells selected for their resistance. In our study, the functional inhibitory effect of caveolin-1 on P-gp transport activity was further investigated. Caveolin-1, as an important structural protein of caveolae, plays a key role in signal transduction and lipid homeostasis. There is a growing body of evidence that a number of proteins, such as Ras, Src, protein kinase C, enos, and the epidermal growth factor receptor, can directly interact with a specific scaffolding domain of caveolin-1, which consequently results in a strong inhibition of their biological activities. 11,26 The proteins interacting with caveolin-1 have a consensus caveolin-binding motif, termed scaffold site, which contains the sequences XXXX XX or XX XXXX, where is an aromatic residue, and X is any amino acid. 27 By BLAST searching, we found that this caveolin-binding motif is also present in human P-gp between amino acid residues 36 and 44 (FSMFRYSNW). This is similar to the equivalent residues in hamster MDR1 (FTMFRYAGW), which has already been identified as a putative caveolin-binding motif. 15 Taken together, this implies that the mechanism of this functional inhibition of the enzymatic activity may be caused by direct interaction between caveolin-1 and P-gp. Nevertheless, other possible mechanisms of the inhibitory effects of caveolin-1 on P-gp transport activity and drug-resistance in cancer cells must be considered. One possible mechanism is that caveolin-1 regulation of cellular cholesterol may be important in the regulation of P-gp drug transport activity. Caveolin-1 has been reported to functionally bind to cholesterol. 28 In addition, recombinant expression of caveolin-1 has been shown to facilitate the transport of newly synthesized cholesterol from the ER to the plasma membrane, where cholesterol rapidly diffuses out of caveolae. 29 Thus, it is possible that the overexpression of caveolin-1 might indirectly affect P-gp transport activity through modulation of the availability of cholesterol at the level of the plasma membrane. 233

234 CAVEOLIN-1 ALTERATION OF MULTIDRUG RESISTANCE 529 Although the precise mechanism by which overexpression of caveolin-1 alters multidrug resistance is worthy of further investigation, our present data suggest that enhancing the expression or function of caveolin-1 could downregulate the multidrug resistance of cancer cells and thus might serve as an alternative mechanism for improving the efficacy of chemotherapy in cancer patients. ACKNOWLEDGEMENTS We thank Dr. E.J. Smart (Kentucky Medical School, Kentucky) for providing the pci-neo and pci-neo-cav-1 expression vectors and Dr. K. Grankvist (Umea University, Sweden) for providing Hs578T cell line. We are grateful to Dr. Zui Pan (Piscataway, NJ) and Dr. Sarah Perrett (Institute of Biophysics, Beijing, P.R. China) for editing the manuscript. REFERENCES 1. Gottesman MM. How cancer cells evade chemotherapy: sixteenth Richard and Hinda Rosenthal Foundation Award Lecture. Cancer Res 1993;53: Bosch I, Croop J. P-glycoprotein multidrug resistance and cancer. Biochim. Biophys. Acta 1996;1288:F Lisanti MP, Tang Z, Scherer PE, Kubler E, Koleske AJ, Sargiacomo M. Caveolae, transmembrane signaling and cellular transformation. Mol Membr Biol 1995;12: Li S, Couet J, Lisanti MP. Src tyrosine kinases, G subunits, and H-Ras share a common membrane-anchored scaffolding protein, caveolin. J Biol Chem 1996;271: Li S, Okamoto T, Chun M, Sargiacomo M, Casanova JE, Hansen SH, Nishimoto I, Lisanti MP. Evidence for a regulated interaction between heterotrimeric G proteins and caveolin. J Biol Chem 1995;270: Scherer PE, Okamoto T, Chun M, Nisimoto I, Lodish HF, Lisanti MP. Identification, sequence, and expression of caveolin-2 defines a caveolin gene family. Proc Natl Acad Sci USA 1996;93: Song SK, Li S, Okamoto T, Quilliam LA, Sargiacomo M, Lisanti MP. Co-purification and direct interaction of Ras with caveolin, an integral membrane protein of caveolae microdomains. J Biol Chem 1996;271: Mineo C, James GL, Smart EJ, Anderson RG. Localization of epidermal growth factor-stimulated Ras/Raf-1 interaction to caveolae membrane. J Biol Chem 1996;271: Li S, Seitz R, Lisanti MP. Phosphorylation of caveolin by Src tyrosine kinases. J Biol Chem 1996;271: Rothberg KG, Heuser JE, Donzell WC, Ying YS, Glenney JR, Anderson RG. Caveolin, a protein component of caveolae membrane coats. Cell 1992;68: Okamoto T, Schlegel A, Scherer PE, Lisanti MP. Caveolins, a family of scaffolding proteins for organizing preassembled signaling complexes at the plasma membrane. J Biol Chem 1998;273: Lavie Y, Fiucci G, Liscovitch M. Up-regulation of caveolae and caveolar constituents in multidrug-resistant cancer cells. J Biol Chem 1998;273: Yang CPH, Galbiati AE, Volonte D, Horwitz SB, Lisanti MP. Upregulation of caveolin-1 and caveolae organelles in Taxol-resistant A549 cells. FEBS Lett 1998;439: Davidson B, Goldberg I, Givant-Horwitz V, Nesland JM, Berner A, Bryne M, Risberg B, Kopolovic J, Kristensen GB, Trope CG, Van D, Reich R. Caveolin-1 expression in ovarian carcinoma is MDR1 independent. Am J Clin Pathol 2002;117: Demeule M, Jodoin J, Gingras D, Beliveau R. P-glycoprotein is localized in caveolae in resistant cells and in brain capillaries. FEBS Lett 2000;466: Uittenbogaard A, Smart EJ. Palmitoylation of caveolin-1 is required for cholesterol binding, chaperone complex formation, and rapid transport of cholesterol to caveolae. J Biol Chem 2000;275: Jonsson O, Motlagh PB, Persson M, Henriksson R, Grankvist K. Increase in doxorubicin cytotoxicity by carvedilol inhibition of P- glycoprotein activity. Biochem Pharmacol 1999;58: Smart EJ, Ying YS, Mineo C, Anderson RGW. A detergent-free method for purifying caveolae membrane from tissue culture cells. Proc Natl Acad Sci USA 1995;92: Hyafil F, Vergely C, Vignaud PD, Grand-Perret T. In vitro and in vivo reversal of multidrug resistance by GF120918, an acridonecarboxamide derivative. Cancer Res 1993;53: Krishan A, Sauereig A, Gordon K, Swinkin C. Flow cytometric monitoring of cellular anthracycline accumulation in murine leukemic cells. Cancer Res 1986;46: Patel VA, Dunn MJ, Sorokin A. Regulation of MDR-1 (P-glycoprotein) by cyclooxygenase-2. J Biol Chem 2002;277: Jodoin J, Demeule M, Beliveau R. Inhibition of the multidrug resistance P-glycoprotein activity by green tea polyphenols. Biochim Biophys Acta 2002;1542: Stouch TR, Gudmundsson O. Progress in understanding the structureactivity relationships of P-glycoprotein. Adv Drug Deliv Rev 2002; 54: Raziani B, Rubin CS, Lisanti MP. Regulation of camp mediated signal transduction via interaction of caveolins with the catalytic subunit of protein kinase A. J Biol Chem 1999;274: Pidgeon GP, Tang KQ, Rice RL, Zacharek A, Li L, Taylor JD, Honn KV. Overexpression of leukocyte-type 12-lipoxygenase promotes w256 tumor cell survival by enhancing v 5 expression. Int J Cancer 2003;105: Anderson RG. The caveolae membrane system. Annu Rev Biochem 1998;67: Couet J, Li S, Okamoto T, Ikezu T, Lisanti MP. Identification of peptide and protein ligands for the caveolin-scaffolding domain. J Biol Chem 1997;272: Murata M, Peranen J, Schreiner R, Weiland F, Kurzchalia T, Simons K. VIP21/Caveolin is a cholesterol-binding protein. Proc Natl Acad Sci USA 1995;92: Smart EJ, Ying YS, Donzell WC, Anderson RGW. A Role for caveolin in transport of cholesterol from endoplasmic reticulum to plasma membrane. J Biol Chem 1996;271:

235 Biochem. J. (2004) 379, (Printed in Great Britain) 697 Kinetic analysis of ligand-induced autocatalytic reactions Jiang-Hong LIU and Zhi-Xin WANG 1 National Laboratory of Biomacromolecules, Center for Molecular Biology, Institute of Biophysics, Academia Sinica, Beijing , People s Republic of China Protein phosphorylation and limited proteolysis are two most common regulatory mechanisms involving the energy-dependent covalent modification of regulatory enzymes. In addition to modifying other proteins, many protein kinases and proteases catalyse automodification reactions (i.e. reactions in which the kinase or zymogen serves as its own substrate), and their activities are frequently regulated by other regulatory ligands. In the present study, a kinetic analysis of autocatalytic reaction modulated by regulatory ligands is presented. On the basis of the kinetic equation, a novel procedure is developed to evaluate the kinetic parameters of the reaction. As an example of an application of this method, the effects of calcium ions on the autoacatalytic activation of trypsinogen by trypsin is re-examined. The results indicate that the binding affinity for Ca 2+ -bound trypsinogen to trypsin is at least two orders of magnitude higher than that for Ca 2+ -free trypsinogen, and therefore that the effect of Ca 2+ ions on K m values for trypsinogen is very much greater than that for the model peptides. Based on the experimental results, one possible molecular mechanism has been proposed. Key words: autophosphorylation, limited proteolysis, protein kinase, trypsin, trypsinogen, zymogen activation. INTRODUCTION Enzyme-catalysed covalent modification is an important mechanism for the regulation of enzyme activity. In such a regulatory process, one enzyme acts to modify the activity of another by chemically modifying the target enzyme. Protein phosphorylation and limited proteolysis are two most common regulatory mechanisms involving the energy-dependent covalent modification of regulatory enzymes. The control of enzyme activity through a reversible covalent modification such as phosphorylation can have a variety of functions: (i) it can change the kinetic and allosteric properties of the enzyme; (ii) by necessity it introduces two interconverting enzymes into the system, which may themselves be subject to control mechanisms; and (iii) further increases in the regulatory potential of an enzyme can be obtained by introducing multiple phosphorylation or by linking two or more phosphorylation/dephosphorylation cycles in sequence [1]. Protein phosphorylation is the most prevalent form of posttranslational covalent modification mechanism that cells employ to regulate enzyme activity. It is a means of superimposing the effect of a stimulus from outside the cell on the prevailing metabolic state within that cell. The enzymes responsible for catalysing this reaction are protein kinases. In addition to phosphorylating other proteins, many protein kinases catalyse autophosphorylation reactions (i.e. reactions in which the kinase serves as its own substrate), and their activities are frequently regulated by phosphorylation and other regulatory ligands [2]. The autophosphorylation reactions can be intramolecular or intermolecular. The intermolecular autophosphorylation involves a bimolecular autocatalytic event. In addition to modulation by other regulatory ligands, the rate of intermolecular autophosphorylation is dependent on the concentration of the protein kinase, and thereby provides an alternative means for regulating particular biological processes. Proteolytic enzymes are normally synthesized and secreted as inactive precusors, which are activated at a physiologically appropriate time and place. These precusors are known as proenzymes, or zymogens. The zymogens must undergo an activ- ation process, usually a limited proteolysis, to attain their catalytic activity. The active forms of zymogens usually have powerful physiological effects, and their synthesis in inactive form permits them to be safely stored until they are required. Zymogen activation is a phenomenon of great importance to our understanding of fundamental biochemical and physiological processes. They are involved in many physiological processes, such as digestion, metabolism, differentiation, immunity, blood coagulation, fibrinolysis, apoptosis and respose to injury [1,3 10]. When the activating enzyme and the activated enzyme coincide, the process is an autocatalytic zymogen activation. Physiological examples of these processes are the activation of trypsinogen, prekallikrein, pepsinogen and human blood coagulation factor XII by trypsin, kallikrein, pepsin and factor XIIa (a form of activated XII), respectively [1,11 13]. Recently, kinetic studies of the autoactivation of protein kinase and zymogen have been reported [14,15]. In the present paper, kinetic analysis of an autocatalytic reaction modulated by a regulatory ligand is presented. As an example of an application of this method, the effect of calcium ions on the autoactivation of trypsinogen by trypsin was re-analysed. Some years ago, Abita et al. [16] examined in detail the effect of Ca 2+ in the trypsin-catalysed hydrolysis of the Lys Ile bond in trypsinogen and several model peptides with sequences related to the N- terminal sequence of bovine trypsinogen. With trypsinogen, when the concentration of Ca 2+ increases from 4 to 50 mm, K m decreased by a factor 3 and k cat was not changed. The trypsincatalysed hydrolysis of the nonapeptide Val-Asp 4 -Lys-Ile-Val- Gly was also Ca 2+ -dependent, and K m decreased by a factor of 4.3; the effect was very similar to that observed for trypsinogen. According to these observations, the authors suggested that the N-terminal hexapeptide is probably randomly arranged at the surface of trypsinogen and floats freely in the surrounding solvent. However, our results indicate that the K m value for the trypsinogen activation was greatly overestimated in the previous study, and that the binding affinity for Ca 2+ -bound trypsinogen to trypsin is at least two orders of magnitude higher than that for Ca 2+ -free trypsinogen. Therefore, the effect of Ca 2+ ions Abbreviation used: TAME, N-α-p-tosyl-L-arginine methyl ester. 1 To whom correspondence should be addressed ( zxwang@sun5.ibp.ac.cn). 235 c 2004 Biochemical Society

236 698 J.-H. Liu and Z.-X. Wang Scheme 1 Mechanism of the autocatalytic reaction L represents regulatory ligand, E and E* represent unmodified and modified enzyme, respectively, and P is the product. on K m values for trypsinogen is very much greater than that for the model peptide, Val-Asp 4 -Lys-Ile-Val-Gly. Based on the experimental results, one possible molecular mechanism has been proposed. THEORETICAL ANALYSIS In the cases of the autocatalytic reactions, the unmodified enzyme serves as a substrate in reactions. The general mechanism of the intermolecular autocatalytic reaction of an enzyme can be written as shown in Scheme 1. In Scheme 1, L represents regulatory ligand, and E and E represent unmodified and modified enzyme, respectively. It is assumed that the other substrates and products for modification, such as ATP, water, etc., are present at constant levels, and can therefore be included in the kinetic constants without loss of generality. Since the concentrations of protein substrate and enzyme are of the same order of magnitude, the steady-state assumption is not satisfactory in this case [17]. When there is equilibrium as far as EL, E EandE EL are concerned, i.e. when k 2 and k 2 are sufficiently small as not to disturb equilibrium, we then have K S = [E ][E] E E, K S = [E ][EL] [E EL] K d = [E][L] [EL], K d = [E E][L] [E EL] The total concentration of enzyme is [T] 0 = [E] 0 + [E ] 0 = [E] + [E ] + [EL] + 2[E E] + 2[E EL] (2) where [T] 0 is the total concentration of trypsin plus trypsinogen, and [E] 0 and [E ] 0 are the initial concentrations of the unmodified and modified enzyme respectively. Let [E T ] = [E ] + [E E] + [E EL] (3) From eqns (1) (3), we have ( ) Kd + [L] [E EL] 2 ([T] 0 + K m )[E EL] [L] + ([T] 0 E T )[E T ][L] = 0 (4) K d + [L] (1) where K m = (K d + [L])K S (5) K d + [L] is the apparent Michealis Menten constant for the intermolecular autocatalytic reaction. The solution of eqn (4) for [E EL] is given by the quadratic formula as [L] { [E EL] = [T]0 + K m 2(K d + [L]) ([T] 0 + K m ) 2 4([T] 0 [E T ])[E T ] } (6) The rate of the modified enzyme formation is given by ( ) d[e T ] = k 2 [E E] + k 2 [E k2 K d EL] = + k 2 [E EL] dt [L] = k cat { [T]0 + K m 2 ([T] 0 + K m ) 2 4([T] 0 [E T ])[E T ] } (7) where k cat = k 2K d + k 2 [L] (8) K d + [L] is the apparent turnover number (catalytic centre activity) for the intermolecular autocatalytic reaction. Eqn (7) can be rewritten as 2d[E T ] [T] 0 + K m ([T] 0 + K m ) 2 4([T] 0 [E T ])[E T ] = k cat dt To integrate this equation, put x = 2[E T ] [T] 0 + ([T] 0 + K m ) 2 4([T] 0 [E T ])[E T ], so that [E T ] = (x + [T] 0) 2 (K m + [T] 0 ) 2 4x and [T] 0 + K m ([T] 0 + K m ) 2 4([T] 0 [E T ])[E T ] = 2[E T ] + K m x = (x + [T] 0) 2 (K m + [T] 0 ) 2 + K m x 2x = (x K m )(x K m 2[T] 0 ) 2x Differentiation of eqn (10) with respect to x gives (9) (10) (11) d[e T ] = x2 + 2K m [T] 0 + K m 2 4x 2 dx (12) Substitution of eqns (11) and (12) into eqn (9) yields x 2 + 2K m [T] 0 + K 2 m x(x K m )(x K m 2[T] 0 ) dx = k cat dt (13) With the boundary condition t = 0, [E T ] = [E ] 0, this integrates to k cat t = ln x x 0 + K m + [T] 0 [T] 0 ln (x K m 2[T] 0 )(x 0 K m ) (x 0 K m 2[T] 0 )(x K m ) (14) c 2004 Biochemical Society 236

237 Ligand-induced autocatalytic kinetics 699 where x 0 = 2[E ] 0 [T] 0 + (K m + [T] 0 ) 2 4[E ] 0 [E] 0 (15) MATERIALS AND METHODS Bovine pancreatic trypsinogen, trypsin and TAME (N-α-p-tosyl- L-arginine methyl ester) were purchased from Sigma Chemical Co. The active-site normality of trypsin was 90%. Traces of chymotryptic activity would not be expected to interfere with the activation of trypsinogen, since the specificity does not fit the activation sites. All other chemicals were local products of analytical grade. The concentration of trypsinogen was determined by measuring the absorbance at 280 nm and using the absorption coefficient M 1 cm 1 [18]. Trypsin activity was routinely assayed by monitoring the increase in absorbance at 245 nm due to hydrolysis of TAME using a PerkinElmer spectrophotometer [19]. Initial rates of the reactions were determined from the linear slope of the progress curves obtained with a molar absorption coefficient ε 245 of 595 M 1 cm 1. All the assays were carried out at 30 C and ph 8.1 in 40 mm Tris/HCl buffer. In kinetic studies of the autocatalytic conversion of trypsinogen into trypsin, aliquots of the incubation mixture of trypsinogen and trypsin were periodically removed and the activity of trypsin was determined at 30 C. The assay system contained 40 mm Tris/HCl (ph 8.1), 1 mm TAME and 10 mm CaCl 2. Since there are many sets of constants that give essentially the same curve, the individual fit for each set of experimental data to eqn (14) cannot give a reliable estimate for the kinetic parameters of the autocatalytic reaction. In practice, commercial trypsinogen always contains a trace amount of contaminating active enzyme. Therefore, the initial concentrations of enzyme species can be written as [E ] 0 = α[t] 0 and [E] 0 = (1 α)[t] 0, where α is a constant. In this case, χ 0 in eqn (14) can be written as x 0 = (2α 1)[T] 0 + (K m + [T] 0 ) 2 4α(1 α)[t] 0 2 (16) and eqn (14) can then be treated as a function with two independent variables, [E T ]and[t] 0, and three parameters, α, K m and k cat. The problem of non-uniqueness of the estimated parameters can be solved by a global-analysis approach when a series of experiments is carried out at different initial concentrations of enzyme [15]. Unlike conventional fitting methods in which data corresponding to each fixed [T] 0 are fit individually, the kinetic parameters obtained by global analysis are constrained by the entire matrix of [E T ]and[t] 0 simultaneously. By combining multiple experiments together in a single analysis, much more dramatic improvement in the fitting results can be obtained. In this study, we use a commercially available computer program for the non-linear regression data analysis, SigmaPlot SigmaPlot s non-linear curve fitter uses a least-squares procedure (Marquardt Levenberg algorithm) to determine the parameters that minimize the sum of the squares of differences between the dependent variable in the equations and the observations. RESULTS As an example, the new method was used to analyse the calciuminduced autoactivation of trypsinogen by trypsin. Trypsinogen, the zymogen form of trypsin, is secreted into the duodenum by pancreatic cells. Trypsin catalyses the activation of trypsinogen in an intermolecular autocatalytic process. The conversion of Table 1 Effect of Ca 2+ concentration on the steady-state kinetic parameters for trypsin-catalysed TAME hydrolysis The kinetic experiments were carried out in 40 mm Tris/HCl buffer (ph 8.1) at 30 C. [E] 0 = 50 nm and [TAME] = µm. k cat and K m were determined from the steadystate kinetic analysis. [Ca 2+ ] (mm) k cat (min 1 ) K m (µm) trypsinogen into trypsin involves the removal of the N-terminal hexapeptide H 2 N-Val-Asp-Asp-Asp-Asp-Lys [20]. This process is strongly stimulated by calcium ions. It has been shown that trypsinogen has two different binding sites for calcium ions while trypsin has only one. The high-affinity calcium-binding site (K d M) is common to both trypsinogen and trypsin. The binding of Ca 2+ on this site induces a stabilization of the conformation of these proteins which prevents the formation of inert proteins in the course of trypsinogen activation and protects trypsin against autolysis. The second site, with a much lower affinity for calcium ions (K d M), is found only in the zymogen and has been assigned to the two aspartyl residues 13 and 14 neighbouring the important Lys-15 Ile-16 bond that is split during activation. Binding of calcium to this site accelerates very clearly the rate of hydrolysis by trypsin of the Lys Ile bond [16,21,22]. To characterize the effect of calcium ions on the trypsincatalysed reaction, the kinetic parameters of trypsin-catalysed TAME hydrolysis were determined first. The initial velocities for the hydrolysis of TAME by trypsin were measured under different Ca 2+ concentrations at ph 8.1, 30 C. By fitting the experimental data to the Michaelis Menten equation, the values of k cat and K m were determined. The steady-state kinetic parameters for the trypsin-catalysed TAME hydrolysis are listed in Table 1. It can be seen from Table 1 that both k cat and K m are independent of the concentration of Ca 2+, indicating that binding of Ca 2+ to trypsin is not required for activity of the enzyme. This result is in agreement with the experimental observation that calcium does not affect activation of a trypsinogen derivative in which the carboxylate groups are blocked [23]. In order to study the stimulation mechanism of calcium ions for the autoactivation of trypsin, the activation kinetics of trypsin was monitored at several fixed concentrations of Ca 2+. The trypsinogen was incubated with trypsin in 100 µl of reaction mixture containing 40 mm Tris/HCl (ph 8.0) and different concentrations of Ca 2+ at 30 C. At defined time intervals, an aliquot (5 µl) was taken from the reaction mixture and assayed for enzyme activity. Enzyme activity assays were carried out under kinetically valid conditions with TAME as a substrate. Figure 1 shows time courses for trypsin autoactivation in the presence of 10 and 50 mm calcium ions. The time required for activation increases at lower enzyme concentration, and the maximal trypsin activity was proportional to the total concentration of trypsin plus trypsinogen, indicating that the reaction went to completion in each case. As an example, fitting of eqn (14) to the experimental data in the presence of 10 mm Ca 2+ is shown in Table 2. In Table 2 the reaction time t is the dependent variable, and [E T ]and[t] 0 are the two independent variables. The best-fitting results were obtained with α = 0.047, K m = µmandk cat = min 1 by using the non-linear regression analysis program SigmaPlot 237 c 2004 Biochemical Society

238 700 J.-H. Liu and Z.-X. Wang Table 2 Determination of kinetic parameters of trypsinogen autoactivation Reaction time t is the dependent variable, and [E T *] and [T] 0 are the two independent variables. t (min) [E T ](µm) [T] 0 (µm) Figure 1 Autocatalytic activation of trypsinogen by trypsin (A) Effect of trypsinogen concentration on the time course for autoactivation in the presence of 10 mm Ca 2+ at 30 C. The symbols represent the experimental data. The total concentrations of trypsinogen plus trypsin are ( ) µm and( ) µm, respectively. The lines are the theoretical curves generated by using eqn (14) with α = 0.047, K m * = µm and k cat * = min 1.(B) Effect of trypsinogen concentration on the time course for autoactivation in the presence of 50 mm Ca 2+ at 30 C. The symbols represent the experimental data. The total concentrations of trypsinogen plus trypsin are ( ) µm and( ) µm, respectively. The lines are the theoretical curves generated by using eqn (14) with α = 0.038, K m * = µmandk cat * = min Similarly, when the time courses for trypsin activation at two different enzyme concentrations were analysed by a global-fitting procedure simultaneously according to eqn (14), the values of K m, k cat and α could then be determined for each fixed concentration of Ca 2+. At the same concentrations of Ca 2+, the kinetic parameters so determined are quite close to those obtained by García-Moreno et al. [24], but are very different from those reported by Abita et al. [16]. In the presence of 50 mm Ca 2+, the value of K m was determined to be 14.5 µm, which is about 30 times lower than that obtained by Abita et al. [16] at 1 C (400 µm). Figure 2 shows the effect of increasing Ca 2+ concentration on the kinetic parameters of trypsin autoactivation. The dominant effect of the Ca 2+ concentration appears to be on K m, but it has no significant effect on k cat. On increasing the concentration of Ca 2+, K m decreases and approaches a limiting value. The kinetic parameters in the absence of Ca 2+ cannot be measured Figure 2 Effect of Ca 2+ concentration on K m * and k cat * (A) Effect of Ca 2+ concentration on K m * for autoactivation of trypsinogen by trypsin. Inset: plot of K m * versus 1/[Ca 2+ ]. (B) Effect of Ca 2+ concentration on k cat * for autoactivation of trypsinogen by trypsin. experimentally since trypsinogen is very unstable under this condition. In the present study, the lowest concentration of Ca 2+ used was 2.5 mm when the tight Ca 2+ -binding site is almost c 2004 Biochemical Society 238

239 Ligand-induced autocatalytic kinetics 701 Reaction mechanism of Ca 2+ -induced trypsinogen auto- Scheme 2 activation saturated. By fitting eqn (5) to the experimental data, the parameters were determined to be K S = µm, K d = mm and K d = mm. Because K d derived from the curve-fitting of the data to eqn (5) had a small negative value with a very large standard deviation, such that its confidence interval would include the value of zero, we suspected that K d might be in fact zero. In this case, the expressions of K m and k cat can be simplified to k cat = k 2 K m = (K d + [L])K S [L] (17) (18) The apparent turnover number k cat is independent of [L] and the plot of K m against [L] is a descending curve. At very low [L] values the plot approaches the K m axis, and at very high [L] values it approaches a limiting value, K S. A straight line will be obtained if K m is plotted against 1/[L] as shown in the inset of Figure 2. When the experimental data of Figure 2 were fitted to eqn (18), a remarkable correspondence was observed. The continuous line in Figure 2 represents the best fit of the experimental data to eqn (18), yielding K S = µm andk d = mm. Since K S, K S, K d and K d are the true dissociation constants, our data suggest that the binding affinity for Ca 2+ -bound trypsinogen to trypsin is at least two orders of magnitude higher than that for Ca 2+ -free trypsinogen. It is to be noted that K d K S = K d K S when K d is very small and K S very large, and the enzyme E can only bind to EL to form the ternary complex E EL, i.e. E cannot form the complex E E, and the mechanism can be written as shown in Scheme 2. DISCUSSION As mentioned above, calcium ions have two diametrically opposite effects on trypsinogen. They markedly increase hydrolysis of the Lys Ile bond near the N-terminus of the chain and almost totally suppress hydrolysis of certain other linkages, which in the absence of calcium are responsible for conversion of the precursor into inert proteins. The Lys Ile bond in the inert proteins is not split by trypsin even in the presence of calcium. This double effect is due to the binding of the calcium at two different sites. The high-affinity Ca 2+ -binding site in trypsin was first identified and described in detail by Bode and Schwager [25] based on analysis of crystal structure and the Ca 2+ content of their crystal. Trypsinogen and trypsin have an apparently identical specific site for Ca 2+ [21]. The binding of Ca 2+ on this site induces a structural change and stabilizes protein towards thermal denaturation or autolysis, but it is not essential for enzyme activity of trypsin. A second Ca 2+ -binding site exists only on the zymogen. This binding site is on the two N-terminal aspartyl residues of trypsinogen, Asp- 13 and Asp-14. Binding of calcium at this site does not seem to involve any structural reorganization and has no effect on the rate of formation of inert proteins, but it is required for complete and efficient activation of trypsinogen [16]. There is no corroborative crystallographic evidence as yet, since Ca 2+ was excluded from the crystallizing solution [26]. With trypsinogen, when the concentration of Ca 2+ increases from 2.5 to 50 mm, K m decreases by a factor of 11 and k cat does not change. Extrapolation of the experimental data to zero and high Ca 2+ concentration suggests that the K m value in the absence of Ca 2+ is very much greater than that at saturating Ca 2+ concentration. Therefore, the occupancy of the second site enhances greatly the binding affinity of trypsinogen to trypsin without changing the rate of decomposition of the trypsinogen trypsin complex. Tryptic hydrolysis of the nonapeptide Val-Asp 4 - Lys-Ile-Val-Gly and of the heptapeptide Val-Asp 2 -Lys-Ile-Val- GlyhavebeenshowntobeCa 2+ -dependent [16,22]. However, Ca 2+ ion binding to Asp-13 and Asp-14 together seems to be insufficient to explain the zymogen Ca 2+ ion effect since the change in K m causedbyca 2+ binding to the N-terminal of trypsinogen is much greater than that to the model peptides. One possible explanation for this result would be that the N-terminal aspartyl residues make an important contribution to the stabilization of the zymogen by forming hydrogen bonds or salt linkages with other side chains of the protein. The binding of Ca 2+ to the N-terminal of zymogen disrupts this interaction and results in exposure of the side chains, which provide an additional docking site for trypsin binding, and therefore increase the binding affinity to trypsin. Most protein kinases share a common molecular organization composed of a regulatory or inhibitory domain and a catalytic kinase domain having many of the conserved features. The holoenzymes are usually in an inactive or basally activated state, and cellular activation of protein kinases occurs as a result of conformational changes induced by binding of regulatory ligands, such as cyclic nucleotides, Ras, Rho, Ca 2+ /calmodulin or cyclin, or phosphorylation by upstream protein kinases [27,28]. Although the events which trigger activation of protein kinases are quite variable, a common molecular mechanism has been proposed [29]. A key feature in control of kinase enzymic activity is a regulated interaction between the inhibitory domain and the catalytic kinase domain. The regulatory ligands bind to the inhibitory domain and disrupt this interaction, thereby relieving initial inhibition of the kinase. The conformational changes at the active site will result in increased accessibility to MgATP, a protein substrate. With some protein kinases, binding of ligand to the regulatory domain can stimulate autophosphorylation of the catalytic domain, and the autophosphorylation is required to fully activate the enzyme. For example, PAK1 (p21-activated protein kinase 1) exists in a closed conformation in the absence of Cdc42 due to an interaction between the inhibitory domain and the kinase domain [30]. The binding of Cdc42 to the inhibitory domain leads to a conformational change. The conformational change withdraws the inhibitory domain from the cleft of the kinase domain and releases the activation loop. The activation loop contains one critical phosphorylation site, Thr-423. In the absence of phosphorylation, this loop is either disordered or in a conformation that is not optimal for catalysis. Cdc42-stimulated autophosphorylation of Thr-423 in the activation loop will activate the enzyme. Therefore, the kinetic analysis described here is also applicable to study the ligand-induced autophosphorylation of protein kinases. This work was supported in part by grants from NIH (R03TW01501), Natural Science Foundation of China ( ) and the Ministry of Science and Technology of China (G ). REFERENCES 1 Cohen, P. (1976) Control of Enzyme Activity, John Wiley and Sons, New York 2 Smith, J. A., Francis, S. H. and Corbin, J. D. (1993) Autophosphorylation: a salient feature of protein kinases. Mol. Cell. Biochem. 127/128, Hadorn, B. (1974) Pancreatic proteinases; their activation and the disturbances of this mechanism in man. Med. Clin. North Am. 58, c 2004 Biochemical Society

240 702 J.-H. Liu and Z.-X. Wang 4 Malcinski, G. M. and Bryant, S. W. (1984) Pattern Formation: a Primer in Developmental Biology, Macmillan, New York 5 MacDonald, H. R. and Nabholz, M. (1986) T-cell activation. Annu. Rev. Cell Biol. 2, Müller-Eberhard, H. J. (1988) Molecular organization and function of the complement system. Annu. Rev. Biochem. 57, Neurath, H. and Walsh, K. A. (1976) The role of proteases in biological regulation. In Proteolysis and Physiological Regulation (Ribbons, E. W. and Brew, K., eds.), pp , Academic Press, New York 8 Löffler, G. and Petrides, P. E. (1988) Physiologische Chemie, pp , Springer, Berlin 9 Thornberry, N. A. and Lazebnik, Y. (1998) Caspases: enemies within. Science 281, Smith, W. L. and Borgeat, P. (1985) The eicosanoids: prostaglandins, thromboxanes, leukotrienes, and hydroxy-eicosaenoic acids. In Biochemistry of Lipids and Membranes (Vance, D. E. and Vance, J. E., eds.), pp , Benjamin-Cunnings, Menlo Park, CA 11 Tans, G., Rosing, J., Berrettini, M., Lammle, B. and Griffin, J. H. (1987) Autoactivation of human plasma prekallikrein. J. Biol. Chem. 262, Karlson, P. (1988) Biochemie, 13th edn, pp , Georg Thieme Verlag, Stuttgart 13 Tans, G., Rosing, J. and Griffin, J. H. (1983) Sulfatide-dependent autoactivation of human blood coagulation Factor XII (Hageman Factor). J. Biol. Chem. 258, Wu, J. W., Wu, Y. and Wang, Z. X. (2001) Kinetic analysis of a simplified scheme of autocatalytic zymogen activation. Eur. J. Biochem. 268, Wang, Z. X. and Wu, J. W. (2002) Autophosphorylation kinetics of protein kinases. Biochem. J. 368, Abita, J. P., Delaage, M. and Lazdunski, M. (1969) The mechanism of activation of trypsinogen. The role of the four N-terminal aspartyl residues. Eur. J. Biochem. 8, Laidler, K. J. and Bunting P. S. (1973) The Chemical Kinetics of Enzyme Action, 2nd edn, pp , Oxford University Press, Oxford 18 Walsh, K. A. (1970) Trypsinogens and trypsins of various species. Methods Enzymol. 19, Wang, S. S. and Carpenter H. (1968) Kinetic studies at high ph of the trypsin-catalyzed hydrolysis of N-α-benzoyl derivatives of L-arginamide, L-lysinamide, and S-2-aminoethyl-L-cysteinamide and related compounds. J. Biol. Chem. 243, Colomb, E. and Figarella, C. (1979) Comparative studies on the mechanism of activation of the two human trypsinogens. Biochim. Biophys. Acta 571, Delaage, M. and Lazdunski, M. (1967) The binding of Ca 2+ to trypsinogen and its relation to the mechanism of activation. Biochem. Biophys. Res. Commun. 28, Delaage, M., Desnuelle, P. and Lazdunski, M. (1967) On the activation of trypsinogen. A study of peptide models related to the N-terminal sequence of the zymogen. Biochem. Biophys. Res. Commun. 29, Radhakrishnan, T. M., Walsh, K. A. and Neurath, H. (1967) Relief by modification of carboxylate groups of the calcium requirement for the activation of trypsinogen. J. Am. Chem. Soc. 89, García-Moreno, M., Havsteen, B. H., Varón, R. and Rix-Matzen, H. (1991) Evaluation of the kinetic parameters of the activation of trypsinogen by trypsin. Biochim. Biophys. Acta 1080, Bode, W. and Schwager, P. (1975) The single calcium-binding site of crystallin bovin beta-trypsin. FEBS Lett. 56, Kossiakoff, A. A., Chambers, J. L., Kay, L. M. and Stroud, R. M. (1977) Structure of bovine trypsinogen at 1.9 Å resolution. Biochemistry 16, Taylor, S. S., Knighton, D. R., Zheng, J., Ten Eyck, L. F. and Sowadski, J. M. (1992) Structural framework for the protein kinase family. Annu. Rev. Cell Biol. 8, Norbury, C. and Nurse, P. (1992) Animal cell cycles and their control. Annu. Rev. Biochem. 61, Kemp, B. E., Parker, M. W., Hu, S., Tiganis, T. and House, C. (1994) Substrate and pseudosubstrate interactions with protein kinases: determinants of specificity. Trends Biochem. Sci. 19, Lei, M., Lu, W., Meng, W., Parini, M. C., Eck, M. J., Mayer, B. J. and Harrison, S. C. (2000) Structure of PAK1 in an autoinhibited conformation reveals a multistage activation switch. Cell 102, Received 8 September 2003/26 November 2003; accepted 23 December 2003 Published as BJ Immediate Publication 5 January 2004, DOI /BJ c 2004 Biochemical Society 240

241 Biochem. J. (2004) 382, (Printed in Great Britain) 433 Thermal and conformational stability of Ssh10b protein from archaeon Sulfolobus shibattae Su XU, Sanbo QIN and Xian-Ming PAN 1 National Laboratory of Biomacromolecules, Institute of Biophysics, Academia Sinica, Beijing , China The secondary structure of the DNA binding protein Ssh10b is largely unaffected by change in temperature between 25 Cand 85 C, indicating that the protein is highly thermostable. Here, we report the temperature-dependent equilibrium denaturation of Ssh10b in the presence of guanidine hydrochloride (GdnHCl). It was found that the transition midpoint values of the temperature (T m ), and changes of enthalpy ( H m ) and entropy ( S m )of Ssh10b unfolding were linearly decreasing with increasing GdnHCl concentration. The true values of the thermodynamic parameters, T m = 402 K, H m = kj mol 1 and S m = kj T 1 mol 1, were obtained by linear extrapolation to 0 M GdnHCl. The value of the heat capacity change of Ssh10b unfolding, C p = kj T 1 mol 1 (approx. 19 J T 1 mol residue 1 ), was obtained from the measured thermodynamic parameters. This is significantly smaller than that of the average value for mesophilic proteins (50 J K 1 mol residue 1 ) or the value calculated from the Ssh10b structural data (64 J T 1 mol residue 1 ). A consequence of the small C p is that the G of Ssh10b is larger than that of mesophilic proteins, while the values of H and T S are smaller. The small C p of Ssh10b appears to result mainly from the presence of compactness in the denatured state. Key words: compact denatured state, enthalpy change, entropy change, equilibrium denaturation, hyperthermophilic protein, thermodynamic parameters. INTRODUCTION Hyperthermophiles constitute a group of micro-organisms with an optimum growth temperature of between 80 C and 100 C. They are generally found close to the root of the phylogenetic tree, which suggests that they preceded their mesophilic counterparts in the course of evolution [1 3]. Proteins isolated to date from the hyperthermophiles are composed of the 20 common amino acids. Furthermore, homologous hyperthermophilic and mesophilic proteins typically show % sequence similarity, and their three-dimensional structures are superimposable [3]. Hyperthermophilic proteins show considerable thermal stability and thereby provide important models for the investigation of fundamental problems in structure biology, such as the stability and folding of proteins [4 7]. However, although a number of hyperthermophilic and thermophilic proteins have been purified and compared with their mesophilic counterparts, progress in understanding the physicochemical basis for the stability of extremely thermophilic proteins has been hampered by the very limited structural and thermodynamic data available [8]. Attempts to obtain thermodynamic information about thermostable proteins have been plagued by irreversibility of unfolding and/or inaccessibility of the unfolding transition to physical measurement. Hence, studies of the thermodynamic stability of thermophilic and hyperthermophilic proteins have been reported for only a few select examples [9 25]. It has been shown that a higher transition temperature (T) in a hyperthermophilic protein could be obtained by three theoretical mechanisms: (1) by shifting the stability curve to increase the overall unfolding free energy at any temperature; (2) by decreasing the heat capacity change of unfolding ( C p ) between the fol- ded and unfolded states to flatten the stability curve; or (3) by shifting the stability curve toward higher temperature [9,10]. Comparison of the change in free energy upon unfolding ( G)versus temperature (T) curves for mesophilic, thermophilic and hyperthermophilic proteins demonstrates that thermophilic and hyperthermophilic proteins use various combinations of these three mechanisms to reach their superior thermodynamic stabilities [3]. For example, the greater protein stability and resistance to higher temperature of Sac7d [11] and Sso7d [12] were found to stem from an increase in stability at all temperatures, with a broader maximum, as compared with their mesophilic homologues (mechanisms 1 and 2, above). The thermophilic homologue of phosphoglycerate kinase [9,10] is stabilized by a small C p (mechanism 2). The maximum stability of cytochrome c-552 [9,10] and cold-shock protein (CspTm) [13] is increased and shifted to higher temperature compared with their mesophilic homologues (mechanisms 1 and 3). In the case of glutamate dehydrogenase from the hyperthermophile Thermotoga maritima [15], interdomain interactions play a significant role in its enhanced stability (mechanism 1). Deutschman and Dahlquist [22] found that changes in both H and C p are important contributors to the thermal stability of TmY as compared with BsY (mechanisms 1 and 2). The free energy of stabilization, G = H T S, of a protein is the difference between the free energy of the folded and the unfolded states of that protein, where H = H 0 + C p (T T 0 ); S = S 0 + C p ln(t/t 0 ); H 0 and S 0 are the standard enthalpy and entropy of unfolding at T 0 = 298 K respectively; and C p is the heat capacity change between the native and unfolded states of the system. The C p is relatively independent of temperature within the range C [26]. Abbreviations used: ASA ap, the change in solvent-accessible surface area on unfolding of apolar groups; ASA pol, the change in solvent-accessible surface area on unfolding of polar groups; cal, calculated; C p, change in heat capacity upon unfolding; DLS, dynamic light scattering; ex, experimentally determined; G, change in free energy upon unfolding; GdnHCl, guanidine hydrochloride; H, change in enthalpy upon unfolding; H 0, standard enthalpy of unfolding at T 0 = 298 K; IPTG, isopropyl β-d-thiogalactoside; S, change in entropy upon unfolding (the addition of a subscript m to any parameter, e.g. S m, indicates the transition midpoint value); S 0, standard entropy of unfolding at T 0 = 298 K. 1 To whom correspondence should be addressed ( xmpan@sun5.ibp.ac.cn). 241 c 2004 Biochemical Society

242 434 S. Xu, S. Qin and X.-M. Pan Thus, H 0, S 0 and C p are the intrinsic parameters for describing the thermodynamic properties of protein folding. Furthermore, these three macroscopic thermodynamic parameters of protein folding have direct microscopic interpretations. H 0 is the measurement of changes (at standard temperature 298 K) in packing forces of side chains, while S 0 reflects the numbers of accessible conformational states between the native and the unfolded states, including the contribution of solvent. C p is the measurement of change in accessible surface area between the native and the unfolded states of the protein. A microscopic understanding of hyperthermophilic protein stability in terms of protein structure should compare the parameters H 0, S 0 and C p of the hyperthermophilic protein with those of the mesophilic protein. The DNA binding protein Ssh10b from the archaeon Sulfolobus shibattae, a member of the sac10b family, is conserved in most thermophilic and hyperthermophilic archaeal genomes that have been sequenced to date. The protein constitutes approx. 4 5 % of the cellular protein, binds double-stranded DNA without apparent sequence specificity and is capable of constraining negative DNA supercoils in a temperature-dependent fashion. This binding ability is weak at 25 C, but is enhanced substantially at 45 C or higher temperature. Ssh10b is a dimeric protein composed of two identical subunits. The monomer of Ssh10b consists of 97 amino acid residues with no disulphide bonds, and the protein is highly thermostable [27]. The crystal structure of Ssh10b reveals a mixed α/β structure comprising four β-strands and two α-helixes [28]. In the present study, we report the temperature-dependent equilibrium denaturation of Ssh10b in the presence of guanidine hydrochloride (GdnHCl), monitored by CD spectra. We found that in the presence of GdnHCl, protein unfolding is fully reversible and the temperature of the unfolding transition becomes accessible to physical measurement. The thermodynamic parameters, H m, T m and S m, for Ssh10b unfolding were found to decrease with increasing GdnHCl concentration. The true values for these thermodynamic parameters were obtained by linear extrapolation to 0 M GdnHCl. The value of C p of Ssh10b unfolding is kj T 1 mol 1, which is significantly smaller than that of the average value for mesophilic proteins. Our results suggest that the small C p of Ssh10b results from the presence of compactness in the denatured state. MATERIALS AND METHODS Materials GdnHCl was from Gibco BRL, IPTG (isopropyl β-d-thiogalactoside) from Merck, and foline phenol was from Sigma. The ssh10b gene was provided by Professor Jinfeng Wang (Institute of Biophysics, Academia Sinica, Beijing, China) and Professor Li Huang (State Key Laboratory of Microbial resources, Institute of Microbiology, Chinese Academy of Sciences, Beijing, China) [27]. Protein purification An overnight culture of Escherichia coli (BL21) cells containing the target plasmid was added to 1 litre of Luria Bertani media plus ampicillin (60 µg/ml) and incubated at 37 C with shaking at 220 r.p.m. At D 600 = 0.8, expression was induced with 1 mm IPTG. After an additional 3 h at 37 C, the cells were harvested by centrifugation (6000 g at 4 C for 20 min). The cell pellet was then resuspended in 5 vol. of buffer (20 mm Na 2 HPO 4 /NaH 2 PO 4, ph 6.8, 1 mm EDTA) and the cells were sonicated on ice at 40 W for 10 min. The lysate was centrifuged at 6000 g for 15 min at 4 C. The supernatant was maintained at 80 C for 20 min in a temperature controller (Huber CC130) and then centrifuged at g for 15 min at 4 C. Ammonium sulphate was added gradually to the supernatant to a final concentration of 70 % (w/v). After centrifugation at 6000 g for 30 min at 4 C, the precipitate was dissolved in 5 ml of buffer A (20 mm Na 2 HPO 4 /NaH 2 PO 4, ph 7.5, 1 mm EDTA) and dialysed against the same buffer. After centrifugation at g, the sample was applied to an UNO S6 column (Biorad), which had been equilibrated in buffer A. Proteins were eluted with buffer B (20 mm Na 2 HPO 4 /NaH 2 PO 4, ph 7.5, 1 M NaCl, 1 mm EDTA) [27]. The concentrated sample was then dialysed against buffer A. The purity of protein was verified by SDS/PAGE. Unfolding studies Unfolding of Ssh10b by GdnHCl was monitored by CD using a Jasco-J720 spectropolarimeter. The CD signal in the region nm (far-uv CD) was monitored using a rectangular quartz cuvette with a pathlength of 1 mm. The temperature was increased from 298 K to 358 K, monitored using a thermometer attached to the thermosatted cell holder. The samples containing various concentrations of GdnHCl were equilibrated for 3 6 h before experiments were performed. Samples were heated at 1 C per minute from 25 Cto85 C in a cell sealed with parafilm to prevent evaporation, while keeping the pressure relatively constant. At the end of each experiment the samples were cooled to room temperature to check the reversibility. Analysis of the denaturation data The thermodynamic properties of Ssh10b were calculated assuming a two-state denaturation process N 2 k 2U. Concentrations of the folded protein [N 2 ] (in dimer units) and the unfolded protein [U] (in monomer units) at different temperatures or denaturant concentrations were calculated [29]: y y U + m U T [D] [N 2 ] = P t /2 y N + m N T [D] y U + m U T [D] y N + m N T [D] y [U] = P t y N + m N T y U + m U T [D] where P t is the total protein concentration in monomer units; y is the experimentally measured signal value at a given temperature T or given denaturant concentration [D]; y N and y U,aretheintercepts; and m N,andm U are the slopes of the native and unfolded baselines, respectively. The apparent equilibrium constant (K) and the corresponding free energy change ( G) at temperature T or denaturant concentration [D] were calculated according to: 2 P t K = y N +m N T [D] y U +m U T [D] [y N +m N T [D] y] 2 y y U +m U T [D] (3) G = RT ln K (4) where R is the gas constant and T is the absolute temperature. According to the linear free energy model [30 33], the changes in free energy, enthalpy, and entropy changes that occur on unfolding are expected to vary linearly with the denaturant concentration, [D]: G = G(H 2 O) m G [D] (5) (1) (2) c 2004 Biochemical Society 242

243 Thermal and conformational stability of Ssh10b 435 H = H (H 2 O) m H [D] (6) S = S(H 2 O) m S [D] (7) where G(H 2 O), H(H 2 O) and S(H 2 O) represent the free energy, enthalpy and entropy changes of unfolding in the absence of denaturant; and m G, m H and m S are the slopes of the transition for the free energy, enthalpy and entropy changes respectively. The midpoint of the transition (C m ), where 50 % of the protein was unfolded, is a function of protein concentration P t : C m = C 0 m + RT ln(p t) m G (8) where C 0 = G(H 2O) m m G is the transition midpoint at protein concentration P t = 1 M. For thermal unfolding, assuming that the heat capacity change ( C p ) between the native and unfolded states of the system is relatively independent of temperature within the range of the transition, then [34,35]: G(T ) = RT ln(k) = H m T S m + C P ((T T m ) ( + T ln 1 T T )) m T where H m,and S m are the enthalpy and entropy changes, respectively, at the transition midpoint, where T = T m. Within the transition range, where T ln(1 T Tm ) T T m T from Equation 9 can be simplified to the van t Hoff plot: ln(k) = S m R H m RT (9) (10) The temperature of the transition midpoint (T m ) is a function of the protein and denaturant concentration, as follows: T m = H m S m R ln(p t ) = H m (H 2 O) m H [D] S m (H 2 O) m S [D] R ln(p t ) (11) Hydrodynamic radius measurements Dynamic light scattering (DLS) measurements were performed at a protein concentration of 10 mg/ml at ph 6.8 at various concentrations of GdnHCl. Before the DLS measurements, the solutions were centrifuged at g for 10 min. All experiments were performed at a 90 scattering angle on a DynaPro-LSR (Protein Solution, Charlottesville, VA, U.S.A.) at a wavelength of nm. During the measurements, the temperature was maintained at 20 C using a built-in temperature controller. The diffusion coefficient (D) was calculated from the autocorrelation function, using the accessory software DYNAMICS provided with the instrument. Since the structure of the protein does not change at GdnHCl concentrations lower than 4 M, and the viscosity of GdnHCl solutions differs significantly from that of pure water, the diffusion coefficient D was corrected not to pure water, but to 0.5 M GdnHCl. The experimentally measured hydrodynamic radius, R ex H, of the protein was determined from the Stokes Einstein relationship [36]: R ex H = kt 6πηD (12) Figure 1 CD spectra of 50 µm Ssh10b proteinin phosphate bufferat ph 6.8 Line 1, native protein measured at 25 C; line 2, native protein measured at 85 C; line 3, denatured protein in 6 M GdnHCl measured at 25 C; line 4, denatured protein in 6 M GdnHCl measured at 85 C; line 5, Ssh10b protein denatured in 7 M GdnHCl and then refolded by 10-fold dilution to give final concentrations of 0.7 M for denaturant and 50 µm for protein; line 6, Ssh10b protein denatured in 3.5 M GdnHCl at 85 C and then refolded by re-equilibrating to 25 C and diluting 7-fold to give final concentrations of 0.5 M for denaturant and 50 µmfor protein. Here k is the Boltzmann s constant, T is the absolute temperature and η is the solvent viscosity. The hydrodynamic radius R H of the protein was calculated using the native protein (in 0.5 M GdnHCl) as a reference molecule: R cal R H = R (N) ex H H (13) RH ex (N) where R ex H (N) is the experimentally determined hydrodynamic radius of the native protein and R cal H (N) is the calculated hydrodynamic radius of the native protein dimer. The theoretical hydrodynamic radius of the native and fully denatured protein were calculated according to the method of Wilkins et al. [36]: R cal H = ( )N (14) for native protein (dimer) and: R cal H = ( )N (15) for fully denatured protein (monomer), where N is the number of residues in the protein. The compaction factor C is defined as [36]: C = [R H (D) R H ]/[R H (D) R H (N)] (16) where R H (D) andr H (N) represent the predicted hydrodynamic radii of the fully denatured and native protein, respectively. RESULTS Denaturant-induced unfolding Figure 1 shows typical CD spectra of the native protein, consistent with the native CD spectrum of Ssh10b reported previously [27], measured at 25 C and ph 6.8 (Figure 1, line 1) and 85 C(Figure 1, line 2). The spectra are nearly identical, indicating that the 243 c 2004 Biochemical Society

244 4 436 S. Xu, S. Qin and X.-M. Pan Table 1 Thermodynamic parameters derived from GdnHCl denaturation for the conformational stability of Ssh10b Protein concentration, P t = 50 µm. T (K) G(H 2 O) (kj mol 1 ) m G (kj mol 1 M 1 ) C m *(M) * Calculated according to equation 8. Figure 2 GdnHCl induced unfolding curves The concentration (P t ) of Ssh10b was 50 µmatph6.8at15 C( ), 20 C( ), 25 C( ), 45 C( ), 65 C( ), 75 C( ), 80 C( ) and 85 C( ). The inset shows the protein concentration dependence of unfolding at 7 µm( ), 50 µm( ) and 70 µm( ). protein is thermally stable up to 85 C. Upon incubation in 8 M urea for 24 h, Ssh10b underwent no reduction in backbone CD signal (results not shown), indicating that Ssh10b is resistant to urea-induced denaturation. At 25 C and ph 6.8 in 6 M GdnHCl, Ssh10b loses approx. 80% of its CD signal (Figure 1, line 3). The CD spectrum of the denatured protein measured at 25 Cwas found to undergo no further change when the concentration of GdnHCl was increased to 8 M, or when the ph was varied between 2 and 10 (results not shown). The fact that the CD spectra do not change with further changes in the denaturing conditions, suggests that the residual structure may not be native-like structure, but rather an ensemble of compact states. The CD signal of the 6 M GdnHCl denatured protein increases with increasing temperature (Figure 1 line 4), implying that there is more residual structure at higher temperature. A similar phenomenon was observed in denaturation of Sac7d [11], but the reason remains unclear. GdnHCl-induced unfolding of Ssh10b was studied at ph 6.8, at several constant temperatures between 15 C and 85 C, by measuring far-uv CD at 222 nm (Figure 2). The transitions were characterized by a single sharp change in the ellipticity that looks like a typical two-state transition. The observed denaturation midpoint decreases with increasing temperature (Table 1), indicating a decrease in the free energy of unfolding with increasing temperature. As shown in the inset to Figure 2, protein concentration dependence of the transition curves was observed at GdnHCl concentrations between 3 and 4.5 M, but a much smaller protein concentration dependence was seen at GdnHCl concentrations higher than 4.5 M. The free energy of unfolding, G(H 2 O), and the m value for the GdnHCl-induced unfolding curve were obtained from the fit to Equation 9. As shown in Figure 3, G(H 2 O) and the m value increased slightly with increasing temperature between 15 C and 45 C, then reached a maximum, before starting to decrease with further increase in temperature. Heat-induced unfolding in the presence of different concentrations of GdnHCl As Ssh10b is thermally stable up to 85 C (Figure 1, lines 1 and 2), the thermal unfolding of the protein was carried out in Figure 3 The unfolding free energy of Ssh10b ( G) versus temperature (T) Inset shows the unfolding m value of Ssh10b versus temperature. The concentration (P t )of Ssh10b was 50 µm. the presence of different concentrations of GdnHCl. At ph 6.8, increasing the temperature from 25 Cto85 C in the presence of GdnHCl results in unfolding of the protein by an apparent twostate mechanism (Figure 4). In certain cases, although the unfolding curve of a protein shows a single observable step, the transition may involve intermediates. Computer simulation of the Ssh10b experimental data could be fitted to both 2-state and 3-state equations within experimental error. However, the 3-state values of G(H 2 O) and H m are only approx. 10% larger than those obtained by 2-state fitting. Thus, the choice of model does not significantly affect the thermodynamic parameters obtained. The values obtained by fitting to a two-state model were used in the subsequent analysis. The van t Hoff plots (Figure 4, inset) demonstrate a good linear relationship between ln(k) and 1/T for each of the GdnHCl concentrations measured. The slope of the van t Hoff plot is H m /R and the intercept is S m /R (Equation 10). The values of H m and S m thus obtained are observed to decrease linearly with increasing GdnHCl concentration (Figure 5), as expected (Equations 6 and 7). Linear extrapolation to zero GdnHCl concentration gives the values H m = kj mol 1 and S m = kj K 1 mol 1. The observed unfolding transition temperature, T m, also appears to decrease linearly with increasing GdnHCl concentration c 2004 Biochemical Society 244