Real-time analysis of conformational control in electron transfer reactions of diflavin oxidoreductases

Transcription

1 Real-time analysis of conformational control in electron transfer reactions of diflavin oxidoreductases A thesis submitted to The University of Manchester for the degree of Doctor of Philosophy (PhD) in the Faculty of Science and Engineering. Tobias Michael Hedison 2016

2 For Celia and Paul Hedison.

3 Contents. Glossary of abbreviations List of Figures List of Tables List of Schemes Abstract Declaration Copyright statement Acknowledgements Preface and thesis structure Introduction Background Protein dynamics Protein dynamics and biological electron transfer Flavin and flavoenzymes Diflavin oxidoreductases Cytochrome P450 reductase (CPR) Physiological roles of CPR Structure of CPR Reaction mechanism of CPR Reaction mechanism of mammalian cytochrome P450s Dynamic properties of CPR Nitric oxide synthase (NOS) Physiological roles of NOS Structure of NOS Reaction mechanism of NOS Dynamic properties of NOS Calmodulin Fluorescence spectroscopy The phenomenon of fluorescence The phenomenon of FRET FRET and protein dynamics Aims of project

4 2. Real-time analysis of conformational control in electron transfer reactions of human cytochrome P450 reductase with cytochrome c Abstract Introduction Results FRET model of CPR domain dynamics Double-mixing stopped-flow assays of ET from NADPH-reduced CPR to cyt c Measurements of CPR domain dynamics by stopped-flow fluorescence spectroscopy Discussion Concluding remarks Experimental Procedures Materials Recombinant protein expression and purification Extrinsic fluorophore labeling Static fluorescence studies Stopped-flow studies and data fitting Correlating calmodulin landscapes with chemical catalysis in neuronal nitric oxide synthase using time-resolved FRET and a 5-deazaflavin thermodynamic trap Abstract Table of Content (TOC) Artwork Introduction Experimental Section Preparation of 5-dFMN reconstituted NOS Conjugation of extrinsic fluorophores to T34C/T110C-CaM Stopped-flow spectrophotometry Results and Discussion A FRET reporter of CaM conformation bound to nnos Electron transfer kinetics in nnos in the presence and absence of CaM Direct monitoring of CaM dynamics during catalytic turnover of nnos New form of nnos containing 5-deazaflavin-mononucleotide (5-dFMN) Internal electron transfer is prevented in 5-dFMN nnos Correlating conformational change with early steps in electron transfer in 5-dFMN nnos NADP + binding remodels the intracam landscape in nnos but over longer time scales

5 3.5 Concluding remarks Supporting Information Experimental Section Results A perspective on conformational control of electron transfer in nitric oxide synthases Abstract Introduction Structure of the NOS isoenzymes Reaction mechanism of NOS Domain dynamics of nitric oxide synthase (NOS) Probing NOS dynamics using intrinsic flavin fluorescence Probing NOS dynamics using external CaM-bound fluorophore fluorescence Probing NOS dynamics using external NOS-bound fluorophore fluorescence Conclusions Conclusions Conclusions Future perspectives References

6 Glossary of abbreviations. 5-dFMN 5-deazaflavin ethylenediaminetetraacetic EDTA mononucleotide acid 5-dRF 5-deazariboflavin enos endothelial NOS A acceptor or absorbance EPR electron paramagnetic resonance A555 Alexa Fluro 555 maleimide EQ state quasi equilibrium state A647 Alexa Fluro 647 maleimide ER endoplasmic reticulum A750 Alexa Fluro 750 maleimide ESI electrospray ionisation ADP adenosine diphosphate ET electron transfer AI auto-inhibitory insert FAD flavin adenine dinucleotide AMP adenosine monophosphate FeCN ferricyanide ATP adenosine triphosphate FMN flavin mononucleotide BWP band width pass FRET Förster resonance energy transfer c concentration H 4 B tetrahydrobiopterin C fluorescence H-transfer hydrogen-transfer CaM calmodulin IC internal conversion CD circular dichroism IET interflavin electron transfer cgmp cyclic guanosine monophosphate inos C MR mean residue concentration IPTG inducible NOS cnos constitutive NOS IR infra-red isopropyl β-d-1- thiogalactopyranoside CNS central nervous system ISC intersystem crossing CPR cytochrome P450 reductase J overlap intergral cryo-em cryo-electron microscopy k rate constant CT charge transfer or C- terminal tail k cat turnover number Cy5 cyanine 5 maleimide K d dissociation constant CYP cytochrome P450 kda kilodalton cyt c cytochrome c k for rate of forward reaction D donor KIE kinetic isotope effect DEAE Diethylaminoethyl K m Michaelis constant DL double labelled k rev rate of reverse reaction DT dithionite K s saturation constant DTT dithiothreitol l cuvette path length E efficiency of energy transfer L-Arg L-Arginine Ɛ molar extinction coefficient for absorbance LB Luria-Bertani E.coli Escherichia coli L-Cit L-Citrulline 6

7 mol equiv mole equivalent S 1 first singlet excited state MRE or [Ɵ] MR MSR NADP + NADPH mean residual ellipicities S 2 second singlet excited state methionine synthase reductase SAXS n refraction index SDS-PAGE oxidised nicotinamide adenine dinucleotide phosphate reduced nicotinamide adenine dinucleotide phosphate SI SiR small angle X-ray scattering sodium dodecyl sulphate polyacrylamide gel electrophoresis supporting information sulfite reductase ND not determined SL single labelled NHA N-hydroxy-L-Arginine SSL solid state labelling NMDA N-methyl-D-aspartate t time nnos NOSoxyFMN neuronal nitric oxide synthase NO nitric oxide tcrfk truncated form of NOS lacking the FAD-binding domain T 1 T-jump first triplet excited state C-terminal riboflavin kinase domain temperature-jump NR1 novel reductase 1 TS transition state P450 BM3 cytochrome P450 BM3 UV-Vis ultraviolet-visible PDA photodiode array WT wild type PDB protein data bank y0 y-ordinate intercept PELDOR pulsed electron-electron double resonance δ-ala δ-aminolevulinic acid PMT photomultiplier tubes ΔTGEE CPR CPR variant lacking four residues in linker region PNS peripheral nervous system κ orientation factor ppm parts per million λ max wavelength maximum QE quasi-equilibrium Φ 0 fluorescence quantum yield r R 0 RAS ROS S 0 distance Förster distance reflective anisotropy spectroscopy reactive oxygen species singlet ground state 7

8 List of Figures. Figure 1.1. The energy landscape of a protein molecule 19 Figure 1.2. Skeletal structures of the flavin cofactors..20 Figure 1.3. Redox and ionic states of flavin cofactors..21 Figure 1.4. Redox partners of CPR Figure 1.5. Structure of wild type rat CPR...26 Figure 1.6. Structures of open and closed conformations of rat CPR Figure 1.7. Structure and molecular architecture of mammalian NOS 36 Figure 1.8. NOS conformational equilibrium during catalysis.40 Figure 1.9. Structure and conformational states of mammalian calmodulin (CaM).. 41 Figure One version of a Jablonski diagram for a hypothetical molecule 42 Figure Relationship between FRET efficiency and donor-acceptor distance separation 44 Figure Set up used for stopped-flow FRET experiments 45 Figure 2.1. Open and closed conformations of rat CPR.52 Figure 2.2. Absorption and fluorescence spectra of CPR-DA..53 Figure 2.3. Cytochrome c reduction by CPR (i) 55 Figure 2.4. Cytochrome c reduction by CPR (ii) 57 Figure 2.5. Cytochrome c reduction by CPR (iii) 58 Figure 2.6. (Quasi) steady-state kinetics 60 Figure 2.7. Stopped-flow fluorescence of CPR-DA versus 40-fold excess NADPH 61 Figure 2.8. Stopped-flow fluorescence of CPR-DA versus stoichiometric NADPH 63 Figure 2.9. Double-mixing stopped-flow fluorescence of 1 NADPH-reduced CPR versus cyt c..64 Figure Schematic representation of the conformational equilibria of CPR.70 Figure 3.1. Structural organization and electron flow through nitric oxide synthase (NOS).77 Figure 3.2. Ligand binding and the dynamic landscape of CaM.82 Figure 3.3. Anaerobic stopped-flow transients obtained at 485 nm on mixing 5 µm NOS (final concentration) with a 20-fold excess of NADPH in the presence or absence of CaM 85 Figure 3.4. Dynamics of nnos-bound CaM during NADPH-driven nnos flavin reduction with native (panels A and B) and 5-dFMN (panel C and D) nnos...91 Figure 3.5. NADP + + nnos binding occurring in the dead time of the stopped-flow instrument..93 Figure 3.6. NADP + binding to nnos driving conformational change of CaM in the nnos CaM complex Figure S3.1. Comparison between native- and DA-T34C/T110C-CaM ( DA-CaM ) bound nnos steady-state turnover numbers 99 8

9 Figure S3.2. Fluorescence excitation and emission spectra of oxidized nnos.101 Figure S3.3. Fluorescence emission changes associated with the binding of A) 0.3 µm D- T34C/T110C-CaM or B) 0.3 µm A-T34C/T110C-CaM (black) with Ca 2+ (red) or both Ca 2+ and 0.3 µm nnos (blue).102 Figure S3.4. Fluorescence emission spectra of donor and acceptor fluorophores bound to T34C/T110C-CaM Figure S3.5. No Inter-CaM FRET across the nnos dimer.104 Figure S3.6. Absorbance changes, recorded by a photodiode array (PDA) stopped-flow instrument, 10 s after mixing of 5 μm (final concentration) of A) native or B) 5-dFMN reconstituted NOS with 20-fold excess NADPH in the presence (red) and absence (black) of CaM Figure S3.7. Time-dependent changes in fluorescence emission of fluorophores bound to the 0.3 μm single-labelled (donor or acceptor labelled - black) or 0.3 μm double-labelled (donor and acceptor labelled red) T34C/T110C CaM-oxidized nnos complex upon reduction with excess NADPH (100 μm, final concentration).107 Figure S3.8. Time-dependent changes in fluorescence emission of fluorophores bound to the 0.3 μm single-labelled (donor or acceptor labelled - black) or 0.3 μm double-labelled (donor and acceptor labelled red) T34C/T110C CaM-oxidized 5-dFMN nnos complex upon reduction with excess NADPH (100 μm, final concentration).108 Figure S3.9. Time-dependent changes in fluorescence emission of fluorophores bound to the 0.3 μm single-labelled (donor or acceptor labelled - black) or 0.3 μm double-labelled (donor and acceptor labelled red) T34C/T110C CaM-oxidized nnos complex upon reduction with excess NADP + (500 μm, final concentration) 110 Figure S3.10 The folding of native and 5-dFMN nnos is identical. 112 Figure S3.11. Reductive titration of native and 5-dFMN reconstituted NOS.113 Figure S3.12 (A) Structures of FMN (black) and 5-dFMN (red) Figure 4.1. Simplified two dimensional depiction of a multidimensional conformational landscape of a protein molecule 117 Figure 4.2. Structure and molecular architecture of NOS..118 Figure 4.3. Redox-dependent binding of CaM to nnos 124 Figure 4.4. nnos-bound CaM dynamics and reaction chemistry are kinetically coupled 125 Figure 4.5. CPR domain dynamics and reaction chemistry are kinetically coupled..126 thesis 9

10 List of Tables. Table 2.1. Oxidative half-reaction kinetics 51 Table 2.2. Comparison of FRET and UV-Vis stopped-flow kinetics..65 Table 3.1. Kinetic parameters extracted from Figure Table 3.2. Rate constants (k), relative fluorescence changes (ΔC), and ordinate intercept (y0) values determined from fitting donor/acceptor fluorescence transients (Figure 3.4B,D) to exponential decay functions.. 91 Table S3.1. Steady state k cat values for nnos turnover assays 100 Table S3.2. Observed KIE values for NADPH-driven nnos flavin reduction 106 Table S3.3. Donor and acceptor fluorophore emission changes extracted from fitting to transients seen in Figure 3.4A and 3.4C (deconvoluted FRET) and Figure S3.7 (native) and S3.8 (5-dFMN).109 Table S3.4. Donor and acceptor fluorophore emission changes extracted from fitting exponential functions to transients seen in Figure 3.6 and Figure S thesis 10

11 List of Schemes. Scheme 1.1. Reductive half-reaction of CPR 28 Scheme 1.2. Catalytic cycles of CYPs 30 Scheme 1.3. Reductive half-reaction of NOS.. 37 Scheme 1.4. Sequential oxidation of L-Arginine catalysed by NOS. 38 Scheme 2.1. The reductive half-reaction of CPR.. 50 Scheme S3.1. Electron flow through the NOS diflavin oxidoreductase to intrinsic and extrinsic electron accepting partners Scheme 4.1. Simplified reductive half reaction of nitric oxide synthase and related diflavin oxidoreductases. 120 Scheme 4.2. The sequential oxidation of L-Arginine (L-Arg) into NO and L-Citrulline (L-Cit) catalysed by NOS.121 thesis 11

12 Abstract. Real-time analysis of conformational control in electron transfer reactions of diflavin oxidoreductases A thesis submitted to The University of Manchester for the degree of Doctor of Philosophy (PhD) in the Faculty of Life Sciences by Tobias M. Hedison, How an enzyme achieves such high rates of catalysis in comparison to its solution counterpart reaction has baffled scientists for many decades. Much of our understanding of enzyme function is derived from research devoted to enzyme chemical reactions and analysis of static three-dimensional images of individual enzyme molecules. However, more recently, a role of protein dynamics in facilitating enzyme catalysis has emerged. It is often challenging to probe how protein motions are correlated to and impact on the catalytic cycle of enzymes. Nevertheless, this subject must be addressed to further our understanding of the roots of enzyme catalysis. Herein, this research question is approached by studying the link between protein domain dynamics and electron transfer chemistry in the diflavin oxidoreductase family of enzymes. Previous studies conducted on the diflavin oxidoreductases have implied a role of protein domain dynamics in catalysing electron transfer chemistry. However, diflavin oxidoreductase motions have not been experimentally correlated with mechanistic steps in the reaction cycle. To address these shortcomings, a real-time analysis of diflavin oxidoreductase domain dynamics that occur during enzyme catalysis was undertaken. The methodology involved specific labelling of diflavin oxidoreductases (cytochrome P450 reductase, CPR, and neuronal nitric oxide synthase, nnos) with external donor-acceptor fluorophores that were further used for time-resolved stopped-flow Förster resonance energy transfer (FRET) spectroscopy measurements. This approach to study enzyme dynamics was further linked with traditional UV-visible stopped-flow approaches that probed enzymatic electron transfer chemistry. Results showed a tight coupling between the kinetics of electron transfer chemistry and domain dynamics in the two diflavin oxidoreductase systems studied. Moreover, through the use of a flavin analogue (5-deazaflavin mononucleotide) and isotopically labelled nicotinamide coenzymes (pro-s/r NADP 2 H), key steps in the reaction mechanism were correlated with dynamic events in calmodulin, the partner protein of nnos. The approaches developed in this project should find wider application in related studies of complex electron-transfer enzymes. Altogether, this research emphasises the key link between protein domain motions and electron transfer chemistry and provides a framework to describe the relationship between domain dynamics and diflavin oxidoreductase function. 12

13 Declaration. No portion of the work referred to in the thesis has been submitted in support of an application for another degree or qualification of this or any other university or other institute of learning. Copyright statement. I. The author of this thesis (including any appendices and/or schedules to this thesis) owns certain copyright or related rights in it (the Copyright ) and he has given The University of Manchester certain rights to use such Copyright, including for administrative purposes. II. Copies of this thesis, either in full or in extracts and whether in hard or electronic copy, may be made only in accordance with the Copyright, Designs and Patents Act 1988 (as amended) and regulations issued under it or, where appropriate, in accordance with licensing agreements which the University has from time to time. This page must form part of any such copies made. III. The ownership of certain Copyright, patents, designs, trade marks and other intellectual property (the Intellectual Property ) and any reproductions of copyright works in the thesis, for example graphs and tables ( Reproductions ), which may be described in this thesis, may not be owned by the author and may be owned by third parties. Such Intellectual Property and Reproductions cannot and must not be made available for use without the prior written permission of the owner(s) of the relevant Intellectual Property and/or Reproductions. IV. Further information on the conditions under which disclosure, publication and commercialisation of this thesis, the Copyright and any Intellectual Property and/or Reproductions described in it may take place is available in the University IP Policy (see in any relevant Thesis restriction declarations deposited in the University Library, The University Library s regulations (see and in The University s policy on Presentation of Theses. 13

14 Acknowledgements. This PhD thesis represents a four year journey of emotional highs and lows. During my postgraduate studies I have been able to travel to different continents, to make many new friends and to drink with some of the best scientists in the world (as well as doing some work here and there). I feel privileged to be in the postion I am in now and for this I owe thanks to numerous people. First and foremost, I would like to thank Prof. Nigel S. Scrutton. When I first joined Nigel s lab as a third year undergraduate project student I had little to no interest in research. After completing this short project with Nigel, the only thing I wanted to do was study enzymology. I thank Nigel for granting me a position in his lab as well as providing an enormous amount of emotional and intellectual support throughout my studies. Thanks should also be given to Dr. Sam Hay. Sam has always been there to answer question after question throughout the years as well as helping me to analyse and interpret much of my data. I owe a great deal of gratitude to my mentor, Dr. Hanan Messiah. Hanan is one of the nicest people I have ever met and it has been a great privilege to work besides her. I will always remember how much time and effort Hanan has put into helping me understand the practical side of enzymology. Furthermore, thanks should be given to Dr. Derren Heyes. Derren has always been there for a good chat or to help me plan experiments or fix equipment. I would like to thank all past and present members of in the Scrutton, Hay, Munro and Leys groups. In particular, thanks should be given to Dr. Karl Fisher and Michiyo Sakuma. Both of which have played a significant part during my studies. Thanks should be given to my girlfriend and fellow scientist Andreea Iorgu. Living and working with Andreea has been fantastic, we discuss and help each other to solve any problems we encounter in the lab on a daily basis. Andreea has made me a better scientist, a calmer person and helped me realise there is much more to life (Te iubesc!). To my family I thank you for believing in me and supporting me through my life. This thesis is for my mother and father who give me the strength to carry on when times get hard. 14

15 Preface and thesis structure. This thesis is presented in the alternative format style with permission from The University of Manchester. The three experimental sections presented in this thesis (Sections 2, 3 and 4) are all published in peer reviewed journals. For consistency throughout this thesis some aspects of the three results sections have been modified. All Figures, Tables, Schemes and Equations are numbered sequentially throughout the thesis. Any supplementary Figures, Tables and Schemes are denoted with the prefix S. The contributions for each of the result sections in this thesis are as follows: Section 2: Publication title: Real-time analysis of conformational control in electron transfer reactions of human cytochrome P450 reductase with cytochrome c. (Published in the FEBS J, September 2015) Authors: Tobias M. Hedison, Sam Hay, Nigel S. Scrutton Contributions: TMH wrote the paper with help from SH and NSS. TMH conducted all the experiments. Analysis and interpretation of the data was conducted by all authors. Section 3: Publication title: Correlating calmodulin landscapes with chemical catalysis in neuronal nitric oxide synthase using time-resolved FRET and a 5-deazaflavin thermodynamic trap (Published in ACS Catalysis, June 2016) Authors: Tobias M. Hedison, Nicole G. H. Leferink, Sam Hay, Nigel S. Scrutton Contributions: TMH wrote the paper with help from NSS. NGHL synthesised the pro-r and pro-s NADP 2 H and the 5-dFMN compounds. TMH conducted all the experiments except for UV-Vis stopped-flow kinetics which were conducted by both NGHL and TMH. All data recorded were analysed and interpreted by TMH with guidance from NSS. Section 4: Publication title: A perspective on conformational control of electron transfer in nitric oxide synthases (In press in the Journal of Nitric Oxide, September 2016) Authors: Tobias M. Hedison, Sam Hay, Nigel S. Scrutton Contributions: TMH wrote the paper and conducted all experiments with help from NSS and SH. 15

16 1. INTRODUCTION 1. Introduction. 1.1 Background. Enzymes are protein molecules that catalyse biochemical reactions. The ability of an enzyme to enhance the rate of catalysis over the respective reference reaction in solution (in some cases rates recorded to be greater than 20 orders of magnitude higher than the non-catalysed solution reaction [1]) enables cells to carry out reactions that would normally not proceed on biological timescales. Owing to this, enzymes are essential for a wide-range of biochemical and physiological functions and therefore, understanding the origin(s) of enzyme catalysis is critical for broader scope. Despite over a hundred years of study, the precise root(s) of enzymatic catalysis are still unknown. This lack of knowledge is seen through difficulties in de novo design of novel enzymes. Designing an enzyme from scratch is the gold standard in our understanding of how enzymes function (attempts have been made using catalytic antibodies [2, 3] and with protein directed evolution [4-7] with little success). Currently, these challenges encountered in the field of enzymology have hampered the potential use of enzymes in healthcare, energy and chemical manufacturing industries. Presently, the best known model of enzyme action is Linus Pauling s transition state (TS) theory [8]. Pauling s model suggests the source of catalysis is accredited to tight binding between the enzyme and the transition state species, which is preferential to that of the enzyme and the substrate or the product molecules [8, 9]. Enzyme transition states typically have femtosecond (fs) lifetimes (time required for electron distribution and chemical bond formation/breaking) [10, 11], while enzyme turnover (k cat ) is typically seen to occur on the millisecond (ms) timescale [12, 13]. This discrepancy between the timescales of chemical steps and turnover makes it challenging to probe the transition state (transition state occupies typically of the total turnover time). However, many studies have shown that stabilisation of this state, through electrostatics, hydrogen bonding and desolvation, are the contributing factors towards chemical steps in enzyme turnover [9, 14-18]. In recent years particular attention has been given to the role of protein dynamics (timedependent movement of atomic coordinates) in enzyme catalysis [12, 13, 19-29]. Current hypotheses infer that fast active site localised protein dynamics (femtosecond-picosecond) contribute to chemical steps in turnover. These dynamics have been predicted from computational and kinetic isotope effects, but the lack of probes for direct observation makes 16

17 1. INTRODUCTION their study an area of intensive research and debate. In contrast, the role of slower large-scale protein dynamics (millisecond-second) in enzyme turnover is well characterised and was first proposed in 1958 by Koshland s induced fit model [30]. Slow dynamics are not directly involved in catalysing chemical steps (however they are shown to influence fast protein motions [19]), but, nonetheless, they are crucial to enzyme function by enabling substrate binding/product release and positioning catalytic residues, cofactors and inorganic ions for subsequent reaction chemistry. Identifying the properties and mechanistic trigger(s) of these slow dynamical processes is essential in understanding how enzymes function and is required for further developments in the field of enzymology [12]. This thesis addresses the relationship between slow timescale protein dynamics and biological electron transfer chemistry catalysed by the diflavin oxidoreductase family of enzymes. Diflavin oxidoreductases are multidomain flavoenzymes which, at the most basic characterisation level, contain a flavin adenine dinucleotide (FAD) binding domain and a flavin mononucleotide (FMN) binding domain separated by a flexible hinge region [31, 32]. A wealth of spectroscopic and structural data indicates that the complex flux of electrons in diflavin oxidoreductases is linked to large-scale protein domain dynamics [32-34]. However, these studies have not provided a direct read-out of conformational change that takes place concurrent with catalysis. In this thesis, to address these shortcomings, traditional UV-Vis and more novel Förster resonance energy transfer (FRET) stopped-flow spectroscopic techniques have been coupled in order to simultaneously monitor diflavin oxidoreductase reaction chemistry and their dynamics. The introduction of this thesis begins with an overview of protein dynamics and how dynamical events influence enzyme catalysis. This will be followed by a description of the two diflavin oxidoreductase systems investigated: cytochrome P450 reductase (CPR), a microsomal membrane anchored enzyme which transfers electrons to cognate P450 proteins; and neuronal nitric oxide synthase (nnos), a haem-containing diflavin oxidoreductase which produces nitric oxide (NO). Finally, strategies used in this thesis for studying protein dynamics in diflavin oxidoreductases will be discussed. This last section will include details on the principles governing the spectroscopic methods used for this study and how they are valuable in the study of biochemical processes. 1.2 Protein dynamics. Structure determination techniques (e.g. X-ray crystallography) are useful in providing atomistic insights into protein function. By determining the structure of a protein, one can identify key features, such as: mechanistically relevant amino acid residues, protein folding architecture, protein-protein interaction surfaces and the presence of cofactor and/or inorganic metal ions. 17

18 1. INTRODUCTION However, many of the structures determined through these methods represent the lowest energy state found within a protein conformational landscape [35-38]. Thus, static structures determined by these approaches are insufficient to provide a complete description of a protein s function [35-38]. The conformational landscape of a protein is a representation of the equilibrium of individual protein conformational substates (Figure 1.1) [35-38]. Conformational landscapes contain both hill and valley features (Figure 1.1A). Valley topographies represent (quasi-)stable conformational states, while the hill features represent the high energy barriers separating them. The interconversion between individual protein substates - described as any time-dependent change in atomic coordinates of a protein molecule - is believed, in part, to contribute towards several biological processes (e.g. enzyme catalysis [19, 23, 24, 39, 40] and signal transduction [41-44]). Therefore, to fully understand the function/mechanism of a protein, one must study protein dynamics. As mentioned in the subsection above (Section 1.1), enzyme molecules are crucial to a number of key biochemical processes. The study of enzyme dynamics is often challenging as enzyme conformational change occurs over a broad range of distance- ( angstrom) and time- (femtosecond-second) scales (Figure 1.1B) [19, 20]. However, due to much interest and debate in the field of enzyme dynamics, the field has significantly progressed in the last decades. In several enzyme systems, it has been observed that protein dynamics may contribute towards the binding of ligand [38, 45, 46] and/or partner proteins [47, 48], enhancing the rate of catalysis [12, 13, 22, 24, 25, 28] and gating chemical steps (such as electron transfer reactions) [47, 49-53] Protein dynamics and biological electron transfer. Biological electron transfer reactions can be categorised into three groups, differentiated by the rate-limiting step during electron transfer reactions: true electron transfer, coupled electron transfer and gated electron transfer [52-55]. True electron transfer reactions are rate limited by the electron transfer event and are governed by the difference in redox potentials of the reactants. Coupled electron transfer reactions are rate limited by the electron transfer event, but are preceded by a thermodynamically unfavourable reaction step. Gated electron transfer reactions are limited by a slow event prior to the electron transfer step. In the case of gated electron transfer, slow events prior to electron transfer can refer to chemical steps (e.g. proton transfer) or to large-scale dynamical changes within the protein molecule. 18

19 1. INTRODUCTION Figure 1.1. The energy landscape of a protein molecule. (A) Simplified two-dimensional depiction of protein s energy landscape. The equilibrium between the conformational states can be perturbed by mutagenesis, temperature, pressure, protein-protein interaction, redox chemistry and ligand/inhibitor binding [35, 36]. (B) Timescale, amplitude and function of several dynamic processes that occur in a protein molecule. Large-scale domain dynamics enable electron transfer proteins to catalyse efficiently inter- and/or intra-protein transfer processes [47-50]. As efficient electron transfer chemistry requires a distance shorter than 14 angstrom (on the basis of the Dutton ruler [52-55]), proteins that catalyse gated electron transfer utilise dynamic events to search for short edge-to-edge distances between redox centres. This dynamical search mechanism is thought to be employed by many electron transfer proteins, including many flavoenzymes which utilise the flavin cofactor as a redox centre (Sections 1.3, 1.4, 1.5 and 1.6). The properties of flavins and flavoenzymes are described below. 19

20 1. INTRODUCTION 1.3 Flavin and flavoenzymes. Flavins are organic compounds which contain a heterocyclic isoalloxazine ring (Figure 1.2) [56-58]. Three well-defined flavin molecules are known - riboflavin, flavin mononucleotide (FMN) and flavin adenine dinucleotide (FAD). Many proteins require flavin molecules to function [57-59]. These proteins systems, which are called flavoenzymes, have been studied extensively over the past 70 years and they are known to be involved in many key biological processes, including: biosynthesis, biodegradation, protein folding, DNA repair, chromatin remodelling, energy production, reduction, oxidation, oxygenation and light sensing [60, 61]. Figure 1.2. Skeletal structures of the flavin cofactors. The atomic numbers of the flavin isoalloxazine ring are shown. FMN- and FAD-bound proteins represent 1-3 % of the total eukaryotic and prokaryotic genome [57, 59, 61]. Due to intrinsic properties of their bound flavin cofactors, which can exist in oxidised, one-electron and two-electron reduced states (Figure 1.3), flavoenzymes are able catalyse a wide variety of redox reaction [57-59]. Flavin cofactors are typically non-covalently attached to their 20

21 1. INTRODUCTION respective apoprotein [59]. However, in a number of flavoenzymes, covalent interactions between cysteine, histidine or tyrosine and the flavin coenzyme are present [59]. This covalent protein-flavin interaction is autocatalytic and has been suggested to enhance the oxidative power of the flavin cofactor, as well as tuning redox potential, facilitating electron transfer, preventing cofactor modification and inactivation, saturating the active site of the flavoprotein and promoting protein stability [59]. Figure 1.3. Redox and ionic states of flavin cofactors. Flavoenzymes react using two half-reactions - a reductive and an oxidative reaction [62]. In the reductive half-reaction, an electron donor (e.g. NADPH) reduces the flavin cofactor, while in the oxidative half-reaction, an electron accepting partner (e.g. protein or molecular oxygen) receives electrons from the reduced flavoenzyme. These half-reactions can typically take place through a ternary complex or a ping-pong mechanism [62]. In the ternary complex of the enzyme, the electron donor and the electron acceptor are present concurrently during turnover, while the ping-pong mechanism functions through a sequential binding and release of the electron donor and electron accepting molecules, respectively [62]. 21

22 1. INTRODUCTION Unlike many other organic compounds, flavins are able to react with molecular oxygen [61, 63, 64]. The manner in which a flavoenzyme reacts with molecular oxygen typically determines its function [61, 63, 64]. This ability is characteristically dictated by the protein environment surrounding a flavin cofactor (hydrogen bonding, hydrophobic interactions, π-stacking, cation-π stacking and accessibility to solvent) [58, 61]. As no general structural rules govern how a flavoenzyme will react with oxygen, it is often challenging to define flavoprotein function based on the proximate flavin environment alone [61]. However, by combining structural studies with spectroscopic methods, both reaction mechanism and function of individual flavoenzymes can be probed. This combinatory approach is typically straightforward in the study of flavoenzymes which are spectroscopically active due to their characteristic optical features and EPR sensitive (paramagnetic semi-quinone species) features of the flavin coenzyme [65]. Out of the flavoenzymes studied to date, the flavin oxygenases, the flavin monooxygenases and the flavin dehydrogenase/electron transferases are the best described in the literature [57-61, 64]. Flavin oxidases and flavin monooxygenases both use molecular oxygen as an electron acceptor. The flavin oxidases react with dioxygen to produce hydrogen peroxide and convert their substrates (typically amines or amino acids). Flavin monooxygenases react with oxygen to produce a flavin (hydro)peroxide intermediate, which subsequently inserts a single atom of oxygen into the respective substrate, while concurrently reducing the other oxygen atom to water. Conversely, flavin dependent dehydrogenase/electron transferases react marginally or not at all with oxygen molecules. This lack of oxygen sensitivity enables flavin dependent dehydrogenases/electron transferases to catalyse complex redox reactions without forming celldamaging radical oxygen species (ROS). Flavin electron transferases represent around half of all proteins involved in electron transfer [66]. One physiologically important family of flavin electron transferases are the diflavin oxidoreductases. The known physiological roles, structure, mechanism and dynamic properties of diflavin oxidoreductases are addressed below. Here, particular focus will be paid to two diflavin oxidoreductases - cytochrome P450 reductase and neuronal nitric oxide synthase which are the subject of this thesis. 1.4 Diflavin oxidoreductases. Diflavin oxidoreductases are flavoenzymes that contain tightly bound FAD and FMN cofactors, separately housed in two discreet domains fused through a flexible hinge region [32]. Members of this family include cytochrome P450 reductase (CPR) [34, 49], the isoforms of nitric oxide synthase (NOS) [33, 67, 68], sulfite reductase (SiR) [69], methionine synthase reductase (MSR) 22

23 1. INTRODUCTION [70], cytochrome P450 BM3 (P450 BM3) [71] and novel reductase one (NR1) [72]. A wealth of spectroscopic and structural data have indicated that diflavin oxidoreductases undergo large-scale structural reorganisation during turnover [32-34, 49, 73]. These dynamics have been inferred to gate electron transfer from NADPH, through FAD and FMN bound cofactors, to a variety of single electron accepting redox centres (depending on the diflavin oxidoreductase). In this thesis, the role of dynamics during catalysis of the NOS and CPR enzymes has been studied. 1.5 Cytochrome P450 reductase (CPR) Physiological roles of CPR. CPR can be described as an endoplasmic reticulum (ER) membrane anchored flavoenzyme that catalyses the transfer of electrons to a multitude to reaction partners [74, 75]. These redox partners are shown in Figure 1.4. The majority of these electron acceptors have physiologically important roles and catalyse a number of key biochemical processes [75-79]. However, due to its promiscuity, CPR can accommodate a number of artificial electron acceptors, such as cytochrome c (cyt c) [80] and ferricyanide (FeCN) [81]. These artificial electron acceptors are often useful to measure CPR activity in vitro and have been employed in a number of studies to examine the electron transfer mechanism of diflavin oxidoreductases [82-89]. Though, since these artificial electron acceptors are not physiologically relevant to the CPR system [90], they will not be discussed in this subsection. The cytochrome P450 (CYP) superfamily represents one of the electron accepting partners of CPR [75, 91]. In humans there are 57 genes encoding CYPs and 1 gene encoding CPR [92]. Of these 57 CYPs, 50 are endoplasmic reticulum tethered and 7 are mitochondrial enzymes (which are involved in sterol biosynthesis) [93, 94]. As CPR is a microsomal enzyme, it will donate electrons to the 50 endoplasmic reticulum located CYPs [92]. Out of the 50 ER CYPs, 20 participate in biosynthesis of endobiotics, 17 metabolise xenobiotic compounds and 13 are orphan enzymes (with no defined physiological function) [93, 94]. All CYPs have a very similar catalytic mechanism that requires the step-wise transfer of electrons from a partner protein to their catalytic heme centre [95, 96]. This stepwise transfer facilitates the mixed-function oxidation role of CYPs where one molecule of dioxygen is consumed per CYP substrate. This dioxygen is split by CYPs. One oxygen atom is inserted into the product, forming a hydroxyl moiety; while the additional oxygen atom is reduced to water (the reaction mechanism of CYPs is further discussed in Section 1.5.4) [96-98]. 23

24 1. INTRODUCTION Figure 1.4. Redox partners of CPR. NADPH donates two electrons to CPR, which donates electrons to numerous partner proteins/small molecules. Physiological electron accepting partners are presented in bold text. Other than CYPs, many of CPR s partner proteins have been thoroughly described in the literature and are known to be essential for a number of key cellular functions. These partners include, but are not limited to: (i) cytochrome b 5 [77], which reduces CYPs (working in combination with CPR and cytochrome b 5 reductase) and regulates key steps in fatty acid metabolism (by transferring electrons to fatty acid desaturase and fatty acid elongase); (ii) squalene monooxygenase [78], which catalyses a key step in sterol biosynthesis (the epoxidation of a C-C bond in squalene to form 2,3-oxidosqualene), and (iii) haem oxygenase [76], an essential enzyme involved in haem catabolism (which degrades heme into bilverdin). The physiological importance of CPR was made evident through both knockout mice models [99, 100] and the phenotypes attributed to common CPR polymorphic variants [ ]. In knockout mice, germline deletion of the CPR gene causes multiple developmental deficiencies and is embryonically lethal. On the other hand, several of the CPR polymorphic variants are known to cause malfunctions in cholesterol biosynthesis and steroidogenesis, often leading to Antley-Bixler syndrome (a rare disorder characterised by skeletal and craniofacial malformations [104]). Due to the significance of the CPR system, numerous research groups have worked on determining how 24

25 1. INTRODUCTION this enzyme functions at a molecular level. Much of the data are reviewed in the subsections below (vide infra) Structure of CPR. CPR contains a 72 kda catalytic portion and an alpha helical 6 kda N-terminal membrane tether [75, 105, 106]. The structure of the catalytic portion of CPR was first determined through the use of X-ray crystallographic methods by Wang and co-workers [105] and is shown in Figure 1.5. CPR has three separate folding domains: an NADP(H)/FAD binding domain, an FMN binding domain and a connecting domain. The NADP(H)/FAD binding domain is located at the C-terminal of CPR and consists of a core, made up of anti-parallel flattened β-barrel, and an NADPH binding site, comprising of five-stranded parallel β-sheets sandwiched by α-helices [105]. The FMN binding domain, which is located at the N-terminal of the protein, consists of a five-stranded parallel β- sheet flanked by five α-helices [105]. Lying between the NADP(H)/FAD and FMN binding domains is the connecting domain as well as a highly dynamic ~15 amino acid linker [84, 105, 107]. On the basis of crystallographic data, there is strong evidence of conformational variation in the CPR system (Figure 1.6). Numerous structures are available for both closed CPR [83, 102, 105], where the FAD and FMN isoalloxazines are in close contact, and open CPR [84, 108, 109], where the distance between the FAD and FMN redox centres is noticeably increased. By combining structural data with both kinetic and computational analysis it has been established that both of these two conformational states are required for functional transfer of electrons catalysed by CPR. Structures determined for wild type rat [105] and human [102] CPR show the enzyme in the more closed form. Due to the proximity of the two flavin cofactors in the closed conformation, electrons are able to efficiently transfer from the FAD to the FMN cofactor [53-55]. However, by analysis of crystal data, it appears that electron transfer from CPR to partner proteins is hindered in this form. This follows because (i) key acidic amino acid residues on the FMN domain of CPR, which are thought to interact with basic residues on the surface of CYP proteins, are occluded in the closed form of CPR and (ii) electron transfer distances between CPR and CYP (determined through computational docking experiments ) are greater than 15 angstrom (the limit for efficient electron transfer [53-55]), and thus CPR-CYP electron transfer in this conformation of CPR is impaired [32]. In a recently published paper, Xia et al. investigated this hypothesis to see if the closed conformation of CPR could catalyse inter-protein electron transfer [83]. The authors constructed a cysteine variant of CPR which could be locked into a closed conformation. As expected, electron transfer from closed CPR to partner proteins was impaired, while interflavin 25

26 1. INTRODUCTION electron transfer rates were somewhat maintained, demonstrating conformational arrangement of CPR is required for its function. Figure 1.5. Structure of wild type rat CPR. The FAD, connecting and FMN domains are shown in dark blue, marine, and light blue ribbons, respectively. The FAD, FMN and NADP + are shown as dark blue, cyan and yellow sticks, respectively. PDB ID 1AMO. By removal of key residues in the linker region of CPR (ΔTGEE CPR), the structure of the open form of CPR has been determined [84]. The distances between the dimethyl benzene rings of FAD and FMN in the structures of open CPR range from angstrom (three structures determined for the open state of CPR). Unlike the disulfide locked closed form of CPR [83], this open variant was unable to catalyse the transfer of electrons from FAD to FMN, but it could catalyse interprotein electron transfer from CPR to CYP proteins. Structures of open conformers of CPR have also been observed using a number of different strategies. These data include the open structure of a yeast-human chimeric CPR protein, which has a 84 angstrom edge-to-edge distance between the two flavin cofactors [108], and the structure of a ΔTGEE variant of CPR in complex with haem oxygenase (Figure 1.6 C), which shows a 30 angstrom distance between the FAD and FMN cofactors and a 6 angstrom distance between the FMN of CPR and the haem of haem oxygenase [109]. Taken together, these data suggest the closed form of CPR is required for inter-flavin electron transfer while the open form of the enzyme is needed for CPR-partner protein electron transfer (Figure 1.6). 26

27 1. INTRODUCTION Figure 1.6. Structures of open and closed conformations of rat CPR. (A) Wild-type (WT) structure of CPR determined by X-ray crystallography (PDB ID 1AMO). (B) Crystallographic structure of the ΔTGEE variant of CPR (PDB ID 3ES9). (C) Crystallographic structure of ΔTGEE variant of CPR bound to haem oxygenase (PDB ID 3WKT). The FAD domain, the connecting domain and the FMN domain are shown as dark blue, marine blue and light blue ribbons, respectively. Haem oxygenase is represented as a red ribbon. The FAD, FMN, AMP and haem cofactors are shown as dark blue, cyan, yellow and red, respectively. Unlike other members of the diflavin oxidoreductase family, CPR has an N-terminal membrane anchor [105, 106]. Over recent years, many groups have used membrane nanodiscs (artificial membranes) to study the roles of both the membrane anchor and different membrane environments (lipid composition) on CPR function [ ]. It has been observed that binding of CPR to a membrane increases the affinity of the FMN cofactor to CPR [112], alters the redox potentials of the flavin cofactors [113], increases the rate of collision (and thus rate of catalysis) [114] and appropriately orientates both CPR and partner proteins enabling electron transfer [84, 113]. Overall, the importance of the N-terminal membrane anchor can be seen through the inability of the solubilised form of CPR (catalytic portion of protein commonly used in many in vitro studies) to efficiently transfer electrons to CYP proteins [105, 116] Reaction mechanism of CPR. Scheme 1.1 shows the reductive half reaction of CPR and related diflavin oxidoreductases. Due to the characteristic UV-Vis and fluorescence spectral features of the flavin chromophores bound to CPR, many rapid-reaction kinetic methods (stopped-flow [51, 82, 117], temperature-jump [118, 119] and laser flash photolysis spectroscopy [120, 121]) can be used to probe CPR reaction chemistry. Out of these techniques, the most utilised has been stopped-flow spectroscopy [82]. By rapidly mixing oxidised CPR with saturating concentrations of NADPH in a stopped-flow instrument, one can track pseudo-first-order kinetics of CPR flavin reduction by following the decay of the oxidised flavin feature, at 454 nm, or the growth and decay of the CPR flavin 27

28 1. INTRODUCTION semiquinone species [82]. Reaction transients recorded from these studies fit to double exponential decay functions which broadly correspond to the rate constants for two- and fourelectron reduction of CPR (k 1 and k 2 in Scheme 1.1). Mechanistically, it is proposed that NADPH binds and transfers a hydride to the N5 position of the CPR bound FAD cofactor. Electrons are subsequently transferred from the FAD hydroquinone to FMN producing a quasi-equilibrum (QE) state a thermodynamic equilibrium of three FAD and FMN redox states (oxidised FAD and hydroquinone FMN; semiquinone FAD and semiquinone FMN; hydroquinone FAD and oxidised FMN). After the initial hydride transfer event, NADP + dissociates and a second NADPH coenzyme binds, which reduces CPR to a FAD hydroquinone and an FMN hydroquinone state [82]. If no electron accepting partner protein is present in solution, an additional phase is observed. This slow phase (slower than enzyme k cat ) is likely to be connected with further oxidation of NADPH, attributed to thermodynamic relaxation through disproportionation reactions and/or conformational change [82]. Scheme 1.1. Reductive half-reaction of CPR (see text for more details). The direct observation of interflavin electron transfer (IET) from FAD to FMN in CPR is compromised in stopped-flow type experiments by the slow rate of electron delivery (the initial hydride transfer event limits the succeeding electron transfer event in CPR) along with the complex multi-exponential nature of CPR reduction [82]. Therefore, novel techniques such as laser photoexcitation of caged electron donors and relaxation kinetic methods have been used to investigate IET in CPR. There are a number of studies on CPR using an array of different photoactivated caged molecules (such as 5-deazariboflavin [120] and thiouredopyrene-3,6,8- trisulfonate [121]). When these molecules are decaged using laser excitation, electrons can be rapidly injected into CPR enabling the rate of FAD FMN electron transfer to be probed. Rapid 28

29 1. INTRODUCTION relaxation kinetic methods, such as temperature-jump (T-jump), on the other hand, can probe interflavin electron transfer in CPR by perturbing the QE form of CPR as a result of rapid heating [118, 119]. This sudden temperature change causes electrons to transfer from the FAD cofactor to the FMN, shifting the three states found in the QE to the oxidised FAD and hydroquinone FMN form of CPR. This process can be monitored using UV-Vis spectroscopic methods and enables direct observation of interflavin electron transfer. All of these aforementioned methods have shown the rate of electron transfer ( s -1 ) is slow relative to the rates proposed by using the Dutton ruler to measure the distance between the FAD and FMN isoalloxazine moieties in the crystal structure of CPR (rates of s -1 implied based on 4 Å distance) [53-55, 105]. Therefore, it was implied that electron transfer catalysed by CPR is gated. Stopped-flow kinetic isotope effect (KIE) studies have shown that electron transfer in CPR is not gated by chemical events (H-transfer) [117]. Thus, it is proposed protein domain dynamics are linked to electron transfer chemistry catalysed by CPR. The oxidative half reaction catalysed by CPR involves the transfer of electrons from the reduced FMN of CPR to one of the CPR partner proteins (see Section and Figure 1.4 for more details). Due to the inability of soluble CPR (N-terminal membrane anchor cleaved CPR) to react with CYP proteins in vitro, there is a lack of detailed kinetic studies describing the mechanism of the CPR oxidative half-reaction [75, 116, 122, 123]. To overcome these limitations, cyt c, which accepts electrons from soluble CPR, is often used as a surrogate electron acceptor to study this reaction [83, 84, 124, 125]. Stopped-flow studies probing the kinetics of electron transfer from CPR to cyt c show the oxidative-half reaction of CPR occurs by a two-step model. The faster of these two phases represents electron transfer from the hydroquinone form of CPR to cyt c, while the slower phase corresponds to electron transfer from the FMN semiquinone form [124]. Despite the challenge of measuring the oxidative half-reaction of CPR with a more physiologically relevant partner protein, Roberts and coworkers have been able to probe such chemistry by using liposome solubilised full length CPR and CYP3A4 [126]. By rapidly mixing full length reduced CPR and ferric CYP3A4 in a stopped-flow instrument, Roberts and coworkers have measured rate constants for the oxidative half-reaction of CPR [126]. Kinetics recorded for electron transfer between CPR and CYP3A4 are remarkably similar to those observed between CPR and cyt c (biphasic transients recorded with similar rate constants for the reduction of electron accepting partner). However, it must be noted that for further understanding of the oxidative half-reaction of CPR more detailed kinetic and spectroscopic studies are required (possibly through the use of nanodisc bilayers artificial membrane environments [ ]). 29

30 1. INTRODUCTION Reaction mechanism of mammalian cytochrome P450s. Numerous structural and mechanistic studies have been performed on cytochrome P450 (CYP) monooxygenases as they have potent physiological roles [127], are often targets for drugs [128] and have potential use in industrial biocatalysis [ ]. CYPs can be categorised into ten separate classes based on their electron donating partner proteins [133]. The class II CYPs (apart from P450 BM3) are all anchored to the microsomal membrane and receive electrons from CPR (sometimes cytochrome b 5 [134]) to their Cys-ligated b haem, which acts as the catalytic core of the enzyme [97, 98, 133]. Scheme 1.2. Catalytic cycle of CYPs (see text for more details). The catalytic cycle of CYPs consists of 8 steps and is shown in Scheme 1.2 [97, 98]. The first step involves substrate (RH) binding to the CYP, causing displacement of water as the 6 th ligand to the CYP haem, shifting the spin state of haem from low to high as well as potentially influencing the redox potential of the haem cofactor [135, 136]. Following substrate binding, electrons are transferred from the CPR partner protein to the CYP, reducing the haem from ferric (Fe 3+ ) to the ferrous state (Fe 2+ ) [135, 137]. Oxygen subsequently binds to the ferrous haem, forming the ferrous O 2 complex (step 3), which can display a resonance structure, where electrons are withdrawn from the haem to the dioxygen producing a ferric superoxide complex [138]. This ferric superoxide complex is further reduced by CPR, producing the ferric-peroxo haem form (step 4), which is pronated (step 5) first to form the ferric-hydroperoxo species (Compound 0) 30

31 1. INTRODUCTION [139] and subsequently (step 6) to produce the ferryl oxo [Fe(IV)=O] porphyrin radical cation species (Compound I) [ ]. In step 7, compound I abstracts a hydrogen from the substrate (RH) producing both a substrate radical and the bound equivalent of a hydroxyl radical (Compound II) [138, 143]. Radical recombination causes the substrate radical and compound II to react (step 7) producing the hydroxylated substrate product (ROH) and finally the cycle is completed by binding of water, which releases the product formed (step 8) [96-98] Dynamic properties of CPR. Theories that domain dynamics gate intra- and inter-cpr electron transfer have been inferred from both transient state kinetic studies (see section 1.5.3) and structural studies (see section 1.5.2). To directly observe such phenomena, the dynamic energy landscape of CPR has been probed through numerous spectroscopic techniques [51, 107, ]. Altogether, these studies have shown CPR exists as a mixture of functionally relevant open and closed conformations, with long and short distances between the isoalloxazine moieties of the two flavin cofactors, respectively. Changes in the distribution of open and closed CPR states (changes in conformational landscape) have implications on catalysis and can be driven by a number of factors including: NADP(H) binding/release, redox state of flavin cofactors, ionic strength of the buffering solutions and hydrostatic pressure. In this subsection the dynamical properties of CPR will be reviewed, explaining how domain motion is related to reaction chemistry. Several of the spectroscopic methods used to the probe the conformational landscape of CPR show the enzyme in a discreet state either more open or more closed. These spectroscopic methods have contributed to the understanding of how solvent perturbations and coenzyme binding influences the conformational landscape of CPR. However, to gain more insight into the broader conformational landscape of CPR, the application of pulsed electron-electron double resonance (PELDOR) electron paramagnetic resonance (EPR) spectroscopy has been vital [51]. PELDOR can be applied to the two-electron reduced (di-semiquinone) state of CPR to measure the distance distribution through the dipole dipole coupling of the two electronic magnetic moments of the FAD and FMN semiquinones. This method has shown CPR does not exist in solution as one discreet state but samples a continuum of conformational states with flavin-flavin distances ranging from angstrom [51]. It is now well established that binding of the NADP(H) coenzyme to CPR shifts the protein from a more open state to a more closed conformation. This has been observed directly by nuclear magnetic resonance [107], electron paramagnetic resonance [51], fluorescence [50, 148], mass spectrometry [144], reflection anisotropy spectroscopy (RAS) [147] and small angle X-ray 31

32 1. INTRODUCTION scattering (SAXS) studies [145], as well as indirectly by T-jump methods (adding NADPH causes a 5-fold increase in rate of interflavin electron transfer, indicating shorter distances between flavinflavin cofactors [118]). Mechanistically speaking, the shortening of interflavin distances by NADPH binding is favourable for interflavin electron transfer events which occur after initial coenzyme binding and hydride transfer events enabling the enzyme to efficiently transfer electrons from NADPH, through FAD, to the FMN cofactor. The relationship between CPR flavin redox chemistry and domain dynamics is an area of intense research and debate. Many spectroscopic techniques have been used to investigate if redox chemistry drives CPR conformational change, including: small angle X-ray scattering [145] and electrospray ionisation (ESI)-ion mobility-mass spectrometry [144]. Despite the limitation of SAXS studies on flavoenzymes (X-ray beams used for these studies generate in situ photoelectrons [ ] that could reduce flavin cofactors), both these studies show CPR occupies a more open conformation in the oxidised form and a more closed conformation in the fully reduced (4-electron) reduced form. The idea that redox chemistry drives functionally relevant conformational change in CPR is interesting. However, to fully understand the role of electron transfer chemistry in driving conformational change of CPR, high-energy transient dynamic states, which appear during catalysis, must be probed. Pudney et al recently found evidence of the relationship between CPR electron transfer chemistry and transient domain dynamics [50]. By specifically labelling CPR with donor and acceptor fluorophores, the authors were able to link pre-steady state reaction kinetics with short-lived dynamic states of CPR through the use of stopped-flow UV-Vis and Forster resonance energy transfer (FRET) kinetics techniques [50]. This combinatory approach of UV-Vis and FRET stopped-flow, to track reaction chemistry and domain dynamics, is the focus of this thesis and has been used to revisit and expand on the work that Pudney et al conducted on CPR as well as to probe the dynamics and reaction chemistry of the structurally related mammalian neuronal nitric oxide synthase (nnos) which is reviewed below. 1.6 Nitric oxide synthase (NOS) Physiological roles of NOS. The mammalian isoforms of nitric oxide synthase (NOS) produce the signalling molecule nitric oxide (NO) from L-Arginine, molecular oxygen and NADPH [152, 153]. NOS-generated NO is essential for many key biochemical processes and aberrant NO production is linked to the pathophysiology of many disease states [154]. Three known isoenzymes of NOS exist neuronal NOS (nnos, NOS-I), inducible NOS (inos, NOS-II) and endothelial NOS (enos, NOS-III). While these three NOS isoforms catalyse the same reaction, they are distinct in a number of ways and 32

33 1. INTRODUCTION are the product of different genes, have discreet tissue-specific locations and are regulated/inhibited by different mechanisms. nnos is predominantly expressed in neuronal tissue, as well as in smooth muscle tissue innervated by peripheral nitrergic nerves (containing NO producing nnos) [155, 156]. Under normal physiological conditions inos is not expressed, however in an immune response (e.g. cytokine or bacterial lipopolysaccharide) inos transcription and translation can be stimulated in any tissue type [156]. enos is mainly expressed in endothelial tissue, but it has also been detected in a number of other locations, including cardiac muscle cells [156]. In this subsection of the introduction, the physiological and pathophysiological roles of the three-tissue specific NOS isoforms will be addressed. In the central nervous system (CNS), NO produced by nnos is thought to be involved in the development of memories as well as in learning and neurogenesis [155, 157, 158]. NO produced by nnos in the peripheral nervous system (PNS) acts as a signalling molecule that activates guanylyl cyclase an enzyme which produces cyclic guanosine monophosphate (cgmp) [156, 159]. When cgmp is generated, it acts on smooth muscle cells (e.g. blood vessels) leading to a number of physiological responses (e.g. penile erection by relaxation of the corpus cavernosum smooth muscle [160, 161]). As mentioned above, inos is only expressed in response to a stimulus. inos produces large quantities of NO and is often expressed in cells of the immune system, such as macrophages [162]. The high amounts of NO produced by inos enables macrophage to carry out their cytostatic and cytotoxic functions by inhibiting a number of key proteins [162] (NO has high affinity to iron containing proteins which it can inhibit) and destroying DNA strands [ ] (by direct interaction or through nitrosative stress) in target cells (e.g. microbes). Similar to nnos, enos also has a role in controlling blood pressure [156, 159], as well as playing a part in vasoprotection and prevention of atherosclerosis [167] (NO produced by enos prevents leukocytes adhesion to blood vessel walls an early step in the development of atherosclerosis). As well as carrying out a number of vital cellular processes, NOS is also involved in the pathophysiology of many disease states [154]. Overstimulation of the NOS isoenzymes leads to elevated productions of NO, which may be damaging to the cell [ ]. Such events tend to occur when there are massive influxes of Ca 2+ into host cells, for example following N-methyl-Daspartate (NMDA) receptor mediated neuronal death after stroke [168, 169]. High concentrations of NO can be directly damaging to the cell (as mentioned above inhibits iron containing enzymes and damages DNA) or react with radical oxygen species (ROS) leading to peroxynitrite production [166], which in turn can cause oxidative damage to a number of biological macromolecules such as proteins, lipids and DNA [154]. 33

34 1. INTRODUCTION Structure of NOS. Despite the lack of an atomistic structure of full length NOS, the X-ray crystal structures of the individual NOS domains [ ] along with recently published cryo-em data [ ], have helped to illustrate the structural organisation of this enzyme (Figure 1.7). NOS is a homodimer, with each monomer subunit containing a reductase and an oxygenase domain. The NOS reductase domain is structurally highly similar to cytochrome P450 reductase [105, 170, 174] (see Section 1.5.2) and contains an C-terminal FAD/NADP(H) binding domain, an FMN binding domain and a connecting domain. The oxygenase domain of NOS is the catalytic core of the enzyme and contains a Cys-ligated b haem and a tetrahydrobiopterin cofactor (H 4 B) [ , 176]. Separating the NOS reductase and oxygenase domain is a calmodulin (CaM) binding site [175, ]. In the absence of Ca 2+ and CaM the FAD and FMN domains of the constitutive NOS isoforms (nnos and enos) are in close proximity, locked in the so called input state (FAD and FMN cofactors juxtaposed enabling efficient interflavin electron transfer) [33, 68] by two unique regulatory sequences: an auto-inhibitory insert (AI) [ ] and a C-terminal tail (CT) [170, ]. CaM binding is essential for NOS catalysis and is thought to free this locked input state, enabling electron transfer from FMN to haem in the enzyme s output state (FMN and oxygenase domains juxtaposed enabling electron transfer from FMN haem) [184, 185, 187, 188]. CaM binds to the three distinct isoforms of NOS with different affinities. The interaction between inos and CaM is virtually irreversible and occur at even very low Ca 2+ concentrations (K d <nm), while binding between the constitutive NOS enzymes (nnos and enos) is reversible and requires higher intracellular Ca 2+ concentrations (K d in nm range) [189, 190]. The mechanism of NOS electron transfer has been extensively studied. NOS catalyses the transfer of electrons from NADPH, through FAD and FMN cofactors, to the catalytic haem centre where NO is produced [33, 68]. On the basis of the structure of the isolated reductase domain, which shows the FAD and FMN cofactors in proximity, electron transfer from the FMN domain to the oxygenase domain is not viable due to the occluded nature of the FMN cofactor [170]. Therefore, it is proposed that large-scale dynamics of the FMN-binding domain ( angstrom), causing shifts between the NOS input and output states, are required for NOS to catalyse electron transfer reactions [33, 68]. These dynamics have been seen by many spectroscopic techniques (see Section 1.6.4). Recently, these different conformations have been detected using cryoelectron microscopy (cryo EM) techniques, which have been used to observe low resolution structural and dynamic data on the three NOS isoforms [ ]. 34

35 1. INTRODUCTION Cryo-EM performed on the NOS isoforms show that the enzyme exists in a dynamic equilibrium of conformational states, and ligand (NADPH, CaM) binding can functionally remodel the equilibrium distribution [ ]. These dynamic states include extended forms of NOS (NOS input state), where the FAD and FMN domains are within close proximity, and compact forms of NOS, where the FMN domain is adjacent to the oxygenase domain (NOS output state) [33, 68]. These structural studies, taken together with the results obtained from additional kinetic and spectroscopic approaches [33, 68], show that large-scale dynamics are required for control of the electron transfer chemistry catalysed by the NOS enzymes. In the section below, the mechanism of NOS electron transfer and the dynamic properties of the enzyme system are discussed in more detail Reaction mechanism of NOS. Stopped-flow kinetic studies of NOS flavin reduction have been performed on both the isolated reductase domain [186, 191, 192] and full length NOS enzymes [184, 193, 194]. These studies show that NOS FAD and FMN reduction by NADPH is complex and consists of multiple overlapping kinetic phases. The mechanism proposed for the reductive half-reaction of the NOS enzyme is shown in Scheme 1.3 and is very similar to that of CPR and other related diflavin oxidoreductases [82, ]. The fastest of the kinetic phases (k 1 in Scheme 1.3) detected using stopped-flow experiments predominantly represents formation of a NADPH-FAD charge-transfer (CT) species after NADPH binding. Following bindings steps, NOS is reduced by NADPH to a 2-electron reduced state (k 2 in Scheme 1.3). Once two-electron reduced, NADP + is released from the enzyme and an additional NADPH coenzyme binds driving the enzyme to a 4-electron reduced state (k 3 in Scheme 1.3). An additional fourth phase is observed after flavin reduction. This additional phase is slower than the k cat of NOS and is thought to represent protein conformational change [82, 191]. There is much interest and debate over the role which the NOS binding partner CaM has during flavin reduction. Many authors suggest that CaM stimulates the rate of flavin reduction with an increase in the observed flavin reduction rate constants [184, 186, 192, 194]. However, in the comprehensive kinetic studies performed by Scrutton and coworkers on the isolated NOS reductase, no such effect is observed [191, 198]. Instead, the authors show that CaM binding appears to alter the amplitude changes associated with each of the four kinetic phases attributed to flavin reduction. These finding are in line with theories suggesting that CaM binding influences the NOS flavin mid-point potentials, a result which has previously been observed by redox potentiometric studies [199], and could explain why CaM binding increases NOS enzymatic steady state turnover rates (cytochrome c reduction) [186, 191, 192]. 35

36 1. INTRODUCTION Figure 1.7. Structure and molecular architecture of mammalian NOS. (A) The structure of the neuronal NOS (nnos) reductase domain (PDB ID 1TLL). The FAD and connecting domains are shown in dark blue, the FMN domain is shown in light blue. The C-terminal tail (CT) and the autoinhibitory (AI) loop are shown in orange and yellow ribbons, respectively. The FAD, FMN and AMP cofactors are shown as dark blue, cyan and yellow sticks respectively. (B) The structure of Ca 2+ -bound CaM bound to a nnos peptide (PDB ID 2O60). CaM is shown in grey, the Ca 2+ ions are shown as yellow spheres and the nnos peptide is shown as a dark blue ribbon. (C) The nnos oxygenase domain dimer (PDB ID 1ZVL). H 4 B, heme and L-Arg are shown as green, red and black sticks, respectively. (D) Domain structure of neuronal, inducible and endothetial NOS enzymes. The FAD and connecting domains are shown in dark blue, the FMN domain is shown in light blue, the haem domain is shown in red and the CaM molecule is shown in grey. The C-terminal tail (CT) is shown in orange and the auto inhibitory (AI) loop is shown in yellow. (E) Simplified molecular architecture of the NOS enzymes. Each NOS monomer contains a reductase portion (FAD and FMN binding domains), an oxygenase domain and a CaM binding site. The colour coding for (E) is the same as (D). 36

37 1. INTRODUCTION Once the NOS FMN cofactor is reduced, electrons transfer from FMN to the haem containing oxygenase domain. In NOS, the FMN haem electron transfer is thought to be cross-monomer [200, 201] and requires the presence of CaM [202]. Due to the complexity associated with deconvoluting multiphasic stopped-flow transients the kinetics of FMN haem electron transfer in NOS have been probed by laser flash photolysis methods [ ]. Interestingly, these methods have shown that the full length forms of NOS transfer electrons from FMN to haem with rates 10-fold slower than those seen in the NOSoxyFMN construct. Discrepancies between the rates observed between the two NOS constructs imply that shuttling between NOS input and output states (dynamic interconversion) is required for electron transfer between the two flavin cofactors and electron transfer from the FMN and the haem. It is believed that this shift between input and output states is rate limiting in NOS catalysis [33, 68]. Scheme 1.3. Reductive half-reactions of NOS (see text for more details). The oxygenase domain of NOS is the catalytic core of the enzyme [ ]. Through the sequential transfer of two electrons from FMN to haem, L-Arginine is oxidised at the NOS haem to generate L-Citrulline and the signalling molecule NO (Scheme 1.4A) [152, 153]. Scheme 1.4B shows the 11 steps required for NOS-catalysed NO production [208]. To initiate NO production, electrons are transferred from the reduced FMN to the ferric haem, producing the ferrous haem state (step 1). Once generated, ferrous haem binds a molecule of dioxygen, giving the ferric haem-superoxy species (step 2) [209], which is subsequently reduced by the electron transfer from the H 4 B cofactor (step 3) [210]. The latter of these steps produces the haem-peroxo species (a compound which has only been seen in NOS at cryogenic temperatures [211]), which is further protonated to form a ferryl haem iron-oxo species (step 4). The NOS haem iron-oxo species reacts directly with L-Arginine to produce the N ω -hydroxy-l-arginine (NHA) intermediate and water, as well as regenerating the NOS ferric haem state (step 5). Like in step 1, the ferric haem is reduced 37

38 1. INTRODUCTION Scheme 1.4. Sequential oxidation of L-Arginine catalysed by NOS. (A) NOS catalysed oxidation of L-Arginine to produce nitric oxide (NO) and L-Citrulline. (B) Catalytic cycle of NOS enzymes (see text for more details). to the ferrous state by electron transfer from the reduced FMN (step 6) [208]. Following reduction, steps 2 through 4 are repeated again leading to the formation of the formation of the 38

39 1. INTRODUCTION haem iron-oxo species (step 7-9) [208]. This species reacts with the N ω -hydroxy-l-arginine (NHA) intermediate leading to the production of ferric haem-no complex and L-Citrulline (step 10). Finally NO is released from the ferric haem-no complex completing the catalytic cycle of the enzyme (step 11) [212, 213] Dynamic properties of NOS. Many spectroscopic and structural determination techniques have shown that NOS enzymes exist in a dynamic equilibrium of conformational states, and ligand (NADPH, CaM) binding can functionally remodel the equilibrium distribution [170, , 198, ]. It is hypothesised that many of the known NOS conformational substates are relevant for catalysing electron transfer from NADPH, through FAD and FMN cofactors, to the catalytic haem centre [33, 68]. Figure 1.8 shows a depiction of the potential conformational equilibrium of the NOS enzymes during catalysis. It is believed that in the NOS resting state (Figure 1.8) the domains sample multiple different conformations, with little to no functional relevance. Upon binding of the reduced coenzyme, NADPH, the FAD and FMN domains of NOS lock together (Figure 1.8). Locking of the FAD and FMN in the NADPH bound state is crucial for interflavin electron and occurs due to interactions between the NADPH coenzyme and the several important amino acids on the auto-inhibitory insert (AI) and the C-terminal tail (CT) [170, ]. Ca 2+ bound CaM interactions with NOS free the enzyme s locked input state, enabling the FMN domain to shuttle between the FAD and the haem domains [184, 185]. CaM has a crucial role in NOS catalysis and enables the enzyme to sample both input and output states, allowing the enzyme to catalyse electron transfer steps [33, 68]. Direct observation of NOS domain dynamics has been seen using cryo-em (see Section 1.6.2) [ ], fluorescence (which is reviewed in Section 5) [199, ], EPR spectroscopy [198, 214, 215] and mass spectrometry (used to identify the NOS oxygenase domain interacts with both the reductase domain and CaM) [216]. EPR measurements performed on NOS have been particularly useful in probing the conformational landscape of NOS. EPR has been used to determine the distances between the semiquinone paramagnetic FMN cofactor and ferric haem cofactors (~ 18.8 angstrom)[214], as well as revealing the distribution of dynamics states of the NOS enzyme [198, 215]. Two key EPR experiments have been performed on NOS showing that the enzyme samples a number of conformational substates [198, 215]. The first of these experiments was performed by Sobolewska-Stawiarz et al by measuring the distance between the FAD and the FMN semiquinone states [198]. Depending on the conditions used (NADP + and/or CaM bound/free NOS), the authors showed that flavin-flavin distance distributions of Å could be 39

40 1. INTRODUCTION detected concurrently in the same sample. This work was subsequently followed by a paper published the same year by Feng and coworkers [215]. By engineering a cysteine knock-in variant of CaM, the authors were able to attach a nitroxide spin label to CaM. This spin-labelled CaM was used to measure the distance between the NOS ferric haem cofactor and the bound CaM, showing that CaM binds to the NOS enzymes in two predominant configurations. 85 % of the CaM-NOS complexes are in an open configuration with large separations between the haem and the CaM. In contrast, the other 15 % of CaM-NOS complexes were shown to be compact, with short distances between the CaM and the haem containing oxygenase domain. Figure 1.8. NOS conformational equilibrium during catalysis. The FAD and connecting domains are shown as dark blue, the FMN domain is shown as light blue, the haem domain is shown as red and CaM is shown as grey. See text for details. Despite the extensive number of studies on NOS dynamics, no studies have been able to detect NOS conformational changes that occur during enzyme turnover. Identifying if dynamic events occur during turnover is crucial for understanding how electron transfer is catalysed by the NOS system. As mentioned in Section Pudney et al have recently developed a stopped-flow FRET method of studying dynamics of the related diflavin oxidoreductase CPR [50]. By attaching fluorophores to CPR the authors were able to track conformational change in CPR, which occur on the same timescale as chemical steps (flavin reduction). In this thesis, NOS domain dynamics will be probed using similar stopped-flow methods in the attempt to identify catalytically relevant NOS conformational changes. 40

41 1. INTRODUCTION Calmodulin Calmodulin is a small (16 kda) Ca 2+ binding proteins that is ubiquitous in eukaryotes [224]. High yields of CaM can be attained from simple expression and purification protocols, thus the properties and function of CaM have been studied extensively [ ]. CaM contains two globular domains separated by a flexible linker region (Figure 1.9). Each of the CaM globular domains has two EF-hands, with each of them binding one Ca 2+ ion. CaM undergoes large conformational changes upon binding Ca 2+ and/or partner proteins (Figure 1.9). The surface of apo-cam is covered in negatively charged amino acids, preventing it from interacting with a number of CaM partner proteins. When intracellular Ca 2+ concentrations are increased, CaM binds to Ca 2+ causing the protein to change conformation, forming an open dumbbell shape, and revealing numerous hydrophobic residues. This open form of CaM binds to a number of target proteins, which are involved in vital biochemical processes such as cell growth, proliferation and apoptosis [232]. One protein which binds CaM is NOS [202]. The conformation of CaM bound to a NOS peptide is shown in Figure 1.9C. Once bound, CaM enables efficient electron transfer between the NOS FMN and the haem cofactors [68, 152, 153]. Therefore, CaM interactions with NOS are essential for NOS-catalysed NO production. Figure 1.9. Structure and conformational states of mammalian calmodulin (CaM). (A) Structure of wild-type CaM (PDB ID 1CLL). (B) Structure of CaM bound to Ca 2+ (PDB ID 1CFC). (C) Structure of CaM bound to both divalent calcium ions and a nnos peptide (PDB ID 2O60). CaM is shown as a grey ribbon, divalent calcium ions are shown as yellow spheres and the nnos peptide is shown as a blue ribbon. 41

42 1. INTRODUCTION 1.7 Fluorescence spectroscopy The phenomenon of fluorescence. Fluorescence is the emission of light from an atom or molecule which has absorbed a photon [233]. Fluorescence typically occurs in organic molecules that are aromatic and is often described through the use of a Jablonski diagram, which is shown in Figure In Figure 1.10 the horizontal lines represent the energy levels (electronic configurations) of an atom or molecule, while the vertical and horizontal arrows represent the transitions between different energy states [233]. In a thermally equilibrated system, all atoms or molecules are present in the singlet ground state (S 0 ). Upon absorbance of a photon of appropriate energy, a valence electron from the atom or molecule is promoted to one of the singlet excited state (excited state electron paired in an antiparallel configuration to a second electron in the ground state) vibrational levels (e.g. S 1 or S 2 ). This is the process of light absorbance and it characteristically occurs on a femtosecond timescale [233]. Figure One version of a Jablonski diagram for a hypothetical molecule. Horizontal lines are representative of energy states, while the vertical and horizontal arrows represent the transitions between the energy states. S 0 is the ground singlet state, S 1 is the first singlet excited state, S 2 is the second singlet excited state and T 1 is the triplet excited state. The red arrows represent the absorbance of light from the singlet ground state to a singlet excited state. The blue dotted arrows represent internal conversion (IC). Fluorescence (purple) or non-radiative emission (black) occur from the lowest singlet excited state, while phosphorescence (orange) occurs from the lowest triplet excited state state. Excited state molecules can undergo spin conversion (green) in a process called intersystem crossing (ISC). 42

43 1. INTRODUCTION Following light absorbance, a number of different processes can occur depending on the properties of the atom or molecule excited [233]. Through the means of internal conversion (IC) and vibrational relaxation, which occur on picosecond timescale, the excited molecule usually ends up in the lowest vibrational level (S 1 ). From the S 1 state the excited state molecule can either: i) return to the ground state releasing a photon (characteristically at a longer wavelength to the photon absorbed) causing fluorescence (which occurs on a nanosecond timescale), ii) nonradiative decay to the ground state or iii) undergo spin conversion in a process called inter-system crossing (ISC), forming a triplet excited state. Emission from the triplet state (T 1 ) is termed phosphorescence and occurs on timescale much slower than those seen for fluorescence. This follows because transition from the T 1 state to the singlet ground state is spin forbidden, slowing down the phosphorescence process to the millisecond to second timescale [233]. As the fluorescent properties of a fluorophore are highly sensitive to the environment, the phenomena of fluorescence are often used in many technologies. In biochemistry, fluorescence has been beneficial in developing DNA sequencing methods, chemical sensing, detecting properties of ligand-protein interactions, and in studying protein dynamics [233]. For the study of protein dynamics, the fluorescence technique Förster resonance energy transfer (FRET) has been particularly valuable. The principles of FRET and how it can be applied to study protein dynamics are addressed below The phenomenon of FRET. Förster resonance energy transfer (FRET) can occur when the emission spectrum of a donor fluorophore overlaps with the excitation (i.e. absorbance) spectrum of an acceptor molecule (this acceptor does not need to have fluorescence properties) [233]. The FRET phenomena is a result of a donor molecule in the excited state transferring energy to an acceptor chromophore, in close proximity ( angstrom in distance), via non-radiative dipole-dipole coupling [233]. The efficiency of the FRET process is related to the distance between the donor and acceptor fluorophores (as well as the amount of donor emission and acceptor absorbance overlap) [233]. Equation 1.1 shows the relationship between the efficiency of energy transfer (E) and the distance (r) between the donor and acceptor fluorophores, where R 0 is the Förster distance (a description of the spectral overlap) [233]. From Equation 1.1, it can be seen that any small change in the donor-acceptor distance alters the efficiency of energy transfer (Figure 1.11). Owing to this relationship, FRET is often referred to as the spectroscopic ruler [234] and has proved effective in the detection of structural and dynamic properties of many biological macromolecules. 43

44 1. INTRODUCTION E = R 0 6 R r6 Equation FRET and protein dynamics. Proteins contain many intrinsic fluorophores (e.g. cofactors, aromatic amino acids) and can also be decorated with a number of external fluorophores (e.g. thiol-reactive fluorophores). Therefore, fluorescence spectroscopy is often very useful in the study of protein molecules. FRET methods (ensemble and single-molecule FRET) have been invaluable in a number of in vivo and in vitro studies to probe protein dynamics [233]. Among other protein systems, FRET has been used to study the dynamic properties of calmodulin (CaM) [225] and ATP synthase [235]. In these studies external donor and acceptor fluorophores were attached to the proteins in specific locations. By adding ligands to these fluorophore labelled protein molecules, changes in the FRET efficiency were observed. These changes in FRET efficiencies were related to distance alterations between the donor and acceptor fluorophores bound to the proteins and could be interpreted as a change in the protein conformational landscape. Figure Relationship between FRET efficiency and donor-acceptor distance separation for an example dye-pair with a R 0 value of 50 angstrom. Stopped-flow based FRET methods have been used in a number of studies to map protein conformational change [50, ]. Stopped-flow spectroscopy is a method typically used to detect transient chemical states that occur during a chemical/enzymatic reaction by following 44

45 1. INTRODUCTION changes in the absorbance (or fluorescence, infra-red) of a sample. However, stopped-flow spectroscopy can also be used to monitor protein dynamics by FRET. The stopped-flow set-up required to monitor FRET is shown in Figure FRET stopped-flow has proved to be particularly useful in probing the conformational change of enzymes. Upon mixing an enzyme with a substrate molecule, stopped-flow FRET can be used to directly visualise transient conformational substates during catalytic turnover by following time-resolved emission changes of the intrinsic or external donor and acceptor protein bound fluorophores. Mapping these conformational changes onto chemical steps during turnover is useful to determine the relationship between enzyme catalysis and dynamics. This approach is used in this thesis to study the relationship between dynamics and reaction chemistry of two diflavin oxidoreductase (CPR, nnos). Figure Set up used for stopped-flow FRET experiments. To perform a stopped-flow FRET measurement the two drive syringes are loaded with the reactants of interest (e.g. fluorophore labelled enzyme and a coenzyme molecule). When the measurement is initiated the two reactants are rapidly mixed together in the sample cell. Light used to excite the fluorophores is continuously passed into the sample cell at a selected wavelength. Upon excitation of the fluorophores, any fluorescence emission from the reactants in the sample cell is transmitted towards the photomultiplier tubes (PMT). To measure fluorescence emission at cetain wavelengths filters are normally placed in front of the PMTs. These can be band-width pass (BWP), long pass or short pass filters, which allow the transmission of light between a certain wavelength, above a certain wavelength or below a certain wavelength, respectively. Using stopped-flow FRET any timedependent changes in fluorescence can be detected after the mixing of the two reactants (dead-time of instrument ~5 ms). 45

46 1. INTRODUCTION 1.8 Aims of project. Numerous structural determination and spectroscopic techniques have shown that large-scale protein domain dynamics gate the transfer of electrons in diflavin oxidoreductases (Section 1.3 and 1.4). However, few studies have focused on detecting transient dynamic species which appear during enzyme turnover. Understanding the properties of these short-lived dynamic substates is not only important for insights into the role of dynamics during diflavin oxidoreductase turnover but may also contribute towards the knowledge required to rationally design synthetic proteins and in the drug discovery process. Here, to study protein domain dynamics, site-directed fluorophore labelling methods are used to modify two mammalian diflavin oxidoreductases - nnos and CPR - with extrinsic Förster resonance energy transfer (FRET) donor and acceptor molecules. Any changes in the emission of the donor and acceptor fluorphores report on local protein dynamics arising from biological events, such as ligand/inhibitor binding and protein-protein interaction. By rapidly mixing fluorophore decorated diflavin oxidoreductase with redox partners (e.g. NADPH or partner protein) in a stopped-flow instrument, time-resolved ( real-time ) dynamics can be probed. The observed rate constants and amplitude changes of the transient dynamic events observed can be compared to reaction kinetics measured by UV-Vis stopped-flow, enabling the relationship between domain dynamics and redox chemistry to be probed. In this thesis, initial studies are focused on the relationship between domain dynamics and electron transfer chemistry in CPR, which is sometimes described as the simplest diflavin oxidoreductase. Previously, studies have shown that the reductive half-reaction of CPR is kinetically coupled to protein domain dynamics [50]. Here, to futher these findings, the reaction chemistry and the domain dynamics associated with the oxidative half-reaction of CPR are studied (Section 2). This work is followed by efforts to understand the relationship between dynamics and electron transfer in NOS - a diflavin oxidoreductase that is structurally more complex than CPR (Section 3 and 4). 46

47 2. REDOX-LINKED DOMAIN DYNAMICS AND CPR 2. Real-time analysis of conformational control in electron transfer reactions of human cytochrome P450 reductase with cytochrome c. Published in: The FEBS Journal First Published: 16 th September 2015 Authors: Tobias M. Hedison, Sam Hay, Nigel S. Scrutton DOI: /febs Running Header: Redox-linked domain dynamics and CPR. Contributions: TMH wrote the paper with help from SH and NSS. TMH conducted all the experiments. Analysis and interpretation of the data was conducted by all authors. 47

48 2. REDOX-LINKED DOMAIN DYNAMICS AND CPR 2.1 Abstract. Protein domain dynamics and electron transfer chemistry are often associated, but real-time analysis of domain motion in enzyme-catalysed reactions and the elucidation of mechanistic schemes that relate these motions to the reaction chemistry are major challenges for biological catalysis research. Previously we suggested that reduction of human cytochrome P450 reductase with the reducing coenzyme NADPH is accompanied by major structural re-orientation of the FMN- and FAD-binding domains through an inferred dynamic cycle of open and closed conformations of the enzyme (PLoS Biol, 2011, e ). However, these studies were restricted to stopped-flow/fret analysis of the reductive half-reaction, and were compromised by fluorescence quenching of the acceptor by the flavin cofactors. Here we have improved the design of the FRET system, by using dye pairs with near-ir fluorescence, and extended studies on human cytochrome P450 reductase to the oxidative half-reaction using a double-mixing stoppedflow assay, thereby analysing in real-time conformational dynamics throughout the complete catalytic cycle. We correlate redox changes accompanying the reaction chemistry with protein dynamic changes observed by FRET, and show that redox chemistry drives a major re-orientation of the protein domains in both the reductive and oxidative half-reactions. Our studies using the tractable (soluble) surrogate electron acceptor cytochrome c provide a framework for analysing mechanisms of electron transfer in the endoplasmic reticulum between cytochrome P450 reductase and cognate P450 enzymes. More generally, our work emphasizes the importance of protein dynamics in intra- and inter-protein electron transfer, and establishes methodology for real-time analysis of structural changes throughout the catalytic cycle of complex redox proteins. KEYWORDS: cytochrome P450 reductase, diflavin oxidoreductase, electron transfer, Förster resonance energy transfer, protein domain dynamics. 2.2 Introduction. Protein dynamics may be described in a similar way to protein folding, using a multi-dimensional conformational landscape [240]. Conformational landscapes encompass hill and valley features: the valley topographies represent the low-energy states, while the hill landscapes correspond to the thermodynamic barriers between them. Use of spectroscopic and crystallographic measurements has shown that energy landscapes may be altered by pressure, temperature, mutagenesis and substrate/inhibitor binding, demonstrating that the study of conformational change is essential to understanding of the functionality of some proteins [12, 28, 35]. However, it is difficult to study the conformational landscape of a protein as these landscapes encompass a broad range of time scales (10 12 to > 1 s) and distance scales (10 2 to > 10 Å). The role of large- 48

49 2. REDOX-LINKED DOMAIN DYNAMICS AND CPR scale domain dynamics that gate biological electron transfer (ET) has been demonstrated in a number of enzyme families [53, 241]. Such domain dynamics have been inferred, or directly shown, to be coupled to ET in a number of related diflavin oxidoreductases, including nitric oxide synthase [177, 179, 198], methionine synthase reductase [242], cytochrome P450 BM3 [71], sulfite reductase [243] and cytochrome P450 reductase (CPR) [34, 50]. Real-time analysis of conformational change during catalysis in such complex redox proteins is a major challenge due to the rich optical properties of multi-centre redox proteins and the inherent complexity of the enzyme catalytic cycle. CPR is a microsomal membrane-anchored oxidoreductase that transfers electrons from NADPH to a variety of haem-containing partner proteins, including cytochrome P450s (CYPs) [75], haem oxygenase [76], squalene monooxygenase [78] and cytochrome c (cyt c) [90]. CPR contains two distinct redox domains, one housing an FAD cofactor and the other housing an FMN cofactor, which are separated by a flexible hinge region [102, 105, 107]. The chemical mechanism of the reaction catalysed by CPR is complex (Scheme 2.1), but is well documented and involves binding of NADPH to the FAD domain, followed by hydride transfer from the nicotinamide to the N5 of the FAD [82]. Following FAD hydroquinone formation, the FMN is reduced by intramolecular ET, and once one- or two-electron-reduced, may act as a one-electron donor for microsomally bound haem proteins such as CYPs [125]. The oxidized FAD may be further reduced by a second NADPH [82], and five oxidation states of CPR may be catalytically relevant: the fully oxidized and one-, two-, three- and four-electron-reduced forms [117]. Cytochrome c may also act as a surrogate electron acceptor, and, like CYPs, accepts electrons from the reduced FMN cofactor [90]. Crystallographic data show that the two flavin cofactors of CPR are positioned with a 4 Å edge-toedge distance between the isoalloxazines [102, 105]. On the basis of the Dutton's ruler, this short distance should allow very rapid inter-flavin ET (approximately s 1 ) [54, 55, 244]. However, temperature-jump perturbation spectroscopy [119] and laser-flash photolysis experiments [120, 121] have shown that ET between FAD and FMN co-factors is relatively slow (approximately 10 s 1 ), suggesting that the inter-flavin ET is adiabatic or gated. Furthermore, phdependent and kinetic isotope studies have shown that the reductive half-reaction catalysed by CPR is not gated by chemical steps (proton transfer) [117], which suggests that conformational changes/protein dynamics are likely to control the rate of inter-flavin ET. Multiple spectroscopic techniques, including mass spectrometry [144], NMR [145], small-angle X- ray scattering [145], neutron scattering [146] and reflection anisotropy [147], as well as construction of an open yeast/human chimeric protein [108], have been used to demonstrate 49

50 2. REDOX-LINKED DOMAIN DYNAMICS AND CPR that CPR exists as a mixture of open and closed conformations, with long and short flavin flavin distances, respectively. Moreover, pulsed electron electron double resonance (PELDOR) spectroscopy measurements of two-electron-reduced (di-semiquinone) forms of several diflavin oxidoreductases has shown that their conformational landscapes are rugged, with multiple conformers present [51, 198, 242]. Many of these aforementioned studies may be interpreted as indicating that perturbation through coenzyme addition, redox state alterations and changes in ionic strength lead to shifts in the conformational landscapes of the enzyme. Further, both the deletion of a hinge region (ΔTGEE) [84] and chemical cross-linking of thiols [83] in variant forms of CPR have shown that inter-flavin ET is inhibited in the open form of the enzyme, while a closed cross-linked form is unable to efficiently reduce cyt c. These data, coupled with studies of ET as a function of solvent viscosity, temperature and pressure [51], have highlighted the potential functional importance of redox-driven conformational exchange between closed and open substates of CPR for inter-flavin ET and subsequent reduction of CYPs. However, these studies do not provide a direct read-out of real-time conformational change concurrent with catalysis. Scheme 2.1. The reductive half-reaction of CPR. NADPH binds to oxidized CPR, and a hydride anion is transferred from the reducing nicotinamide coenzyme NADPH to FAD. After FAD hydroquinone formation, electrons are equilibrated across the FAD and FMN cofactors to form a quasi-equilibrium state ( QE ). With excess NADPH, further reduction occurs, leading to formation of an enzyme species that contains FAD and FMN hydroquinone forms. We recently reported an analysis of conformational change in CPR in the reductive half-reaction (reduction of the flavins by NADPH) using stopped-flow Förster resonance energy transfer (FRET) studies [50]. The study was not extended to the complete reaction cycle, as this was not possible using a single-mixing stopped-flow method. Our previous study indicated coupling of dynamics to chemical catalysis during flavin reduction. However, the quenching of fluorescence emission prevented detailed analysis of the conformational transitions, such as direct spatial mapping of transitions from open to closed forms, and vice versa. Here we address the limitations of the previous work. Specifically, we have developed a doublemixing stopped-flow method that enables real-time analysis of conformational change throughout the complete catalytic cycle (reductive and oxidative half-reactions). The limitations 50

51 2. REDOX-LINKED DOMAIN DYNAMICS AND CPR imposed by fluorescence quenching are largely overcome by use of longer-wavelength FRET pairs. Use of these new FRET pairs enabled analysis of the oxidative half-reaction (and thus also the complete reaction cycle) with the soluble electron acceptor cyt c, which is widely used as an electron acceptor in turnover studies of diflavin oxidoreductases [124, 125], including CPR [82, 86, 245]. This methodology allowed the UV-Vis kinetics of CPR redox chemistry to be followed concurrently with a time-based analysis (approximately 5 ms to 10 s) and spatial analysis (20 80 Å) of CPR domain dynamics. 2.3 Results. The rationale behind the present study was to establish a simple FRET model that reports directly on protein domain dynamics (i.e. opening and closing of CPR) and is not compromised by fluorescence changes attributed to quenching by the redox cofactors. Having established a suitable model, we set out to develop a double-mixing stopped-flow assay that can access the kinetics of ET from CPR to the soluble redox acceptor cyt c, which, together with previous studies of the reductive half-reaction [50], provides a complete analysis of ET for the whole catalytic cycle. The new FRET model and double-mixing stopped-flow assays were then combined to investigate the real-time dynamic changes throughout the CPR/cyt c catalytic cycle FRET model of CPR domain dynamics. We took advantage of naturally occurring cysteine residues in human CPR (C228, C472 and C566) to attach the fluorophores for FRET analysis. Using mass spectrometry, we showed previously that these residues are labelled readily [50]. Although there are two accessible cysteine residues in the FAD-binding domain (C472 and C566), modification of one prevents labelling at the other due to their proximity. Both residues are similarly separated from C228, and, as discussed previously, this simplifies the FRET analysis [50]. The location of these cysteine residues dictates that the distance between fluorophore dyes is changed as CPR adopts more open or closed conformations (Figure 2.1). In Figure 2.2A, the absorbance spectra of CPR and cyt c in various states of reduction are shown. As the spectral features of the haem and flavin cofactors bound to CPR and cyt c span the UV and most of the visible spectrum, these two chromophores may cause significant quenching of dye fluorescence emission. Therefore, the fluorescence labels Cy5 (donor, D) and Alexa750 (acceptor, A), which emit in the far-red/near-ir region, and are thus unlikely to be significantly quenched by haem or flavin, were selected to monitor domain dynamics in this investigation (Figure 2.2B). 51

52 2. REDOX-LINKED DOMAIN DYNAMICS AND CPR Figure 2.1. Open and closed conformations of rat CPR. The open structure was generated from PDB ID 3ES9_A and the closed structure was generated from PDB ID PDB 1AMO_A. The C228 C556 and C228 C472 sulfur sulfur distances are 46.9 and 51.5 Å, respectively, in the open form, and 51.2 and 57.3 Å, respectively, in the closed form, i.e. differences of 4.3 and 5.8 Å for the C228 C556 and C228 C472 sulfur sulfur distances, respectively. The distances between C472 and C556 on the FAD-binding domain are 14.4 and 14.7 Å for 1AMO_A and 3ES9_A, respectively. Both of these distances are too short to be observed by traditional FRET measurements, and may be ignored in this investigation. Binding of the Cy5/Alexa750 pair to CPR (two dyes per protein, referred to as CPR-DA) was shown to have an efficiency of 19 ± 3% (eight experiments), based on the extinction coefficients of dyes and protein (Figure 2.2B). Moreover, from the absorption and emission spectra, the pair was calculated as having a Förster radius (R 0 ) of approximately 68 Å (Figure 2.2C,D), which is slightly larger than the inter-cysteine (S S) distances calculated from the open and closed conformers of rat CPR (94% sequence homology to human CPR), which are 51.2 and 57.3 Å in the closed form and 46.9 and 51.5 Å in the open form (Figure 2.1) [84, 105]. However, the distances between the dyes in the various conformers of CPR are likely to be greater than inter-cysteine distances due to the size of the linker between the cysteine and the fluorescent moiety present on the maleimide labels (approximately 10 Å). Thus, as the dye-dye distance is likely to be similar to R 0, the Cy5/Alexa750 dye pair is expected to be sensitive to conformational changes associated with domain dynamics in CPR. 52

53 2. REDOX-LINKED DOMAIN DYNAMICS AND CPR Figure 2.2. Absorption and fluorescence spectra of CPR-DA. (A) Absorption spectra for oxidized CPR (black), CPR reduced with one equivalent of NADPH (red) or 20 equivalents of NADPH (blue), oxidized cyt c (magenta) or sodium dithionite-reduced cyt c (green). (B) Absorption spectra for Cy5 and Alex750 bound to CPR in an approximately 1:1 ratio (black), and for CPR singly labelled with Cy5 (red dashed line) or Alex750 (red dotted line). (C) UV-Vis spectrum for Alexa750 (black) and emission spectrum for Cy5 excited at 560 nm (red). (D) Calculated spectral overlap for the spectra shown in (C). The spectral overlap gives a Förster radius, R 0, of 68 Å, calculated as R 0 = [κ 2 Φ 0 n 4 J] 1/6, where κ 2 = 2/3, Φ 0 = 0.28, n = 1.33 and J = m 1 cm 1 nm 4. (E) Fluorescence emission spectra for CPR-D (black), CPR-DA (red), an equal mix of CPR-D and CPR-A (blue), and unlabelled CPR (magenta) when excited at 650 nm. These spectra are normalized to 670 nm, and the emission of the doubly labelled sample is scaled for CPR concentration. (F) Fluorescence emission spectra for CPR-DA (black) and CPR-DA mixed with a 20-fold excess of reduced cyt c (red) or oxidized cyt c (blue) (20 μm), when excited at 650 nm. Experimental conditions were typically 1 μm labelled CPR in 50 mm potassium phosphate buffer (ph 7.0) at 25 C. 53

54 2. REDOX-LINKED DOMAIN DYNAMICS AND CPR Figure 2.2E shows the fluorescence emission spectra of CPR-D (donor-labelled) and CPR-DA when excited at 650 nm (donor excitation maximum). A significant change in emission at 670 and 780 nm was observed for CPR-DA, corresponding to a decrease in donor emission and an increase in acceptor emission compared with the CPR-D spectrum. These changes in fluorescence emission are indicative of FRET from donor to acceptor fluorophores. Furthermore, when an equimolar mixture of singly labelled CPR-D and CPR-A (acceptor-labelled) is excited at 650 nm, negligible FRET is observed. This lack of significant inter-molecular FRET from CPR-D to CPR-A suggests that, as expected, there is no significant dimerization of oxidized CPR. Also, when non-labelled CPR was excited at 650 nm, minimal intrinsic flavin fluorescence was observed. Figure 2.2F shows the emission spectra of CPR-DA in the presence of 20 μm oxidized cyt c or 20 μm reduced cyt c. The lack of change in dye emission indicates there is no quenching of dye fluorescence by the haem present in cyt c. Combined, the data indicate that labelling of CPR using the long-wavelength dye pair Cy5/Alexa750 gives rise to a FRET signal that is probably responsive to distance changes, and is not affected by cofactor quenching of the signal or inter-molecular FRET Double-mixing stopped-flow assays of ET from NADPH-reduced CPR to cyt c Double-mixing assay for cyt c reduction. Before monitoring redox-linked domain dynamics of CPR, the oxidative kinetics of the enzyme were studied by pre-steady state UV-Vis stopped-flow spectroscopy. A double-mixing regime was developed to follow the reduction of cyt c by NADPH-reduced CPR. The double-mixing stoppedflow method allowed rapid mixing of CPR and NADPH in the first mixing event. This leads to reduction of CPR; with one stoichiometric equivalent of NADPH, the end point is a mixture of reduced CPR forms determined by the potentials of the flavin and nicotinamide coenzyme couples (Scheme 2.1). After the first mix, the reduced forms of CPR were then incubated in the ageing loop of the stopped-flow spectrometer for various ageing times before finally being mixed with oxidized cyt c. Photodiode array spectra were initially recorded to follow the reaction of cyt c with NADPHreduced CPR. The photodiode array data, shown in Figure 2.3A, indicate that the oxidation of CPR may be followed by monitoring the rapid increase in absorbance of the haem α-band at 550 nm. Transients monitoring the reduction of cyt c at 550 nm were found to fit to a double exponential function (Figure 2.3C). Furthermore, under the stated conditions, the 600 nm feature (Figure 2.3A) of the flavin neutral di-semiquinone has a minimal spectral contribution (Figure 2.3B). 54

55 2. REDOX-LINKED DOMAIN DYNAMICS AND CPR Figure 2.3. Cytochrome c reduction by CPR (i). (A) Stopped-flow photodiode array spectra showing cyt c reduction by 1 NADPH-reduced CPR. (B) Absorbance difference spectra, relative to the spectrum at 5 ms, derived from the data in (A). (C) Single wavelength transient at 550 nm derived from the data in (A) fitted to a double-exponential function. Experimental conditions were 2 μm CPR mixed with 2 μm NADPH and aged for 2 s prior to a second mix with 20 μm cyt c. All samples were suspended in 50 mm potassium phosphate buffer (ph 7.0) at 25 C Optimizing the ageing time between stoichiometric reduction of CPR and subsequent reduction of cyt c. Optimal double-mixing conditions were established by varying the ageing time after mixing of equimolar amounts of oxidized CPR and NADPH. Figure 2.4A shows an example transient of the first mix in the stopped-flow experiment: stoichiometric reduction of CPR by NADPH (denoted 1 NADPH-reduced CPR), followed by monitoring the neutral semiquinone species at 600 nm. The 55

56 2. REDOX-LINKED DOMAIN DYNAMICS AND CPR trace was shown to fit to a double-exponential function over the 4 s time scale recorded, with rate constants for the first and second phases of 30.0 ± 1.7 (64%) and 4.6 ± 0.5 s 1 (36%), respectively. While these apparent rate constants are likely to be a convolution of multiple steps, the faster step probably reflects the initial ET from FAD hydroquinone to FMN, with the slower phase reflecting establishment of the quasi-equilibrium species shown in Scheme 2.1. After the first mix, the two-electron-reduced CPR present in the ageing loop was subsequently mixed with ferric cyt c at various ageing times, and cyt c reduction was monitored at 550 nm (Figure 2.4B). These data were also fitted to a double-exponential function, and Figure 2.4C,D show how the observed rate constants and amplitude changes exhibit a hyperbolic dependence on the ageing time ( 1 s), saturating on the same timescale as the QE state is established. At earlier ageing times, before the two-electron-reduced CPR has equilibrated, the cyt c reduction kinetics are both faster and have higher yield, but the reaction kinetics are very sensitive to ageing time (the mechanistic reasons for which remain uncertain). As the observed rate constants and amplitude changes associated with the kinetic phases of CPR oxidation are invariant at ageing times > 1 s (Figure 2.4C,D), and the largest population of the CPR di-semiquinone species is formed during this time (Figure 2.4A), a 2 s ageing time was selected for subsequent experiments. As the concentration of reduced cyt c is readily determined from the fitted change in absorbance (amplitude) at 550 nm, the stoichiometry of the reaction was determined. At short ageing times, approximately one molecule of cyt c is reduced per CPR, with this efficiency decreasing at longer ageing times. This reaction stoichiometry suggests that the one-electron-reduced CPR does not reduce cyt c at this experimental time scale (4 s) Double-mixing studies with a fixed ageing time and variable NADPH concentration. Figure 2.5 illustrates the kinetics of cyt c reduction by NADPH-reduced CPR that has been incubated in the ageing loop for 2 s. When equimolar NADPH and CPR are mixed together (Figure 2.5A,B), two second-order kinetic phases are observed, with rate constants of 1.54 ± 0.04 and 0.33 ± 0.01 μm 1 s 1. Alternatively, for experiments with 2:1 and 4:1 ratios of NADPH to CPR, two kinetic phases are observed, with the observed rate constants both saturating at higher cyt c concentrations (Figure 2.5C,E). When CPR is reduced using a 20-fold excess of NADPH (Figure 2.5G), the observed rate constants from two kinetic phases are both largely independent of cyt c concentration; while the second phase does show a small decrease in observed rate constant, this probably arises from the poor fitting of the data due to complications resulting from multiple turnover of CPR (data not shown). 56

57 2. REDOX-LINKED DOMAIN DYNAMICS AND CPR Figure 2.4. Cytochrome c reduction by CPR (ii). (A) Stopped-flow transient monitoring flavin reduction in CPR by equimolar NADPH (4 μm). (B) Double-mixing stopped-flow transients showing the dependence of cyt c reduction by 1 NADPH-reduced CPR on ageing time. The colour of each trace represents the ageing time as indicated by coloured arrows in (A). Transients monitoring cyt c reduction by 1 NADPH-reduced CPR were measured at 600 nm, and fitted to a double-exponential function. (C, D) The dependence of the observed rate constants and amplitudes (A i, Equation 2.2) on ageing time is shown in (C) and (D), respectively. For the amplitude data, the first kinetic phase is shown in black, the second in red, and the sum in blue. Experimental conditions are equivalent to those in Figure 2.3, with ageing times between 0.03 and 10 s. The reaction stoichiometry was again determined by taking the sum of the fitted amplitudes of each kinetic phase (Figure 2.5B,D,F,H), and, in all cases, cyt c was reduced in sub-stoichiometric amounts relative to the added NADPH. These data suggest that some oxidation state(s) of CPR are unable to reduce cyt c, and/or electron flux through CPR (from NADPH to FAD to FMN) is slow relative to the experimental time scale (approximately 10 s). The pseudo steady-state flux of electrons through CPR to cyt c (V app ) was determined by calculating the fractional amplitude (A i )- weighted sum of the rate constant (k i ) for each stopped-flow experiment as shown: n V app = i=1 A i k i Equation

58 2. REDOX-LINKED DOMAIN DYNAMICS AND CPR Figure 2.5. Cytochrome c reduction by CPR (iii). The cyt c concentration dependence of the rate constants (A, C, E, G) and amplitudes (B, D, F, H) describing the reduction of cyt c by CPR reduced using equimolar NADPH (A, B), or twofold (C, D), fourfold (E, F) and 20-fold (G, H) equivalents of NADPH in a double-mixing stoppedflow experiment with an ageing time of 2 s. For the amplitude data, the first kinetic phase is shown in black, the second in red, and the sum in blue. Experimental conditions are equivalent to those in Figure 2.3, and data fitting is described in Table 2.1. When the NADPH concentration is low, the electron flux is dependent on cyt c concentration, whereas, at higher levels of NADPH, multiple turnovers of CPR are possible, and the electron flux 58

59 2. REDOX-LINKED DOMAIN DYNAMICS AND CPR becomes cyt c-independent and approaches steady-state levels: the electron flux with 20:1 NADPH:CPR is 4.9 ± 0.1 s 1, whereas the k cat is 10.4 ± 0.4 s 1 (Figure 2.6). As the electron flux is not slow during these stopped-flow experiments, it appears that the sub-stoichiometric reduction of cyt c probably arises from accumulation of one or more oxidation states of CPR over the experimental time scale that are unable to reduce cyt c. Kinetic phase Parameters NADPH equivalents k for ( s -1 ) ± ± 5.1 k rev ( s -1 ) - 0 a 2.6 ± 2.7 1st K S (µm) ± ± 8.1 k for / K S (µm -1 s -1 ) ± ± 0.5 k (µm -1 s -1 ) 1.54 ± k for ( s -1 ) ± ± 0.6 k rev ( s -1 ) - 0 a 1.2 ± 0.4 2nd K S (µm) ± ± 22.8 k for / K S (µm -1 s -1 ) ± ± 0.07 k (µm -1 s -1 ) 0.33 ± Table 2.1. Oxidative half-reaction kinetics. Fitting parameters for the data in Figure 2.5. The stoichiometric data demonstrated second-order kinetics and were fitted to a linear function, while other data were fitted to Equation 2.3. a, Not significantly larger than zero so value fixed to 0. 59

60 2. REDOX-LINKED DOMAIN DYNAMICS AND CPR Figure 2.6. (Quasi) steady-state kinetics. (A) Electron transfer flux during cyt c haem reduction by CPR reduced with equimolar NADPH (black) or twofold (red), fourfold (blue) and 20-fold (magenta) equivalents of NADPH, determined by fitting the data in Figure 2.5 to Equation 2.1. Data are fitted to a polynomial function (solid lines) to guide the eye. (B) Steady-state kinetics of 2 nm wild-type CPR (black) and donoracceptor labelled CPR (red) with 100 μm NADPH and excess cyt c, measured by absorption at 550 nm. Experiments were performed in 50 mm potassium phosphate buffer (ph 7.0) at 25 C. The data in (B) are fitted to a Michaelis Menten equation with k cat = 10.4 ± 0.4 s 1 and K m = 10.2 ± 1.7 μm (wild-type) and 10.3 ± 0.8 s 1 and 8.8 ± 2.5 μm (CPR-DA) Measurements of CPR domain dynamics by stopped-flow fluorescence spectroscopy. The conformational landscape associated with the reductive half-reaction of CPR has previously been studied using fluorescence stopped-flow spectroscopy with fluorophore-labelled enzyme (Alexa488/Cy5) [50]. These experiments were performed using excess NADPH, and demonstrated that domain dynamics are kinetically linked to reaction chemistry. Initially, a similar experiment was performed here using CPR labelled with Cy5 and Alexa750 (CPR-DA). The CPR-DA exhibits similar UV-Vis reaction kinetics to the unlabelled enzyme, with steady-state k cat and K m values of 10.3 ± 0.8 s 1 and 8.8 ± 2.5 μm (the values for unlabelled enzyme are 10.4 ± 0.4 s 1 and 10.2 ± 1.7 μm; Figure 2.6B), and observed reductive half-reaction rate constants of 18.7 ± 0.7 and 4.3 ± 0.2 s 1 for the first and second NADPH reduction of CPR, respectively (the values for unlabelled protein are = 14.9 ± 0.1 and 4.0 ± 0.1 s 1 ). Figure 2.7A shows stopped-flow transients monitoring the fluorescence emission (650 nm excitation of the donor) from the singly labelled enzymes CPR- D and CPR-A and the doubly labelled CPR-DA enzyme upon reduction using a 40-fold excess of NADPH. CPR-D exhibits two distinct kinetic phases: the first shows fluorescence quenching, with a rate constant of 15.8 ± 0.7 s 1, while the second shows de-quenching, with a rate constant of 4.3 ± 0.3 s 1. As there was no acceptor fluorophore present in this sample, it is likely that the quenching arises from conformational rearrangement of the enzyme and/or quenching by the flavin neutral (di)semiquinone species. The donor fluorescent transient for CPR-DA displays a similar emission 60

61 2. REDOX-LINKED DOMAIN DYNAMICS AND CPR Figure 2.7. Stopped-flow fluorescence of CPR-DA versus 40-fold excess NADPH. (A) Labelled CPR stoppedflow fluorescence transients. Fluorescence donor transients are indicated in black for singly labelled protein (CPR-D) and red for doubly labelled protein (CPR-DA), whereas acceptor transients are indicated in blue for singly labelled protein (CPR-A) and green for doubly labelled protein (CPR-DA). (B) Acceptor and corrected donor emission. (C) CPR donor:acceptor ratio (representing domain dynamics) plotted against an example UV-Vis transient, measured at 600 nm (CPR reaction kinetics). (D) Schematic of domain dynamics, showing the predominant conformational sub-state during the reductive half-reaction. The FAD/NADP + -binding domain, the FMN-binding domain and the connecting domain are represented by red, green and blue rounded rectangles, respectively. Experimental conditions were 50 mm potassium phosphate buffer (ph 7.0), 1 μm labelled/15 μm unlabelled CPR, at 25 C. Donor and acceptor fluorophores were excited at 650 and 750 nm, respectively, and fluorescence data were smoothed using a five-point moving average. behaviour and kinetics to the CPR-D (Table 2.2), but there is a reduction in the magnitude of dye quenching. This difference in fluorescence quenching between the doubly and singly labelled samples may possibly be attributed to FRET between the donor and acceptor fluorophore. CPR-A (Figure 2.7A) shows no changes in fluorescence emission when excited at 750 nm (a significant improvement over our previous study in which Cy5 was used as the acceptor [50]), while the acceptor transient recorded for CPR-DA upon 650 nm excitation exhibits two distinct kinetic 61

62 2. REDOX-LINKED DOMAIN DYNAMICS AND CPR phases that may also be attributed to FRET between the donor and acceptor fluorophore (Table 2.2). To interpret the FRET changes associated with protein dynamics only, the CPR-DA donor emission was corrected by subtracting the CPR-D emission [50]. The deconvoluted FRET changes associated with the donor and acceptor fluorophores were fitted to a double-exponential function (Figure 2.7B). Moreover, the relative donor:acceptor (D:A) emission (FRET response) was calculated (Figure 2.7C), and the data were fitted to a double-exponential function with rate constants of 26.0 ± 0.8 and 3.1 ± 0.1 s 1. Based on crystal structures of open and closed forms of rat CPR [84, 105], we expect the D A distance to reduce by approximately 5 Å when CPR opens (Figure 2.1). The two kinetic phases in Figure 2.7B,C may thus be attributed to closing and opening of CPR on the same time scale as the reaction kinetics, as illustrated in the schematic shown in Figure 2.7D. Figure 2.8 shows the FRET changes associated with the reduction of CPR by stoichiometric NADPH. As also seen in studies using a 40-fold excess of NADPH, the singly labelled CPR-D shows a quenching of fluorescence over the time scale recorded (Figure 2.8A). This same pattern of donor quenching is seen in the CPR-DA sample, but the magnitude of quenching is reduced. The relatively small difference in emission between the CPR-D and CPR-DA samples may be attributed to FRET between the donor and acceptor dyes. The deconvoluted donor transient (Figure 2.8B) cannot be adequately fitted by Equation 2.2 due to the poor signal-to-noise ratio. However, as the CPR-A sample (Figure 2.8A) shows no change in fluorescence emission over the course of the experiment, emission from the CPR-DA acceptor is likely to arise from FRET from the donor, and shows fluorescence quenching with two kinetic phases (Table 2.2). The D:A emission ratio shows a FRET change that may be attributed to closing of CPR, with a rate constant of 47.9 ± 1.1 s 1. These rate constants are similar to those obtained in UV-Vis studies of the reductive half-reaction of CPR when mixed with equimolar NADPH (Table 2.2). Examples of fluorescence double-mixing stopped-flow transients monitoring the reduction of cyt c by equimolar NADPH-reduced CPR are shown in Figure 2.9A. The deconvoluted donor and acceptor fluorescence spectra for the two cyt c concentrations measured illustrates how acceptor and donor emission mirror one another over the 4 s measured. The reduction of ferric cyt c by equimolar NADPH-reduced CPR-DA (Figure 2.9D) is similar to that for the unlabelled sample, with second-order rate constants of 1.01 ± 0.01 and 0.34 ± 0.02 μm 1 s 1 for the first and second kinetic phases, respectively. These cyt c-dependent rate constants illustrate that the oxidative kinetics of CPR are unaffected by the bulky fluorophores bound to it. The relative D:A emission, shown in 62

63 2. REDOX-LINKED DOMAIN DYNAMICS AND CPR Figure 2.9C, demonstrates that the two domains slowly open after the reduction of cyt c by NADPH-reduced CPR has occurred (Figure 2.9E). Figure 2.8. Stopped-flow fluorescence of CPR-DA versus stoichiometric NADPH. A) Labelled CPR stopped-flow fluorescence transients. Fluorescence donor transients are indicated in black for singly labelled protein (CPR- D) and red for doubly labelled protein (CPR-DA), whereas acceptor transients are indicated in blue for singly labelled protein (CPR-A) and green for doubly labelled protein (CPR-DA). (B) Acceptor and corrected donor emission. (C) CPR donor:acceptor ratio (representing domain dynamics) plotted against an example UV-Vis transient, measured at 600 nm (CPR reaction kinetics). (D) Schematic of domain dynamics showing the predominant conformational sub-state during the reductive half-reaction. The FAD/NADP + -binding domain, the FMN-binding domain and the connecting domain are represented by red, green and blue rounded rectangles, respectively. Experimental conditions were 50 mm potassium phosphate buffer (ph 7.0), 1 μm labelled/15 μm unlabelled CPR, at 25 C. Donor and acceptor fluorophores were excited at 650 and 750 nm, respectively, and fluorescence data were smoothed using a five-point moving average. 63

64 2. REDOX-LINKED DOMAIN DYNAMICS AND CPR Figure 2.9. Double-mixing stopped-flow fluorescence of 1 NADPH-reduced CPR versus cyt c. (A) Labelled CPR stopped-flow fluorescence transients. Fluorescence donor transients are indicated in black for singly labelled protein (CPR-D) and red for doubly labelled protein (CPR-DA), whereas acceptor transients are indicated in blue for singly labelled protein (CPR-A) and green for doubly labelled protein (CPR-DA). (B) Acceptor and corrected donor emission. (C) CPR donor:acceptor ratio (representing domain dynamics) for both reactions with 20 μm cyt c (black) and 40 μm cyt c (red). In (C), data are vertically offset to show the individual traces. (D) UV-Vis kinetics of 1 NADPH-reduced CPR-DA with cyt c. The black and red datasets represent the first and second kinetic phases, respectively. Both datasets are fitted to linear trends with second-order rate constants of 1.01 ± 0.01 and 0.34 ± 0.02 μm 1 s 1 for the first- and second-order rate constants, respectively. (E) Schematic of domain dynamics showing the predominant conformational substate during the oxidative half-reaction of CPR. The FAD/NADP + -binding domain, the FMN-binding domain and the connecting domain are represented by red, green and blue rounded rectangles, respectively. The redox partner protein cytochrome c is shown as a purple oval. Experimental conditions were 50 mm potassium phosphate buffer (ph 7.0), 1 μm labelled CPR (4 μm labelled CPR for UV-Vis measurements), at 25 C. Donor and acceptor fluorophores were excited at 650 and 750 nm, respectively, and fluorescence data were smoothed using a five-point moving average. 64

65 2. REDOX-LINKED DOMAIN DYNAMICS AND CPR k 1obs (s -1 ) k 2obs (s -1 ) CPR vs. 40-fold NADPH CPR-D 15.8 ± ± 0.3 CPR-DA 18.0 ± ± 0.3 Corrected CPR-DA donor 11.8 ± ± 1.0 CPR-DA acceptor 36.0 ± ± 0.1 D:A Emission 26.0 ± ± 0.1 Flavin reduction 14.9 ± ± 0.1 CPR vs. stoichiometric NADPH CPR-D 34.1 ± ± 0.1 CPR-DA 27.7 ± ± 0.1 Corrected CPR-DA donor ND ND CPR-DA acceptor 47.5 ± 1.0 ND D:A Emission 47.9 ± 1.1 ND Flavin reduction 30.6 ± ± 0.2 1:1 NADPH:CPR vs. 20 um cyt c CPR-D ± ± 0.02 CPR-DA ± ± 0.02 Corrected CPR-DA donor 1.2 ± 0.1 ND CPR-DA acceptor ND ND D:A Emission 0.53 ± 0.02 ND cyt c reduction 22.7 ± ± 0.2 1:1 NADPH:CPR vs. 40 um cyt c CPR-D ± ± 0.03 CPR-DA ± ± 0.02 Corrected CPR-DA donor 1.6 ± 0.1 ND CPR-DA acceptor ND ND D:A Emission 0.43 ± 0.03 ND cyt c reduction 43.1 ± ± 0.3 Table 2.2. Comparison of FRET and UV-Vis stopped-flow kinetics. Rate constants determined by fitting fluorescence transients in Figures to Equation 2.2. Flavin and cyt c reduction were monitored by absorbance at 600 and 550 nm, respectively, under the same conditions as fluorescence measurements. ND, not detected. 65

66 2. REDOX-LINKED DOMAIN DYNAMICS AND CPR 2.4 Discussion. Homo sapiens cytochrome P450 reductase (CPR) plays an important physiological role in the transfer of electrons from NADPH to many single electron-accepting haem-containing proteins including CYPs, and is also commonly used as a model system to understand the mechanism of ET in the diflavin oxidoreductase enzyme family [31]. CPR is thought to exist in a mixture of dynamically exchanging conformational states that may be described in terms of an energy or conformational landscape [50, 51, 83, 84]. Although real-time analysis of the conformational landscape during the complete catalytic cycle of CPR (or any enzyme system) is lacking, previous studies using fluorescently labelled CPR have suggested that domain dynamics are kinetically linked to the reductive half-reaction of the enzyme [50]. In the present study, we monitored domain dynamics during the complete catalytic cycle of CPR, while also addressing technical concerns relating to the quenching of extrinsic dye fluorescence by the flavin cofactors observed previously [50]. By using a pair of fluorescent dyes (Cy5 and Alexa750) that are active in the IR region and covalently attached to wild-type CPR, we were able to monitor conformational change during both the reductive and oxidative half-reactions of CPR with NADPH and cyt c, respectively. There are no detectable changes in fluorescence emission from either Cy5 or Alexa750 in CPR-DA that suggest domain re-orientation when oxidized CPR interacts with either ferric or ferrous cyt c (Figure 2.2F). This suggests that cyt c does not bind to oxidized CPR and/or that the interactions between these two proteins upon binding do not cause significant structural rearrangement of CPR. A recent NMR study of the interaction between the FMN domain of CPR and cyt c has demonstrated that transient dynamic complexes are formed between the two proteins [124]. Taken together, these data are consistent with a model whereby ET from CPR to cyt c occurs in a diffusion-controlled manner through formation of a collisional complex that does not require domain re-organization in CPR. UV-Vis double-mixing stopped-flow experiments showed that the oxidative half-reaction kinetics of CPR with cyt c are dependent on both the oxidation state of CPR (Figure 2.5) and the (ageing) time delay between addition of NADPH and cyt c to CPR (Figure 2.4). During these experiments, an initial burst phase is followed by slower pseudo-steady-state turnover [124, 125]. We focus here on the burst phase kinetics, but note that the apparent rate constants extracted may be convoluted with contributions from multiple turnovers. When CPR is mixed with equimolar NADPH, subsequent cyt c reduction occurs in two phases, which are both strictly second-order with respect to cyt c concentration (Figure 2.5A). These kinetics are also dependent on the ageing time (Figure 2.4), suggesting that the rate of inter-protein ET is limited by the diffusion-controlled 66

67 2. REDOX-LINKED DOMAIN DYNAMICS AND CPR encounter of CPR and cyt c, and that the multiple kinetic phases and dependence on ageing time may arise from different oxidation and/or conformational states of CPR. When CPR is reduced by multiple equivalents of NADPH, the cyt c reduction kinetics are more complex, again showing two phases that now exhibit saturation behaviour (Figures 2.4E and 2.5C), consistent with the inter-protein ET becoming rate-limited by a preceding chemical step. Upon reduction by two or four NADPH equivalents, the faster saturating rate constants (k 1for, Table 2.1) are comparable to inter-flavin ET rate constants determined previously using temperature-jump [119] and laser flash photolysis methods [120, 121]. It is likely that turnover under saturating NADPH is (partly) rate-limited by inter-flavin ET in CPR. Indeed, with a 20-fold excess of NADPH, cyt c reduction by CPR becomes first-order, and the burst-phase electron flux determined using stopped-flow experiments is similar to k cat (Figure 2.6). Taken together, the oxidative half-reaction kinetics determined here strongly suggest that interprotein ET occurs via a collisional complex despite exhibiting saturation behaviour, and the pseudo-first-order (saturation) kinetics are likely to arise due to rate-limiting inter-flavin ET in CPR. It is also noteworthy that, while the CPR to cyt c ET appears fastest when CPR is reduced with only one or two electrons (Figure 2.5), a comparison of the apparent second-order rate constants (k for /K S for the saturating cases; Table 2.1) shows that the diffusion-controlled rate of inter-protein ET is unlikely to be dependent on CPR redox state. It is not obvious why two kinetic phases are observed in all experiments, but these may arise from different rates of ET from the CPR FMN semiquinone and hydroquinone species to cyt c, for example, as CPR is not completely fourelectron-reduced by saturating NADPH [117]. Use of the FRET pair Cy5 and Alexa750 improves on previous work [50] by exploiting fluorophores that absorb and emit in the near-ir region, thus minimizing spectral overlap with, and fluorescence quenching by, the CPR flavin cofactors. However, the limitations to this approach include the relatively low quantum yield of these fluorophores (0.28 for Cy5 and 0.12 for Alexa750), and the poor red sensitivity of fluorescence detectors. Nevertheless, the relative magnitude of donor and acceptor emission changes and the FRET efficiency (D:A fluorescence emission ratio) are comparable to those in our previous study using green/red fluorophores [50]. While the domain dynamics in CPR are likely to be complex, involving multiple conformers [51] that may dynamically interconvert via rotating and swinging motions [246], FRET experiments only observe elements of these dynamics, as they report on the distance between two points on the protein. In CPR, the two fluorophore labelling sites have been suggested to move further apart, thus decreasing the FRET efficiency, when the protein closes to form more compact 67

68 2. REDOX-LINKED DOMAIN DYNAMICS AND CPR conformation(s) [50]. Here, we interpret these FRET data similarly to [50] within the context of a simplified two-state model involving open and closed conformations, but do not mean to infer that multiple conformations of the enzyme are not present. Upon reduction of CPR-DA with NADPH, the enzyme appears to close (Figures 2.6 and 2.7). This closing occurs with comparable kinetics to the UV-Vis absorbance changes associated with the reduction of CPR to the di-semiquinone state (Table 2.2), suggesting that flavin reduction and conformational change are kinetically coupled. This closing event was not previously observed in FRET experiments of fluorescently labelled CPR [50], and was uncovered here due to the increased time resolution and signal-to-noise ratio that we were able to achieve using the Cy5/Alexa750 dye pair. The closing of CPR on this time scale probably enables efficient ET between FAD and FMN, and may be driven by NADPH binding rather than flavin reduction, as NADP + binding has been shown to cause CPR to form more closed conformations [50, 51]. Upon further reduction of CPR-DA by excess NADPH, the enzyme re-opens with similar kinetics to conversion of the di-semiquinone state to a further/fully reduced state (Figure 2.7). The conformation(s) of this new open state may differ from the fully oxidized state, but result in a similar mean dye dye separation and thus similar FRET efficiency. The kinetic coupling of flavin reduction and protein opening in CPR was observed previously [50], and suggests that redox change may act as a driver for conformational change in this enzyme. This opening event may have a functional consequence if it makes the FMN more accessible to partner cytochromes (assuming that ET from CPR to cyt c occurs only from the FMN), thus facilitating faster interprotein ET. To investigate CPR dynamics during the oxidative half-reaction, two-electron-reduced CPR-DA was mixed with cyt c in a double-mixing stopped-flow FRET experiment. Under these conditions, CPR is observed to open with kinetics that are slower than the cyt c reduction monitored by UV-Vis absorption, appearing to be first-order with respect to cyt c concentration, (Figure 2.9 and Table 2.2). It had previously been suggested that ET from a closed form of CPR would be slow, based on analysis of a disulfide locked variant of the enzyme [83]. However, this variant was also impaired in its ability to transfer electrons between the flavin cofactors, suggesting that it may be locked in a non-productive conformation by the disulfide cross-link [83] and/or that efficient inter-flavin ET requires conformational sampling [51]. The data in Figures 2.4 and 2.5 show that cyt c is reduced by two-electron-reduced CPR, which consequently must be in a closed conformation (Figure 2.8). Unfortunately, we were unable to measure significant differences in FRET within CPR upon mixing fully reduced CPR-DA with cyt c, as the FRET efficiencies of the fully reduced and re-oxidized CPR 68

69 2. REDOX-LINKED DOMAIN DYNAMICS AND CPR samples are similar (Figures 2.7 and 2.8). However, it is quite likely that four-electron-reduced CPR also undergoes some form of structural rearrangement during/after reduction of cyt c. Further, Figure 2.4C shows that inter-protein ET is fastest at short ageing times when CPR is twoelectron-reduced but still in an open conformation. While inter-protein ET between CPR and cyt c has not been evolutionarily optimized (they are not physiological partners), it appears that interprotein ET to cyt c may be faster when CPR is in an open conformation. There may be differences in the mechanism and/or control of inter-protein ET to physiological partner proteins such as CYPs, and it is possible that CPR may only reduce these proteins when in an open conformation, i.e. leaky inter-protein ET from closed forms of CPR is minimized under physiological conditions. The use of single- and double-mixing UV-Vis and FRET stopped-flow experiments provides new insight into the dynamics and chemistry of the full catalytic cycle of the reaction of CPR with cyt c, which is shown diagrammatically in Figures 2.7D and 2.8D for the reductive half-reaction and Figure 2.9E for the oxidative half-reaction. In this dynamic model of catalysis, NADPH binding and/or FAD reduction cause CPR to adopt a more compact closed conformation. The subsequent slower inter-flavin ET may be accompanied by further structural rearrangement, with some twoelectron-reduced species adopting more compact conformations. Further reduction of CPR by a second NADPH is accompanied by conformational changes resulting in a more open form of the enzyme, consistent with our previous study [50]. Inter-protein ET to cyt c may probably occur from both open and closed conformations of one-, two-, three- and four-electron-reduced CPR, but ET may be fastest from more open conformations, and conformational rearrangement of CPR subsequent to the inter-protein ET event may be relatively slow Concluding remarks. The model presented in this work (Figure 2.10) provides a structural framework to describe the conformational control of inter- and intra-protein ET in CPR that is likely to be applicable to all mammalian diflavin oxidoreductases, given the structural and mechanistic similarities within this enzyme family. The study emphasizes the importance of protein dynamics in intra- and interprotein electron transfer, and establishes a methodology for real-time analysis of structural changes throughout the catalytic cycle of complex redox proteins. 69

70 2. REDOX-LINKED DOMAIN DYNAMICS AND CPR Figure Schematic representation of the conformational equilibria of CPR. The NADP/FAD-binding domain is shown in red, the FMN-binding domain is shown in green, and the connecting domain is shown in blue. The NADP + -free form of CPR adopts predominantly more open conformations, with relatively larger distances between the flavin cofactors. Upon coenzyme addition (NADP + /NADPH), oxidized CPR makes a transition to more compact forms, with relatively shorter inter-flavin distances. Transfer of a hydride anion from NADPH to FAD causes a shift in the conformational sub-states to predominantly more compact CPR conformations. These compact forms favour electron transfer from FADH 2 to FMN (i.e. short electron transfer distances relative to more open conformations). When CPR is reduced with excess NADPH, an additional opening phase is observed after initial closing triggered by (stoichiometric) NADPH reduction. This opening allows the enzyme to transfer electrons rapidly from the FMN domain of CPR to partner proteins. Electron transfer between cofactors and partner proteins may occur in multiple conformational states. However, the model implies that these reactions will be more rapid in selected conformational sub-states in which donor acceptor distances are shortened. 70

71 2. REDOX-LINKED DOMAIN DYNAMICS AND CPR 2.6 Experimental Procedures Materials. All reagents were analytical grade and were purchased from Sigma-Aldrich (Gillingham, Dorset, U.K.), unless otherwise stated Recombinant protein expression and purification. Homo sapiens cytochrome P450 reductase (CPR), lacking the 60 amino acid membrane-binding N- terminus, was expressed from plasmid pet15b (Millipore, Watford, Hertfordshire, U.K.) in Escherichia coli BL21 (DE3) plys (New England BioLabs, Hitchin, Hertfordshire, U.K.), and purified by DEAE Sepharose anion exchange chromatography followed by nickel Sepharose affinity chromatography, as described previously [86, 245]. Prior to use, CPR was oxidized using a few grains of potassium ferricyanide, and immediately passed through an Econo-Pac 10DG desalting column (Bio-Rad, Hemel Hempstead, U.K.) to remove surplus oxidant. The oxidized CPR concentration was determined by absorption spectroscopy using an absorption coefficient (ε) of 22 mm 1 cm 1 at 454 nm [82] Extrinsic fluorophore labeling. Maleimide labelling sites on CPR have previously been identified by mass spectrometry as C228, C472 and C566 [50]. To label CPR, fluorescent dyes were incubated with the enzyme in 50 mm potassium phosphate buffer (ph 7.0) at room temperature for 5 h. The dyes that were used to label CPR were Cy5 mono-maleimide (Cy5, GE Healthcare, Little Chalfont, U.K.) and Alexa Fluor C5 750 maleimide (Alexa750, Life Technologies, Paisley, U.K.). For labeling with a 1:1 ratio of Cy5 to Alexa750, 10 μm of CPR was incubated with 200 μm of both fluorophores. Following labeling, the unreacted dyes were removed from the labeled protein by passing samples through a desalting column (Econo-Pac 10DG desalting column; Bio-Rad, Hemel Hempstead, U.K.) Static fluorescence studies. Fluorescence emission spectra were recorded on an Edinburgh Instruments (Livingston, West Lothian, U.K.) FLS920 fluorometer equipped with double excitation and emission monochromators, a red-sensitive cooled photo-multiplier detector, and a 450 W xenon arc lamp. Spectra were recorded using 0.5 nm excitation and 5 nm emission slit widths in 1 ml fluorescent quartz cells (Starna Scientific Ltd, Hainault, U.K.) with a 10 mm excitation path length. Fluorescence emission data were collected at 25 C in 50 mm potassium phosphate buffer (ph 7.0) with approximately 1 μm of CPR. 71

72 2. REDOX-LINKED DOMAIN DYNAMICS AND CPR Stopped-flow studies and data fitting. Kinetic studies were performed using a SX20 stopped-flow spectrometer (Applied Photophysics Ltd, Leatherhead, Surrey, U.K.) housed within an anaerobic glove box (Belle Technology, Weymouth, Dorset, U.K., < 2 ppm. O 2 ). Experiments were performed in 50 mm potassium phosphate buffer (ph 7.0) at 25 C. Buffers were de-gassed by bubbling with oxygen-free nitrogen before placing in the glove box, where they were left overnight to remove all traces of oxygen. The reductive half-reaction of CPR was monitored at 454 or 600 nm, as previously described [82]. Reactions were initiated by mixing either an equimolar amount of NADPH or a 20-fold stoichiometric equivalent of NADPH with 15 μm CPR (final concentration), unless stated otherwise. To follow the oxidative half-reaction of CPR, the haem α-band of bovine heart cyt c was monitored at 550 nm (Δε 550 = 21.1 mm 1 cm 1 ). A double-mixing stopped-flow method was used to study the oxidative half-reaction kinetics of CPR. For both the reductive and oxidative half-reactions, monitored by UV-Vis stopped-flow spectroscopy, 4 12 transients were recorded for each experiment. Individual transients were fitted to Equation 2.2, where A i is the amplitude and k i is the rate constant of the ith exponential component, and ΔA is the total amplitude change: n A = i=1 A i exp( k i t) Equation 2.2 The results of these experiments are presented as mean rate constants ±1 standard deviation. Fluorescence emission stopped-flow transients monitoring both the the donor and acceptor were recorded by dual-channel detection, using a R1104 red-sensitive photomultiplier detector (Applied Photophysics, Leatherhead, Surrey, U.K.) for the acceptor channel. To monitor donor and acceptor fluorophores, 670 nm bandpass and 750 nm highpass filters (Thor Labs, Ely, U.K.), respectively, were used. All experiments were performed using approximately 1 μm CPR (final concentration) to ensure no inner filter effects were present. Between 9 and 15 or 20 and 40 repeats were performed for the donor and acceptor fluorophores, respectively. Due to the low quantum yield of the dyes, the transients were averaged and smoothed using a five-point moving average. These fluorescent transients were fitted to Equation 2.2. To extract the fluorescence change associated with FRET alone, the percentage of emission from CPR-D was subtracted from the percentage emission of the corresponding fluorophore in a FRET pair (CPR-DA); this was performed due to quenching of the donor emission by aromatic amino acids, cofactors and the flavin semiquinone absorbance feature at 600 nm. Rate constants for fluorescence data are presented with a standard error of fit. Stopped-flow control studies indicated that NADPH (1 mm) 72

73 2. REDOX-LINKED DOMAIN DYNAMICS AND CPR did not reduce the FRET dyes (1 μm) directly over the time scale of the experiments (10 s) in absence of CPR/cyt c (data not shown). Rate constants for the UV-Vis stopped-flow kinetics were plotted against cyt c concentration, and fitted to either a linear function (one-step model) or a hyperbolic function (two-step model) (Equation 2.3): k obs = k rev + k for [cyt c] K S +[cyt c] Equation 2.3 Equation 2.3 allowed the kinetic parameters [rate of reverse reaction (k rev ), rate of forward reaction (k for ) and substrate saturation constant (K S )] to be determined for the two-step model of cyt c reduction by reduced CPR. Steady-state turnover of CPR was followed by observing the reduction of cyt c at 550 nm under saturating conditions of NADPH (200 μm). At least five traces were recorded for each cyt c concentration. The K m and k cat values for cyt c were determined by fitting data to the Michaelis Menten equation. 73

74 3. REDOX-LINKED DOMAIN DYNAMICS AND NOS 3. Correlating calmodulin landscapes with chemical catalysis in neuronal nitric oxide synthase using time-resolved FRET and a 5- deazaflavin thermodynamic trap. Published in: ACS catalysis First Published: 28 th June 2016 Authors: Tobias M. Hedison, Nicole G.H. Leferink, Sam Hay, Nigel S. Scrutton DOI: /acscatal.6b01280 Running Header: Redox-linked domain dynamics and NOS. Contributions: TMH wrote the paper with help from NSS. NGHL synthesised the pro-r and pro-s NADP 2 H and the 5-dFMN compounds. TMH conducted all the experiments except for UV-Vis stopped-flow kinetics which were conducted by both NGHL and TMH. All data recorded were analysed and interpreted by TMH with guidance from NSS. 74

75 3. REDOX-LINKED DOMAIN DYNAMICS AND NOS 3.1 Abstract. A major challenge in enzymology is the need to correlate the dynamic properties of enzymes with, and understand the impact on, their catalytic cycles. This is especially the case with large, multicenter enzymes such as the nitric oxide synthases (NOSs), where the importance of dynamics has been inferred from a variety of structural, single molecule and ensemble spectroscopic approaches, but where motions have not been correlated experimentally with mechanistic steps in the reaction cycle. Here we take such an approach. Using time resolved spectroscopy employing absorbance and Förster resonance energy transfer (FRET) and by exploiting the properties of a flavin analogue (5-deazaflavin mononucleotide (5-dFMN)) and isotopically labeled nicotinamide coenzymes, we correlate the timing of CaM structural changes when bound to neuronal nitric oxide synthase (nnos) with the nnos catalytic cycle. We show that remodeling of CaM occurs early in the electron transfer sequence (FAD reduction), and not at later points in the reaction cycle (e.g. FMN reduction). Conformational changes are tightly correlated with FAD reduction kinetics and reflect a transient opening then closure of the bound CaM molecule. We infer that displacement of the C-terminal tail on binding NADPH and subsequent FAD reduction are the likely triggers of conformational change. By combining the use of cofactor/coenzyme analogues and time resolved FRET/absorbance spectrophotometry we show how the reaction cycles of complex enzymes can be simplified enabling detailed study of the relationship between protein dynamics and reaction cycle chemistry an approach that can also be used with other complex multicenter enzymes. KEYWORDS: nitric oxide synthase, calmodulin, Förster resonance energy transfer, protein dynamics, flavoenzyme, flavin analogue Table of Content (TOC) Artwork. 75

76 3. REDOX-LINKED DOMAIN DYNAMICS AND NOS 3.2 Introduction. Underpinning the function of all enzymes is the concept of protein conformational landscapes, knowledge of which is essential also in the rational design of synthetic proteins [5, 247] and in the drug discovery process [45, 248]. Many structure determination techniques (e.g., X-ray crystallography) have produced frozen snapshots of proteins that provide mechanistic insights into function. The drawback is that these snapshots typically represent only the lowest energy state of the many substates found within the conformational landscape of a dynamic protein. It is widely perceived that the interchange between individual conformers contributes, in part, toward the ability of an enzyme to enhance the rate of catalysis [12, 22, 24, 25, 28] and gate chemical steps [49, 50, 249], as well as to impart specificity for substrates [38, 45, 46]. The interchange between different conformations occurs over several angstroms, on time scales that span picoseconds to seconds. Many biophysical techniques can be used to probe the modulation of energy landscapes: for example, following the binding of ligands/inhibitors/partner-proteins [19, 33, 38, 250, 251], or changes in temperature, ionic strength, and pressure [50, 51, 198, 252]. However, fewer studies have focused on the detection of transient conformations that appear during enzyme catalysis. Likewise, little information is available on the mechanistic trigger(s) for these conformational transitions (e.g., substrate binding/product release and/or chemical steps). Understanding the nature and impact of short-lived high-energy conformational states on enzyme function is currently a topic of major interest. A protein whose dynamic properties have been studied extensively is calmodulin (CaM) [ ]. CaM is a small, ubiquitous protein involved in the regulation of many biological processes in eukaryotes [224]. By binding Ca 2+, CaM undergoes major conformational change, shifting from a closed form to an extended dumbbell shape by the separation of two globular and structurally related calcium-binding domains. Once CaM adopts this open conformation, numerous hydrophobic residues on its surface are exposed, increasing the affinity of CaM for a variety of partner proteins [232]. One such partner is the flavohemoprotein neuronal nitric oxide synthase (nnos), for which CaM binding is essential for function [152, 202]. nnos is one of three tissue-specific isoforms of nitric oxide synthase (NOS), all of which are homodimers, and produce L-Citrulline and the signaling molecule nitric oxide (NO) from NADPH, dioxygen and L-Arginine [152]. Despite the lack of atomic level structural data for full-length nnos holo-enzyme, a structural and mechanistic model (Figure 3.1) has emerged from a wealth of spectroscopic data [68, , 191, , , 222, 223, ] and determined structures of component NOS domains [ ]. 76

77 3. REDOX-LINKED DOMAIN DYNAMICS AND NOS Figure 3.1. Structural organization and electron flow through nitric oxide synthase (NOS). The FAD domain is shown in bright orange, the FMN domain in light orange, the heme domain in red and the partner protein calmodulin (CaM) in in gray. Ligand (NADPH and CaM) binding is implicated in shifting the conformational equilibrium of nnos and thus regulating electron transfer during the catalytic cycle. Dimerization of nnos occurs at an interface between the heme domains. Each nnos monomer comprises a reductase and an oxygenase domain separated by a CaM binding region. The reductase domain is structurally similar to cytochrome P450 reductase (CPR) [105, 174] and encompasses FAD- and FMN-binding domains separated by a connecting domain and a flexible linker. Individual nnos monomers dimerize at the interface of the oxygenase domains, which contain a regulatory PDZ domain and a heme domain. The heme domains also contain tightly bound heme and tetrahydrobiopterin (H 4 B) cofactors. On binding to NOS, Ca 2+ - bound CaM undergoes a conformational transition from an elongated open structure to a compact, spherical shaped conformation [175, 258]. These structural transitions are accompanied by changes in nnos as the protein occupies energetically more favored conformations [178, 223]. Large-scale domain motion (> 10 Å) influenced by CaM binding is believed to regulate electron transfer in NOS enzymes (Figure 3.1) [33, 170, 198]. Catalysis is initiated by the binding of the reducing coenzyme NADPH to the FAD-binding domain and hydride transfer to the N5 position of the FAD isoalloxazine ring [191]. Interflavin electron transfer from the FAD cofactor to FMN then follows [117, 191]. On reduction of the FMN, electrons are transferred to the heme domain in the partner monomer of the enzyme dimer [200, 201, 256] where NO is produced. FMN to heme electron transfer requires the presence of CaM, which is essential also for NO production [187, 202, 256]. On the basis of spectroscopic and structural data, shuttling of the FMN-domain 77

78 3. REDOX-LINKED DOMAIN DYNAMICS AND NOS between the FAD and heme domains is thought to be a feature of the natural catalytic cycle (Figure 3.1) [33, 170]. The two flavin cofactors are juxtaposed when electrons are transferred from the FAD to FMN domains, allowing for efficient interflavin electron transfer in the input state (Figure 3.1) [33, 170, 174]. Given the relatively large distance between the FMN and heme cofactors in the input state, electron transfer between the reductase and oxygenase domains does not occur unless there is movement of the FMN domain toward the oxygenase domain (the so-called output state). These shifts from input to output states are regulated by CaM-binding [187] and have been studied using a variety of spectroscopic techniques, including fluorescence [ , 259], single-molecule fluorescence [223], electron paramagnetic resonance [198, 214, 215], cryoelectron microscopy [177, 179, 260] and temperature jump spectroscopy [203]. However, to the best of our knowledge, there are no reported studies that have directly visualized by time resolved spectroscopy conformational substate transitions during catalytic turnover. To address this limitation, we have labeled CaM with Alexa555 and Alexa647 fluorophores at defined locations to investigate CaM dynamics during enzyme turnover. CaM labeling was used because the nnos flavohemoprotein contains 24 cysteine residues and 90 lysine residues, many of which are solvent exposed, preventing specific labeling of NOS with external fluorophores. Using Förster resonance energy transfer (FRET), we have studied the spatial and temporal dynamics of CaM bound to nnos when nnos is reduced with NADPH using rapid mixing stoppedflow spectrophotometry. We demonstrate also how use of a flavin analogue (5-deazaflavinmononucleotide; 5-dFMN) [261, 262] in place of FMN can simplify the correlation of structural transitions with NOS reaction chemistry. Our approach is general and mutatis mutandis could be adopted with other complex multicenter redox enzymes. 3.3 Experimental Section. The Supporting Information (SI) (Section 3.6) contains details of chemical suppliers and established methods for cloning, expression and purification of CaM, nnos and the C-terminal kinase domain of FAD synthetase. Likewise, the SI documents contain methods for the enzymatic synthesis of 5-deazaflavin mononucleotide (5-dFMN) from 5-deazariboflavin along with conditions and instrumentation used for nnos steady-state turnover assays, static fluorescence and circular dichroism measurements. 78

79 3. REDOX-LINKED DOMAIN DYNAMICS AND NOS Preparation of 5-dFMN reconstituted NOS. The FMN cofactor was removed from NOS using immobilized metal affinity chromatography [263]. Briefly, 10 mg of purified NOS was applied to a 5 ml Ni-NTA column (GE Healthcare, Little Chalfont, U.K.) equilibrated with 40 mm HEPES buffer (ph 7.6) and supplemented with 10% glycerol and 150 mm NaCl at room temperature. The FMN was removed by washing the column with 25 ml of buffer supplemented with 2 M KBr at room temperature. Next, the column was transferred to a cold room (4 C) and washed with 25 ml of cold buffer before elution of the apoprotein with buffer supplemented with 260 mm imidazole. For reconstitution with 5-dFMN, the eluted apoprotein was added to a solution containing ~0.5 mm 5-dFMN, 0.2 mm FAD, 0.2 mm tetrahydrobiopterin (H 4 B), and 2 mm dithiothreitol (DTT) contained in 50 mm Tris-HCl (ph 8.0) buffer, supplemented with 10% glycerol and 150 mm NaCl, and incubated overnight at 4 C with gentle mixing. The reconstituted protein was concentrated, and excess flavins and other cofactors were removed by applying the concentrated protein solution to an Econo-Pac DG10 column (Bio- Rad, Hempstead, U.K.) equilibrated in 40 mm HEPES buffer (ph 7.6), and supplemented with 10% glycerol and 150 mm NaCl. For anaerobic experiments, the final desalting step was performed in a Belle Technology anaerobic glovebox. Protein concentrations were determined at 444 nm in the presence of CO, using a molar extinction coefficient of 74 mm -1 cm -1 (A 444 A 500 ) Conjugation of extrinsic fluorophores to T34C/T110C-CaM. A solid-state labeling methodology (SSL) was followed to achieve high-efficiency labeling of T34C/T110C-CaM with Alexa555 and Alexa647 maleimide [264, 265]. T34C/T110C-CaM (5 µm) was incubated initially in 40 mm HEPES buffer (ph 7.1), supplemented with 5 mm dithiothreitol (DTT), 1 mm CaCl 2, 150 mm NaCl, and 10 % glycerol for 2 h, at 4 C. The protein was then recovered from solution by ammonium sulfate precipitation and pelleted by centrifugation. To remove traces of DTT, the T34C/T110C-CaM pellet was washed with 40 mm HEPES buffer (ph 7.1) supplemented with 5 M ammonium sulfate, 1 mm CaCl 2, 150 mm NaCl and 10 % glycerol, and recovered by centrifugation. The protein pellet was then resuspended in 40 mm HEPES buffer (ph 7.6) supplemented with 1 mm CaCl 2 and 400 mm NaCl containing 100 µm of the desired fluorophore(s). Labeling reactions were incubated overnight, at 4 C, in the dark. Excess fluorophore was separated from conjugated T34C/T110C-CaM by passing the sample down an Econo-Pac DG10 gel filtration column (Bio-Rad, Hempstead, U.K.) equilibrated with the desired buffer. For all anaerobic experiments this final step was performed in a nitrogen-purged glovebox using oxygen-free buffer. Each fluorophore (100 μm) was mixed with the protein to label T34C/T110C-CaM with an approximate 1:1 ratio of Alexa555 and Alexa

80 3. REDOX-LINKED DOMAIN DYNAMICS AND NOS Stopped-flow spectrophotometry. All stopped-flow measurements were performed using an Applied Photophysics Ltd (Leatherhead, U.K.) SC18MV instrument. The sample handling unit was placed inside a Belle Technology anaerobic glovebox (< 5 ppm of O 2 ). All buffers and solutions were degassed by bubbling with oxygen-free nitrogen prior to entering the glovebox and left overnight to equilibrate to ensure removal of all traces of oxygen. All stopped-flow experiments were performed at 10 C in 40 mm HEPES buffer (ph 7.6), supplemented with 10% glycerol and 150 mm NaCl. For UV-vis measurements of nnos flavin reduction, reactions were initiated by mixing 5 μm NOS with 100 μm NADPH (final concentrations) in the presence or absence of CaM and 1 mm Ca 2+. Under these conditions the observed rates are independent of the coenzyme concentration [191]. Multiple wavelength studies were carried out using a photodiode array (PDA) detector (Applied Photophysics Ltd., Leatherhead, U.K.). In single wavelength studies, flavin reduction by NADPH was monitored at 485 nm, a heme isosbestic point [184, 193, 194]. For kinetic isotope effect (KIE) measurements single-wavelength studies of flavin reduction were performed using pro-r and pro- S NADP 2 H. All measurements were repeated at least five times and are plotted as an average ± 1 standard deviation. Stopped-flow Förster resonance energy transfer (FRET) measurements were performed by mixing 0.25 µm nnos/0.25 µm labeled T34C/T110C-CaM with 100 µm NADPH or 500 µm NADP + (final concentrations) in the presence of 0.5 mm Ca 2+ and 5 μm H 4 B. Dual-channel fluorescence was recorded with a 2 mm excitation path length using two photomultiplier tubes (PMT), including a R1104 red-sensitive photomultiplier detector (Applied Photophysics Ltd., Leatherhead, U.K.) to increase signal to noise for the acceptor channel. The donor channel was fitted with a 600 ± 5 nm bandwidth pass (ThorLabs, Ely, U.K.) filter while the changes in fluorescence emission of the acceptor were monitored using a 650 nm cut-on filter (ThorLabs, Ely, U.K.). All measurements were repeated at least five times and are plotted as an average ± 1 standard deviation. Data were analyzed and interpreted using methods previously described [50, 249]. This analysis involved subtracting the percentage emission of the single labeled CaM-bound fluorophore (Donor- or Acceptor-T34C/T110C-CaM) from the percentage emission of the corresponding fluorophore in the double labeled CaM (DonorAcceptor-T34C/T110C-CaM) to extract the fluorescence changes associated with FRET alone. nnos tryptophan emission changes upon binding NADP + were monitored by mixing 0.2 µm nnos/0.2 µm CaM (final concentration) with varying concentrations of NADP +, in the presence of 0.5 mm Ca 2+. All emission changes associated with tryptophan were recorded using a stopped- 80

81 3. REDOX-LINKED DOMAIN DYNAMICS AND NOS flow cell with a 2 mm excitation path length. Tryptophan was excited at 295 nm, and emission changes were followed using a 340 nm cutoff filter. All kinetic traces were fitted to standard exponential decay functions using Origin Pro (software). 3.4 Results and Discussion A FRET reporter of CaM conformation bound to nnos. Our experimental model for detecting both CaM dynamics and nnos-bound CaM conformational change during the catalytic cycle of nnos is shown in Figure 3.2. Since wild-type CaM has no native cysteine residues, we introduced two solvent-exposed Cys residues (T34C/T110C-CaM) using site-directed mutagenesis, thereby allowing the addition of fluorophores at two specific locations in CaM. Labeling efficiency was > 90 %, and there was no recordable nonspecific fluorophore CaM conjugation (data not shown). This double-cysteine-containing CaM protein has been used previously to monitor CaM dynamics in a variety of published fluorescence studies [225, 259, ], and it is useful here for probing intra CaM dynamics on binding to nnos. This follows because (i) the position of the fluorophore binding sites, one on each of the N- and C- terminal calcium-binding globular domains of CaM, allows small changes in CaM conformation to be detected, (ii) the two maleimide labeling sites are located far away from the calcium-binding pockets on calmodulin and have little or no reported effect on the CaM-calcium interaction, and (iii) the effect of mutagenesis, as well as addition of the bulky fluorophore molecule to CaM, has no noticeable effect on the known catalytic ability of CaM to stimulate nnos steady-state turnover (Figure S3.1 in the Supporting Information). On the basis of predicted fluctuations in distance between the two fluorophore labeling sites when CaM shifts between open and closed conformations (~15-60 Å; Figure 3.2A), we selected the fluorophore pair of Alexa555 (Donor, D) and Alexa647 (Acceptor, A) with a calculated Förster radius (R 0 ) of 47 Å. This allows the monitoring of subtle changes in CaM dynamics by Förster resonance energy transfer (FRET). Note, however, that not all labeled CaM molecules will undergo FRET due to the random nature of the labeling strategy. The selection of the fluorophore-pair took into account also the fluorescence excitation spectra of both fluorophores (λ max values of 555 and 645 nm for donor and acceptor fluorophores, respectively), which are red-shifted from the nnos and CaM intrinsic fluorophores (flavins and aromatic amino acids). Thus, the analysis of FRET data (a reporter of CaM dynamics) is not compromised by undesirable fluorescence emission (Figure S3.2 in the Supporting Information). 81

82 3. REDOX-LINKED DOMAIN DYNAMICS AND NOS Figure 3.2. Ligand binding and the dynamic landscape of CaM. (A) Structures of apo (PDB_1CLL, shown on the left), Ca 2+ -bound (PDB_1CFC, shown in the middle) and both Ca 2+ /nnos-peptide-bound forms of CaM (PDB_2O60, shown on the right). Divalent calcium ions are shown as yellow spheres, and the nnos peptide is represented as an orange ribbon. The distances between the α-carbon atoms of the two fluorophore labeling sites (Cys34 and Cys110; highlighted in red) are 27, 52.4 and 12.4 Å for the apo, Ca 2+ -bound and the Ca 2+ /nnos-bound forms, respectively. (B) Normalized fluorescence emission spectra showing ratiometric changes in the donor and acceptor emission. Samples: Alexa555-Alexa647 labeled T34C/T110C-CaM (black); T34C/T110C-CaM plus Ca 2+ (red); T34C/T110C-CaM plus Ca 2+ and nnos (blue). All data in panel (B) were normalized to the emission maxima of their respective donor-only sample, and any emission changes observed from directly exciting the acceptor were corrected for by subtracting away from the double labeled sample containing the same amount of acceptor (see Figures S3.3 and S3.4 in the Supporting Information). Conditions are described in the Experimental Section. Prior to conducting stopped-flow FRET studies to monitor the dynamics of nnos-bound CaM, we recorded and analyzed fluorophore labeled-t34c/t110c-cam fluorescence emission to see if our experimental model for tracking CaM dynamics fit to previously published ideas of how the conformational landscape of CaM adjusts on ligand binding. Relevant CaM structures determined by X-ray crystallography [268] and NMR spectroscopy [269] are shown in Figure 3.2, along with the ratiometric changes in the equimolar donor to acceptor labeled (DA) T34C/T110C-CaM emission when CaM binds Ca 2+ and nnos (see Figure S3.3 and S3.4 in the Supporting Information for unprocessed data, including single fluorophore labeled D/A-T34C/T110C emission under the same conditions, and for information on data normalization). The altered fluorescence data 82

83 3. REDOX-LINKED DOMAIN DYNAMICS AND NOS indicate a change in the conformational landscape of apo-cam on binding Ca 2+. This shift in conformation from a closed to a more open form is evident from the reduced FRET efficiency: an increase in donor emission at 570 nm, and decrease in acceptor emission at 670 nm relative to fluorophore emission in the Ca 2+ -free DA-T34C/T110C-CaM when fluorophores were excited at 555 nm. When the fluorophore labeled apo-t34c/t110c-cam was incubated with both Ca 2+ and nnos, an increase in the acceptor emission along with an anticorrelated response in donor emission were observed. This recorded FRET change is indicative of CaM forming a compact conformer with short Cys34-Cys110 distances, previously observed from published structural data of NOS [175] and nnos peptides [270] bound to CaM. Prior to performing time-resolved experiments, we assessed the possibility for inter-cam FRET (FRET from CaM-CaM across the nnos homodimer), which in principle could affect the interpretation of nnos-bound intra-cam FRET data (i.e., FRET within the individual nnos-bound CaM proteins). Two separate batches of T34C/T110C-CaM were labeled with either donor or acceptor fluorophores, giving two separate and single-labeled CaM samples (D- and A- T34C/T110C-CaM). An equimolar mix of the D and A single labeled T34C/T110C-CaM was added to 1 mol equiv of nnos, and no FRET was observed across the nnos dimer (Figure S3.5 in the Supporting Information). These data are in agreement with recently published cryo-em structures [177] of native NOS proteins, which show the distances (> 90 Å) between the two nnos-bound CaM molecules to be outside the range for efficient energy transfer between the fluorophore pair Electron transfer kinetics in nnos in the presence and absence of CaM. We probed the kinetics of NADPH-dependent nnos flavin reduction by transient absorption stopped-flow spectrophotometry under pseudo-first-order conditions (20-fold excess NADPH) [191]. nnos is a complex enzyme with multiple cofactors (FAD, FMN, and heme) that are rich in optical features spanning the UV-visible spectrum. To study reduction of the nnos flavins (FAD and FMN) by stopped-flow spectrophotometry, we followed the quenching of flavin absorbance at 485 nm, an isosbestic point for heme [184, 193, 194]. Stopped-flow traces reporting on nnos (±CaM) flavin reduction, along with their respective exponential fits, are presented in Figure 3.3. Transients were fit optimally to five exponential terms (Figure 3.4B) over a time-scale of 2 µs to 200 s. However, given that the nnos-cam steady-state turnover values, k cat, for NO formation and NADPH consumption at 10 C are ~0.1 and 0.3 s -1, respectively, we focus here only on the first four kinetic phases, as the slow fifth phase (with a rate constant of ~0.01 s -1 ) is not relevant to catalysis (Figure 3.3). Four resolvable kinetic phases have been seen previously in studies of the 83

84 3. REDOX-LINKED DOMAIN DYNAMICS AND NOS isolated nnos reductase domain [191, 192] and in line with the study by Knight and co-workers [191], we observed no appreciable increase in the rate of flavin reduction in full length nnos in the presence of CaM in comparison with its absence. CaM does nonetheless have an effect on the relative amplitudes of the individual kinetic phases. These CaM-dependent influences on the amplitudes of kinetic phases are significant in all four phases (Figure 3.3 and Table 3.1) and as previous studies have implied [199], it is likely that CaM has a role in governing nnos flavin redox potentials. Alongside a CaM-induced structural change of nnos that will affect the electron transfer geometry/distance, any alteration in redox potentials might also account for observed stimulation of electron transfer rates from nnos FMN to the extrinsic partner protein cytochrome c in comparison to CaM-free nnos (Table S3.1 in the Supporting Information). The involvement of the four observed kinetic phases observed in stopped-flow studies of nnos flavin reduction by NADPH in the enzyme catalytic cycle was probed by kinetic isotope effect (KIE) measurements using site-specifically deuterated NADPH (pro-r and pro-s NADP 2 H). KIE values were significant for the first three phases observed in stopped-flow studies of flavin reduction when excess pro-r (but not pro-s) NADP 2 H is mixed with nnos (Table S3.2 in the Supporting Information). The KIE values observed for the first three kinetics phases suggest that they report on hydride transfer from NADPH to FAD. These phases however do not map to discrete mechanistic steps, as the electron transfer steps are reversible and coupled. This accounts for the observation of primary KIEs in each of the first three phases measured in our stopped-flow studies. Below we demonstrate that interflavin electron transfer is predominantly associated with the fourth kinetic phase in studies with nnos that contains 5-dFMN rather than the conventional FMN (vide infra) Direct monitoring of CaM dynamics during catalytic turnover of nnos. We have used FRET stopped-flow spectroscopy to detect transient nnos-bound CaM conformations that appear during the reaction cycle of nnos. Donor/acceptor fluorophoreconjugated T34C/T110C-CaM (DA-T34C/T110C-CaM) was bound to nnos in equimolar concentration and rapidly mixed with NADPH. The time-resolved fluorescence emission was followed on the same time-scale as the nnos reaction chemistry (vide supra). nnos-bound DA- T34C/T110C-CaM was also mixed with buffer only in order to assess any potential photobleaching over the time scale of the rapid mixing experiments (200 s). No changes in emission were recorded over this time period for the acceptor fluorophore, and very small changes (~1%) were observed for the donor fluorescence. Since these emission changes were minor, they were omitted from subsequent analysis of all fluorescence emission data. FRET data representing the 84

85 3. REDOX-LINKED DOMAIN DYNAMICS AND NOS single-turnover conformational changes of nnos-bound CaM on reduction with NADPH are shown in Figure 3.4A,B. Here the represented time-resolved donor and acceptor fluorescence emission changes are a deconvolution from other contributions to the emission response (Figure S3.7 in the Supporting Information), most notably the spectral changes in the nnos bound chromophores (haem and flavins). We extracted the fluorescence emission change associated with donor-acceptor fluorophore FRET by subtracting the single-labeled (donor or acceptor) T34C/T110C-CaM bound nnos sample relative emission from the relative emission of the corresponding fluorophore in the double-labeled (donor and acceptor) T34C/T110C-CaM bound nnos sample (for more information see refs [50, 249]). Changes in the donor and acceptor emission were anti-correlated, indicative of changes in FRET efficiency, and were fit to a fourexponential decay function (Figure 3.4A). Figure 3.3. Anaerobic stopped-flow transients obtained at 485 nm on mixing 5 µm NOS (final concentration) with a 20-fold excess of NADPH in the presence or absence of CaM: (A) native NOS mixed with NADPH; (B) native NOS mixed with pro-r NADP 2 H; (C) native NOS mixed with pro-s NADP 2 H; (D) 5-dFMN-substituted NOS mixed with NADPH. Data and respective fits to an equation describing four sequential exponential processes are shown. Measurements were performed at least twice with different NOS preparations. Representative transients shown are the average of six to eight individual traces. FRET data, which are presented as a ratio of donor to acceptor emission (Figure 3.4B), clearly exemplify an opening of the compact CaM protein, which is bound to nnos initially in the oxidized 85

86 3. REDOX-LINKED DOMAIN DYNAMICS AND NOS form when it is rapidly mixed with NADPH. This opening is kinetically coupled to early stages of the flavin reduction chemistry (k 1 ), involving the formation of a mixture of enzyme species (i.e., predominantly a distribution of FAD hydroquinone and oxidized FAD-NADPH charge transfer (CT) species; see ref [191] for a more detailed discussion of the reaction mechanism). Following the formation of this predominantly more open CaM substate, time-dependent emission changes in the donor and acceptor fluorescence show CaM to close, revealing a more compact conformer with shorter interfluorophore distances. This subsequent closing of the transiently opened nnosbound CaM conformer occurs in a single kinetic process with a rate constant of 15.6 s -1, which is similar to the k 2 value observed for flavin reduction (Table 3.1 and 3.2). This predominantly closed CaM conformer appears to be structurally similar to the form of CaM bound to oxidized nnos (i.e., it has the same FRET emission properties). The third rate constant (k 3 ) for flavin reduction appears not to be associated with intra-cam conformational change on binding to nnos (at least within the detection limits of the instrumental setup used). The observed rate constants associated with the fourth (k 4 ) and fifth (k 5 ) kinetic phases of nnos flavin reduction correlate closely with observed kinetic phases that report on intracam dynamics (i.e., a well-defined decrease in CaM-bound fluorophore FRET efficiency is observed). There are no crystallographic data for full length nnos: thus, the orientation that CaM adopts in the nnos calmodulin binding site is not known. Structures only exist for the isolated reductase component [170, 174], the oxygenase domain [ ] and the FMN domain bound to CaM [175]. Consequently, how the structure of nnos is affected by the change in CaM structure is not known. Current models invoke a complex landscape for the FAD, FMN, and oxygenase domains, and the mechanism of electron transfer is thought to involve conformational sampling of the FMN domain to facilitate electron transfer from the FAD domain to the heme oxygenase component. CaM is implicated, albeit in an ill-defined way, in this conformational sampling mechanism and is known to assist electron transfer from the FMN domain to the heme oxygenase. What we see for the first time from the current study is that CaM itself undergoes complex structural transitions during the catalytic cycle of nnos, and these dynamic changes are likely central to the conformational sampling mechanism and the coordination of electron delivery from the FAD domain to the heme oxygenase. Our FRET studies suggest that changes in the CaM conformation are coupled to early stages of flavin reduction, specifically FAD (but not FMN) reduction. To test this idea further, we have exploited the use of the FMN derivative 5-deazaflavin mononucleotide (5-dFMN) [261, 262], which cannot stabilize the flavin semiquinone species. Incorporation of this analogue in nnos is 86

87 3. REDOX-LINKED DOMAIN DYNAMICS AND NOS therefore expected to block FAD to FMN transfer in nnos but still enable reduction of the FAD domain by NADPH. The availability of 5-dFMN nnos would therefore simplify the redox chemistry in stopped-flow studies with NADPH and enable study of conformational change by FRET, while (importantly) retaining the overall structural integrity of nnos New form of nnos containing 5-deazaflavin-mononucleotide (5-dFMN). We simplified stopped-flow analysis of electron transfer in nnos by preventing electron flow to the FMN and oxygenase domains, enabling more precise correlation of CaM conformational change with redox chemistry. This was achieved by substituting the natural FMN cofactor with the flavin biomimetic 5-dFMN. The affinity of the FMN cofactor for NOS is known to be less than that of the tightly bound FAD and heme cofactors [223]. The FMN co-factor was removed from nnos using the weak chaotropic agent potassium bromide [263]. Apo-flavoproteins can generally be reconstituted with flavin derivatives modified at the isoalloxazine moiety, since the major interactions of the protein with the cofactor are associated with the N10 side chain [263, 271]. 5- dfmn is structurally similar to conventional FMN and has been used previously as a biomimetic with other flavoproteins [271]. By incubating 5-dFMN with FMN-depleted nnos the FMN binding site was replenished with 5-dFMN. Differences in the spectral properties of native and 5-dFMN bound nnos (specifically, a decrease in absorbance at ~450 and ~380 nm and an increase in absorbance at ~400 and ~340 nm; Figure S3.11 in the Supporting Information) are entirely consistent with the known absorption properties of 5-dFMN and FMN (Figure S3.12 in the Supporting Information). Moreover, the secondary structure of 5-dFMN nnos, determined by far- UV circular dichroism (CD), is essentially identical with that of native nnos (Figure S3.10 in the Supporting Information). The overall protein secondary structure is therefore retained following replacement of FMN with 5-dFMN. The activity of the reconstituted 5-dFMN nnos enzyme was investigated using a variety of enzymatic steady-state turnover assays. In addition to reduction of the heme in the oxygenase domain, the diflavin reductase domain of nnos is able to reduce artificial electron acceptors such as cytochrome c (cyt c) and ferricyanide. Cyt c accepts electrons only from the FMN cofactor of nnos, but ferricyanide accepts electrons from both the FAD and FMN cofactors (Scheme S3.1 in the Supporting Information) [184]. Using steady-state assays of cyt c and ferricyanide reduction, as well as monitoring NADPH depletion and NO formation, electron flow through the various NOS cofactors can be established. Exchange of the FMN cofactor for 5-dFMN completely abolished steady-state cytochrome c reduction and NO formation, and almost completely eliminated NADPH consumption (~1 % activity remained in comparison with native nnos, which is attributed 87

88 3. REDOX-LINKED DOMAIN DYNAMICS AND NOS to reoxidation of nnos-bound FADH 2 by molecular oxygen) [272]. Ferricyanide reduction is maintained in 5-dFMN nnos (Table S3.1 in the Supporting Information), consistent with the known ability of the FAD domain to transfer electrons to ferricyanide. That activities known to be dependent on interflavin electron transfer are abolished is consistent with our expectation that interflavin electron transfer is blocked in 5-dFMN nnos as a result of the very unfavorable reduction potential for 5-dFMN semiquinone formation [261]. This was further corroborated by anaerobic reductive titration of native nnos and 5-dFMN NOS with both NADPH (electrons enter only via the FAD domain) and dithionite (DT), where electrons can access each of the flavin/heme cofactors directly. With NADPH (Figure S3.11 in the Supporting Information), the absorbance changes are consistent with loss of oxidized flavin and appearance of the blue flavin semiquinone species, which, as reported previously, is a consequence of interflavin electron transfer following initial reduction of FAD to FADH 2 by NADPH [82, 191]. With 5-dFMN nnos (reduced by NADPH), depletion of oxidized flavin is observed, but there is minimal appearance of the blue semiquinone signature. This is consistent with the expected block on electron transfer to the FMN domain. In contrast, reductive titration with DT (Figure S3.11) results in full reduction (flavins and haem) of nnos and 5-dFMN nnos. These findings are consistent with a variety of previously published papers that emphasize the different redox chemistry of flavin and 5-deazaflavin [261, 262] in relation to semiquinone stabilization, which is an obligate intermediate in the catalytic cycle of NOS [191] Internal electron transfer is prevented in 5-dFMN nnos. Similar to the transient absorption stopped-flow studies of native nnos, we measured the singleturnover NADPH-dependent reduction of 5-dFMN nnos by rapid mixing of oxidized 5-dFMN nnos with excess NADPH (20 mol equiv) [191]. Stopped-flow transients measured over 10 s are shown in Figure 3.3. The blue shift in 5-dFMN spectral features allows monitoring of the nnos-bound FAD co-factor alone at 485 nm (and see accompanying spectra recorded with a photodiode array in Figure S3.6 in the Supporting Information). In contrast to comparable experiments performed with nnos (vide supra), where flavin reduction was described by four kinetic phases over 10 s, in the case of 5-dFMN nnos flavin reduction (monitored at 485 nm) flavin reduction occurred in three kinetic phases. Observed rate constants determined by fitting to a triple-exponential equation and associated amplitude changes are similar to those recorded for the native enzyme (k 1 to k 3 ). The slower fourth phase seen for nnos is absent in studies with 5-dFMN NOS, consistent with the inability to transfer electrons from FAD to 5-dFMN. 88

89 3. REDOX-LINKED DOMAIN DYNAMICS AND NOS The slow fourth phase observed with nnos has an observed rate constant similar to that of k cat in steady-state NO production and NADPH depletion assays (Table S3.1 in the Supporting Information). This suggests that this kinetic phase contributes to rate limitation in steady-state turnover, together with other established steps in the catalytic cycle (notably FMN to heme electron transfer) [33]. We emphasize again that this is a kinetic phase associated with an observed spectral change and does not report on a single mechanistic step. As with other diflavin reductases such as the well-characterized cytochrome P450 reductase [82], the kinetic phases observed for nnos report on the formation of a distribution of enzyme intermediates. In all likelihood, the slow fourth phase reports not only on interflavin electron transfer but also on FMN to heme electron transfer in a single kinetic process (but heme reduction is not observed at the wavelength we monitored). The finding that interflavin electron transfer is relatively slow in nnos is in line with a previous study on the isolated nnos reductase domain. This study suggested that interflavin electron transfer might also be gated by NADP + release [191], consistent with steadystate and isotope effect studies reported by others [273]. Over extended time scales (200 s) stopped-flow studies with 5-dFMN NOS also revealed an additional (fourth) kinetic phase (Figure 3.4D). This step has the same kinetics as formation of the native nnos EQ state (~0.01 s -1 for native and 5-dFMN nnos), albeit with a different associated amplitude (0.006 and for native and 5-dFMN nnos). This EQ state has been discussed in many studies with diflavin oxidoreductases as a signal that likely results from further conformational change and/or further oxidation of NADPH attributed to thermodynamic relaxation through disproportionation reactions [82, 191]. The formal attribution of this phase to mechanistic processes is complicated. It is not relevant to steady-state catalysis and for that reason we have chosen not to comment on it further in this work Correlating conformational change with early steps in electron transfer in 5-dFMN nnos. With the availability of 5-dFMN NOS we were able to investigate conformational changes linked to FAD reduction, in the absence of electron transfer to the FMN and oxygenase domains. Stopped-flow FRET data for the reaction between 5-dFMN nnos-bound to donor-acceptor labeled CaM and NADPH are shown in Figures 3.4C,D. Anti-correlated donor and acceptor emission transients along with FRET data (i.e., the ratio of donor and acceptor emission) were optimally fit to four exponential decays over 200 s. Rate constants associated with the four phases are similar to those recorded for donor-acceptor labeled CaM bound to native nnos (Table 3.2). In addition, the directionality of conformational change (i.e., opening and closing 89

90 3. REDOX-LINKED DOMAIN DYNAMICS AND NOS of CaM) for 5-dFMN nnos-bound CaM mimics that for nnos. Some differences in FRET amplitudes (degree of opening or closing ) between nnos and 5-dFMN nnos-bound CaM can be seen in the third and fourth phases. In nnos enzyme these phases are associated with the kinetics of interflavin electron transfer and formation of the EQ state, respectively. NOS reductant k 1 (s -1 ) ΔA 1 k 2 (s -1 ) ΔA 2 k 3 (s -1 ) ΔA 3 k 4 (s -1 ) ΔA 4 native NADPH (67.1) (0.003) 41.2 (27.3) (0.003) 3.5 (1.4) (0.008) 0.53 (0.28) (0.005) native pro-s NADP 2 H (96.9) (0.004) 38.3 (26.6) (0.001) 2.5 (0.1) (0.006) 0.29 (0.08) (0.011) native pro-r NADP 2 H 23.2 (0.6) (0.004) 5.4 (0.1) (0.001) 1.3 (0.1) (0.001) 0.23 (0.07) (0.001) 5- dfmn native + CaM native + CaM native + CaM 5- dfmn + CaM NADPH NADPH pro-s NADP 2 H pro-r NADP 2 H NADPH (13.4) (30.5) (5.3) 50.9 (3.8) (19.4) (0.004) (0.009) (0.002) (0.007) (0.014) 22.4 (4.2) 20.9 (40.4) 9.9 (1.6) 8.8 (0.3) 53.0 (6.2) (0.007) (0.012) (0.003) (0.004) (0.002) 4.8 (0.7) 4.1 (5.7) 1.5 (0.4) 0.68 (0.17) 5.5 (2.5) (0.001) (0.06) (0.001) (0.006) (0.001) ND 0.10 (0.06) 0.09 (0.06) 0.06 (0.03) ND (0.017) (0.026) (0.001) Table 3.1. Kinetic parameters extracted from Figure 3.3. Observed rate constants (k) and absorbance changes (ΔA) determined by fitting transients in Figure 3.3 to exponential decay functions. Estimated errors are given in parentheses. ND = not detected The data indicate that interflavin electron transfer/fmn reduction do not drive CaM conformational change, as this step is blocked in 5-dFMN nnos. We infer therefore that the observed FRET conformational changes are associated with early steps in the electron transfer sequence, which relate either to NADPH-binding and/or reduction of the FAD. We note that the extended C-terminal tail of nnos (and other NOS isoforms) needs to be displaced from the NADPH-binding site on mixing nnos with NADPH [184, 186, 187, 192, 274] and this might be a mechanism for connecting binding and/or electron transfer events with the observed FRET signals that report on intra-cam motions. ND ND 90

91 3. REDOX-LINKED DOMAIN DYNAMICS AND NOS Figure 3.4. Dynamics of nnos-bound CaM during NADPH-driven nnos flavin reduction with native (panels A and B) and 5-dFMN (panels C and D) nnos. (A) and (C) show the time-resolved anticorrelated emission changes of donor and acceptor fluorophore labeled T34C/T110C-CaM bound in equimolar concentrations to nnos and 5-dFMN nnos, respectively, on mixing with excess NADPH in a stopped-flow instrument. (B) and (D) show NADPH reduction of native and 5-dFMN nnos recorded over 200 s (black) for the reaction between 5 μm nnos bound to equimolar DA-T34C/T110C-CaM on mixing with excess NADPH along with the ratio of donor to acceptor fluorophores (blue) representing defined CaM opening (increased Cys-Cys distances) and closing (decreased Cys-Cys distances) steps during turnover. The first four rate constants associated with nnos flavin reduction are relevant to enzyme turnover and have been labeled accordingly as flavin reduction ; the slow phase is not relevant to steady-state turnover and has been termed the EQ state (see main text for a more detailed discussion). The area between the black dotted lines in (D) (~ 1-11 s) corresponds to the interflavin electron transfer step that is lost in the 5-dFMN nnos variant. See the Experimental Section (Section3.3) for details on conditions and instrumentation used. NOS k 1 (s -1 ) ΔC 1 k 2 (s -1 ) ΔC 2 k 3 (s -1 ) ΔC 3 k 4 (s -1 ) ΔC 4 y0 native 274 (47) (0.01) 15.6 (2.1) 0.09 (0.01) 0.17 (0.05) (0.00) (0.002) (0.002) 1.10 (0.02) 5- dfmn 205 (38) (0.02) 33.2 (5.2) 0.11 (0.01) 0.35 (0.36) (0.00) (0.001) (0.003) 1.13 (0.02) Table 3.2. Rate constants (k), relative fluorescence changes (ΔC), and ordinate intercept (y0) values determined from fitting donor/acceptor fluorescence transients (Figure 3.4B,D) to exponential decay functions. Estimated errors are given in parenthesis. 91

92 3. REDOX-LINKED DOMAIN DYNAMICS AND NOS NADP + binding remodels the intracam landscape in nnos but over longer time scales. We also investigated the effects of coenzyme binding in the absence of nnos reduction on the modulation of the intra-cam conformational landscape. In this case, NADP + was used rather than NADPH to prevent reduction of nnos flavins. Quenching of intrinsic nnos tryptophan fluorescence emission was used to monitor the binding of NADP +. By titrating nnos (bound to DA- T34C/T110C-CaM) with NADP +, we were able to measure a dissociation constant, K d, of 106 ± 27 μm for the enzyme-coenzyme complex. By following Trp fluorescence in the stopped-flow instrument when rapidly mixing saturating concentrations of NADP + with the nnos DA- T34C/T110C-CaM complex the observed rate of coenzyme binding was obtained. The expected fluorescence changes associated with coenzyme-enzyme interaction occurred within the dead time (1.5 ms) of the stopped-flow instrument (Figure 3.5). Consequently, and in line with other diflavin oxidoreductases [50], we conclude that NADP + binding to nnos is rapid (> 500 s -1 at 10 C). To study the effects of NADP + binding on nnos-bound intra-cam dynamics, donor-acceptor fluorophore labeled-t34c/t110c-cam bound to 1 mol equiv of nnos was mixed with saturating concentrations of NADP +. FRET stopped-flow transients representing the nnos-bound intra-cam dynamics for the reaction between NADP + and equimolar concentrations of nnos and T34C/T110C-CaM are shown in Figure 3.6. No FRET changes were observed in the dead time of the stopped-flow instrument, indicating that there are no binding induced nnos-bound CaM conformational changes. However, conformational change was observed only over extended time scales (up to 200 s) and transients reporting on nnos-bound CaM conformational change were analyzed using a double-exponential function (Figure 3.6). The two observed kinetic phases have rate constants similar in value to k 3 and k 4 recorded for CaM dynamics in NADPH-reduced native and 5-dFMN nnos (Table 3.2 and S3.4 in the Supporting Information). The response to NADP + binding is therefore different from that observed in stopped-flow studies with NADPH, where CaM dynamics were observed on shorter time scales and were associated with NADPH binding and FAD reduction. This suggests either: (i) that the mode of interaction of NADP + and NADPH with nnos is sufficiently different so as to elicit different responses in the remodeling of the CaM landscape or (ii) that redox changes in the FAD are also required to drive the relatively fast remodeling of the CaM landscape in stopped-flow studies of nnos reduction with NADPH. Either way, it is clear that NADPH binding/fad reduction is the primary trigger for the remodelling of CaM rather than internal electron transfer to FMN/haem. 92

93 3. REDOX-LINKED DOMAIN DYNAMICS AND NOS Figure 3.5. NADP + + nnos binding occurring in the dead time of the stopped-flow instrument. (A) Fluorescence emission changes of the nnos tryptophans in the nnos:cam complex upon binding the oxidized coenzyme NADP +. The insert in (A) shows the tryptophan emission changes associated with NADP + binding to the nnos:cam complex recorded in the dead time of the stopped-flow instrument. In the inset the black trace represents the mixing of the nnos:cam complex with buffer alone, while red and blue are labeled accordingly with the final NADP + concentrations used in the stopped-flow study. (B) Static titration data (black) and the emission changes recorded in the dead time of the stopped-flow instrument (red) for the NADP + -dependent changes in the tryptophan emission of the nnos:cam complex. Data in (B) were fitted to a hyperbolic binding function, which gives an apparent K d value for the nnos:cam-nadp + complex of 106 ± 27 μm. Figure 3.6. NADP + binding to nnos driving conformational change of CaM in the nnos-cam complex. (A) Time-resolved anticorrelated emission changes of donor and acceptor fluorophore labeled T34C/T110C-CaM bound in equimolar concentrations to nnos when mixed with excess NADP + (500 μm final concentration) in a stopped-flow instrument. (B) Ratio of donor to acceptor fluorophores presented in (A) representing the opening of nnos-bound CaM (i.e., an increase in the Cys-Cys distance). See the Experimental Section (Section 3.3) for details on conditions used and instrument setup. 93

94 3. REDOX-LINKED DOMAIN DYNAMICS AND NOS 3.5 Concluding remarks. It has been known for some time that ligand-protein interactions affect the conformational landscape of enzyme molecules [30, 38]. Ligand-induced conformational changes have been documented previously in the diflavin oxidoreductase family and by NADP + binding in nnos [198]. Moreover, crystallographic studies of NOS and other diflavin oxidoreductases have also implicated a role for ligand-induced conformational change in these enzymes, particularly in relation to conformational sampling of the FMN binding domain in relation to the FAD and oxygenase domains (i.e., the so-called input and output states of nnos). A major and largely unmet challenge has been the need to correlate protein motions (i.e. in realtime) across the catalytic cycle so that a more holistic understanding of the role of dynamics in enzyme mechanisms can be inferred. In this paper we have begun to address this limitation by correlating the dynamics of CaM bound to the reaction chemistry of nnos during what is a highly complex reaction cycle. This has enabled us to pinpoint major conformational changes in CaM as a function of time to correlate these changes with specific chemical steps in the reaction cycle and identify/suggest the mechanistic triggers for these major conformational adjustments. This extends appreciably our current understanding of nnos dynamics inferred from structural, singlemolecule, EPR, electron microscopy and kinetic approaches, and it paves the way for similar analyses on other complex redox systems where knowledge of dynamics in relation to chemistry is important in advancing the mechanistic description of catalysis. Our analysis has shown that major remodeling of the CaM landscape occurs during early phases of electron transfer during the NADPH-dependent reduction of nnos. A combination of isotope effects, FRET studies, and use of 5-dFMN to block internal electron transfer has enabled more precise mapping of kinetic phases observed in stopped-flow experiments to domain-specific redox changes. Although these phases cannot be attributed to a single mechanistic step, we have demonstrated that binding and/or redox changes associated with the FAD domain are correlated with major remodeling of the bound CaM in nnos. This remodeling is likely triggered in part through displacement of the C-terminal tail from the NADPH binding site by the incoming nicotinamide coenzyme, but other factors might also be in play. The approaches we have developed should find wider application in related studies with NOS isoforms where simplification of the reaction chemistry and strategic positioning of fluorescence reporters can be used to inform not only on the dynamics of CaM, but also on the relative orientations and timedependent conformational remodeling of other domains in NOS enzymes. 94

95 3. REDOX-LINKED DOMAIN DYNAMICS AND NOS 3.6 Supporting Information Experimental Section Materials. All reagents were of analytical grade and were purchased from Sigma-Aldrich (Gillingham, Dorset, U.K.), unless otherwise stated. 5-deazariboflavin (5-dRF) was synthesized by Salford Ultrafine Chemicals and Research Limited (Manchester, U.K.). [4(R)-2H]NADPH (pro-r NADP2H) and [4(S) 2H] NADPH (pro-s NADP2H) were prepared and characterized as described previously [275]. Alexa Fluor 555 C2 (A555) and Alexa Fluor 647 C2 (A647) maleimide were purchased from Thermo Fisher Scientific (Loughborough, U.K.). Pre-cast Sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE) gels were purchased from Bio-Rad (Hempstead, U.K.) Cloning, expression, and purification of tcrfk. A synthetic gene encoding FAD synthetase from C. ammoniagenes [276] was obtained from GenScript. A truncated version encoding the C-terminal riboflavin kinase domain (tcrfk) [277] was amplified using the oligonucleotide primers tcrfk_forward (5 - GGGATTCATATGTTTTATGTTACCGGTCCG-3 ) and rfs-his_reverse (5 - CACACTCGAGTGATTCGGCTTGCAGAAAG-3 ), and cloned between the NdeI and XhoI restriction sites of the pet22b(+) vector (Novagen, Prudhoe, U.K.) resulting in plasmid pet-tcrfk-his. Escherichia coli BL21(DE3) cells harbouring the pet-tcrfk-his plasmid were grown in 1 litre cultures of Luria-Bertani medium supplemented with 50 µg ml -1 carbenicillin at 37 C until and OD 600 of ~ 0.6 was reached. Protein expression was induced by the addition of 0.3 mm isopropyld-1-thiogalactopyranoside (IPTG), and incubation was continued at 25 C for 16 h before harvesting the cells by centrifugation. The cell pellet was re-suspended in 50 mm Tris-HCl (ph 8.0) supplemented with 200 mm NaCl and 10 mm imidazole, and the cells were broken by ultrasonication. The lysate was cleared by centrifugation and loaded onto a Ni-IDA Agarose (Generon, Maidenhead, U.K.) column equilibrated with lysis buffer. The column was washed with 75 mm imidazole in the same buffer, and the tcrfk protein was eluted with 250 mm imidazole. The purified protein was dialysed against 50 mm Tris-HCl (ph 8.0) supplemented with 200 mm NaCl and stored at -20 C. Protein concentrations were determined using a calculated molar extinction coefficient ( 280 ) of mm -1 cm -1 at 280 nm Expression and purification of native NOS. Native His 6 -tagged rat neuronal NOS was overexpressed in E. coli and purified by Ni-IDA Agarose affinity chromatography and 2,5 -ADP-Sepharose (GE Healthcare, Little Chalfont, U.K.) affinity 95

96 3. REDOX-LINKED DOMAIN DYNAMICS AND NOS chromatography as described previously [198, 278]. Protein concentrations were determined at 444 nm in the presence of carbon monoxide (CO), using a molar extinction coefficient of 74 mm -1 cm -1 (A 444 A 500 ) for the ferrous heme-co adduct [279]. Immediately prior to kinetic experiments, native NOS was incubated with 0.2 mm FAD, 0.2 mm FMN, 0.2 mm tetrahydobiopterin (H 4 B) and 2 mm DTT for ~30 min on ice. Next, a few grains of potassium ferricyanide were added and excess reagents were removed by applying the protein solution onto a Econo-Pac DG10 gel filtration column (Bio-Rad, Hempstead, U.K.) equilibrated in 40 mm HEPES (ph 7.6) buffer supplemented with 10 % glycerol and 150 mm NaCl. For anaerobic experiments, the final desalting step was performed in an anaerobic glove box (Belle Technology, Weymouth, U.K.) under nitrogen atmosphere in which oxygen levels were kept below 5 ppm Cloning, expression, and purification of native and T34C/T110C calmodulin (CaM). The CaM variant T34C/T110C was constructed using the wild-type CaM containing pcoladuet-1 plasmid template by using the QuikChange TM method (Agilent, Stockport, U.K.). Both recombinant mammalian CaM and the T34C/T110C were expressed in E. coli BL21(DE3) and purified using a single Phenyl-Sepharose hydrophobic interaction chromatography (GE Healthcare, Little Chalfont, U.K.) step as described before [198, 226]. Protein concentrations were determined at 276 nm using an extinction coefficient of 276 = mm -1 cm Biosynthesis of 5-deaza-FMN. For the biosynthesis of 5-deaza-FMN (5-dFMN), a solution containing 0.5 mm 5-dRF, 1 mm ATP, 10 mm MgCl 2 and 10 µm tcrfk in 50 mm Tris-HCl (ph 8.0) was incubated at 37 C for 120 min while shaking. The reaction was stopped by incubating the solution for 5 min at 95 C, and any insoluble material was removed by centrifugation. The reaction products were identified by thinlayer chromatography (TLC). Samples taken during the reaction, and the reference compounds FAD, FMN, riboflavin, and 5-dRF were applied onto a silica gel 60 F 254 TLC plate (Merck Millipore, Nottingham, U.K.). Butanol:acetic acid:water (12:3:5) was used as mobile phase and fluorescent spots were visualized by UV illumination Circular dichroism (CD) measurements. Circular dichroism was performed on an Applied Photophysics (Leatherhead, U.K.) Chirascan qcd spectrometer at 25 o C using a sealed cuvette with a 0.1 mm path length. Measurements were performed with 5 µm native or 5-dFMN nnos in 40 mm HEPES (ph 7.6), supplemented with 150 mm NaCl and 10 % glycerol. Mean residual ellipicities (MRE, [Ɵ] MR ) were calculated using eq S3.1. Where Ɵ is degrees of ellipicities, l is the cuvette path length in cm and C MR is the mean residue 96

97 3. REDOX-LINKED DOMAIN DYNAMICS AND NOS concentration. C MR is calculated using eq. 3.2 where n is the number of amino acids and c is the concentration of the protein in molar. [θ] MR = 100θ/(C MR l) Equation S3.1 C MR = nc Equation S Static fluorescence measurements. Fluorescence emission spectra were recorded on an Edinburgh Instruments (Livingston, U.K.) FLS920 fluorometer equipped with double excitation and emission monochromators, a redsensitive cooled photo-multiplier detector, and a 450 W xenon arc lamp. Spectra were recorded using 0.5 nm excitation and 5 nm emission slit-widths in 1 ml fluorescent quartz cells (Starna Scientific Ltd, Hainault, U.K.) with a 10 mm excitation path length. Fluorescence emission data were collected at 25 C in 40 mm HEPES (ph 7.6), 150 mm NaCl and 10 % glycerol, unless otherwise stated. For Ca 2+ free measurements a calcium sponge (Molecular Probes, Thermo Fisher Scientific Loughbourough, U.K.) was used to remove the divalent ion from all buffers Steady-state activity measurements. The steady-state turnover of native- and 5-dFMN reconstituted-nnos was determined at 10 o C on assay mixtures containing catalytic amounts of NOS, 10 mm L-Arginine, 0.5 mm CaCl 2, and 0.1 mm NADPH, in the presence or absence of 7 µm CaM in assay buffer. Steady-state NADPH oxidation rates were determined at 340 nm ( 340 = 6.22 mm -1 cm -1 ), cytochrome c reduction at 550 nm ( 550 = 21.1 mm -1 cm -1 ) in the presence of 10 µm bovine heart cytochrome c, ferricyanide reduction at 420 nm ( 420 = 1.04 mm -1 cm -1 ) in the presence of 1 mm potassium ferricyanide, and NO formation was monitored at 401 nm ( 401 = 38 mm -1 cm -1 ) in the presence of 10 µm oxyhemoglobin. All steady state assays were conducted in 40 mm HEPES (ph 7.6) supplemented with 150 mm NaCl, 10 % glycerol. 97

98 3. REDOX-LINKED DOMAIN DYNAMICS AND NOS Results Steady State Turnover Scheme S3.1. Electron flow through the NOS diflavin oxidoreductase to intrinsic and extrinsic electron accepting partners. Black and blue arrows are representative of electron flow through nnos and from nnos to redox partners, respectively [184]. CaM is required for cross-monomer electron transfer from nnos FMN to heme [200, 256]. CaM along with all electron accepting partners that require CaM-nNOS binding are shown in italics. 98

99 3. REDOX-LINKED DOMAIN DYNAMICS AND NOS Figure S3.1. Comparison between native- and DA-T34C/T110C-CaM ( DA-CaM ) bound nnos steady-state turnover numbers. Black, red and blue bars represent NADPH consumption, NO formation and cyt c reduction assays, respectively. Data are normalized to the native nnos assay turnover numbers. 99

100 3. REDOX-LINKED DOMAIN DYNAMICS AND NOS NOS FeCN *cyt c NADPH NO reduction (s -1 ) reduction (s -1 ) consumption (s -1 ) formation (s -1 ) Native 32.8 (1.1) 0.4 (0.03) 0.05 (0.00) ND Native + CaM 32.5 (0.7) 1.6 (0.6) 0.27 (0.01) dFMN 28.2 (0.4) ND (0.001) ND (0.002) 5-dFMN nnos + CaM 28.2 (0.4) ND (0.001) ND Table S3.1. Steady state k cat values for nnos turnover assays. All data are recorded at 10 o C. *cyt c was shown to react with NADPH without nnos present, so steady-state cyt c reduction rates were calculated by subtracting rates of solution reaction from those with enzyme present. Estimated errors are given in parenthesis. ND, not detected. 100

101 3. REDOX-LINKED DOMAIN DYNAMICS AND NOS Static Fluorescence Measurements. Figure S3.2. Fluorescence excitation and emission spectra of oxidized nnos. Excitation wavelengths used in fluorescence experiments (555 and 645 nm) to monitor nnos-bound CaM dynamics are red shifted from intrinsic nnos flavin chromophore excitation spectra (black). nnos emission was recorded with a 450 nm excitation wavelength while excitation data was recorded with a 525 nm emission wavelength. All 1 μm nnos samples were in 40 mm HEPES (ph 7.6) supplemented with 150 mm NaCl, 1 mm CaCl 2, 10 % glycerol in the presence of 5 µm H 4 B. Data are normalized to λ max values for both excitation and emission spectra. The appearance of nnos-blue semiquinone (which has absorbance features between 550 and 650 nm) is minimal during single-turnover stopped-flow measurements of nnos flavin reduction thus there are no complications from direct excitation of nnos semiquinone flavin species [191]. 101

102 3. REDOX-LINKED DOMAIN DYNAMICS AND NOS Figure S3.3. Fluorescence emission changes associated with the binding of (A) 0.3 µm D-T34C/T110C-CaM or (B) 0.3 µm A-T34C/T110C-CaM (black) with Ca 2+ (red) or both Ca 2+ and 0.3 µm nnos (blue). Data show both a quenching in the emission as well as a ~3 nm red-shift in the λ-maxima of both donor and acceptor fluorophores when fluorophore labeled T34C/T110C-CaM binds to nnos in the presence of Ca 2+ relative to the Ca 2+ free or Ca 2+ bound fluorophore labeled T34C/T110-CaM. D-T34C/T110C- and A-T34C/T110C-CaM samples were excited at 555 nm and 645 nm, respectively. Data are plotted relative to the emission maxima of the D/A-T34C/T110C CaM only samples (black) which was normalized to 100 %. All samples were in 40 mm HEPES (ph 7.6) supplemented with 150 mm NaCl, 1 mm CaCl 2, 10 % glycerol in the presence of 5 µm H 4 B. 102

103 3. REDOX-LINKED DOMAIN DYNAMICS AND NOS Figure S3.4. Fluorescence emission spectra of donor and acceptor fluorophores bound to T34C/T110C-CaM. (A), (B) and (C) are representative of apo-cam, Ca 2+ -bound CaM and nnos with equimolar Ca 2+ -bound CaM, respectively. Black is 0.15 μm donor only labeled sample (D-T34C/T110C-CaM), red is 0.3 μm donoracceptor labeled CaM (DA-T34C/T110C-CaM) and blue is 0.15 μm acceptor labeled CaM (A-T34C/T110C- CaM). In (C) the magenta line is 0.3 μm nnos bound to 0.3 μm unlabeled T34C/T110C-CaM, excited at 555 nm. All data are plotted relative to the emission maxima (~ 570 nm) of the donor only sample (black). All data were recorded using a 555 nm excitation wavelength and all samples were in 40 mm HEPES (ph 7.6), 150 mm NaCl, 1 mm CaCl 2, 10 % glycerol in the presence of 5 µm H 4 B. 103

104 3. REDOX-LINKED DOMAIN DYNAMICS AND NOS Figure S3.5. No Inter-CaM FRET across the nnos dimer. The black spectrum in (A) shows the fluorescence emission from 0.15 µm of D-T34C/T110C CaM and 0.15 µm A-T34C/T110C CaM bound to 0.3 µm oxidized nnos (black). The red spectrum in (A) show the emission from 0.15 µm single labeled D-T34C/T110C CaM bound to one equivalent of nnos (dotted) as well as the emission spectrum of 0.15 µm A-T34C/T110C sample bound to equimolar nnos (dash red). The sum of the two aforementioned spectra in (A) is shown as a solid red line. All data in (A) are normalized to the concentration of donor fluorophore in the D-T34C/T110C CaM bound nnos containing sample (dotted red line). (B) shows the difference in relative fluorescence emission between the solid red and black spectra shown in (A) (red spectrum black spectrum) to demonstrate that with using the A555-A647 fluorophore pair (R 0 = 47 Å) there is no recordable FRET between the two calmodulins bound to the nnos dimer. All data were recorded using a 555 nm excitation and all samples were in 40 mm HEPES (ph 7.6), 150 mm NaCl, 1 mm CaCl 2, 10 % glycerol in the presence of 5 µm H 4 B. 104

105 3. REDOX-LINKED DOMAIN DYNAMICS AND NOS Stopped-flow UV-Vis Measurements. Figure S3.6. Absorbance changes, recorded by a photodiode array (PDA) stopped-flow instrument, 10 s after mixing of 5 μm (final concentration) of (A) native or (B) 5-dFMN-reconstituted NOS with 20-fold excess NADPH in the presence (red) and absence (black) of CaM. All experiments were conducted at 10 o C in 40 mm HEPES (ph 7.6) supplemented with 150 mm NaCl, 1 mm CaCl 2, 10 % glycerol with 1 mm CaCl

106 3. REDOX-LINKED DOMAIN DYNAMICS AND NOS NOS NADPH k 1 k 2 k 3 k 4 Native pro-r 8.9 (2.8) 7.6 (5) 2.7 (1.1) 2.3 (1.4) NADP 2 H Native pro-r 4.3 (0.6) 2.4 (4.6) 6 (8.5) 1.7 (1.3) +CaM NADP 2 H Native pro-s 1.1 (0.6) 1.1 (1.1) 1.4 (0.6) 1.8 (1.0) NADP 2 H Native +CaM pro-s NADP 2 H 1.4 (0.2) 2.1 (4.1) 2.7 (3.8) 1.1 (1.0) Table S3.2. Observed KIE values for NADPH-driven nnos flavin reduction. Estimated errors are given in parenthesis. 106

107 3. REDOX-LINKED DOMAIN DYNAMICS AND NOS Stopped-flow FRET measurements. Figure S3.7. Time-dependent changes in fluorescence emission of fluorophores bound to the 0.3 μm singlelabeled (donor or acceptor labeled - black) or 0.3 μm double-labeled (donor and acceptor labeled red) T34C/T110C CaM-oxidized nnos complex upon reduction with excess NADPH (100 μm, final concentration). No emission changes were recorded in the dead-time of the stopped-flow instrument and all data were normalized to percentage of emission change, relative to the emission at seconds. The difference between the fluorophore emissions in the single- (donor or acceptor labeled T34C/T110C CaM) and doublelabeled samples (donor and acceptor labeled T34C/T110C CaM) were normalized to percentage emission change at time seconds and is shown as the blue transient. Donor emission changes are monitored using a 600 nm (+/- 10 nm) band width pass (BWP) with a 555 nm excitation wavelength and are shown in panel (A). Acceptor emission changes are followed using a 650 nm long-pass (cut-on) filter, data collected for the acceptor only single labeled sample (A-T34C/T110C CaM-nNOS) were collected using a 645 nm excitation wavelength while the emission changes for the acceptor in the double labeled sample were collected with a 555 nm excitation wavelength. All fluorescence changes associated with the acceptor are shown in panel (B). All experiments were conducted in 40 mm HEPES (ph 7.6), 150 mm NaCl, 1 mm CaCl 2, 10 % glycerol with 5 μm H 4 B. Transients shown here are averages of 8 traces. 107

108 3. REDOX-LINKED DOMAIN DYNAMICS AND NOS Figure S3.8. Time-dependent changes in fluorescence emission of fluorophores bound to the 0.3 μm singlelabeled (donor or acceptor labeled - black) or 0.3 μm double-labeled (donor and acceptor labeled red) T34C/T110C CaM-oxidized 5-dFMN nnos complex upon reduction with excess NADPH (100 μm, final concentration). No emission changes were recorded in the dead-time of the stopped-flow instrument and all data were normalized to percentage of emission change, relative to the emission at seconds. The difference between the fluorophore emissions in the single- (donor or acceptor labeled T34C/T110C CaM) and double-labeled samples (donor and acceptor labeled T34C/T110C CaM) were normalized to percentage emission change at time seconds and is shown as the blue transient. Donor emission changes are monitored using a 600 nm (+/- 10 nm) band width pass (BWP) with a 555 nm excitation wavelength and are shown in panel (A). Acceptor emission changes are followed using a 650 nm long-pass (cut-on) filter, data collected for the acceptor only single labeled sample (A-T34C/T110C CaM-5-dFMN nnos) were collected using a 645 nm excitation wavelength while the emission changes for the acceptor in the double labeled sample were collected with a 555 nm excitation wavelength. All fluorescence changes associated with the acceptor are shown in panel (B). All experiments were conducted in 40 mm HEPES (ph 7.6), 150 mm NaCl, 1 mm CaCl 2, 10 % glycerol with 5 μm H 4 B. Transients show here are averages of 8 traces. 108

109 3. REDOX-LINKED DOMAIN DYNAMICS AND NOS NOS k 1 (s -1 ) ΔC 1 k 2 (s -1 ) ΔC 2 k 3 (s -1 ) ΔC 3 k 4 (s -1 ) ΔC 4 y0 native SL Donor 504 (86) 9.8 (3.7) 11.7 (1.3) -4.9 (0.8) 0.19 (0.03) 2.8 (0.3) (0.001) 7.3 (0.1) 91.3 (0.8) DL Donor 663 (262) 3.5 (2.3) 11.6 (1.8) -3.5 (0.7) 0.19 (0.01) 1.7 (0.1) (0.001) 4.9 (0.1) 95.4 (0.9) Donor (FRET) 259 (125) -2.9 (0.7) 14.2 (5.2) 1.8 (1.0) 0.25 (0.16) -0.9 (0.5) (0.003) -2.5 (0.2) (1.3) SL Acceptor 438 (116) 9.5 (2.8) 17.2 (1.2) -5.1 (0.4) 0.13 (0.02) 1.0 (0.1) (0.001) 2.8 (0.2) 98.0 (1.0) DL Acceptor 417 (68) 21.6 (4.5) 15.8 (1.2) (1.2) 0.15 (0.01) 2.8 (0.3) (0.001) 7.7 (0.2) 92.1 (1.2) Acceptor (FRET) 297 (61) 8.1 (1.9) 15.1 (2.4) -6.4 (1.2) 0.17 (0.02) 1.7 (0.2) (0.002) 5.0 (0.3) 94.3 (2.3) 5- dfmn SL Donor 239 (68) 6.3 (1.3) 38.8 (6.0) -5.9 (2.0) 0.22 (0.1) 1.1 (0.2) (0.001) 9.2 (0.2) 90.3 (1.2) DL Donor 295 (116) 3.4 (1.1) 37.2(12.2) -3.7 (0.9) 0.61 (1.2) 0.7 (0.2) (0.001) 5.7 (0.2) 95.0 (1.2) Donor 240 (79) -2.7 (1.3) 28.9 (20.1) 1.7 (1.1) 1.13 (1.14) -0.4 (0.1) (0.004) -3.7 (0.2) (1.3) (FRET) SL Acceptor 201 (39) 5.7 (1.0) 31.6 (4.2) -6.1 (1.0) 0.04 (0.02) 1.2 (0.5) (0.002) 2.8 (0.4) 97.6 (0.9) DL Acceptor 189 (19) 15.7 (1.6) 32.5 (2.1) (1.6) 0.09 (0.04) 1.8 (0.6) (0.001) 9.8 (0.6) 90.6 (2.2) Acceptor (FRET) 186 (42) 10.1 (1.4) 33.4 (6.0) -8.8 (1.2) 0.18 (0.08) 0.9 (0.3) (0.002) 7.0 (0.4) 93.0 (1.9) Table S3.3. Donor and acceptor fluorophore emission changes extracted from fitting to transients seen in Figure 3.4A and 3.4C (deconvoluted FRET) and Figure S3.7 (native) and S3.8 (5-dFMN). Estimated errors are given in parenthesis. SL, single labeled; DL, double labeled 109

110 3. REDOX-LINKED DOMAIN DYNAMICS AND NOS Figure S3.9. Time-dependent changes in fluorescence emission of fluorophores bound to the 0.3 μm singlelabeled (donor or acceptor labeled - black) or 0.3 μm double-labeled (donor and acceptor labeled red) T34C/T110C CaM-oxidized nnos complex upon reduction with excess NADP + (500 μm, final concentration). No emission changes were recorded in the dead-time of the stopped-flow instrument and all data were normalized to percentage of emission change, relative to the emission at seconds after mixing. The difference between the fluorophore emissions in the single- (donor or acceptor labeled T34C/T110C CaM) and double-labeled samples (donor and acceptor labeled T34C/T110C CaM) were not calculated since little changes were observed in the single labeled samples (black) which are attributed to photo-bleaching (see results and discussion). Donor emission changes are monitored using a 600 nm (+/- 10 nm) band width pass (BWP) with a 555 nm excitation wavelength and are shown in panel (A). Acceptor emission changes shown in (B) were followed using a 650 nm long-pass (cut-on) filter, data collected for the acceptor only single labeled sample (A-T34C/T110C CaM-nNOS) were collected using a 645 nm excitation wavelength while the emission changes for the acceptor in the double labeled sample were collected with a 555 nm excitation wavelength. All fluorescence changes associated with the acceptor are shown in panel (B). All experiments were conducted in 40 mm HEPES (ph 7.6), 150 mm NaCl, 1 mm CaCl 2, 10 % glycerol with 5 μm H 4 B. Transients show here are averages of 8 traces. 110

111 3. REDOX-LINKED DOMAIN DYNAMICS AND NOS k 1 (s -1 ) ΔC 1 k 2 (s -1 ) ΔC 2 y0 SL D ND ND ND ND ND DL D 1.24 (1.2) -0.7 (0.2) (0.003) -3.9 (0.7) (1.1) Donor (FRET) 1.24 (1.2) -0.7 (0.2) (0.003) -3.9 (0.7) (1.1) SL A ND ND ND ND ND DL A 0.53 (0.97) 1.7 (1.8) (0.057) 1.8 (1.1) 97.8 (2.5) Acceptor (FRET) 0.53 (0.97) 1.7 (1.8) (0.057) 1.8 (1.1) 97.8 (2.5) Donor/Acceptor 0.43 (0.56) (0.012) (0.011) (0.013) 1.10 (0.02) Table S3.4. Donor and acceptor fluorophore emission changes extracted from fitting exponential functions to transients seen in Figure 3.6 and Figure S3.9. Estimated errors are given in parenthesis. SL, single labeled; DL, double labeled. ND, not determined. 111

112 3. REDOX-LINKED DOMAIN DYNAMICS AND NOS Circular dichroism (CD) measurements. Figure S3.10. The folding of native and 5-dFMN nnos is identical. (A) SDS-PAGE gel of native and 5-dFMN nnos showing no alteration (proteolysis) in nnos monomer after unfolding treatment with potassium bromide and subsequent binding of 5-dFMN binding (see Supplementary Experimental Section). (B) Circular dichroism of native (black) and 5-dFMN nnos (red) showing secondary structure is identical between the two enzymes. CD experiments were performed on 5 μm of both native and 5-dFMN nnos in 40 mm HEPES (ph 7.6) supplemented with 150 mm NaCl and 10 % glycerol. 112

113 3. REDOX-LINKED DOMAIN DYNAMICS AND NOS Absorbance measurements. Figure S3.11. Reductive titration of native and 5-dFMN reconstituted NOS. NOS (~3 µm) was anaerobically titrated with NADPH (~1 µm aliquots) and dithionite (DT) in the presence of 5 mm L-Arginine and 5 µm H 4 B in 40 mm HEPES (ph 7.6) buffer supplemented with 10% glycerol and 150 mm NaCl. Spectra were taken after each addition. (A) Native NOS titrated with NADPH, (B) 5-dFMN NOS titrated with NADPH, (C) Native NOS titrated with dithionite, and (D) 5-dFMN substituted NOS titrated with DT. The insets show the observed difference spectra upon addition of reductant. Spectra were recorded at ambient room temperature on a Cary UV-Vis spectrometer with μm of flavin samples. 113

114 3. REDOX-LINKED DOMAIN DYNAMICS AND NOS Figure S3.12. (A) Structures of FMN (black) and 5-dFMN (red). (B) The absorbance spectra of flavin mononucleotide (FMN; black) and 5-deazaflavin mononucleotide (5-dFMN; red) in 40 mm HEPES (ph 7.6) supplemented with 150 mm NaCl and 10 % glycerol. Spectra were recorded at ambient room temperature on a Cary UV-Vis spectrometer with μm of flavin samples. Spectra in (B) are normalized to molar extinction coefficient. 114

115 4. REVIEW: PROTEIN DYNAMICS AND NOS CATALYSIS 4. A perspective on conformational control of electron transfer in nitric oxide synthases. Published in: Journal of Nitric Oxide. First Published: In press on 9 th September 2016 Authors: Tobias M. Hedison, Sam Hay, Nigel S. Scrutton. DOI: /j.niox Running Header: Protein dynamics and NOS catalysis. Contributions: TMH wrote the paper and conducted all experiments with help from NSS and SH. 115

116 4. REVIEW: PROTEIN DYNAMICS AND NOS CATALYSIS 4.1 Abstract. This perspective reviews single molecule and ensemble fluorescence spectroscopy studies of the three tissue specific nitric oxide synthase (NOS) isoenzymes and the related diflavin oxidoreductase cytochrome P450 reductase. The focus is on the role of protein dynamics and the protein conformational landscape and we discuss how recent fluorescence-based studies have helped in illustrating how the nature of the NOS conformational landscape relates to enzyme turnover and catalysis. KEYWORDS: nitric oxide synthase, cytochrome P450 reductase, diflavin oxidoreductase, fluorescence, protein dynamics. 4.2 Introduction. Many experimental and computational techniques have shown the importance of protein conformational change in both cell signalling [41-44] and enzyme catalysis [19, 23, 24, 39, 40]. It is now commonly believed that X-ray crystallography-derived structural data, although valuable, is insufficient in describing the function of many proteins. This has led to the notion that the structure-function dogma should be expanded to encompass protein dynamics (structuredynamics-function relationship) [35]. The dynamic profile of a protein can be thought of as a multidimensional conformational landscape, which comprises of hill and valley features representing high and low energy protein sub-states, respectively (Figure 4.1) [35, 280]. These population sub-states can be easily perturbed, but also voluntarily controlled, by mutagenesis, temperature, pressure, protein-protein interaction, redox chemistry and ligand/inhibitor binding [35]. The field of protein dynamics is gaining increasing attention, as greater insight into protein function is required, in order to be able to develop target-directed pharmaceuticals [45, 248] and to rationally design enzymes for bio-catalytic purposes [5, 247]. However, the study of conformational changes associated with enzyme turnover is challenging, as dynamics occur over a broad range of time and distance scales, from sub-ångstrom localised vibrations (femtoseconds) to large domain reorganisation (seconds) [20, 24]. Nitric oxide synthase (NOS) is proposed to make defined conformational changes during catalysis. NOS produces the small molecule nitric oxide (NO), which has a broad range of physiological roles from vasodilation to neurotransmission [281]. The three tissue specific NOS isoenzymes are homodimers that function by transferring electrons, which originate from NADPH, to the catalytic NOS haem porphyrin centre. Each NOS monomer is made up of i) a C-terminal reductase domain, 116

117 4. REVIEW: PROTEIN DYNAMICS AND NOS CATALYSIS comprising discrete FAD/NADP(H) and FMN binding domains, ii) an N-terminal oxygenase domain, which contains a tightly bound haem B and a tetrahydrobiopterin (H 4 B) molecule and iii) a binding site for calmodulin (CaM), which links the reductase and oxygenase domains (Figure 4.2). Figure 4.1. Simplified two dimensional depiction of a multidimensional conformational landscape of a protein molecule. The valley features of the landscape represent (quasi)stable conformational states that interconvert via higher energy barriers or hills [35, 280]. Temperature, pressure, mutagenesis, protein-protein interactions, reaction chemistry and ligand/inhibitor binding all influence the conformational landscape of a protein molecule (see text). NOS is believed to shuttle between the so called input and output conformational states, which have different orientations of the reductase and oxygenase domains [170]. These two states are thought to be functionally relevant in gating the precise flux of electrons from NADPH to the catalytic haem centre. In recent years there has been a variety of spectroscopic approaches used to probe NOS conformational change. Much of these data are summarised in recently published review articles on the general biochemistry and biophysics of NOS [33, 67, 68, 152, 190, 282, 283]. Herein, as an alternative perspective, we present how recently published fluorescence spectroscopic data have helped in illustrating the nature of the NOS conformational landscape related to catalysis. We offer an overview of the information gathered from single molecule and ensemble fluorescence spectroscopy studies for the three tissue specific NOS isoenzymes and also for the related diflavin oxidoreductase cytochrome P450 reductase (CPR). 4.3 Structure of the NOS isoenzymes. Despite the lack of an atomistic structure of full length NOS, the X-ray crystal structures of the individual NOS domains [ ], along with recently published cryo-em data [ ], have helped to illustrate the structural organisation of this enzyme (Figure 4.2). The three tissuespecific NOS isoenzymes are homodimeric proteins which bind and are activated by CaM. Each NOS reductase domain contains distinct FAD and FMN binding subdomains (Figure 4.2A) and this 117

118 4. REVIEW: PROTEIN DYNAMICS AND NOS CATALYSIS domain is homologous to CPR, a microsomal membrane-bound diflavin oxidoreductase, which transfers electrons to a multitude of partner proteins, e.g. cognate cytochrome P450 enzymes (CYPs) [50, 249]. Other members of the diflavin oxidoreductase family include methionine synthase reductase (MSR), a key enzyme in folate and methionine metabolism [242] and the reductase domain of the bacterial CYP cytochrome P450 BM3 (P450 BM3) [71]. Figure 4.2. Structure and molecular architecture of NOS. A) The structure of the NADP + -bound neuronal NOS diflavin reductase domain (PDB ID 1TLL). B) The structure of holo-cam bound to a nnos-peptide (PDB ID 2O60). C) The crystal structure of the nnos oxygenase domain dimer (PDB ID 1ZVL). CaM binds between the oxygenase and reductase domains of NOS. D) Structural organisation of the functional NOS dimer. The NOS FAD and FMN binding domains are shown in dark blue and light blue, respectively. CaM is shown in grey and the NOS oxygenase domain is shown in red. The NOS reductase and oxygenase domains are connected by a CaM binding site. All three NOS isoforms bind to CaM, which is essential for NOS catalysis. However, the binding between both the constitutive NOS (cnos) proteins [neuronal NOS (nnos) and endothelial NOS (enos)] and CaM is Ca 2+ dependent and reversible, while inducible NOS (inos)-cam interactions occur regardless of intracellular calcium concentrations [152]. Many spectroscopic studies have shown the flux of electrons in NOS isoenzymes occurs from NADPH, through FAD and FMN cofactors, to the catalytic haem centre where NO is produced (see below). However, based on crystallographic data of the isolated NOS reductase domain in 118

119 4. REVIEW: PROTEIN DYNAMICS AND NOS CATALYSIS complex with NADP +, where the FAD and FMN cofactors are in proximity, the electron transfer from reductase to oxygenase domains is thought not to be possible, due to the occluded location of the FMN [170]. This structure suggests that a large-scale shuttling of the NOS FMN domain (over ~70 Å), between the FAD and haem domains, is required for intra-nos electron transfer. These types of large-scale domain dynamics have been observed at atomistic detail in variants of CPR using X-ray crystallographic techniques [84, 105, 109]. Moreover, this conformational control has recently been detected also in NOS using cryo-electron microscopy (cryo-em), which has provided low resolution information on the molecular architecture of the protein [ ]. 4.4 Reaction mechanism of NOS. Due to the characteristic flavin and haem absorbance features, many UV-Vis based spectroscopic approaches, from conventional stopped-flow techniques [186, 191, 192, 284] to novel laser flash photolysis [ ], have been used to probe NOS and related diflavin oxidoreductase redox chemistry. Flavin reduction is typically monitored in the stopped-flow by rapidly mixing the enzyme with NADPH and following the quenching of the ~450 nm oxidised flavin feature or the growth and decay of the semiquinone species [82, 191, 284]. Due to the multiphasic nature of the stopped-flow traces for NOS reduction, the assignment of chemical steps to individual phases observed in the stopped-flow has not been possible [191, 284]. However, Scheme 4.1 shows the simplified mechanism for NADPH-driven flavin reduction, which has been advanced for NOS [191, 284] and is conceptually similar to that described for other related diflavin oxidoreductases [82]. Upon binding of NADPH to NOS, a hydride anion is transferred from the C4 position of the coenzyme to the N5 position of the oxidised NOS bound FAD cofactor. Following FAD reduction, reducing equivalents are transferred from the FAD hydroquinone to FMN, yielding the so called quasi-equilibrium (QE) state - a thermodynamic equilibrium where electrons partition between the FAD and FMN cofactors. In the QE state, three predominant redox states are present (FAD hydroquinone and FMN oxidised; FAD semiquinone and FMN semiquinone; FAD oxidised and FMN hydroquinone). Following QE state formation, a second NADPH coenzyme binds to NOS and drives the equilibrium towards the FAD and FMN hydroquinone species. The role of CaM during flavin reduction has been investigated. In full length nnos [284], and in studies of isolated nnos reductase domain [191], the binding of CaM has only small/marginal effects on the observed rates of interflavin electron transfer. However, CaM binding does alter the mid-point potential of NOS bound flavin [199], which will perturb the QE state and likely contributes to stimulated rates of cytochrome c reduction in steady-state assays in the presence of CaM [186, 191, 192, 284]. 119

120 4. REVIEW: PROTEIN DYNAMICS AND NOS CATALYSIS Scheme 4.1. Simplified reductive half reaction of nitric oxide synthase and related diflavin oxidoreductases (see text for more details). The chemistry catalysed by diflavin oxidoreductases diverges to some extent after FMN reduction. CPR and MSR transfer reducing equivalents from FMN to their respective partner proteins, while NOS and cytochrome P450 BM3 (P450 BM3) shuttle electrons from reduced FMN to their haem binding oxygenase domain. For NOS catalysis, the transfer of electrons from reduced FMN to haem is thought to be cross-monomer [200, 201] and requires the presence of CaM [202]. This FMN to haem electron transfer is difficult to study by stopped-flow based methods due to the slow rate of electron delivery along with the complexity of transients recorded. Thus, novel laser flash photolysis approaches, which can rapidly inject electrons into complex redox systems, have been used to track NOS FMN to haem electron transfer. In particular, studies of FMN to haem electron transfer using laser flash photolysis of carbon monoxide dissociation on partially reduced NOS have been used to access the dynamics and the chemistry catalysed by the enzyme. These measurements were initially performed on the truncated NOSoxyFMN construct [204, 205] and then subsequently on the full length NOS enzyme [203, 206, 207]. Discrepancies between the rates observed between these two forms of NOS constructs demonstrated that shuttling between NOS input and output states (dynamic interconversion) is rate limiting in NOS catalysis, providing evidence for a role of protein dynamics in gating NOS catalysis. After sequential electron transfer from FMN to haem, the signalling molecule NO along with the amino acid L-Citrulline is produced at the NOS oxygenase domain, from L-Arginine and molecular oxygen (Scheme 4.2) [285]. 120

121 4. REVIEW: PROTEIN DYNAMICS AND NOS CATALYSIS Scheme 4.2. The sequential oxidation of L-Arginine (L-Arg) into NO and L-Citrulline (L-Cit) catalysed by NOS. Initially L-Arg is oxidised to N-hydroxy-L-Arginine (NHA) which is further oxidised to L-Cit and NO. Electrons transfer from the NOS FMN hydroquinone to the NOS oxygenase domain is thought to be cross dimer and only occurs when CaM is bound. 4.5 Domain dynamics of nitric oxide synthase (NOS) Probing NOS dynamics using intrinsic flavin fluorescence. Using fluorescence to probe changes in the microenvironment of flavin chromophores (FAD and FMN), which are non-covalently bound to NOS, has proven to be a robust method of detecting conformational change in diflavin oxidoreductases [148]. Gachhui et al [217] were the first to investigate NOS conformational change by observing FAD and FMN fluorescence steady state emission. The authors showed binding of Ca 2+ /CaM to the nnos reductase protein caused an increase in flavin emission at ~530 nm, when absorbance features of isoalloxazines were excited [217]. These CaM-dependent alterations in flavin emission have been observed in a range of experiments and are attributed to an increase in solvent exposure of the NOS FMN cofactor upon binding to CaM [199, 218, 219]. NADP + binding has the opposite effect on NOS flavin fluorescence [199]. These data suggest that CaM binding promotes conformational freedom of the NOS FMN domain, while coenzyme binding hinders this mobility. This basic idea is consistent with a number of reports that suggest an auto-inhibitory loop and a C-terminal tail region of NOS make contact with the NADP(H) binding site [ , 184, 186]. In cnos isoforms this interaction between 121

122 4. REVIEW: PROTEIN DYNAMICS AND NOS CATALYSIS coenzyme and the protein prevents movement of the FAD and FMN domains, fixing the protein in the input state. When CaM binds to cnos, this interaction is disrupted, enabling conformational freedom of the FMN domain. Overall, these steady-state fluorescence measurements have been useful in detecting how mutagenesis alters the conformational landscape of the enzyme, as well as showing that shuttling of the FMN domain between the input and output states limits catalysis (e.g. as demonstrated using solvent viscosity studies) [218, 219]. The study of NOS by flavin fluorescence lifetime decay measurements offers additional information on protein conformational change over steady state fluorescence emission measurements. Flavin lifetime measurements were first used to study the neuronal NOS (nnos) holoenzyme. These studies showed that the recorded flavin fluorescence decay fits to a multiexponential model, indicating the presence of multiple flavin microenviroments (conformations) which do not interconvert on the fluorescence timescale [220]. Both cofactors (H 4 B and haem) and CaM/Ca 2+ -dependent changes in the fluorescence lifetime decays were observed, giving clues into how the nnos oxygenase and reductase domains fold back onto one another and into the role of CaM in shuttling the FMN domain [220]. Cryo-EM structure of the full length enzyme [ ] has recently confirmed what was inferred from this fluorescence study thereby demonstrating the capabilities of fluorescence spectroscopy to detect dynamic changes. NOS fluorescence lifetimes were recently revisited by Salerno and co-workers [221, 222]. The authors probed the differences between the inos holoenzyme and the inosoxyfmn complex, allowing them to assign specific phases seen in the multiexponential fluorescence lifetime decay to certain conformational states related to NOS catalysis [222]. This was taken further in subsequent work published by the same authors on the nnos isoform [221]. Combined, these studies with inos and nnos showed the presence of three major conformational states of the NOS enzymes. Two of these states are the well-defined input (lifetime of ~100 ps) and output (lifetime of ~1 ns) states, where the FMN domain is closer to the FAD domain or haem domain, respectively. The third state detected was defined as an intermediate state ( open state with a lifetime of ~4.3 ns), where the FMN domain is distanced from the FAD and the haem domains. This open state was reported to be heterogeneous (tail of fluorescence decays could fit to both single and multiple exponential functions) [221] and similar to recently published cryo-em [ ] and EPR [198, 214, 215] studies. These studies therefore reveal the complex and rugged nature of the NOS conformational landscape, which is also altered by CaM binding. 122

123 4. REVIEW: PROTEIN DYNAMICS AND NOS CATALYSIS Probing NOS dynamics using external CaM-bound fluorophore fluorescence. The commercial availability of extrinsic cysteine and lysine-binding fluorophores has significantly contributed towards the understanding of protein conformational change. Labelling a protein with fluorophores in specific locations enables one to probe conformational change by a number of fluorescence based methods. However, unlike CPR [50, 249], NOS has multiple solvent exposed cysteine and lysine residues, some of which are crucial for structural integrity. Therefore, it is challenging to site-specifically label NOS with extrinsic fluorophores [223]. Alternatively, wild-type CaM does not have any innate cysteine residues, so site-specific labelling of CaM variants offers a tractable approach to the study of CaM-NOS dynamics by fluorescence spectroscopy. A number of well characterised CaM Cys knock-in variants have been created that can be labelled with extrinsic fluorophores [215, 259, 265, 270, 284, 286, 287]. These variants, which have no reported effect on NOS turnover, have been used to study the kinetics of CaM-NOS association as well as deciphering the antiparallel-orientation (N-terminal of NOS binds to C-terminal of CaM and vice versa) and the compact configuration (short distance between C- and N-terminal) that CaM adopts when bound to NOS enzymes [259, 287]. However, much of this work has focused on the use of truncated versions of NOS and there has been little focus on the interactions between full length NOS and CaM. In an attempt to assess any differences in the redox-dependent association of CaM and nnos, we have labelled one of these commonly used CaM variants (T34C- CaM) with the cysteine binding Alexa-647 maleimide fluorophore. In this study the T34C-CaM bound Alexa-647 maleimide fluorescence was seen to quench upon binding to nnos, likely due to Förster resonance energy transfer (FRET) from the fluorophore to chromophore(s) in nnos. Figure 4.3 shows the titrations of Alexa-647 labelled T34C-CaM with both oxidised and aerobic dithionite-reduced nnos. Due to the complex nature of the tight binding between CaM and nnos, as well as the presence of different NOS-CaM binding states (previously observed by EPR [215], fluorescence lifetime and single-molecule studies [286]), dissociation constants (K d ) values were not calculated for the CaM-nNOS interaction. However, by inspection of the two different CaM-nNOS titration curves, a clear variation can be observed. These data provide compelling evidence that redox-chemistry drives conformational changes in the nnos enzyme, which in turn has a knock-on effect on CaM binding. These redox-dependent conformational changes have been seen in the related diflavin oxidoreductase CPR. When CPR is reduced to two- or fourelectron reduced forms, the FAD and FMN containing domains move relative to one another, and similar behaviour could explain why the interactions between NOS and CaM are altered by redox chemistry [144, 146]. We have recently probed the redox-dependent conformational changes of 123

124 4. REVIEW: PROTEIN DYNAMICS AND NOS CATALYSIS nnos-bound CaM by using a novel stopped-flow FRET technique [284], which was initially developed to track temporally-resolved changes in CPR domain organisation (see section below [50, 249]). We subsequently labelled a double cysteine-containing CaM variant (T34C/T110C-CaM) with both donor and acceptor fluorophores (Alexa-555 and Alexa-647). By monitoring timeresolved changes in fluorescence emission by stopped-flow spectroscopy, we were able to detect defined conformational changes in the CaM-NOS complex and showed that CaM dynamics are kinetically coupled to key mechanistic steps during nnos turnover. These two fluorescence studies show that redox chemistry, along with cofactor binding, drives conformational change in the NOS enzymes, and is essential for gating electron transfer from NADPH to the NOS catalytic haem centre. Figure 4.3. Redox-dependent binding of CaM to nnos. (A) shows the UV-visible absorbance spectra of oxidised (black) and dithionite-reduced (red) nnos (~5 µm). The dithionite reduced NOS was shown to be stable for over 2 hours under air and the absorbance is not perturbed by the addition of a 5-fold excess of T34C-CaM (data not shown). (B) shows the titration of 5 nm A647-T34C-CaM with oxidised (black) and dithionite-reduced (red) nnos monitored by quenching of the fluorescence emission from the A647 bound to CaM. All materials were of analytic grade and purchased from Sigma-Aldrich, except Alexa Fluor 647 C2 maleimide (A647)m which was purchased from Thermo fisher scientific. Both recombinant rat neuronal nitric oxide synthase and the T34C-CaM variant were expressed and purified as previously described [198, 226, 278, 284]. Purified nnos was oxidised with ferricyanide and passed down a desalting column to remove excess oxidising agent. To reduce the oxidised nnos, excess sodium dithionite mixed with the oxidised for of nnos, which was subsequently passed down a desalting column equilibrated with the buffering solution. A647 maleimide was bound to T34C-CaM in the dark using previously published protocols [284]. Fluorescence measurements were made with an Edinburgh Instruments (Livingston, U.K.) FLS920 fluorometer. Emission spectra were taken using 5 nm excitation and 5 nm emission slit-widths in 1 ml fluorescence quartz cells (Starna Scientific Ltd, Hainault, U.K.) with a 10 mm excitation path length. Data were collected at room temperature in 40 mm HEPES (ph 7.6), 150 mm NaCl, 1 mm CaCl 2 and 10 % glycerol. Under these conditions, no fluorophore photo-bleaching was observed 124

125 4. REVIEW: PROTEIN DYNAMICS AND NOS CATALYSIS Figure 4.4. nnos-bound CaM dynamics and reaction chemistry are kinetically coupled. The graph shows the transient reaction chemistry of nnos (relative absorbance, red) and dynamics of nnos-bound CaM (relative FRET efficiency, grey) recorded when mixing excess NADPH with nnos (bound to CaM), under pseudo first order conditions. The schematic presented here shows the temporal resolved dynamics of nnos-bound CaM when nnos is reduced with NADPH. This figure is adapted in part from ref. [284] Probing NOS dynamics using external NOS-bound fluorophore fluorescence. As discussed above, wild-type NOS domain dynamics are difficult to probe using FRET-based experiments due to the high number of potential fluorophore labelling sites. Recently, He et al. constructed a Cys-lite variant of the NOS reductase domain [223] that is unreactive towards maleimide fluorophores. This subsequently enabled these authors to further mutate the variant to allow the successful site-specifically labelling of NOS with fluorophores. Single molecule FRET measurements [223] were used to monitor the distances between the nnos FAD and FMN binding domains, allowing the conformational landscape of NOS to be probed. The main finding from this study is that CaM alters the distribution of NOS conformational states, increasing the distances between the FAD and FMN binding domains as well as increasing the rate of conversion between protein sub-states and narrowing the distribution of NOS conformers [223]. These alterations in the NOS conformational landscape apparently enables the enzyme to form more productive geometries, increasing the flux of electrons from NADPH to the haem catalytic centre, as is evident by the stimulation of NOS catalytic activity by CaM [186, 192, 284]. CPR has far fewer solvent exposed cysteine residues than NOS (3 cf. 14) and can be labelled with two extrinsic fluorophores [249]. Labelling of CPR with a mixture of donor and acceptor fluorophores allows the tracking of conformational changes that affect the FAD-FMN separation [50, 249] and we have used this approach to investigate single-turnover domain dynamics that occur during CPR catalysis. Coenzyme (NADP + ) binding causes CPR to adopt a closed 125

126 4. REVIEW: PROTEIN DYNAMICS AND NOS CATALYSIS conformation with shorter distances between the FAD and FMN cofactors [50] and stopped-flow experiments have illustrated that CPR undergoes defined conformational changes ( opening and closing ) during turnover, which appear to occur on the same timescale as redox chemistry (Figure 4.5) [50, 249]. These time-resolved domain dynamics of CPR had not been detected prior to these measurements, and may play a role in gating electron transfers in CPR as well as in the NOS isoenzymes. Figure 4.5. CPR domain dynamics and reaction chemistry are kinetically coupled. The graph shows the transient reaction chemistry (relative absorbance, red) and dynamics (relative FRET efficiency, blue) of CPR when mixing NADPH with the enzyme under pseudo first order conditions. The schematic shows the structures of the open (PDB ID 3ES9) and closed (PDB ID 1AMO) CPR structures with the FAD and FMN binding domains in dark blue and light blue, respectively. This figure is adapted in part from ref. [249]. 4.6 Conclusions Various studies of NOS and related diflavin reductases now give a largely consistent picture where these enzymes exist in multiple conformational states that can be described by a rugged conformation landscape, which is perturbed by CaM and coenzyme binding, flavin (and possibly haem) redox state and also solvent environment (viscosity, ionic strength, etc.). Stopped-flowbased FRET studies of CPR and CaM-nNOS have recently begun to show how enzyme chemistry (flavin reduction) maps onto conformational changes in these enzymes and now allows a mapping between conformational and catalytic landscapes in diflavin oxidoreductases. 126