Quantitation of Proteins and Antibodies in Serum by LC-MS/MS

Transcription

1 Quantitation of Proteins and Antibodies in Serum by LC-MS/MS Jim Walters, Ph.D. September, 20, 2017 Merck KGaA Darmstadt, Germany

2 The life science business of Merck KGaA, Darmstadt, Germany operates as MilliporeSigma in the U.S. and Canada.

3 JIM WALTERS, PH.D. APPLIED TECHNOLOGY SPECIALIST ANALYTICAL MILLIPORESIGMA Jim Walters is an Applied Analytical Technology Specialist at MilliporeSigma. This role was preceded by 10+ years as a Principal R&D Scientist in the Analytical R&D Department at MilliporeSigma in St. Louis, MO. He received his Ph.D. at the University of South Carolina in 2000 from the Biomedical Science program in the Department of Microbiology and Immunology. He completed a post-doctoral research fellowship in Biomedical Mass Spectrometry at the Washington University Department of Chemistry in St. Louis, MO under Professor Michael Gross. In 2006, Jim joined the Analytical R&D Department at MilliporeSigma in St. Louis. Jim and his colleagues in Analytical R&D were on the fore front of a stable-isotope-labeled intact protein program for use as internal standards in quantitative proteomic workflows. Jim has over 19 years of experience applying liquid chromatography-mass spectrometry to biotechnology and proteomics, commercializing numerous products at MilliporeSigma. In his current role, he supports Account Managers via interfacing with end-users to educate and inform, empowering them to select the best solutions for their analytical needs.

4

5 Outline Why Quantitative Mass spectrometry (MS) vs ELISA Quantitative MS workflow Why SIL proteins/abs as internal standards Expression and characterization of SIL proteins and antibodies Quantitative MS assays using SIL proteins and antibodies Protein Certified Reference Materials (Thyroglobulin)

6 Serum Protein Measurement Methods ELISA / LBA Pros: High sensitivity High throughput Traditional methodology Minimal sample prep LC-MS/MS Pros: Highly selective Faster assay development Ability to multiplex Can be combined with enrichment Cons: Assay specific reagents Long lead times Poor standardization Specificity concerns Difficult to multiplex Cons: Expensive instrumentation Extensive sample prep Requires an internal standard

7 General Design of an Internal Standard A compound purposely added to both samples and standards at a known and/or consistent concentration in order to provide a basis for comparison in quantitation and used for correction of potential variable effects along the workflow. Not be present in any of the samples Similar as possible in physiochemical properties to the target analyte Added as early on in the procedure as possible recovery during transfer and clean-up variability in extraction efficiency injection volume variability matrix suppression for LC-MS, preferably an isotopically labeled version of the analyte (SIL)

8 Typical Quantitative Protein/Ab Workflow by MS Protein Fractionation 1D or 2D Gels Abundant Protein Depletion Antibody Enrichment Sample Preparation Dissolution/ Denaturation Protein Extract Digestion Enzymatic Chemical Peptide Digest LC-MS Add SIL peptide Peptide Fractionation Anti-peptide antibodies (SISCAPA) Cation-exchange LC QQQ Mass Spectrometer SIL peptides are typically added late in the workflow

9 Typical Quantitative Protein/Ab Workflow by MS Protein Fractionation 1D or 2D Gels Abundant Protein Depletion Antibody Enrichment Sample Preparation Dissolution/ Denaturation Protein Extract Digestion Enzymatic Chemical Peptide Digest LC-MS Add SILuMab or SIL Protein Peptide Fractionation Anti-peptide antibodies (SISCAPA) Cation-exchange LC QQQ Mass Spectrometer SIL proteins are added as an initial step in the workflow

10 % Bias % CV Accuracy and Precision with Three SIL-IS s Hongyan Li et al; Anal. Chem. 2012, 84,

11 Desired SIL-Protein Properties High protein purity Matches native protein sequence High incorporation of stable isotopes Similar PTM to native protein (ie, glycosylation) Digestion kinetics same as native protein Similar enrichment or fractionation as native protein

12 SIL Protein and SILuMab Development Available Cell Lines CHO HEK E. coli

13 SIL-Protein Purity by SDS-PAGE Purity greater than 95% achieved

14 Characterization of SILUMab and a representative SiluPROT Sequence confirmation by peptide mapping SILuMab Heavy Chain Apolipoprotein A1 (APOA1) High sequence coverage obtained

15 RP-LC-MS Analysis of Intact SIL-IGF1 Sequence of IGF-1 consisting of 70 amino acids in a single chain and three intramolecular disulfide bonds Sequence and structure verified by RP-LC-UV-MS

16 Isotope Incorporation 13 C 6 15 N 2 Lys in SILuMab and 13 C 6 15 N 4 Arg in SIL-Thyroglobulin x 20 VVSVLTVLHQDQLNGK Heavy: [M+3H] Light: [M+3H] % Incorporation > 99% VIFDANAPVAVR Heavy: [M+2H] Light: [M+2H] % Incorporation = 98.2% Incorporation > 98%

17 SIL Protein Characterization: PTM s Glycosylation of SIL Thyroglobulin Glycosylation similar to native human protein

18 SIL Protein Digestion Kinetics APOA1 in Human Serum, TFE Denaturation Kinetics are similar after 4 hours of digestion

19 SIL Protein Digestion Kinetics APOA1 in Human Serum, Urea Denaturation (FASP) Kinetics are similar after denaturation

20 Desired SIL-Protein Properties High protein purity Matches native protein sequence High incorporation of stable isotopes Similar PTM to native protein (ie, glycosylation) Digestion kinetics same as native protein Similar enrichment or fractionation as native protein

21 Protein Internal Standard Workflow Protein Fractionation 1D or 2D Gels Abundant Protein Depletion Antibody Enrichment Sample Preparation Dissolution/ Denaturation Protein Extract Digestion Enzymatic Chemical Peptide Digest LC-MS Add SILuMab or SIL Protein Peptide Fractionation Anti-peptide antibodies (SISCAPA) Cation-exchange LC QQQ Mass Spectrometer SIL protein should normalize enrichment variability

22 SIL Protein Characterization: Ligand Binding Affinity SIL-Infliximab vs Remicade on Biacore TNF- sensorgrams SIL-Infliximab KD = 0.15 nm Remicade KD = 0.17 nm Ligand binding equivalent to therapeutic antibody

23 Desired SIL-Protein Properties High protein purity Matches native protein sequence High incorporation of stable isotopes Similar PTM to native protein (ie, glycosylation) Digestion kinetics same as native protein Similar enrichment or fractionation as native protein

24 Immuno-affinity enrichment LC-MS assay Erythropoietin (EPO) in dog serum 200 µl Serum 10 ng SIL-EPO Wash PBS In plate Digestion Trypsin LC-MS Anti hu-epo Capture Plate QQQ Mass Spectrometer Beagle serum with h-epo at 1 pg/ml to 10 ug/ml, 50 ng/ml SIL-EPO

25 Immuno-affinity enrichment LC-MS assay Erythropoietin (EPO) in dog serum Light Peptides Heavy Peptides Variable capture efficiency, saturation above 1 ug/ml

26 A v e ra g e A re a R a tio (L :H ) Immuno-affinity enrichment LC-MS assay Erythropoietin (EPO) in dog serum A r e a R a tio s SLTTLLR.y5 VYSNFLR.+2y5 VNFYAWK.+2y C o n c e n tra tio n E P O (lig h t) (p g /m L ) Response normalized by SIL protein ISTD

27 Comparison of Peptide and Protein Internal Standards Erythropoietin (EPO) in dog serum Bioshell Accuracy Table [EPO] ng/ml Protein IS Peptide IS VNFYAWK SLTTLLR VYSNFLR VNFYAWK SLTTLLR VYSNFLR Red highlight: >20% deviation from expected Greater accuracy achieved with SIL protein ISTD

28 Therapeutic Antibodies and How SIL mabs can help Therapeutic monoclonal antibodies (mabs) represent the majority of biotherapeutics now on the market as they are inherently more selective and less toxic. 68 antibody based therapeutics are FDA approved and >50 more in large clinical studies. Additionally, just last week the FDA approved the first antibody biosimilar for the treatment of cancer (Bevacizumab). Two motivations each with a different quantitative solution: Pharmacokinetic (PK) characteristics of therapeutic antibodies are assessed during preclinical development. Particularly during early preclinical animal studies wherein multiple structural variants of a drug candidate may need to be evaluated in various animal species to enable prioritization for further development a more universal approach required. Therapeutic drug monitoring - to guide dosing for favorable outcomes and limit toxicity - a specific approach required 28

29 Universal Peptide Strategy Quantification of Human MAb in Pre-Clinical Model Plasma Surrogate tryptic peptide from Fc region of human MAb Second tryptic peptide from light chain of human MAb Human Therapeutic MAb Native antibodies circulating in Animal Plasma/Serum Generalized preclinical PK assay employing surrogate peptides from constant regions

30 Immuno-affinity enrichment LC-MS PK assay Humira (adalimumab) in monkey serum 100 µl Serum Wash In plate Digestion Anti hu-fc Capture Plate 200 ng SILuMab PBS Trypsin LC-MS QQQ Mass Spectrometer Monkey serum with ADA at 10 ng/ml to 50 ug/ml, 2 ug/ml SILuMab

31 Human Mab in Monkey Serum Assay Statistics Accuracy: %, Precision: <10% Range: 100 ng/ml to 50 ug/ml

32 Therapeutic Drug Monitoring of Infliximab Surrogate peptides from Infliximabspecific variable regions Human serum with INFX at 100 ng/ml to 100 ug/ml, 25 ug/ml SILuMab

33 Human Therapeutic Mab in Drug Monkey Monitoring Serum of Infliximab Assay Statistics GLEWVAEIR (HC) Accuracy: %, Precision: <15% Range: 500 ng/ml to 100 ug/ml

34 SILuMab Standards SigmaAldrich.com/silutions

35 27 SILuProt Standards and unlabeled counterparts available: Product Name Cat. No. SILu Prot APOA1 Apolipoprotein A-I MSST0001 SILu Prot PTX3 Pentraxin-related protein MSST0003 SILu Prot VEGFA Vascular endothelial growth factor A MSST0005 SILu Prot CLU Clusterin MSST0007 SILu Prot MAPK1 Mitogen activated protein kinase 1 MSST0009 SILu Prot ALB Albumin MSST0011 SILu Prot AMBP Alpha-1 microglycoprotein MSST0013 SILu Prot B2M Beta-2-microglobulin MSST0015 SILu Prot IL6 Interleukin 6 MSST0017 SILu Prot MAPK3 Mitogen activated protein kinase 3 MSST0019 SILu Prot CRP C-reactive protein MSST0021 SILu Prot APOA2 Apolipoprotein A-II MSST0029 SILu Prot MAPT Microtubule-associated protein tau-441 MSST0031 SILu Prot IFNG Interferon Gamma MSST0039 SIL-Thyroglobulin Certified Reference Material T-109 SigmaAldrich.com/silutions

36 Protein Certified Reference Materials (CRM s) Reference material, which has been established to be fit for its intended use in a measurement process, and is characterized by a metrologically valid procedure for one or more specified properties, accompanied by a certificate that provides the value of the specified property, its associated uncertainty, and a statement of metrological traceability. all measurements made must have a directly traceable link back to the defined quantity

37 Protein CRM: Thyroglobulin a 660 kda, dimeric protein produced by the thyroid and used entirely within the thyroid gland. Thyroglobulin protein accounts for approximately half of the protein content of the thyroid gland. Tg is used by the thyroid gland to produce the thyroid hormones thyroxine (T4) and triiodothyronine (T3) Thyroglobulin levels in the blood are mainly used as a tumor marker for certain kinds of thyroid cancer. Following thyroidectomy, Tg levels are monitored. A subsequent elevation of the thyroglobulin level is an indication of recurrence of thyroid carcinoma.

38 Clinical Story for Tg Testing Generally determined by Immunoassay Presence of circulating autoantibodies interfere with IA 10% of healthy population 20% of thyroid cancer patients Resulted in LC-MS based lab developed tests Complex and more expensive Peptide based IS Calibrated using Tg reference material

39 Characterization of SIL Thyroglobulin Sequence Confirmation >95% coverage

40 Characterization of Thyroglobulin SIL Incorporation Peptide m/z XIC Area % Incorporation Heavy ,673,222 SQAIQVGTSWK > 98%* Light Not detected Heavy ,568,759 LALQFTTNPK > 98%* Light Not detected Heavy ,671,426 FLQGDHFGTSPR > 98%* Light Not detected SGNPNYPYEFSR Heavy ,853,909 > 98%* Light ,788

41 Concentration (µg/ml) Homogeneity and Stability assessments Statistical analysis of the production lot by LC-MS/MS Tg CRM E1 E2 M1 M2 L1 L2 rep rep rep average %CV 1.9% 9.8% 6.2% 1.1% 7.5% 10.7% Homogeneity_u = 1.53% Accelerated stability Weeks Real time stability in progress Refrigerate Room Temp 40 C Freeze/ Thaw %CV

42 Metrological Traceability Chain: Thyroglobulin value assignment Definition of unit of measure Primary reference measurement Primary calibrator Manufactures raw material Final product Secondary reference measurement Manufactures measurement CRM used as control End-users routine measurement on patient samples

43 AAA Workflow and Improvements Gravimetrically dispense samples Calibrated with NIST AA SRM Add IS Speed vac Hydrolyze to amino acids Speed vac -BSA (controls) -Samples Accutag label UPLC-UV analysis Gravimetric dilution of NIST BSA-SRM Weight corrected recovery and back calculations

44 Accuracy AND Precision BSA Results Expanded Uncertainity (95%) NIST BSA (927e) (mg/ml) Lower Limit Mean Upper Limit BSA measured by AAA at MilliporeSigma u = 0.2 STDdev/SQRT of n n = 114 U = 0.5 u * k K=2 Lower Limit Mean Upper Limit Expanded Uncertainity (95%) expressed as % recovery with an estiamted expanded uncertainity of 95%, n=114 u = 0.16 STDdev/SQRT of n n=114 U = 0.31 u * k k=2 Lower Limit Mean Upper Limit Expanded Uncertainity (95%) expressed as mg/ml with an estiamted expanded uncertainity of 95%, n=114 N = 114

45 TG reference material consensus based

46 AAA Thyroglobulin Experiments Assayed using different hydrolysis chambers Assayed using 4 different starting concentrations ( µg/ml) Repeat all steps above 2-4 times Perform multiple times by different scientist at different concentrations u = 0.4 STDdev/SQRT of n n = 69 U = 0.8 u * k K=2 Lower Limit Mean Upper Limit Expanded Uncertainity (95%) % recovery of TG-RM Only ~ 70% recovery!!!!

47 From BCR 457 Report g/l g/l Difference of g/l u = 0.4 STDdev/SQRT of n n = 69 U = 0.8 u * k K=2 Lower Limit Mean Upper Limit Expanded Uncertainity (95%) = g/l Back calculation of concentration

48 Thryoglobulin CRM Production Bulk Protein AAA Human source Prepare raw material stock solution Assign concentration with Amino Acid Analysis (AAA) NIST traceable -SRM-2389a, SRM-927 AAA Grav. Dil. Gravimetric dilution of AAA assigned Stock Traceable to SI Diluent - no endogenous Tg Disp. Seal Dispense final solution into ampoules and flame-seal under inert atmosphere Control temperature exposure Verif. Concentration and homogeneity LC-MS/MS Traceable to IRMM BCR -457 T 109 and T-113: 10µg/mL in SigMatrix TM

49 SIL-Tg Certificate of Analysis

50 Protein vs Peptide Internal Standards: General Design Samples equal one of three different sources of Tg (reference standard, commercially available natural source or recombinant Tg) spiked into either neat or one of three serum matricies. SIL-rTg introduced into each sample prior to denaturation, then the csil mixture (8 peptides) was added just prior to digestion, and the tsil mixture (8 peptides) was added immediately following digestion. The potential calibrators or internal standards all carry different isotopic labels so that all can be spiked into the same sample. After LC-MS/MS analysis it becomes an intensive data analysis exercise to determine how well each standard performs as a calibrator or internal standard. All standards were quantified using the AAA method previously described and diluted and added to sample gravimetrically. Shuford, C.M. et al. Absolute Protein Quantification by Mass Spectrometry: Not as Simple as Advertised; Anal. Chem., 2017, 89 (14), pp

51 Protein vs Peptide Calibration Standards in neat matrix Absolute quantification of three neat human Tg sources (stg, ctg, rtg) using either tryptic SILs, cleavable SILs or SIL-rTg as the calibration standard. SIL-rTG performed superior to tryptic or cleavable peptides and were significantly less variable. The imprecision in the mean quantity determined across 24 measurements for each standard: SIL-rTG: ~11% csil: ~50% tsil: ~88% Control for digestion is key to reducing variance and optimal control and minimum variance is best obtained by controlling denaturation through use of SIL proteins. Shuford, C.M. et al. Absolute Protein Quantification by Mass Spectrometry: Not as Simple as Advertised; Anal. Chem., 2017, 89 (14), pp

52 Protein vs Peptide Internal Standards in neat matrix Absolute quantification of two neat human Tg sources (stg and ctg) by using rtg as the external calibration standard and either tryptic SILs, cleavable SILs or SIL-rTg as the internal standard. In this experiment all 3 forms of ISs may be equally effective in achieving absolute quantification, at least in the absence of matrix effects However, these peptides were chosen to monitor (and synthesized) based on several criteria that made them amenable to the experimental workflow. If starting assay development without any preliminary empirical data, likely only SIL-rTG IS would perform consistently and allow for streamlining assay development. Shuford, C.M. et al. Absolute Protein Quantification by Mass Spectrometry: Not as Simple as Advertised; Anal. Chem., 2017, 89 (14), pp

53 Protein vs Peptide IS s: 39 Peptide experiment in neat matrix Absolute quantification of neat human Tg sources by using rtg as the external calibration standard and SILrTg as the IS for 39 signature peptides. Although some minor bias is observed when quantifying the human source Tg using rtg as the external calibrator and SIL-rTG as the internal standard this is an efficient way to screen peptides during method development to help choose peptides that do not show bias due to minor digestion efficiency differences. Shuford, C.M. et al. Absolute Protein Quantification by Mass Spectrometry: Not as Simple as Advertised; Anal. Chem., 2017, 89 (14), pp

54 Protein vs Peptide Internal Standards in serum matrix Absolute quantification of two human-derived Tg sources spiked into 3 different serum matrices using rtg as the external calibrator and either tsils, csils, or SIL-rTg as the IS. Quantification of the FSP peptide displayed a matrix effect with chicken and human serum which was also dependent on the denaturing condition used when either tsil or csil was used as an internal standard. SIL-rTg as the internal standard, any matrix effect was normalized for the difference in digestion efficiency between matrices. Additionally the VIF did not experience these same matrix effects. Shuford, C.M. et al. Absolute Protein Quantification by Mass Spectrometry: Not as Simple as Advertised; Anal. Chem., 2017, 89 (14), pp

55 Summary Produced a pipeline of highly characterized SIL proteins (and counterparts with naturally occurring isotopic distribution) and antibodies produced in human and e. coli cells and characterized for quantitative MS applications Use of SIL proteins and SILuMab standards reduces error and variability associated with enrichment and enzymatic digestion and can address the shortcomings of LBA s and peptide IS s associated with long assay development times SIL Proteins/Ab standards are the only form of IS likely to normalize for digestion-based matrix effect and to normalize for protein-enrichment variance SIL proteins/ab standards have the added advantage that they can be added at the first step in sample preparation and, thereby, should reduce imprecision relative to other forms of internal standardization added later in the workflow (e.g. surrogate synthetic peptides) Our growing pipeline of protein CRM s provide an extremely accurate value assignment with associated uncertainty and a C of A required for more stringent regulatory requirements

56 Acknowledgements Lab Corp Chris Shuford Russell Grant Patricia Holland Capabilities include: Analytical R&D Team Kevin Ray Pegah Jalili Judy Cao Gordon Nicol Mark Angeles Jim Blasberg Specific applications Custom sample prep products Partnerships / collaborations Product samples Method development Product recommendations Bundle workflow products SigmaAldrich.com