Targeted mass spectrometry Marina Zajec Dept. of Neurology and - PowerPoint PPT Presentation

Targeted mass spectrometry Marina Zajec Dept. of Neurology and Clinical Chemistry Lab. of Neuro-Oncology/Clinical&Cancer Proteomics

Outline  Introduction to targeted mass spectrometry • When to use targeted mass spectrometry? • What is required for a targeted mass spectrometry experiment? • Selected Reaction Monitoring vs. Parallel Reaction Monitoring  Examples with real data 1. Quantitation of low levels HSP90α by Parallel Reaction Monitoring 2. Development of targeted mass spectrometry assay to detect M-protein in multiple myeloma patient serum

LC-MS/MS proteomics strategies Shotgun proteomics - discovery Targeted proteomics  Proteins in the mixture are digested  Mass spectrometer is analyzing a and the resulting peptides are preselected group of proteins separated by liquid chromatography and analyzed by mass spectrometry  By use of internal standard  Spectra are generated from all quantitative values of proteins can be detectable proteins in a sample acquired  stable isotope labelled  Results are interpreted by database (SIL) reference spiked into the sample searching  Semi-quantitative analysis

When to use targeted approach?  when predetermined sets of proteins need to be measured across multiple samples in a consistent, reproducible and quantitatively precise manner Examples: Picotti P., Aebersold R. Nature, 2012.

What is required for a targeted proteomics experiment? 1. Protein(s) of interest, based on: Proteins of interest  Previous experiments (e.g. Shotgun proteomics)  Scientific literature  Prior knowledge Target peptides 2. Selection of the target peptides  Optimally represent the protein set – proteotypic peptides Targeted analysis

Target peptides  Measured as surrogates for proteins  Need to fulfill certain criteria: • Unique to the target protein – proteotypic peptides • No variable modifications (e.g. methionine present in the amino acid sequence) • No ragged ends in the sequence (KK, KR, RK, RR) • Optimal length: 7-15 amino acids

Target peptides - examples 1. KRNGGGR Ragged ends 2. RNGGGKK 3. LEPADFAVYYCQR 4. YGSSPLIFGGGTR OK 5. ASTLESGVPSR 6. FLIYK Too short/long 7. FSGSGSGTAFTLTISSLQPDDFATYYCQQYDSPPYTFGQGTK

Selected Reaction Monitoring (SRM) vs. Parallel Reaction Monitoring (PRM) Targeted proteomics SRM Triple quadrupole PRM Orbitrap (Fusion and Lumos) Coon et al. MCP, 2012.

Method optimization - SRM 1. Selection of optimal transitions  select the fragment ions for each precursor-ion charge state that provide the highest signal intensity and lowest level of interfering signals 2. Retention time assignment – scheduling 3. Collision energy optimization  maximize the SRM signal response for specific peptides or fragments

PRM compared to SRM  High specificity In PRM all product ions are monitored providing high confidence of peptide identification. The high resolution mass analyzer increases specificity (narrower mass window) compared to a SRM.  Reduced interference Compared to SRM, PRM provides data with high mass accuracy, which allows the removal of noise of interfering signals.  Reduced assay development time PRM-based targeted proteomics requires less effort in assay development compared to SRM as fragment ions can be selected post acquisition.

Examples with real data  Quantitation of low levels HSP90α by Parallel Reaction Monitoring • Comparing selectivity, sensitivity and repeatability of SRM, PRM, and immunoassay  Development of targeted mass spectrometry assay to detect M-protein in multiple myeloma patient serum • Clinical application of targeted mass spectrometry • Personalized proteomics

Example 1

HSP90 (low ng/mL level) quantification 43 control sera PRM SRM ELISA (2 microtiter plates) (2D-LC) (2D-LC) Triple quadrupole High resolution MS SRM by Xevo TQs PRM by Orbitrap Fusion anti-HSP90 Performed by C. Guzel

Selection of stable isotope labeled peptides for quantification MPEETQTQDQPMEEEEVETFAFQAEIAQLMSLIINTFYSNKEIFLRELISNSSDALDKIRYESLTDPSKLDSGKELHINLI PNKQDRTLTIVDTGIGMTKADLINNLGTIAKSGTKAFMEALQAGADISMIGQFGVGFYSAYLVAEKVTVITKHNDDE QYAWESSAGGSFTVRTDTGEPMGRGTKVILHLKEDQTEYLEERRIKEIVKKHSQFIGYPITLFVEKERDKEVSDDEAEE KEDKEEEKEKEEKESEDKPEIEDVGSDEEEEKKDGDKKKKKKIKEKYIDQEELNKTKPIWTRNPDDITNEEYGEFYKSLT Primary structure of NDWEDHLAVKHFSVEGQLEFRALLFVPRRAPFDLFENRKKKNNIKLYVRRVFIMDNCEELIPEYLNFIRGVVDSEDLP LNISREMLQQSKILKVIRKNLVKKCLELFTELAEDKENYKKFYEQFSKNIKLGIHEDSQNRKKLSELLRYYTSASGDEMV HSP90 α SLKDYCTRMKENQKHIYYITGETKDQVANSAFVERLRKHGLEVIYMIEPIDEYCVQQLKEFEGKTLVSVTKEGLELPED EEEKKKQEEKKTKFENLCKIMKDILEKKVEKVVVSNRLVTSPCCIVTSTYGWTANMERIMKAQALRDNSTMGYMAA KKHLEINPDHSIIETLRQKAEADKNDKSVKDLVILLYETALLSSGFSLEDPQTHANRIYRMIKLGLGIDEDDPTADDTSA AVTEEMPPLEGDDDTSRMEEVD  No known modifications or problematic cleavage sites  High quality stable isotope labeled peptides are required for correct quantitation Performed by C. Guzel

SRM/PRM assay 2D-LC approach (SCX prefractionation ) SCX-LC RP-LC n = 43 control sera SRM/PRM + stable isotopes (references) light heavy 10 SCX y t proteins i peptides mRP C18 s n e 5 t n i prefractionation 0 0 2 4 6 SpeedVac digestion concentration Calculation concentration (ratio heavy/light) 13 C/ 15 N label Based on 2 HSP90 α peptides: YIDQEELNK DQVANSAFVER Performed by C. Guzel

Targeted method  4 target HSP90 peptides (time scheduled): light heavy light heavy Performed by C. Guzel

SRM vs PRM on serum digest ∼ 60 ng/mL HSP90 Interfering peak SRM SRM heavy peptide light peptide PRM PRM heavy peptide light peptide • High resolution data obtained by PRM • No or less interfering peak detected by PRM Performed by C. Guzel

Distribution fragments obtained by SRM and PRM 100 Y ID Q E E L N K y 5 e n d o g e n o u s SRM Y ID Q E E L N K y 6 e n d o g e n o u s Y ID Q E E L N K y 7 e n d o g e n o u s p e rc e n ta g e (% ) peptide y5 y6 y7 YIDQEELNK 50 632.33 760.38 875.41 (endogenous) mean %CV 31.8 47.7 3.9 0 1 11 21 31 41 Pure compound (reference) s a m p le n o . 100 YIDQEELNK y5 endogenous PRM YIDQEELNK y6 endogenous YIDQEELNK y7 endogenous percentage (%) 50 peptide y5 y6 y7 YIDQEELNK 632.33 760.38 875.41 (endogenous) mean %CV 13.3 9.9 1.1 0 1 11 21 31 41 Pure compound (reference) sample no. Performed by C. Guzel

Distribution ratio’s of fragments obtained by SRM and PRM 1 0 0 D Q V A N S A F V E R y 7 e n d o g e n o u s SRM D Q V A N S A F V E R y 8 e n d o g e n o u s D Q V A N S A F V E R y 9 e n d o g e n o u s p e rc e n ta g e (% ) peptide y7 y8 y9 5 0 DQVANSAFVER 822.41 893.45 992.52 (endogenous) mean %CV 26.9 15.0 22.3 0 1 1 1 2 1 3 1 4 1 s a m p le n o . Pure compound (reference) 100 D Q V A N S A F V E R y 7 e n d o g e n o u s PRM D Q V A N S A F V E R y 8 e n d o g e n o u s D Q V A N S A F V E R y 9 e n d o g e n o u s p e rc e n ta g e (% ) 50 peptide y7 y8 y9 DQVANSAFVER 822.41 893.45 992.52 (endogenous) mean %CV 3.7 2.5 2.9 0 1 11 21 31 41 Pure compound (reference) Performed by C. Guzel s a m p le n o .

Comparison of HSP90 levels by SRM, PRM and ELISA SRM PRM DQVANSAFVER YIDQEELNK ELISA 5 0 0 5 0 0 4 0 0 4 0 0 H S P 9 0 (n g /m L ) H S P 9 0 (n g /m L ) 3 0 0 3 0 0 2 0 0 2 0 0 1 0 0 1 0 0 L L O Q S R M L O Q S R M 0 L L O Q P R M /E L IS A 0 L O Q P R M /EL ISA 0 1 0 2 0 3 0 4 0 0 1 0 2 0 3 0 4 0 sam p le no s a m p le n o . Peptide LOD (ng/mL) LLOQ (ng/mL) Peptide LOD (ng/mL) LLOQ (ng/mL) SRM 5.6 17.4 SRM 6.7 20.4 PRM 1.0 2.9 PRM 1.3 3.8 ELISA 0.4 1.2 ELISA 0.4 1.2 Performed by C. Guzel

Correlation SRM vs ELISA YIDQEELNK DQVANSAFVER C D 4 0 0 4 0 0 Y = 1.6*X - 3.8 Y = 0.799*X + 27.1 2 = c o n c e n tra tio n H S P 9 0 E L IS A (n g /m L ) 0.764 c o n c e n tra tio n H S P 9 0 E L IS A (n g /m L ) R 2 = 0.652 R 3 0 0 3 0 0 2 0 0 2 0 0 1 0 0 1 0 0 0 0 0 1 0 0 2 0 0 3 0 0 4 0 0 0 1 0 0 2 0 0 3 0 0 4 0 0 co n ce n tra tion H S P 9 0 D Q V A N S A F V E R (n g /m L) con cen tratio n H S P 90 Y ID Q E E LN K (ng /m L) Correlation PRM vs ELISA F 5 0 0 E 5 0 0 Y = 0.845*X + 13.3 4 0 0 c o n c e n tra tio n H S P 9 0 E L IS A (n g /m L ) 2 = 0.878 R 4 0 0 c o n c e n tra tio n H S P 9 0 E L IS A (n g /m L ) Y = 0.726*X + 20 2 = 0.811 R 3 0 0 3 0 0 2 0 0 2 0 0 1 0 0 1 0 0 0 0 1 0 0 2 0 0 3 0 0 4 0 0 5 0 0 0 con cen tratio n H S P 90 Y ID Q E E LN K (ng /m L) 0 1 0 0 2 0 0 3 0 0 4 0 0 5 0 0 Performed by C. Guzel co n ce n tra tion H S P 9 0 D Q V A N S A F V E R (n g /m L)

Bland-Altman plots (method comparison) 1 0 0 SRM vs ELISA YIDQEELNK 5 0 % D iffe re n c e A v e ra g e 0 + 9 5 % 2 0 0 4 0 0 significant different: YES (p < 0.0001) B ia s -5 0 - 9 5 % -1 0 0 YIDQEELNK 1 0 0 PRM vs ELISA 5 0 + 9 5 % % D iffe re n c e A v e ra g e significant different: NO 0 B ia s 2 5 0 5 0 0 (p = 0.1581) - 9 5 % -5 0 -1 0 0 Performed by C. Guzel

Summary  PRM is highly reproducible compared to SRM assay to determine HSP90 concentrations in SCX fractionated sera at relative low ng/mL level  PRM could be used as an attractive alternative for ELISA to quantify multiple proteins highly reproducible in biological samples including sera  Targeted mass spectrometry could be used for personalized cancer diagnostics and follow-up (e.g. in multiple myeloma)