High Resolution Mass Spectrometry (HRMS) in Discovery ...

HIGH RESOLUTION MASS SPECTROMETRY (HRMS) IN DISCOVERY PROTEOMICS

Michael L. Merchant, PhDSchool of Medicine, University of LouisvilleLouisville, KY

A clinical proteomics perspective

Learning Objectives• After this presentation, you should be able to:

1. Provide a working description of the relationship of resolution and mass accuracy as it related to protein identification.

2. Provide an overview of the LCMS‐based proteomic workflow. 3. Describe factors mitigating the discovery process in clinical

proteomics.4. Provide an example of the application of high‐resolution

mass spectrometry (HRMS) to biomarker discovery and protein post‐translational modification

Overview of topics to be discussed• Introduction• Overview of the LCMS workflow• Sample handling and preparation• Data dependent acquisition (DDA) workflow• ‐versus the data independent acquisition (DIA) workflow

• Informatics• Current applications

• Biomarker discovery• Quantitative Phosphoproteomics• Top‐down mass spectrometry

• Conclusions

• University of Louisville– Jon Klein, MD, PhD– Ken McLeish, MD– Michael Brier, PhD– Jian Cai, PhD– James Hribar, PhD– Danny Wilkey, BA– Ming Li, BS

• The Ohio State Univ.– Brad Rovin, MD

• University of Pennsylvania– Harv Feldman, MD– Peter Yang, PhD

• University of Washington– Jonathan Himmelfarb, MD

• Disclosures• NIH (NIDDK)

• R01 DK091584• U01 DK085673

• Owner/Partner• Pharos Medicine, LLC.

Acknowledgements

Introduction – terms and definitions• Resolution

• Low (1000)• Medium (10,000‐20,000)• High (≥ 50,000)

• Mass Accuracy• Low (1.0‐0.1 Da)• Medium (0.1‐0.01 Da)• High (0.01‐0.001 Da)

• Tandem Mass Spectrometry• MS2

• Precision Proteomics• Application of high resolution methods at the MS1 and MS2 (or MSn) level of analysis

• Proteospecific‐• Amino acid sequence that is specific to the species being studied

• Data dependent analysis• MS2 fragmentation targeting based on MS1 information

• Data independent analysis• MS2 fragmentation data collected systematically and independent of MS1 information

• Qualitative analysis• Index of proteins present in the sample

• Quantitative analysis• Associating relative or absolute values of abundance to the Qual analysis.

Mann M , and Kelleher N L PNAS 2008;105:18132-18138Egertson JD et. Al. Nature Methods 2013 10(8) 744-2528

Proteomic workflow

Traditional “ Bottom‐up” Approach Bottlenecks

• Amount of isolated protein• Sample complexity• Protein separation methods• Protease digestion• Peptide separation methods• Sensitivity, mass accuracy, and scan speed of mass spectrometer

• Prior knowledge of PTM• Protein database• False positive assignments

Protein Peptidesprotease

Mass Spectrometer

Bench work(Variability)

Informatics

Peptide assignment

Protein assignment

In Silico(Statistics)

Mann M , and Kelleher N L PNAS 2008;105:18132-18138

©2008 by National Academy of Sciences

Precision Proteomics

Impact of mass spectrometer in contemporary proteomics

Factors mitigating discovery progress

Research question

Sample handling

Mass spectrometer

Informatics

• Defining the question that proteomics will answer

• Sample handling• Isolation of protein‐containing structure• Protein extraction• Protein separation• Protein digestion

• Mass spectrometry analysis• Quantification

• Informatics• Identification• Bioinformatics/Interpretation

The research question

• Expression and quantitative proteomics• Define and quantify the protein components

• Functional proteomics.• Define the interactions among proteins

yielding knowledge about Protein Interaction Networks (PIN)

• Define the mechanisms by which proteins communicate with each other yielding knowledge about Protein Signaling Networks (PSN)

• Minimizing unnecessary variance and systematic errors

• Power the study correctly• Correcting for multiple hypothesis• Replicates

• a) Biological b) Technical

• Detergents• a) SDS or NP‐40 b) Protease MAX

• Chaotropes• a) Urea or b) GuanHCL

• Buffers• a) ammonium bicarbonates/acetates or b) TRIS, c) Trizma,

d) Hepes, e) MES

• Inhibitors• a) Protease, b) Phosphatase, c) Deacetylase, or d)

Bacteriastat

• Proteases• a) Trypsin, b) Arg‐C, c) Lys‐C, or d) Asp‐N

Expression

PSN

PIN

The research design

Protein Expression System

Lysis

Subcellular Fractionation

Chromatography

Ion Exchange

Affinity

Antibody Lectin Metal Chelate

Reversed Phase

SECHILIC

Electrophoresis

NativeDenaturing

1‐DE 2‐DE

Addressing complexity before mass measurement

Overview of LCMS workflowAcquisition of parent “peptide” mass to charge (m/z) and fragment m/z values

Algorithmic matching of observed to predicted fragment masses

sample nanoLC LTQ‐Orbitrap‐ELITE

nanoLC‐column

nanospray source

20120524_ABC_Pooled_2d_Rovin_30_Recal #19133 RT: 68.93 AV: 1 NL: 2.51E6T: FTMS + c NSI Full ms [300.00-2000.00]

300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900m/z

0

5

10

15

20

25

30

35

40

45

50

55

60

65

70

75

80

85

90

95

100

Rel

ativ

e A

bund

ance

20120524_ABC_Pooled_2d_Rovin_30_Recal #19137 RT: 68.94 AV: 1 NL: 5.26E3T: ITMS + c NSI d Full ms2 [email protected] [275.00-2000.00]

300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900m/z

0

5

10

15

20

25

30

35

40

45

50

55

60

65

70

75

80

85

90

95

100

Rel

ativ

e A

bund

ance

y1y2y3

b3b2b1

NN

NO H

O

OH

H O

H O

NH 2

R 1

R 2

R 3

R 4

z1z2z3

c3c2c1

NN

NOH

O

OH

H O

H O

NH2

R1

R2

R3

R4

0

MS1*

MS2

FragmentIntensity

m/z

Curates database

Advantages of HRMS in proteomic discovery research

Qualitative/quantitative methods using HRMS‐data dependent analysis (DDA)

MS1‐based quantification

• Feature analysis• Differential LCMS (dLCMS)

• Enzymatic labeling• Oxygen‐18 (18O)

• Chemical labeling• Reductive methylation• ICAT• Chemical synthesis with stable isotope labeled amino acids

• Metabolic labeling• SILAC (13C‐, 15N‐ labeled amino acids)• 13C‐, 15N‐ labeled amino acids growth media

MS2‐based quantification

• Label Free• Spectral counting

• Isobaric chemical tags• Reporter ion quantification• iTRAQ• TMT

McLeish_13_20121203_Recal #11491 RT: 49.10 AV: 1 NL: 2.32E7T: FTMS + p NSI Full ms [300.00-2000.00]

400 600 800 1000 1200 1400 1600 1800m/z

0

5

10

15

20

25

30

35

40

45

50

55

60

65

70

75

80

85

90

95

100

Relative Abu

ndan

ce

636.6413m/zz=+3


400 600 800 1000 1200 1400 1600 1800m/z

0

5

10

15

20

25

30

35

40

45

50

55

60

65

70

75

80

85

90

95

100Relative Abu

ndan

ce

636.6425m/zz=+3 Urine proteomic study

Acute kidney injury following cardiac surgeryElution of identical peptide in two patient samples

Single MS spectra can contain >100 peptide featuresHigh complexity

1) analysis of only 16 of 100 peptides is common2) repeat analysis may miss 5 of the 163) many will have closely isobaric species

Caveats to increased sensitivity


600 610 620 630 640 650 660 670 680m/z

0

5

10

15

20

25

30

35

40

45

50

55

60

65

70

75

80

85

90

95

100

Relative Abu

ndan

ce

636.6413m/zz=+3


600 610 620 630 640 650 660 670 680 690m/z

0

5

10

15

20

25

30

35

40

45

50

55

60

65

70

75

80

85

90

95

100

Relative Abu

ndan

ce

636.6425m/zz=+3

Increased sensitivity can contribute to co‐isolation of closely isobaric peptides

c₁₇⁺1779.94

y₁₅⁺1548.90

c₆⁺663.45

c₁₁⁺, z₁₁⁺1175.68

z₁₂⁺1246.77

c₇⁺, z₇⁺734.53

c₁₇²⁺890.33

c₈⁺849.35

500 1000 1500

m/z

0

10

20

30

40

50

60

70

80

Inte

nsity

[cou

nts]

(10^

3)

Extracted from: C:\Users\dwilkey\Desktop\Temp Proteome Discoverer Analyses\McLeish_20121203\McLeish_6_20121203_Recal.raw #11743 RT: 49.54 ITMS, ETD, z=+3, Mono m/z=636.64252 Da, MH+=1907.91301 Da, Match Tol.=1.2 Da

Sequence: VFNNIGADLLTGSESENK, Charge: +3, Monoisotopic m/z: 636.64252 Da, MH+: 1907.91301 Da, RT: 49.54 min,Identified with: Mascot (v1.27); IonScore:48, Exp Value:1.5E‐003, Ions matched by search engine: 11/102Fragment match tolerance used for search: 1.2 DaFragments used for search: c; y; z; z+2

SCY1‐like protein 2 [SCYL2_HUMAN]

HRMS in clinical proteomic studies• Label‐free identification and quantification of biomarkers of human disease (urine proteomics)• Addressing over‐abundant proteins

• Post‐translational modifications• Cell signaling and the phosphoproteome

• Direct analysis of whole proteins via top‐down MS• Histones and epigentics

URINE PROTEOMICS AND LUPUS NEPHRITISEfforts to support diagnosis without renal biopsies

Renal disease, urine protein, and depletionPatient No.

10 15 34 49 71 81 105

Perc

ent T

otal

Spec

tal C

ount

s

Albumin

Sero

trans

ferrin

a2-M

acro

globu

lin

a1-A

ntitr

ypsin

Hapto

globin

a1-a

cid G

lycop

rotei

n

Apolip

opro

tein

A-I

Apolip

opro

tein

A-II%

Spe

ctral

Count

0

20

40

60

80

100

~80% of urinary MS/MS datais from abundant serum proteins

Scaffold IDs; 99% Protein Probability, 95% Peptide Probability with at Least Two High Confidence Peptides

All Proteins

Decoy IDsNo Decoy IDs

# Proteins # Identified Spectra # Spectra % Identified

LTQ Analysis 266 773 94157 0.94Orbitrap Elite Analysis 620 35543 145551 24

TIC for 0.15M Salt Step of SCX Fractionation

LTQ LTQ‐Orbitrap

RT: 20.00 - 120.00

20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 105 110 115 120Time (min)

0

5

10

15

20

25

30

35

40

45

50

55

60

65

70

75

80

85

90

95

100

Rel

ativ

e A

bund

ance

NL: 3.55E7TIC F: ITMS + c NSI Full ms [300.00-2000.00] MS 20120523_ABCPooled_2d_Rovin_30

RT: 5.00 - 105.00

5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 105Time (min)

0

5

10

15

20

25

30

35

40

45

50

55

60

65

70

75

80

85

90

95

100

Rel

ativ

e A

bund

ance

NL: 2.55E9TIC F: FTMS + c NSI Full ms [300.00-2000.00] MS 20120524_ABC_Pooled_2d_Rovin_30_Recal

Scan #19133

E+7 E+9

MS Scan #1913320120524_ABC_Pooled_2d_Rovin_30_Recal #19133 RT: 68.93 AV: 1 NL: 2.51E6T: FTMS + c NSI Full ms [300.00-2000.00]

300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900m/z

0

5

10

15

20

25

30

35

40

45

50

55

60

65

70

75

80

85

90

95

100

Rel

ativ

e A

bund

ance

Zoom from 990‐1070m/z

20120524_ABC_Pooled_2d_Rovin_30_Recal #19133 RT: 68.93 AV: 1 NL: 1.40E6T: FTMS + c NSI Full ms [300.00-2000.00]

995 1000 1005 1010 1015 1020 1025 1030 1035 1040 1045 1050 1055 1060 1065 1070m/z

0

5

10

15

20

25

30

35

40

45

50

55

60

65

70

75

80

85

90

95

100

Rel

ativ

e A

bund

ance

1049.5465m/z

MS2 Spectrum for 1049.5465m/z Peak20120524_ABC_Pooled_2d_Rovin_30_Recal #19137 RT: 68.94 AV: 1 NL: 5.26E3T: ITMS + c NSI d Full ms2 [email protected] [275.00-2000.00]

300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900m/z

0

5

10

15

20

25

30

35

40

45

50

55

60

65

70

75

80

85

90

95

100

Relative Abu

ndan

ce

y₁₄⁺1490.75

y₈⁺847.48

y₁₆⁺1660.78

b₁₀²⁺‐H₂O, b₅⁺438.22

[M+2H]²⁺‐NH₃‐H₂O1032.61

b₆⁺551.31

b₈⁺‐H₂O705.44

y₁₃⁺1375.72b₈⁺

723.37

b₁₀⁺893.45

b₁₆⁺‐NH₃1548.71

y₁₂⁺1276.73

y₁₁⁺1205.58

400 600 800 1000 1200 1400 1600 1800 2000

m/z

0

1000

2000

3000

4000

5000

6000

Inte

nsity

[cou

nts]

, , , , ,

Peroxiredoxin‐6 [PRDX6_HUMAN] Monoisotopic m/z: 1049.54653 Da (‐3.54 mmu/‐3.37 ppm) Mascot IonScore:68Sequence: PGGLLLGDVAPNFEANTTVGR, Charge: +2

Observed

Matched

Phosphoproteome Studies

Identification of low abundant (rare) , protein post‐translational modifications to gain information on molecular signaling events…

Studying protein phosphorylation events using affinity enrichment…needle in a haystack

Engholm‐Keller, K and Larsen, MR J. Proteomics 75 (2011) 317‐328

Reductive methylation:an inexpensive and effective quantification strategy

Wilson‐Grady, JT, Haas, W, and Gygi, SP Methods 61 (2013) 277‐286

Integrated workflow for pS/pT/pY detection and quantification


Informatics analysis of phosphoproteome data


HRMS AND TOP DOWN PROTEOMICS

Analysis of protein post‐translational modifications from the whole protein level.

FTMS selected ion monitoring (932‐942m/z) for CAIICAII_FTMS_121116105929 #1 RT: 7.02 AV: 1 NL: 5.13E5T: FTMS + p NSI SIM ms [932.00-942.00]

936.95 937.00 937.05 937.10 937.15 937.20 937.25 937.30 937.35 937.40 937.45 937.50 937.55 937.60 937.65m/z

0

5

10

15

20

25

30

35

40

45

50

55

60

65

70

75

80

85

90

95

100

Rel

ativ

e A

bund

ance

CAII_FTMS_121116105929 #1 RT: 7.02 AV: 1 NL: 5.13E5T: FTMS + p NSI SIM ms [932.00-942.00]

932.0 932.5 933.0 933.5 934.0 934.5 935.0 935.5 936.0 936.5 937.0 937.5 938.0 938.5 939.0 939.5 940.0 940.5 941.0 941.5 942.0m/z

0

5

10

15

20

25

30

35

40

45

50

55

60

65

70

75

80

85

90

95

100

Rel

ativ

e A

bund

ance

937.302

937.979

936.721 938.655CAII ‐H2O

CAII + Na+1

CAII + 2Na+1

CAII

Z=+31HCD CAII_121116103808 #1 RT: 58.90 AV: 1 NL: 5.99E4T: FTMS + p NSI sid=25.00 Full ms2 [email protected] [255.00-1800.00]

300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500m/z

0

5

10

15

20

25

30

35

40

45

50

55

60

65

70

75

80

85

90

95

100

Rel

ativ

e A

bund

ance

923.85779z=5

1175.15967z=6

1007.42407z=7

1086.89429z=7

740.93079z=4

836.81531z=3539.28973

z=1

1268.04236z=6793.13452

z=3337.19217

z=1 1409.98962z=5

593.14636z=5

897.69824z=?

1360.49829z=4513.45264

z=5 648.56311z=4

702.35559z=1

438.24109z=1

1215.91504z=3

1295.22046z=6

365.18753z=?

1450.20984z=? 1547.8

z=

MS2Fragmentation

Spectrum

Modified histones

Histones

HPLC 12% SDS PAGE

Abs 280n

m

Time (min)

Isolate/enrichHistones

Arg‐Cdigestion

Modified and unmodified histone peptide fragments

Arg‐C digestion

Extracthistones

LCMS Analysis

Isolated Cells

TopDown Bottom

Up

Analysis of whole protein modification

NFUT1-130724_50-54min_131002 #1623-1997 RT: 45.84-52.48 AV: 125 NL: 2.62E5T: FTMS + p NSI Full ms [400.00-1200.00]

400 450 500 550 600 650 700 750 800 850 900 950 1000 1050 1100 1150 1200m/z

0

5

10

15

20

25

30

35

40

45

50

55

60

65

70

75

80

85

90

95

100

Rel

ativ

e A

bund

ance

557.21947z=1

699.02972z=22 731.55547

z=21768.73218

z=?668.07246z=23

809.24424z=19

640.19428z=24 853.25669

z=18614.66522

z=25 903.44664z=1 960.84996

z=?535.23806z=1

590.44805z=26

1023.83925z=1 1097.89761

z=?

877.49016z=?

1183.43113z=?986.48150

z=?935.52312

z=? 1138.63628z=?

1061.13676z=?

512.42583z=?451.23101

z=?

NFUT1-130724_50-54min_131002 #1623-1997 RT: 45.84-52.48 AV: 125 NL: 1.26E5T: FTMS + p NSI Full ms [400.00-1200.00]

690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707m/z

0

5

10

15

20

25

30

35

40

45

50

55

60

65

70

75

80

85

90

95

100

Rel

ativ

e A

bund

ance

699.02972z=22

698.25671z=22

699.66607z=22697.71043

z=?

700.30285z=22

697.02799z=? 700.93922

z=22

701.53006z=?

696.29910z=22

702.80224z=?

695.61880z=22

702.12165z=?

694.38839z=?

695.16229z=? 703.39289

z=?693.75450

z=?693.07001

z=? 704.16485z=?

705.34821z=1

692.34305z=?

691.38481z=?

707.39389z=?706.16880

z=?

mRP‐HPLC

Chargestates

PTMstates

Conclusions• The sensitivity and analytic power of current HRMS platforms are pushing the boundaries of discovery work.

• Given the increased sensitivity of state‐of‐the‐art mass spectrometers, sample (mis‐)handling can significantly impact the quality of the data regardless of the proteomic end goal.

• With regards to clinical proteomics experiments, maximizing reproducible sample handling is vital to the success of all clinical proteomic projects.

• LCMS based proteomics methods are important tools that can be used to aid in the study human health and disease.

• Efficient utilization of proteomic resources are often times best achieved through collaboration.

Self‐Assessment Questions1. Which of the following is not a source of variability in proteomic discovery research‐

A) sample complexityB) peptide separation methodsC) sensitivity, mass accuracy, and scan speed of the mass spectrometerD) amount of isolated proteinE) none of the above

2. Which of the following is not a statistical bottleneck proteomic discovery research‐A) prior knowledge of PTMB) protein databaseC) false positive assignmentsD) peptide separation methodsE) none of the above

3. Advantages of high resolution mass spectrometers over older low resolution mass spectrometers include‐A) detection of lower abundant peptidesB) greater coverage of peptide fragmentation dataC) better detection of protein post‐translational modificationsD) characterization of whole proteins and post‐translational patternsE) all of the above

High Resolution Mass Spectrometry (HRMS) in Discovery ...

Documents