Introduction to Proteomics - UNIL · 2017-08-31 · 4 MS of peptides vs.proteins • Small peptides...
Transcript of Introduction to Proteomics - UNIL · 2017-08-31 · 4 MS of peptides vs.proteins • Small peptides...
Introduction to Proteomics
• Mass Spectrometry for proteomics
• Protein identification by MS/MS
• Proteomics workflows and applications
1
• Mass Spectrometry for proteomics
2
3
Mass spectrometry : essential functions
SAMPLE ION
GENERATION
ION SEPARATION
ION DETECTION
ION SOURCE MASS ANALYZER DETECTOR
• ESI : Electrospray Ionisation
• MALDI : Matrix Assisted Laser Desorption/Ionization
• Quadrupoles • Ion traps • Time-of-flight with reflectron • TOF/TOF • Orbitrap • FT-ICR
• Faraday cup • Scintillation counter • Electromultiplier • High-energy dynodes with electron multiplier • Array (detector) • FT-MS
4
MS of peptides vs.proteins
• Small peptides « fly » much better than big proteins
– Higher ionisation efficiency sensitivity – Signal intensity inversely proportional to mass
• MS in proteomics is mostly (but not only) MS of peptide fragments after protein digestion (typically w. TRYPSIN)
5
MS of peptides : general concepts
• MS : measures charged species (ions)
• MS : we ALWAYS measure M/Z , not Mr
• Molecular mass (Mr) M/Z :
• Work mostly in positive (+) ion mode :
=> peptides protonated (MH+) due to acidic conditions and ionisation process
• Single or multiple (Z>1) charge states can be observed f (compound, ionisation mode, instrument)
M
Z
Mr + (Ma * Z)
Z =
Ma : mass of the ionizing adduct (typically H+ for positive MS) Ma(H+)= 1.0072 u
6
MALDI-TOF of a tryptic digest of BSA
+TOF MS: 50 MCA scans from Sample 1 (BSA Digest 100 fmol) of BSA Digest 100 fmol MS ...a=3.56217430068478150e-004, t0=3.64725878201043440e+001, Thresholded
Max. 1305.0 counts.
800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800m/z, amu
0102030405060708090
100110120130140150160170180190
Intensit
y, coun
ts
927.59
847.59
1440.00 1479.98 1567.94
1640.161022.56869.07
1305.871163.77 1481.981296.861249.77
1050.55871.07 1024.56 1417.931283.91789.53 1142.86 1443.011073.03 1595.951386.76 1824.09857.141108.71 1292.95978.60 1790.101501.84 1616.92
YLYEIAR
LSQKFPK
LVNELTEFAK
FKDLGEEHFK
HPEYAVSVLLR
HLVDEPQNLIK
LGEYGFQNALIVR
KVPQVSTPTLVEVSR
DAFLGSFLYEYSR
? ?
?
? ? ?
7
1+ versus 2+ ions
1160 1162 1164 1166 1168 1170 1172 1174 m/z, amu
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80 84
Intensity, counts
1163.76
1164.76
1165.74
1166.78 1162.70
∆m=1.0 Da
1.0 Da
1.0 Da
1+
580.0 581.0 582.0 583.0 584.0 585.0 586.0 m/z, amu
0
20
40
60
80
100
120
140
160
180
200
Intensity, counts
582.28
582.77
583.27
∆m=0.5 Da
0.5 Da
0.5 Da
2+
LVNELTEFAK Mr = 1162.6234
MH+ MH22+
8
Modes of measurement : MS & MS/MS
Ion production (ionisation)
Ion separation
Ion detection
Ion production (ionisation)
Ion separation – isolation of “parent” ion
Ion fragmentation (CID)
Ion separation – separate fragment ions
Ion detection - measure fragment ions
MS
Detect all ions present
Tandem MS (MS/MS)
Fragment a specific ion and measure fragments
9
• CID = collision induced dissociation
• Precursor ion = parent ion : the one being fragmented
• Daughter ions = fragment ions produced by CID
• Tandem mass spectrometry = MS/MS •- here : the combination of ion selection / CID / fragment analysis
• ESI of tryptic peptides typically generates 2+ - 3+ charged ions due to the presence of Lys or Arg at the C terminal end of the peptides
• y- and b- ion series fragments are usually observed in MS/MS fragmentation spectra of tryptic peptides.
MS/MS Glossary and facts
10
Modes of measurement : MS & MS/MS
MS
MS/MS ( Tandem MS )
Ion production
Ion separation
Detection
Ion production
Ion separation
Detection
Fragmen- tation (CID)
+ + +
+ + +
+ + +
+ + +
11
C
O
C
H
R
N
H
O
C N H C
H
R
H
C
H
R
N
H
C
O
OH
b y
2 1 3
1 2
a x 1 2
c z 1 2
Peptide backbone fragmentation in the gas phase (1)
12
C
O
C
H
R
N
H
O
C N H C
H
R
H
C
H
R
N
H
C
O
OH
2 1 3
O
C N H C
H
R
H
1
C
O
C
H
R
N
H
C
H
R
N
H
C
O
OH
2 3
H
N H C
H
R
H
1
C
O
C
H
R
N
H
O
C C
H
R
N
H
C
O
OH
2 3
N
H
O
C N H C
H
R
H
1
H
+
C
O
C
H
R
C
H
R
N
H
C
O
OH
2 3
+
Peptide Fragmentation In the Gas phase (2) Ion structure
b y 2 1
a x 2 1
c z 2 1 H
+
+
+ +
H
2H+
13
y- , b- fragment ion series
AA1 AA2 AA3 AA4 AA5 OH H
b1
b2
b3
b4
b5
y1
y2
y3
y4
y5
14
QQ-TOF
LIT 3D-IT
QQQ
LIT-OT
transmit scan select
transmit fragment
transmit scan select
trap select fragment scan
trap select fragment scan
transmit scan select
transmit fragment scan
scan
trap select fragment scan
Q: quadrupole IT : ion trap LIT : linear ion trap TOF : time-of-flight OT : orbitrap
ESI- MS/MS-capable instruments used in proteomics
LOW RES
HIGH RES
15
Experimental set-up : nanoLC-MS/MS
[Organic solvent]
t
C18 Column L = 15-25 cm ID = 75 µm
Mass spectrometer
T-splitter
< 1 %
HPLC pump
> 99 % (waste)
sample
200-300 nl/min
m/z
400 600 800 1000 1200
582.32135 z=2
25 30 35 40 45 50 Time (min)
K.LVNELTEFAK.T Measured : 1162.6253 Da Calculated : 1162.6234 Da
LC-MS/MS data acquisition steps - 1
1. Determination of peptide mass
Chromatographic separation of peptides (time)
16
25 30 35 40 45 50 Time (min)
200 400 600 800 1000 m/z
951.50293
595.27759
708.40369
837.48389
494.26
365.21
218.49
N E L T
E F
L V N E L T E F A K
m/z
Mr=1162.625
Ion isolation Collision Induced Dissociation (CID)
LC-MS/MS : what happens inside the instrument -2
2. Isolation , fragmentation, fragment analysis
17
18
Peptide LC elution
36 38 40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 70 72 74 Time, min
Survey scan Detect precursors
500 600 700 800 900 1000 m/z, amu
547.34
651.36
464.27
820.5
1 2
3
MS/MS of
100 200 400 600 800 1000
129.11
571.88
147.12
260.21 239.16 563.37 886.57 84.09 502.34 770.45
785.49
m/z, amu
1
MS/MS of
200 400 600 800 1000
129.11
226.13
490.27 325.20 198.13
589.34 175.12 70.07
542.34 84.09 389.74 183.15 882.52 813.45
m/z, amu
2
Excl
ude
1
2 Data Dependent Acquisition (DDA)
• Protein identification by MS/MS
19
20
MS/MS-based protein identification : concept
Protein sample
Protein fragments (5-30 AA peptides)
Exact masses of peptides Fragmentation (MS/MS) spectrum of each peptide
Protein sequence(s)
Protein fragment sequences (same protease specificity)
Calculated exact masses of peptides Calculated fragmentation spectrum of each peptide
Specific protease e.g. trypsin
MS
software
software
software
Best Match(es)
experimental In silico
21
Matching a peptide : F A D L S E A A N R
23
Mascot ID graph
Mascot Score Histogram Ions score is -10*Log(P), where P is the probability that the observed match is a random event. Individual ions scores > 34 indicate identity or extensive homology (p<0.05). Protein scores are derived from ions scores as a non-probabilistic basis for ranking protein hits. « Threshold »
24
Mascot peptide (ion) score
More in-depth information : For those interested, a more technical description of the calculation of the MASCOT score is given in : http://www.matrixscience.com/help/scoring_help.html
• All steps of score calculation are not known (proprietary algorithm)
• Based on MOWSE algorithm
• Measures degree of matching of MS/MS spectrum vs. theoretical spectrum (# matched fragments, % of matched fragments,… )
• Also takes into account natural distribution of masses and thus « uniqueness » of a peptide mass in database
LIMITATIONS • Not corrected for multiple testing problem • Bias against small peptides
25
Mascot output
1. VIME_HUMAN Mass: 53676 Score: 326 Matches: 5(4) Sequences: 5(4) emPAI: 0.40 Vimentin OS=Homo sapiens GN=VIM PE=1 SV=4 Check to include this hit in error tolerant search Query Observed Mr(expt) Mr(calc) Delta Miss Score Expect Rank Unique Peptide 75 627.7522 1253.4899 1253.5598 -0.0699 0 80 1e-06 1 U R.LGDLYEEEMR.E 86 662.2811 1322.5476 1322.6102 -0.0626 0 86 2.4e-07 1 U R.EEAENTLQSFR.Q 95 714.8277 1427.6408 1427.7045 -0.0637 0 84 3.7e-07 1 U R.SLYASSPGGVYATR.S 115 556.9347 1667.7823 1667.8366 -0.0543 0 27 0.17 1 U R.ETNLDSLPLVDTHSK.R 117 578.9162 1733.7267 1733.8076 -0.0809 1 49 0.00081 1 U R.LQDEIQNMKEEMAR.H
Click Protein view
Click Peptide view
mass error
Peptide (ion) score
Protein score
Calculated From sequence
26
Mascot output – peptide match details
Many solutions proposed for validation : manual, semi-manual, training sets, mass accuracy, statistical validation
27
Peptide validation problem
There are good hits, bad hits, and everything in between…
?
True positives with weak spectra
False positives
Homologous sequences
Non peptides Contaminants (keratins,..) Modified peptides …
28
Monoisotopic mass of neutral peptide Mr(calc): 1113.622 Fixed modifications: Carbamidomethyl (C) Ions Score: 69 Expect: 3.8e-07 Matches (Bold Red): 17/90 fragment ions using 32 most intense peaks
A good hit (Mascot score=69) MS/MS Fragmentation of VMLAANIGTPK Match to Query 525: 1113.616556 from(557.815554,2+)
29
Monoisotopic Mr(calc): 1478.7994 Fixed modifications: Carbamidomethyl (C) Ions Score: 13 Expect: 18 Matches (Bold Red): 10/110 fragment ions using 26 most intense peaks
A bad hit (Mascot score=13) MS/MS Fragmentation of VSIALSSHWINPR Found in KLOT_MOUSE, Klotho precursor - Mus musculus (Mouse) Match to Query 2: 1478.838768 from(740.426660,2+)
*
* *
*
* *
* Strong unmatched peaks
*
30
! We are not identifying proteins, but peptides ! 1) Apply principle of parsimony (Occam’s razor) : within a family, list protein sequence which can explain the most of the identified peptides 2) To highlight the presence of a member of a protein family, at least one discriminating (unique) peptide must be present (what if it is a borderline hit ?)
Protein inference problem
Protein 1 Protein 2 Protein 3
Peptide A
Peptide B
Peptide C
Peptide D
Peptide E
Protein 4
Worst case : two distinct homologous members of the same family found in two samples, each with a weak discriminating peptide….what to say ?
Both hits are reported Reported Not reported !
31
Further processing of Mascot results ( shotgun experiment, one or more samples)
Mascot Search Results Significant hits: ALBU_BOVIN (P02769) Serum albumin precursor (Allergen Bos d 6). ALBU_CANFA (P49822) Serum albumin precursor (Allergen Can f 3). VWF_PIG (Q28833) Von Willebrand factor precursor (vWF) (Fragment). CIQ3_BOVIN (P58126) Voltage-gated potassium channel protein KQT-like 3 RYR2_RABIT (P30957) Ryanodine receptor 2 (Cardiac muscle-type ryanodin K2CA_BOVIN (P04263) Keratin, type II cytoskeletal 68 kDa, component IA ALFB_RABIT (P79226) Fructose-bisphosphate aldolase B (EC 4.1.2.13) (Li
Mascot Search Results Significant hits: ALBU_BOVIN (P02769) Serum albumin precursor (Allergen Bos d 6). ALBU_CANFA (P49822) Serum albumin precursor (Allergen Can f 3). VWF_PIG (Q28833) Von Willebrand factor precursor (vWF) (Fragment). CIQ3_BOVIN (P58126) Voltage-gated potassium channel protein KQT-like 3 RYR2_RABIT (P30957) Ryanodine receptor 2 (Cardiac muscle-type ryanodin K2CA_BOVIN (P04263) Keratin, type II cytoskeletal 68 kDa, component IA ALFB_RABIT (P79226) Fructose-bisphosphate aldolase B (EC 4.1.2.13) (Li
220507
Uncategorized
Sample
Uncategorized
Sample220507_QS_DEJ_TA
P
220507_QS_DEJ_R
AP-Protein name Number
of similar matches
Accession numbers
Protein molecular weight (AMU)
TAP Igol-
Heat shock protein SSA1 (Heat shock protein YG100) - Saccharomyces cere 1 HSP71_YE 69641.3 10 260S ribosomal protein L4-A (L2A) (RP2) - Saccharomyces cerevisiae (Baker' 2 RL4A_YEA 39074 8 9Elongation factor 1-alpha (EF-1-alpha) (Translation elongation factor 1A) (Eu 1 EF1A_YEA 50014.6 10 8Uncharacterized protein YNL157W - Saccharomyces cerevisiae (Baker's yea 1 YNP7_YEA 18034.1 0 5Homocitrate synthase, mitochondrial precursor (EC 2.3.3.14) - Saccharomyc 2 HOSM_YE 48577.5 6 2RNA-binding protein SRO9 (Suppressor of RHO3 protein 9) - Saccharomyce 1 SRO9_YEA 48041.5 2 3GTP-binding protein GTR1 - Saccharomyces cerevisiae (Baker's yeast) 1 GTR1_YEA 35831.2 8 060S ribosomal protein L10 (L9) (Ubiquinol-cytochrome C reductase complex 1 RL10_YEA 25344.1 4 2Glyceraldehyde-3-phosphate dehydrogenase 3 (EC 1.2.1.12) (GAPDH 3) - Sa 1 G3P3_YEA 35728.6 2 340S ribosomal protein S15 (S21) (YS21) (RP52) (RIG protein) - Saccharomyc 1 RS15_YEA 15984.2 4 1Elongation factor 3A (EF-3A) (EF-3) (Translation elongation factor 3A) (Euka 1 EF3A_YEA 115978.2 5 1
Sample 1
Sample 2
Mascot
-Statistical Validation -Protein assignment (parsimony) -Sample alignment
Scaf
fold
1 2
Final list (.xls)
32 Free Scaffold viewer : www.proteomesoftware.com
Data analysis and distribution software : Scaffold
Protein ID probability Percentage of total spectra Nr. assigned spectra Nr. unique peptides Nr. unique spectra Unweighted spectrum count Quantitative value
Filtering parameters
33
Liu, H.; Sadygov, R. G.; Yates, J. R., 3rd, A model for random sampling and estimation of relative protein abundance in shotgun proteomics. Anal Chem 2004, 76, (14), 4193-201.
Spectral counting and quantification
• multiple spectra matches for abundant peptides
• spectral counting reflects relative abundance of proteins between samples
• poorly reliable at low spectral count
34
Spectral counting: Scaffold® http://www.proteomesoftware.com/
Ex. Scaffold report
# matched spectra
Spiked myoglobin (ratio 1:2) in E.coli lysate
CAVEATS: • Normalisation needed • Oversampling necessary • Accurate at medium/high spectral counts , unreliable at low spectral counts • Samples must be reasonably similar
Probability of hitting a protein by unbiased sampling in MS/MS runs ~ concentration Relative quantitation
Improvements possible using: - low-scoring spectra matched to confidently identified peptide sequences (Zhou et al. J Proteome Res. 2010 9:5698-704.) - MS/MS TIC (Scaffold: average, total or TOP 3)
m/z
time
Chromatographic Separation (reversed-phase)
Tandem mass spectra of 50-2000 peptides
Q7Z5Y2 Mass: 118789 Total score: 178 Peptides matched: 6 Rho-interacting protein 3. Mr(calc) Score Peptide 930.48 42 EGLTVQER 1032.54 11 NWIQTIMK 1206.63 29 FSLCILTPEK 1369.75 24 LSTHELTSLLEK 1406.77 55 FFILYEHGLLR 1775.88 16 QVPIAPVHLSSEDGGDR
Protease digestion Peptide extraction
Nano-HPLC
MS/MS
Summary : Typical Analytical Workflow
Database matches DHX9_HUMAN ATP-dependent RNA helicase A NFM_HUMAN Neurofilament triplet M protein Q9BQG0 Hypothetical protein MYO6_HUMAN Myosin VI. TP2A_PIG DNA topoisomerase II, alpha isozyme Q7Z5Y2 Rho-interacting protein 3. FLIH_HUMAN Flightless-I protein homolog. TP2B_MOUSE DNA topoisomerase II, beta isozyme S3B1_HUMAN Splicing factor 3B subunit Q8VCW5 Similar to alpha internexin neuronal Q8CHF9 MKIAA0376 protein (Fragment).
Protein sequence database
Database searching Software (MASCOT) Output :
•Protein identification in simple/complex mixtures
•Extensive sequence coverage and peptide mapping
•Analysis of modified peptides possible
Biological question
35
• Proteomics workflows and applications
36
37
Expression proteomics : analysis of protein expression levels and their changes
Typical questions : * what distinguishes a lymphocyte from a neuron ? * which proteins are newly induced in a cell after a specific stimulus ? • Protein levels : main end product of gene activation, functionally active molecules
• Transcriptomics (cDNA, Affymetrix oligo chips, RNAseq,…) vs. proteomics • Comprehensive • Higher throughput, fast(er) • More sensitive • Assumption : [mRNA] ~ [protein]
Schwanhäusser B. et al. Nature 473, 337-342 (2011)
Classes of experiments - I
38
2) Functional proteomics : studies on subsets of the proteome to map interactions and functional relationships Typical question : * which groups of proteins bind to each other and function together ? * how does the protein composition of an organelle change in determined conditions • Major proteomics application • Protein-protein interactions => active complexes ( molecular machines) • In vivo (keep PTMs, consider natural abundances, occurrences). • Strategy : guilty by association • Others : 2-hybrid screens (genetic), FACS, fluorescence
Classes of experiments - II
39
3) Modification proteomics: the analysis of post-translational modifications (PTMs). Typical question : how is protein activity modulated by covalent chemical modification ? •Very numerous and varied PTMs •Affect activity, targeting, degradation, … •Combinatorial •Dynamic •Heterogeneous •Various stoichiometry •Proteomics and particularly MS are the sole “universal” techniques to study them
Classes of experiments - III
40
• Classical proteomics : – 2D-PAGE and MS (MALDI-TOF)
• More and more : LC-MS/MS-based “shotgun” approaches for
large scale protein identification and quantification
– Compared to 2D-PAGE : greater
dynamic range and higher proteome coverage
– Can be combined with isotope labelling to achieve relative protein quantification
Evolution of proteomics technique
digest
Fraction.
LC-MS
Glyco enrich
organelle
Phospho enrich Isotope
label
Protein complex
SHO
TGU
N
ID QUANT
41
WORKFLOW 1 :
„classical“ 2D-PAGE
+ Protein ID by MALDI-TOF-MS
(+) ID and quantification PTM separated
(-) Reproducibility : poor Dynamic Range limited PTM separated
42
1 1 2 2 3A 3A 3B 3B 4 4
5 6 6 7 7 8 8
9 9 10 10 11 11 12 12
1
14
8
15
Example: adaptation of bacteria to growth conditions
Normal medium Low Glucose
Wick LM, et al, Environ Microbiol 3: 588-599, 2001
E.Coli adaptation to low glucose by modulation of 15 proteins
Trypsin digestion
MS Protein ID
43
WORKFLOWS 2 :
General shotgun protein ID
44
Shotgun sequencing from complex mixtures
Denaturation, Proteolytic digestion
Db search
Multiprotein complex
List of identified proteins 1. P45218 2. P21543 3. Q12588 4. P32651 5. Q01245 6. ….
Complex peptide mixture (1000-20000 species)
Nano rp-LC-MSMS
45
TIC: from 151002_ACO_B4strep.wiff Max. 5.3e5 cps.
5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 105 110 115Time, min
0.0
5.0e4
1.0e5
1.5e5
2.0e5
2.5e5
3.0e5
3.5e5
4.0e5
4.5e5
5.0e5
5.3e5
Inten
sity, cp
s
653.36
133.06
526.26
406.23303.13 852.44
203.11199.19402.54186.12472.25
541.1186.10536.34 449.15574.66
1064.42 629.30665.38494.25
330.18 527.26
615.4186.10563.30
599.32
409.18
A VERY complex mixture – direct analysis (no separation)
46
Mascot Search Results User : MQ Email : Search title :151002_ACO_B4strep.wiff : Angelos frac B4 IP strept MS data file : C:\DOCUME~1\paf\LOCALS~1\Temp\mas5D.tmp Database : Sprot 4028 (114033 sequences; 41888693 residues) Taxonomy : Mammalia (mammals) (23838 sequences) Timestamp : 17 Oct 2002 at 08:09:30 GMT Significant hits: ALBU_BOVIN (P02769) Serum albumin precursor (Allergen Bos d 6). DNM1_HUMAN (P26358) DNA (cytosine-5)-methyltransferase 1 (EC 2.1.1.37) AC15_HUMAN (P35251) Activator 1 140 kDa subunit (Replication factor C IF16_HUMAN (Q16666) Gamma-interferon-inducible protein Ifi-16 (Interfe K1CJ_HUMAN (P13645) Keratin, type I cytoskeletal 10 (Cytokeratin 10) ( K22E_HUMAN (P35908) Keratin, type II cytoskeletal 2 epidermal (Cytoker ACF7_HUMAN (Q9UPN3) Actin cross-linking family protein 7 (Macrophin) ( AC14_HUMAN (P35250) Activator 1 40 kDa subunit (Replication factor C 4 ALBU_FELCA (P49064) Serum albumin precursor (Allergen Fel d 2). AC15_MOUSE (P35601) Activator 1 140 kDa subunit (Replication factor C DYHC_MOUSE (Q9JHU4) Dynein heavy chain, cytosolic (DYHC) (Cytoplasmic AC12_HUMAN (P35249) Activator 1 37 kDa subunit (Replication factor C 3 EF11_CRIGR (P20001) Elongation factor 1-alpha 1 (EF-1-alpha-1) (Elonga RYR3_HUMAN (Q15413) Ryanodine receptor 3 (Brain-type ryanodine recepto K2C1_HUMAN (P04264) Keratin, type II cytoskeletal 1 (Cytokeratin 1) (K PLE1_RAT (P30427) Plectin 1 (PLTN) (PCN). AHNK_HUMAN (Q09666) Neuroblast differentiation associated protein AHNA TRYP_PIG (P00761) Trypsin precursor (EC 3.4.21.4). ACF7_MOUSE (Q9QXZ0) Actin cross-linking family protein 7 (Microtubule CENF_HUMAN (P49454) CENP-F kinetochore protein (Centromere protein F) ALBU_HUMAN (P02768) Serum albumin precursor. PLE1_HUMAN (Q15149) Plectin 1 (PLTN) (PCN) (Hemidesmosomal protein 1) TRI4_HUMAN (Q15650) Activating signal cointegrator 1 (ASC-1) (Thyroid NF1_HUMAN (P21359) Neurofibromin (Neurofibromatosis-related protein N NEBU_HUMAN (P20929) Nebulin
. . .
A VERY complex mixture still gives results, but...
47 450 500 550 600 650 700 750 800 850 900
m/z, amu
0 10 20
40
60
80
100
120
140
160
180
Intensity, counts
18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48
Time, min
0.0 2.0e4 4.0e4 6.0e4 8.0e4 1.0e5 1.2e5 1.4e5 1.6e5 1.8e5 2.0e5 2.2e5 2.4e5 2.6e5 2.7e5
Intensity, cps
120.08 157.15
545.75 581.28 545.75
652.36 216.11 681.35
670.84 131.09 343.18 498.55 738.39 555.22 153.08 561.29 569.27 548.74 499.72 489.53 445.07 499.70 469.25 469.24 651.85
445.10 445.09 648.38
........ how deep are we going ?
analyzed missed
48
Fractionation to reduce complexity (1)
Db search
Complex mixture
Protein IDs 1. P45218 2. P21543 3. Q12588 4. P32651 5. Q01245 …. .... .... .... 1545. Q34258
LC-MS/MS
GeLC-MS workflow
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Trypsin digestion
IgG HC 52 kDa
IgG LC 24 kDa
A
B
C
D
E
F
Bait + -__
LC-MS/MS Trypsin digestion
db search (+) (-)
AHNK AHNKANXA2 ANXA2GNAI2 GNAI25NTD ACTG ACTGGBB1 GBB1GNAI3 GNAI3CD44 CALM CALMHSP7C HSP7C
LYN LYNSTOM STOMCD59 ECHB ECHB
RAB35 RAB35PLEC1 PLEC1VPP1
CAD13 CAD13CLH1 CLH1
SNP23 SNP23
bait interactors
Ex: protein‐protein interactions analysis by affinity purification
49
• Ideal interface to biology • Analytical and micropreparative • Robust • Solid phase chemistry of proteins • Easy, low-tech • Removal of contaminants :
– At the loading point – After migration during fix / staining steps
Why is SDS-PAGE such a good preparation method?
• protein digestion in gel: non quantitative • peptide sequence recovery: usually incomplete • whole protein recovery: poor
Disadvantages
50
51
Db search
Protein IDs 1. P45218 2. P21543 3. Q12588 4. P32651 5. Q01245 …. .... .... .... 1545. Q34258
Complex mixture
260404RPChistonesFRAC001:1_UV1_280nm 260404RPChistonesFRAC001:1_Conc 260404RPChistonesFRAC001:1_Fractions 260404RPChistonesFRAC001:1_Inject
-30.0
-20.0
-10.0
0.0
10.0
20.0
30.0
mAU
0.0 5.0 10.0 15.0 20.0 25.0 mlA1 A2 A3 A4 A5 A6 A7 A8 A9A10 A11 A12 B12 B11 B10 B9 B8 B7 B6 B5 B4 B3 B2 B1 C1 C2 C3 C4 C5 C6 C7 C8 C9 C10 C11 C12 D12 D11 D10 D9
SCX LC or Peptide IEF
LC-MS/MS
Trypsin digestion
Peptide fractionation
Fractionation to reduce complexity (2)
Peptide-based fractionation
WORKFLOWS 3:
quantification strategies
52
53
Comparison : A B ? Which proteins change in amount and how much ?
Applications : - Healthy vs. diseased tissues
- Healthy vs. diseased body fluids
- Drug treated / untreated cells
- Stimulated / unstimulated cells
- Mutants / wt cells
…….
Relative protein quantification
quantification at peptide level 54
Techniques for large scale quantitative proteomics
2D Electrophoresis
MS-based methods
Label-free methods
Labelling methods
Chemical Labelling -ICAT -iTRAQ -ICPL -TMT -Di-ME -18O ….
Metabolic Labelling -SILAC - ISIS -15N ….
Classical (Label-free)
DIGE (Labelling)
MS-based Spectral counting
Targeted SRM/MRM
quantification at protein level
55
∆m
Relative quantification by stable isotope labelling
Sample A Light
Sample B Heavy
mix analyse
H
L
Labelling strategies : • Chemical (side chains : C, K, N-term) ICAT, iTRAQ, ICPLP,…
• Metabolic ( K, R, all )
SILAC,…
• Enzymatic Trypsin + 18O, …
Co-analyse
Eliminate analytical variability
56
SILAC experiment workflow
1:1
+ -
P
A S G Q A F E L I L S P R
id
Data analysis software !
MaxQuant
57
…K SLTNDWEDHLAVK H…
Heat shock protein HSP 90β
13C6-Lys labelling
13C0 13C6
762 763 764 765 766 767 768 769 770 771 772 773 m/z
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
105
110
115
Rel
ativ
e Ab
unda
nce
764.38 767.39
764.88
767.89
765.38 768.39
765.88
768.89 766.88 769.39
766.36
SILAC peaks
∆m=3.0 Da
58
Chemical labelling : Isobaric Tags (iTRAQ)- multiplex quantification
Control Treated
Harvest cells
in vitro chemical labeling
Same peptide from 4 samples has same mass (isobaric)
quantification by tags in MS/MS spectra at fixed M/Z
Trypsin digestion
Multiplexed Protein Quantitation in Saccharomyces cerevisiae Using Amine-reactive Isobaric Tagging Reagents Ross P., Pappin D. et al. Molecular & Cellular Proteomics 3.12 - 2004
59
• Metabolically ( during protein synthesis )
Incorporation of one or more labelled amino acid (+) “native” proteins (+) compatible w. purifications (+) accurate (-) need cultivatable organism (-) limited multiplexing (max. 3)
• Chemically ( post protein synthesis )
“specific” chemical modification of AA side chain (+) any sample can be done (+) higher multiplex (iTRAQ max 8-plex) (-) side (or incomplete) reactions (-) separate purifications (-) less accurate
How to label ? Pros and cons
60
Metabolic labelling Chemical labelling
LC-MS/MS
Sample A Sample B
Protein extraction and trypsin digestion
Light labeling Heavy labeling
Mix A and B
Label free
Relat
ive in
tensit
y
Light Light
Heavy Heavy
m/z
Labeling Analytical variability minimized Number of samples limited (2-8)
Label free Number of samples unlimited Simpler sample preparation Analytical variability Computationally heavy (XIC)
Relat
ive in
tensit
y
LC-MS/MS
Protein extraction and trypsin digestion
m/z Relat
ive in
tensit
y
m/z
Sample A Sample B Sample A Sample B
61 Figures from: Mueller L.N., Brusniak M.Y., Mani D.R., Aebersold R. Journal of Proteome Research, 2008, 7(1), 51-61.
Signal processing in MS quantification
Peptides A/A* Sample X/Y (m/z | Tr | z)
MS/MS spectrum
62
Quantitation summary
Application Multiplexing Accuracy (process)
Quantitative proteome coverage
Linear dynamic rangea
Ease of use
Metabolic protein labeling
• Complex biochemical workflows • Cell culture systems only 2-3 +++ ++ 1–2 logs +
Chemical protein labeling (MS) • Medium to complex biochemical workflows 2-3 +++ ++ 1–2 logs +
Chemical peptide labeling (MS) • Medium complexity biochemical workflows 2-3 ++ ++ 2 logs +
Chemical peptide labeling (MS/MS) • Medium complexity biochemical workflows 2-8 ++ ++ 2 logs +
Enzymatic labeling (MS) • Medium complexity biochemical workflows 2 ++ ++ 1–2 logs ++
Spiked peptides • Medium complexity biochemical workflows • Targeted analysis of few proteins multiple ++ + 2 logs ++
Label free (ion intensity)
• Simple biochemical workflows • Whole proteome analysis multiple + +++ 2–3 logs ++
Label free (spectrum counting)
• Simple biochemical workflows • Whole proteome analysis multiple + +++ 2–3 logs +++
a In MRM mode, dynamic range may be extended to 4–5 logs
Adapted from: Bantscheff M. et al. Anal Bioanal Chem (2007) 389:1017–1031.
63
WORKFLOWS 4:
„Modificomics“ f.ex. Phosphoproteomics
64
Caveat
Protein identification
IS NOT
protein characterisation
Two peptides are enough to identify a protein but
we are still identifying two peptides, not the entire protein
Highly similar sequences cannot be distinguished
For finding PTMs extensive sequence coverage is essential !!!
Modification ∆ Mass Residue Origin Proteolysis Various Any PTM, artefact
Dehydration - 18.0106 N, Q, S, T, Y PTM, artefact
Glycosylation (N-, O-, simple/complex) Various N, S, T, (Q) PTM
Phosphorylation + 79.9663 S, T, Y PTM
Sulfonation + 79.9568 S,T,Y,C PTM
Acetylation + 42.0106 N-term or K PTM, derivative
Carbamidomethylation + 57.0215 C Derivative
Methylation + 14.0156 K, R, D, E, … PTM, artefact
Ubiquitination (mono, di-, poly, K48, K63, ..) Various / + 114.043 K PTM
Sumoylation (SUMO-1, -2, -3) Various K PTM
Oxidation + 15.9949 C, M, W PTM, artefact
ADP-ribosylation + 541.0611 R,C,N,S,E PTM
Myristoylation + 210.1984 N-term G, K, C PTM
Palmitoylation + 238.2297 C, K, S, T, N-term PTM
Prenylation (farnesyl-, geranylgeranyl- ) Various CaaX (C-term) PTM
Nitrosylation + 28.9902 C PTM
….. Almost 100 known…. 65
Some common PTMs
PTM characterization by MS
AA1 AA2 AA3 AA4 AA5 OHH
b1
b2
b3
b4
b5
y1
y2
y3
y4
y5
AA1 AA2 AA3 AA4 AA5 OHH
b1
b2
b3
b4
b5
y1
y2
y3
y4
y5
AA 1
AA 2
AA 3
AA 4
AA 5
OH H
b 1
b 2
b 3
b 4
b 5
y 1
y 2
y 3
y 4
y 5
AA 1
AA 2
AA 3
AA 4
AA 5
OH H
b 1
b 2
b 3
b 4
b 5
y 1
y 2
y 3
y 4
y 5
b3*=b3+∆M
M*=M+∆M
M
b4*=b4+∆M b5*=b5+∆M
y3*=y3+∆M y4*=y4+∆M
y5*=y5+∆M
ETYGDMADCCEK
ETYGDMoxADCCEK 66
67
Common issues in PTM analysis
• Protein sequence coverage → can be increased by multi-enzyme digestion, linked to abundance issue
• Labile PTMs / MS suitability → enzyme inhibitors, PTM derivatisation, use of alternative MS fragmentation (for ex. ETD) • Abundance → PTM enrichment • Artefacts → appropriate sample preparation, control experiments
• Isobaric PTMs → high resolution MS, specific fragments (for ex. immonium ions) • Unknown (untargeted) PTMs → error-tolerant search, blind search • Localization → use of alternative MS fragmentation, localization algorithms • Connectivity → middle-down / top-down analysis
68
Phosphoproteomics
• Kinase cascades all aspects of cell regulation
• Functional assay of the proteome
69
Questions in phosphorylation analysis
• Is a protein of interest phosphorylated ?
• Which proteins are phosphorylated in a cell (or in a precise pathway ?)
• Mapping phosphorylation sites : exact residues
• Quantitation of changes in response to a stimulus
• Effect on physiological protein activity
70
1) Quantity problem : abundance of the protein to analyse is often low and phosphorylation is substoichiometric, especially when purifying from in vivo
Scale up prep, P-peptide enrichment
2) Binding of phosphopeptides to metal and other surfaces : what is its real impact ? Probably sequence
dependent Inert HPLC systems ; injection with Phosphoric acid 3) Low ( or no ) signal intensity due to acidic nature : highly variable depending on seq. Derivatisation 4) Are we digesting with the good enzyme ? Phosphorylated regions are sometimes (often?) in problematic
regions of proteins : very acidic, K/R-poor sequences (example plectin ) Digestion with multiple proteases 5) Bad fragmentation due to neutral loss : highly variable depending on peptide sequence
Choice of MS instrument, neutral loss MS, MSn, multi-stage activation
Problems with phosphopeptide analysis
71
Phosphoproteomics by TiO2 P-peptide enrichment
P
P
P
P
Trypsin
ID : P-proteins P-sites
P P
Raw digest
P P
P P
P P
P P
Peptide fractionation (SCX, IEF, …)
Phosphopeptide Enrichment (TiO2 column)
LC-MS/MS (all fractions)
Variants / options • SILAC label
• α-P-Tyr protein enrichment
• Asp, Glu esterification
72
Metal affinity enrichment makes P-peptides detectable
799.0 1441.8 2084.6 2727.4 3370.2 4013.0Mass (m/z)
1970.4
0
10
20
30
40
50
60
70
80
90
100
% Inte
nsity
4700 Reflector Spec #1 MC[BP = 1859.8, 1970]
1858.7
568
1687.6
488
1581.5
453
1345.5
874
846.30
43
2283.8
948
1555.5
520
1773.6
974
1199.6
309
868.29
60821
.7419
984.68
00
1399.4
635
1880.7
428
1465.6
097
1029.6
702
2008.7
428
1709.6
327
1603.5
292
1544.6
104
1247.5
176
2280.9
314
1829.5
911
935.75
04
2088.6
843
2354.8
398
3033.1
445
1753.6
036
2641.8
477
2226.8
958
2901.0
007
2460.0
518
1154.0
935
2844.0
037
1088.4
283
2586.9
165
2976.1
724
2519.8
750
1951.7
654
3435.2
961
1657.4
292
3492.2
837
3862.5
703
3383.0
786
3326.1
038
2764.0
093
3097.0
986799.0 1441.8 2084.6 2727.4 3370.2 4013.0
Mass (m/z)
2808.7
0
10
20
30
40
50
60
70
80
90
100
% Inte
nsity
4700 Reflector Spec #1 MC[BP = 2089.7, 2809]
2088.7
200
2901.0
273
1993.9
058
2071.3
652
2510.8
892
1782.6
104
1581.5
905
2917.0
127
1451.5
465
1045.3
555
2806.3
408
1687.6
956
1210.5
200
2457.8
291
1391.5
679
1256.4
673
998.45
37
880.56
94
2968.0
063
2753.9
824
812.58
58
2216.8
076
939.45
19
1100.4
554
3383.1
111
1532.6
592
2577.8
735
2131.7
046
1154.4
034
1859.7
871
1730.7
300
2655.8
967
1906.6
333
1337.5
195
2860.0
571
3031.1
201307
5.9375
1628.5
876
3197.0
142
P1
P3
P2
Raw sample
After TiO2 column
MALDI-TOF analysis of a digested phosphoprotein
73
Multi-enzymatic strategy
Ex: POM1_SCHPO (S. pombe, fission yeast)
1. SEQ: sequence covered with trypsin digestion 2. SEQ: additional sequence covered with chymotrypsin digestion 3. SEQ: additional sequence covered with Lys-C digestion 4. SEQ: additional sequence covered with Glu-C digestion => Total sequence coverage: 95.9 % S : phosphosite found with trypsin S : additional phosphosite found with chymotrypsin S : additional phosphosite found with Lys-C SS : ambiguous phosphosite localization => Total number of phosphosites : 41
Semi-specific search, 4 missed cleavages allowed:
• Protein sequence coverage can be improved by using different digestion enzymes: Trypsin (K,R) Chymotrypsin (F, L, W, Y) Lys-C (K) Gluc-C (D, E) Arg-C (R) Combination of 2 enzymes
MGYLQSQKAV SLGDENTDAL FKLHTSNRKS ANMFGIKSEL LNPSELSAVGSYSNDICPNR QSSSSTAADT SPSTNASNTN ISFPEQEHKD ELFMNVEPKGVGSSMDNHAI TIHHSTGNGL LRSSFDHDYR QKNSPRNSIH RLSNISIGNNPIDFESSQQN NPSSLNTSSH HRTSSISNSK SFGTSLSYYN RSSKPSDWNQQNNGGHLSGV ISITQDVSSV PLQSSVFSSG NHAYHASMAP KRSGSWRHTNFHSTSHPRAA SIGNKSGIPP VPTIPPNIGH STDHQHPKAN ISGSLTKSSSESKNLSTIQS PLKTSNSFFK ELSPHSQITL SNVKNNHSHV GSQTKSHSFATPSVFDNNKP VSSDNHNNTT TSSQVHPDSR NPDPKAAPKA VSQKTNVDGHRNHEAKHGNT VQNESKSQKS SNKEGRSSRG GFFSRLSFSR SSSRMKKGSKAKHEDAPDVP AIPHAYIADS STKSSYRNGK KTPTRTKSRM QQFINWFKPSKERSSNGNSD SASPPPVPRL SITRSQVSRE PEKPEEIPSV PPLPSNFKDKGHVPQQRSVS YTPKRSSDTS ESLQPSLSFA SSNVLSEPFD RKVADLAMKAINSKRINKLL DDAKVMQSLL DRACIITPVR NTEVQLINTA PLTEYEQDEINNYDNIYFTG LRNVDKRRSA DENTSSNFGF DDERGDYKVV LGDHIAYRYEVVDFLGKGSF GQVLRCIDYE TGKLVALKII RNKKRFHMQA LVETKILQKIREWDPLDEYC MVQYTDHFYF RDHLCVATEL LGKNLYELIK SNGFKGLPIVVIKSITRQLI QCLTLLNEKH VIHCDLKPEN ILLCHPFKSQ VKVIDFGSSCFEGECVYTYI QSRFYRSPEV ILGMGYGTPI DVWSLGCIIA EMYTGFPLFPGENEQEQLAC IMEIFGPPDH SLIDKCSRKK VFFDSSGKPR PFVSSKGVSRRPFSKSLHQV LQCKDVSFLS FISDCLKWDP DERMTPQQAA QHDFLTGKQDVRRPNTAPAR QKFARPPNIE TAPIPRPLPN LPMEYNDHTL PSPKEPSNQASNLVRSSDKF PNLLTNLDYS IISDNGFLRK PVEKSRP
151
101151201251301351401451501551601651701751801851901951
10011051
MGYLQSQKAV SLGDENTDAL FKLHTSNRKS ANMFGIKSEL LNPSELSAVGSYSNDICPNR QSSSSTAADT SPSTNASNTN ISFPEQEHKD ELFMNVEPKGVGSSMDNHAI TIHHSTGNGL LRSSFDHDYR QKNSPRNSIH RLSNISIGNNPIDFESSQQN NPSSLNTSSH HRTSSISNSK SFGTSLSYYN RSSKPSDWNQQNNGGHLSGV ISITQDVSSV PLQSSVFSSG NHAYHASMAP KRSGSWRHTNFHSTSHPRAA SIGNKSGIPP VPTIPPNIGH STDHQHPKAN ISGSLTKSSSESKNLSTIQS PLKTSNSFFK ELSPHSQITL SNVKNNHSHV GSQTKSHSFATPSVFDNNKP VSSDNHNNTT TSSQVHPDSR NPDPKAAPKA VSQKTNVDGHRNHEAKHGNT VQNESKSQKS SNKEGRSSRG GFFSRLSFSR SSSRMKKGSKAKHEDAPDVP AIPHAYIADS STKSSYRNGK KTPTRTKSRM QQFINWFKPSKERSSNGNSD SASPPPVPRL SITRSQVSRE PEKPEEIPSV PPLPSNFKDKGHVPQQRSVS YTPKRSSDTS ESLQPSLSFA SSNVLSEPFD RKVADLAMKAINSKRINKLL DDAKVMQSLL DRACIITPVR NTEVQLINTA PLTEYEQDEINNYDNIYFTG LRNVDKRRSA DENTSSNFGF DDERGDYKVV LGDHIAYRYEVVDFLGKGSF GQVLRCIDYE TGKLVALKII RNKKRFHMQA LVETKILQKIREWDPLDEYC MVQYTDHFYF RDHLCVATEL LGKNLYELIK SNGFKGLPIVVIKSITRQLI QCLTLLNEKH VIHCDLKPEN ILLCHPFKSQ VKVIDFGSSCFEGECVYTYI QSRFYRSPEV ILGMGYGTPI DVWSLGCIIA EMYTGFPLFPGENEQEQLAC IMEIFGPPDH SLIDKCSRKK VFFDSSGKPR PFVSSKGVSRRPFSKSLHQV LQCKDVSFLS FISDCLKWDP DERMTPQQAA QHDFLTGKQDVRRPNTAPAR QKFARPPNIE TAPIPRPLPN LPMEYNDHTL PSPKEPSNQASNLVRSSDKF PNLLTNLDYS IISDNGFLRK PVEKSRP
151
101151201251301351401451501551601651701751801851901951
10011051
74
Unknown PTMs – mining unassigned MS/MS spectra
- A large proportion of MS/MS spectra are not assigned in proteomics samples:
Ex. From: Chalkley R J et al. Mol Cell Proteomics 2005;4:1189-1193
⇒ error-tolerant search (Mascot) can be used on proteins identified by a first pass search
• The selected enzyme becomes semi-specific
• The complete list of modifications is tested, serially
• For a protein, the set of substitutions that can arise from single base substitutions is tested
• Only one of the above is allowed per peptide.
- Other tools: MS-Alignment, InsPecT, MODa, …(Note: some tools use de novo sequencing)
- All possible unknown PTMs cannot be searched in the classical way (too large search space)
75
Unknown PTMs – error-tolerant search (Mascot) - Results of error-tolerant search must be interpreted with caution: many artefacts (PTM identity or position) !
- Many PTMs can be explained by sample preparation artefacts (oxidation, carbamylation, propionamide, …)
!?
- Ex (PTMs from error-tolerant search in italics):
- R + carbamidomethyl !?
76
Key concepts
Proteome : complexity, plasticity, dynamic range
Proteomics : more challenging than genomics but direct access to cell functions
2D-PAGE : differential display, then MS for ID
LC & MS : many workflows to ID and quantify proteomes to depths of 4000-5000 proteins
Applications : Expression analysis
PTMs
Protein-protein interactions
Protein trafficking, compartments, tissues,…
Biomarker discovery
….
77
Take home message-1
Many new possibilities in large scale protein analysis PTM’s
PTM are one of the most exciting and difficult « new » fields Huge variety and complexity of PTMs; no general workflow exists
Quantitative proteomics
Quantitation is now feasible on a significant fraction of the proteome Several methods available; data quality and throughput are variable. Choice is often based on the experimental system and design
78
Take home message-2
• Some choices crucial for success:
Biological question : what are we looking for ? Model system Sample preparation (!)
Abundance of protein of interest Complexity of mixture Enrichment mechanism
Data analysis : not sooooooooooo easy ! If we get results, can we interprete them ? If we get results, are they going to be useful ?
79
• Mallick, P., & Kuster, B. (2010). Proteomics: a pragmatic perspective. Nature Biotechnology, 28(7), 695-
709.
• Nesvizhskii, A. I., Vitek, O., & Aebersold, R. (2007). Analysis and validation of proteomic data generated by tandem mass spectrometry. Nature methods, 4(10), 787-97.
• Köcher, T., & Superti-Furga, G. (2007). Mass spectrometry-based functional proteomics: from molecular machines to protein networks. Nature methods, 4(10), 807-15.
Some good articles to start reading
• Cox, J., & Mann, M. (2007). Is proteomics the new genomics? Cell, 130(3), 395-8.
Discusses limits and potential of proteomics and compares proteomics with genomics (incl. mRNA protein
correlation)
• Cravatt, B. F., Simon, G. M., & Yates, J. R. (2007). The biological impact of mass-spectrometry-based proteomics. Nature, 450(7172), 991-1000.
Very good overview presenting examples of proteomics studies which had a significant impact on a field of biology
• Yates, J. R., Ruse, C. I., & Nakorchevsky, A. (2009). Proteomics by mass spectrometry: approaches, advances, and applications. Annual review of biomedical engineering, 11(c), 49-79.
Focus on MS technology and its applications
Some good reviews -1
Some good reviews - 2
• Jensen, O. (2006). Interpreting the protein language using proteomics.
Nat Rev Mol Cell Biol, 7: 39-403. • Witze, E.S., Old, W.M., Resing, K.A., & Ahn, N.G. (2007). Mapping protein post-translational modifications
with mass spectrometry. Nature Methods, 4(10): 798-806.
Give an overview of proteomics techniques used for PTM characterization in cells • Kamath, K.S., Vasavada M.S. & Srivastava S. (2011). Proteomic databases and tools to decipher post-
translational modifications. J. Proteomics, 75: 127-144.
Review of databases and other tools useful for study of PTMs
• Ong, S.E. & Mann, M. (2005). Mass spectrometry-based proteomics turns quantitative.
Nat Chem Biol, 1(5): 252-262.
General introduction to quantitative proteomics
• Bantscheff, M., Schirle, M., Sweetman, G., Rick, J. & Kuster, B. (2007). Quantitative mass spectrometry in proteomics: a critical review. Anal Bioanal Chem, 389: 1017–1031.
• Elliott, M.H., Smith, D.S., Parkera, C.E. & Borchers, C. (2009). Current trends in quantitative proteomics. J. Mass. Spectrom. 44: 1637–1660.
Review and comparison of analytical techniques used in quantitative proteomics
80