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

22

MASCOT Search www.matrixscience.com

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