Post-translational modifications

Post-translational Modification of ProteinsExpanding Nature’s Inventory (2006)by C.T. Walsh

ISBN 0-9747077-3-2

Genome↓

Transcription – mRNA↓

Translation – proteins + co-translational modifications

↓Post-translational modifications

Just for introduction

Post-translational modifications I

• Enzymatic processing• N-, O-, C-linked glycosylation – Asn, Ser/Thr/Hyl/Hyp,

Trp• Phosphorylation – Tyr, Ser, Thr, His, Asp• Acylation

acetylation of the N-terminusfatty acid anchors on Cys

• Cross-linkage – Lys, Trp, Tyr, Met• Oxydation – Cys, Met, Trp, Tyr, His• Methylation – N-terminus, Arg, Lys• Ubiquitination - Lys

Post-translational modifications II

• Cannot be predicted – though consensus sequences have been reported for some of them

• Organism-dependent• Can be tissue- or location-specific• Stable or dynamic – high and low level• Alters biological activity, and physical

properties• May alter the immune response

↓

How to find PTMs?

Sometimes proteins give us a clue:• higher/lower/wider band on SDS-PAGEFor example, +~5 kDa on SDS-PAGE may indicate 1 occupied N-linked glycosylation site

• multiple spots with IEF or on 2D-gel• higher MW with gel-filtration or native gel-elfo• altered hydrophobicity• altered immunochemistry• altered biological activity• “blank” cycle in Edman sequencing (with proper

standard & altered chromatography the modification could be identified: Identification of glycosylation sites in mucin peptides by edman degradation. Zachara NE, Gooley AA. Methods Mol Biol. 2000;125:121-8. Review.

How to find PTMs?

Sometimes we’re searching for something specific

• using a known donor and radioactive labeling

ATP for phosphorylation (Why cannot we use this approach for glycosylation?)

• using modification-specific derivatization and staining or enrichment

Periodate oxidation and labeling for glycosylation

• using modification-specific antibodiesAgainst phosphoTyr

• using modification-specific enzymes Phosphatase or glycosidase treatment

• Mass spectrometry Mass Spectrometry Mass Spectrometry!

Sample preparation • Does the modification survive?• Losses due to enzymatic reactionsPhosphatase activity for phosphopeptides

• Losses due to chemical reactionsAcid-sensitive modifications, such as phosphoHis

• Losses due to different physical properties

Fatty-acylated or glycopeptides

• Do we introduced some modifications?* Biological artifacts* Chemical side reactions

Some modifications that occur in vivo and can be introduced in vitro

• Carbamidomethylation of N-termini and Lys side-chains

• Formylation of Lys side-chains• Asn cyclization; -aspartame formation;

deamidation• Oxidation • Glu, Asp methyl ester formationAnd the list goes on…

Finding the N-termini

• 80-90% of the eukaryotic proteins is acetylated

If 2nd aa: G, A, S, C, T, P, V - Met-1 is clipped of; Ac- is added to G,

A, S, TIf 2nd aa: E, D, Q, M, I, L, W, F – persisting Met gets acetylated

Finding the N-termini

• N-terminal sequencing• Positive enrichment (Wells-group);

subtiligase-mediated selective N-terminal labeling with biotin-containing probe; enrichment; release with TEV protease; MS-analysis

“Global sequencing of proteolytic cleavage sites in apoptosis by specific labeling of protein N termini.” Mahrus S, Trinidad JC, Barkan DT, Sali A, Burlingame AL, Wells JA. Cell. 2008 Sep 5;134(5):866-76.

• Negative enrichment – next slide

Finding the N-termini

• McDonald L, et al., Positional proteomics: selective recovery and analysis of N-terminal proteolytic peptides.Nat Methods. 2005 Dec;2(12):955-7. Epub 2005 Nov 18.

Modified N-terminus

Modified side-chain

Methylation (mono, di, tri - +14, +28, +42 Da)

• N-terminus, Lys, Arg, His• Trimethyl – acetyl = 36 mmuAccurate mass measurement helps;Fragmentation is different too

• Glu (Asp) may form Me-ester – upon CBB staining (MeOH + acid) + 14 Da

7000

6000

5000

4000

3000

2000

1000

0

644.0643.8643.6643.4643.2643.0642.8642.6m/z

14

12

10

8

6

4

2

0

x103

795.0794.8794.6794.4794.2794.0793.8793.6m/z

642.77672 (10+)

b57 + 42 Da

Calculated masses 6418.6718 - with Ac 6418.7082 - with Me3

Experimental value 6418.6973 Me3b57 (-1.6 ppm)

Mass calculated: 7134.8569 Experimental: 7134.8463 (-1.4 ppm)

y64

793.65655 (9+)

FT ICR MS

700

600

500

400

300

200

100

0

Ion

Cou

nts

400350300250200150100

m/z100

80

60

40

20

0

Ion

Cou

nts

800700600500400

m/z

72.08

87.05147.11 169.10

186.13

214.13

260.20

280.20(2+)

294.85(3+)308.86(3+)

329.74(2+)

373.26(2+)

745.50

623.36599.39

462.79(2+)

479.83(2+)

571.28

486.31

422.77(2+)

427.77(2+)

44

1.5

4(2

+)

45

8.2

9

471.29(2+)

314.5(3+)

320.2(3+)

354.73(2+)

141.1

V

N y1

SV-28-H2O

a2-NH3

a2

b2

y2

y42+

[MH3-NH3-NMe3]3+

[MH3-2NH3]3+

[MH3-NH3]3+

y52+

[a7-NH3-NMe3]2+ y6

2+

y72+

[MH2-CO2H-NMe3]2+

[MH2-NH3]2+

MH22+

SVK*ESVK*EI

SV

K*E

-28

[MH2-2xNH3]2+

y6

b6-NH3-NMe3

512.36

SVK*EI-28

SVK*EI-28-NMe3

Asn-Val-Ser-Val-Lys(Me3)-Glu-Ile-Lys

- 59 Da

Disulfide-bridges

• in membrane and secreted proteins important 3D structure feature

• Some toxins and antibacterial peptides feature characteristic Cys network

• -S-S- prone to shuffling @ basic pH

How to find & determine disulfide linkages?

• Alkylate the free SHs; digest the protein; fractionate the digest; reduce/or oxidize the fractions; look for altered chromatographic behavior; identify the components

* Optimal pH for alkylation is good for shuffling too* Digestion may be hindered by a compact structure* Multiple links could be present within one

cleavage product* If oxidation is used, Met, Trp also may be altered

A “novel” methodology for assignment of disulfide bond

pairings in proteins.• Wu J, Watson JT. Protein Sci. 1997 Feb;6(2):391-8.

• Reduction by TCEP (works at low pH); cyanylation of nascent –SHs; fractionation of partially reduced forms; cleavage at the N-termini of the modified Cys-residues (with NH4OH)

* Fractionation of complex, partially reduced mixtures could be a problem…

Digestion

@ low pH

Reduction/alkylation

Reduction/alkylation

I

m/z

I

m/z

I

m/z

I

m/z

I

m/z

MS/MS

Reduction/alkylation

Digestion

@ low pH

Reduction/alkylation

Assigning disulfide-bridges

Synthetic Ac-TIMP-1(Ser175)126-184

ECTVFPCLSI PCKLQSGTHC LWTDQLLQGS EKGFQSRHLA CLPREPGLCS WQSLRSQIA

Where are the disulfide bridges?

* digestion with trypsin @ pH 6;

* with pepsin in acid

Bodi, N. et al., J. Pept. Sci. 9, 430-441 (2003).

300028002600240022002000m/z

inte

nsity

12001000800600m/z

inte

nsity

1600140012001000800600m/z

[38-

44]

fre

e S

H

[33-

37]

[45-

55]

Fre

e S

H

[38-

55]

41-

49

[38-

44]-

[45-

55]

41-

49

Ac[

1-13

]-[4

5-55

]2-

7, 1

2-49

[14-

32]-

[38-

44]

20-

41

[33-

37]

[38-

44]

S*

[45-

55]

S*

Ac[

1-13

] 1

S-S

, 1 S

*

Ac[

1-13

] 3

S*

[45

-59]

1 S

*

30002800260024002200200018001600m/z

Ac[

1-13

] 1

free

SH

, 2 S

*

[38

-55]

2 S

* [

14-3

2] 1

S*

[38

-59]

2 S

*

[14

-37]

1 S

*

MALDI-TOF analysis of the tryptic digest

PSD

In-source reduction

Disulfide bridges in TIMP-1 C-terminal domain: [38-44]-S-S-[45-55], PSD of MH+ at m/z 2082.1

1000

800

600

400

200

0

Inte

nsi

ty

200015001000500m/z

2X1

75

.6, y

1,

y 1*

25

1.6

, b2

25

5.2

, y2*

-NH

3

32

2.6

, b3

77

6.3 8

10

.08

41

.7 12

42

.1* 1

27

6.4*

13

08

.9*

11

0.1

, H

19

28

, b6+

H2O

4,

b 10*

+H

2O

His-Leu-Ala-NH-CH-CO-Leu-Pro-Arg

CH2

CH2

Glu-Pro-Gly-Leu-NH-CH-CO-Ser-Trp-Gln-Ser-Leu-Arg*

S

S

810

776

842.11242.4

1276.5

1308.5

38

45

MALDI-PSD/CID yields characteristic triplets

4000

3000

2000

1000

0

Ion

Cou

nts

600500400300200100

m/z1200

800

400

0

Ion

Cou

nts

1000900800700600

m/z

400

300

200

100

0

Ion

Cou

nts

1600150014001300120011001000

m/z

Figure 9. Low energy CID of m/z 871.79 (most abundant ion in 9+ cluster), MW (monoisotopic) 7835.03 Da. This molecule was identifiedas disulfide linked peptides [443-496] and [519-531] of myosin heavy chain. Sequence and fragmentation are shown on the next page.

327.21

y3**y3**

y3**456.25

y4**

490.24y4 569.34

y5**

120.08F

201.12b2

637.32

y5 682.44y6**

763.10 (3+)b19

811.49y7**

796.13(3+)b20

872.18 (3+)

b22

706.41 (3+)b17

924.5y8**

1144.16 (2+)

1307.77 (2+)

1068.62

b9 b19b22

No characteristic triplets in ESI-CID

Fragment ions used for the identification of disulfide linked myosin peptides. (see low energy CID spectrum in Figure 9.)

b-ions myosin heavy chain peptide [443-496] 763.10 (3+) 1144.16 (2+) 796.13 (3+) 872.18 (3+) 201.12 1068.62 706.41 (3+) 1307.77 (2+) 2 9 17 19 20 22 VTRINQQLDTKQPRQYFGVLDIAGFEIFDFNSLEQLCINFTNEKLQQFFNHHhS 5 4 637.32 490.24 y-ions

myosin heavy chain peptide [519-531] DLAACIELIEKPhS 8 7 6 5 4 3 924.5 811.49 327.21 682.44 456.25 569.34 **y-ions

Disulfide BridgesThe ProteinProspector mass modification search can be used in conjunction with MS-Bridge to find peptides with disulfide bridges. For this example, mass shifts between 0-2000 Da were considered. (1023.3276+4).

KLSWADLIVFAGNCALESMGFK+4

VSFADLVVLGGCAAIEK

Glycosylationhttp://glycores.ncifcrf.gov/

Reference: Essentials of Glycobiology by Varki et al.

N-linked

AsnXxxSer/Thr/Cys

Further processing

N-linked glycosylation

• consensus sequence• GlcNAc2Man3 – coreoligomannose structure – just Man unitscomplex sugars– GlcNAc-Gal–SA

antennaehybrid structurescore fucosylationsulfate, phosphate modifications

• PNGase F removes all N-linked structures; Asn Asp

N-linked glycosylation

• Incredible heterogeneity: a site may be only partially occupied and may display numerous different carbohydrates

• species-, tissue-, cell-type-specific modification, physiological changes, diseases may alter the sugars

certain structures are immunogenic

Gal 1-3 capping, Fuc 1-3 on inner GlcNAc;blood group determinants

Characterization of N-linked carbohydrate pool

• PNGase digestion; sugar/protein separation; carbohydrate analysis without derivatization

* High pH anion exchange chromatography with pulsed amperometric detection

* NMR* MS• After modifying the reducing terminus(Advantage of the derivatization: improved “visibility”)* Electrophoresis with a combination of endoglycosidase cocktails “Elucidation of N-linked oligosaccharide structures of recombinant human factor VIII using

fluorophore-assisted carbohydrate electrophoresis.” Kumar HP, Hague C, Haley T, Starr CM, Besman MJ, Lundblad RL, Baker D. Biotechnol Appl Biochem. 1996 Dec;24 ( Pt 3):207-16.

* Capillary electrophoresis

No information on the modification sites and site heterogeneity

N-linked glycoproteins

• Periodate oxidation & color-reaction on gels

• Periodate oxidation & capture on hydrazide resin; release with PNGase F

• Western blot with protein- & sugar-specific antibody

Non-specific background; co-migrating proteins will interfere

N-linked glycopeptides(analysis of purified proteins only)

• Identification from diagnostic fragments:* HexNAc m/z 204* HexHexNAc m/z 366 Precursor scan, or „ping-pong” acquisition• Identification from oligosaccharide

heterogeneity• enrichment by HILIC or lectin-

chromatography

K.F. Medzihradszky Meth. Enzymol. 405, 116-138 (2005).

human lecithin:cholesterol acyltranferase and apolipoprotein D, tryptic digest, LC/MS analysis

MHTVNGYVN*R

Recombinant Factor VIII, 50 kDa subunit

AG(Man8GlcNAc2)NVSNIIPASATLNADVR

peptide+GlcNAc

About the structures of N-linked glycopeptides

• from the measured mass, and the CID spectrum the modified peptide can be identified + the size and class of the sugar

Automated N-glycopeptide identification using a combination of single- and tandem-MS. Goldberg D, Bern M, Parry S, Sutton-Smith M, Panico M, Morris HR, Dell A. J Proteome Res. 2007 Oct;6(10):3995-4005.

• the identity of the sugar units and their linkage positions CANNOT be determined

• NMR, exo- and endoglycosidases are needed to complete the job

[MHNa2]3+ of QV(Man10GlcNAc2)NIT and [MH2Na] 3+ of QV(Man9GlcNAc2)NITGK

Scheme 1. Fragments observed in the low energy CID spectrum of [MH2Na]3+ of glycopeptide QV(Man9GlcNAc2)NITGK (Figure 6)

+Na 2159.81 1080.41(2+) +Na 671.18 +Na Gln Man1-2Man1 1997.77 999.39(2+)

36

Man1 Val Man1

Man1 36

Man1-4GlcNAc-GlcNAc-Asn 6 Man1-2Man1- 2Man1 Ile +Na +Na Thr 163.05 2483.87 1242.44(2+) 1684.46 2321.85 1161.43(2+) 962.49 Gly +Na 347.08 Lys +Na 1360.41 +Na Man1-2Man1 1036.30

36

Man1 Man1

Man1 36

Man1-4GlcNAc 6 Man1-2Man1 - 2Man1 204.08 +Na 1522.41 +Na 1198.36

One component from the previous slide

K.F. Medzihradszky Meth. Enzymol. 405, 116-138 (2005).

O-linked sugars

• No consensus sequence• No common core structure• No universal enzyme-elimination works (NaOH)sugars have to be reduced upon release

Detection is problematic – because of heterogeneity & variable site occupancy

Site assignment is even harder

O-linked sugars

Other O-linked core structures

• FucHarris, R.J. & Spelmann, M.W. (1993) Glycobiology, 3, 219-224.

• GlcNishimura, H et al., (1989) J. Biol. Chem. 264, 20320-20325.• Man – in yeast____________________________________

• GlcNAc – single unit; INSIDE the cell

Characterization of O-linked carbohydrate pool

• Release with NaOH or hydrazynolysis• Reducing end is usually derivatized i)

to prevent peeling ii) to increase “visibility”

• Methods just like for N-linked sugars• NMR; HPAE-PAD; capillary and gel-

elfo; glycosidase cocktails; MS

No information on site occupancy and heterogeneity

O-linked glycopeptides

• Known protein - Elucidation of O-glycosylation structures of the

beta-amyloid precursor protein by liquid chromatography-mass spectrometry using electron transfer dissociation and collision induced dissociation. Perdivara I, Petrovich R, Allinquant B, Deterding LJ, Tomer KB, Przybylski M. J Proteome Res. 2009 Feb;8(2):631-42.

• Known sugar structure - Zsuzsanna Darula and Katalin

F. Medzihradszky “Affinity Enrichment and Characterization of Mucin Core-1 Type Glycopeptides from Bovine Serum” Mol. Cell. Proteomics, Nov 2009; 8: 2515 - 2526.

• Perseverance & luck - Crina I.A. Balog, Oleg M.

Mayboroda, Manfred Wuhrer, Cornelis H. Hokke, André M. Deelder, and Paul J. Hensbergen. “Mass spectrometric identification of aberrantly glycosylated human apolipoprotein C-III peptides in urine from Schistosoma mansoni-infected individuals.” MCP published January 13, 2010, 10.1074/mcp.M900537-MCP200

500

400

300

200

100

0

Inte

nsi

ty

12001000800600400200m/z

274

292

SA

453

SA-Gal 657

SA-Gal-GalNAc

697.4 751.2

812.8

944.6

1045.4

1126.4(2+)

(2+)

(2+)

-SA

-SA-Gal

-SA-Gal-GalNAc

(3+)(3+)

1888.2 + GalNAcGalSA

Typical CID of an O-linked glycopeptide

CID fragmentation of O-linked glycopeptides

J. Am. Soc. Mass Spectrom. 7, 1996, 319-328.

CID vs ETD

NH2-CH(R1)-CO-NH-CH(R2)-CO-…-NH-CH(Rn)-COOH

The weakest bonds go first;glycosidic bond; phosphate …

radical ions fragment, side-chains usually survive

b

y

c

z.

ETD of KTFMLQASQPAPT(GalNAcGalSA)HSSLDIK

Inter-alpha-trypsin inhibitor heavy chain H1 precursor

O-glycosylation within the cell

• Regulatory modification of nuclear and cytoplasmic proteins

• Single GlcNAc, no extension• 1 glycosyl-transferase (at least 21 for

secreted O-linked GalNAc); 1 cytosolic N-Acetyl--D-glucosaminidase

• Poorly understood due to lack of effective methods for enrichment and detection.

O-GlcNAc-detection/enrichment

1) Labeling with radioactive Gal and Gal-transferase

1/a) same enzyme + GalNAz (azido-derivative) + biotin-label + streptavidin - Enrichment and site mapping of O-linked N-acetylglucosamine by a combination of chemical/enzymatic tagging, photochemical cleavage, and electron transfer dissociation mass spectrometry. Wang Z, Udeshi ND, O'Malley M, Shabanowitz J, Hunt DF, Hart GW.Mol Cell Proteomics. 2010 Jan;9(1):153-60.2)

2) Immuno-precipitation with O-GlcNAc-specific antibodies

3) Lectin-affinity chromatography

Lectin-affinity Chromatography

•WGA lectin has affinity for GlcNAc, but affinity to a single GlcNAc moiety is low: millimolar1.

WGA

Good recoveryof glycoprotein

Complex glycosylation

GlcNAcGlcNAcGlcNAc

protein

No recovery

WGAPSVPVSerGSAPGR

O-GlcNAcO-GlcNAcylation

Load Wash Elute

Load Wash Elute

Peptide Mixture

Non-enriched

O-GlcNAc modified enriched

UV

abs

orba

nce

214

Time (minutes isocratic HPLC)

Solution?Isocratic chromatography on a long!!! column

Vosseller, K. et al. Mol Cell Proteomics (2006) 5 5: p.923-934

CID Analysis of O-GlcNAc-Modified Peptides

Chalkley, R. J. and Burlingame, A. L. J. Am. Soc. Mass Spectrom. (2001) 12 p.1106-1113

•O-glycosidic link is significantly more labile under CID conditions than peptide backbone.

•Modification site identification using CID often not possible.

ECD MS Spectrum of GlcNAc-modifiedPeptide from Spectrin

Vosseller, K et al. Mol Cell Proteomics (2006) 5 5: p.923-924

One of the most important regulatory events:

turns proteins on and off

induces or prevents other post translational

modifications in the same protein

signaling pathways : phosphorylation cascades

PhosphorylationBiological significance

Difficulties

a) Dynamic process : kinase vs. phosphatase

– both must be blocked during isolation

b) Phosphorylation often @ low level (<5%)

c) Lower ionization efficiency – signal of

phosphopeptides suppressed

Enrichment is a must at protein level

at peptide level

Confirming the presence of phosphorylation

Western blot

pTyr large enough for sequence independent

recognition - works well

pSer, pThr – not reliable

Dyes – questionable reliability

Phosphatase treatment + isoelectric focusing – pI

shift

In vitro/in vivo assay with radioactive phosphate

Determining site of phosphorylation

Mikrosequencing

Mutation studies

Mass spectrometry

MS spectrum : + 80 Da shift

MS/MS fragmentation :

pSer, pThr - H3PO4 loss : –98 Da

pTyr – always retains the modification

immonium ion at m/z 216 Da

Enrichment methods

Ion exchange on SCX

IMAC : Fe(3+), Ga(3+)…

binding at low pH

methyl-esterification prior to IMAC

TiO2, ZrO2

Immunoprecipitation (only for pTyr)

Reproducible isolation of distinct, overlapping segments of the phosphoproteome

 

                                                                                                                        

               Overlap between phosphopeptide isolation methods on the level of identified phosphorylation sites.

B. Bodenmiller et al, Nat. Methods 2007

Large scale vs. 1 protein

large scale phosphorylation studies: large amounts of a complex mixture is

analyzed compensating for losses during sample preparation

leads to more PTM identification but low end results are incidental

single protein samples:sample amount is usually limitedmore challengingcombination of techniques may be necessary

MALDI-TOF MS without enrichment

1200 1700 2200 m /z

0

5000

10000

15000

20000

25000

a.i.

/O = /bruker/2006/06novem ber/ke061110_02/2S R ef/pdata/1 A dm inistrator F ri N ov 10 11:56:05 2006

P P

PP

Enrichment with TiO2

1400 1900 2400 m /z

0

10000

20000

30000

40000

50000

60000

70000

a.i.

/O = /bruker/2006/06novem ber/ke061109_01/2S R ef/pdata/1 A dm inistrator F ri N ov 10 10:15:26 2006

P

P

P

P

P P

P

P

P

CID of CDSSPDpSAEDVR140x10

3

120

100

80

60

40

20

0

Inte

nsity

800600400200m/z

40x103

30

20

10

0

Inte

nsity

140012001000800600400200m/z

456

.7 M

-98

3+

518

.3 y

4

389

.3

b3

324

.1 b

2

274

.3 y

2

213

.1

PD

175

.1 y

1 435

.9 y

8-9

82+

597

.2 P

Dp

SA

ED

-98

695

.2 P

Dp

SA

ED

498

.2 b

4

871

.4 y

7

477

.3 M

## 3

+

109

4.4

b

10-9

8

870

.5 y

7

968

.4 y

8

119

2.4

b

10

389

.2 y

3

324

.1 b

2

274

.2 y

2

462

.2 b

42#

715

.4 M

## 2

+

136

0.6

M

-py

684

.4 M

-98

2+

126

2.5

M

-py-

98

756

.3 y

6

658

.4 y

6-9

8m/z 733.7 (2+)

m/z 489.6 (3+)

Some new resultsSci Signal. 2010 Jan 12;3(104):ra3. “Quantitative phosphoproteomics reveals widespread full

phosphorylation site occupancy during mitosis.”Olsen JV, Vermeulen M, Santamaria A, Kumar C, Miller ML, Jensen LJ, Gnad F, Cox J, Jensen TS, Nigg EA, Brunak S, Mann M.

“… quantified 6027 proteins and 20,443 unique phosphorylation sites and their dynamics.”

Some other tricks to study phosphorylation

• Shokat-method: utilizing analog-specific kinases

“Design and use of analog-sensitive protein kinases.” Blethrow J, Zhang C, Shokat KM, Weiss EL. Curr Protoc Mol Biol. 2004 May;Chapter 18:Unit 18.11.

• ERLIC-chromatography – single step enrichment & fractionation of acidic/hydrophilic modifications

“Simultaneous characterization of glyco- and phospho-proteomes of mouse brain membrane proteome with electrostatic repulsion hydrophilic interaction chromatography (ERLIC).” Zhang H, Guo T, Li X, Datta A, Park JE, Yang J, Lim SK, Tam JP, Sze SK. Mol Cell Proteomics. 2010 Jan 4. [Epub ahead of print]