Y l U i it S h l f M di iYale University School of MedicineW.M. Keck Foundation

Biotechnology Research Laboratory

Christopher Colangelo, Ph.D.

Quantitative Protein Profiling &Quantitative Protein Profiling &Quantitative Protein Profiling & Quantitative Protein Profiling & Targeted Protein Expression Targeted Protein Expression

NIDA NIDA External Advisory Board (EAB) MeetingExternal Advisory Board (EAB) Meeting

December 3, 2008December 3, 2008

Goals of Protein Profiling Core

• Improve and develop new protein profiling technologies to p p p p g gidentify proteins that play key roles and/or are biomarkers in response to substance abuse

• 1 Use existing DIGE MudPIT iTRAQ and SILAC pipelines1. Use existing DIGE, MudPIT, iTRAQ, and SILAC pipelines on larger number of samples.

• 2. Better Protein Identification• 3. More Accurate Protein Quantitation (replicates)• 4. Faster Sample Turnaround• 5 Verification of potential markers (Targeted Proteomics)5. Verification of potential markers (Targeted Proteomics)

MudPITProtein ID& PTM

Phosphoprotein Profiling

Discovery

Protein ID

TiO2

PP

2D-ProteinSeparation DIGEiTRAQ

ICATSILAC

Quantitative Proteomics

MRM T t d P t iTargeted Proteomics Verification/

Validation

Validation

20

BioanalyticalQuantitation(small molecule)

Bioanalytical

Small Molecule & Med. Chem. MS

0C

onc

Bioanalytical

Growth of NIDA iTRAQ Protein Profiling Service35

64%

30

35

i ( l )68%

20

25

ples

iTRAQ (4‐plex)iTRAQ (8‐plex)

61%

15

20

er of S

amp

50%

57%

61%

5

10

Num

b

100%

0

5

2005 (April ‐ Dec) 2006 2007 20082005 (April  Dec)  2006 2007 2008Year

Red Values = Percentage (%) of NIDA samples/Total samples run during calendar year

iTRAQ® Reagents - 8plex

Isobaric Tag: Total mass = 305

Reporter Group

N

N

O

ON

O

Balance GroupPeptide Reactive

N N N N113 114 115 116

OPeptide Reactive

Group (PRG)

N N N13

2CHN

15

+

N N+ + +

N

117 118 119 121N

133CH

15

+++ +

N15

iTRAQ® Reagents Workflow

8 samplesidentical m/z

Dig

est

Mix MS113

114

192

191

PRG

PRG

+

+

+

+S1

S2S3

S4

S5 S7 Den

atur

e &

D 114

115

116

117

118

191

190

189

188

187

PRG

PRG

PRG

PRG

PRG

+

+

+

+

+

+

+

+

+

+

1197.61

S5S6

S7S8

Para

llel

D 118

119

121

187

186

184

PRG

PRG

PRG

+

+

+

+

+

+MS/MS

iTRAQ® Reagents

[Figure used by permission from ABI]

Workflow for iTRAQ Protein Profiling Analysis

Control Treated

100 ug of control/treated samples are reduced, alkylated, and digested with trypsin

iTRAQ labeling

Control Treated

114 115 116 117

g

SCX

2D chromatographicti f tid

Strong cation exchange

RP

separation of peptidesReverse-phase

QSTAR Elite Mass Spectrometer

Yale Protein Expression Database (YPED)

Database Searching and Quantitation

Yale Protein Expression Database (YPED): Key FeaturesKey Features

Web-based Client Built on Open-Source PlatformWeb based Client Built on Open Source Platform Web-based viewing of results

Data and Technology IntegrationData Upload from multiple technologies integration of theData Upload from multiple technologies, integration of the results, enables data review

Data Archive and AnalysisIntuitive Web interface for users to query analyze and visualizeIntuitive Web interface for users to query, analyze and visualize dataAllows comparison of results across samples

Public Repository and Data Preservation (Sept 2008)Public Repository and Data Preservation (Sept 2008)Stores and Exports data in industry standard format (MIAPE)Chosen by Yale Digital Landscape as the best Yale scientific database for development and beta testing

“YPED: a web-accessible database system for protein expression analysis.”Shifman, M.A., Li, Y., Colangelo, C.M., Stone, K.L., Wu, T.L., Cheung, K.H., Miller, P.L., Williams, K.R. J Proteome Research 2007, 6(10), 4019-4024.

Postsynaptic Density (PSD)

• Multi-protein complex organized into functional assemblies

• Site of synaptic plasticity• Receptor complexes localized by scaffold proteins, which

links to signaling moleculeslinks to signaling molecules• Proteomic studies of the rodent PSD reveals diverse

classes of proteins (Li et al. 2004, Jordan et al. 2004, Yoshimura et al. 2002, Collins et al. 2005), , )

• Large-scale changes to the PSD after drug exposure is poorly studied

iTRAQ Workflow for Rat Brain PSD samples

Saline treated Exposed: 30 mg/kg cocaineSaline-treated Exposed: 30 mg/kg cocaine

Cortex Striatum Hippocampus Cortex Striatum Hippocampus

iTRAQ Analysis

iTraq Summary of PSD analysis

• Cortical PSD:Cortical PSD:– 323 proteins identified– 29 proteins , 63 proteins

Cortical Hippocampal2760 127

• Striatal PSD:– 560 proteins identified– 57 proteins , 67 proteins

175

25160

57 proteins , 67 proteins

• Hippocampal PSD:– 580 proteins identified Striatal

112

– 33 proteins , 45 proteins

YPED Data Results 8-plex iTRAQ on Cortical, Striatal, and Hippocampal PSD from naïve rats

Principal Component Analysis on 8‐plex iTRAQ dataset from  Mouse Brain PSD from various brain regions. 

StriatumStriatum

i

Cortex

Hippocampus

iTRAQ ProteinPilot False Discovery Rate

8-plex iTRAQ on Cortical, Striatal, and Hippocampal PSD from naïve ratsfrom naïve rats

Scatter plot for the cortical PSD

11611

118

121119Calmodulin peptide

Hippocampal PSD

33.5

114 113 ti

Cortical PSD Striatal PSD

Scatter plot for the cortical PSD replicates

113

114

116 115 117

11.5

22.5

114:113 ratio

115:113 ratio

Additional calmodulin peptides were reviewed to substantiate this expression change.

00.5

1

0 100 200 300p g

Ratios shows that the expected deviation is ±20%

Upgraded Protein Identification and Quantitation Software

• Upgraded to MASCOT Cluster version 2.2 • Offers improved Phosphopeptide Identification• Migrated MASCOT to an duo quadcore LINUX cluster• Migrated MASCOT to an duo quadcore LINUX cluster• This allows processing on 8 nodes, which was needed given the

additional data generated from the QSTAR Elite, LTQ-Orbitrap, and 4000 QTRAP mass spectrometersOrbitrap, and 4000 QTRAP mass spectrometers

• ProteinPilot 2.0.1• Protein Identification and Quantitation software• Enables simultaneous identification and quantitation for

ICAT™, iTRAQ™, and SILAC™ reagents• Performs Protein Grouping and Calculates False Discovery Rates

• Mascot Distiller – Quantitation Toolbox (SILAC)• We were a beta test site for the MASCOT Distiller Quantitation

T lb d h l d h d l t f SILAC b dToolbox and helped push development for SILAC based quantitation

Targeted Proteomics Platform

• Utilizes Multiple Reaction Monitoring (MRM)

• Uses– Utilizes information from MS based discovery– Rapid develop of an assay to monitor proteinRapid develop of an assay to monitor protein

dynamics (eg. Expression changes or PTMs in experimental systems)

• AdvantagesAdvantages– High dynamic ranges (105)– Hundreds of samples– Quantitation – Relative or Absolute (heavy

idi t i tid )ABI 4000 QTRAP and idiotypic peptides)ABI 4000 QTRAP and Waters NanoAcquity UPLC

Yale Center for Clinical Investigation

Schematic of the ESI Triple Quadrupole MS System in MRM Operation Mode

Fragment Filter

MS Orifice y12

1243.8+ &1236 8+Filter

p

All Ionse-

gpeptides

840.9++ & 837.5++

fragment

Electrospray Ion Detector

1236.8+peptides

Ion Guide

CollisionCell

Q1st

AnalyticalQuadrupole

AmplifiedSignal

(current)

3rd

AnalyticalQuadrupole

Fragment peptideSelect Peptide Select Fragment Detect Fragment

– Highest specificity and sensitivity for detecting components in a complex mixture– Largest linear dynamic range for quantitation– Requires triple quadrupole MS capability– Well accepted as the MS technique for quantitation (Pharmaceutical Industry)

Targeted Proteomics Workflow

Peptide synthesisHPLC purification

Proteinmixture

Enzymatic Digest

Heavy idiotypic peptide

pAmino acid analysis

LC-MRMChromatogramMix

Heavy Internal MRM

Analysis

Heavy Internal Standard

Analyte

Schematic of Multiple Reaction Monitoring Scan (MRM)

YPED Synthetic Peptide Database

MRM targeted Proteomics of Rho GEF LFC Ser646 (WIL11613 )

Heavy internal standard peptide calibration curve for the 534.3/849.4 and 534.3/720.4 MRM transitions of ser 646 peptide L*ESpFESLR (L* = Heavy Leucine Amino Acid (13C6, 15N),

Sp = phosphoserine)

ser 646L*ESpFESLR

534 3 / 849 4 (y7 H3PO4 )

534.3 / 720.4 (y6‐H3PO4)

Sp = phosphoserine)534.3 / 849.4 (y7‐H3PO4 )

5342+ 849+

(Internal Std) 5342+ 720+ 

(Internal Std)Heavy Internal Standard 

L*ESpFESLRp

5302+  849+ 

(LFC ctrl)5302+ 720+ 

(LFC ctrl)LFC control mouse brain sample

ser 646LESpFESLR

Integration of MultiQuant Results in YPED

Small Molecule Bioanalytical Chemistry

Plasma levels of Lipitor in Mouse Plasma18

12

14

16

g/m

l)8

10

cent

ratio

n (n

gPE Series 275 HRes LC System

2

4

6

Con

c

0

1 hr, 10 mg/kg

1 hr, 30 mg/kg

1hr, 100 mg/kg

24 hr, 10 mg/kg

24 hr, 30 mg/kg

24 hr, 100 mg/kg

API 5000Yi T, Rao DA, Tang PC, Wang Y, Cuchara LA, Bothwell AL, Colangelo

CM, Tellides G, Pober JS, Lorber MI. Transplantation. 2008 Sep 15;86(5):719-27.

Acknowledgements

Ken Williams (Director)Kathy Stone Erol Gulcicek T Kiet LamTuKiet Lam Terence Wu

Mary Lopresti Tom AbbottJean KanyoJean Kanyo

Kathrin Wilczak-Havill Edward Voss

Matt Berberich Mark ShifmanMark Shifman Ji Young LeeHongyu Zhao Erika Andrade

Alex StipanovichAlex Stipanovich

Yale Center for Clinical Investigation