HTP Construct Optimization using Bioinformatics Coupled with Amide Hydrogen Deuterium Exchange (DXMS) and HTP NMR screening

Yuanpeng (Janet) Huang

Northeast Structural Genomics Consortium(NESG)

Rutgers University

PSI-1 NESG-RUTGERS HUMAN PROTEIN PRODUCTION in E. Coli (2000.7.1-2005.7.1)

Total PDB/Cloned Targets = 1.4%

Cloned Target: 905 Analytical Expressed & Soluble: 215Purification:    224Aggregation:     33Crystal Trial: 63NMR Screening:    51PDB (Xray, NMR): 6,7

PSI-2 BIOMEDICAL THEME – Human Cancer Protein Interaction Network (HCPIN)

Huang, et al (2008) Targeting the human cancer pathway protein interaction network by structural genomics Molecular & Cellular Proteomics 7: 2048-2060

EFFORTS TO IMPROVE PSI-2 HUMAN PROTEIN PRODUCTIVITY

Target Selection Select proteins validated by SwissProt Exclude proteins annotated or predicted to be

secreted or TM Gene synthesis and RT-PCR Construct Design and Optimization

Identify disordered regions by DisMeta prediction DXMS analysis

SG of extracellular & membrane-bound HCPIN targets

-- Chiang, Rossi, Gurla, Montelione, & Anderson, in preparation.

EFFORTS TO IMPROVE PSI-2 HUMAN PROTEIN PRODUCTIVITY

Target Selection Select proteins validated by SwissProt Exclude proteins annotated or predicted to be

secreted or TM Gene synthesis and RT-PCR Construct Optimization

Identify disordered regions by DisMeta prediction DXMS analysis

SOME PARTIALLY DISORDERED PROTEIN STRUCTURES SOLVED BY NESG

Interfere with Structural Determination Efforts

Identify disordered regions

DisMeta - Disorder Prediction MetaServer

DXMS - 1H/2H exchange mass spectrometry

www-nmr.cabm.rutgers.edu/bioinformatics/disorder

Secondary Structure Prediction

Disorder Prediction Server Results

Summary of Disorder predictions

SyR11

Residue 50

Disorder Prediction MetaServer

1-155 full length

50-155truncated

difference

Digest

Digest

LC-MS

LC-MS

H2O

D2O

m/z574.5 575.0 575.5 576.0 576.5 577.0 577.5 578.0 578.5 579.0 579.5 580.0

575.76

576.24

576.76

577.26577.78

0

100

Rela

tive A

bund

ance

Non-deuterated

m/z574.5 575.0 575.5 576.0 576.5 577.0 577.5 578.0 578.5 579.0 579.5

576.30

576.78

577.28

575.82

577.78

578.24578.74579.240

100

Rela

tive A

bund

ance

Deuterated

Peptide Mass shift based on D2O exposure duration

H/D Exchange MS: Concept

Quench

(pH ~2.5, -80°C)

(on ice)Centroid of peak = 576.7

Centroid of peak = 576.13

Sharma, et al Construct optimization for protein NMR structure analysis using amide hydrogen/deuterium exchange mass spectrometry, Proteins 2009 (in press)

H. Zheng

WR33 DXMS Analysis

> 50%

< = 25%

> 25%

> 60%

> 70%

> 80%

Mouse Homologue, Kobayashi N., et.al

C. elegans, WR33, NESG,

R. Tejero, J. Aramini

CNN

C

Human Homologue, HR387, NESG,

J. Aramini

C

Target

HTP Construct Design (DisMeta Prediction)

Multiple Alternative Constructs

ProteinProduction

NMR Screening

Xtal Screening

PROTEIN PRODUCTION PROTOCOL

Construct OptimizationBy DXMS/DisMeta

HTP Human Protein Construct Design

pdb hits (>80%)

2. Find multiple target regions(DisMeta prediction)

3. Propose alternative constructs

1. Select target domains•PDB hit regions (<80% seq. id) •PFAM domains

pdb hits (<80%) PFAM

Total number of constructs (domain) ≈ # of target regions × # of alternative constructs

target region

disordered

Propose Alternative Constructs

Identify multiple target regions (TRs) for each target domain All TRs with length > 50aa and cover at least 80% of

the target domainAdjust TR for disorder and helix/strand regions

For each TR (S,E)Propose 1-4 alternative constructs (S-5, S)x(E, E+5)Remove the ones intersect with helix/strand Adjust N/C ends (-2:2)

target domain

disordered

target domain

helix✗

Total number of constructs (domain) ≈ # of target regions × # of alternative constructs

Target

HTP Construct Design (DisMeta Prediction)

Multiple Alternative Constructs

ProteinProduction

NMR Screening

Xtal Screening

PROTEIN PRODUCTION PROTOCOL

Construct OptimizationBy DXMS/DisMeta

NMR SALVAGE PROTOCOL

ER553

61 - 199 + C tag 63 - 199 + C tag 59 - 199 + C tag 65 - 199 + C tag57 - 199 + C tag

ER541

1-103 37-162

1-146Full length

1-170Full length

Salvage using DisMeta predictions

Failed Competition

BjR38

ER553

SaR32 VpR68

LkR15

Salvage using DX-MS results

1-207 1-137

1-136 1-10759-199 10 – 90

Micro-probe

PROGRESS ON PSI-2 RUTGERS HUMAN PROTEIN PRODUCTION (2005.7.1-PRESENT)

Total PDB/Cloned Targets (Constructs) = 3%(1.5%)

Cloned Target (Construct): 367 (734)Analytical Expressed & Soluble: 216Purification:    132Aggregation:     74Crystal Trial: 49NMR Screening (Good, Promising):    55(11, 12)PDB (Xray, NMR): 7,4

PROGRESS ON PSI-2 RUTGERS HUMAN PROTEIN PRODUCTION (2005.7.1-PRESENT)

Total PDB/Cloned Targets (Constructs) = 3%(1.5%)

Cloned Target (Construct): 367 (734)Analytical Expressed & Soluble: 216Purification:    132Aggregation:     74Crystal Trial: 49NMR Screening (Good, Promising):    55(11, 12)PDB (Xray, NMR): 7,4

HTP robotic NMR micro cryoprobe screening using microgram quantities of protein

Protein samples inMicrotubes assessed and scored prior to loading theautomatic sample changer

B600 with samples loadedfor data collection

1D proton spectrum with water suppressionassess Signal-to-noise, foldedness of protein

2D NH-HSQC spectrum to evaluatethe feasibility for structure determination

Archieval of the raw data along with thespectral images, quality scores and stabilityinto SPINE database.

Setup

&

Run

Data

Archival

Virtual 96-well

SPINE-based Tools

Bruker Icon-NMR

GVT Swapna

HTP Buffer Optimization

MJ001 MJ002 MJ003 MJ004 MJ005 MJ006 MJ007 MJ008 MJ009 MJ010 MJ011 MJ012

A CcR27.008 CcR27.009 CcR27.010 CcR27.011 CcR27.012 CcR27.013 CcR27.014 CcR27.015 CcR27.016 CcR27.017 CcR27.018 CcR27.019

B StR82.012 StR82.013 StR82.014 StR82.015 StR82.016 StR82.017 StR82.018 StR82.019 StR82.020 StR82.021 StR82.022 StR82.023

C SaR31.005 SaR31.006 SaR31.007 SaR31.008 SaR31.009 SaR31.010 SaR31.011 SaR31.012 SaR31.013 SaR31.014 SaR31.015 SaR31.016

D VpR162.008 VpR162.009 VpR162.010 VpR162.011 VpR162.012 VpR162.013 VpR162.014 VpR162.015 VpR162.016 VpR162.017 VpR162.018 VpR162.019

E MaR182.008 MaR182.009 MaR182.010 MaR182.011 MaR182.012 MaR182.013 MaR182.014 MaR182.015 MaR182.016 MaR182.017 MaR182.018 MaR182.019

F SvR198.008 SvR198.009 SvR198.010 SvR198.011 SvR198.012 SvR198.013 SvR198.014 SvR198.015 SvR198.016 SvR198.017 SvR198.018 SvR198.019G IR12.004 IR12.005 IR12.006 IR12.007 IR12.008 IR12.009 IR12.010 IR12.011 IR12.012 IR12.013 IR12.014 IR12.015

MJ001 MJ002 MJ003 MJ004 MJ005 MJ006 MJ007 MJ008 MJ009 MJ010 MJ011 MJ012

A p p p

B p p P P P P P P p p P

C p p P p P p p p p

D p p P P P P P p p P

E p p p P p p p

F p p P P p P p p P p

G p p P P P p P p p

1D Proton Experiment Good = Perform 2D HSQC Bad = Stop WorkMJ001 MJ002 MJ003 MJ004 MJ005 MJ006 MJ007 MJ008 MJ009 MJ010 MJ011 MJ012

A p p p

B p p P P P P P P p p P

C p p P p P p p p p

D p p P P P P P p p P

E p p p P p p p

F p p P P p P p p P p

G p p P P P p P p p

Proteins with Good HSQCPrecipitation (button testing)

Robotic screening using12 Buffers varying pH, NaCl, Arginine, Acetonitrile, Zn, Ca, Detergent among others.

Clear Cloudy precipitated

Twice the time using 1/20th of the sample

1 mm micro probe and 1.7 mm micro cryoprobe

It is now routinely used in NESG HTP structure production pipeline

• low yield eukaryotic proteins

3D Structure Determination using microgram quantities of protein

Aramini et al (2007) Microgram-scale protein structure determination by NMR. Nature Methods. 4:491-3

SUMMARY Current protocol of construct optimization is focused

on identification of disorder regions DisMeta

Very fast and no-cost HTP construct design protocol is developed together with DisMeta

DXMS More reliable, experimental evidence on disordered regions Useful for identification of disordered regions when the prediction

is not satisfactory Automated analysis of DXMS data is under development

Human protein production in E. Coli is improved NMR micro cryoprobe Efficient and cost-effective HTP robotic NMR screening NMR Structure determination become feasible for human

proteins with low expression and low solubility

ACKNOWLEDGEMENTS

Salvage by DXMSWill BuchwaldAsli ErtekinSeema SharmaHaiyan ZhengPeter Lobel

BioinformaticsJohn EverettJessica LockeBinchen MaoSai Tong

Protein Production Thomas ActonLi-Chung MaRitu ShastryGVT SwapnaRong XiaoLi Zhao

Gaetano T. Montelione