110/15/07BCB 444/544 F07 ISU Dobbs #23 - Protein Tertiary Structure Prediction BCB 444/544 Lecture...

1BCB 444/544 F07 ISU Dobbs #23 - Protein Tertiary Structure Prediction 10/15/07

BCB 444/544

Lecture 23

Protein Tertiary Structure Prediction

#23_Oct15


Mon Oct 15 - Lecture 23


• Chp 15 - pp 214 - 230

Wed Oct 17 & Thurs Oct 18 - Lecture 24 & Lab 8

(Terribilini)

RNA Structure/Function & RNA Structure Prediction

• Chp 16 - pp 231 - 242

Fri Oct 18 - Lecture 25

Gene Prediction • Chp 8 - pp 97 - 112

Required Reading (before lecture)


New Reading & Homework Assignment

ALL: HomeWork #4 (emailed & posted online Sat AM)

Due: Mon Oct 22 by 5 PM (not Fri Oct 19) Read:

Ginalski et al.(2005) Practical Lessons from Protein Structure Prediction, Nucleic Acids Res. 33:1874-91. http://nar.oxfordjournals.org/cgi/content/full/33/6/1874 (PDF posted on website)

• Although somewhat dated, this paper provides a nice overview of protein structure prediction methods and evaluation of predicted structures.

• Your assignment is to write a summary of this paper - for details see HW#4 posted online & sent by email on Sat

Oct 13

http://mi.caspur.it/PMDB/help.php


Seminars Last Week

Dr. Klaus Schulten (Univ of Illinois) - Baker Center

Seminar The Computational Microscope

2:10 PM in E164 Lagomarcino http

://www.bioinformatics.iastate.edu/seminars/abstracts/2007_2008/Klaus_Schulten_Seminar.pdf

• Check out links on Schulten's website (videos, etc) • http://www.ks.uiuc.edu/~kschulte/

• Great seminar - amazing simulations of dynamics in proteins and large macromolecular assemblies

• Very computationally intensive - very impressive demonstration of power of computation to produce insights not attainable using only experimental approaches

http://www.oxfordjournals.org/nar/database/summary/855



http://www.rcsb.org/pdb/search/searchModels.do


Seminars this Week

BCB List of URLs for Seminars related to Bioinformatics:http://www.bcb.iastate.edu/seminars/index.html

• Oct 18 Thur - BBMB Seminar 4:10 in 1414 MBB • Sachdeve Sidhu (Genentech) Phage peptide and

antibody libraries in protein engineering and ligand selection

• Oct 19 Fri - BCB Faculty Seminar 2:10 in 102 ScI• Lyric Bartholomay (Ent, ISU) TBA

http://expasy.org/tools/

http://www.bcb.iastate.edu/courses/BCB691-F2007.htm#Oct%2019


Protein Sequence & Structure: Analysis

• Diamond STING Millennium - Many useful structure analysis tools, including Protein Dossier http://trantor.bioc.columbia.edu/SMS/

• SwissProt (UniProt)Protein knowledgebase

http://us.expasy.org/sprot

• InterProSequence analysis tools

http://www.ebi.ac.uk/interpro

http://predictioncenter.gc.ucdavis.edu/

http://www.ks.uiuc.edu/~kschulte/




http://www.bcb.iastate.edu/seminars/index.html



Chp 14 - Secondary Structure Prediction

SECTION V STRUCTURAL BIOINFORMATICS

Xiong: Chp 14

Protein Secondary Structure Prediction

• √Secondary Structure Prediction for Globular Proteins

• √Secondary Structure Prediction for Transmembrane Proteins

• √Coiled-Coil Prediction


Where Find "Actual" Secondary Structure? In the PDB


How Does Predicted Secondary Structure Compare with Actual? (An example)

Query MAATAAEAVASGSGEPREEAGALGPAWDESQLRSYSFPTRPIPRLSQSDPRAEELIENEEGOR V CCCCHHHHHHHHCCHHHHHHCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCHHHHHHCCCCFDM CCCCCCCCCCCCCCCCCEECCCCCCCCCHHHCCCCCCEECCCCCCCCCCHHHHHHHHCCCCDM CCCCHHHHHHCCCCCCCEECCCCCCCCCHHHCCCCCCEECCCCCCCCCCHHHHHHHHCCC

DSSPAuthor

Predicted - Using 3 methods (from CMD server, Jernigan Group, ISU)

Actual - Calculated from PDB coordinates by DSSP or author:


Chp 15 - Tertiary Structure Prediction


Xiong: Chp 15


• Methods• Homology Modeling• Threading and Fold Recognition• Ab Initio Protein Structural Prediction• CASP


Structural Genomics - Status & Goal

~ 20,000 "traditional" genes in human genome (recall, this is fewer than earlier estimate of

30,000)

~ 2,000 proteins in a typical cell> 4.9 million sequences in UniProt (Oct 2007)

> 46,000 protein structures in the PDB (Oct 2007)

Experimental determination of protein structure lags far behind sequence determination!

Goal: Determine structures of "all" protein folds in nature, using combination of experimental structure determination methods (X-ray crystallography, NMR, mass spectrometry) & structure prediction


Structural Genomics Project

TargetDB: Database of Structural Genomics

Targets

http://targetdb.pdb.org

http://nar.oxfordjournals.org/cgi/content/full/33/6/1874


PMDB: Protein Model Database http://mi.caspur.it/PMDB/help.php

also, via NAR's Molecular Biology Database Collection http://www.oxfordjournals.org/nar/database/summary/855

Database of Theoretical Structures?

Theoretical structural models (predicted) are no longer accepted by the PDB (since 10/15/06); but, it is possible to search for models deposited earlier:

http://www.rcsb.org/pdb/search/searchModels.do

http://www.ent.iastate.edu/dept/faculty/bartholomay/

http://www.ent.iastate.edu/dept/faculty/bartholomay/

http://en.wikipedia.org/wiki/List_of_protein_structure_prediction_software


Protein Structure Prediction or Protein Folding Problem

"Major unsolved problem in molecular biology"

In cells: spontaneousassisted by enzymesassisted by chaperones

In vitro: many proteins can fold to their "native" states spontaneously & without assistance

but, many do not!


Deciphering the Protein Folding Code

• Protein Structure Prediction or• Protein Folding Problem

Given the amino acid sequence of a protein, predict its 3-dimensional structure (fold)

• Inverse Folding ProblemGiven a protein fold, identify every amino acid sequence that can adopt its 3-dimensional structure


Protein Structure Prediction

Structure is largely determined by sequence

BUT:• Similar sequences can assume different structures• Dissimilar sequences can assume similar structures• Many proteins are multi-functional 2 Major Protein Folding Problems:

1- Determine folding pathway 2- Predict tertiary structure from sequence

Both still largely unsolved problems


Steps in Protein Folding

1- "Collapse"- driving force is burial of

hydrophobic aa’s (fast - msecs)

2- Molten globule - helices & sheets form, but

"loose" (slow - secs)

3- "Final" native folded state - compaction

& rearrangement of 2' structures

Native state?- assumed to be lowest free energy- may be an ensemble of structures


Protein Dynamics

• Protein in native state is NOT static

• Function of many proteins requires conformational changes, sometimes large, sometimes small

• Globular proteins are inherently "unstable"

(NOT evolved for maximum stability)

• Energy difference between native and denatured state is very small (5-15 kcal/mol)

(this is equivalent to ~ 2 H-bonds!)

• Folding involves changes in both entropy & enthalpy


Difficulty of Tertiary Structure Prediction

Folding or tertiary structure prediction problem can be formulated as a search for minimum energy conformation

• Search space is defined by psi/phi angles of backbone and side-chain rotamers

• Search space is enormous even for small proteins!

• Number of local minima increases exponentially with number of residues

Computationally it is an exceedingly difficult problem!


Tertiary Structure Prediction Methods

2 (or 3) Major Methods:1. Comparative Modeling:

• Homology Modeling (easiest!) • Threading and Fold Recognition (harder)

2. Ab Initio Protein Structural Prediction (really hard)


Comparative Modeling?

Comparative modeling - term is

sometimes used interchangeably with homology modeling, but also sometimes used to mean both:

• homology modeling

• threading/fold recognition


Ab Initio Prediction

1. Develop energy function

• bond energy• bond angle energy• dihedral angle energy• van der Waals energy• electrostatic energy

2. Calculate structure by minimizing energy function • usually Molecular Dynamics (MD) or Monte Carlo (MC)

Ab initio prediction - impractical for most real (long) proteins• Computationally? very expensive• Accuracy? Usually poor for all except short peptides

(but much improvement recently!)

Provides both folding pathway & folded structure


Comparative Modeling

Provide folded structure only

Two types:

1) Homology modeling

2) Threading (fold recognition)

Both rely on availability of experimentally determined structures that are "homologous" or at least structurally very similar to target


Homology Modeling

1. Identify homologous protein sequences (-BLAST)2. Among available structures (in PDB), choose one

with closest sequence to target as template(can combine steps 1 & 2 by using PDB-BLAST)

1. Build model by placing target sequence residues in corresponding positions on homologous structure & refine by "tweaking" modeled structure (energy minimization)

2. Homology modeling - works "well"1. Computationally? "relatively" inexpensive2. Accuracy? higher sequence identity better

model

1. Requires ~30% sequence identity with sequence for which structure is known


Threading - Fold RecognitionIdentify “best” fit between target sequence & template structure

1. Develop energy function2. Develop template library3. Align target sequence with each template in library &

score4. Identify top scoring template (1D to 3D alignment)5. Refine structure as in homology modeling

Threading - works "sometimes"1. Computationally? Can be expensive or cheap, depends

on energy function & whether "all atom" or "backbone only" threading is used

2. Accuracy? in theory, should not depend on sequence identity (should depend on quality of template library & "luck")

Usually, higher sequence identity to protein of known structure better model


Threading: the Motivation

• Basic premise:

• Statistics from Protein Data Bank (>46,000 structures)

• Thus, chances for a protein to have a native-like structural fold in PDB are quite good

• Note: Proteins with similar structural folds could be either homologs or analogs

The number of unique structural folds in nature is fairly small (probably 2000-3000)

Prior to Structural Genomics Project, 90% of "new" structures submitted to PDB were similar to existing folds in PDB - suggesting that almost all folds in nature have been identified


1. Align target sequence with template structures

in fold library (usually from the PDB)

2. Calculate energy score to evaluate "goodness of fit" between target sequence & template structure

3. Rank models based on energy scores

Target Sequence

Structure Templates

ALKKGF…HFDTSE

Steps in Threading


Threading Goal - & Issues

• Structure database - must be "complete"

• Can't build a good model if there is no good template in library!

• Sequence-structure alignment algorithm:

• Bad alignment Bad score!

• Energy function or Scoring Scheme:

• Must distinguish correct sequence-fold alignment from incorrect sequence-fold alignments

• Must distinguish “correct” fold from close decoys

• Prediction reliability assessment - How determine

whether predicted structure is correct? (or even close?)

Find “correct” sequence-structure alignment of a target sequence with its native-like fold in template library (usually derived from PDB)


Threading: Template database

• Build a database of structural templates e.g., ASTRAL domain library derived from the

PDB

Sometimes, supplement with additional decoys e.g., generated using ab initio approach such as Rosetta (Baker)


Threading: Energy function

• Two main methods (& combinations of these)

• Structural profile (environmental) physicochemical properties of amino acids

• Contact potential (statistical) based on contact statistics from PDB

famous one: Miyazawa & Jernigan (ISU)


Protein Threading: Typical energy function

How well does a specific residue fit structural environment?

What is "probability" that two specific residues are in contact?

Alignment gap penalty?

Total energy: Ep + Es + Eg

Goal: Find a sequence-structure alignment that minimizes energy function


A Local Example: Rapid Threading Approach for Protein Structure Prediction

Kai-Ming Ho, Physics Haibo Cao

Yungok Ihm Zhong Gao

James MorrisCai-zhuang

Wang Drena Dobbs, GDCB

Jae-Hyung LeeMichael

TerribiliniJeff Sander

Cao H, Ihm Y, Wang, CZ, Morris, JR, Su, M, Dobbs, D, Ho, KM (2004)

Three-dimensional threading approach to protein structure recognition

Polymer 45:687-697








Motivations for & Assumptions of Ho Threading Algorithm

Goal: Develop a threading algorithm that:• Is simple & rapid enough to be used in high throughput

applications• Is relatively "insensitive" to sequence similarity

between target protein sequence & sequence of template structure

(to enhance detection of remote homologs & structures that are similar due to convergent evolution)

• Can be used to answer questions such as:What are predicted structures of all "unassigned" ORFs in Arabidopsis?Does Arabidopsis have a protein with structure similar to mammalian Tumor Necrosis Factor (TNF)?

Assumptions:• Native state of a protein is lowest free energy state• Hydrophobic interactions drive protein folding


Simplify: Template structure representation

,1=ijC 5.6≤ijr Åif (contact)

,0=ijC Otherwise

A neighbor in sequence (non-contact)

i

j

1

N

Template structure ( contact matrix) C NN ×

Yungok Ihm


Simplify: Target Sequence Representation

• Miyazawa-Jernigan (MJ) model: inter-residue contact energy M(i,j) is a quasi-chemical approximation based on pair-wise contact statistics extracted from known protein structures in the PDB: 20 X 20 matrix = 210 values ("letters")

• Li-Tang-Wingreen (LTW): factorize the MJ interaction

matrix to reduce the number of parameters associated with amino acids from 210 to 20 q values

• Hydrophobic-Polar (HP): represent amino acids as either H (hydrophobic) or polar (P); Dill et al demonstrated the utility of this simple binary alphabet representation: 2 values

Compare results with 210 vs 20 vs 2 letter representations

How low can we go?


Simplify: Energy Function

• Interaction “counts” only if two hydrophobic amino acid residues are in contact

• At residue level, pair-wise hydrophobic interaction is dominant:

E = i,j Cij Uij

Cij : contact matrix

Uij = U(residue I, residue J)

MJ: U = Uij

LTW: U = Qi*Qj

HP: U = {1,0}

Yungok Ihm

Energy calculation: Contact energy

Miyazawa-Jernigan (MJ) matrix:

210 parametersStatistical potential

Li-Tang-Wingreen (LTW):

20 parameters

})){(2~

( βαα +++= jiij qqCM

Contact Energy: )(1

ijjijic CQCQEN

ij

β+∑=

=2604.0,6797.0

−=−=−−=

βααii qQ

with

C M F I L

CMFILVW

046 054 -020 049 -001 006057 001 003 -008052 018 010 -001 -004

=M

iq

€

Qi~ solubility

~ hydrophobicity

contact matrix C

Yungok Ihm

ij

1

N

Template Structure

β+∑==

N

ij

jijic QCQE1

Contact Energy

Contact Matrix

Sequence

AVFMRIHNDIVYNDIANTTQ

Sequence Vector

)6497.0 ,1197.1 ,9897.0 ,7997.0(

),.....,,,(

== EFVA QQQQS

otherwise(a neighbor in sequence)

,0

56 if ,1

ij

ijij

C

rC Å

Scoring Function

Summary of Ho Threading Procedure

Yungok Ihm

Can complexity be further reduced?Consider simplifying structure representation, too

ALKKGF…HFDTSE

Sequence – Structure (1D – 3D problem)

(1D – 2D problem)

(1D – 1D problem)

Sequence – Contact Matrix

Sequence – 1D Profile

Haibo Cao

Examine eigenvectors of contact matrix

∑=

= N

i

ii

iii

TV

TVr

1

2

2

)~

(

)~

(

λ

λ

211

2

1

1

~~

)~

(~~

TVTV

TVVTCTT

i

N

i

i

N

i

iii

λλ

λ

≅=

=≡

∑

∑

=

=

Hydrophobic Contacts

iλiV :i-th eigenvector

C

1V :eigenvector with largest eigenvalue

:i-th eigenvalue of

:fraction of hydrophobic contacts from i-th eigenvectorir:protein sequence of the template structureT

C :contact matrix

Haibo Cao

Represent contact matrix by its dominanteigenvector (1D profile)

• First eigenvector (with highest eigenvalue) dominates the overlap between sequence and structure

• Higher ranking (rank > 4) eigenvectors are “sequence blind”

Haibo Cao


Threading Alignment StepThreading Alignment Step - - now fast! now fast! Align Align target sequence vector (1D)target sequence vector (1D) with with eigenvector profile of eigenvector profile of template structure template structure (1D)(1D)

1VP =1D Profile

Maximize the overlap between the

Sequence (S) and the profile (P) allowing gapsPS •

Calculate contact energy

using the alignment: Ec

New profile CPP =

Cao et al Polymer 45 (2004)


Parameters for alignment?

• Gap penalty: Insertion/deletion in helices or

strands is strongly penalized; smaller penalties for in/dels in loops

Gap penalties apply to alignment score only, not to energy calculation

• Size penalty: If a target residue and aligned

template residue differ in radius by > 0.5Å and if residue is involved in > 2 contacts, alignment is penalized

Size penalties apply to alignment score only, not to energy calculation

Loop

Helix

ALKKGFG…HFDTSE

Yungok Ihm


How incorporate secondary structure?

• Predict secondary structure of target sequence (PSIPRED, PROF, JPRED, SAM, GOR V)

N+ = total number of matches between predicted & actual secondary structure of template

N- = total number of mismatches

Ns = total number of residues selected in alignment

“Global fitness” : f = 1 + (N+ - N-) / Ns

Emod = f * Ethreading

Yungok Ihm

How much better is this “fit” than random?

Eshuffle : Shuffled Sequence vs Structure

Erelative = Emod – Eshuffled

Yungok Ihm

Avg E score for same sequence shuffled (randomized) many times

E score modifed to reflect fit with predicted 2' structure


Performance Evaluation? "Blind Test"

CASP5 Competition (CASP7 is most recent)

(Critical Assessment of Protein Structure Prediction)

Given: Amino acid sequence

Goal: Predict 3-D structure (before experimental results published)


Typical Results: (well, actually, our BEST Results):

HO = #1-Ranked CASP5 Prediction for this Target

• Target 174

• PDB ID = 1MG7

Actual Structure

Predicted Structure

T174_1

T174_2

Cao, Ihm, Wang, Dobbs, Ho


• FR Fold Recognition • (targets manually assessed by Nick Grishin)

• -----------------------------------------------------------

• Rank Z-Score Ngood Npred NgNW NpNW Group-name • 1 24.26 9.00 12.00 9 12 Ginalski • 2 21.64 7.00 12.00 7 12 Skolnick Kolinski • 3 19.55 8.00 12.50 9 14 Baker • 4 16.88 6.00 10.00 6 10 BIOINFO.PL • 5 15.25 7.00 7.00 7 7 Shortle • 6 14.56 6.50 11.50 7 13 BAKER-ROBETTA • 7 13.49 4.00 11.00 4 11 Brooks • 8 11.34 3.00 6.00 3 6 Ho-Kai-Ming • 9 10.45 3.00 5.50 3 6 Jones-NewFold • -----------------------------------------------------------

• FR NgNW - number of good predictions without weighting for multiple models• FR NpNW - number of total predictions without weighting for multiple models

Overall Performance in CASP5 Contest

~8th out of 180 (M. Levitt, Stanford)


CASP - Check it out!

Critical Assessment of Protein Structure Prediction

http://predictioncenter.gc.ucdavis.edu/

• CASP7 contest - 2006:• http://www.predictioncenter.org/casp7/Casp7.html

• Provides assessment of automated servers for protein structure prediction (LiveBench, CAFASP,

EVA) & URLs for them

• Related contests & resources:

• Protein Function Prediction (part of CASP)

• CAPRI = Critical Assessment of Predicted Interactions

• New: CASPM = CASP for M = Mutant proteins

• Predict effects of small (point) mutations, e.g., SNPs


Another Convenient List of Links for Protein Prediction Servers

http://en.wikipedia.org/wiki/List_of_protein_structure_prediction_software


Chp 13 - Protein Structure Visualization, Comparison & Classification


Xiong: Chp 13

Protein Structure Visualization, Comparison & Classification

• Protein Structural Visualization

Protein Structure Comparison• Protein Structure Classification


Protein Structure Comparison Methods

3 Basic Approaches for Aligning Structures (see Xiong textbook for details)

1. Intermolecular 2. Intramolecular 3. Combined

But, very active research area - many recent new methods

3 Popular Methods: 1. DALI = Distance Matrix Alignment of Structures

(Holm)• FSSP Database

2. SSAP = Sequential Structure Alignment Program (Orengo)1. CATH Database

• CE = Combinatorial Extension (Bourne)• VAST at NCBI

URLS:

http://en.wikipedia.org/wiki/Structural_alignment_software


Another local example: Combining Structure Prediction, Machine Learning & "Real" (wet-lab) Experiments to Investigate the Lentiviral

Rev Protein: A Step Toward New HIV Therapies

Susan Carpenter (Washington State Univ)

Wendy SparksYvonne Wannemuehler

Drena Dobbs, GDCBJae-Hyung LeeMichael Terribilini

Kai-Ming Ho, Physics Yungok IhmHaibo CaoCai-zhuang Wang

Gloria Culver, BBMBLaura Dutca

5410/15/07BCB 444/544 F07 ISU Dobbs #23 - Protein Tertiary Structure Prediction

ProvirusCytoplasm

Nucleus

Late: Structural ProteinsProgeny RNA

Macromolecular interactions mediated by

Rev protein in lentiviruses (HIV & EIAV)

pre-mRNA AAAA

(protein-protein)

(protein-protein)

(protein-protein)

NUCLEAR EXPORT

AAAARevRevRevRevNUCLEAR IMPORT

SpliceosomeSpliceosome

AAAA

Early: Regulatory Proteins

Tat

RevRev

RevRev MULTIMERIZATIONAAAARevRev

RNA BINDINGRevRev

(protein-RNA)

Susan Carpenter


Rev is essential for lentiviral replication

• Rev is a small nucleoplasmic shuttling protein

(HIV Rev 115 aa; EIAV Rev 165 aa)

• Recognizes a specific binding site on viral RNA:

Rev Responsive Element (RRE)

• Interacts with CRM1 to export incompletely spliced viral RNAs from nucleus to the cytoplasm

• Specific domains of Rev mediate nuclear localization, RNA binding, and nuclear export

• Critical role of Rev in lentiviral replication makes it an attractive target for antiviral (AIDs) therapy


Problem: no high resolution Rev structure! not even for HIV Rev, despite intense effort ($$)

• Why?? • Rev aggregates at concentrations needed for NMR or

X-ray crystallography

• What about insights from sequence comparisons? • "undetectable" sequence similarity among Revs from

different lentiviruses (eg, EIAV vs HIV <10%)

• But: • We know that lentiviral Rev proteins are functionally

"homologous" - even in highly diverse lentiviruses


• Computationally model structures of lentiviral Rev proteins

- using structural threading algorithm (with Ho et al)

• Predict critical residues for RNA-binding, protein interaction - using machine learning algorithms (with Honavar et al )

• Test model and predictions - using genetic/biochemical approaches (with Carpenter &

Culver)- using biophysical approaches (with Andreotti & Yu groups)

Initially: focus on EIAV Rev & RRE

Hypothesis: Rev proteins from diverse lentiviruses share structural features critical for function

Approach:


HIV-1 Rev

Functional domains: EIAV vs HIV Rev

1 31 165

EIAV Rev

NES NLS

RRDRW

ERLE

KRRRK

RBM Folding?

exon 1 exon 2

NES - Nuclear Export SignalNLS - Nuclear Localization SignalRBM - putative RNA Binding Motif

1 116

NESNLS/RBM

RQARRNRRRRWR


Predicted EIAV Rev Structure

Yungok Ihm


EIAV HIVFIV

SIV Dimer HIV Dimer

Comparison of Predicted Rev Structures

Yungok Ihm


A

Predicted Structure HIV Rev

N-terminus

B

NMR Structure HIV Rev N-terminal

Peptide (Battiste & Williamson)

C

OverlayAlignment of Predicted

& NMR Structures

Predicted vs Experimental Structure of

N-terminal region of HIV Rev

Yungok Ihm


Location of functional residues EIAV Rev?

Yungok Ihm

Putative RBM

NESLeu36,45,49: On surface,

consistent with rolein nuclear export

Leu95 & Leu109:Buried in core, critical

hydrophic contacts for fold?


Mutate hydrophobic residues predicted to be critical for helical packing in core

L65

L95

L109

Yungok Ihm

Single Ala Mutation L A

Single AspMutation L D

Negligible effect on Rev activity

Dramatic change in Rev activity?

Insert charged aa in hydrophobic core

Double AlaMutation LL AA

Reduction in Rev activity?

L65 vs L95 & L109

Single mutants: Leu to Ala Leu to Asp

Double mutants: Leu to Ala


aaa

50100150

L65ADL95ADL109ADL65AL95AL65AL109AL95AL109ASingle MutationsDouble MutationsControls

Act

ivity

of

Rev

Str

uctu

ral M

utan

ts

Sha

m

RI

pcD

NA

3

Functional Analysis of Rev Structural Mutants in vivo (CAT assay)

Wendy Sparks


Functional domains: EIAV vs HIV Rev

HIV-1 Rev

- RNA interaction - Protein interactionNES - Nuclear Export SignalNLS - Nuclear Localization SignalRBM - putative RNA Binding Motif

Green

Red

1 116

NESNLS/RBM

RQARRNRRRRWR

EIAV Rev

NES NLS

RRDRW

ERLE

KRRRK

RBM Folding?


Putative RNA-binding Motifs & Predicted RNA-binding Residues Mapped onto Predicted EIAV Rev Structure

61 71 81 91

ARRHLGPGPT QHTPSRRDRW IREQILQAEV LQERLEWRIR …++ +++++++ ++++++++++ + +

121 131 141 151 161 HFREDQRGDF SAWGDYQQAQ ERRWGEQSSP RVLRPGDSKRRRKHL + ++++ ++ +++ +++++++++++++++

Michael Terribilini

Yungok Ihm

KRRRK

RRDRW

ERLE


Express & purify MBP-ERev deletion mutants

60

42

3022

Mark

er

MB

P

1-1

65

31

-16

5

31

-14

5

57

-16

5

57

-14

5

57

-12

4

12

5-1

65

14

6-1

65

MBP-ERev

1-16531-165

31-145

57-165

57-145

57-124125-165

146-165

NES NLS

1 31 57 125 146 165RBM Folding?

Jae-Hyung Lee

MBP

MBP

MBP

MBP

MBP

MBP

MBP

MBP


MBP-ERev binds specifically to RRE in vitro

sense antisense

31

-16

5

BS

A MB

P1

-16

5

BS

A

MB

P

1-1

65

31

-16

5 Cold RRE

No p

rote

in

No c

old

RR

E

UV crosslinking Competition

Undigested32P-RRE

Jae-Hyung Lee

PREDICTED:

Structure

Protein binding residues

RNA binding residues

KRRRK

RRDRW

VALIDATED:



EIAV Rev: Binding Predictions vs Experiments

++

131 141 151 161 QRGDFSAWGDYQQAQERRWGEQSSPRVLRPGDSKRRRKHL++++++++++ ++ +++ ++++++ + ++++++++++++++++++++

61 71 81 91

ARRHLGPGPTQHTPSRRDRWIREQILQAEVLQERLEWRI+++++++++++++++ ++++++++++++++++

41 51GPLESDQWCRVLRQSLPEEKISSQTCI++++++++ ++

Lee et al (2006)J Virol 80:3844

Terribilini et al (2006)PSB 11:415

57-1

65

MB

PW

T

31-1

65

31-1

45

145-1

65

RRDRW

ERLE KRRR

K

NES

57 125 145 16531 FOLD?

NLS/RBM

RBM

Jae-Hyung Lee


AADAA

AALA

KAAAK

Roles of Putative RNA Binding Motifs?

NES NLS

RRDRW

ERLE

KRRRK

RBD

ERDE

RBD

1 31 57 124 146 165

Jae-Hyung Lee

Rev RNA Binding Motifs: Predicted vs Experiment

AADAA AALA KAAAK

ERDE

PREDICTED:

Structure



KRRRK

RRDRW

VALIDATED:



++


61 71 81 91



RRDR

WERLE KRRRK

NES

57 125 145 16531

KA

AA

K

AA

DA

A

AA

LA

ER

DE

WT NLS

RBM FOLD?

NLS/RBM

Jae-Hyung Lee

KRRRK

RRDRW

Summary: Predictions vs Experiments


61 71 81 91





RRDRW ERLE

KRRR

K

NES

57 125 145 16531

FOLD NLS/RBMRBM

ERLE


Conclusions & Future Directions

Combination of computational & wet lab approaches revealed that:• EIAV Rev has a bipartite RNA binding domain• Two Arg-rich RBMs are critical

• RRDRW in central region (but not ERLE)• KRRRK at C-terminus, overlapping the NLS

• Based on computational modeling, the RBMs are in close proximity within the 3-D structure of protein

• Lentiviral Rev proteins & their cognate RRE binding sites may be more similar in structure than has been appreciated



Future: Computational: Use Rev-RRE model system to discover "predictive rules" for protein-RNA recognition

Experimental?


Experimentally determine the structure of Rev-RRE complex !!!

BCB 444/544 F07 ISU Dobbs #23 - Protein Tertiary Structure Prediction

Building “Designer” Zinc Finger DNA-binding Proteins J Sander, P Zaback, F Fu, J Townsend, R Winfrey

D Wright, K Joung, L Miller, D Dobbs, D Voytas

Wright et al (2006)Nature Protocols

Sander et al (2007)Nucleic Acids Res


Chp 16 - RNA Structure Prediction


Xiong: Chp 16 RNA Structure Prediction (Terribilini)

• Introduction• Types of RNA Structures• RNA Secondary Structure Prediction Methods• Ab Initio Approach• Comparative Approach• Performance Evaluation

110/15/07BCB 444/544 F07 ISU Dobbs #23 - Protein Tertiary Structure Prediction BCB 444/544 Lecture...

Documents

Transcript of 110/15/07BCB 444/544 F07 ISU Dobbs #23 - Protein Tertiary Structure Prediction BCB 444/544 Lecture...