A bioinformatic approach to study escape mutations of HIV1: analysis of gag genes Thursday, August 1

1
A bioinformatic approach to study escape
mutations of HIV-1 analysis of gag genes
Thursday, August 17th, 2006XVI International
AIDS Conference

• H. Peters, M. Mendoza, R. Capina, M. Luo, X. Mao,
  M. Gubbins, I. MacArthur, B. Sheardown, J.
  Kimani, J. Ndinya-Achola, S. Njenga, J. Bwayo, S.
  Thavaneswaran, F.A. Plummer

Public Health Agency of Canada, National
Microbiology Laboratory, Winnipeg, Canada,
University of Manitoba, Medical Microbiology,
Winnipeg, Canada, University of Nairobi,
Department of Medical Microbiology, Nairobi,
Kenya
2
Introduction

• Study Subjects
• Pumwani Sex Worker Cohort in Nairobi, Kenya
  about 2300 members spanning 21 years
• First described by Frank Plummer at World Aids
  Conference in Berlin, 1993
• Approximately five percent of members have been
  observed to be resistant to HIV infection

http//aidsinfonyc.org/hivplus/issue3/ahead/africa
n.html
3
Background

• A proper understanding of viral evolution is
  critical in understanding virus-host relationship
• Many studies have shown that CTL restricted
  selection is a significant driving force of viral
  evolution
• MHC Class I alleles are an excellent host factor
  to begin the study of this complex relationship
• Individually and on a population scale

4
Rationale

• Identify positively selected amino acids
• QUASI selection mapping program
• Associations between the presence of a positive
  selection and a specific HLA allele can provide
  information about
• Epitope locations, escape variants
• Objective
• To study suspected escape mutations, try to
  establish their functional significance, and
  classify them
• Identify possible HLA class I epitopes

5
Identifying Positively Selected Residues

• Phylogenetic analysis (MEGA? 3.1)
• QUASI Selection Mapping
• Tests observed replacement mutations
• Positive selection
• Replacement gtgt silent, p lt 0.05

BMC Bioinformatics (2001) 21
6
HLA Associations

• K is dominant C-Terminal anchor to A3 Supertype
  (according to SYFPEITHI)
• Includes A03
• K28Q could represent possible escape variant
• Hans-Georg Rammensee, et al. Immunogenetics
  (1999) 50 213-219

• Pearson Chi-Square Test
• Significant associations to positive selection
  are clues to epitope locations
• Evidence of selection pressure
• Los Alamos Database
• RK9 epitope to A03

Selection
Q
Residues 18 to 30 of p17 gag
http//hiv-web.lanl.gov/content/immunology/maps/ma
ps.html
HIV Molecular Immunology 2005, Editors Bette T.
M., et al. Los Alamos National Laboratory,
Theoretical Biology and Biophysics, LA-UR 06-0036
7
Results HLA Associations

• Our population
• 23 of 32 individuals positive for A03 have K28Q
• Indicates that K28Q could lie within an epitope
• Consistent with Los Alamos Database

p 6.66e-018
8
Functional Significance?

• p17 K28Q
• No change in mean CD4 count, p 0.157
• This mutation likely does not affect viral
  fitness
• P24 A31G
• Correlates to B5703, and also is associated with
  a higher mean CD4 count (p 0.038)
• Implies that the pressure exerted by 5703, in the
  form of the positively selected residue,
  restricts viral activity
• Consistent with the known protective effects of
  B57
• Costello, C. et al. (1999) AIDS 131990-91

9
Results

• p7 R7K
• Associated to lower average CD4 count (p 0.016)
• Associated to Cw4 (p 0.033)
• Associated negatively to B5703 (p 0.025), and
  to Cw1801/02 (p 0.033)
• Negative HLA association
• Leslie, A. et al. (2005) JEM 2016 891-902

10
Results

• p17 T81A
• Associated to lower average CD4 count (p 0.007)
• Associated to B5802 (p 8.31e-03)
• Associated negatively to A0201 (p 0.004), and
  to B5801 (p 2.27e-04)
• Kiepiela, P., et al. (2004) Nature 432 769775

11
Results

• p24 I58V
• Lower CD4 count (p 0.018)
• Associated to B5802 (p 1.4e-04)
• Negatively associated to Cw07 (p 0.002)
• p24 I91n (neutral mutation)
• Lies within CypA binding region
• Associated to higher CD4 count (p 5.6e-04)
• Associated to B3501 (p 0.004)
• Gao, G., et al. (2000) Journal of Biological
  Chemistry. 27520 15232-38

12
Implications

• Predict viral evolution based on the HLA
  composition of the population
• Large scale sequence analysis
• QUASI allows for very fast and very easy
  establishment of positively selected amino acids
• Epitope mapping
• Can make functional epitope studies
  faster/cheaper by using fewer peptide

13
Acknowledgements

• The women of the Pumwani Sex Worker Cohort
• The Bill and Melinda Gates Foundation
• Dr. Ma Luo and Dr. Frank Plummer
• The Departments of Medical Microbiology of the
  Universities of Manitoba and Nairobi
• The organizers of the XVI International AIDS
  conference