Limits...
Broad and Gag-biased HIV-1 epitope repertoires are associated with lower viral loads.

Rolland M, Heckerman D, Deng W, Rousseau CM, Coovadia H, Bishop K, Goulder PJ, Walker BD, Brander C, Mullins JI - PLoS ONE (2008)

Bottom Line: In a subgroup of 270 chronically infected individuals, we found that lower viral loads and higher CD4 counts were associated with a larger predicted epitope repertoire.Additionally, in Gag and Rev only, more epitopes were restricted by alleles associated with low viral loads than by alleles associated with higher viral loads.The favorable impact on markers of disease status of the propensity to present more HLA binding peptides and specific proteins gives impetus to vaccine design strategies that seek to elicit responses to a broad array of HIV-1 epitopes, and suggest a particular focus on Gag.

View Article: PubMed Central - PubMed

Affiliation: Department of Microbiology, University of Washington School of Medicine, Seattle, Washington, USA.

ABSTRACT

Background: HLA class-I alleles differ in their ability to control HIV replication through cell-mediated immune responses. No consistent associations have been found between the breadth of Cytotoxic T Lymphocytes (CTL) responses and the control of HIV-1, and it is unknown whether the size or distribution of the viral proteome-wide epitope repertoire, i.e., the intrinsic ability to present fewer, more or specific viral epitopes, could affect clinical markers of disease progression.

Methodology/principal findings: We used an epitope prediction model to identify all epitope motifs in a set of 302 HIV-1 full-length proteomes according to each individual's HLA (Human Leukocyte Antigen) genotype. The epitope repertoire, i.e., the number of predicted epitopes per HIV-1 proteome, varied considerably between HLA alleles and thus among individual proteomes. In a subgroup of 270 chronically infected individuals, we found that lower viral loads and higher CD4 counts were associated with a larger predicted epitope repertoire. Additionally, in Gag and Rev only, more epitopes were restricted by alleles associated with low viral loads than by alleles associated with higher viral loads.

Conclusions/significance: This comprehensive analysis puts forth the epitope repertoire as a mechanistic component of the multi-faceted HIV-specific CTL response. The favorable impact on markers of disease status of the propensity to present more HLA binding peptides and specific proteins gives impetus to vaccine design strategies that seek to elicit responses to a broad array of HIV-1 epitopes, and suggest a particular focus on Gag.

Show MeSH

Related in: MedlinePlus

Distribution of epitopes by HLA alleles and by protein.Distribution of epitopes among HIV-1 proteins for HLA alleles associated with lowest/highest viral loads. The ratio of predicted epitopes predicted for each protein corresponded to the number of epitope-fulfilling motifs identified in each protein over the total number of epitopes identified for the whole proteome. (A) Shows the distribution of epitopes for “good” alleles, i.e., those associated with the lowest viral loads in the cohort (lowest quartile: VL<125,437; mean = 65,384; median = 58,229). (B) Shows the epitope distribution for “bad” alleles, those associated with the highest viral loads in the cohort (highest quartile: VL>320,643; mean = 971,587; median = 531,208). For each allele belonging to a quartile, average values per allele were calculated based on the viral loads of HLA-matched individuals). (C) Illustrates the percentage of epitopes restricted by “good” and “bad” HLA alleles for each protein.
© Copyright Policy
Related In: Results  -  Collection


getmorefigures.php?uid=PMC2170517&req=5

pone-0001424-g002: Distribution of epitopes by HLA alleles and by protein.Distribution of epitopes among HIV-1 proteins for HLA alleles associated with lowest/highest viral loads. The ratio of predicted epitopes predicted for each protein corresponded to the number of epitope-fulfilling motifs identified in each protein over the total number of epitopes identified for the whole proteome. (A) Shows the distribution of epitopes for “good” alleles, i.e., those associated with the lowest viral loads in the cohort (lowest quartile: VL<125,437; mean = 65,384; median = 58,229). (B) Shows the epitope distribution for “bad” alleles, those associated with the highest viral loads in the cohort (highest quartile: VL>320,643; mean = 971,587; median = 531,208). For each allele belonging to a quartile, average values per allele were calculated based on the viral loads of HLA-matched individuals). (C) Illustrates the percentage of epitopes restricted by “good” and “bad” HLA alleles for each protein.

Mentions: Next, we ranked HLA alleles by the average viral loads of subjects in the Durban cohort: the quartile with the lowest viral loads (<125,437 viral copies; mean = 65,384; median = 58,229) included 12 alleles, herein referred as “good” alleles; the quartile with the highest viral loads (>320,643 viral copies; mean = 971,587; median = 531,20) included 12 “bad” alleles. Interestingly, the distribution of predicted epitopes among HIV-1 proteins revealed that “good” HLA alleles focused more on Gag (Figure 2A) and less on Nef (Figure 2B). For “good” HLA alleles, predicted Gag epitopes increased 1.69 fold (p = 0.036) compared to the distribution found for “bad” HLA alleles, while predicted Nef epitopes decreased 2.35 fold (p = 0.038). When analyzed by individual protein, Gag- and Rev-specific repertoires showed more epitopes restricted by “good” HLA alleles than by “bad” ones, whereas there were more epitopes restricted by “bad” HLA alleles than by “good” ones in Nef, Env, Pol, Tat, Vif, Vpu, and also Vpr (albeit marginally) (Figure 2C). Nef- and Gag-specific epitope repertoires showed similar percentages of epitopes restricted by “good” alleles, however, the proportion of epitopes restricted by “bad” alleles was significantly higher in Nef compared to its proportion in Gag.


Broad and Gag-biased HIV-1 epitope repertoires are associated with lower viral loads.

Rolland M, Heckerman D, Deng W, Rousseau CM, Coovadia H, Bishop K, Goulder PJ, Walker BD, Brander C, Mullins JI - PLoS ONE (2008)

Distribution of epitopes by HLA alleles and by protein.Distribution of epitopes among HIV-1 proteins for HLA alleles associated with lowest/highest viral loads. The ratio of predicted epitopes predicted for each protein corresponded to the number of epitope-fulfilling motifs identified in each protein over the total number of epitopes identified for the whole proteome. (A) Shows the distribution of epitopes for “good” alleles, i.e., those associated with the lowest viral loads in the cohort (lowest quartile: VL<125,437; mean = 65,384; median = 58,229). (B) Shows the epitope distribution for “bad” alleles, those associated with the highest viral loads in the cohort (highest quartile: VL>320,643; mean = 971,587; median = 531,208). For each allele belonging to a quartile, average values per allele were calculated based on the viral loads of HLA-matched individuals). (C) Illustrates the percentage of epitopes restricted by “good” and “bad” HLA alleles for each protein.
© Copyright Policy
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC2170517&req=5

pone-0001424-g002: Distribution of epitopes by HLA alleles and by protein.Distribution of epitopes among HIV-1 proteins for HLA alleles associated with lowest/highest viral loads. The ratio of predicted epitopes predicted for each protein corresponded to the number of epitope-fulfilling motifs identified in each protein over the total number of epitopes identified for the whole proteome. (A) Shows the distribution of epitopes for “good” alleles, i.e., those associated with the lowest viral loads in the cohort (lowest quartile: VL<125,437; mean = 65,384; median = 58,229). (B) Shows the epitope distribution for “bad” alleles, those associated with the highest viral loads in the cohort (highest quartile: VL>320,643; mean = 971,587; median = 531,208). For each allele belonging to a quartile, average values per allele were calculated based on the viral loads of HLA-matched individuals). (C) Illustrates the percentage of epitopes restricted by “good” and “bad” HLA alleles for each protein.
Mentions: Next, we ranked HLA alleles by the average viral loads of subjects in the Durban cohort: the quartile with the lowest viral loads (<125,437 viral copies; mean = 65,384; median = 58,229) included 12 alleles, herein referred as “good” alleles; the quartile with the highest viral loads (>320,643 viral copies; mean = 971,587; median = 531,20) included 12 “bad” alleles. Interestingly, the distribution of predicted epitopes among HIV-1 proteins revealed that “good” HLA alleles focused more on Gag (Figure 2A) and less on Nef (Figure 2B). For “good” HLA alleles, predicted Gag epitopes increased 1.69 fold (p = 0.036) compared to the distribution found for “bad” HLA alleles, while predicted Nef epitopes decreased 2.35 fold (p = 0.038). When analyzed by individual protein, Gag- and Rev-specific repertoires showed more epitopes restricted by “good” HLA alleles than by “bad” ones, whereas there were more epitopes restricted by “bad” HLA alleles than by “good” ones in Nef, Env, Pol, Tat, Vif, Vpu, and also Vpr (albeit marginally) (Figure 2C). Nef- and Gag-specific epitope repertoires showed similar percentages of epitopes restricted by “good” alleles, however, the proportion of epitopes restricted by “bad” alleles was significantly higher in Nef compared to its proportion in Gag.

Bottom Line: In a subgroup of 270 chronically infected individuals, we found that lower viral loads and higher CD4 counts were associated with a larger predicted epitope repertoire.Additionally, in Gag and Rev only, more epitopes were restricted by alleles associated with low viral loads than by alleles associated with higher viral loads.The favorable impact on markers of disease status of the propensity to present more HLA binding peptides and specific proteins gives impetus to vaccine design strategies that seek to elicit responses to a broad array of HIV-1 epitopes, and suggest a particular focus on Gag.

View Article: PubMed Central - PubMed

Affiliation: Department of Microbiology, University of Washington School of Medicine, Seattle, Washington, USA.

ABSTRACT

Background: HLA class-I alleles differ in their ability to control HIV replication through cell-mediated immune responses. No consistent associations have been found between the breadth of Cytotoxic T Lymphocytes (CTL) responses and the control of HIV-1, and it is unknown whether the size or distribution of the viral proteome-wide epitope repertoire, i.e., the intrinsic ability to present fewer, more or specific viral epitopes, could affect clinical markers of disease progression.

Methodology/principal findings: We used an epitope prediction model to identify all epitope motifs in a set of 302 HIV-1 full-length proteomes according to each individual's HLA (Human Leukocyte Antigen) genotype. The epitope repertoire, i.e., the number of predicted epitopes per HIV-1 proteome, varied considerably between HLA alleles and thus among individual proteomes. In a subgroup of 270 chronically infected individuals, we found that lower viral loads and higher CD4 counts were associated with a larger predicted epitope repertoire. Additionally, in Gag and Rev only, more epitopes were restricted by alleles associated with low viral loads than by alleles associated with higher viral loads.

Conclusions/significance: This comprehensive analysis puts forth the epitope repertoire as a mechanistic component of the multi-faceted HIV-specific CTL response. The favorable impact on markers of disease status of the propensity to present more HLA binding peptides and specific proteins gives impetus to vaccine design strategies that seek to elicit responses to a broad array of HIV-1 epitopes, and suggest a particular focus on Gag.

Show MeSH
Related in: MedlinePlus