Advertisement
  • Loading metrics

Protective HLA alleles are associated with reduced LPS levels in acute HIV infection with implications for immune activation and pathogenesis

  • Daniel T. Claiborne ,

    Contributed equally to this work with: Daniel T. Claiborne, Eileen P. Scully

    Roles Conceptualization, Data curation, Formal analysis, Writing – original draft

    Affiliation Emory Vaccine Center, Yerkes National Primate Research Center, Emory University, Atlanta, Georgia, United States of America

  • Eileen P. Scully ,

    Contributed equally to this work with: Daniel T. Claiborne, Eileen P. Scully

    Roles Conceptualization, Data curation, Formal analysis, Writing – original draft

    Affiliation Ragon Institute of MGH, MIT and Harvard, Cambridge, Massachusetts, United States of America

  • Christine D. Palmer,

    Roles Data curation, Writing – review & editing

    Affiliation Ragon Institute of MGH, MIT and Harvard, Cambridge, Massachusetts, United States of America

  • Jessica L. Prince,

    Roles Conceptualization, Data curation, Writing – review & editing

    Affiliation Emory Vaccine Center, Yerkes National Primate Research Center, Emory University, Atlanta, Georgia, United States of America

  • Gladys N. Macharia,

    Roles Data curation, Formal analysis, Investigation, Methodology

    Affiliation Human Immunology Laboratory, International AIDS Vaccine Initiative, London, United Kingdom

  • Jakub Kopycinski,

    Roles Data curation, Formal analysis, Methodology, Supervision

    Affiliation Human Immunology Laboratory, International AIDS Vaccine Initiative, London, United Kingdom

  • Clive M. Michelo,

    Roles Data curation, Formal analysis, Investigation, Methodology

    Affiliation Zambia-Emory HIV Research Project, Lusaka, Zambia

  • Howard W. Wiener,

    Roles Formal analysis, Validation

    Affiliation Department of Epidemiology, University of Alabama at Birmingham, Birmingham, Alabama, United States of America

  • Rachel Parker,

    Roles Data curation, Investigation, Methodology, Validation

    Affiliation Department of Pathology and Laboratory Medicine, Emory University, Atlanta, Georgia, United States of America

  • Krystelle Nganou-Makamdop,

    Roles Data curation, Investigation

    Affiliation Vaccine Research Center, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, Maryland, United States of America

  • Daniel Douek,

    Roles Conceptualization, Resources, Supervision, Validation

    Affiliation Vaccine Research Center, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, Maryland, United States of America

  • Marcus Altfeld,

    Roles Conceptualization, Funding acquisition, Supervision

    Affiliation Virus Immunology Unit, Heinrich-Pette-Institut, Hamburg, Germany

  • Jill Gilmour,

    Roles Funding acquisition, Supervision

    Affiliation Human Immunology Laboratory, International AIDS Vaccine Initiative, London, United Kingdom

  • Matt A. Price,

    Roles Supervision, Validation

    Affiliations International AIDS Vaccine Initiative, New York, New York, United States of America, Department of Epidemiology and Biostatistics, University of California at San Francisco, San Francisco, California, United States of America

  • Jianming Tang,

    Roles Conceptualization, Data curation, Methodology, Supervision, Validation, Writing – review & editing

    Affiliation Department of Medicine, University of Alabama at Birmingham, Birmingham, Alabama, United States of America

  • William Kilembe,

    Roles Data curation, Resources

    Affiliation Zambia-Emory HIV Research Project, Lusaka, Zambia

  • Susan A. Allen,

    Roles Data curation, Funding acquisition, Resources

    Affiliations Zambia-Emory HIV Research Project, Lusaka, Zambia, Department of Pathology and Laboratory Medicine, Emory University, Atlanta, Georgia, United States of America

  •  [ ... ],
  • Eric Hunter

    Roles Conceptualization, Funding acquisition, Supervision, Writing – review & editing

    ehunte4@emory.edu

    Affiliations Emory Vaccine Center, Yerkes National Primate Research Center, Emory University, Atlanta, Georgia, United States of America, Department of Pathology and Laboratory Medicine, Emory University, Atlanta, Georgia, United States of America

  • [ view all ]
  • [ view less ]

Protective HLA alleles are associated with reduced LPS levels in acute HIV infection with implications for immune activation and pathogenesis

  • Daniel T. Claiborne, 
  • Eileen P. Scully, 
  • Christine D. Palmer, 
  • Jessica L. Prince, 
  • Gladys N. Macharia, 
  • Jakub Kopycinski, 
  • Clive M. Michelo, 
  • Howard W. Wiener, 
  • Rachel Parker, 
  • Krystelle Nganou-Makamdop
PLOS
x

Abstract

Despite extensive research on the mechanisms of HLA-mediated immune control of HIV-1 pathogenesis, it is clear that much remains to be discovered, as exemplified by protective HLA alleles like HLA-B*81 which are associated with profound protection from CD4+ T cell decline without robust control of early plasma viremia. Here, we report on additional HLA class I (B*1401, B*57, B*5801, as well as B*81), and HLA class II (DQB1*02 and DRB1*15) alleles that display discordant virological and immunological phenotypes in a Zambian early infection cohort. HLA class I alleles of this nature were also associated with enhanced immune responses to conserved epitopes in Gag. Furthermore, these HLA class I alleles were associated with reduced levels of lipopolysaccharide (LPS) in the plasma during acute infection. Elevated LPS levels measured early in infection predicted accelerated CD4+ T cell decline, as well as immune activation and exhaustion. Taken together, these data suggest novel mechanisms for HLA-mediated immune control of HIV-1 pathogenesis that do not necessarily involve significant control of early viremia and point to microbial translocation as a direct driver of HIV-1 pathogenesis rather than simply a consequence.

Author summary

During acute HIV infection, there exists a complex interplay between the host immune response and the virus, and the balance of these interactions dramatically affects disease trajectory in infected individuals. Variations in Human Leukocyte Antigen (HLA) alleles dictate the potency of the cellular immune response to HIV, and certain well-studied alleles (HLA-B*57, B*27) are associated with control of HIV viremia. However, though plasma viral load is indicative of disease progression, the number of CD4+ T cells in the blood is a better measurement of disease severity. Through analysis of a large Zambian acute infection cohort, we identified HLA alleles that were associated with protection for CD4+ T cell loss, without dramatic affect on early plasma viremia. We further link these favorable HLA alleles to reduction in a well-known contributor to HIV pathogenesis, the presence of microbial products in the blood, which is indicative of damage to the gastrointestinal tract, a process which accelerates disease progression in HIV infected individuals. Ultimately, these results suggest a new mechanism by which the cellular immune response can combat HIV-associated pathogenesis, and further highlight the contribution of gut damage and microbial translocation to accelerating disease progression, even at early stages in HIV infection.

Introduction

The factors underlying HIV-1 pathogenesis are a complex interplay between the host and virus. We have previously shown that viral characteristics, such as the replicative fitness of the transmitted variant, are significant predictors of early immune activation and further disease progression in HIV-1 infected individuals [1]. With respect to host factors, a series of genome-wide association studies (GWAS) have identified the HLA I locus to be the primary site of polymorphisms significantly affecting HIV-1 disease outcome [24]. Variations at this locus have also been linked to the efficiency of T cell control of viremia [5,6], and viremia is a strong predictor of HIV-1 disease progression [710]. Specific HLA class I alleles, such as B*57 and B*27, are consistently enriched in HIV-1 controllers, reinforcing the role of these genes in viral suppression [11]. However, the role of HLA may extend beyond viral control; in a multi-country HIV cohort study, the HLA class I allele, B*81, was shown to be associated with protection from CD4+ T cell decline without significant control of plasma viremia [12], suggesting that there are additional mechanisms of HLA-mediated CD4+ T cell protection distinct from early control of plasma viral load. We propose that additional immunogenetic factors with this protective phenotype exist, and furthermore, that these factors alter early events post infection, before the establishment of a viral load set point.

The most potent clinical indicator of HIV-1 pathogenesis and the severe immune depletion of AIDS is the CD4+ T cell count in the peripheral blood. CD4+ T cell decline during HIV-1 infection has been shown to be more accurately predicted by levels of chronic immune activation than viral load, underlining the importance of both host and virus in pathogenesis [1315]. The determinants of this chronic immune activation are multiple, and include viral activity along with microbial translocation [1618]. Microbial translocation is a major contributor to immune activation and disease progression in HIV-1 infection [19]. The early loss of gut mucosal integrity with the infection and depletion of CD4+ T cells locally is thought to facilitate the transit of bacterial components into the bloodstream [20]. Studies have linked both microbial DNA and LPS levels to T cell activation and amplification of viremia [21].

However, there are differences in baseline levels of detectable microbial translocation in specific populations, with evidence of ongoing immune activation in HIV seronegative MSM [22], suggesting that some of this pathogenesis may have host determinants. Further work has demonstrated that the composition of the microbial community in the female genital tract has a significant effect on the local inflammatory environment [23]. In addition, specific HLA alleles have been linked to particular microbial community phenotypes, and in turn to inflammatory disease susceptibility [2427]. Taken together, there is strong evidence for the importance of microbial translocation in immune activation and HIV-1 pathogenesis, and that the composition of the microbiome is partially determined by host immunogenetics.

To elucidate novel mechanisms of immune protection, we sought to identify additional HLA alleles associated with the phenotype of CD4+ T cell preservation, irrespective of their associations with plasma viremia. This approach allows us to uncover alternate mechanisms by which the host cellular immune response alters disease trajectory following acute infection. We identified 4 HLA class I alleles and 2 HLA class II alleles significantly associated with protection from rapid CD4+ T cell decline even when correcting for early set point viral load, sex of the infected individual, and viral replicative capacity (vRC). Furthermore, the HLA class I alleles we identified were associated with a reduction in levels of circulating LPS during acute infection, and appear to be associated with CD8+ T cell responses targeting more conserved regions of Gag. Early LPS levels were a significant predictor of CD4+ T cell decline independent of plasma viral load, and were further associated with cellular immune activation, specifically in the T cell compartment. Collectively, these data suggest that microbial translocation is a driver of immune activation rather than a consequence, establishing early gut damage and microbial translocation as a potential target for therapeutic intervention at both the acute and chronic stages of HIV infection.

Results

HLA alleles can protect from CD4+ T cell decline without controlling early plasma viral load

To identify the contribution of protective immunogenetic factors to disease progression in a Zambian heterosexual transmission cohort, we performed HLA class I and class II typing to 4-digit resolution in a group of 127 acutely infected subjects (S1 Table). Detailed clinical data was collected and plasma viral load and CD4+T cell counts measured longitudinally starting from the time of HIV-1 diagnosis (median of 44 days post estimated date of infection) and continuing at 3-month intervals thereafter up to 6 years post infection. We have previously defined viral replicative capacity (vRC) for the transmitted viruses in this cohort, as measured by an in vitro replicative fitness assay [1,28] and that data was integrated into our analysis.

We employed Cox proportional hazards models coupled with a stepwise backward variable selection approach, to identify both HLA class I and II alleles associated with protection from significant CD4+ T cell decline, which we defined as the time to a CD4+ T cell count less than 300 cells/mm3. Sex of the infected participant and vRC of the transmitted virus, which represent additional host and viral characteristics previously shown to significantly affect HIV-1 pathogenesis [1,2931], were added to the model as static covariates during the backwards selection process.

In a multivariable Cox proportional hazards model, 4 HLA class I alleles (B*1401, B*57, B*5801, and B*81) and 2 HLA class II alleles (DQB1*02 and DRB1*15) were found to be independent protective factors delaying the loss of CD4+ T cells (Table 1).

thumbnail
Table 1. HLA alleles associated with significant protection from CD4+ T cell decline.

https://doi.org/10.1371/journal.ppat.1007981.t001

As is implied by their independent predictive nature, HLA alleles affecting CD4+ T cell decline were additive in their effects and formed distinct profiles even in the absence of sex and vRC as covariates (Fig 1A). Furthermore, these HLA alleles generated additive scores when divided into HLA class I or class II alleles only (S1 Fig). In contrast, a significant additive effect of the protective alleles was not observed on set point viral load (Fig 1B), suggesting that these alleles may affect immunopathogenesis without a substantial impact on early plasma viremia in this cohort.

thumbnail
Fig 1. Multiplicities of protective HLA alleles and their additive effects on CD4+ T cell decline and set point viral loads.

HLA-B*1401, B*57, B*5801, B*81, HLA-DQB1*02, and HLA-DRB1*15 were defined as CD4-protective in this cohort. (A) Kaplan-Meier survival curves, with an endpoint defined as CD4 counts < 300/μl, demonstrating the effect of carriage of increasing numbers of protective HLA alleles on CD4+ T cell decline. P-values represent significantly different decline trajectories between adjacent groups and were generated from a Cox proportional hazards model. (B) Mean set point viral loads between individuals carrying increasing numbers of protective HLA alleles.

https://doi.org/10.1371/journal.ppat.1007981.g001

To further interrogate this observation, we sought to isolate the effects from the influence of viral load. Plasma viremia, and specifically viral load set point, is a consistent and strong predictor of CD4+ T cell decline in this and multiple other HIV infection cohorts [1,32,33]. Classically, HLA alleles that are protective in the context of HIV-1 pathogenesis have been associated with significant control of plasma viremia, especially early in infection [4,34]. As the initial analyses (Table 1) assessed HLA alleles for their effect on CD4+ T decline without controlling for set point viral load we built a multivariable Cox proportional hazards model where set point viral load was added as a covariate. We found that all 6 HLA alleles remained significant predictors with set point viral load in the model (S2 Table). Indeed, only B*57 was associated with a statistically significant decrease in early set point viral loads when analyzed in isolation (S2 Fig), an association which has been described previously [35]. These data suggest that these particular HLA alleles are exerting their protective effect on CD4+ decline via an alternative mechanism and not simply through the well-described control of early plasma viremia.

Protective HLA class I alleles are associated with a reduction in markers of gut damage in early HIV infection

Inflammation is an alternative driver of disease progression, independent of viral load [36], and is frequently associated with gut damage and bacterial translocation [37]. We therefore investigated associations between the protective HLA alleles identified in this study and inflammatory markers linked to pathogenesis. We found a striking association between the identified protective HLA class I alleles (B*1401, B*57, B*5801, and B*81) and lower plasma LPS levels at the earliest time of sampling (median of 44 days post estimated date of infection; Fig 2A). This difference in LPS levels was specific for class I alleles; we observed no association between protective HLA class II alleles and circulating LPS (Fig 2B). The grouped protective class I alleles were also associated with a moderate 0.31 log reduction in plasma viral loads at the time of LPS sampling (Fig 2C). However, in a generalized linear model, carriage of protective class I alleles is a significant predictor of LPS levels at seroconversion (p = 0.008), whereas plasma viral loads are not (p = 0.66) (Fig 2D).

thumbnail
Fig 2. Protective HLA class I, but not HLA class II, alleles are associated with reduced levels of circulating lipopolysaccharide early after HIV infection.

Lipopolysaccharide (LPS) levels were measured in the plasma for 124 seroconvertors a median of 44 days after the estimated date of infection. (A and B) The levels of LPS in the plasma were compared between individuals carrying any of the defined protective HLA class I (panel A) or HLA class II alleles (panel B). Statistical comparisons were made using the Student’s t test (A and C) or an ANOVA (B) and reported p-values are one-tailed. (C) Comparison of viral load at the time of LPS measurement between individuals with or without protective HLA class I alleles. (D) A graphical depiction of a generalized linear model showing the relative associations between both plasma viral load and the presence of protective HLA class I alleles with LPS levels measured in the plasma. The distance between horizontal lines represent the difference in means between individuals with (light blue) and without (gray) protective HLA class I alleles whereas the direction of the lines represents regression between the two continuous variables: viral load (x-axis) and LPS in the plasma (y-axis).

https://doi.org/10.1371/journal.ppat.1007981.g002

Protective HLA class I alleles were also associated with reduced levels of LPS, soluble CD14, and intestinal fatty acid-binding protein (I-FABP) at 6 months post seroconversion (Fig 3A–3C) in a subset of individuals with samples available for analysis (n = 30). These data indicate that the LPS association is durable and is reinforced by other markers of gut integrity and microbial translocation. Furthermore, individuals carrying these protective HLA-I alleles, but not those lacking them, demonstrated a significant reduction in IL-10 levels over time (Fig 3D), consistent with published literature linking the presence of circulating microbial products with increased production of IL-10 by monocytes [38,39].

thumbnail
Fig 3. Associations between protective HLA-I alleles and markers of gut damage and microbial translocation at 6 months post HIV infection.

(A–C) Comparison of levels of lipopolysaccharide, soluble CD14, and intestinal fatty acid binding protein (I-FABP) measured at 6 months post HIV infection between those with and without protective HLA-I alleles for a subset of the cohort (n = 30). Statistical comparisons were made using the Student’s t test and reported p-values are one-tailed. (D) Changes in IL-10 levels as measured in the plasma from seroconversion (mean of 44 days post infection) to 6 months post infection between individuals with (light blue) and without (gray) protective HLA-I alleles. Statistical comparison was made using a multivariate analysis of variance (MANOVA). Bars represent the standard error of the mean.

https://doi.org/10.1371/journal.ppat.1007981.g003

Other factors such as excessive alcohol use can significantly affect gut integrity and bacterial dysbiosis leading to increased levels of microbial translocation [4042]. In this cohort, data on alcohol consumption that was previously collected [43,44] was used to generate scores identifying individuals who consumed alcohol excessively and those that did not. We observed a significant elevation in plasma LPS at seroconversion in individuals that consumed alcohol excessively (S3 Fig). Importantly, in a generalized linear model, excessive alcohol consumption and protective HLA class I alleles were independent predictors of plasma LPS levels near seroconversion (S3 Fig), suggesting the association between HLA class I and LPS is not significantly confounded by alcohol consumption in this cohort.

The MHC gene locus contains additional genes that play key roles in the innate and adaptive immune systems[45]. Many of the polymorphisms in this region are in strong linkage disequilibrium (LD) [46], raising the possibility that the observed associations between these 4 HLA class I alleles and microbial translocation may be attributed to other genes in this locus. To rule out this possibility, we analyzed high resolution ImmunoChip data to look for other polymorphisms in the MHC locus in strong LD with these four HLA haplotypes. We observed no SNPs with r2 values >0.8 (S4 Fig). This suggests that the observed reduction in microbial translocation is likely attributable to the presence of these 4 protective HLA class I alleles and not other immune-related genes in the MHC locus.

Lower levels of plasma LPS in acute HIV infection are directly associated with protection from early CD4+ T cell decline

We next assessed whether a direct relationship between LPS levels and CD4+ T cell decline existed in this cohort. While translocation of microbial products, such as LPS, has been linked to immune activation and disease progression during chronic HIV-1 infection [19], we sought to specifically establish a link between very early microbial translocation and subsequent CD4+ T cell loss. In a Kaplan-Meier survival analysis with an endpoint defined as CD4+ T cell counts falling below 300 cells/mm3, individuals in the lowest 50th percentile of LPS levels at seroconversion exhibited significantly slower CD4+ T cell decline (Fig 4). This association was independent of viral load at the time of LPS sampling (S3 Table). Thus, this measure of microbial translocation early during HIV-1 infection is directly associated with the kinetics of longitudinal CD4+ T cell decline.

thumbnail
Fig 4. Circulating LPS levels measured during acute infection are associated with longitudinal CD4+ T cell decline.

A Kaplan-Meier survival analysis comparing CD4+ T cell decline between individuals with LPS levels above or below the median for the cohort (n = 124). Thirteen individuals with CD4-protective HLA class I alleles were contained in the >50% LPS group, and 21 in the <50% LPS group. Statistics were generated from the log-rank test.

https://doi.org/10.1371/journal.ppat.1007981.g004

Systemic LPS early in infection is associated with deleterious T cell phenotypes

In chronic HIV-1 infection, systemic LPS levels are correlated with increased cellular immune activation, specifically with an increased frequency of activated CD38+/HLA-DR+ CD8 T cells [19]. The mechanisms of this effect may include both direct stimulation of T cells by microbial products [47], as well as secondary effects mediated by innate recognition and inflammatory cytokine production [22,48].

In order to determine the effect of early microbial translocation on subsequent T cell activation in this cohort, cryopreserved PBMCs available from a subset of individuals collected 3 months after the estimated date of infection (median 97 days post estimated date of infection and median 65 days post systemic LPS measurements) were assessed by flow cytometry for markers of T cell activation and antigen experience/exhaustion. We found that LPS levels at seroconversion directly correlated with PD-1 expression on central memory and effector memory CD4+ T cells an average of two months later (Fig 5A and 5B). Similarly, in the CD8+ T cell compartment, increased LPS in the periphery also correlated with the percentage of CD38 and HLA-DR double positive cells, increased percentages of Ki67+ cells, and CD38 expression levels (assessed by MFI) (Fig 5C–5E). Furthermore, the associations between LPS and these activated T cell phenotypes were independent of viral loads at the time of PBMC collection (S4 Table), providing further evidence for early microbial translocation as an independent predictor of immune activation and a potential driver of immunopathogenesis.

thumbnail
Fig 5. The levels of circulating LPS at seroconversion predict levels of cellular immune activation in the T cell compartment at 3 months post infection.

Cryopreserved PBMCs isolated 3 months (mean of 97 days) post infection were stained for markers of immune activation, exhaustion, and cellular turnover as well as for the delineation of memory subsets in CD4 and CD8 T cells for a subset of the cohort for whom cryopreserved PBMCs were available (n = 19). Five of these 19 individuals carried CD4-protective HLA class I alleles. (A–B) Direct associations between LPS levels and the percentage of PD-1+ cells in central memory (CD45RO+/CCR7+) and effector memory (CD45RO+/CCR7-) CD4+ T cells. (C–E) Direct associations between LPS levels and markers of activation (CD38+/HLA-DR+, CD38 MFI) and turnover (Ki67+) in CD8+ T cells. Correlation statistics were generated using linear regression. Solid lines indicate trend lines and dashed lines represent 95% confidence bands.

https://doi.org/10.1371/journal.ppat.1007981.g005

Individuals with protective HLA class I alleles more efficiently target conserved regions of Gag

In order to gain insight regarding the quality of the CD8+ T cell response restricted by these protective HLA class I alleles, we analyzed IFNγ ELISpot data for a subset of individuals in this cohort for which early cryopreserved PMBCs were available. When PBMCs were interrogated with potential T cell epitope (PTE) Gag pools [49,50], we observed no difference in the magnitude of responses between those with or without protective HLA class I alleles, even when controlling for the relatedness of the transmitted gag sequence to the peptides contained in the PTE Gag pools (Fig 6A). Furthermore, we observed no differences in the breadth of Gag targeting when analyzing the number of specific epitopes targeted between HLA-I groups (Fig 6B). However, when peptide pools were optimized for sequence conservation, we observed that individuals with the protective HLA class I alleles identified in this study exhibited significantly higher IFNγ ELISpot responses to a pool of highly conserved Gag peptides (Fig 6C)[5153]. Targeting more conserved regions of Gag, which are unlikely to rapidly escape due to fitness constraints [5458], may help to explain the sustained protective effect of these alleles with the respect to longitudinal CD4+ T cell preservation.

thumbnail
Fig 6. Individuals with CD4-protective HLA class I alleles more efficiently target conserved regions of Gag.

(A) A graphical depiction of a generalized linear model showing the relative association between the presence of CD4-protective HLA class I alleles with IFNγ ELISpot responses to Gag PTE peptide pools from cryopreserved PBMCs collected a median of 45 days post infection. A correction factor for transmitted Gag sequence similarity to the PTE peptide pools was added as a covariate. The distance between horizontal lines represent the difference in means between individuals with (light blue, n = 6) and without (gray, n = 27) protective HLA class I alleles whereas the direction of the lines represents regression between the two continuous variables: Gag sequence similarity to peptide pools (x-axis) and IFNγ ELISpot responses reported as spot forming units per million PBMCs (y-axis). (B) Cryopreserved PBMCs collected at one-year post infection (median 339 days post infection) were interrogated with peptide pools containing 300 clade C consensus 10-mer PTE Gag peptides, with subsequent deconvolution to identify single peptide responses. The number of individual Gag peptide responses for individuals with (blue bar, n = 6) and without (gray bar, n = 12) CD4-protective HLA class I alleles are depicted. (C) A graphical depiction of a generalized linear model showing the relative association between the presence of CD4-protective HLA class I alleles with IFNγ ELISpot responses to Gag HIVconsv peptides [51] from cryopreserved PBMCs collected a median of 45 days post infection. A correction factor for transmitted Gag sequence similarity to the 3 Gag regions covered by the 31 HIVconsv Gag peptides was added as a covariate. The distance between horizontal lines represent the difference in means between individuals with (light blue, n = 6) and without (gray, n = 27) protective HLA class I alleles whereas the direction of the lines represents regression between the two continuous variables: Gag sequence similarity to HIVconsv Gag peptides (x-axis) and IFNγ ELISpot responses reported as spot forming units per million PBMCs (y-axis).

https://doi.org/10.1371/journal.ppat.1007981.g006

Discussion

In this study, we have utilized detailed clinical, virologic, and immunogenetic data from a well-defined heterosexual transmission cohort [1] to identify protective HLA alleles associated with attenuated CD4+ T cell decline irrespective of reduced early set point viral loads. Selected HLA class I (HLA-I) alleles were associated with significantly lower levels of systemic LPS during the early stages of infection, suggesting that this may be an unrecognized mechanism of protection associated with cellular immune responses restricted by favorable HLA alleles. This association was durable throughout the early stages of infection, as protective HLA class I alleles were favorably associated with additional markers of gut integrity and microbial translocation at subsequent time points.

The discovery of HLA alleles associated with protection from rapid CD4+ T cell decline without significant control of early plasma viremia is not without precedent. HLA-B*81 has been shown to be highly protective in this and in other African cohorts without significant control of early set point viral loads [12]. The importance of set point viral loads not withstanding [32,33], the study by Prentice et al., as well as the current study, serve to highlight other critical determinants of longitudinal pathogenesis as measured by CD4+ T cell counts. We have previously demonstrated that higher viral replication capacity was linked with increased inflammation and immune activation early in infection, setting the scene for subsequent pathogenesis [1]. We concluded that elevated immune activation, linked to transmission of viruses with higher replication capacity, was responsible for more rapid CD4+ T cell decline. In this present study, we observe a similar phenomenon, where the identified HLA-I alleles were protective with respect to CD4+ T cell decline independent of early set point viral loads and viral replicative capacity. These alleles were also associated with lower levels of microbial translocation as measured by systemic LPS. This leads us to hypothesize that these protective alleles may be exerting their effects of preserving CD4+ T cell counts by limiting early immune activation, perhaps by controlling in situ viral replication in key tissues, such as the gastrointestinal tract, a site of early HIV replication post transmission [5961].

In contrast to the class I alleles identified in this study, the protective HLA class II alleles (DQB1*02 and DRB1*15) were not associated with reduced levels of LPS in early infection. The protective mechanism of HIV-specific CD4+ T cells is less clear than CD8+ T cells, especially since HIV-specific CD4+ T cells are preferentially infected and depleted by HIV during infection [62]. However, Ranasinghe et al. have previously identified one of these same alleles, DRB1*15, as significantly associated with control of plasma viremia in a subtype B infected cohort [63]. In that study, DRB1*15 alleles were shown to present a greater range of HIV-derived peptides, suggesting that these class II alleles may prime broader CD4+ T cell responses against HIV. It is not clear why we did not observe a similar association between DRB1*15 and control of viremia, but it is most likely due to the distinct nature of the cohorts studied (subtype B infected and of European descent compared to subtype C infected African populations) as well as differences in the 4-digit typing of the alleles identified, where DRB1*1503 was the most common variant in this Zambian cohort and DRB1*1502 was most prevalent in the cohort studied by Ranasinghe et al.

Translocation of bacterial-derived products into the systemic circulation was first shown to be associated with pathogenesis in HIV-1 infection by Brenchley et al., specifically in the chronic stages of infection [19]. It has been proposed that specific loss of Th17 CD4+ T cells in gut-associated lymphoid tissue (GALT) contributes to structural damage seen at the mucosal barrier, allowing subsequent translocation of bioactive bacterial products into the periphery [37,64,65]. Bacterial products such as LPS have been associated with increased immune activation, a driver of HIV pathogenesis that is independent of plasma viral loads. Although depletion of total CD4+ T cells (including Th17) in the gut lamina propria occurs early during acute infection, it has been difficult to unravel the cause/effect relationship between microbial translocation and immune activation in the context of ongoing viral replication. Several in vitro and in vivo studies have attempted to understand this relationship. In vitro studies have shown that LPS alone, as well as other TLR agonists, can drive central and effector memory CD4+ T cells into cell cycle and lead to expression of activation markers on CD8+ T cells [47]. Moreover, in vivo studies of a high-risk HIV-negative MSM cohort have shed light on the specific effects of microbial translocation and systemic LPS on cellular immune activation [22,48]. In these studies, HIV-negative individuals with detectable but subclinical levels of endotoxemia exhibited significantly elevated levels of proinflammatory cytokines such as TNFa, IP-10, and MCP-1, as well as significantly lower CD4/CD8 T cell ratios. Fluctuations in naturally occurring endotoxemia in the referenced HIV-negative cohort were also associated with altered CD4+ T cell proliferation profiles and altered monocyte function. Specifically, elevated LPS was associated with a reduced proliferative capacity in the CD4+ T cell compartment, which has clear implications for CD4+ T cell decline in HIV infection.

It is not known if the findings presented in the current study can be generalized to other populations, such as MSM, living with HIV. Innate immune cell persistence in the gut is linked to the composition of the microbiome [66] and specific microbiota features have been shown to enhance inflammation [67] underlining the significance of gut barrier maintenance and a healthy microbiome. Moreover, there is evidence that both sex and sexual preferences can influence the composition of the microbiome, leaving open questions about the intersection between gut dysbiosis, HIV infection, and systemic inflammation[6871]. Nevertheless, in our study cohort of both men and women, the findings linking LPS to outcomes were independent of the sex of the participants, suggesting a more generalized phenomenon potentially linked to inflammation.

We demonstrate in this current study that LPS levels in the plasma measured at or near seroconversion predict longitudinal CD4+ T cell decline, with individuals in the lowest 50th percentile of LPS values protected from rapid loss of CD4+ T cells. In addition, we observed correlations between circulating LPS levels at seroconversion and markers of T cell activation (CD38 MFI and CD38/HLA-DR co-expression), antigen experience/exhaustion (PD-1 expression, specifically on CD4+ memory T cells), and proliferation (Ki67 expression) at a subsequent 3-month time point, suggesting a causal link between circulating LPS and immune activation, which itself is strongly correlated with disease progression irrespective of viral loads [36]. The associations between circulating LPS and cellular immune activation are largely unaffected by the inclusion of plasma viral load at the time of PBMC sampling in a multivariable general linearized model. These data suggest that, at some of the earliest time points post infection, LPS is a driver of immune activation and establishes LPS as a predictor of CD4+ T cell decline that is independent of plasma viremia.

Interestingly, we did not observed any difference between individuals in terms of magnitude or breadth of epitope recognition as measure by IFNγ ELISpot response between those that expressed or lacked CD4-protective HLA class I alleles when using broad potential T cell epitope (PTE) Gag pools [49,50]. However, we did observe a more robust response to conserved regions of the Gag protein [51,52] in individuals expressing CD4-protective HLA-I alleles. This data demonstrates that the quality of the cytotoxic T lymphocyte response restricted by these HLA-I alleles may be responsible for their efficacy. Collectively, these data suggest that CD4-protective HLA-I alleles, perhaps through their ability to target biologically relevant epitopes, can initiate cellular immune responses capable of maintaining gut integrity and limiting the translocation of microbial products into the systemic circulation. This is possibly via robust control of viral replication at specific tissue sites, at a time when we do not observe significant control of plasma viremia. This hypothesis is supported by a study by Altfeld et al., where the authors observed a striking disconnect between cytotoxic T lymphocyte (CTL) responses measured in the peripheral blood and responses measured in the lymph nodes after structured treatment interruptions [72]. Indeed, a subset of the measured CTL responses were found exclusively in the lymph node, suggesting the cellular immune response can be compartmentalized. In addition, the influence of HLA genotype on the composition of the gut microbiome may contribute; with specific HLA types supporting the growth of more favorable microbial communities [25].

In summary, these data definitively link the presence of microbial products in the circulation during acute/early infection with subsequent CD4+ T cell decline and cellular immune activation. The levels of bacterial products found in the periphery, as measured by plasma LPS levels are modulated by protective HLA class I alleles at both seroconversion and 6 months post infection. These findings demonstrate an unappreciated mechanism of protection by favorable HLA alleles, beyond simply controlling systemic viral load, and further establish microbial translocation as a driver of HIV pathogenesis. Moreover, they offer additional supporting data that efforts to ameliorate gut damage and dysbiosis, and the subsequent immune activation this dysregulation initiates in HIV-infected patients, will have positive impacts on disease outcome.

Methods

Ethics statement

All participants in the Zambia Emory HIV Research Project (ZEHRP) discordant couples cohort in Lusaka, Zambia were adults at the time of participation and enrolled in human subjects protocols approved by both the University of Zambia Research Ethics Committee and the Emory University Institutional Review Board. Prior to enrollment, individuals received counseling and signed a written informed consent form agreeing to participate.

Study subjects

The subjects included in this study were selected from the ZEHRP cohort based on being recently infected with HIV-1. All subjects were initially seronegative partners within serodiscordant cohabitating heterosexual couples, who knew their HIV status and were provided counseling and condoms, but subsequently seroconverted. All subjects were antiretroviral therapy naïve and were identified a median of 44 (IQR = 33–49) days after the estimated date of infection (EDI). The algorithm used to determine the EDI has been previously described [73]. Additional demographic and cohort data can be found in S1 Table. All subjects were infected by HIV-1 subtype C viruses.

Viral loads and CD4 count measurements

Early set point viral load (VL) for newly infected individuals was defined as the earliest stable nadir VL value measured between 3 and 9 months post infection and which did not show a significant increase in value within a 3–4 month window. HIV plasma VL was determined at the Emory Center for AIDS Research Virology Core Laboratory using the Amplicor HIV-1 Monitor Test (version 1.5; Roche). CD4+ T cell counts were based on T-cell immunophenotyping, with assays done using the FACScount System (Beckman Coulter Ltd., London, United Kingdom) in collaboration with the International AIDS Vaccine Initiative.

Genotyping

Genomic DNA was extracted from whole blood or buffy coats (QIAamp blood kit; Qiagen). HLA genotyping relied on a combination of PCR-based techniques, involving sequence-specific primers (Invitrogen) and sequence-specific oligonucleotide probes (Innogenetics), as described previously [74]. Ambiguities were resolved by direct sequencing of three exons in each gene, using kits (Abbott Molecular, Inc.) designed for capillary electrophoresis and the ABI 3130xl DNA Analyzer (Applied Biosystems).

SNP genotyping with the Illumina ImmunoChip was processed at a genomics core facility (University of Alabama at Birmingham); SNP alleles were inferred using the joint calling and haplotype phasing algorithm implemented in BEAGLECALL[75]. We completed a series of data cleaning and quality control procedures for SNPs in the xMHC region, excluding SNPs based on the following criteria: (i) duplication, (ii) missingness (call rate<98.5%), (iii) minor allele frequency <0.025 in SCs and <0.015 in SPs and (iv) deviation from Hardy–Weinberg equilibrium (P<10−6). Data processing and quality control procedures have been described previously for this data set[46].

Generation of Gag-MJ4 chimeras and in vitro replication assay

Gag-MJ4 chimeras were generated from frozen plasma isolated at the seroconversion time point for 127 subjects as previously described[1]. Briefly, viral RNA was extracted from 140 ul of plasma using the Qiagen Viral RNA extraction kit (Qiagen). Combined RT-PCR and first round PCR were performed in a single reaction, and gag genes were amplified using a nested second round PCR. Patient-derived gag genes were joined with the MJ4 long terminal repeat portion via splice-overlap-extension PCR, and gag-LTR amplicons were cloned into the MJ4 proviral vector using NgoMIV and BclI endonuclease restriction enzymes.

The in vitro replication assay used to generate viral replication capacity (vRC) values has been described extensively [76]. Briefly, replication competent virus was made via transfection of 293T cells (American Type Culture Collection). Virus supernatants were harvested, titered on the TZM-bl indicator cell line, and used to infect the GXR25 T cell line (gift from Dr. Mark Brockman, Simon Fraser University, Burnaby, BC, Canada) at a constant multiplicity of infection. Supernatants from infected cultures were harvested every 2 days, and cultures were split to maintain healthy cell confluency. The extent of viral particles released into the supernatant was quantified via a radiolabeled reverse transcriptase assay. vRC values were generated by dividing the slope of replication of each virus by the slope of replication of wild-type MJ4.

Cox proportional hazards models to identify protective HLA and HLA scores

Cox proportional hazards models with a backwards variable selection strategy were used to identify protective HLA alleles. All HLA alleles represented by at least 3 or more individuals in the cohort were included in the initial model. Sex of the seroconvertor as well as viral replication capacity of the transmitted founder virus (based on the gag gene alone) were added as static covariates present in every iteration of the model. HLA alleles contributing least significantly to the model (highest p-value in the model) were iteratively excluded in a stepwise fashion until all variables in the model had p-values < 0.05. Linkage disequilibrium between specific HLA alleles was accounted for during the variable selection process.

HLA scores presented in Fig 1 were generated by adding the number of HLA class I (B*1401, B*57, B*5801, B*81) and HLA class II (DQB1*02, DRB1*15) alleles identified to be CD4-protectve in this cohort for each individual. In S1A Fig and S1B Fig, only CD4-protective HLA class I alleles or CD4-protective HLA class II alleles were summed, respectively. Scores of 0, 1, 2, or 3 refer to any individual having that number of CD4-protective HLA class I and/or class II alleles.

Evaluation of plasma analytes

Measurement of LPS levels was performed using the LAL Chromogenic Endotoxin Quantification kit (American Diagnostica). Levels of sCD14 (R&D systems) were measured using a standard ELISA based assay according to the manufacturer’s instructions. Plasma levels of IL-10 were measured using MILLIPLEX Human Cytokine/Chemokine detection kit (EMD Millipore). Samples were run in duplicate with all individuals on the same plate and wells with low bead count or coefficient of variance >30% were excluded from subsequent analysis. Plates were read on the Bio-Plex 3D Suspension Array System (Bio-Rad).

Intestinal fatty acid binding protein (I-FABP) was measured using a commercially available ELISA DuoSet assay (R&D Systems, Minneapolis, MN, USA) according to the manufacturer's instructions with minor adjustments. Plasma samples were diluted to 10% in diluent from the R&D Systems soluble CD14 ELISA kit (DC140) and plates were blocked with Sigma Blocking Buffer.

In order to maintain normal distributions and to lessen the effect of outliers, extreme positive values for each analyte measured in plasma were Winsorized to the 90th percentile.

Flow cytometry

Cryopreserved PBMCs were stained for flow cytometric analysis in a manner identical to that described in [1]. The following antibodies were used to distinguish T cell populations and activation phenotypes: CD3-APC/Cy7 clone SP34-2, HLA-DR-PerCP/Cy5.5 clone G46-6, CCR5-PE clone 3A9, Ki67- AlexaFluor700 clone B56, CCR7-PE/Cy7 clone 3D12 (BD Biosciences); CD4- APC clone OKT4, CD8-BV605 clone RPA-T8, PD-1-BV421 clone EH12.2H7 (BioLegend); CD45RO-PE/Texas Red clone UCHL1, CD27-PE/Cy5 clone 1A4CD27 (Beckman Coulter); CD38-FITC clone AT-1 (Stemcell).

All flow cytometry data was collected on an LSRII (BD Biosciences) cytometer using FACSDiVa version 6.3.1 software. Analysis of the data performed using FlowJo version 9.7.5 software (TreeStar).

Alcohol use data and classifications

Data on alcohol consumption was collected for individuals in the ZEHRP cohort as described previously [43,44]. In order to categorize individuals into those that frequently drank alcohol to excess and those that did not, we made use of data collected from questions regarding how often in the last year individuals got drunk. Initial scores were denoted as follows: 1 = never, 2 = less than monthly, 3 = monthly, 4 = weekly, 5 = daily or almost daily. For the purpose of this study, individuals with scores 2 or less were denoted as “drunk less than monthly” and individuals with scores 3 or greater were denoted as “drunk monthly or more”.

IFNγ ELISpot assays measuring Gag-specific responses

Cryopreserved PBMCs collected at the first study visit post infection (median 45 post estimated date of infection) from 33 individuals were thawed, rested overnight, and interrogated with two PTE Gag peptide pools (HIV-1 PTE Gag Peptide Pool from NIAID, DAIDS) and a pool of 31 15-mer peptides spanning 3 highly conserved regions in Gag [51] via an IFNγ ELISpot assay as previously described [77,78]. Spot forming units per million PBMCs for the two PTE pools were summed. Correction factors for sequence conservation of the transmitted Gag amino acid sequence was calculated for each of the 33 individuals, specific for either the PTE or conserved Gag peptide pools. For the PTE pool correction factor, all 320 PTE peptides were aligned to full length Gag amino acid sequences derived from viral RNA isolated from the seroconversion time point (median of 45 days post estimated date of infection) for all 33 individuals. Scores were subsequently calculated by summing the number of PTE peptides that were at least 90% conserved with respect to the individual’s Gag sequence. For the conserved epitope pool correction factor, Gag amino acid sequences from infected individuals were aligned to the 3 distinct regions covered by this pool of 31 15-mer peptides [51]. Percent similarity to these 3 regions was calculated by extracting the pairwise identity based on the sequence alignment performed using Geneious bioinformatics software (Biomatters, Aukland, NZ). The nucleotide sequences corresponding to the amino acid sequences used for these calculations have previously been deposited under GenBank accession nos. KP715723–KP715849.

Peptide specific IFNγ ELISpot responses to measure breadth were performed on a separate subset of cryopreserved PBMCs collected a median of 339 days post infection for a total of 18 individuals in the cohort [79,80]. PBMCs were thawed and expanded ex vivo using a bi-specific CD3/CD4 antibody [81]. Expanded cells were then used to interrogate responses to a matrix of peptide pools with subsequent deconvolution to single peptides [82].

Statistical analyses

All statistical analysis was performed using JMP Pro, version 12 (SAS Institute Inc., Cary, NC). All bivariate continuous correlations were performed using standard linear regression. One-way comparison of means was performed using the Student’s t-test, and one-tailed p-values are reported. Standard ANOVA was used for the comparison of means between more than two variables. MANOVA was used to assess changes between means of two groups over time. Generalized Linear Models were used to test the predictive nature of two or more continuous or categorical variables against a single continuous outcome. Kaplan-Meier survival curves and Cox proportional hazards models were performed using an endpoint defined as a single CD4+ T cell count reading less than 300 cells/μl, unless otherwise specified, and statistics reported for survival analyses are generated from the log-rank test.

Supporting information

S1 Fig. Effects of HLA class I and class II allele scores on CD4+ T cell decline.

HLA class I score consist of the following alleles: HLA-B*1401, B*57, B*5801, and B*81. HLA class II scores consist of the following alleles: HLA-DQB1*02 and HLA-DRB1*15. Scores were calculated by summing the number of CD4-protective HLA class I or class II alleles for each individual for panel A and B, respectively. (A and B) Kaplan-Meier survival curves, with an endpoint defined as CD4 counts < 300/μl, demonstrating the effect of carriage of increasing numbers of protective HLA alleles on CD4+ T cell decline. P-values represent the comparison of decline trajectories between adjacent groups and were generated from a Cox proportional hazards model.

https://doi.org/10.1371/journal.ppat.1007981.s001

(DOCX)

S2 Fig. Univariable comparisons between protective HLA alleles and set point viral load.

HLA alleles initially defined as being protective for CD4+ T cell decline were evaluated for differences in mean set point viral loads between those carrying the allele (light blue bars) and those not carrying the allele (gray bars). Statistics based on individual t tests. ns (not significant); * p < 0.05; ** p < 0.01; *** p < 0.001.

https://doi.org/10.1371/journal.ppat.1007981.s002

(DOCX)

S3 Fig. Alcohol consumption is an independent predictor of plasma LPS levels near seroconversion.

Alcohol consumption data was collected through a revised version of the Alcohol Use Disorders Identification Test (AUDIT) questionnaire for couples in the ZEHRP cohort. Individuals with alcohol consumption data were divided into two groups, those that reported getting drunk monthly or more than monthly in the last year (n = 40), and those that reported getting drunk less than monthly or never in the last year (n = 58). (A) The graph depicts the comparison of the mean LPS levels early after HIV infection between these two groups. Statistics based on the Student’s t test, two-tailed p-value. (B) The table depicts a generalized linear model with protective HLA class I alleles and excessive alcohol consumption as predictors of plasma LPS levels near seroconversion.

https://doi.org/10.1371/journal.ppat.1007981.s003

(DOCX)

S4 Fig. SNPs in linkage disequilibrium with four protective HLA class I alleles.

(A–D) Graphs depict the linkage disequilibrium (LD) between individual SNPs within the MHC locus and four HLA class I alleles: B*1401, B*57, B*5801, and B*81. The r2 values for the linkage disequilibrium association between SNPs and each HLA class I allele are plotted on the y-axis. SNPs with r2 values >0.8 (above dotted line) are considered to be in strong LD.

https://doi.org/10.1371/journal.ppat.1007981.s004

(DOCX)

S2 Table. The effect of protective HLA alleles on CD4+ T cell decline is independent of set point viral load.

https://doi.org/10.1371/journal.ppat.1007981.s006

(DOCX)

S3 Table. Accelerated pathogenesis associated with increased LPS levels is independent of plasma viral loads at the time of LPS sampling.

https://doi.org/10.1371/journal.ppat.1007981.s007

(DOCX)

S4 Table. Associations between the levels of LPS and cellular immune activation are largely independent of plasma viremia at the time of PBMC collection.

https://doi.org/10.1371/journal.ppat.1007981.s008

(DOCX)

Acknowledgments

The investigators thank all the volunteers in Zambia who participated in this study and all the staff in the Rwanda Zambia HIV Research Group who made this study possible. The investigators would like to thank Jon Allen and Paul Farmer for technical assistance, sample management, and database management.

References

  1. 1. Claiborne DT, Prince JL, Scully E, Macharia G, Micci L, Lawson B, et al. Replicative fitness of transmitted HIV-1 drives acute immune activation, proviral load in memory CD4 +T cells, and disease progression. Proceedings of the National Academy of Sciences. 2015 Feb 17;:201421607–10.
  2. 2. Fellay J, Shianna KV, Ge D, Colombo S, Ledergerber B, Weale M, et al. A Whole-Genome Association Study of Major Determinants for Host Control of HIV-1. Science. 2007 Aug 17;317(5840):944–7. pmid:17641165
  3. 3. The International HIV Controllers Study. The Major Genetic Determinants of HIV-1 Control Affect HLA Class I Peptide Presentation. Science. American Association for the Advancement of Science; 2010 Dec 9;330(6010):1551–7. pmid:21051598
  4. 4. Martin MP, Carrington M. Immunogenetics of HIV disease. Immunological Reviews. 2013 Jul 1;254(1):245–64. pmid:23772624
  5. 5. Fellay J, Ge D, Shianna KV, Colombo S, Ledergerber B, Cirulli ET, et al. Common Genetic Variation and the Control of HIV-1 in Humans. McCarthy MI, editor. PLoS Genet. 2009 Dec 24;5(12):e1000791–12. pmid:20041166
  6. 6. Kloverpris HN, Stryhn A, Harndahl M, van der Stok M, Payne RP, Matthews PC, et al. HLA-B*57 Micropolymorphism Shapes HLA Allele-Specific Epitope Immunogenicity, Selection Pressure, and HIV Immune Control. Journal of Virology. 2011 Dec 28;86(2):919–29. pmid:22090105
  7. 7. Semple M, Loveday C, Weller I, Tedder R. Direct measurement of viraemia in patients infected with HIV-1 and its relationship to disease progression and zidovudine therapy. J Med Virol. 1991 Sep;35(1):38–45. pmid:1940882
  8. 8. Mellors JW, Kingsley LA, Rinaldo CRJ, Todd JA, Hoo BS, Kokka RP, et al. Quantitation of HIV-1 RNA in plasma predicts outcome after seroconversion. Ann Intern Med. 1995 Apr;122(8):573–9. pmid:7887550
  9. 9. Katzenstein DA, Hammer SM, Hughes MD, Gundacker H, Jackson JB, Fiscus S, et al. The relation of virologic and immunologic markers to clinical outcomes after nucleoside therapy in HIV-infected adults with 200 to 500 CD4 cells per cubic millimeter. AIDS Clinical Trials Group Study 175 Virology Study Team. N Engl J Med. 1996 Oct;335(15):1091–8. pmid:8813039
  10. 10. O'Brien WA, Hartigan PM, Martin D, Esinhart J, Hill A, Benoit S, et al. Changes in plasma HIV-1 RNA and CD4+ lymphocyte counts and the risk of progression to AIDS. Veterans Affairs Cooperative Study Group on AIDS. N Engl J Med. 1996 Feb;334(7):426–31. pmid:8552144
  11. 11. Walker BD, Yu XG. Unravelling the mechanisms of durable control of HIV-1. Nature Publishing Group. Nature Publishing Group; 2013 Jul 1;13(7):487–98.
  12. 12. Prentice HA, Porter TR, Price MA, Cormier E, He D, Farmer PK, et al. HLA-B*57 versus HLA-B*81 in HIV-1 Infection: Slow and Steady Wins the Race? Journal of Virology. 2013 Mar 8;87(7):4043–51. pmid:23365442
  13. 13. Bofill M, Mocroft A, Lipman M, Medina E, Borthwick NJ, Sabin CA, et al. Increased numbers of primed activated CD8+CD38+CD45RO+ T cells predict the decline of CD4+ T cells in HIV-1-infected patients. AIDS. 1996 Jul;10(8):827–34. pmid:8828739
  14. 14. Gougeon ML, Lecoeur H, Dulioust A, Enouf MG, Crouvoiser M, Goujard C, et al. Programmed cell death in peripheral lymphocytes from HIV-infected persons: increased susceptibility to apoptosis of CD4 and CD8 T cells correlates with lymphocyte activation and with disease progression. The Journal of Immunology. 1996 May;156(9):3509–20. pmid:8617980
  15. 15. McCune JM. The dynamics of CD4+ T-cell depletion in HIV disease. Nature. 2001 Apr;410(6831):974–9. pmid:11309627
  16. 16. Brenchley JM, Price DA, Douek DC. HIV disease: fallout from a mucosal catastrophe? Nat Immunol. 2006 Mar;7(3):235–9. pmid:16482171
  17. 17. Paiardini M, Muller-Trutwin M. HIV-associated chronic immune activation. Immunological Reviews. 2013 Jul;254(1):78–101. pmid:23772616
  18. 18. Zevin AS, McKinnon L, Burgener A, Klatt NR. Microbial translocation and microbiome dysbiosis in HIV-associated immune activation. Current Opinion in HIV and AIDS. 2016 Mar;11(2):182–90. pmid:26679414
  19. 19. Brenchley JM, Price DA, Schacker TW, Asher TE, Silvestri G, Rao S, et al. Microbial translocation is a cause of systemic immune activation in chronic HIV infection. Nature Medicine. 2006 Nov 19;12(12):1365–71. pmid:17115046
  20. 20. Estes JD, Harris LD, Klatt NR, Tabb B, Pittaluga S, Paiardini M, et al. Damaged Intestinal Epithelial Integrity Linked to Microbial Translocation in Pathogenic Simian Immunodeficiency Virus Infections. Trkola A, editor. PLoS Pathog. 2010 Aug 19;6(8):e1001052–15. pmid:20808901
  21. 21. Ericsen AJ, Lauck M, Mohns MS, DiNapoli SR, Mutschler JP, Greene JM, et al. Microbial Translocation and Inflammation Occur in Hyperacute Immunodeficiency Virus Infection and Compromise Host Control of Virus Replication. Douek DC, editor. PLoS Pathog. 2016 Dec 7;12(12):e1006048–18. pmid:27926931
  22. 22. Palmer CD, Tomassilli J, Sirignano M, Romero-Tejeda M, Arnold KB, Che D, et al. Enhanced immune activation linked to endotoxemia in HIV-1 seronegative MSM. AIDS. 2014 Sep;28(14):2162–6. pmid:25003719
  23. 23. Anahtar MN, Byrne EH, Doherty KE, Bowman BA, Yamamoto HS, Soumillon M, et al. Cervicovaginal Bacteria Are a Major Modulator of Host Inflammatory Responses in the Female Genital Tract. Immunity. Elsevier Inc; 2015 May 19;42(5):965–76. pmid:25992865
  24. 24. Gomez A, Luckey D, Yeoman CJ, Marietta EV, Berg Miller ME, Murray JA, et al. Loss of Sex and Age Driven Differences in the Gut Microbiome Characterize Arthritis-Susceptible *0401 Mice but Not Arthritis-Resistant *0402 Mice. Highlander SK, editor. PLoS ONE. 2012 Apr 24;7(4):e36095–11. pmid:22553482
  25. 25. Marietta E, Rishi A, Taneja V. Immunogenetic control of the intestinal microbiota. Immunology. 2015 Jun 3;145(3):313–22. pmid:25913295
  26. 26. De Palma G, Capilla A, Nova E, Castillejo G, Varea V, Pozo T, et al. Influence of Milk-Feeding Type and Genetic Risk of Developing Coeliac Disease on Intestinal Microbiota of Infants: The PROFICEL Study. Heimesaat MM, editor. PLoS ONE. 2012 Feb 3;7(2):e30791–10. pmid:22319588
  27. 27. Scher JU, Sczesnak A, Longman RS, Segata N, Ubeda C, Bielski C, et al. Expansion of intestinal Prevotella copri correlates with enhanced susceptibility to arthritis. eLife. 2013 Nov 5;2:205–20.
  28. 28. Prince JL, Claiborne DT, Carlson JM, Schaefer M, Yu T, Lahki S, et al. Role of Transmitted Gag CTL Polymorphisms in Defining Replicative Capacity and Early HIV-1 Pathogenesis. Lifson J, editor. PLoS Pathog. 2012 Nov 29;8(11):e1003041–15. pmid:23209412
  29. 29. Farzadegan H, Hoover DR, Astemborski J, Lyles CM, Margolick JB, Markham RB, et al. Sex differences in HIV-1 viral load and progression to AIDS. The Lancet. 1998 Nov;352(9139):1510–4.
  30. 30. Sterling TR, Lyles CM, Vlahov D, Astemborski J, Margolick JB, Quinn TC. Sex differences in longitudinal human immunodeficiency virus type 1 RNA levels among seroconverters. J INFECT DIS. 1999 Sep;180(3):666–72. pmid:10438353
  31. 31. Sterling TR, Vlahov D, Astemborski J, Hoover DR, Margolick JB, Quinn TC. Initial plasma HIV-1 RNA levels and progression to AIDS in women and men. N Engl J Med. 2001 Mar;344(10):720–5. pmid:11236775
  32. 32. Pantaleo G, Fauci AS. Immunopathogenesis of HIV infection. Annu Rev Microbiol. 1996;50:825–54. pmid:8905100
  33. 33. Langford SE, Ananworanich J, Cooper DA. Predictors of disease progression in HIV infection: a review. AIDS Research and Therapy. London: BioMed Central; 2007;4:11–1. https://doi.org/10.1186/1742-6405-4-11 pmid:17502001
  34. 34. Carrington M, Walker BD. Immunogenetics of Spontaneous Control of HIV. Annu Rev Med. 2012 Feb 18;63(1):131–45.
  35. 35. Migueles SA, Sabbaghian MS, Shupert WL, Bettinotti MP, Marincola FM, Martino L, et al. HLA B*5701 is highly associated with restriction of virus replication in a subgroup of HIV-infected long term nonprogressors. Proceedings of the National Academy of Sciences of the United States of America. The National Academy of Sciences; 2000 Mar 14;97(6):2709–14. pmid:10694578
  36. 36. Deeks SG. Immune activation set point during early HIV infection predicts subsequent CD4+ T-cell changes independent of viral load. Blood. 2004 Apr 22;104(4):942–7. pmid:15117761
  37. 37. Klatt NR, Funderburg NT, Brenchley JM. Microbial translocation, immune activation, and HIV disease. Trends in Microbiology. Elsevier Ltd; 2013 Jan 1;21(1):6–13. pmid:23062765
  38. 38. Brockman MA, Kwon DS, Tighe DP, Pavlik DF, Rosato PC, Sela J, et al. IL-10 is up-regulated in multiple cell types during viremic HIV infection and reversibly inhibits virus-specific T cells. Blood. 2009 Jul 9;114(2):346–56. pmid:19365081
  39. 39. Said EA, Dupuy FP, Trautmann L, Zhang Y, Shi Y, El-Far M, et al. Programmed death-1–induced interleukin-10 production by monocytes impairs CD4(+) T cell activation during HIV infection. Nature Medicine. 2010 Apr 1;16(4):452–9. pmid:20208540
  40. 40. Bjarnason I, Ward K, Peters T. THE LEAKY GUT OF ALCOHOLISM: POSSIBLE ROUTE OF ENTRY FOR TOXIC COMPOUNDS. The Lancet. Elsevier; 1984 Jan;323(8370):179–82.
  41. 41. Webel AR, Sattar A, Funderburg NT, Kinley B, Longenecker CT, Labbato D, et al. Alcohol and dietary factors associate with gut integrity and inflammation in HIV-infected adults. HIV Med. 2016 Nov 9;18(6):402–11. pmid:27860212
  42. 42. Bishehsari F, Magno E, Swanson G, Desai V, Voigt RM, Forsyth CB, et al. Alcohol and Gut-Derived Inflammation. Alcohol Res. National Institute on Alcohol Abuse and Alcoholism; 2017;38(2):163–71. pmid:28988571
  43. 43. Coldiron ME, Stephenson R, Chomba E, Vwalika C, Karita E, Kayitenkore K, et al. The Relationship Between Alcohol Consumption and Unprotected Sex Among Known HIV-discordant Couples in Rwanda and Zambia. AIDS and Behavior. Springer US; 2007 Aug 18;12(4):594–603. pmid:17705032
  44. 44. Davey DJ, Kilembe W, Wall KM, Khu NH, Brill I, Vwalika B, et al. Risky Sex and HIV Acquisition Among HIV Serodiscordant Couples in Zambia, 2002–2012: What Does Alcohol Have To Do With It? AIDS and Behavior. Springer US; 2017 Feb 27;21(7):1892–903. pmid:28243934
  45. 45. Horton R, Wilming L, Rand V, Lovering RC, Bruford EA, Khodiyar VK, et al. Gene map of the extended human MHC. Nat Rev Genet. 2004 Dec 1;5(12):889–99. pmid:15573121
  46. 46. Prentice HA, Pajewski NM, He D, Zhang K, Brown EE, Kilembe W, et al. Host genetics and immune control of HIV-1 infection: fine mapping for the extended human MHC region in an African cohort. Genes Immun. 2014 May 1;15(5):275–81. pmid:24784026
  47. 47. Funderburg N, Luciano AA, Jiang W, Rodriguez B, Sieg SF, Lederman MM. Toll-Like Receptor Ligands Induce Human T Cell Activation and Death, a Model for HIV Pathogenesis. Unutmaz D, editor. PLoS ONE. 2008 Apr 2;3(4):e1915–7. pmid:18382686
  48. 48. Palmer CD, Romero-Tejeda M, Sirignano M, Sharma S, Allen TM, Altfeld M, et al. Naturally Occurring Subclinical Endotoxemia in Humans Alters Adaptive and Innate Immune Functions through Reduced MAPK and Increased STAT1 Phosphorylation. The Journal of Immunology. 2016 Jan 8;196(2):668–77. pmid:26643479
  49. 49. Li F, Malhotra U, Gilbert PB, Hawkins NR, Duerr AC, McElrath JM, et al. Peptide selection for human immunodeficiency virus type 1 CTL-based vaccine evaluation. Vaccine. 2006 Nov;24(47–48):6893–904. pmid:16890329
  50. 50. Malhotra U, Li F, Nolin J, Allison M, Zhao H, Mullins JI, et al. Enhanced Detection of Human Immunodeficiency Virus Type 1 (HIV-1) Nef-Specific T Cells Recognizing Multiple Variants in Early HIV-1 Infection. Journal of Virology. 2007 Apr 25;81(10):5225–37. pmid:17329342
  51. 51. Létourneau S, Im E-J, Mashishi T, Brereton C, Bridgeman A, Yang H, et al. Design and Pre-Clinical Evaluation of a Universal HIV-1 Vaccine. Nixon D, editor. PLoS ONE. 2007 Oct 3;2(10):e984–11. pmid:17912361
  52. 52. Borthwick N, Ahmed T, Ondondo B, Hayes P, Rose A, Ebrahimsa U, et al. Vaccine-elicited Human T Cells Recognizing Conserved Protein Regions Inhibit HIV-1. Molecular Therapy. 2014 Feb;22(2):464–75. pmid:24166483
  53. 53. Rosario M, Bridgeman A, Quakkelaar ED, Quigley MF, Hill BJ, Knudsen ML, et al. Long peptides induce polyfunctional T cells against conserved regions of HIV-1 with superior breadth to single-gene vaccines in macaques. Eur J Immunol. 2010 May 14;40(7):1973–84. pmid:20468055
  54. 54. Kelleher AD, Long C, Holmes EC, Allen RL, Wilson J, Conlon C, et al. Clustered Mutations in HIV-1 Gag Are Consistently Required for Escape from Hla-B27–Restricted Cytotoxic T Lymphocyte Responses. J Exp Med. 2001 Feb 5;193(3):375–86. pmid:11157057
  55. 55. Ferguson AL, Mann JK, Omarjee S, Ndung’u T, Walker BD, Chakraborty AK. Translating HIV Sequences into Quantitative Fitness Landscapes Predicts Viral Vulnerabilities for Rational Immunogen Design. Immunity. Elsevier Inc; 2013 Mar 21;38(3):606–17. pmid:23521886
  56. 56. Boutwell CL, Carlson JM, Lin TH, Seese A, Power KA, Peng J, et al. Frequent and Variable Cytotoxic-T-Lymphocyte Escape-Associated Fitness Costs in the Human Immunodeficiency Virus Type 1 Subtype B Gag Proteins. Journal of Virology. 2013 Mar 8;87(7):3952–65. pmid:23365420
  57. 57. Leslie AJ, Pfafferott KJ, Chetty P, Draenert R, Addo MM, Feeney M, et al. HIV evolution: CTL escape mutation and reversion after transmission. Nature Medicine. 2004 Feb 8;10(3):282–9. pmid:14770175
  58. 58. Altfeld M, Addo MM, Rosenberg ES, Hecht FM, Lee PK, Vogel M, et al. Influence of HLA-B57 on clinical presentation and viral control during acute HIV-1 infection. AIDS. 2003 Dec;17(18):2581–91. pmid:14685052
  59. 59. Veazey RS, DeMaria M, Chalifoux LV, Shvetz DE, Pauley DR, Knight HL, et al. Gastrointestinal tract as a major site of CD4+ T cell depletion and viral replication in SIV infection. Science. 1998 Apr;280(5362):427–31. pmid:9545219
  60. 60. Brenchley JM, Schacker TW, Ruff LE, Price DA, Taylor JH, Beilman GJ, et al. CD4 +T Cell Depletion during all Stages of HIV Disease Occurs Predominantly in the Gastrointestinal Tract. J Exp Med. 2004 Sep 20;200(6):749–59. pmid:15365096
  61. 61. Mehandru S, Poles MA, Tenner-Racz K, Horowitz A, Hurley A, Hogan C, et al. Primary HIV-1 infection is associated with preferential depletion of CD4+ T lymphocytes from effector sites in the gastrointestinal tract. J Exp Med. 2004 Sep;200(6):761–70. pmid:15365095
  62. 62. Douek DC, Brenchley JM, Betts MR, Ambrozak DR, Hill BJ, Okamoto Y, et al. HIV preferentially infects HIV-specific CD4+ T cells. Nature. 2002 May;417(6884):95–8. pmid:11986671
  63. 63. Ranasinghe S, Cutler S, Davis I, Lu R, Soghoian DZ, Qi Y, et al. Association of HLA-DRB1&ndash;restricted CD4+ T cell responses with HIV immune control. Nature Medicine. Nature Publishing Group; 2013 Jun 23;19(7):930–3. pmid:23793098
  64. 64. Raffatellu M, Santos RL, Verhoeven DE, George MD, Wilson RP, Winter SE, et al. Simian immunodeficiency virus-induced mucosal interleukin-17 deficiency promotes Salmonella dissemination from the gut. Nature Medicine. 2008 Apr;14(4):421–8. pmid:18376406
  65. 65. Klatt NR, Estes JD, Sun X, Ortiz AM, Barber JS, Harris LD, et al. Loss of mucosal CD103+ DCs and IL-17+ and IL-22+ lymphocytes is associated with mucosal damage in SIV infection. Mucosal Immunol. Society for Mucosal Immunology; 2012 May 30;5(6):646–57. pmid:22643849
  66. 66. Hensley-McBain T, Wu MC, Manuzak JA, Cheu RK, Gustin A, Driscoll CB, et al. Increased mucosal neutrophil survival is associated with altered microbiota in HIV infection. Douek DC, editor. PLoS Pathog. Public Library of Science; 2019 Apr 11;15(4):e1007672–24. pmid:30973942
  67. 67. Neff CP, Krueger O, Xiong K, Arif S, Nusbacher N, Schneider JM, et al. Fecal Microbiota Composition Drives Immune Activation in HIV-infected Individuals. EBioMedicine. German Center for Neurodegenerative Diseases (DZNE); 2018 Apr 1;30(C):192–202.
  68. 68. Armstrong AJS, Shaffer M, Nusbacher NM, Griesmer C, Fiorillo S, Schneider JM, et al. An exploration of Prevotella-rich microbiomes in HIV and men who have sex with men. Microbiome; 2018 Nov 3;:1–16.
  69. 69. Li SX, Sen S, Schneider JM, Xiong K-N, Nusbacher NM, Moreno-Huizar N, et al. Gut microbiota from high-risk men who have sex with men drive immune activation in gnotobiotic mice and in vitro HIV infection. Silvestri G, editor. PLoS Pathog. 2019 Apr 4;15(4):e1007611–26. pmid:30947289
  70. 70. Noguera-Julian M, Rocafort M, Guillén Y, Rivera J, Casadellà M, Nowak P, et al. Gut Microbiota Linked to Sexual Preference and HIV Infection. EBioMedicine. The Authors; 2016 Mar 1;5(C):135–46.
  71. 71. Tuddenham SA, Koay WLA, Zhao N, White JR, Ghanem KG, Sears CL, et al. The Impact of Human Immunodeficiency Virus Infection on Gut Microbiota α-Diversity: An Individual-level Meta-analysis. Clinical Infectious Diseases. 2019 Mar 28;360:119–13.
  72. 72. Altfeld M, van Lunzen J, Frahm N, Yu XG, Schneider C, Eldridge RL, et al. Expansion of pre-existing, lymph node-localized CD8+ T cells during supervised treatment interruptions in chronic HIV-1 infection. J Clin Invest. 2002 Mar 15;109(6):837–43. pmid:11901192
  73. 73. Haaland RE, Hawkins PA, Salazar-Gonzalez J, Johnson A, Tichacek A, Karita E, et al. Inflammatory Genital Infections Mitigate a Severe Genetic Bottleneck in Heterosexual Transmission of Subtype A and C HIV-1. Trkola A, editor. PLoS Pathog. 2009 Jan 23;5(1):e1000274–13. pmid:19165325
  74. 74. Tang J, Tang S, Lobashevsky E, Myracle AD, Fideli U, Aldrovandi G, et al. Favorable and Unfavorable HLA Class I Alleles and Haplotypes in Zambians Predominantly Infected with Clade C Human Immunodeficiency Virus Type 1. Journal of Virology. American Society for Microbiology; 2002 Aug 1;76(16):8276–84. pmid:12134033
  75. 75. Browning BL, Yu Z. AR TICLE Simultaneous Genotype Calling and Haplotype Phasing Improves Genotype Accuracy and Reduces False-Positive Associations for Genome-wide Association Studies. The American Journal of Human Genetics. The American Society of Human Genetics; 2009 Dec 11;85(6):847–61. pmid:19931040
  76. 76. Claiborne DT, Prince JL, Hunter E. A Restriction Enzyme Based Cloning Method to Assess the In vitro Replication Capacity of HIV-1 Subtype C Gag-MJ4 Chimeric Viruses. JoVE. 2014;(90):1–18.
  77. 77. Peters BS, Jaoko W, Vardas E, Panayotakopoulos G, Fast P, Schmidt C, et al. Studies of a prophylactic HIV-1 vaccine candidate based on modified vaccinia virus Ankara (MVA) with and without DNA priming: Effects of dosage and route on safety and immunogenicity. Vaccine. 2007 Mar;25(11):2120–7. pmid:17250931
  78. 78. Gill DK, Huang Y, Levine GL, Sambor A, Carter DK, Sato A, et al. Equivalence of ELISpot Assays Demonstrated between Major HIV Network Laboratories. Kaul R, editor. PLoS ONE. 2010 Dec 14;5(12):e14330–10. pmid:21179404
  79. 79. Hayes PJ, Cox JH, Coleman AR, Fernandez N, Bergin PJ, Kopycinski JT, et al. Adenovirus-based HIV-1 vaccine candidates tested in efficacy trials elicit CD8+ T cells with limited breadth of HIV-1 inhibition. AIDS. 2016 Jul;30(11):1703–12. pmid:27088318
  80. 80. Spentzou A, Bergin P, Gill D, Cheeseman H, Ashraf A, Kaltsidis H, et al. Viral Inhibition Assay: A CD8 T Cell Neutralization Assay for Use in Clinical Trials of HIV‐1 Vaccine Candidates. J INFECT DIS. 2010 Mar;201(5):720–9. pmid:20132004
  81. 81. Wong JT, Colvin RB. Selective reduction and proliferation of the CD4+ and CD8+ T cell subsets with bispecific monoclonal antibodies: evidence for inter-T cell-mediated cytolysis. Clin Immunol Immunopathol. 1991 Feb;58(2):236–50. pmid:1824687
  82. 82. Fiore-Gartland A, Manso BA, Friedrich DP, Gabriel EE, Finak G, Moodie Z, et al. Pooled-Peptide Epitope Mapping Strategies Are Efficient and Highly Sensitive: An Evaluation of Methods for Identifying Human T Cell Epitope Specificities in Large-Scale HIV Vaccine Efficacy Trials. Lu S, editor. PLoS ONE. 2016 Feb 10;11(2):e0147812–8. pmid:26863315