Formulated the hypotheses, contributed to the design of the research, analyzed the data, interpreted the results, and wrote the manuscript: PWH. Contributed to framing the hypotheses, the design of the research, supervised the Treg measurements, and provided access to ACTG control samples: ALL. Performed the T cell activation and cytokine flow cytometry measurements: ES. Performed the Treg measurements: JAM CB. Provided access to the low-level viremia and HIV-specific antibody measurements, analyzed these data, and contributed to the interpretation of these measures: HH. Provided access to the cytokine flow cytometry data and contributed to the interpretation of these measures: BE. Supervised the HIV-specific antibody measurements and contributed to the interpretation of these results: PJN MPB. Provided access to patient samples and contributed to the design and analysis of the data: JNM. Contributed to framing the hypotheses and interpreting the results: JMM. Provided patient samples and contributed to framing the hypotheses, designing the research, and interpreting the results: SGD.
The authors have declared that no competing interests exist.
HIV-infected individuals maintaining undetectable viremia in the absence of therapy (HIV controllers) often maintain high HIV-specific T cell responses, which has spurred the development of vaccines eliciting HIV-specific T cell responses. However, controllers also often have abnormally high T cell activation levels, potentially contributing to T cell dysfunction, CD4+ T cell depletion, and non-AIDS morbidity. We hypothesized that a weak T regulatory cell (Treg) response might contribute to the control of viral replication in HIV controllers, but might also contribute to generalized immune activation, contributing to CD4+ T cell loss. To address these hypotheses, we measured frequencies of activated (CD38+ HLA-DR+), regulatory (CD4+CD25+CD127dim), HIV-specific, and CMV-specific T cells among HIV controllers and 3 control populations: HIV-infected individuals with treatment-mediated viral suppression (ART-suppressed), untreated HIV-infected “non-controllers” with high levels of viremia, and HIV-uninfected individuals. Despite abnormally high T cell activation levels, controllers had lower Treg frequencies than HIV-uninfected controls (P = 0.014). Supporting the propensity for an unusually low Treg response to viral infection in HIV controllers, we observed unusually high CMV-specific CD4+ T cell frequencies and a strong correlation between HIV-specific CD4+ T cell responses and generalized CD8+ T cell activation levels in HIV controllers (P≤0.001). These data support a model in which low frequencies of Tregs in HIV controllers may contribute to an effective adaptive immune response, but may also contribute to generalized immune activation, potentially contributing to CD4 depletion.
The HIV vaccine field has returned “back to basics” after a T cell-mediated immunity vaccine recently failed to prevent HIV infection and actually increased the risk of infection in important subgroups of individuals
It is important to recognize this heterogeneity as some mechanisms of viral control may prevent both initial infection and clinical progression better than others. For example, high T cell activation and low regulatory T cell (Treg) responses in highly exposed HIV-uninfected individuals have been consistently associated with an increased risk of subsequent HIV infection
We hypothesized that an unusually low Treg response to viral infection might allow some HIV controllers to maintain strong antiviral immune responses at the cost of at the cost of abnormally high generalized immune activation, potentially contributing to CD4+ T cell decline even in the absence of clinically detectable viremia. To address these hypotheses, we measured frequencies of activated (CD38+ HLA-DR+), regulatory (CD4+CD25+CD127dim), HIV-specific, and CMV-specific T cells in a large cohort of HIV controllers. We compared these data to those observed in three well characterized control populations: HIV-infected individuals with treatment-mediated viral suppression, untreated HIV-infected “non-controllers” with high levels of viremia, and HIV-uninfected individuals.
A total of 52 HIV controllers with plasma HIV RNA levels <75 copies/ml in the absence of antiretroviral therapy, 176 ART-suppressed participants, 72 untreated HIV-infected non-controllers with plasma HIV RNA levels >10,000 copies/ml, and 38 HIV-uninfected participants contributed to these studies. Most were men between 40 and 50 years of age, although compared to other HIV-infected groups, HIV controllers were more likely to be women (P = 0.006,
Characteristic | HIV-uninfectedN = 38Median (IQR) | HIV-infectedControllersVL<75 copies/mlN = 52Median (IQR) | HIV-infectedAntiretroviral-treatedVL<75 copies/mlN = 176Median (IQR) | HIV-infectedUntreatedVL>104 copies/mlN = 72Median (IQR) |
Age, years | 43 (37 to 42) | 48 (45 to 52) | 46 (41 to 52) | 44 (40 to 49) |
Female gender, no. (%) | 8 (22) | 16 (31) | 28 (16) | 12 (17) |
CD4 count, cells/mm3 | - | 683 (466 to 942) | 449 (302 to 652) | 251 (169 to 395) |
Plasma HIV RNA level, log10 copies/ml | - | <1.9 | <1.9 | 4.5 (4.2 to 4.9) |
Hepatitis C seropositive, no. (%) | - | 24 (71) |
47 (27) | 23 (36) |
Duration of HIV Diagnosis, years | - | 16 (10 to 19) | 13 (8 to 17) | 13 (9 to 16) |
VL, Plasma HIV RNA Level.
Hepatitis C virus serology was unavailable for 18 of 52 controllers.
We and others have previously reported that most HIV controllers maintain strikingly high frequencies of CD4+ and CD8+ T cells producing interferon (IFN)-γ and interleukin (IL)-2 in response to HIV Gag peptides
The frequency of activated (CD38+ HLA-DR+) CD8+ T cells (
We hypothesized that a low Treg response to HIV infection might explain why most HIV controllers maintain high HIV-specific T cell responses but also high generalized T cell activation levels. To assess this possibility, we sampled cryopreserved peripheral blood mononuclear cells (PBMC) from 20 HIV controllers, 20 ART-suppressed, and 20 untreated non-controllers, and 34 healthy HIV–uninfected controls and compared the frequencies of CD25+CD127dim CD4+ Tregs between groups. Despite having higher frequencies of activated CD4+ and CD8+ T cells than HIV-uninfected controls, the HIV controllers had a lower median frequency of Tregs (3.9% vs. 4.9%, P = 0.014,
It is surprising that HIV controllers have lower Treg frequencies and counts than HIV-uninfected individuals since higher levels of antigen stimulation and inflammation would be expected to cause greater expansion of Tregs
Since unusually low Treg responses in HIV controllers might allow for both stronger adaptive HIV-specific immune responses and generalized T cell activation, we hypothesized that there would be a strong relationship between these two latter factors. Among HIV controllers, higher frequencies of CD4+ T cells producing both IFN-γ and IL-2 in response to stimulation with HIV Gag peptides were strongly associated with higher frequencies of activated CD4+ T cells (rho: 0.36, P = 0.012) and activated CD8+ T cells (rho: 0.55, P<0.001,
The association between the frequency of activated (CD38+ HLA-DR+) CD8+ T cells and the frequency of CD4+ T cells producing both IFN-γ and IL-2 after stimulation with overlapping HIV Gag (
The frequency of HIV-specific CD8+ T cells were less consistently associated with the frequency of activated T cells. In general, there was little evidence for an association between the frequency of HIV-specific CD8+ T cells producing both IFN-γ and IL-2 and the frequency of activated CD4+ or CD8+ T cells. However, higher frequencies of activated CD8+ T cells tended to be associated with higher frequencies of CD8+ T cells producing IFN-γ but not IL-2 in response to HIV Nef (rho: 0.42, P = 0.025,
We next hypothesized that an unusually low Treg response in HIV controllers might also contribute to higher adaptive immune responses directed at other chronic viral infections. We chose to focus on cytomegalovirus (CMV) since CMV is nearly ubiquitous in HIV infected individuals, is typically controlled to nearly undetectable levels in individuals with intact immune systems, yet elicits high frequencies of CMV-specific T cells even in HIV-uninfected individuals
(
A wealth of data now suggest that most HIV controllers maintain control of viral replication at least in part through potent HIV-specific T cell responses
A theoretical model to describe the potential positive and negative consequences of low Treg frequencies in HIV controllers is presented. While a low Treg response might increase HIV-specific T cell responses, contributing to the clearance of HIV-infected cells and the maintenance of extremely low levels of viral replication, a low Treg response might also increase generalized T cell activation, contributing to CD4+ T cell decline and other inflammation-associated comorbidities even in the presence of very low levels of viral replication.
Multiple mechanisms have been proposed to explain why HIV controllers maintain low to undetectable levels of viral replication in the absence of therapy. While it is possible that some HIV controllers may simply be infected with defective viruses
It is important to acknowledge this heterogeneity in the mechanisms of viral control in HIV controllers as some mechanisms are likely to be associated with more negative inflammatory consequences than others. While other cohorts have not observed increased T cell activation levels in HIV controllers
Alternatively, HIV controllers may be enriched for host genetic factors associated with strong innate and/or weak Treg responses to viral infection. Indeed, we found that HIV controllers had significantly lower frequencies of CD25+CD127dim CD4+ Tregs in peripheral blood than HIV-uninfected individuals despite much higher levels of T cell activation. While we cannot exclude the possibility that HIV controllers preferentially retain Tregs in lymphoid tissues, a recent study also found low frequencies of Tregs in tissues of HIV controllers
Our results differ from another recent report describing preserved Treg frequencies (as defined by FoxP3 expression) in the peripheral blood of a much smaller cohort of 12 HIV controllers
Consistent with the hypothesis that HIV controllers are predisposed to a weak Treg response to chronic viral infections, we observed significantly higher CMV-specific CD4+ T cell responses in HIV controllers than non-controllers and HIV-uninfected individuals. While we cannot exclude the possibility that greater CMV shedding explains the higher CMV-specific CD4+ T cell responses in HIV controllers, CMV shedding tends to be lower in individuals with higher CD4+ T cell counts and lower plasma HIV RNA levels
In summary, we have observed that while most elite controllers maintain high HIV-specific T cell responses, most also have abnormally high generalized T cell activation levels, which may occasionally contribute to significant CD4 depletion even in the absence of clinically detectable viremia. Furthermore, those with the highest HIV-specific T cell responses have the highest levels of generalized immune activation, suggesting possible inflammatory consequences of T cell-mediated control of HIV replication. An unusually low regulatory T cell response to HIV infection may well explain this phenomenon. Perhaps the best immune response to HIV infection is one that maintains control of viral replication while minimizing negative inflammatory consequences. Some elite controllers are able to maintain this balance and understanding the mechanisms of control in these individuals is likely to have important implications for HIV vaccine research.
HIV-infected adults were sampled from the Study of the Consequences of the Protease Inhibitor Era (SCOPE), a clinic-based cohort of over 1000 chronically HIV-infected individuals at the University of California San Francisco. From this cohort, we evaluated three distinct groups of HIV-infected individuals: (1) HIV controllers, defined as HIV-seropositive individuals maintaining plasma HIV RNA levels <75 copies/ml in the absence of therapy (episodes of clinically detectable viremia in the previous year were allowed if they were followed by undetectable values); (2) “ART-suppressed” individuals maintaining plasma HIV RNA levels <75 copies/ml on antiretroviral therapy; and (3) untreated HIV “non-controllers” with plasma HIV RNA levels above 10,000 copies/mL. T cell activation data have been previously reported on 30 of the 52 HIV controllers and all of the ART-suppressed and untreated patients in the current report
In addition to the above participants, untreated HIV-infected participants with plasma HIV RNA levels between 75 and 10,000 copies/ml were sampled from the SCOPE cohort. HIV-negative individuals were also sampled from a trial of post-exposure prophylaxis following a non-occupational exposure to HIV
Given limited PBMC availability, cryopreserved PBMC from different SCOPE participant-timepoints were sampled for the measurement of both Treg frequency and T cell activation levels in 20 HIV controllers, 20 HAART-suppressed participants, and 20 non-controllers. Only specimens on participants with CD4+ T cell counts >350 cells/mm3 were selected for these analyses to ensure adequate overlap between groups. For the Treg analyses, cryopreserved PBMC were also sampled from 34 healthy HIV-uninfected controls from the AIDS Clinical Trials Group 5015 study
All participants provided written informed consent and this research was approved by the institutional review board of the University of California, San Francisco.
Freshly collected, EDTA-anticoagulated whole blood was analyzed by four-color flow cytometry on a Beckman Coulter Epics XL flow cytometer. Blood was stained on a Beckman Coulter Prep Plus and lysed on a Beckman Coulter TQ Prep. Activated (CD38+/HLA-DR+) T cells were identified with FITC-conjugated anti-HLA-DR, PE-conjugated anti-CD38 (both from BD Bioscience), PC5-conjugated anti-CD3 and PE-texas red conjugated anti-CD4 or CD8 (Beckman Coulter). The activation markers CD38 and HLA-DR were gated from the CD3+CD4+ or CD3+CD8+ cells on a 2-dimensional dot plot where quadrant gates, set on an isotype control, were used to define positive and negative populations. T cell activation levels were reported as the percentage of CD4+ and CD8+ T cells expressing both HLA-DR and CD38.
Fresh whole blood was stimulated with overlapping peptide pools (15-amino-acid peptides overlapping by 11 amino acids) of the HIV-1 p55 Gag, Pol, Nef, Env, or CMV pp65 protein (BD Biosciences, San Jose, CA) for 6 h in the presence of brefeldin A, as reported recently
Cryropreserved PBMC were evaluated using 4-color flow cytometry. Mouse anti-human monoclonal antibodies (CD4, CD8, CD25, CD45RO, and CD127) conjugated to fluorescein isothiocyanate (FITC), phycoerythrin (PE), PerCP, and allophycocyanin (APC) from BD Biosciences (San Jose, CA) or Coulter Immunology (Miami, FL) were used to stain the PBMC preparations. Non-specific antibody binding to Fc receptors was blocked by pre-incubation of the cells with Fcγ-receptor block (Miltenyi Biotec, Auburn, CA). All samples were evaluated within 24-hours of staining using a FACSCalibur™ flow cytometer. Logical gating was used to identify the frequency of T regulatory (CD4+/CD25+/CD127dim) T lymphocyte populations (
A “de-tuned” enzyme immunoassay (Organon Tecnika Vironostika [OTV], BioMerieux) was used to measure semiquantitative HIV antibody levels on a subset of HIV controllers
Continuous variables were compared between groups with Kruskal Wallis tests followed by Wilcoxon ranksum tests for pairwise comparisons. Dichotomous variables were compared between groups with chi square and Fisher's exact tests. Relationships between continuous variables were assessed with Spearman's rank order correlation coefficients. Adjusted differences between groups were assessed with linear regression, calculating standard errors with heteroskedasticity-consistent covariance matrix estimators and log-transforming outcomes when necessary to satisfy model assumptions
(TIF)
The authors would like to thank Dr. Dennis Hartigan-O'Connor for his thoughtful discussion of this work.