“HIV controllers” (HICs) are rare individuals in whom HIV-1 plasma viral load remains undetectable without antiretroviral treatment. This spontaneous viral control in HICs is usually associated to strong functional HIV-specific CD8+ T cell responses. Accordingly, we have recently shown that CD8+ T cells from HICs strongly suppress ex vivo HIV-1 infection of autologous CD4+ T cells, suggesting a crucial role of this response in vivo. Knowledge of the mechanisms underlying the CD8+ T cell antiviral activity might help to develop effective T cell-based vaccines. In the present work, we further characterized the HIV-suppressive capacity of CD8+ T cells in 19 HICs. CD8+ T cells from 14 of the 19 HICs showed strong HIV-suppressive capacity ex vivo. This capacity was stable over time and was partially effective even on other primate lentiviruses. HIV-suppressive capacity of CD8+ T cells correlated strongly with the frequency of HIV-specific CD8+ T cells, and in particular of Gag-specific CD8+ T cells. We also identified five HICs who had weak HIV-suppressive CD8+ T cell capacities and HIV-specific CD8+ T cell responses. Among these five HICs, at least three had highly in vitro replicative viruses, suggesting that the control of viremia in these patients is not due to replication-defective viruses. These results, on the one hand, suggest the importance of Gag responses in the antiviral potency of CD8+ T cells from HICs and, on the other hand, propose that other host mechanisms may contribute to restraining HIV infection in HICs.

At least partial control of HIV can be achieved by CD8+ T cells (1, 2, 3). Some HLA class I molecules, particularly alleles B27 and B57, have been linked to better control of HIV infection (4, 5, 6). The presence of Gag-specific CD8+ T cells and the breadth of their specificities have also been linked to low HIV viremia (7, 8, 9). One of the most compelling indications of the pressure exerted by CD8+ T cell responses is the emergence of variants that escape recognition by these cells (10, 11, 12, 13, 14). However, most HIV-infected individuals have uncontrolled viremia and eventually progress to AIDS despite strong CD8+ T cell responses.

Rare individuals called “HIV controllers” (HICs)5 spontaneously and durably control HIV infection in the absence of therapy, possibly illustrating what truly effective CD8+ T cell responses can achieve (15, 16). HICs have extremely low and stable amounts of viral DNA in their PBMC (17) and undetectable plasma viral load (18). The protective HLA alleles B27 and B57 are overrepresented among these individuals (4, 6, 19, 20, 21). Despite very low levels of Ag in blood (22), most but not all HICs have high frequencies of HIV-specific CD8+ T cells that preferentially target the viral Gag protein (20, 23, 24). Studies of CD8+ T cell responses in HICs have revealed important characteristics of functional HIV-specific CD8+ T cells in HIV infection. Contrary to cells from viremic individuals, HIV-specific CD8+ T cells from HICs can, upon stimulation with their cognate Ag, proliferate and generate a multifunctional response that includes perforin expression, degranulation, and chemokine/cytokine secretion (25, 26, 27). This could be related to a peculiar activation phenotype of these cells (21) and to constitutive telomerase activity that protects them against senescence (28). However, how much of this is the cause and how much the consequence of viral control and low-level immune activation remains to be determined. We have recently shown that CD8+ T cells from most HICs are endowed with a striking capacity to suppress HIV infection ex vivo (21), a property that is likely to be relevant in vivo. To further characterize this HIV-suppressive activity we extended our analysis to a larger group of 19 HICs and evaluated the relationship between this activity and HIV-specific CD8+ T cell responses.

Nineteen patients diagnosed with HIV-1 infection at least 10 years previously who had never received antiretroviral treatment and in whom > 90% of plasma HIV RNA assays gave values<400 copies/ml were studied (Table I): 8 have been described elsewhere (21), and 11 were newly recruited from the ANRS EP36 national monitoring program on HIV controllers. The subjects were serologically HLA typed by complement-mediated lymphocytotoxicity testing (InGen One Lambda). All had very weak and stable DNA load (Table I).

Table I.

Characteristics of HIV controllers included in the study

PatientaSexAge (years)HIV DiagnosedMode of InfectionbSubtype of Infecting VirusHLAMedian CD4 Count (first–last) (cells/ml)Viral Follow-up (years)Median RNA Viral Load (copies/ml)Viral Load Blips (50–400 copies/ml)cHIV DNA (log copies/106 cells)
A6 69 1985 Hem A2/A74/B27/B57 1041 (775-1528) 11 <50 7/16 NA 
A9 45 1992 MMS A2/A29/B27/B57 844 (1107-522) 15 <50 0/13 NA 
A12 46 1986 IDU NA A1/A2/B44/B57 632 (1151-632) 10 <50 4/13 1.67 
A21 45 1987 Unknown A1/A2/B51/B57 839 (1217-785) 10 <50 9/18 1.62 
A7 39 1983 Hem A2/B44/B57 613 (601-408) 11 <50 7/18 1.77 
A2 49 1985 Het A3/A23/B7/B57 709 (599-493) <50 1/12 0.5 
A1 49 1988 Het A2/B27/B57 958 (1290-849) 12 <50 1/15 0.5 
A4 48 1987 IDU A2/A32/B27/B60 597 (602-649) 10 <50 0/15 1.67 
B5 59 1989 MMS A2/A68/B14/B57 890 (1000-937) 12 <50 0(2)/16 1.31 
A15 50 1987 Het A3/A30/B27/B51 640 (883-330) 11 <50 2/20 1.04 
A3 42 1993 Het A3/A30/B57/B63 1035 (960-1033) 11 <50 0/11 1.04 
A18 44 1996 Het NA A2/A31/B39/B57 896 (665-877) 11 <50 3(1)/16 1.04 
A17 46 1989 IDU A2/B7/B57 1194 (601-1196) 11 <50 0/8 1.94 
A11 44 1989 IDU A3/A30/B7/B57 504 (390-556) 11 <50 2/19 1.28 
A16 50 1985 IDU A24/A32/B27/B62 976 (1107-1142) <50 0/17 2.08 
A13 47 1992 MMS A3/A29/B35/B44 727 (653-828) 11 <100 0/21 1.64 
A23 38 1994 Unknown NA A43/A74/B57/B72 771 (902-773) <50 0/11 0.5 
A22 40 1991 Het NA A11/A30/B13/B27 800 (800-717) 10 <50 0/9 NA 
A19 39 1994 Het A2 A2/A68/B60/B14 928 (790-808) 12 <50 0/10 1.57 
PatientaSexAge (years)HIV DiagnosedMode of InfectionbSubtype of Infecting VirusHLAMedian CD4 Count (first–last) (cells/ml)Viral Follow-up (years)Median RNA Viral Load (copies/ml)Viral Load Blips (50–400 copies/ml)cHIV DNA (log copies/106 cells)
A6 69 1985 Hem A2/A74/B27/B57 1041 (775-1528) 11 <50 7/16 NA 
A9 45 1992 MMS A2/A29/B27/B57 844 (1107-522) 15 <50 0/13 NA 
A12 46 1986 IDU NA A1/A2/B44/B57 632 (1151-632) 10 <50 4/13 1.67 
A21 45 1987 Unknown A1/A2/B51/B57 839 (1217-785) 10 <50 9/18 1.62 
A7 39 1983 Hem A2/B44/B57 613 (601-408) 11 <50 7/18 1.77 
A2 49 1985 Het A3/A23/B7/B57 709 (599-493) <50 1/12 0.5 
A1 49 1988 Het A2/B27/B57 958 (1290-849) 12 <50 1/15 0.5 
A4 48 1987 IDU A2/A32/B27/B60 597 (602-649) 10 <50 0/15 1.67 
B5 59 1989 MMS A2/A68/B14/B57 890 (1000-937) 12 <50 0(2)/16 1.31 
A15 50 1987 Het A3/A30/B27/B51 640 (883-330) 11 <50 2/20 1.04 
A3 42 1993 Het A3/A30/B57/B63 1035 (960-1033) 11 <50 0/11 1.04 
A18 44 1996 Het NA A2/A31/B39/B57 896 (665-877) 11 <50 3(1)/16 1.04 
A17 46 1989 IDU A2/B7/B57 1194 (601-1196) 11 <50 0/8 1.94 
A11 44 1989 IDU A3/A30/B7/B57 504 (390-556) 11 <50 2/19 1.28 
A16 50 1985 IDU A24/A32/B27/B62 976 (1107-1142) <50 0/17 2.08 
A13 47 1992 MMS A3/A29/B35/B44 727 (653-828) 11 <100 0/21 1.64 
A23 38 1994 Unknown NA A43/A74/B57/B72 771 (902-773) <50 0/11 0.5 
A22 40 1991 Het NA A11/A30/B13/B27 800 (800-717) 10 <50 0/9 NA 
A19 39 1994 Het A2 A2/A68/B60/B14 928 (790-808) 12 <50 0/10 1.57 
a

Patients are sorted in function of the suppressive capacity of their CD8+ T cells (log p24 decrease, Fig. 1 C). Horizontal rule separates strong and weak responders.

b

Hem, hemophilia; Het, heterosexual sex; IDU, injection drug use; MMS, male-male sex.

c

Number of viral load between 50 and 400 copies per total determinations with detection limit <50 copies/ml. Viral loads of >400 RNA copies/ml are indicated in parentheses.

All of the subjects gave their written informed consent.

Total DNA was extracted from whole blood with QIAamp DNA minikits (Qiagen), according to the manufacturer’s instructions. HIV-1 DNA was then quantified by real-time PCR (LTR amplification; Agence Nationale de Recherches sur le SIDA) (29). Four PCRs, each testing 1 μg of total DNA, were performed per extract in this ultrasensitive assay (threshold of 10 copies/million leukocytes) (30).

CD4+ and CD8+ cells were purified (>97%) from freshly isolated PBMC by positive and negative selection, respectively, with Ab-coated magnetic beads (Miltenyi Biotec). CD4+ cells were stimulated for 3 days with phytohemagglutinin (PHA) at 2 μg/ml in the presence of IL-2 (Chiron) at 100 IU/ml. The culture medium was RPMI 1640 containing 10% FCS and penicillin/streptomycin (100 U/ml). CD8+ T cells were kept in culture without mitogens or cytokines.

CD4+ T cells (105) were superinfected with HIV-1 BaL (R5) in triplicate at a multiplicity of infection (moi) of 10−3.6 in 96-well plates with a spinoculation protocol (31). For some experiments, SIVagm.Gril, SIVmac.251, and HIV-2.SBL and autologous primary viruses were used for infection. For coculture, 105 CD4+ T cells were mixed with 105 CD8+ T cells (CD8/CD4 ratio of 1:1) at the moment of infection. After challenge the cells were washed and cultured for 14 days. Viral replication was monitored every 3–4 days in supernatants by p24 or p27 ELISA (Zeptometrix). Infectivity assays were conducted in the presence of 100 IU/ml IL-2. We have previously shown that the presence of this cytokine during the infectivity assays did not affect the suppressive capacity of unstimulated CD8+ T cells (21).

Activated CD4+ lymphocytes (5 × 104) were superinfected with HIV-1 BaL (R5) as described above. Various dilutions of virus (moi of 10−1.6 to 10−2.6) were used in parallel to obtain similar levels of infection in each individual/experiment. CD4+ T cells were culture in the presence or absence of unstimulated CD8+ T cells (CD8/CD4 ratio of 1:1).

Seventy-two hours after infection, cells were harvested and stained with CD4-ECD (SFCI12T4D11) and CD8-PC5 (B9.11). Cells were then permeabilized (Cytofix/Cytoperm fixation and permeablization kit; BD Biosciences) and stained with KC57-FITC (FH190-1-1) to detect intracellular HIV Ags. Abs were from Beckman Coulter. Flow cytometry was performed with a Cytomix FC500 and CXP acquisition software (Beckman Coulter).

Between 2 and 5 × 106 CD4+ T cells from each patient were activated with PHA and IL-2 as described above. Viral production in culture supernatants was monitored for 28 days by p24 ELISA. When required, CD4+ T cells were reactivated on day 10 with CD8-depleted PHA-preactivated allogeneic PBMC, PHA, and IL-2. Virus-containing supernatants from CD4 T cell cultures were titrated on mixed PHA-activated CD4+ T cells from two blood donors.

IFN-γ secretion by HIV-specific CD8+ T cells was quantified ex vivo with an ELISPOT assay using appropriate stimuli (32). We used a set of 124 peptides corresponding to known optimal CTL epitopes derived from the HIV-1 Env, Gag, Pol, and Nef proteins (National Institutes of Health HIV Molecular Immunology Database; www.hiv.lanl.gov/content/immunology/index.html). These peptides were synthesized by Neosystem and used at a final concentration of 2 μg/ml. For each subject, optimal peptides were tested depending on the results of HLA typing with an average of 36 ± 9 peptides tested per subject. IFN-γ spot-forming cells (SFCs) were counted with a KS-ELISPOT system (Carl Zeiss Vision) and expressed as SFCs/106 PBMC after subtracting the background of control unstimulated cells. Wells were considered positive if they contained at least 50 SFCs/106 PBMC and exhibited at least twice the background level.

Depletion of CD8+ T cells producing IFN-γ upon stimulation with HIV peptides was performed with an IFN-γ secretion assay enrichment kit (Miltenyi Biotec) as recommended by the manufacturer. Briefly, purified CD8+ T cells were stimulated for 6 h with appropriate pools of specific HIV peptides. Subsequently, the cells were labeled (5 min at 4°C) with an IFN-γ catch reagent that attached to the cell surface of all leukocytes. The cells were then incubated for 45 min at 37°C to allow IFN-γ secretion. The secreted IFN-γ was captured by the IFN-γ catch reagent on the positive, secreting cells. These cells were subsequently labeled with a second IFN-γ-specific Ab conjugated to R-PE. The IFN-γ-secreting cells were magnetically labeled with anti-PE magnetic beads and depleted by magnetic field separation. Purity of the depleted fractions was evaluated by flow cytometry.

The following Abs were used: CD8-ECD or -PC5 (clone B9.11), CD3-PC5 (UCHT1), CD45RO-ECD (UCHL1), HLA-DR-ECD (Immu-357), and CD38-FITC (T16), all from Beckman Coulter; and CD27-FITC (M-T271) from BD Biosciences.

HIV-specific CD8+ T cells were identified by using soluble PE- or allophycocyanin-labeled peptide-HLA class 1 multimers (Proimmune; Beckman Coulter Immunomics). The following epitopes were used: the HLA-A*0201-restricted peptide ligands SLYNTVATL (Gag 77–85) and ILKEPVHGV (Pol 476–484), the A*0301-restricted peptide ligands RLRPGGKKK (Gag 20–28) and QVPLRPMTYK (Nef 73–82), the B*2705-restricted peptide ligand KRWIILGLNK (Gag 263–272), and the B*5701-restricted peptide ligands KAFSPEVIPMF (Gag 162–172), TSTLQEQIGW (Gag 240–249), and QASQDVKNW (Gag 308–316). PBMC were incubated with pentamers (1 μg/ml) for 30 min and then with relevant Abs for 15 min. Cells were washed in Cell Wash (BD Biosciences) plus 1% BSA, incubated for 10 min with FACS lysing solution (BD Biosciences). After washing, cells were fixed in 1% paraformaldehyde for flow cytometry with an Epics XL (Beckman Coulter) or a FACSCanto flow cytometer (BD Biosciences) and analyzed with RXP software (Beckman Coulter).

The proliferative capacity of HIV-specific CD8+ T cells was evaluated by flow cytometry. PBMC were stained with 0.35 μM CFSE (Molecular Probes) for 10 min at 37°C, and, after washing they were stimulated for 5 days with 2 μg/ml peptide or medium alone. After labeling with pentamer, anti-CD8, and anti-CD3 Abs, PBMC were fixed in 1% paraformaldehyde for flow cytometry as described above.

All values throughout the text are means ± SD. Values of p were calculated with the rank sum test. Correlations were identified by simple linear regression analysis and Spearman’s rank correlation test. SigmaStat 3.5 software was used (Systat Software).

In a previous study we found that undetectable viremia in 9 out of 10 HICs was associated with a remarkably strong capacity of their circulating CD8+ T cells to control in vitro HIV-1 infection of autologous CD4+ T cells (21). To extend this observation, we used the same viral suppression assay to assess the ex vivo anti-HIV capacity of CD8+ T cells from 19 HICs, 11 of whom were newly recruited for this study and 8 were retested (Table I). Viral replication was readily detected in the supernatants of purified CD4+ T cells from all 19 HICs after PHA activation and challenge with HIV-1 BaL (Fig. 1,A). A marked reduction in HIV-1 infection (undetectable in eight HICs) was generally observed when autologous unstimulated CD8+ T cells from HICs were added to the culture (Fig. 1,A). The associated CD8+ T cell-mediated decrease in the level of HIV proteins was due to the absence of infected CD4+ T cells in the coculture (Fig. 1,B). As a whole, the HIV-suppressive capacity of CD8+ T cells from HICs (2.79 ± 1.31 log p24 decrease, CD8/CD4 vs CD4) was much stronger than that of cells both from viremic individuals (0.82 ± 0.53 log p24 decrease, CD8/CD4 vs CD4), confirming our previous results (21), as well as from HAART-treated individuals with undetectable viral load (0.62 ± 0.63 log p24 decrease, CD8/CD4 vs CD4) (Fig. 1,C). In particular, CD8+ T cells from 14 of the 19 HICs suppressed HIV far more strongly (log p24 decrease >2) than did cells from both viremic and treated individuals (Fig. 1,C). These subjects are referred to below as strong responder HICs. Longitudinal analysis (>12 mo) of CD8+ T cell antiviral activity in five strong responder HICs included in our previous study suggested that this HIV-suppressive capacity is a stable characteristic (Table II). In contrast, here we identified five “weak responder” HICs (Table I) whose CD8+ T cells could not efficiently control HIV infection of autologous CD4+ T cells (log p24 decrease <2) (Fig. 1, B and C): the HIV-suppressive capacity of these subjects’ CD8+ T cells was not stronger than that of viremic or HAART-treated patients (Fig. 1,C). We have reported that susceptibility of CD4+ T cells from HICs to in vitro HIV infection was not different than that of cells from healthy blood donors (21), and no significant differences were found either between weak responder and strong responder HICs (p = 0.331) (Fig. 1 A).

FIGURE 1.

A, PHA-activated CD4+ T cells were infected, in the absence (•) or presence (○) of autologous unstimulated CD8+ T cells (1:1 ratio), with replicative HIV-1 BaL. Circles represent the average (n = 3 independent infections) peak p24 values detected in culture supernatants for each individual. Horizontal lines indicate median values. B, PHA-activated CD4+ T cells from HICs A6 and A13 were superinfected with HIV-1 Bal and left alone (central panels) or cocultured with autologous unstimulated CD8+ T cells (right panels). Three days later the level of infection was determined by quantifying intracellular p24 on CD8 cells. C, The HIV-suppressive capacity of CD8+ T cells, as determined by the log fold decrease in the level of secreted p24 (CD4 vs CD4/CD8 cell cultures), was compared in the 19 HICs (•), 13 chronically HIV-infected subjects with viremia >7000 copies/ml (▴), and 8 HAART-treated patients with virologic control (plasma HIV RNA <50 copies/ml) for >23 mo (▪). Horizontal lines indicate median values. D, p24 production in culture supernatants (mean ± SD, n = 3) at the peak of viral replication after superinfection of CD4+ T cells from HIV controllers with equivalent infectious doses (moi of 10−3.6) of HIV-1 Bal or filtered supernatants containing autologous HIC viruses. CD4+ T cells were cultured alone (filled bars) or in the presence of non prestimulated CD8+ T cells (open bars).

FIGURE 1.

A, PHA-activated CD4+ T cells were infected, in the absence (•) or presence (○) of autologous unstimulated CD8+ T cells (1:1 ratio), with replicative HIV-1 BaL. Circles represent the average (n = 3 independent infections) peak p24 values detected in culture supernatants for each individual. Horizontal lines indicate median values. B, PHA-activated CD4+ T cells from HICs A6 and A13 were superinfected with HIV-1 Bal and left alone (central panels) or cocultured with autologous unstimulated CD8+ T cells (right panels). Three days later the level of infection was determined by quantifying intracellular p24 on CD8 cells. C, The HIV-suppressive capacity of CD8+ T cells, as determined by the log fold decrease in the level of secreted p24 (CD4 vs CD4/CD8 cell cultures), was compared in the 19 HICs (•), 13 chronically HIV-infected subjects with viremia >7000 copies/ml (▴), and 8 HAART-treated patients with virologic control (plasma HIV RNA <50 copies/ml) for >23 mo (▪). Horizontal lines indicate median values. D, p24 production in culture supernatants (mean ± SD, n = 3) at the peak of viral replication after superinfection of CD4+ T cells from HIV controllers with equivalent infectious doses (moi of 10−3.6) of HIV-1 Bal or filtered supernatants containing autologous HIC viruses. CD4+ T cells were cultured alone (filled bars) or in the presence of non prestimulated CD8+ T cells (open bars).

Close modal
Table II.

Log p24 decrease (CD4 vs CD4/CD8 of 1:1) during follow-up (>12 mo) of HICs

HICnaMedian ValueFirst SampleLast Sample
A1 3.5 Sept 2005 4.3 May 2008 3.5 
A3 3.1 Jan 2006 3.1 Jan 2007 3.3 
A4 3.8 Jan 2005 4.2 Apr 2008 3.5 
A6 3.8 Sept 2005 4.3 Nov 2007 3.8 
B5 3.5 Oct 2005 1.2 May 2008 3.5 
HICnaMedian ValueFirst SampleLast Sample
A1 3.5 Sept 2005 4.3 May 2008 3.5 
A3 3.1 Jan 2006 3.1 Jan 2007 3.3 
A4 3.8 Jan 2005 4.2 Apr 2008 3.5 
A6 3.8 Sept 2005 4.3 Nov 2007 3.8 
B5 3.5 Oct 2005 1.2 May 2008 3.5 
a

Number of blood samples analyzed.

To determine whether the weak HIV-suppressive activity observed in certain HICs was due to our use of a laboratory-adapted HIV strain, we analyzed the capacity of nonstimulated CD8+ T cells from weak responders A13 and A19 and from strong responder A21 to suppress superinfection of their own CD4+ T cells by autologous viruses previously obtained in primary culture of these individuals’ cells (see below). CD8+ T cells from strong responder A21 equally controlled CD4+ T cell superinfection by HIV-BaL and by autologous virus (Fig. 1,D). In contrast, the weak CD8-mediated HIV suppression in subject A13 was not improved when his autologous virus was used to challenge his CD4+ T cells (0.01 vs 0.16 log p24 decrease with HIV-BaL and the autologous virus, respectively) (Fig. 1,D). CD8+ T cells from weak responder A19 showed a stronger capacity to inhibit infection by autologous viruses (0.33 vs 1.76 log p24 decrease for HIV BaL and autologous virus infection, respectively) (Fig. 1,D), although the level of suppression did not reach that observed in strong responders. Interestingly, while most HICs were infected by subtype B viruses, subject A19 was infected by HIV-1 subtype A2 (Table I). Therefore, although the use of nonautologous viruses might lead to an underestimation of the HIV-suppressive activity of CD8+ T cells, it was unlikely to explain the differences observed between weak and strong responders.

We examined whether the difference between strong and weak responder HICs was associated with a difference in the magnitude of HIV-specific CD8+ T cell responses. To quantify the HIV-specific CD8+ T cell response, we used the standard determination of the frequency of IFN-γ-secreting CD8+ T cells upon stimulation with appropriate HLA-defined optimal HIV-1 Env, Gag, Pol, and Nef peptides in an ELISPOT assay. The numbers of IFN-γ-secreting cells were heterogeneous (Fig. 2,A), in agreement with recent reports (20, 24). The highest frequencies of HIV-specific CD8+ T cells were observed in strong responders (8517 ± 4038 vs 1058 ± 903 SFCs/106 PBMC in weak responder HICs, p = 0.0014) (Fig. 2 A). The frequency of HIV-specific CD8+ T cells in HICs was not significantly different, as a whole (6843 ± 4866 SFCs/106 PBMC), from that observed in chronically viremic patients (4616 ± 4148 SFCs/106 PBMC for 18 patients with >3 years of infection and plasma viral load >1000 RNA copies/ml, p = 0.20). The magnitude of the CD8+ T cell response in weak responder HICs was similar to that in HAART-treated patients (865 ± 1071 SFCs/106 PBMC for 11 patients with >2 years of treatment and plasma viral load <50 RNA copies/ml, p = 0.50; and Ref. 32).

FIGURE 2.

A, Frequencies of HIV-specific IFN-γ-secreting CD8+ T cells in strong responder HICs (log p24 decrease >2) (SR) and in weak responder HICs (log p24 decrease <2) (WR). An average of 36 ± 9 peptides were tested in each subject, depending on the results of HLA typing. Each symbol corresponds to the sum of SFCs/106 PBMC obtained with individual peptides described as being restricted by HLA Ags. Horizontal lines are median values for each group. B, Correlation between the HIV-suppressive capacity of CD8+ T cells from HICs (log p24 decrease as shown in Fig. 1 C) and their frequency of IFN-γ-producing CD8+ T cells upon HIV peptide stimulation. Each symbol represents one HIC. Vertical dashed line separates weak responder and strong responder HICs. C, Percentage of HIV-specific cells (based on HIV multimer and CD8 expression) from strong and weak responder HICs that expressed ex vivo HLA-DR and CD38, coexpressed CD27 and CD45RA, or proliferated (and lost CFSE labeling) after 5 days of peptide stimulation. Each symbol represents one specificity for one HIC. Horizontal lines are mean values for each group.

FIGURE 2.

A, Frequencies of HIV-specific IFN-γ-secreting CD8+ T cells in strong responder HICs (log p24 decrease >2) (SR) and in weak responder HICs (log p24 decrease <2) (WR). An average of 36 ± 9 peptides were tested in each subject, depending on the results of HLA typing. Each symbol corresponds to the sum of SFCs/106 PBMC obtained with individual peptides described as being restricted by HLA Ags. Horizontal lines are median values for each group. B, Correlation between the HIV-suppressive capacity of CD8+ T cells from HICs (log p24 decrease as shown in Fig. 1 C) and their frequency of IFN-γ-producing CD8+ T cells upon HIV peptide stimulation. Each symbol represents one HIC. Vertical dashed line separates weak responder and strong responder HICs. C, Percentage of HIV-specific cells (based on HIV multimer and CD8 expression) from strong and weak responder HICs that expressed ex vivo HLA-DR and CD38, coexpressed CD27 and CD45RA, or proliferated (and lost CFSE labeling) after 5 days of peptide stimulation. Each symbol represents one specificity for one HIC. Horizontal lines are mean values for each group.

Close modal

Interestingly, we found a strong correlation between the frequency of IFN-γ-producing CD8+ T cells upon peptide stimulation and the HIV-suppressive capacity of unstimulated CD8+ T cells (Spearman 0.835, p < 0.00001) (Fig. 2,B). This supports the possibility that the ex vivo anti-HIV activity of CD8+ T cells from HICs is driven by HIV-specific cells, in keeping with an MHC class I-mediated mechanism (21). This correlation further distinguished strong and weak responder HICs (Fig. 2 B).

We explored whether differences could also be observed between strong and weak responder HICs at the phenotypical level of their HIV-specific CD8+ T cells. Due to the low frequency of these cells in weak responders, we could perform these analyses only in three of them. HIV-specific CD8+ T cells from strong responder HICs possessed a discordant activation phenotype with high expression of the activation marker HLA-DR associated with a low CD38 expression (Fig. 2,C), in keeping with our previous study (21). In contrast, the expression of both activation markers was low in the cells from weak responders (Fig. 2,C), a phenotype that is found in HAART subjects (21). HIV-specific CD8+ T cells from both strong and weak responders had high proliferative potential (Fig. 2,C), which is a hallmark of a high-quality HIV-specific CD8+ T cell response in HICs (26). Interestingly, we found in weak responders an increase of a subpopulation of HIV-specific CD8+ T cells characterized by the coexpression of CD27 and CD45RA (Fig. 2 C). We have recently reported that this subpopulation is characteristically abundant in HIV patients treated during acute primary HIV infection and may represent a stable quiescent long-term memory pool (33).

The response to Gag contributed most (average, 51.8%) to the total HIV-specific CD8+ T cell response (Fig. 3,A). In strong responder HICs the contribution of the Gag response was 56.8% on average compared with 37.9% in weak responder HICs (p = 0.14). Responses to Nef peptides also contributed significantly to the overall CD8+ T cell response in HICs (average, 31.8% in strong responders and 31.2% in weak responders) (Fig. 3,A). The contributions of Env and Pol responses were much smaller (7.9% and 8.7%, respectively, in the whole HIC population) (Fig. 3 A). The contributions of the responses to the different HIV proteins were not different in HICs than in viremic patients (not shown), although a tendency was observed to a greater contribution of Gag responses in strong responder HICs than in viremics (average, 38%; p = 0.08). The magnitude of the Gag response was higher in HICs (3682 ± 2969 SFCs/106 PBMC) than in viremics (1703 ± 2061 SFCs/106 PBMC, p = 0.05). In contrast, Gag responses contributed less and Nef responses more to the total HIV-specific CD8+ T cell response in HAART-treated patients (14.3% and 60.6% of Gag and Nef responses) than in HICs (p = 0.014 and p = 0.025, respectively).

FIGURE 3.

A, Percentage of the HIV-specific CD8+ T cell response that was due to CD8+ T cells secreting IFN-γ upon stimulation with Gag, Nef, Env, and Pol peptides. Each symbol represents one HIC. Circles represent strong responders; squares represent weak responders. Horizontal dashed lines are mean values for each group. B and C, Correlation between the HIV-suppressive capacity of CD8+ T cells from HICs and their frequency of IFN-γ-producing CD8+ T cells upon stimulation with Gag and Nef peptides, respectively. Each symbol represents one HIC. Vertical dashed line separates weak responder and strong responder HICs.

FIGURE 3.

A, Percentage of the HIV-specific CD8+ T cell response that was due to CD8+ T cells secreting IFN-γ upon stimulation with Gag, Nef, Env, and Pol peptides. Each symbol represents one HIC. Circles represent strong responders; squares represent weak responders. Horizontal dashed lines are mean values for each group. B and C, Correlation between the HIV-suppressive capacity of CD8+ T cells from HICs and their frequency of IFN-γ-producing CD8+ T cells upon stimulation with Gag and Nef peptides, respectively. Each symbol represents one HIC. Vertical dashed line separates weak responder and strong responder HICs.

Close modal

We then examined the influence of the specificity of HIC CD8+ T cells on the efficiency of HIV suppression. The correlation between the HIV-suppressive capacity of nonstimulated CD8+ T cells and the frequency of IFN-γ-producing CD8+ T cells upon peptide stimulation was strongest for Gag peptides (Spearman 0.907, p < 0.00001) (Fig. 3,B). This correlation was unlikely to be due to a bias for HLA-B57-restricted Gag responses since in the 13 individuals carrying this HLA allele, HIV-specific CD8+ T cell responses targeting HLA-B57 restricted Gag epitopes represented, on average, 26 ± 12% of their total response. Other responses were either restricted by HLA-B57 but not directed at Gag (19 ± 18%) or restricted by other alleles and directed at Gag (28 ± 28% of the response) or at other proteins (27 ± 20%). As mentioned, Nef was also a main target of the HIV-specific CD8+ response; however, only a weak correlation was found with the magnitude of Nef responses (Spearman 0.473, p = 0.040) (Fig. 3 C). Furthermore, this correlation with Nef responses was completely lost when weak responder HICs were excluded from the analyses (Spearman 0.070, p = 0.797). Most interestingly, in the group of strong responder HICs, the CD8+ T cell anti-HIV capacity still correlated more tightly with the magnitude of Gag responses (Spearman 0.812, p < 0.00001) than with the total frequency of IFN-γ-producing CD8+ T cells (Spearman 0.634, p = 0.007). Overall these results suggest that the numbers of CD8+ T cells responding to Gag epitopes influence the capacity of CD8+ T cells from HICs to suppress HIV infection of autologous CD4+ T cells.

To evaluate more directly the impact of Gag responses in the HIV-suppressive activity of CD8+ T cells from strong responder HICs, we first tried to compare the HIV-suppressive capacity of FACSAria-sorted pentamer-positive cell fractions. Unfortunately, and despite a fairly good viability, functionality of these cells was compromised. Hence, we compared the relative weight of Gag and Nef responses by assessing the HIV-suppressive capacity of CD8+ T cell fractions depleted of either one response or the other. These experiments were performed with cells from three HICs (A3, A6, and A11) with similar numbers of HIV-specific cells (11,270, 12,473, and 12,612 SFC/106 PBMC, respectively) and a contribution of the Gag response to the total HIV-CD8+ T cell response close to 50% (49.8%, 63.0%, and 55.3%, respectively). CD8+ T cells isolated from HICs were stimulated with 1) a pool of all the optimal HIV-1 peptides that were recognized in individual ELISPOT assays (not shown); (2) a pool of Gag peptides only; (3) a pool of Nef peptides only. As shown in Fig. 4,A, the suppression of HIV infection observed when autologous unstimulated CD8+ T cells from strong responder HICs were added to CD4+ T cell cultures was lost when the CD8+ T cells that produced IFN-γ upon stimulation with the complete pool of recognized HIV peptides were removed. In the case of A3, both the cell fractions depleted of Gag-specific or Nef-specific CD8+ T cells retained strong HIV-suppressive capacity (Fig. 4,B). For A11, depletion of Gag-specific cells caused the nearly complete loss of HIV-suppressive capacity, whereas depletion of Nef-specific cells had no effect (Fig. 4,B). For A6, the depletion of Gag-specific cells also caused a strong loss of HIV-suppressive capacity (Fig. 4 B). Removal of Nef-specific cells occasioned a more modest loss of HIV-suppressive capacity. In summary, although the respective contributions of Gag and Nef responses were difficult to quantify precisely, Gag-specific CD8+ T cells seemed to strongly contribute to the HIV-suppressive capacity of CD8+ T cells in all three strong responder HICs evaluated, in agreement with the correlations described above. In contrast, the contribution of Nef responses was more variable.

FIGURE 4.

A, p24 production in culture supernatants (mean ± SD, n = 3) at the peak of viral replication after superinfection of CD4+ T cells from A6 with HIV-1 Bal. CD4+ T cells were cultured alone or in the presence (CD4/CD8 of 1:1) of non-prestimulated CD8+ T cells or CD8+ T cells depleted of HIV-specific CD8+ T cells. These results are representative of experiments with 5 HICs. B, HIV-suppressive capacity of CD8+ T cells (mean ± SD, n = 3), as determined by the log fold decrease in the level of secreted p24 (CD4 vs CD4/CD8, 1:1 cell cultures), after depletion of HIV-specific cells (black bars), Gag-specific (open bars), or Nef specific (gray bars) cell fractions. Pie charts at the bottom represent the relative contribution of Gag and Nef responses to the total HIV-specific CD8+ T cell response. C, CD4+ T cells from HIV controllers were infected with replicative HIV-1.BaL (black bars), SIVagm.Gril (open bars), SIVmac.251 (gray bars), or HIV-2.SBL (patterned bars) and cultured alone or with autologous unstimulated CD8+ T cells. Viral replication was monitored by p24 or p27 ELISA on culture supernatants. Bars indicate the level of suppression at the peak of viral replication when CD8+ T cells were present in the culture (mean ± SD, n = 3). n.d., Experiment not done.

FIGURE 4.

A, p24 production in culture supernatants (mean ± SD, n = 3) at the peak of viral replication after superinfection of CD4+ T cells from A6 with HIV-1 Bal. CD4+ T cells were cultured alone or in the presence (CD4/CD8 of 1:1) of non-prestimulated CD8+ T cells or CD8+ T cells depleted of HIV-specific CD8+ T cells. These results are representative of experiments with 5 HICs. B, HIV-suppressive capacity of CD8+ T cells (mean ± SD, n = 3), as determined by the log fold decrease in the level of secreted p24 (CD4 vs CD4/CD8, 1:1 cell cultures), after depletion of HIV-specific cells (black bars), Gag-specific (open bars), or Nef specific (gray bars) cell fractions. Pie charts at the bottom represent the relative contribution of Gag and Nef responses to the total HIV-specific CD8+ T cell response. C, CD4+ T cells from HIV controllers were infected with replicative HIV-1.BaL (black bars), SIVagm.Gril (open bars), SIVmac.251 (gray bars), or HIV-2.SBL (patterned bars) and cultured alone or with autologous unstimulated CD8+ T cells. Viral replication was monitored by p24 or p27 ELISA on culture supernatants. Bars indicate the level of suppression at the peak of viral replication when CD8+ T cells were present in the culture (mean ± SD, n = 3). n.d., Experiment not done.

Close modal

We have already reported a broad capacity of CD8+ T cells from strong responder HICs to effectively control superinfection by different HIV-1 subtypes (21). Interestingly, CD8+ T cells from strong responder HICs also partially suppressed infection of CD4+ T cells by other human lentiviruses such as HIV-2, SIVmac, and SIVagm (Fig. 4 C). At least some of the HIV-1 epitopes recognized by HIV-specific CD8+ T cells from strong responders HICs were conserved in the other lentiviruses used in our experiments (not shown). In accordance with a MHC-mediated mechanism, the capacity to suppress SIV infection was totally lost when CD8+ T cells were separated from autologous CD4+ T cells by semipermeable membranes (not shown), as was shown in the case of HIV-1 (21).

A recent report by Hatano et al. suggested that low level viral replication is ongoing in most HICs (22). We thus examined whether differences between strong responder and weak responder HICs might exist at a virological level. Ultrasensitive viral load tests were not available for this study. However, given the long documented virological follow-up of the patients in the study, we had access to multiple RNA viral load determinations for all HICs (Table I). The length of the follow-up and the number of viral load determinations were similar for strong responder and weak responder HICs (p = 0.309 and p = 0.515, respectively; Table I). Interestingly, historical plasma viral load results showed that small blips of viral RNA were more frequently detected during follow-up among strong responder HICs than among HICs with weak CD8+ T cell responses, who appeared to control HIV infection more tightly (p = 0.016; Fig. 5,A and Table I).

FIGURE 5.

A, Frequency of viral load determinations with values >50 HIV RNA copies/ml of plasma during follow-up and (B) total HIV-DNA in blood cells at inclusion, for strong responder HICs (SR, •) and weak responder HICs (WR, □). Each symbol represents one HIC. C, Correlation between peak p24 production detected in the supernatant of 105 CD4+ T cells from weak responder (□) and strong responders HICs (•) upon PHA stimulation (mean of three values) and HIV-suppressive capacity of CD8+ T cells from 16 HICs. Each symbol represents one HIC. The dashed line represents the background level. D, Kinetics of viral replication (3, 7, and 10 days postinfection) after infection (1.2 ng of p24/106 cells) of CD4+ T cells from a single healthy blood donor. Viruses from subjects A13 and A19 were obtained at 10 and 14 days, respectively, of culture of PHA-activated CD4+ T cells. Viruses from A22 were obtained after 8 days of culture of PHA-activated CD4+ T cells and 5 additional days of culture in the presence of heterologous PHA-activated CD4+ T cells. Open bars represent laboratory-adapted viruses, gray bars primary isolates, and black bars HIC-derived viruses. The mean and SD are shown (n = 3).

FIGURE 5.

A, Frequency of viral load determinations with values >50 HIV RNA copies/ml of plasma during follow-up and (B) total HIV-DNA in blood cells at inclusion, for strong responder HICs (SR, •) and weak responder HICs (WR, □). Each symbol represents one HIC. C, Correlation between peak p24 production detected in the supernatant of 105 CD4+ T cells from weak responder (□) and strong responders HICs (•) upon PHA stimulation (mean of three values) and HIV-suppressive capacity of CD8+ T cells from 16 HICs. Each symbol represents one HIC. The dashed line represents the background level. D, Kinetics of viral replication (3, 7, and 10 days postinfection) after infection (1.2 ng of p24/106 cells) of CD4+ T cells from a single healthy blood donor. Viruses from subjects A13 and A19 were obtained at 10 and 14 days, respectively, of culture of PHA-activated CD4+ T cells. Viruses from A22 were obtained after 8 days of culture of PHA-activated CD4+ T cells and 5 additional days of culture in the presence of heterologous PHA-activated CD4+ T cells. Open bars represent laboratory-adapted viruses, gray bars primary isolates, and black bars HIC-derived viruses. The mean and SD are shown (n = 3).

Close modal

HIV-1 DNA level in blood cells, which is a stable parameter that gives an estimation of the HIV-1 reservoir size (34), was available for most HICs (Table I). Despite the differences in the frequency of viral RNA blips mentioned above, proviral DNA levels were very low in all the HICs, regardless of the strength of their CD8+ T cell responses (Fig. 5,B). We then investigated whether autologous viral replication might be activated upon stimulation of CD4+ T cells from HICs. Surprisingly, replication-competent viruses were more readily detected in the supernatants of activated CD4+ T cells from weak responders than from strong responders (Fig. 5,C). Moreover, autologous virus production upon CD4+ T cell stimulation correlated negatively with the HIV-suppressive capacity of CD8+ T cells (Spearman −0.635, p = 0.01). We obtained enough autologous viruses from weak responder HICs A13, A19, and A22 to test their infectivity. These viruses were able to spread and infect heterologous CD4+ T cells as efficiently as other laboratory-adapted strains and primary isolates (Fig. 5 D). Their titers (6.1, 5.6, and 5.5 50% tissue culture-infective dose/ml for vA13, vA19, and vA22, respectively) were also similar (5.4 TCID50/ml for both BaL and NL4.3, and 6.1 TCID50/ml for v30007). Therefore, at least some HICs with weak CD8+ T cell responses carry viruses highly replicative in vitro. This is in agreement with recent reports showing that defective or attenuated viruses do not generally account for the control of viral replication in HICs (35, 36, 37).

Here we show that the HIV-suppressive capacity of CD8+ T cells from HIV controllers is stable over time and is associated with the magnitude of HIV-specific CD8+ T cell responses, in particular to those directed against Gag. We also identify a group of HICs who carry infectious viruses and are able to durably control HIV infection despite a weak HIV-suppressive capacity of their CD8+ T cells.

Most of the HIC subjects in our study (14 of 19) had CD8+ T cells with marked and stable HIV-suppressive capacities (strong responder HICs, p24 log decrease >2) that we have never observed in viremic (21) or HAART-treated individuals. The protective HLA alleles B27 and/or B57 were present in all strong responder HICs. However, CD8+ T cells from a subgroup of HICs had only weak HIV-suppressive capacity. In agreement with recent reports (20, 24), the HICs we studied had heterogeneous levels of HIV-specific CD8+ T cells, as estimated by the frequency of IFN-γ-producing CD8+ T cells. The magnitude of the HIV-specific CD8+ T cell response correlated strongly with the capacity of CD8+ T cells from HICs to control HIV infection of autologous CD4+ T cells in vitro. Accordingly, the lowest frequencies of IFN-γ-producing CD8+ T cells were found in weak responder HICs.

Some underestimation of the CD8+ T cell response in HICs may come for the use of peptides derived from consensus sequences for ELISPOT determinations, or of a laboratory-adapted HIV strain for HIV-suppression analyses. However, CD8+ T cells from two weak responder HICs had limited suppressive capacity even when autologous viruses were used, which further supported a truly weak CD8+ T cell response in these individuals. We cannot exclude that control of viremia in weak responders may be due to robust HIV-specific CD8+ T cell responses in lymphoid tissues, and actually Ferre and collegues have recently shown that HICs have polyfunctional HIV-specific T cell responses in rectal mucosa that were frequently stronger than in blood (38). However, the presence in this study of a few HICs with very weak responses both in the blood and in the rectal mucosa is interesting. Although a weak high quality CD8+ T cell response might be sufficient to control viremia in vivo, it seems unlikely to be the case in weak responder HICs. The absence of viral blips during the follow-up of weak responders and our finding that at least some of these HICs carry viruses that are highly infectious in vitro and readily detectable upon in vitro activation endorse the idea of a very tight and active host-restraint of HIV-1 infection. Our phenotypical analyses of the HIV-specific CD8+ T cells in weak responders showed an increased proportion of a CD27+CD45RA+ subset of cells, previously observed in patients treated during primary HIV infection, and that might represent a quiescent and stable memory pool able to proliferate and acquire effector capacities upon Ag stimulation (33). Although these cells may provide an effective response in the eventuality of viral replication, their increased proportion in weak responder HICs together with the low expression of HLA-DR suggest a long period without antigenic stimulation of the CD8+ T cell response.

Therefore, an alternative mechanism is probably responsible for controlling HIV-1 in these HICs. The lower antiviral activity of CD8+ T cells in weak responder HICs did not seem to be compensated for by other cell populations within PBMC (e.g., NK cells or γδ T cells), as illustrated by HIV-suppressive experiments where nonstimulated PBMC (depleted of CD4+ cells), used instead of CD8+ T cells, were also unable to control HIV superinfection of autologous CD4+ T cells (not shown). Interestingly, persistent lack of low-level detectable viremia in one HIC has been recently associated to low levels of HIV Abs and remarkably low levels of T cell activation (22). Further virologic studies (such as viral sequencing or determination of tissular viral replication) and the analysis of innate responses and regulatory T cells (39) might help to identify new mechanisms of control in HICs.

Unlike the cells from weak responder HICs, CD8+ T cells from strong responder HICs had a broad capacity to suppress superinfection of their own CD4+ T cells by a wide range of HIV-1 strains (21) and, at least partially, by other lentiviruses. This could be related to the presence of HIV-specific CD8+ T cells targeting epitopes located within highly conserved regions of the virus. Responses against Gag and Nef epitopes together accounted for the bulk of total CD8+ T cell responses in strong responder HICs, and no phenotypic differences were observed between Gag-specific and Nef-specific CD8+ T cell responses in these individuals (21). Interestingly, we observed a strong correlation between the HIV-suppressive capacity of CD8+ T cells in strong responder HICs and the number of Gag-specific CD8+ T cell responses. Moreover, the analysis of the relative HIV-suppressive capacity of the Gag response in three strong responder HICs showed that, for all three HICs, Gag-specific CD8+ T cells possess the strongest antiviral capacities. Thus, Gag responses appear to be strongly involved in the antiviral potency of CD8+ T cells. This is in agreement with a report showing evidence of CD8+ T cell selective pressure on gag in HICs (40). Increasing evidence suggests that Gag-specific CD8+ and CD4+ T cell responses are associated with better control of HIV viremia (7, 8, 9, 20, 41). Gag epitopes are presented on the surface of infected CD4+ T cells early after viral entry, before DNA integration and viral protein synthesis (42), and this might allow Gag-specific CD8+ T cells to recognize and eliminate infected cells before the infection is properly established and before Nef-mediated down-regulation of MHC class I molecules occurs (43). Other factors such as functional avidity (41, 44) or lytic granule loading (45) might contribute to an enhanced HIV-suppressive capacity of Gag-specific CD8+ T cells.

No correlation was found between HIV-suppressive capacity of CD8+ T cells in strong responder HICs with Nef-specific CD8+ T cell responses. However, this observation does not preclude a contribution of responses targeting Nef (or other viral proteins) to the HIV-suppressive capacity of CD8+ T cells. Actually, our experiments of selective depletion of HIV-specific cell fractions showed variable capacities (from strong to none) of Nef-specific CD8+ T cells from HICs to suppress HIV infection, perhaps depending on the frequency of the Nef responses that were targeting epitopes restricted by HLA-B57. Along these lines, escaping mutations are also found in Nef epitopes in HICs, although less frequently than in Gag epitopes (46).

Escaping mutations in epitopes located within structurally important regions of the virus could limit the capacity of the virus to mutate to escape immune pressure, as variations in these regions have a fitness cost (13, 47). Although we did not directly address this issue, the difficulties to detect HIV-1 replication in the culture supernatants of activated CD4+ T cells from strong responder HICs might reflect the impact of the pressure exerted by CD8+ T cell responses on viral fitness. Nevertheless, we cannot exclude that, given the extraordinary antiviral potency of CD8+ T cells from strong responder HICs, the few remaining CD8+ T cells in the >97% pure CD4+ T cell fractions used in these experiments were enough to efficiently suppress autologous virus replication.

Several important questions await answers; that is, mainly whether the potent CD8+ T cell response observed in most HICs precedes or follows initial viremic control, and how such a potent CD8+ T cell response is maintained. The association presented here between blips in plasma viral RNA and stronger CD8+ T cell responses in HICs must be considered with care because of the limited number of weak responder HICs, but it is tempting to speculate that CD8+ T cell control of HIV might involve a feedback mechanism whereby occasional blips (or ongoing low-level viral replication) are needed to boost the antiviral response. The increased telomerase activity in these cells would further ensure their persistence (28). Two scenarios can be envisaged: 1) if viremia is controlled by a common mechanism in weak and strong responder HICs, the presence of the protective HLA B27 and B57 alleles may help to sustain control over time, in the eventuality of viral escape, through the establishment of a robust CD8+ T cell response; 2) different mechanisms are responsible for initial control of HIV infection in weak and strong responder HICs. Detailed longitudinal studies of HICs will be necessary to answer these questions.

We thank all the members of the French National Agency for Research on AIDS and Viral Hepatitis (ANRS) EP36 HIV Controllers study group for helpful discussions. We thank Chiraz Hamimi for technical help. We also thank Dr. Laurence Meyer, Dr. Daniel Séréni, Dr. Caroline Lascoux, Dr. Olivier Taulera, Jeannine Delgado, Dr. François Bricaire, Dr. Michèle Bentata, Dr. Pascale Kousignian, Michèle Pauchard, Dr. Alain Krivitzky, Patricia Honoré, Marie-Thérèse Rannou, Dr. Jean-Paul Viard, Dr. David Zucman, Nadège Velazquez, and all the other physicians and nurses who cared for the patients. We especially thank the subjects who participated in this study for their cooperation.

The authors have no financial conflicts of interest.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1

This work was financially supported by the Agence Nationale de Recherches sur le SIDA. S.Y.S. was supported by Korea Science and Engineering Foundation and the Institut Pasteur Korea.

5

Abbreviations used in this paper: HIC, HIV controller; moi, multiplicity of infection; PBMC, peripheral blood mononuclear cell; SFC, spot-forming cell.

1
Borrow, P., H. Lewicki, B. H. Hahn, G. M. Shaw, M. B. Oldstone.
1994
. Virus-specific CD8+ cytotoxic T-lymphocyte activity associated with control of viremia in primary human immunodeficiency virus type 1 infection.
J. Virol.
68
:
6103
-6110.
2
Friedrich, T. C., L. E. Valentine, L. J. Yant, E. G. Rakasz, S. M. Piaskowski, J. R. Furlott, K. L. Weisgrau, B. Burwitz, G. E. May, E. J. Leon, et al
2007
. Subdominant CD8+ T-cell responses are involved in durable control of AIDS virus replication.
J. Virol.
81
:
3465
-3476.
3
Jin, X., D. E. Bauer, S. E. Tuttleton, S. Lewin, A. Gettie, J. Blanchard, C. E. Irwin, J. T. Safrit, J. Mittler, L. Weinberger, et al
1999
. Dramatic rise in plasma viremia after CD8+ T cell depletion in simian immunodeficiency virus-infected macaques.
J. Exp. Med.
189
:
991
-998.
4
Flores-Villanueva, P. O., E. J. Yunis, J. C. Delgado, E. Vittinghoff, S. Buchbinder, J. Y. Leung, A. M. Uglialoro, O. P. Clavijo, E. S. Rosenberg, S. A. Kalams, et al
2001
. Control of HIV-1 viremia and protection from AIDS are associated with HLA-Bw4 homozygosity.
Proc. Natl. Acad. Sci. USA
98
:
5140
-5145.
5
Kiepiela, P., A. J. Leslie, I. Honeyborne, D. Ramduth, C. Thobakgale, S. Chetty, P. Rathnavalu, C. Moore, K. J. Pfafferott, L. Hilton, et al
2004
. Dominant influence of HLA-B in mediating the potential co-evolution of HIV and HLA.
Nature
432
:
769
-775.
6
Migueles, S. A., M. S. Sabbaghian, W. L. Shupert, M. P. Bettinotti, F. M. Marincola, L. Martino, C. W. Hallahan, S. M. Selig, D. Schwartz, J. Sullivan, M. Connors.
2000
. HLA B*5701 is highly associated with restriction of virus replication in a subgroup of HIV-infected long term nonprogressors.
Proc. Natl. Acad. Sci. USA
97
:
2709
-2714.
7
Edwards, B. H., A. Bansal, S. Sabbaj, J. Bakari, M. J. Mulligan, P. A. Goepfert.
2002
. Magnitude of functional CD8+ T-cell responses to the Gag protein of human immunodeficiency virus type 1 correlates inversely with viral load in plasma.
J. Virol.
76
:
2298
-2305.
8
Kiepiela, P., K. Ngumbela, C. Thobakgale, D. Ramduth, I. Honeyborne, E. Moodley, S. Reddy, C. de Pierres, Z. Mncube, N. Mkhwanazi, et al
2007
. CD8+ T-cell responses to different HIV proteins have discordant associations with viral load.
Nat. Med.
13
:
46
-53.
9
Riviere, Y., M. B. McChesney, F. Porrot, F. Tanneau-Salvadori, P. Sansonetti, O. Lopez, G. Pialoux, V. Feuillie, M. Mollereau, S. Chamaret, et al
1995
. Gag-specific cytotoxic responses to HIV type 1 are associated with a decreased risk of progression to AIDS-related complex or AIDS.
AIDS Res. Hum. Retroviruses
11
:
903
-907.
10
Chopera, D. R., Z. Woodman, K. Mlisana, M. Mlotshwa, D. P. Martin, C. Seoighe, F. Treurnicht, D. A. de Rosa, W. Hide, S. A. Karim, et al
2008
. Transmission of HIV-1 CTL escape variants provides HLA-mismatched recipients with a survival advantage.
PLoS Pathog.
4
:
e1000033
11
Martinez-Picado, J., J. G. Prado, E. E. Fry, K. Pfafferott, A. Leslie, S. Chetty, C. Thobakgale, I. Honeyborne, H. Crawford, P. Matthews, et al
2006
. Fitness cost of escape mutations in p24 Gag in association with control of human immunodeficiency virus type 1.
J. Virol.
80
:
3617
-3623.
12
Rousseau, C. M., M. G. Daniels, J. M. Carlson, C. Kadie, H. Crawford, A. Prendergast, P. Matthews, R. Payne, M. Rolland, D. N. Raugi, et al
2008
. HLA class I-driven evolution of human immunodeficiency virus type 1 subtype c proteome: immune escape and viral load.
J. Virol.
82
:
6434
-6446.
13
Schneidewind, A., M. A. Brockman, R. Yang, R. I. Adam, B. Li, S. Le Gall, C. R. Rinaldo, S. L. Craggs, R. L. Allgaier, K. A. Power, et al
2007
. Escape from the dominant HLA-B27 restricted CTL response in Gag is associated with a dramatic reduction in HIV-1 replication.
J. Virol.
81
:
12382
-12393.
14
Goepfert, P. A., W. Lumm, P. Farmer, P. Matthews, A. Prendergast, J. M. Carlson, C. A. Derdeyn, J. Tang, R. A. Kaslow, A. Bansal, et al
2008
. Transmission of HIV-1 Gag immune escape mutations is associated with reduced viral load in linked recipients.
J. Exp. Med.
205
:
1009
-1017.
15
Deeks, S. G., B. D. Walker.
2007
. Human immunodeficiency virus controllers: mechanisms of durable virus control in the absence of antiretroviral therapy.
Immunity
27
:
406
-416.
16
Saez-Cirion, A., G. Pancino, M. Sinet, A. Venet, O. Lambotte.
2007
. HIV controllers: how do they tame the virus?.
Trends Immunol.
28
:
532
-540.
17
Lambotte, O., F. Boufassa, Y. Madec, A. Nguyen, C. Goujard, L. Meyer, C. Rouzioux, A. Venet, J. F. Delfraissy.
2005
. HIV controllers: a homogeneous group of HIV-1-infected patients with spontaneous control of viral replication.
Clin. Infect. Dis.
41
:
1053
-1056.
18
Dinoso, J. B., S. Y. Kim, R. F. Siliciano, J. N. Blankson.
2008
. A comparison of viral loads between HIV-1-infected elite suppressors and individuals who receive suppressive highly active antiretroviral therapy.
Clin. Infect. Dis.
47
:
102
-104.
19
Bailey, J. R., K. G. Lassen, H. C. Yang, T. C. Quinn, S. C. Ray, J. N. Blankson, R. F. Siliciano.
2006
. Neutralizing antibodies do not mediate suppression of human immunodeficiency virus type 1 in elite suppressors or selection of plasma virus variants in patients on highly active antiretroviral therapy.
J. Virol.
80
:
4758
-4770.
20
Emu, B., E. Sinclair, H. Hatano, A. Ferre, B. Shacklett, J. N. Martin, J. M. McCune, S. G. Deeks.
2008
. HLA class I-restricted T-cell responses may contribute to the control of human immunodeficiency virus infection, but such responses are not always necessary for long-term virus control.
J. Virol.
82
:
5398
-5407.
21
Sáez-Cirión, A., C. Lacabaratz, O. Lambotte, P. Versmisse, A. Urrutia, F. Boufassa, F. Barre-Sinoussi, J. F. Delfraissy, M. Sinet, G. Pancino, A. Venet.
2007
. HIV controllers exhibit potent CD8 T cell capacity to suppress HIV infection ex vivo and peculiar CTL activation phenotype.
Proc. Natl. Acad. Sci. USA
104
:
6776
-6781.
22
Hatano, H., E. L. Delwart, P. J. Norris, T. H. Lee, J. Dunn-Williams, P. W. Hunt, R. Hoh, S. L. Stramer, J. M. Linnen, J. M. McCune, et al
2009
. Evidence for persistent low-level viremia in individuals who control HIV in the absence of antiretroviral therapy.
J. Virol.
83
:
329
-335.
23
Gea-Banacloche, J. C., S. A. Migueles, L. Martino, W. L. Shupert, A. C. McNeil, M. S. Sabbaghian, L. Ehler, C. Prussin, R. Stevens, L. Lambert, et al
2000
. Maintenance of large numbers of virus-specific CD8+ T cells in HIV-infected progressors and long-term nonprogressors.
J. Immunol.
165
:
1082
-1092.
24
Pereyra, F., M. M. Addo, D. E. Kaufmann, Y. Liu, T. Miura, A. Rathod, B. Baker, A. Trocha, R. Rosenberg, E. Mackey, et al
2008
. Genetic and immunologic heterogeneity among persons who control HIV infection in the absence of therapy.
J. Infect. Dis.
197
:
563
-571.
25
Betts, M. R., M. C. Nason, S. M. West, S. C. De Rosa, S. A. Migueles, J. Abraham, M. M. Lederman, J. M. Benito, P. A. Goepfert, M. Connors, et al
2006
. HIV nonprogressors preferentially maintain highly functional HIV-specific CD8+ T cells.
Blood
107
:
4781
-4789.
26
Migueles, S. A., A. C. Laborico, W. L. Shupert, M. S. Sabbaghian, R. Rabin, C. W. Hallahan, D. Van Baarle, S. Kostense, F. Miedema, M. McLaughlin, et al
2002
. HIV-specific CD8+ T cell proliferation is coupled to perforin expression and is maintained in nonprogressors.
Nat. Immunol.
3
:
1061
-1068.
27
Zimmerli, S. C., A. Harari, C. Cellerai, F. Vallelian, P. A. Bart, G. Pantaleo.
2005
. HIV-1-specific IFN-γ/IL-2-secreting CD8 T cells support CD4-independent proliferation of HIV-1-specific CD8 T cells.
Proc. Natl. Acad. Sci. USA
102
:
7239
-7244.
28
Lichterfeld, M., D. Mou, T. D. Cung, K. L. Williams, M. T. Waring, J. Huang, F. Pereyra, A. Trocha, G. J. Freeman, E. S. Rosenberg, et al
2008
. Telomerase activity of HIV-1-specific CD8+ T cells: constitutive upregulation in controllers and selective increase by blockade of PD ligand 1 in progressors.
Blood
112
:
3679
-3687.
29
Schvachsa, N., G. Turk, M. Burgard, D. Dilernia, M. Carobene, M. Pippo, M. Gomez-Carrillo, C. Rouzioux, H. Salomon.
2007
. Examination of real-time PCR for HIV-1 RNA and DNA quantitation in patients infected with HIV-1 BF intersubtype recombinant variants.
J. Virol. Methods
140
:
222
-227.
30
Avettand-Fenoel, V., N. Mahlaoui, M. L. Chaix, C. Milliancourt, M. Burgard, M. Cavazzana-Calvo, C. Rouzioux, S. Blanche.
2007
. Failure of bone marrow transplantation to eradicate HIV reservoir despite efficient HAART.
AIDS
21
:
776
-777.
31
O'Doherty, U., W. J. Swiggard, M. H. Malim.
2000
. Human immunodeficiency virus type 1 spinoculation enhances infection through virus binding.
J. Virol.
74
:
10074
-10080.
32
Lacabaratz-Porret, C., A. Urrutia, J. M. Doisne, C. Goujard, C. Deveau, M. Dalod, L. Meyer, C. Rouzioux, J. F. Delfraissy, A. Venet, M. Sinet.
2003
. Impact of antiretroviral therapy and changes in virus load on human immunodeficiency virus (HIV)-specific T cell responses in primary HIV infection.
J. Infect. Dis.
187
:
748
-757.
33
Lecuroux, C., I. Girault, A. Urrutia, J. M. Doisne, C. Deveau, C. Goujard, L. Meyer, M. Sinet, A. Venet.
2009
. Identification of a particular HIV-specific CD8+ T cell subset with a CD27+ CD45RO/RA+ phenotype and memory characteristics after initiation of HAART during acute primary HIV infection.
Blood
113
:
3209
-3217.
34
Avettand-Fenoel, V., T. Prazuck, L. Hocqueloux, A. Melard, C. Michau, R. Kerdraon, E. Agoute, C. Rouzioux.
2008
. HIV-DNA in rectal cells is well correlated with HIV-DNA in blood in different groups of patients, including long-term non-progressors.
AIDS
22
:
1880
-1882.
35
Blankson, J. N., J. R. Bailey, S. Thayil, H.-C. Yang, K. Lassen, J. Lai, S. K. Gandhi, J. D. Siliciano, T. M. Williams, R. F. Siliciano.
2007
. Isolation and characterization of replication-competent HIV-1 from a subset of elite suppressors.
J. Virol.
81
:
2508
-2518.
36
Lamine, A., A. Caumont-Sarcos, M. L. Chaix, A. Sáez-Cirión, C. Rouzioux, J. F. Delfraissy, G. Pancino, O. Lambotte.
2007
. Replication-competent HIV strains infect HIV controllers despite undetectable viremia (ANRS EP36 study).
AIDS
21
:
1043
-1045.
37
Miura, T., M. A. Brockman, C. J. Brumme, Z. L. Brumme, J. M. Carlson, F. Pereyra, A. Trocha, M. M. Addo, B. L. Block, A. C. Rothchild, et al
2008
. Genetic characterization of human immunodeficiency virus type 1 in elite controllers: lack of gross genetic defects or common amino acid changes.
J. Virol.
82
:
8422
-8430.
38
Ferre, A. L., P. W. Hunt, J. W. Critchfield, D. H. Young, M. M. Morris, J. C. Garcia, R. B. Pollard, H. F. Yee, Jr., J. N. Martin, S. G. Deeks, and B. L. Shacklett. 2008. Mucosal immune responses to HIV-1 in elite controllers: a potential correlate of immune control. Blood Epub ahead of print on Dec. 23, 2008: doi: 10.1182/blood-2008-10-182709.
39
Chase, A. J., H. C. Yang, H. Zhang, J. N. Blankson, R. F. Siliciano.
2008
. Preservation of FoxP3+ regulatory T cells in the peripheral blood of human immunodeficiency virus type 1-infected elite suppressors correlates with low CD4+ T-cell activation.
J. Virol.
82
:
8307
-8315.
40
Bailey, J. R., T. M. Williams, R. F. Siliciano, J. N. Blankson.
2006
. Maintenance of viral suppression in HIV-1-infected HLA-B*57+ elite suppressors despite CTL escape mutations.
J. Exp. Med.
203
:
1357
-1369.
41
Chen, H., A. Piechocka-Trocha, T. Miura, M. A. Brockman, B. D. Julg, B. M. Baker, A. C. Rothchild, B. L. Block, A. Schneidewind, T. Koibuchi, et al
2009
. Differential neutralization of human immunodeficiency virus (HIV) replication in autologous CD4 T cells by HIV-specific cytotoxic T lymphocytes.
J. Virol.
83
:
3138
-3149.
42
Sacha, J. B., C. Chung, E. G. Rakasz, S. P. Spencer, A. K. Jonas, A. T. Bean, W. Lee, B. J. Burwitz, J. J. Stephany, J. T. Loffredo, et al
2007
. Gag-specific CD8+ T lymphocytes recognize infected cells before AIDS-virus integration and viral protein expression.
J. Immunol.
178
:
2746
-2754.
43
Collins, K. L., B. K. Chen, S. A. Kalams, B. D. Walker, D. Baltimore.
1998
. HIV-1 Nef protein protects infected primary cells against killing by cytotoxic T lymphocytes.
Nature
391
:
397
-401.
44
Almeida, J. R., D. A. Price, L. Papagno, Z. A. Arkoub, D. Sauce, E. Bornstein, T. E. Asher, A. Samri, A. Schnuriger, I. Theodorou, et al
2007
. Superior control of HIV-1 replication by CD8+ T cells is reflected by their avidity, polyfunctionality, and clonal turnover.
J. Exp. Med.
204
:
2473
-2485.
45
Migueles, S. A., C. M. Osborne, C. Royce, A. A. Compton, R. P. Joshi, K. A. Weeks, J. E. Rood, A. M. Berkley, J. B. Sacha, N. A. Cogliano-Shutta, et al
2008
. Lytic granule loading of CD8+ T cells is required for HIV-infected cell elimination associated with immune control.
Immunity
29
:
1009
-1021.
46
Bailey, J. R., T. P. Brennan, K. A. O'Connell, R. F. Siliciano, J. N. Blankson.
2009
. Evidence of CD8+ T cell mediated selective pressure on HIV-1 nef in HLA-B*57+ elite suppressors.
J. Virol.
83
:
88
-97.
47
Streeck, H., M. Lichterfeld, G. Alter, A. Meier, N. Teigen, B. Yassine-Diab, H. K. Sidhu, S. Little, A. Kelleher, J. P. Routy, et al
2007
. Recognition of a defined region within p24 Gag by CD8+ T cells during primary human immunodeficiency virus type 1 infection in individuals expressing protective HLA class I alleles.
J. Virol.
81
:
7725
-7731.