Cell-mediated immunity is affected early in human immunodeficiency virus type 1 (HIV-1) infection. HIV-1—specific CD4+ T cell proliferative responses are not measurable in most patients but have been reported in long-term nonprogressors and in patients treated with highly active antiretroviral therapy (HAART) during primary infection. However, treatment with HAART generally does not restore HIV-1—specific CD4+ T cell responses in chronically infected patients. In this study, HIV-1—specific CD4+ T cell responses in 10 HIV-1—infected patients who began HAART with low CD4 cell count nadirs and experienced significant immune reconstitution were studied. Surprisingly, 5 of these patients had proliferative responses to ⩾1 HIV-1 gene product, compared with 0 of 8 chronically infected patients who started HAART when their CD4 cell counts were still relatively high. These results suggest that, in some patients with advanced HIV-1 infection, treatment with HAART can lead not only to significant increases in CD4 cell counts but also to the restoration of HIV-1—specific responses.
CD4+ T cell dysfunction is a prominent feature of infection with human immunodeficiency virus type 1 (HIV-1). Responses to recall antigens and HIV-1 proteins are lost relatively early in the course of the disease . Treatment with highly active antiretroviral therapy (HAART) has been shown to increase CD4+ T cell counts and to improve the proliferative response to recall antigens [2–4]. Proliferative responses to HIV-1 proteins have been detected in long-term nonprogressors (LTNPs) [5, 6] but not in untreated patients with progressive disease. Pioneering studies by Rosenberg et al.  demonstrated that patients treated with HAART during primary HIV-1 infection could develop CD4+ T cell responses to the HIV-1 p24 protein, a result that has been confirmed by other groups [1, 7]. In contrast, there have been conflicting reports as to whether HAART can restore HIV-1—specific responses in chronically infected patients [4, 7–10].
Here, we present the results of a cross-sectional study comparing HIV-1—specific proliferative responses in patients treated with HAART before the development of marked T cell depletion with those of patients with low CD4 cell count nadirs who have experienced significant immune reconstitution after the initiation of HAART.
Patients and Methods
The inclusion criteria for patients receiving HAART were as follows: HIV-1 seropositivity for ⩾2 years before the study; CD4 cell counts >350 cells/µL at the time of study; a stable HAART regimen for ⩾1 year before the study; and virus load <500 copies/mL. Patients in the low CD4 cell count nadir category had CD4 cell count nadirs of <75 cells/smL, whereas patients in the high CD4 cell count nadir cohort had CD4 cell count nadirs >250 cells/µL. LTNPs had been seropositive for >6 years and had not been receiving any antiretroviral therapy for ⩾2 years before the study. They had maintained CD4 cell counts >600 cells/µL and virus loads <5000 copies/mL while not receiving therapy.
Heparinized blood was centrifuged on Ficoll-Hypaque gradients, to obtain peripheral blood mononuclear cells (PBMC). PBMC were resuspended at a concentration of 106 cells/mL in RPMI medium with 10% human AB serum, and 2 × 105 cells in 200 µL were placed in 96-well round-bottomed plates. In some experiments, PBMC were depleted of either CD4+ or CD8+ T cells with antibody-conjugated magnetic beads (Dynal). Recombinant HIV-1 p24 antigen and p66 antigen were obtained from Protein Sciences and were used at concentrations of 2 µg/mL and 1 µg/mL, respectively. The antigens were produced with baculovirus vectors in insect cells, and control protein produced in the baculovirus system was obtained from the same company and was used at concentrations of 2 µg/mL and 1 µg/mL. Recombinant HIV-1 p17 antigen was obtained from Austral Biologicals and was used at a concentration of 2 µg/mL. Candida albicans antigen was obtained from Greer Laboratories and was used at a concentration of 40 µg/mL. Tetanus antigen was the kind gift of Dr. David Schwartz (Dept. of Molecular Microbiology and Immunology, Johns Hopkins University School of Public Health, Baltimore). It was used at a concentration of 2.7 flocculation units/mL
Cells were incubated with the indicated antigens for 6 days and then were pulsed with 1 µCi per well of 3H-thymidine (Amersham) for 12–18 h before being harvested onto fiberglass filters (Packard Instrument Company). After drying, the filters were read on a matrix beta plate reader (Packard Instrument Company) to quantify 3H-thymidine incorporation. The stimulation index (SI) for the p66 and p24 antigens was determined by dividing counts per minute (cpm) obtained from PBMC stimulated with antigens by cpm obtained from PBMC stimulated with control protein. The SIs for Candida, tetanus, and p17 antigens were obtained by dividing the cpm obtained with the antigens by the cpm obtained from cells cultured in medium alone. On the basis of results obtained from uninfected blood donors, an SI >5.0 was considered to be a positive response.
To evaluate the reconstitution of helper T cell responses in patients who began to receive HAART at different stages of disease, we studied proliferative responses to HIV-1 antigens and classic recall antigens in 3 groups of chronically infected adults: patients who started to receive HAART with CD4 cell counts >250 cells/µL (high CD4 cell count nadir), patients who started to receive HAART with CD4 cell counts <75 cells/µL (low CD4 cell count nadir), and LTNPs who maintained high CD4 cell counts without therapy. The clinical data from these patients are summarized in table 1. For the patients with low CD4 cell count nadirs who received HAART, the mean CD4 cell count nadir was 29 cells/µL (range, 8–65 cells/µL), and the current CD4 cell counts were >350 cells/µL (mean, 497 cells/µL; range, 355–926 cells/µL). All patients had received anti-retroviral therapy for ⩾1 year and had had virus loads <500 copies/mL for a mean of 1.62 years.
In the case of the patients with relatively high nadirs who received HAART, the mean CD4 cell count nadir was 402 cells/µL (range, 260–514 cells/µL). Current mean CD4 cell count was 855 cells/µL (range, 607–1087 cells/µL). All patients had been receiving HAART for ⩾ 1 year and had had undetectable virus loads (<500 copies/mL) for a mean of 2.60 years. LTNPs had been infected for >6 years and had maintained CD4 cell counts >600 cells/µL and low virus loads while not receiving therapy. Three of 5 of these patients were treatment naive. The remaining 2 patients had been receiving dual-nucleoside regimens. Both patients had not been receiving therapy for >2 years before study entry and had maintained high CD4 cell counts and low virus loads.
Responses to HIV-1 antigens
Of 8 patients who began to receive HAART with relatively high CD4 cell count nadirs, none demonstrated a proliferative response (SI >5) to the HIV-1 Gag antigen p24 (figure 1A). In contrast, 5 of 10 treated patients with low CD4 cell count nadirs and 4 of 5 LTNPs had detectable p24-specific T cell proliferative responses. Of the patients with low CD4 cell count nadirs, the 5 responders had SIs (range, 5.7–30.0; mean, 18.8) that were similar to those of LTNPs (SI, 9.2–42.5; mean, 23.0).
The patients who responded to the Gag p24 antigen were tested for responses to HIV-1 p17, another Gag subunit. Of the responders, 4 of 4 patients with low CD4 cell count nadirs who received HAART and 2 of 2 LTNPs also responded to the HIV-1 p17 antigen (figure 1B). In contrast, 0 of 4 patients with relatively high CD4 cell count nadirs who received HAART responded to this antigen.
Similar results were seen with responses to HIV-1 pol antigen p66: 2 of 10 treated patients with low CD4 cell count nadirs and 3 of 5 LTNPs had a proliferative response to the p66 antigen (figure 1C). All the patients who responded to the p66 antigen also had p24-specific responses. The mean SI in the 2 patients with low CD4 cell count nadirs was 14.0, which was similar to the mean SI of 15.9 seen in the 3 responding LTNPs. HIV-1—seronegative control subjects and patients with relatively high CD4 cell count nadirs who received HAART had negative p66 proliferative responses.
Responses to non—HIV-1 recall antigens
Although patients with high CD4 cell count nadirs did not develop HIV-1—specific proliferative responses while receiving HAART, they did show responses to classical recall antigens. Five of 8 patients with high nadirs who received HAART responded to the recall antigen tetanus toxoid. The median SI in this group was 11.0. In addition, 3 of 10 patients with low nadirs who received HAART and 6 of 6 LTNPs had proliferative responses, with group median SIs of 3.4 and 15.3, respectively (figure 1D). In the case of Candida, all but one patient responded. The median SI of 35.7 seen in patients with low CD4 cell count nadirs was similar to the median SI of 33.9 seen in patients with high nadirs and the median SI of 27.8 seen in LTNPs (figure 1E). Taken together, these results confirm that patients receiving HAART who start therapy with relatively high CD4 cell counts have T cell responses to classic recall antigens but not to HIV-1 antigens. In light of this, the appearance of HIV-1—specific T cell responses in patients who start to receive HAART while experiencing late-stage disease is even more surprising.
To verify that the proliferative responses observed were mediated by CD4+ T cells, cells from 2 responders from the group of patients with low CD4 cell count nadirs who received HAART were depleted of either CD4+ or CD8+ T cells and were compared with unfractionated PBMC in proliferation assays. In both patients, CD8-depleted, but not CD4-depleted, cells were reactive to p24 and p66 antigens (data not shown), which provides strong support for the conclusion that the proliferative assay measures CD4+ T cell responses.
Although it is clear that treating patients with HAART during primary HIV-1 infection results in the development of p24-specific CD4+ T cell responses [1, 6, 7], studies of the effect of HAART on HIV-1—specific CD4+ T cell responses in chronically infected patients have yielded mixed results. Treatment of chronically infected patients with saquinavir and ritonavir resulted in the development of p24-specific proliferative responses in 10 of 41 patients . In another longitudinal study, 4 of 8 chronically infected patients receiving zidovudine, lamivudine, and ritonavir developed p17-specific responses, whereas only 1 of 8 patients responded to p24 . In AIDS Clinical Trials Group 315, which used the same antiretroviral drugs, there was no significant improvement in proliferative responses to p24, p66, or gp120 after 12 weeks of therapy .
Rinaldo et al.  reported that, in a cohort of chronically infected patients treated with a regimen of zidovudine, lamivudine, and indinavir, there was only a transient improvement in CD4+ T cell reactivity to the HIV-1 p66 protein and no improvement at all to either the p24 or gp120 proteins. In the Spanish EARTH-1 study, 0 of 8 patients with early HIV-1 infection who received stavudine, lamivudine, and ritonavir therapy had proliferative responses to gp120, gp160, or p24 . In a recent study, only 1 of 5 patients who started to receive hydroxyurea, didanosine, and indinavir after seroconversion had p24-specific responses, compared with 7 of 8 patients who received this regimen before seroconversion .
Taken together, these studies suggest that treatment with HAART does not restore HIV-1—specific CD4+ T cell responses in most chronically infected patients. In contrast, multiple studies have shown an improvement in the proliferative response to classical recall antigens in similar patient populations [2–4]. These results suggest that different mechanisms may be responsible for the lack of responsiveness seen to the different antigens in untreated, chronically infected individuals. T cells specific for recall antigens, such as tetanus and cytomegalovirus antigens, may be either trapped in lymph nodes or temporarily rendered anergic because of widespread immune activation in untreated individuals . T cells specific for HIV-1 antigens appear to be physically deleted or permanently inactivated early in the course of the disease, because treatment with HAART usually does not restore HIV-1—specific responses. A recent study showing the presence of cytokine-secreting, Gag-specific CD4+ T cells in untreated patients  would argue against the complete deletion of HIV-1—specific T cells. Specific deletion of T cells with high-affinity receptors or a block in a distal step of T cell activation are possible alternatives.
Our study differs from those outlined above in that it focused on patients with very low CD4 cell count nadirs—patients who would have been excluded from several of the above studies [4, 8, 10]. We show that 5 of 10 patients who had low CD4 cell count nadirs before the initiation of HAART and who subsequently experienced immune reconstitution had proliferative responses to one or more HIV-1 protein. Although we did not study the patients before the initiation of HAART, we think it is unlikely that their proliferative responses preceded the initiation of therapy, given that they had had very low CD4 cell counts and high virus loads at this stage and that a negative correlation between virus loads and p24-specific responses has been demonstrated in untreated patients . In contrast to the low CD4 cell count nadir group, in whom HIV-1—specific T cells were found, 0 of 8 patients who were treated with HAART before the development of marked T cell depletion had HIV-1—reactive CD4+ T cell responses. This finding is remarkable, given that these patients had higher CD4 cell counts than did patients with low CD4 cell count nadirs and responded as well to Candida and tetanus antigens.
The reason for this difference in response is unclear but may be related to the mechanisms involved in immune reconstitution. Treatment with HAART results in an initial redistribution of memory T cells to the periphery, followed later by the generation of naive T cells [2, 12]. Studies looking at T cell receptor excision circles as a marker of recent thymic emigrants have provided evidence of thymic regeneration in patients receiving HAART . It is possible that patients with very low CD4 cell count nadirs are more likely to undergo significant thymic regeneration, peripheral expansion of naive cells, or both. A fraction of the newly generated naive cells will be HIV-1 specific. In patients with higher nadirs, there is less of a stimulus for T cell production or peripheral expansion of naive cells, leaving the patient with a repertoire depleted of HIV-1—specific cells. Indeed, a recent study has shown significant levels of T cell receptor excision circles are found only in patients with low CD4 cell count nadirs . These patients would thus be more likely to have CD4+ T cells with HIV-1—specific receptors. Agents that stimulate thymopoesis or naive T cell expansion may, therefore, be helpful in inducing HIV-1—specific responses in chronically infected patients receiving HAART.
We thank David Schwartz (Johns Hopkins University School of Public Health, Baltimore) for the kind gift of tetanus antigen and for technical advice and Scott Barnett (Johns Hopkins University School of Medicine, Baltimore) for patient recruitment.