The present study investigated immune restoration in patients at intermediate stages of human immunodeficiency virus (HIV) disease after initiation of highly active antiretroviral therapy (HAART). A progressive increase in both memory and naive CD4+ T cells was observed from the first weeks of therapy, concomitant with a decrease in the expression of activation markers on CD8+ T cells. The early-activation marker CD69 remained, however, overexpressed on T cells after suboptimal stimulation in vitro, indicative of persistent immune activation. The percentage of interleukin (IL)-2-producing CD4+ T cells significantly increased from 9 months of HAART. In most patients, CD4+ T cells recovered an ability to produce IL-2 on stimulation, similar to that of HIV-seronegative controls. Reversal of T-cell anergy may be a key event in immune restoration for achieving long-term clinical benefit with HAART.
Infection with human immunodeficiency virus type 1 (HIV-1) is associated with functional alterations of CD4+ T lymphocytes and a progressive loss of CD4+ T cells [1–3]. Early impairment of T-cell function during the clinical-latency phase of HIV disease is characterized by defective T-cell proliferation and decreased interleukin (IL)-2 production in response to recall antigens, HIV proteins, and stimulation through the T-cell receptor (TCR)/CD3 complex [3–5]. Highly active antiretroviral therapy (HAART) results in a dramatic and prolonged reduction in plasma virus load and a progressive increase in the number of circulating CD4+ T lymphocytes in most treated patients [6–8]. Whereas the early increase in peripheral CD4+ T-cell numbers may be related to redistribution of sequestered lymphocytes, repopulation with naive CD4+ cells is also observed in treated patients with advanced HIV disease, as peripheral destruction of the cells is decreased and T-cell renewal occurs [9–11]. Suppression of virus replication is also associated with a decrease, within the first months of therapy, in the percentage of circulating CD4+ and CD8+ T cells expressing activation markers [12–15].
The decline in morbidity and mortality that is observed in patients with advanced HIV disease undergoing HAART strongly suggests that the increase in circulating CD4 T-cell counts is associated with enhanced T-cell function and specific immune responses [16, 17]. Thus, a recent report indicates that treated patients may recover T-cell reactivity against cytomegalovirus and Mycobacterium tuberculosisantigens . The aim of the present study was to prospectively assess immune restoration after initiation of HAART in patients at intermediate stages of HIV disease. In addition to analyzing changes in subsets of naive and memory CD4+ T cells and in the activation status of circulating T lymphocytes, we investigated functional T-cell recovery by assessing cytokine-producing CD4+ and CD8+ T lymphocytes at a single-cell level.
Patients and Methods
The study was designed to prospectively monitor immune functions in patients at intermediate stages of HIV disease after initiation of triple-combination antiretroviral therapy. Fifteen patients were included from February 1997 to September 1997, on the basis of HIV-RNA plasma levels >4.3 log10 copies/mL and CD4 cell counts >150 cells/μL. All patients had previously received nucleoside analogues, with the exception of stavudine, and were naive of protease inhibitors. After prescreening, the patients began a triple-drug combination including stavudine, lamivudine, and indinavir. The study group comprised 15 men with a mean age of 40 years (range, 29–49 years). The mean CD4+ T-cell count was 288 ± 106 cells/μL (mean ± SD) at baseline. The mean baseline HIV plasma virus load was 4.7 ± 0.4 log10 copies/mL (range, 4.3–5.5 log10 copies/mL), as assessed by means of the bDNA assay with a sensitivity threshold of 2.7 log10 copies/mL (Chiron, Emeryville, CA).
All patients completed at least 12 months of follow-up. Blood samples were obtained twice prior to initiation of HAART, at days 15 and 0, and at day 15 and months 1, 2, 4, 6, 9, and 12 of therapy. Antiretroviral treatment remained unchanged throughout the period of follow-up in all patients. Blood samples were also collected from 12 healthy HIV-seronegative donors. Informed consent was obtained from patients and seronegative controls.
Phenotypic analysis of freshly obtained peripheral blood lymphocytes
Surface markers of peripheral blood lymphocytes were analyzed by flow cytometry using 3-color direct immunofiuorescence and monoclonal antibodies (mAbs) conjugated with either fluorescein isothiocyanate (FITC), phycoerythrin (PE), or peridinal chlorophyll protein (PerCP). The mAbs anti-CD3 PerCP, anti-CD4 FITC, anti-CD4 PE, anti-CD8 PE, anti-CD8 PerCP, anti-CD28 PE, anti-CD69 PE, anti-CD25 PE, anti-CD45RO PE, anti-CD45RA PE, anti-HLA-DR PE, and anti-CD38 PE were obtained from Becton Dickinson (Mountainview, CA). Anti-CD62L FITC mAb was obtained from Immunotech (Marseille, France). Stained cells were analyzed by using a FACScan flow cytometer (Becton Dickinson) and the Cell Quest software (Becton Dickinson). Fluorescence parameters were collected by using 4 decades of logarithmic amplification. At least 2000 events were analyzed in each experiment.
Activation of T lymphocytes after stimulation with soluble anti-CD3 mAb
The expression of the early activation marker CD69 is optimally induced after cross-linking of the T-cell receptor with immobilized anti-CD3 mAb or PHA stimulation. To use a more sensitive assay to assess the activation status of T cells in vivo, we analyzed the expression of CD69 after suboptimally stimulating the cells with soluble anti-CD3 mAb . Fresh peripheral blood mononuclear cells (PBMC) were obtained by centrifugation of EDTA-anticoagulated venous blood over MSL (milieu pour séparation des lymphocytes; Eurobio, Les Ulis, France). PBMC (2 × 106/mL) were then cultured at 37ºC in 24-well plates in RPMI 1640 containing 10% heat-inactivated fetal calf serum and antibiotics. The cells were stimulated with either 5 μg/mL of anti-CD3 mAb UCHT-1 (Immunotech) in solution or with 5 μg/mL of PHA (Wellcome, Dartford, UK). After 18 h of culture, the cells were stained with a combination of anti-CD4 FITC, anti-CD69 PE, and anti-CD8 PerCP mAbs. The expression of the CD69 antigen was then determined separately in CD4+ and CD8+ T-cell subsets.
Single-cell analysis of T-cell cytokine production
For single-cell analysis of cytokine production by T cells, freshly obtained PBMC (2 × 106/mL) were stimulated with a combination of 50 ng/mL phorbol myristate acetate (PMA) and 500 ng/mL ionomycin (Sigma Chemical Co., St Louis) in the presence of brefeldin (10 μg/mL; Sigma) for 22 h at 37ºC. The cells were saturated with 5% normal human AB serum prior to incubation with anti-CD8 FITC and anti-CD3-PerCP mAbs, fixation with PBS containing 4% formaldehyde (Sigma), and immunofluorescence staining for intracellular cytokines. For staining, the cells were washed with PBS-BSA 0.2% and subsequently incubated with PE-conjugated mAbs against IL-2, interferon (IFN)-γ, IL-4, or isotypic controls in the presence of PBS containing 0.5% saponin and 0.2% BSA (Sigma) for 30 min at 4ºC. Anticytokine mAbs and isotypic control mAbs were from Becton-Dickinson. The cells were washed twice in saponin buffer and once with PBS-NaN3, resuspended in PBS, and immediately processed by cytofluorometry. CD3+CD8− and CD3+CD8+ cells were analyzed separately for cytokine expression. Analysis of CD3+CD8− cells was preferred to that of CD3+CD4+ cells because the CD4 antigen is down-modulated after CD4 T-cell activation. No cytokine was detected in unstimulated cells.
Quantitative variables were expressed as mean ± SD. Comparisons between HIV-seropositive individuals and seronegative controls were performed by using the unpaired Student's t test. Data obtained at different time points of follow-up were compared by means of the nonparametric Wilcoxon rank test. Statistical significance was considered for P values <.05.
Plasma levels of HIV-RNA decreased below the limit of detection of 2.7 log copies/mL in all 15 patients in the study within 2–8 weeks of HAART. Virus suppression was persistent throughout the period of follow-up in 13 of the 15 patients. In 2 patients, HIV-RNA was detected in plasma at levels below 4 log copies/mL, from month 9 of therapy.
Longitudinal changes in T-cell phenotype
The absolute number of circulating CD4+ T cells increased from 288 ± 106 × 106 cells/μL (mean ± SD) at baseline to 526 ± 221 cells/μL after 12 months of HAART (figure 1). The increase in CD4 cells was significant from day 15 of therapy (P = .010). It was strictly dependent on an increase in the absolute number of lymphocytes for the first 6 months of HAART, after which a significant increase in the percentage of CD4+ T lymphocytes was also observed. Thus, the percentage of CD4+ T cells was 16.4 ± 6.3 at baseline, 16.3 ± 6.1 at 2 months (P > .05), 18.1 ± 6.0 at 4 months (P> .05), 19.0 ± 6.9 at 6 months (P = .014), 20.2 ± 7.6 (P = .010) at 9 months, and 20.8 ± 6.8 (P = .002) at 12 months of HAART. Expansion of both naive CD45RA+/CD62L+ and memory CD45RO+/CD4+ T cells occurred continuously from day 15 of HAART (figure 1). The proportion of cells exhibiting a naive phenotype among CD4+ T cells increased from 36.7% ± 14.5% at baseline to 39.3% ± 13.6% after 12 months of HAART, although the difference did not reach statistical significance (P> .05). No increase in naive CD4+ T cells was observed in patients with a low representation of the CD4+/CD45RA+ population prior to therapy. Thus, the mean number of CD4+/CD45RA+ cells at baseline in 5 patients in whom naive cells did not increase after 2 months of HAART (54 ± 20 cells/μL) was significantly lower than that (137 ± 64 cells/μL) in 10 patients in whom CD4+/CD45RA+ cells increased during that time (mean increase: 67 ± 44 cells/μL; P = .009).
Absolute counts of CD8+ T lymphocytes increased from the first weeks of therapy to reach a maximum after 2 months of HAART (data not shown). The relative percentage of CD8+ T lymphocytes decreased from 63.5% ± 10.1% at baseline to 53.5% ± 11.4% at month 12 of HAART (P = .0007). The mean CD4/CD8 ratio increased significantly from 0.28 ± 0.15 at baseline to 0.37 ± 0.22 at month 4 (P = .014) and 0.43 ± 0.28 after 12 months of HAART (P = .0012).
As previously reported by others [12–14], a rapid decrease in the percentage of CD8+ cells expressing the CD38 activation marker occurred from day 15 of HAART. The decrease in expression of HLA-DR on CD8+ T cells occurred later, from month 4 (figure 2). Noticeably, however, the percentage of CD8+ T cells expressing activation markers remained high in treated patients after 12 months of HAART in comparison with HIV-seronegative controls (52.5% ± 16.6% vs. 32.0% ± 11.0%; P = .0009). The proportion of CD4+ T cells expressing HLA-DR decreased steadily to reach significance from month 4 (19.2 ± 6.3 at baseline; 16.1 ± 5.6 at month 4, P = .014). The proportion of CD4+ T cells expressing CD25 did not significantly decrease within 12 months of HAART. We further investigated the expression of the early activation marker CD69 on T cells, by comparing suboptimal stimulation with soluble anti-CD3 mAb and optimal stimulation with PHA in vitro. Prior to HAART, CD4+ and CD8+ T cells both overexpressed CD69 after stimulation with soluble anti-CD3. The calculated ratio between the mean fluorescence intensity of CD69 expression on CD4+ T cells stimulated with anti-CD3 and that of cells stimulated with PHA was 0.50 ± 0.20 in patients and 0.34 ± 0.11 in seronegative controls (P = .03). This ratio was 0.69 ± 0.31 and 0.34 ± 0.09 (P = .003) for CD8+ T cells. No significant reduction in anti-CD3-induced overexpression of CD69 was seen in treated patients after 12 months of HAART (data not shown).
We also examined the expression of the accessory molecule CD28 on CD4+ and CD8+ T cells. More than 90% of CD4+ T cells of patients expressed CD28, except for 1 patient. In the latter individual, 75% of CD4+ T cells expressed CD28 at baseline, a percentage that did not subsequently improve despite sustained suppression of virus replication. The percentage of CD8+ T cells lacking CD28 expression decreased steadily from 65.2 ± 13.9 prior to HAART to 55.0 ± 18.6 after 12 months of therapy (P = .003).
Longitudinal changes in cytokine production by T cells
The function of T cells of patients undergoing HAART was investigated by determining the proportion of CD4+ and CD8+ T cells producing IL-2, IFN-7, and IL-4 after stimulation of freshly obtained PBMC with PMA and ionomycin, as described in Methods. Prior to initiation of HAART, IL-2-producing CD4+ T cells were significantly decreased in patients in comparison with HIV-seronegative controls (table 1). There was no relationship between plasma levels of HIV-RNA at baseline and the reduced ability of CD4+ T cells to produce IL-2 (data not shown). The proportion of IL-2-producing CD8+ T cells was also reduced in the patients' group. The proportion of T cells producing IFN-γ did not differ between patients and controls (table 1). Figure 3 depicts the longitudinal changes in the mean percentage of IL-2-producing CD4+ T cells in patients undergoing HAART. Within the first 6 months of therapy, the proportion of CD4+ T cells producing IL-2 in response to PMA and ionomycin remained low despite sustained virus suppression and increasing CD4+ T-cell counts. From 9 months of HAART, however, we observed a significant increase in the percentage of IL-2-producing CD4+ T cells (figures 3 and 4, table 1). Thus, the percentage of IL-2-producing CD4+ T cells was 56 ± 12 at month 9 and 57 ± 9 at month 12 of HAART as compared with 43 ± 14 at baseline (P = .030 and .004, respectively). Figure 5 shows the individual profiles of IL-2 production at various time points in the study. Although the percentage of IL-2-producing CD4+ T cells increased in most patients between baseline and month 9, it decreased in a few patients at month 9 to return to baseline values at month 12. Thirteen of 15 patients recovered a production of IL-2 by CD4+ T cells, similar to that of HIV-seronegative controls, after 12 months of HAART. At 12 months, the mean percentage of IL-2-producing CD4+ T cells in the 15 patients of the study did not differ from that of healthy controls (P = .9 according to unpaired t test). In 2 patients, the percentage of IL-2-producing CD4+ T cells remained below the range observed in healthy controls, 1 of whom had escaped antiretroviral treatment. When analyzed on an individual basis, the best reconstitution of IL-2 production was not associated with the best recovery of naive or memory CD4+ T cells.
The mean proportion of IL-2-producing CD8+ T cells increased between baseline and month 12 of HAART in the 15 patients of the study, although the difference from pretreatment values did not reach statistical significance (table 1, figure 4). We observed no difference in the proportion of CD4+ or CD8+ T cells producing IFN-γ during follow-up (table 1). The proportion of IL-4-producing CD4+ T cells remained <5% in all patients prior to and during HAART (data not shown).
The present study was aimed at prospectively assessing immune restoration in patients at intermediate stages of HIV disease undergoing HAART. We observed a progressive increase in both memory and naive CD4+ T cells and a decrease in the expression of activation markers on CD8+ T cells from the first weeks of therapy. Overexpression of the early activation marker CD69 on T cells after suboptimal stimulation with soluble anti-CD3 mAb persisted throughout the 12 months of follow-up, showing a continuous state of immune activation under HAART despite prolonged virus suppression. A remarkable result was the observation of a delayed restoration to normal values of the ability of stimulated CD4+ T cells to produce IL-2.
To study immune restoration under HAART, we selected a group of asymptomatic patients with a mean number of circulating CD4+ T cells of 300 cells/μL. Within the first 2 weeks of HAART, the total lymphocyte counts increased, resulting in an increase in the absolute number of both CD4+ and CD8+ T cells. Cells of both naive and memory phenotypes contributed to the initial rise in CD4+ T lymphocytes. The latter results are in contrast to the delayed appearance of naive T cells reported in patients with more advanced HIV disease receiving HAART . The initial increase in memory CD4+ T cells under HAART results from the redistribution of cells previously sequestered in lymphoid and inflamed tissues . In addition to retrafficking, the expansion of peripheral T-cell populations may result from peripheral expansion, thymic input, and/or from the occurrence of revertant memory T cells [10, 20, 21]. In agreement with the results of Connors et al., we found that the increase in CD4+/CD45RA+/CD62L+ naive T cells was minimal in patients with a low representation of that population prior to initiation of therapy . The data support the hypothesis of an expansion of preexisting cells. Recent results indicate that thymic tissue and hematopoietic microenvironment may remain functional in some patients with HIV disease [23, 24]. The contribution of thymic regeneration to immune reconstitution in patients undergoing HAART is currently under investigation .
The percentage of CD4+ T cells among lymphocytes increased significantly from month 6 and that of CD8+ T cells decreased from month 4, resulting in an increase in the CD4/CD8 ratio, although the latter remained far below normal values after 12 months of follow-up. To our knowledge, only IL-2 therapy may restore a CD4/CD8 ratio above 1.0 in patients receiving antiretroviral therapy . As previously observed by others [12–15], a decrease in the expression of the CD38 activation marker on CD8+ cells occurred as the HIV virus load decreased. The decrease in the percentage of CD4+ T cells expressing HLA-DR was moderate, and no significant decrease in the proportion of CD4+ T cells expressing CD25 was observed after 12 months of HAART The persistence of a chronic immune activation in successfully treated patients was further shown by the continuous overexpression of CD69 on T cells stimulated with soluble anti-CD3 mAb. CD69 is rapidly induced on T cells after triggering through the CD3/TCR complex . Whereas the induction of CD69 is strictly correlated with the extent of CD3/TCR cross-linking, the use of soluble anti-CD3 mAb appears to be a sensitive means to assess the activation status of T cells . The persistence of immune activation, despite the absence of detectable plasma HIV-RNA with the currently available assays, may relate to low degree of continuous virus replication .
CD8+ T cells lacking expression of the CD28 costimulatory molecule have been suggested to represent a population of activated differentiated cytotoxic effector cells driven by virus antigens [28, 29]. CD8 lymphocytosis, which is commonly seen in HIV-seropositive individuals, consists of CD28− cells expressing activation markers such as HLA-DR, CD38, and CD45RO [29, 30]. We observed a decrease in CD8+CD28− T cells during HAART. CD28 expression was low on CD4+ T cells in only 1 of 15 patients at baseline, consistent with previous work from our laboratory in which we observed a reduced expression of CD28 in no more than 25% of unselected HIV-seropositive individuals (unpublished data). No change in CD28 expression on CD4+ T cells was seen during follow-up.
A specific aim of the present study was to assess longitudinal changes in Thl-type cell function under HAART. Consistent with findings by others, the proportion of CD4+ T cells producing IL-2 was significantly reduced in patients in comparison with healthy controls, when measured at baseline [3, 31]. We found no relationship between plasma levels of HIV-RNA and the defective ability of CD4+ T cells to produce IL-2 after stimulation. The reduction in IL-2 production by CD8+ T cells was close to significance when patients' values at baseline were compared with those of HIV-seronegative controls. We found no significant difference in the proportion of T cells producing IFN-γ between patients and controls. Several authors reported a decrease in the ability of T cells to produce IFN-γ in HIV disease [32, 33]. Fan and colleagues reported an increase in IFN-γ mRNA expression in CD4+ but not in CD8+ T lymphocytes of HIV-infected individuals . When analyzed at a single-cell level, increased IFN-γ production was observed in CD8+ T cells .
In the present study, we have demonstrated, for the first time, a significant improvement in IL-2 production by CD4+ T cells in patients undergoing HAART. The increase in the proportion of IL-2-producing CD4+ cells occurred only following 9 months of sustained virus suppression with therapy. The relative increase in IL-2 production differed between patients when analyzed on an individual basis, independently of the recovery of naive CD4 T cells. After 12 months of HAART, the percentage of IL-2-producing CD4+ cells did not differ between patients and healthy controls. Thirteen of 15 patients in the study recovered a normal production of IL-2 by CD4+ T cells after 12 months of HAART. In 2 patients, the percentage of IL-2-producing CD4+ T cells remained below the range observed in healthy controls; 1 of these 2 patients had escaped antiretroviral treatment.
Only a few reports on the functional recovery of CD4+ T cells in patients receiving HAART are available at present [12, 13, 15, 18, 35]. Whether normalization of the CD4 TCR repertoire occurs in patients successfully treated with HAART remains to be established inasmuch as conflicting results have been published in this regard [22, 36]. T-cell reactivity in response to mitogens and recall antigens was found to improve within 1–6 months of HAART [13, 15]. In severely immuno-suppressed patients, the recovery of CD4 T-cell reactivity to the tuberculin and cytomegalovirus recall antigens was related to sustained profound reductions in virus load over time and to increased numbers of memory and naive CD4+ T cells .
The ultimate goal of therapy in HIV disease is to achieve virus suppression and a functional restoration of the immune system. Our observations that HAART may result in a significant increase in both memory and naive CD4+ T cells and in reversal of T-cell anergy in patients at intermediate stages of HIV disease emphasize that immunodeficiency associated with AIDS is reversible, and may be indicative of a long-term clinical benefit in patients with sustained virus suppression.
We thank Catherine Demouchy for excellent technical assistance and Joelle Nataff, Isabelle Carriere, and Philippe Mariot for help in performing the study. We are grateful to Dr. L. Belec for performing plasma virus load assays. We gratefully thank the patients for providing blood samples.