Main

Apoptosis or programmed cell death is a physiologically determined event. It is an active process, which, in many cases, requires gene activation, the initiation of protein synthesis, and activation of endogenous endonucleases. Characteristic morphologic changes include nuclear and cytoplasmic condensation(1, 2).

Apoptosis of PBMCs and of CD4+ and CD8+ T lymphocytes after in vitro culture has been shown in HIV-infected adults and children, and is enhanced by activation(3–7). Recent studies have shown a relationship between PBMC apoptosis and CD4+ T cell depletion after overnight or more prolonged in vitro culture(8, 9).

In this cross-sectional study, we measured both CD4+ and CD8+ T cell apoptosis immediately ex vivo and after overnight culture, using a simple assay flow cytometric assay based upon the light scatter characteristics of apoptotic cells. We correlated apoptosis with severity of HIV disease and also with response to antiretroviral medication. Our patients were mainly infants and children, in whom few studies of HIV-induced apoptosis have been performed.

METHODS

Patient selection. All children attending the Colorado Pediatric AIDS Clinical Trials Unit were eligible for study. Patients were recruited between July 1993 and October 1994. They were selected at random, based on availability and when blood samples were being collected for clinically indicated studies. In the early part of the study, HIV-infected parents of patients attending the clinic were also studied. Informed consent was obtained from a parent or guardian and from adolescent patients. Pediatric control subjects consisted of children undergoing elective outpatient surgical procedures, requiring general anesthesia. Informed consent was obtained from parents or guardian. HIV serology was performed on all pediatric controls to confirm seronegativity. Patients and controls were matched for age and time of blood draw. A small group of seronegative children with cystic fibrosis was also studied to evaluate the effects of low grade chronic bacterial infection. The study was approved by the Colorado Multiple Institutional Review Board.

Clinical information was recorded prospectively, and patients were classified for disease severity with the assistance of a physician who was unaware of apoptosis data. Disease classification followed CDC guidelines for pediatric and adult patients(10, 11). Permission was obtained from participating adults for access to their medical records. Adolescents who had become HIV-infected under 12 y of age were classified in the pediatric age group. Adolescents infected later were classified as adults.

The adult and pediatric disease classifications stratify patients according to both disease severity and CD4+ T lymphocyte depletion. For the pediatric group, asymptomatic patients are classified as “N,” those with mild symptoms or signs as “A,” those with moderate symptoms or signs as “B,” and those most severely affected as “C.” CD4+ T lymphocyte depletion is corrected for age and divided into category “1” for no depletion, “2” for intermediate, and “3” for severe depletion.

Reagents. FCS was purchased from Hyclone Laboratories Inc., Logan, UT; penicillin-streptomycin and Dulbecco's PBS from Life Technologies, Inc., Grand Island, NY. RPMI 1640 with glutamine and sodium bicarbonate, Histopaque, acridine orange, and ethidium bromide were purchased from Sigma Chemical Co., St. Louis, MO. Cells were cultured in RPMI 1640 supplemented with 10% FCS and penicillin at 50 U/mL and streptomycin at 50 μg/mL. Goat anti-mouse antibody was obtained from Jackson Immunoresearch, West Grove, PA. Ficoll-Paque was obtained from Pharmacia Biotechnica, Upsala, Sweden. Twenty-four-well plates were obtained from Costar, Cambridge, MA. Hoechst 33342 dye was the generous gift of Richard Duke, Immunology Core Laboratory, University of Colorado Health Science Center. Biotin-16-dUTP, and TdT with buffers were purchased from Boehringer Mannheim Corporation, Indianapolis, IN.

Antibodies. BMA-031 (pan anti-T cell receptor ab MAb) was the generous gift of Roland Kurrle, Behringwerke AG, Marburg, FRG. Leu 3a MAb was the generous gift of the Sloan-Kettering Institute. OKT8 and OKT3 MAbs were derived from cell lines obtained from the ATCC. All were of the IgG1 isotype. Biotinylation and FITC conjugation were performed in the laboratory according to established protocols. Titering of MAbs was done on PBMCs from HIV-seronegative control subjects. Streptavidin-PE and FITC were purchased from Tago Pharmaceuticals, Burlingame, CA. Leu 3a-PE and Leu 2a-PE antibodies were purchased from Becton Dickinson, San Jose, CA. Murine isotype control antibody (IgG2a-FITC; IgG1-PE) was purchased from Olympus, Lake Success, NY.

Sample processing. PBMCs were isolated from heparinized blood samples by Ficoll-Paque or Histopaque density gradient centrifugation. Samples were immediately stained for CD4 and CD8 surface markers. Time zero was defined as the time of separation of PBMCs before overnight culture. Cells from patients and control subjects were stimulated by incubating with BMA-031 (pan anti-T cell receptor MAb) at 100 μg/mL in PBS for 40 min on ice. Unstimulated and stimulated PBMCs from both patients and controls were then cultured overnight in 24-well plates, previously coated with goat anti-mouse antibody, at 37 °C and 5% CO2. PBMCs exposed to 500 rads of γ-irradiation were used as a positive control.

Nonspecific staining was excluded by incubating cells for 10 min in staining solution containing human γ-globulin at 1 mg/mL, before staining for surface markers, and then by staining an aliquot with an isotype control antibody. CD4 and CD8 markers were stained with biotinylated MAbs by incubating at 37 °C for 30 min. Cells were then counterstained with streptavidin-PE or FITC and incubated at 4 °C in the dark for 20 min. In some experiments, cells were stained with directly PE-conjugated CD4 and CD8 MAbs. After washing, samples were fixed in 1% paraformaldehyde in PBS. They were protected from light at room temperature for 30 min and then at 4 °C until flow cytometric analysis, usually within 24 h.

Detection of apoptosis. Morphology. PBMCs, previously fixed in 1% paraformaldehyde, were examined morphologically to confirm that apoptotic cells were present in clinical samples. Cells were stained with the Hoechst 33342 dye (10 μL of cell suspension mixed with 10 μL of dye at 25 μg/mL in PBS) and examined by fluorescent microscopy (Nikon Diaphot-TMD), using an UV filter. Apoptosis was identified by the characteristic increased density of nuclear chromatin(12). The microscopist was unaware of the identity of the specimens. Two hundred cells were counted.

Alternatively, in our sorting experiment (see below), we used acridine orange (3 μg/mL) and ethidium bromide (10 μg/mL) in PBS mixed with an equal volume of cell suspension to differentiate between live and dead cells. Nuclei of “viable” cells took up acridine orange only and had an “open” chromatin pattern. Live apoptotic cells had condensed nuclear chromatin, but excluded ethidium bromide. Cells with an open nuclear chromatin pattern, but with ethidium bromide uptake, were regarded as necrotic. Dead apoptotic cells had condensed nuclear chromatin and took up ethidium bromide(12).

Flow cytometric assay based on changes in scatter. The scatter characteristics of either CD4+ or CD8+ cells were analyzed. Cells with diminished forward scatter (decreased size) and increased side scatter (increased granularity) were considered apoptotic. A FACScan (Becton Dickinson) was used for patient flow cytometry. Ten thousand scatter events were recorded. Standard compensation techniques were used. Events were accumulated at the same flow rate for each patient and control pair to decrease intraassay variability. Data were stored in list mode and analyzed using Lysys software (Becton Dickinson). Percentage of apoptotic cells in controls was subtracted from that in HIV-positive patients to compensate for small interassay variabilities in gating.

For validation of the scatter assay, PBMCs were exposed to 500 rads of γ-irradiation and, after a 48-h incubation at 37 °C and 5% CO2, were stained with an anti-CD4 MAb (Leu 3a-PE) or an isotype control MAb. The isotype control antibody was used to exclude nonspecific uptake of MAb by dying or dead cells. Cells positive for the CD4 surface marker were then sorted into apoptotic and viable populations by scatter characteristics, using a Coulter EPICS 700 cell sorter, and analyzed morphologically as above.

TdT assay for DNA fragmentation. Cells were stained with Leu 3a-PE and isotype control MAbs, fixed in 1% paraformaldehyde, and analyzed by the scatter assay. They were then stored in 75% ethanol for 12-36 h. The TdT assay was performed as described by Gorczyca et al.(13) and adapted by Su et al.(14). The TdT reaction was carried out in the presence of biotin-16-dUTP. FITC-avidin was used to detect biotin-16-dUTP incorporated at DNA ends. Flow cytometric analysis was by FACScan. Ten thousand events were recorded. DNA fragmentation of CD4+ T cells was measured. Correction was made for nonspecific uptake of biotin-16-dUTP and streptavidin-FITC background staining in CD4+ PBMCs.

Statistical analysis. Statistical analyses were by analysis of variance for multiple HIV-infected subgroups and the t test where two groups were compared. The paired t test was used for comparison between patients and controls. The Wilcoxon one-way analysis was also performed for subgroups, but is not reported as results were similar to analysis of variance. Data are presented, primarily from time zero (i.e. immediately after PBMC isolation), as means and SEM.

RESULTS

Patients. Twenty-two children, four adolescents, and nine adults were enrolled in the cross-sectional study. The demographic characteristics of patients and control subjects are shown in Tables 1 and 2. Nineteen patients were vertically infected. Patients tended to be younger than control subjects. On four occasions, seronegative adults were used as controls for seropositive children. In all cases, the patients were over 6 y of age, by which time adult CD4+ T cell counts are seen. One pediatric patient, classified as “N2” was excluded from time zero analysis because of poor staining in both patient and control samples.

Table 1 Demographic characteristics of HIV-infected patients
Full size table
Table 2 Disease classification for HIV-infected patients
Full size table

When adult and pediatric patients were compared, there were no significant differences in either CD4+ or CD8+ T cell apoptosis both at time zero or after overnight incubation (CD4+ T cells p = 0.42 and 0.88; and CD8+ T cells p = 1 and 0.72). Therefore, adult and pediatric data were combined for correlation with lymphocyte subsets and antiretroviral therapy.

The mean time between blood collection and processing for HIV-infected patients was 3.6 h (SEM 0.6) and 3.3 h (SEM 0.6) for seronegative patients (p = 0.28). There was no correlation between CD4+ or CD8+ T cell apoptosis and length of time between blood draw and sample processing for patients and control subjects.

Eight patients had experienced an intercurrent infection within 30 d of being studied. The mean time interval between resolution of the infection and the apoptosis assay was 9.5 d (SEM 1.4). There was no relationship between either CD4+ or CD8+ T cell apoptosis and resolution of intercurrent infection (p = 0.79 and 0.8, respectively). One patient, classified as “N2,” had a lower respiratory infection at the time of study. Six patients in subgroup “C3” had chronic sinusitis, and two also had bronchiectasis. None was overtly symptomatic at the time of study.

Detection of apoptosis. Morphologic assessment confirmed the presence of apoptotic cells in clinical samples, but revealed no significant difference between patients and control subjects at time zero. After overnight culture, however, significantly more apoptosis was seen in HIV-infected patients than in control subjects (Table 3).

Table 3 Mean percentage of peripheral blood mononuclear cells with apoptotic morphology in HIV-infected and seronegative patients
Full size table

To assess both surface phenotype and apoptosis in a mixed population of PBMCs, a simple flow cytometric technique was developed based on the light scatter characteristics of apoptotic compared with live cells. The results of sorting γ-irradiated CD4+ T cells into “apoptotic” and “viable” populations by differences in scatter are shown in Figure 1. Of note, before and after sorting, approximately 75% of apoptotic cells were dead, illustrating the importance of a flow cytometric assay that detects both live and dead apoptotic cells. Only 6% of cells from either the viable or apoptotic regions were necrotic, suggesting that necrotic cells are unlikely to confound the quantitation of apoptotic cells.

Figure 1
figure 1

Scatter characteristics of irradiated and nonirradiated CD4+ T cells. (A) Nonirradiated PBMCs; (B) irradiated PBMCs. Irradiated PBMCs were stained with an anti-CD4 MAb, and positive cells were then sorted by scatter characteristics into R1 (viable) and R2 (apoptotic) subpopulations. Apoptotic cells have decreased forward and increased side scatter. Cells from each region were examined by fluorescent microscopy for apoptotic morphology after staining with acridine orange and ethidium bromide.

Full size image

DNA fragmentation occurs commonly in cells undergoing apoptosis. TdT repairs fragmented DNA. The TdT assay, in which DNA is repaired with labeled nucleotides, has been developed to measure apoptosis(13, 14). It was compared with the scatter assay for four samples in which apoptosis was induced by beauvericin. Although the scatter assay tended to show higher levels of apoptosis, the two methods correlated well (R2 = 0.89; p = 0.06; n = 4).

The two assays were also compared in PBMCs isolated immediately ex vivo from seven patients and seven seronegative controls. Each pair was studied individually. There was a good correlation between the assays for differences in CD4+ T cell apoptosis within each pair (R2 = 0.68; p = 0.021). A similar excellent correlation was found for CD8+ T cell apoptosis (R2 = 0.79; p = 0.006; n = 4).

The scatter-based flow cytometric assay thus provides a simple method for quantifying apoptosis in phenotypically identified subpopulations of PBMCs, and correlates well with quantitation based on apoptotic morphology or on DNA fragmentation.

HIV-seropositive patients have significantly more CD4+ and CD8+ T cell apoptosis than seronegative control subjects. The scatter-based flow cytometric assay was used to measure CD4+ and CD8+ T cell apoptosis in PBMCs analyzed immediately ex vivo and after overnight incubation. An example of CD8+ T cell apoptosis is shown in Figure 2. The two seropositive patients showed a higher percentage of apoptosis than the seronegative control patient, with the more severely affected “C3” patient having more apoptosis than the “N2” patient.

Figure 2
figure 2

Increased percentage of CD8+ T cell apoptosis at time zero and after overnight incubation in HIV-infected children. (A) Time zero; (B) after overnight incubation. PBMCs from a seronegative patient and two seropositive patients were stained with anti-CD8 MAb immediately ex vivo and after an overnight incubation. The γ-irradiated PBMCs (500 rads) were used as a positive control. Gates R1 and R2 were drawn and the FL2 histograms were gated on R1 + R2. R3 was then drawn and its scatter characteristics were analyzed in R1 and R2. N, asymptomatic; C, severely symptomatic; 2, mild CD4+ T cell depletion; 3, severe CD4+ T cell depletion.

Full size image

As a group, HIV-infected patients had a significantly higher percentage of CD4+ and CD8+ T cell apoptosis than seronegative control subjects both at time zero and after overnight culture (Table 4). T cell receptor ligation significantly enhanced CD4+ T cell apoptosis in HIV-seropositive patients (p = 0.002; paired t test; n = 31). HIV-infected children with more severe disease have significantly more CD4+ T cell apoptosis.

Table 4 Percentage CD4+ and CD8+ T cell apoptosis in HIV-seronegative controls and seropositive patients
Full size table

Children with the most advanced disease (subgroup “C”) had a significantly higher percentage of apoptotic CD4+ T cells than those with milder disease (p = 0.015; n = 23). There was a tendency to a stepwise increase in percentage of apoptotic cells with disease severity (Fig. 3). HIV-seronegative cystic fibrosis patients had negligible apoptosis at time zero, suggesting that the low grade persistent bacterial infection occurring in “C3” patients with chronic sinusitis and bronchiectasis does not contribute to the increased percentage of CD4+ and CD8+ T cell apoptosis.

Figure 3
figure 3

Percentages of CD4+ and CD8+ T cell apoptosis at time zero correlate with disease severity in HIV-seropositive children. (A) Percentage CD4+ T cell apoptosis; (B) percentage CD8+ T cell apoptosis. N, asymptomatic; A, mild symptoms; B, moderate symptoms; C, severe symptoms; CF, seronegative children with cystic fibrosis. Apoptosis from seronegative controls was subtracted from that of HIV-seropositive patients. Statistical analysis by analysis of variance.

Full size image

The same correlation with disease severity was seen after overnight incubation in unstimulated CD4+ T cells (p = 0.04; n = 22), and nonparametric analysis confirmed this trend (Wilcoxon: p = 0.1). CD8+ T cell apoptosis at time zero (p = 0.5) and after overnight incubation (p = 0.6) was similar in all subgroups. T cell receptor ligation did not enhance subgroup differences in either the CD4+ or CD8+ T cell populations.

CD4+ and CD8+ T cell apoptosis correlate with CD4+ T cell depletion in adult and pediatric HIV disease. The percentage of CD4+ and CD8+ T cell apoptosis was directly related to CD4+ T cell depletion. Patients with the lowest CD4+ T cell counts (subgroup 3) had the highest percentage of apoptosis (Fig. 4). Similar results were seen after overnight incubation, as well as when adult and pediatric patients were analyzed separately (data not shown).

Figure 4
figure 4

Percentages of CD4+ and CD8+ T cell apoptosis at time zero are directly correlated with CD4 T cell depletion in HIV-infected patients. (A) Percentage CD4+ T cell apoptosis; (B) percentage CD8+ T cell apoptosis. Apoptosis from seronegative controls was subtracted from that of HIV-seropositive patients. Adult and pediatric patients were combined for analysis. Statistical analysis by analysis of variance. 1, CD4+ T cells >25% or ≥1500 (<12 mo); ≥1000 (1-5 y); ≥500 (6-12 y and adults). 2, CD4+ T cells 15-24% or 750-1499 (<12 mo); 500-999 (1-5 y); 200-499 (6-12 y and adults). 3, CD4+ T cells <15% or <750 (<12 mo); <500 (1-5 y); <200 (6-12 y and adults) (CD4+ T cells/mL).

Full size image

Antiretroviral therapy and apoptosis; correlation with treatment response. Fourteen patients were on monotherapy at the time of analysis; seven on zidovudine; six on didanosine, and one on lamivudine. All five patients on combination therapy were receiving zidovudine with two also on didanosine and three on zalcitabine.

There was a trend toward less CD4+ T cell apoptosis with increased time on medication, although this did not reach statistical significance (R2 = 0.09; p = 0.2; n = 19). Patients who responded to antiretroviral therapy with the greatest increase in CD4+ T cell percentage tended to have the lowest percentage of CD4+ and CD8+ T cell apoptosis [R2 = 0.15 (p = 0.1) and 0.34 (p = 0.34), respectively] (Fig. 5). The mean time on an unchanged regimen was 0.92 (SEM 1.4) years.

Figure 5
figure 5

Percentage CD4+ and CD8+ T cell apoptosis at time zero are inversely correlated with response to antiretroviral therapy. (A) Percentage of CD4+ T cell apoptosis; (B) percentage of CD8+ T cell apoptosis. Mean duration of unchanged antiretroviral therapy was 0.92 (SEM 1.4) y. Control apoptosis was subtracted from patient apoptosis.

Full size image

Patients on antiretroviral therapy tended to have more CD4+ T cell apoptosis than those not on therapy, possibly reflecting the increased disease severity in those requiring therapy [7.4% (1.7) versus 2.7% (2.1); p = 0.09]. CD8+ T cell apoptosis was similar in the two groups.

The mean increase in CD4+ T cell apoptosis above seronegative control values for 11 subjects not on antiretroviral therapy was 2.3% (SEM 1.8), 7.0% (SEM 1.6) for 14 patients on one drug, and 3.9% (SEM 2.7) for five patients on two drugs (p = 0.17). The latter finding suggests that two antiretroviral agents may be more effective than one.

DISCUSSION

Apoptosis in HIV infection. The present study makes a number of new observations relevant to HIV disease. It shows, for the first time, that the rate or extent of both CD4+ and CD8+ T cell apoptosis is correlated with CD4+ T cell depletion in HIV-infected children. This finding is in agreement with data previously reported for HIV-infected adults(8, 9). Bäumler et al.(7) have recently reported increased spontaneous CD4+ and CD8+ T cell apoptosis in children. Our work also extends prior studies in demonstrating increased percentages of CD4+ and CD8+ T cell apoptosis in PBMCs analyzed immediately ex vivo without the need for overnight incubation or stimulation. This finding provides strong evidence that apoptosis occurs in HIV-infected children and adults in vivo.

A number of questions arise from these observations. The first question is how HIV infection leads to increased apoptosis of CD4+ and CD8+ T cells. Although this study does not address mechanism, it is clear from our prior work and that of others that a number of mechanisms are possible and could induce death of both infected and uninfected T cells(15). The second question, addressed in part by these studies, is whether apoptosis plays a role in the progression of asymptomatic HIV infection to AIDS.

The central question is -why there is an increased percentage of CD4+ and CD8+ T cell apoptosis with advanced HIV disease. CD4+ T cell depletion is a poor prognostic sign. Numerous paediatric studies have correlated low CD4+ T cell counts with more severe manifestations of HIV disease(16). CD4+ T cell depletion is also associated with a high viral load, which is itself a recently characterized poor prognostic marker. Whether apoptosis correlates directly with viral load or is a reflection of antiviral immunologic activity remains to be determined. A recent adult study suggests that the CD4+ T cell count and viral load are independent predictors of severe clinical disease(17). It is, therefore, relevant to examine the relationship of both with apoptosis.

Apoptosis as a marker of disease activity. The results of this study show a correlation between CD4+ and CD8+ T cell apoptosis and both disease progression and CD4+ T cell depletion, suggesting that apoptosis plays a role in the pathogenesis of HIV disease. Prior studies in adult patients on the relationship between apoptosis and CD4+ T cell depletion have yielded conflicting results. Meyaard et al.(6) found no correlation between apoptosis and either CD4+ and CD8+ T cell counts or a change from non-syncytium-inducing to syncytium-inducing viral phenotype. However, this study relied on the use of cryopreserved cells from a limited number of adult patients. More recently, Carbonari et al.(8) found significantly more apoptosis in cultured PBMCs from four patients with CD4+ T cell counts less than 400/μL than in nine with higher counts. The relationship between PBMC death and CD4+ T cell depletion has been confirmed by Pandolfi et al.(9), who showed a correlation after prolonged in vitro culture. Finally, both Carbonari et al.(8) and Pandolfi et al.(9) found increased PBMC apoptosis in patients who subsequently showed clinical progression, suggesting that apoptosis may be a useful clinical marker for patients at risk of progression.

We have also correlated CD8+ T cell apoptosis with CD4+ T cell depletion. We speculate that the significance of this finding could be that it reflects increased CD8+ immunologic activity in response to progression of disease.

Apoptosis as a marker of therapeutic activity. Although our study of antiretroviral therapy was done on only a small sample of patients, these data showed a trend toward less CD4+ and CD8+ T cell apoptosis in those children in whom antiretroviral therapy resulted in an elevation of CD4+ T cell percentage. These data are in agreement with a recent study of prednisolone therapy given to asymptomatic HIV-positive adults. Andrieu et al.(18) showed a sustained increase in CD4+ T cell counts accompanied by diminished PBMC. Although this study was also done on only a small number of patients, these data data suggest that CD4+ T cell apoptosis may provide another marker of therapeutic efficacy, and argue for inclusion of apoptosis in the laboratory analyses of adults and children undergoing trials of combination antiretroviral therapy.

Detection of apoptosis by the scatter assay. We detected apoptotic cells in a newly characterized flow cytometric assay by diminished forward and increased side scatter. This assay has the advantages of low cost, simplicity, and safety. In addition, because scatter data are generally collected and saved in all flow cytometric protocols, analyses can be done retrospectively.

Diminished forward scatter and increased side scatter reflecting decreased cell size and increased granularity was one of the earliest flow cytometric characteristics of apoptosis recognized(19–22). Wesselborg and Kabelitz(23) have shown that changes in scatter precede DNA fragmentation and might therefore be a sensitive marker of apoptosis. A similar observation was made by Carbonari et al.(8) who detected apoptosis by a combination of decreased forward and increased side scatter and reduced CD45 expression. After irradiating cells to induce apoptosis, many cells with decreased forward scatter still expressed a high level of CD45, suggesting that scatter changes might precede loss of the CD45 marker in apoptotic PBMCs.

The fact that changes in the light scatter properties of a cell early in the apoptotic process may explain why we were able to detect increased apoptosis in ex vivo PBMCs from HIV-infected patients. In contrast, both Meyaard et al.(6) and Carbonari et al.(8), assaying for apoptosis either by DNA fragmentation or diminished CD45 expression, showed no significant increase in apoptosis without overnight culture. In addition, our own work analyzing morphology by fluorecence of DNA binding dyes, was unable to detect increased apoptosis in PBMCs analyzed immediately ex vivo.

Many other techniques for the flow cytometric detection of apoptosis have been described(24–26). The advantages of the scatter assay include low cost, simplicity, and safety, because these cells are fixed in paraformaldehyde. The scatter assay does not involve the purchase of expensive reagents nor additional laboratory time. Our sorting experiment demonstrated that, for irradiated cells, as many as half the apoptotic cells expressing surface marker may be dead. Dead cells, therefore, may contribute significantly to measurable apoptosis and might be excluded in assays specific for live apoptotic cells. By measuring the scatter characteristics of only those cells stained for a specific surface marker, one automatically excludes debris and can quantify apoptosis in defined cell populations.

CONCLUSION

We have, for the first time, demonstrated a correlation between both CD4+ T cell apoptosis and severity of disease in HIV-infected children. We have also found significant correlations between CD4+ and CD8+ T cell apoptosis and CD4+ T cell depletion. Our demonstration of this phenomenon directly after PBMC isolation suggests that it occurs in vivo and is a reflection of disease activity.

The trend toward less CD4+ and CD8+ T cell apoptosis in patients with the best response to antiretroviral therapy, as measured by increased CD4+ T cell percentage, argues that the quantitation of apoptosis may have a role in the monitoring of therapeutic efficacy.