Main

Several viruses causing infectious diseases of infancy may be associated with impairment of various immunologic functions of either B or T cell immunity(1). These acquired forms of immunodeficiency are generally transient in nature, even though in a few cases they may result in permanent anergy(1). As a clinical consequence, viral induced immunodeficiencies lead to a great variety of manifestations that differ in the severity of clinical signs and long-term outcome. The more remarkable example of a severe viral induced immunodeficiency is represented by AIDS, associated with HIV-1 infection, that has caused almost 2 million deaths in the past years(2). Measles and varicella may lead to life-threatening complications that still account for over 1 million deaths per year, due to the wide spread of these viruses in the overall population(3, 4). Children who have recently suffered from these diseases may have immunologic abnormalities, including decreased cell-mediated hypersensitivity and low percentages of circulating CD4+ lymphocytes, which in a few cases may persist for a prolonged time(5–7).

Programmed cell death or apoptosis is an active phenomenon involved in either the differentiation processes within the immune system or homeostasis of tissue turnover(8, 9). In the immune system, apoptosis limits clonal expansion and participates in the intrathymic negative selection of the TCR repertoire(10). Hallmarks of this process are changes in cell morphology and DNA degradation in nucleosomes of multiples of 200 bp by the activation of endogenous nucleases(11, 12). Peripheral blood T cells are resistant to apoptosis, even though it has been reported that CD4 cross-linking primes T cells for apoptosis after subsequent TCR stimulation(13–15). In addition, prolonged stimulation may lead to the activation of the apoptotic cell death program, suggesting that mature T cells can be triggered, under certain circumstances, to undergo apoptosis(16).

A dysregulation of apoptosis is now emerging as a possible cause of human diseases(17, 18). In HIV-infected individuals, enhanced spontaneous or activation-induced apoptosis of T lymphocytes is a well established feature(19, 20). In certain neurodegenerative disorders, such as spinal muscular atrophy(21) and Alzheimer's disease, a failure in the regulatory control of apoptosis is believed to play a pathogenetic role(22, 23).

The Fas/Apo-1 (CD95) surface molecule is involved in the process, and its ligation mediates DNA fragmentation(24–26). In CD4+ and CD8+ populations, obtained from cytomegalovirus and VZV seropositive subjects, and cultured in the presence of cytomegalovirus or VZV antigen, the expression of Fas increased significantly(27).

To ascertain whether a dysregulation of apoptosis may also be associated with the partial anergy that follows infection with anergizing viruses, PBMC from children, who had been affected by measles or VZV infection, were examined for spontaneous or activation-induced apoptosis.

METHODS

Subjects. Fifteen previously healthy children (3 girls and 12 boys; mean age, 72 mo; age range, 10 mo to 12 y), who had received a diagnosis of measles or varicella during the previous 6 mo, were enrolled into the study. Diagnosis was based on the typical clinical features and on the presence of specific antibodies. No clinical or laboratory evidence of HIV infection or clinical signs of an underlying immunodeficiency were present. Fifteen age-matched healthy control subjects were also studied. No signs suggestive of measles or varicella were noted in the clinical history of the control subjects, nor were they immunized against measles.

Immunophenotype analysis and proliferation assays. Immunophenotyping was performed by flow cytometry (FACS-can® Becton Dickinson, San Jose, CA) using the following MAb: CD3 (Leu-4), CD4 (Leu-3a), CD8 (Leu-2a), CD19 (Leu-12), CD56 (Leu-19) (Becton Dickinson, San Jose, CA), CD95 (Fas) (Immunotech, Marseilles, France). HLA class II molecules (HLA-DR; Becton Dickinson) were detected on CD3+ cells by two-color immunofluorescence.

The proliferative response was evaluated in triplicate cultures of 2× 105 PBMC, obtained by Ficoll-Hypaque (Biochrom, Berlin, Germany) density gradient centrifugation by standard procedures. Cultures were pulsed with 1 μCi/well of [3H]thymidine (Amersham International, Buckinghamshire, England) for the last 16 h and then harvested. The following stimuli were used: 8 μg/mL PHA (Difco Laboratories, Detroit, MI), 1:200 final dilution Pokeweed mitogen (Life Technologies Ltd, Paisley, Scotland), 8.5 μg/mL concanavalin A (Difco Laboratories), 20 ng/mL PMA + 0.5 mM ionomycin (Sigma Chemical Co., St. Louis, MO). CD3 cross-linking was performed by precoating culture plates with 10 and 1 ng/mL anti-CD3 purified MAb (4B6, gift of Dr. C. Morimoto, Dana Farber Cancer Institute, Boston, MA). In experiments performed with T cell blasts, PBMC were stimulated for 48 h with 8μg/mL PHA.

Cell cycle analysis and DNA fragmentation assays. The percentage of apoptotic PBMC was quantitated by PI (Sigma Chemical Co.) staining and flow cytometry on a FACScan cytometer (Becton Dickinson) as previously described(28). In brief, PBMC (1 × 106) from patients and control subjects were cultured for 72 h in medium alone or in the presence of the appropriate stimulant. CD3 and CD4 cross-linking was performed by precoating tissue culture plates with anti-CD3(4B6; 10 ng/mL) and/or CD4 MAbs (ascites, 19THY5D7, dilution 1:200). After culturing, PBMC were washed twice in phosphate buffer solution and then incubated with PI solution containing 0.1% (vol/vol) Triton X-100 (Sigma Chemical Co.), 3.4 mM trisodium citrate, 50 μg/mL PI. After gentle mixing, cells were put on ice in the dark for 1 h and analyzed by the Cell FIT-DNA software (Becton Dickinson).

DNA fragmentation was determined on PBMC, resting or cultured as above described, by analyzing the ladder appearance characteristic of apoptosis. Low molecular weight DNA was obtained from 4 × 106 cells. At the indicated time, PBMC were washed twice in phosphate buffer solution and the cell pellet was lysed in 0.5 mL hypotonic buffer (5 mM Tris-HCl, 20 mM EDTA, 0.5% Triton X-100, pH 8), and centrifuged at 27 000 × g for 20 min. Fragmented DNA in the supernatant was precipitated overnight at -20°C in 650 μL of isopropanol and 100 μL of 5 M sodium chloride. After centrifugation at 27 000 × g for 20 min, the precipitates were air-dried and resuspended in 10 mM Tris, 1 mM EDTA, pH 7.4. Separation of ethidium bromide-stained DNA fragments was performed by 1% agarose gel electrophoresis. The loading buffer contained 15 mM EDTA, 2% SDS, 50% glycerol, and 0.5% bromphenol blue.

Cell morphology and viability. Cell morphology was evaluated by light microscopy of cells cytocentrifuged and stained by May-Grunwald Giemsa. Loss of cell volume, chromatin condensation along the membrane, and nuclear fragmentation were considered indicative of apoptosis.

Statistical analysis. The statistical significance of differences was evaluated by t test, or by Wilcoxon rank sum test for unpaired data, when indicated. Correlation analysis was performed by regression line and F test.

RESULTS

Immunophenotyping and proliferation assays. Immunophenotype analysis revealed that 7 out of 15 children, who had suffered measles or VZV infection, had lower numbers of circulating CD4+ lymphocytes than age-matched control subjects, ranging between 0.43 and 1.02 × 103 cells/mL (Table 1). Eight patients had a lower CD4/CD8 ratio than the control group, values ranging between 0.37 and 0.85. No difference between measles or varicella patients groups was appreciable. In seven patients the expression of the activation marker HLA class II molecules on CD3+ cells was studied. No difference between patients and control subjects was found (3.1 ± 0.8% versus 4.5 ± 1.5%, respectively). We next evaluated the proliferative response of PBMC after stimulation with suboptimal or optimal concentrations of immobilized anti-CD3 MAb, and PMA and ionomycin by assessing [3H]thymidine uptake in short-term cultures. PBMC from four patients had a lower proliferative response to a suboptimal concentration of anti-CD3. Two of these four patients also had a low number of CD4+ lymphocytes, as reported in Table 1. Table 2 illustrates the results in these four patients. Optimal stimulation through CD3 cross-linking or PMA and ionomycin induced a normal or close to normal response. No appreciable differences between patients and controls were noted when potent mitogens, such as PHA or concanavalin A, were used (data not shown).

Table 1 Lymphocyte subpopulations and CD4:CD8 ratio in the patients' group
Full size table
Table 2 Proliferative response in four low responder patients and controls
Full size table

Cell death in peripheral blood mononuclear cells from children recently infected with measles or VZV. Cell death was investigated in children who had suffered from measles or varicella in the last 6 mo.Figure 1 shows a representative experiment performed to evaluate DNA fragmentation of PBMC. Agarose gel of DNA obtained from patient's PBMC resulted in a clear ladder pattern, indicative of DNA fragmentation in oligonucleosomes containing multiples of 200 bp. Maximal DNA fragmentation was noted after a 72-h incubation with immobilized anti-CD4 MAb (lane 2). In patient's PBMC, DNA fragmentation was also observed after CD3 and CD4 co-stimulation (lane 3) or after an incubation with medium alone (lane 1), but at a lesser extent than that observed after stimulation with anti-CD4 MAb. The ladder appearance was not seen in DNA obtained from freshly isolated PBMC(data not shown) or in DNA from control cells.

Figure 1
figure 1

Representative experiment of six distinct assays showing a DNA fragmentation pattern as observed in PBMC from children affected by measles or VZV within the previous 6 mo. The experiment shown was performed 2 mo after the acute disease. The maximal effect was observed in patients after a 72-h stimulation with anti-CD4 MAb (lane 2). A lower effect was seen after culture in the absence of any stimulus (lane 1) as well as after PBMC stimulation with anti-CD3 and -CD4 MAbs (lane 3). No DNA degradation was found in control PBMC.

Full size image

Cellular DNA content was also measured through cell cycle analysis by PI staining of PBMC. Figure 2 indicates that patients' PBMC underwent apoptosis after a 72-h incubation in medium alone. The percentage of apoptotic cells were higher in measles and VZV patients than in control subjects. Mean values ± SEM were 11.29 ± 3.93% in measles patients (p < 0.01), 8.98 ± 2.00% in varicella patients(p < 0.01), and 2.33 ± 0.41% in control subjects. Again, no apoptotic cells were found in freshly isolated PBMC. No correlation was found between the percentage of apoptotic cells and the time that had elapsed from the acute disease.

Figure 2
figure 2

Percentage of apoptotic cells in PBMC from patients affected by measles, varicella or age-matched control subjects. PBMC were cultured in vitro for 72 h in the absence of any stimulus. DNA content was measured by PI staining and flow cytometry in 15 distinct experiments. Bars indicate SEM. *p, calculated by Wilcoxon rank sum test, <0.001 vs controls.

Full size image

We next investigated whether cell stimulation through MAb against CD4, or CD3 and CD4 molecules in co-stimulation experiments, increased cell apoptosis. A representative experiment (Fig. 3) shows that the maximal apoptotic effect was observed after a 72-h CD4 stimulation. CD3 and CD4 costimulation also increased cell apoptosis, but to a lesser extent than CD4 stimulation alone. In controls, only a slight increase of apoptotic cells after stimulation with anti-CD4 or anti-CD3 and CD4 MAbs was observed, mean values ± SEM being 3.81 ± 0.97% and 3.42 ± 0.57%, respectively. Conversely, in the overall group of patients the percentages of apoptotic cells after stimulation with anti-CD4 or anti-CD3 and CD4 MAbs were 12.97 ± 2.97% (p < 0.01) and 9.72 ± 2.33%(p < 0.02), respectively (Fig. 4). In individual patients these stimulations resulted in increase of the percentage of apoptotic cells compared with that obtained by incubation with medium alone. When the proliferative response to CD3 cross-linking was correlated to the percentage of apoptotic cells, a trend to a negative correlation was found, even though it was not statistically significant due to the limited number of patients in each group (r2 = 0.29; p = 0.14). However, three out of the four patients with a low proliferative response to CD3 cross-linking, identified in the Table 2 as patients 1, 3, 4, and 13, also had high percentages of apoptotic cells, values being 15.4, 13.5, and 11.3%, respectively. Again, no difference between measles and varicella patients was noted. Three patients were tested three times in a 6-mo follow-up for spontaneous or activation induced apoptosis. Results in each patient were homogeneous, indicating that the phenomenon is persistent over time (data not shown).

Figure 3
figure 3

Representative experiment showing activation induced apoptosis in PBMC from patients. Cells were cultured with immobilized anti-CD4 or anti-CD3 and CD4 MAbs, as indicated. DNA content was measured by PI staining and flow cytometry. In patients' PBMC, the maximal effect was noted after a 72-h stimulation with anti-CD4 MAb. No apoptosis was found in freshly isolated patients' PBMC or in control subjects' PBMC. Apoptotic cells(Ao) are indicated.

Full size image
Figure 4
figure 4

Increased percentage of apoptotic cells in patients after PBMC stimulation with anti-CD4 or anti-CD3 and CD4 MAbs, compared with controls. DNA content was measured by PI staining and flow cytometry in 13 distinct experiments. PBMC activation through either stimuli induced a higher increase in the number of apoptotic cells in patients than in controls. No apoptosis was found in freshly isolated PBMC. Bars indicate SEM.*p < 0.02; **p < 0.01.

Full size image

The morphologic analysis of cells revealed the presence of a considerable number of cells showing loss of cell volume, chromatin condensation, and nuclei fragmentation, indicative of cell apoptosis (Fig. 5). Freshly isolated PBMC were not apoptotic, and a 72-h stimulation with anti-CD3 and -CD4 MAbs increased the number of apoptotic cells.

Figure 5
figure 5

Morphology of cytocentrifuged preparations of patient's or control's PBMC. (A) Patient cells immediately isolated from peripheral blood; (B) Patient PBMC cultured for 72 h in the absence of any stimulus; (C) Patient PBMC cultured for 72 h in the presence of anti-CD4 MAb; (D) Control PBMC cultured for 72 h in the presence of anti-CD4 MAb. A number of cells from patient in cultured samples show chromatin condensation, nuclear fragmentation, and change in cell volume, suggestive of apoptosis. Arrows indicate apoptotic cells.

Full size image

Fas expression on CD3+ and CD4+ lymphocytes. Surface Fas molecules mediate cell apoptosis, and their expression is up-regulated during cell activation. To define whether the increased spontaneous and activation-induced apoptosis in the patients' group paralleled an abnormal expression of Fas, CD3+, and CD4+ lymphocytes were assayed for Fas expression. Figure 6 shows that no difference between patients and controls was found in the Fas expression of CD3+ and CD4+ cells, when they were either unstimulated or stimulated for 72 h through anti-CD3 and -CD4 MAbs. Similarly, no difference between patients and control subjects was observed in the intracellular expression of the protective bcl-2 molecule (data not shown). Moreover, incubation of PBMC with a saturating concentration of anti-CD95 MAb did not substantially modify the percentages of apoptotic cells, either when they were unstimulated or after CD3 and CD4 cross-linking (data not shown).

Figure 6
figure 6

Surface expression on CD3+ or CD4+ lymphocytes of Fas antigen. No difference was appreciable between patients and controls.

Full size image

DISCUSSION

Our results indicate that PBMC from children, affected in the previous 6 mo by the common measles or VZV infections, are more prone than control cells to die rapidly by an apoptotic pathway. This phenomenon was observed as long as 6 mo after the acute disease, and it was associated in 40% of cases with a persistent low level of circulating CD4+ cells. Although virusinduced immune suppression is a well recognized condition associated with several viral infections, the pathogenetic mechanisms are unknown(1).

Apoptosis is a physiologic process, that plays a key role in the maintenance of homeostasis in the immune response(29, 30). It is involved in tolerance, in autoreactive cell deletion, and in the deletion of activated cells to limit clonal expansion(12, 31). A dysregulation of the process is now emerging as a possible pathogenetic mechanism for various autoimmune, neurologic, and infectious diseases(17, 18, 32). AIDS is the more remarkable example of a viral induced immunodeficiency caused by accelerated apoptosis of peripheral blood lymphocytes(15). This has been ascribed to CD4 cross-linking mediated by the virus envelope protein gp120, inappropriate function of accessory cells, or superantigen activity induced by the virus(15, 33–35). Recent evidence indicates that infection of human thymus with measles virus induces significant thymocyte death, thus supporting the hypothesis that accelerated apoptosis may be involved in the pathogenesis of viral induced immune suppression(36).

As for the measles and VZV life cycle, both viruses can be isolated from mononuclear cells of infected patients only for a short period of time(37, 38). After initial viremia, measles virus undergoes extensive multiplication in the reticuloendothelial system, followed by virus clearance within a few days(4). A persistent measles viral infection, lasting several years, has been documented in the brain of patients with subacute sclerosing panencephalitis, a very rare neurodegenerative disease that affects 0.6-2.2 per 100 000 naturally infected patients(39, 40). VZV belongs to the family of herpesviruses, and in contrast to measles, more frequently remains in a latent form in infected individuals. Transcripts of the viral genome may be identified in ganglia and neuronal cells years after the first contact with the virus(41). The mechanisms responsible for maintenance of latency are not known. Because measles and VZVs can be isolated from the peripheral blood only for a short period of time, a possible explanation for the observation of accelerated apoptosis is that these viruses may be sequestrated in lymphoid or other reservoir organs for a longer time. Eventually, they may indirectly trigger bystander cells to undergo apoptosis in a similar fashion to that which occurs during HIV infection. Indeed, a direct cytopathic effect of HIV is not per se responsible for triggering cell apoptosis(34).

The clinical relevance of our findings remains to be established. It may explain the defect of cell-mediated immunity observed in these patients, and may predispose to secondary infections. Several mechanisms may underlie a failure to induce a productive T cell immune response. Virtually, structural or functional deficiencies of all the molecules involved in the different phases of cell activation may result in an immune defect, as demonstrated by congenital immune defects(42). Moreover, experimental evidence indicates that the absence of the costimulatory signal, provided by members of the B7 family, results in a functional clonal inactivation(43). This condition, called anergy, is a complex phenomenon requiring a differential association of the TCR with specific intracytoplasmic protein kinases, distinct from those involved in the productive immunity(44). Whether a similar mechanism is also implicated in the defect of cell-mediated immunity after infections with measles or VZVs is not known at present. Even though no T cells bearing HLA class II molecules were found in our patients, an additional explanation of the accelerated apoptosis observed in our patients is also that a few cells are in an activation stage, which results in cell death after a further in vitro stimulation through CD3 and CD4 receptors.

Molecular mechanisms that regulate cell apoptosis are not clearly defined at present, Fas/Fas ligand and bcl-2 molecules participate in the process by exerting opposite effects(45, 46). Previous studies indicate that acute viral infections, such as infectious mononucleosis and varicella, may alter the expression of the protective bcl-2 molecule(47, 48), which may be related to the reemergence of the cell death program. TCR-induced apoptosis is mediated by Fas/Fas ligand interaction, and the surface expression of these molecules is up-regulated after stimulation. Expression of Fas on peripheral blood mononuclear cells cultured in vitro with cytomegalovirus and VZV increased significantly, and the surface expression correlated with DNA fragmentation(27). Taken together these observations argue in favor of a major role of these two molecules in regulating the cell death process after an acute viral infection. In our study we found equivalent expression of the Fas protein in either unstimulated CD3+ and CD4+ cells or cells stimulated through CD3 and/or CD4 cross-linking in controls and patients. Similarly, we found comparable expression of bcl-2 in patients' and control subjects' PBMC. Even though in preliminary experiments we could find an increased Fas-induced cytotoxicity in patients, we cannot exclude an involvement of Fas/Fas ligand molecules in the alteration herein described. However, apoptosis is a complex phenomenon under control of a wide number of genes(23, 49, 50). Cytokines and cytokine receptors participate in the regulatory network of apoptosis. In mice, IL-2 and IL-2 receptor α-chain knockouts result in a mulitsystem autoimmune disease, related to defective elimination of previously activated T cells(51). IL-2 and IL-12 both inhibit apoptosis after CD4 cross-linking with the viral protein gp120(52). Therefore, it remains to be established whether the functional abnormalities herein described are related to a more complex immune disorder due to alterations of the cytokine network.

In conclusion, our data indicate that an accelerated cell death process occurs after measles or VZV infections. Whether this phenomenon is a properly regulated physiologic event, leading to the elimination of lymphocytes activated by the viruses, is not clear. An alternative hypothesis is that the accelerated apoptosis may contribute to the immune deficiency that follows some acute viral diseases in infancy.