INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The presence of human immunodeficiency virus (HIV) in the lung may cause an interstitial lung disease (ILD) characterized by a marked infiltration of T cells in the pulmonary interstitium and alveolar spaces (1). The T-cell alveolitis, although present at all stages of HIV infection, is most pronounced in patients with early-middle stage disease (2) and is mainly secondary to the compartmentalization of CD8+ cytotoxic T lymphocytes (CTL) with anti-HIV activity (3). Lung CTL are central cells in controlling HIV infection because they are able to recognize HIV structural proteins on the surface of infected cells. The effectiveness of lung CTL in controlling HIV infection varies during different phases of HIV disease. In particular, a progressive decline of pulmonary CTL activity may be observed with the progression of the disease (2). So far, the cellular basis of this phenomenon is not fully understood. We demonstrated an unexpected in vivo infectivity of HIV toward pulmonary CD8+ T cells (4). On the basis of these data, the hypothesis has been formulated that the infection of CD8 T cells might be implicated in the local CTL dysfunction, and ultimately in the progression of the disease.

The pattern of cytokines released by the alveolar macrophages (AMs) is pivotal in determining the effects of the pulmonary immune response to HIV. In particular, compelling evidence suggests that interleukin-15 (IL-15) is a central regulator of the intrapulmonary proliferation of CD8+ T cells (5). This cytokine, which shares many biologic activities with IL-2 including the ability to costimulate CTL proliferation (6), upregulates the expression of coligands that favor the contact between T cells and antigen-presenting cells (APC) in the lung (7). This in turn triggers the in situ activation and proliferation of T cells, favoring the local accumulation of CD8+ T cells. Besides inducing T-cell proliferation, IL-15 shows other functional effects on T cells. In particular, IL-15 represents an important cofactor in the survival of preactivated T cells (8).

Previous studies have demonstrated that increased apoptosis of T-cell subsets can be detected in the peripheral blood from HIV-infected patients. In particular, it is known that the CD45R0+/CD8+ T-cell subset, which may be expanded in the peripheral blood of HIV-infected individuals, is more prone to in vitro apoptosis (9). In addition, there are data indicating that apoptosis of not only CD4+ T cells but also of CD8+ T cells occurs in secondary lymph nodes of HIV-infected patients (10, 11). In the present report we analyzed the susceptibility to apoptosis of T cells isolated from the lung of HIV-infected patients and the regulatory role of IL-15 on T-cell apoptosis. Our study demonstrates that pulmonary CD8+ T cells accumulating in the lungs of HIV-infected patients are preactivated Tc1 cells (12) prone to an excessive, spontaneous, and activation-induced apoptosis and that IL-15 may prevent this phenomenon, acting as an inhibitor of apoptosis pathways.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Study Populations

Eighteen HIV-seropositive patients (14 men and four women; eight smokers and 10 nonsmokers; average age 35.5 ± 5.5 yr) underwent bronchoalveolar lavage (BAL) evaluation. According to the technical recommendations and guidelines for the standardization of BAL procedures (13), in all patients a complete morphologic and immunologic analysis was available, including cell recovery, differential count of macrophages, lymphocytes, neutrophils, and eosinophils, and flow cytometry analysis of CD3, CD4, and CD8 BAL T-cell populations.

Seven healthy adult control subjects were selected (five men and two women; average age 31.5 ± 4.1 yr; two nonsmoking, healthy subjects, and five evaluated for complaints of cough without lung disease subjects [three smokers and two nonsmokers]). They showed normal physical examinations, chest X-rays, lung function tests, and BAL cell numbers.

Purification of Pulmonary T Cells

Lung T cells were enriched from the BAL cell suspensions by rosetting with neuraminidase-treated sheep red blood cells (SRBC) followed by Ficoll-Hypaque gradient separations (7).

Monoclonal Antibodies and Cytokines

The commercially available conjugated or unconjugated mAbs used belonged to the Becton Dickinson and PharMingen series and included: CD3, CD4, CD8, CD95, and CD120b. Anti-IL-15 M110 (IgG1) and anti-IL-15 receptor (IL-15R) (IgG1) mAbs were kindly provided by Dr. A. Troutt (Immunex Co., Seattle, WA). Anti-IL-2, anti-IL-4, anti-tumor necrosis factor-alpha (anti-TNF-), and anti-interferon gamma (anti-IFN-) mAbs were purchased from PharMingen (San Diego, CA). Anti-hCXCR3 mAb was purchased from R&D Systems (Minneapolis, MN). Anti IL-12R2 mAb was kindly provided by Dr. F. Sinigaglia (Roche Ricerche, Milan). Human IL-15 (hIL-15) was kindly provided by Dr. A. Troutt (Immunex).

Phenotypic Evaluation of BAL Cells

The frequency of BAL cells positive for the aforementioned reagents was determined as previously reported (7). Both BAL lymphocytes and pulmonary macrophages were gated in flow cytometry analysis using physical characteristics of cells and expression of the T-associated CD3 and macrophage-associated CD68 antigens in the area of lymphocytes and pulmonary macrophages, respectively. The expression of cytoplasmic cytokine was evaluated after permeabilization of cell membranes (7). The mean fluorescence intensity (MFI) was used to compare the positivity of cytokine and cytokine receptors on different cell populations. To evaluate whether the shift of the positive cell peak was statistically significant, the Kolmogorov-Smirnov test for analysis of histograms was used (7).

Induction and Analysis of Spontaneous and CD3-Mediated Apoptosis by Lung T Cells

To investigate the propensity of lung T cells to undergo apoptosis, highly purified BAL T cells were cultured at the concentration of 1 × 10⁶ cells/ml in RPMI medium in 24-well plates at 37° in 5% CO₂. In separate experiments BAL T cells were stimulated with anti-CD3 mAb (0.1 µg/ml) or phorbolmyristate acetate (PMA) (10 ng/ml) and ionomycin (1 µg/ml). After 24 h of incubation, apoptotic T cells were recognized using the method of Nicoletti and coworkers (14) or an annexin-V-based method (R&D Systems, Minneapolis, MN).

Effect of IL-15 on the Apoptosis of BAL T Cells

To assess whether IL-15 protects lung T cells from undergoing spontaneous apoptosis, BAL T cells were cultured for 24 h with medium alone and IL-15 (10, 100, and 1,000 ng/ml). The frequency of apoptotic BAL T cells was determined as reported previously.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Table 1 reports differential BAL cell counts and T-cell subsets in HIV-infected subjects and control subjects. Compared with normal subjects, cell recovery was significantly higher in HIV-infected patients and the absolute number of lymphocytes was increased, documenting a lymphocytic alveolitis. CD8+ T cells represented the majority of BAL T cells in HIV-infected patients, whereas less than 3% of T cells were CD4+.

BAL T Cells Accumulating in the Lung of Patients with HIV Infection are Tc1 Cells

Flow cytometry analysis showed that BAL T cells accumulating in the alveolar spaces of patients with HIV infection were preactivated CD45R0+/CD69+ T cells bearing CXCR3, IL-12R2, and interferon gamma (IFN-) (Figure 1), a pattern which is characteristic of Tc1 cells (12). As a confirmation, cytoplasmic IL-4 was not expressed by BAL T cells (less than 1% of BAL T cells were IL-4+ in all BAL samples). Concerning other activation-related molecules, they bore the  chain of the IL-15 receptor (IL-15R), the  chain of the IL-2R (CD132), CD95, and TNF-receptor type II (CD120b) at high density. Normal BAL T cells were CD45R0+ T lymphocytes which did not express IFN- nor cytokine (IL-15R, IL-12R, CD120a, and CD120b) or chemokine receptors (CXCR3) (Figure 2).

View larger version (31K): [in this window] [in a new window]   Figure 1.   The flow cytometry profile of BAL T cells recovered from one representative HIV-infected patient with lymphocytic alveolitis. Thirty-seven percent of BAL cells were lymphocytes, (A) and the CD4/CD8 ratio was 0.04 (B). Panels C to H show that BAL T lymphocytes are CD45R0+CD69+CXCR3+ IL-12R2+ cells bearing cytoplasmic IFN-. (Tc1 cells). They also express IL-15R (I ) and CD95 (K ). The positivity for the different molecules was determined as reported in METHODS. A consistent pattern of expression was seen in 12 patients with HIV infection and lymphocytic alveolitis.

View larger version (32K): [in this window] [in a new window]   Figure 2.   The flow cytometry profile of BAL T cells recovered from one representative healthy subject. Eight percent of BAL cells were lymphocytes, and the CD4/CD8 ratio was 2.1. Panels A to I show that normal BAL T lymphocytes express CD45R0 antigen, but not IFN- nor cytokine (IL-15R, IL-12R) or chemokine receptors (CXCR3). The positivity for the different molecules was determined as reported in METHODS. A consistent pattern of expression was seen in all control subjects.

T Cells of Most Patients with T-cell Alveolitis are Prone to Apoptosis and Show Decreased Survival Following Activation

In a parallel set of experiments we evaluated the predisposition of pulmonary T cells to undergo programmed cell death by culturing them for 24 to 36 h. Figure 3 shows data on DNA fluorescence of propidium iodide-stained BAL T cells. The percentage of apoptotic BAL T cells from HIV-infected patients was 32 ± 19 compared with 8 ± 4 of BAL T cells from control subjects (p < 0.05). The extent of apoptosis was variable in HIV-infected patients, varying from 6 to 68%.

View larger version (25K): [in this window] [in a new window]   Figure 3.   Quantification of apoptosis by flow cytometric analysis (D) of DNA fragmentation (A, B, and C ). Panels A, B, and C show histograms of BAL T cells from two representative HIV-infected subjects (B and C ) and from an HIV-seronegative subject (A). Panel D details the proportion of BAL T cells undergoing spontaneous apoptosis in 18 patients with HIV infection after a 24-h incubation in medium. The gray area identifies apoptosis values estimated in normal T cells. A percentage ranging from 6% to 68% of BAL T cells from HIV-infected patients underwent spontaneous apoptosis.

To determine whether the intensity of apoptosis in lung T cells was related to the expression of proapoptotic molecules, the percentage of lung T cells prone to cell death was plotted against the ex vivo expression of Fas by lung T cells. The extent of apoptosis was strongly correlated with levels of apoptosis. In six patients with > 40% apoptotic cells the MFI of Fas histograms was 162 ± 22 whereas in eight patients with < 20% apoptotic cells the MFI was 54 ± 21 (p < 0.001).

To evaluate the influence of T-cell activation on BAL T-cell apoptosis, the annexin based method was used as detailed in METHODS. Figure 4 shows a representative subject ( panels A and B) and illustrates the summary of data obtained from 11 patients with T-cell alveolitis ( panel C ). 29% ± 19 of lung T lymphocytes showed susceptibility to spontaneous cell death; after activation with CD3 mAb or ionomycin the number of T cells undergoing cell death was significantly higher than in medium-cultured T cells (48 % ± 11 and 52% ± 11, respectively; p > 0.05 in both cases).

View larger version (22K): [in this window] [in a new window]   Figure 4.   Quantification of the proportion of lung T cells undergoing apoptosis by flow cytometric analysis using the annexin-V-based method reported in METHODS. BAL T cells were cultured for 24 h in medium alone or after stimulation with anti-CD3 mAb (0.1 µg/ml) or PMA (10 ng/ ml) and ionomycin (1 µg/ml). Panels A and B show medium and CD3 stimulated T cells of a representative HIV-infected subject. Panel C details the proportion of lung T cells undergoing apoptosis in 11 HIV-infected patients. Anti-CD3 and PMA-ionomycin stimulation significantly increase the apoptotic rate.

IL-15 Protects BAL T Cells from Spontaneous and CD3-mediated Apoptosis

Despite the increased apoptosis of pulmonary CD8 T cells, most HIV-infected patients develop a T-cell alveolitis, even when CD4 T cells dramatically drop. This suggests that non-T molecules influence the persistence of the CTL response, preventing or retarding apoptosis of activated lung T cells. Because it is known that IL-15 suppresses the experimentally induced DNA fragmentation, in a parallel set of experiments we evaluated whether AMs obtained from patients with T-cell alveolitis express IL-15 (Figure 5) and the putative role of this macrophage-derived cytokine in local T-cell homeostasis (Figure 6).

View larger version (29K): [in this window] [in a new window]   Figure 5.   Expression of IL-15 and mTNF- by AM from two representative HIV-infected patients and an HIV-seronegative normal subject. Thirty-one percent and 27% of BAL cells were lymphocytes and the CD4/CD8 ratio was 0.03 and 0.09 in Cases 4 and 13, respectively. Phenotypic analysis was determined as reported in METHODS. A consistent pattern of expression of mTNF- (A and C ) and cytoplasmic IL-15 (B and D) was seen in all HIV- infected patients with ILD whereas AMs recovered from five normal subjects did not express these molecules (E and F  ).

View larger version (16K): [in this window] [in a new window]   Figure 6.   Effect of IL-15 on the survival of lung T cells. The quantification of apoptosis was made by flow cytometric analysis using the annexin-V-based method reported in METHODS. IL-15 is a potent inhibitor of spontaneous apoptosis of lung T cells; its protective properties are dose-dependent, because lower concentrations of the cytokine are less effective in protecting T cells from lethal death.

Profiles shown in Figure 5 indicate that AMs of HIV- infected patients are preactivated cells. In fact, purified AMs obtained from HIV-infected subjects showed an upmodulation of IL-15 ( panels A and C ) and membrane tumor necrosis factor-alpha (mTNF-) ( panels B and D) expression with respect to normal AMs ( panels E and F ). The MFI of IL-15 (182 ± 17) and mTNF- (179 ± 34) was higher in HIV-infected patients with T-cell alveolitis than in normal subjects (25 ± 5 and 35 ± 6, respectively; p < 0.001).

To study the effects of IL-15 on the survival of pulmonary T cells, BAL T cells isolated from 10 patients with HIV infection were cultured for 24 h in the presence of medium alone or different concentrations of IL-15, and the analysis of relative apoptosis rate was performed. As shown in Figure 6, IL-15 had a protective effect on BAL T cells, which become more resistant to cell death. Furthermore, the effect of IL-15 was dose-dependent because lower concentrations of the cytokine were less effective in protecting T cells from undergoing lethal apoptosis. In five patients we evaluated whether the antiapoptotic efficacy of the cytokine was inhibited by a neutralizing anti-IL-15 antibody. In the presence of 1,000 ng/ml of IL-15 the number of BAL T cells undergoing apoptosis was 13 ± 4 (range from 8 to 19%); the addition of an anti-IL-15 mAb to the culture system led to a significant increase in the apoptosis rate (35 ± 14; range between 19 and 50%; p < 0.05) whereas the addition of an irrelevant antibody had no effect.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The current study provides the first evidence that pulmonary CD8+ T cells accumulating in the lungs of HIV-infected patients are preactivated Tc1 cells prone to spontaneous and activation-induced apoptosis. Indeed, the degree of apoptosis was strikingly heterogeneous in different HIV-infected patients, because T-cell survival in some patients was normal and altered in others (Figure 1). Given this wide heterogeneity, differences in survival are likely to influence not only lymphocytosis, which may take place in the lung of most patients (15), but also the capabilities of lung CTL to clear HIV and, in turn, the natural history of HIV infection of the pulmonary microenvironment.

We observed that the apoptosis of pulmonary T cells correlates with the immune activation state of the lung T-cell compartment. HIV spreads early to resident immunocompetent cells which represent a continuous source of virions during the disease; the persistence of HIV in the lung is considered the primary mechanism of chronic stimulation of the pulmonary immune system. In this context, it is likely that the unceasing expression of HIV antigens leads to an uncontrolled activation of the pulmonary immune system, which in turn may program lung T cells for anergy/apoptosis. The apoptotic process is an essential step for the rapid clearance of activated T cells and thus regulates T-cell lymphocytosis associated with most viral infection (16). Nevertheless, this physiologic process could be detrimental when, as in the case of HIV infection, viral antigens persist at sites of infection. It should be noted that the concomitant occurrence of apoptosis and an excessive T-cell activation is not specific of the ILD caused by HIV. In fact, there are data indicating that other diffuse lung diseases characterized by a massive T-cell infiltration of pulmonary parenchyma and persistent antigenic stimulation, including sarcoidosis and hypersensitivity pneumonitis, are associated with a chronic T-cell activation, an upregulation of Fas expression (17, 18), and an increased degree of apoptosis (19).

As reported previously, we have demonstrated the infection of the pulmonary CD8 T-cell compartment by HIV (4, 20). Because it is known that the production of HIV antigens may accelerate Fas-mediated apoptosis of T cells (21), it is tempting to relate the productive infection to the apoptosis rate of lung CD8+ cells. Lung CD8+ T cells supporting high viral replication could be deleted either directly or by influencing the activation of the pulmonary immune system. In this context, the chronic overexpression of Fas by lung T cells and the influence of some viral proteins on the Fas-mediated cell death pathway could contribute to an excessive apoptosis (22). As a confirmation of this possibility, we noticed a direct relationship between the in vivo expression of Fas by BAL T cells and their predisposition to undergo spontaneous apoptosis. Further studies aimed at verifying these hypotheses are in progress in our laboratory. In particular, we are determining whether there is a relationship between apoptosis and the proportion of T cells that are HIV-infected, and whether productive HIV infection of the pulmonary microenvironment may modulate the expression of molecules that are involved in the apoptosis of lung CD8+ T cells, including Fas/CD95 and CD2, an antigen that regulates lung lymphocyte survival in the animal model (23). Furthermore, we are evaluating whether, as recently suggested (24), apoptosis of lung CD8+ T cells may be mediated by AMs through the interaction between mTNF- expressed by AMs (Figure 4) and tumor necrosis factor receptor type II (TNFRII) expressed by BAL T cells (Figure 1) or through a Fas/FasL interaction (25). It is expected that the use of blocking antibodies that disrupt mTNF- or Fas/FasL interactions in coculture experiments of lung CD8+ T cells and AMs may show whether macrophages are able to influence the survival of pulmonary T cells.

Despite the in vitro apoptosis shown by lung T cells, most of the patients described in this study had a T-cell alveolitis (Table 1). This fact suggests that molecules blocking apoptosis are actively produced in patients with intrapulmonary accumulation of CD8+ T cells. IL-2 is known to inhibit T-cell apoptosis (26), but pulmonary CD4+ T cells, i.e., the cell source of IL-2, were virtually absent in our patients. Thus, it is conceivable that other molecules surrogate IL-2 in providing antiapoptotic transduction signals for CD8+ T cells. In a search for other molecules that regulate the inhibition of T-cell death in HIV lung, we verified whether IL-15, which is known to trigger the growth of the CD8 T-cell pool and CTL recruitment (27, 28), also has apoptosis-protective properties in the lung. In fact, the survival of lung T cells, which abundantly express the  and  chains of the IL-2 receptor and IL-15R, was significantly influenced by the presence of IL-15, its protective effect being dose-dependent.

The demonstration that IL-15 acts as a potent inhibitor of cell death for pulmonary T cells extends our understanding of the complex mechanisms which regulate T-cell homeostasis at the pulmonary level. Considering that AMs from patients with T-cell alveolitis express IL-15, it is possible that these cells may have a role in protecting T cells from lethal apoptosis in vivo. However, the matter is not that clear, because AMs also express proapoptotic molecules, including mTNF-our present data and (29)that in theory could be detrimental for the control of HIV infection. The puzzle is also complicated by the fact that IL-15 promotes the entry of M-tropic HIV in T cells (30), suggesting a mechanism by which this cytokine favors the transition of HIV displaying the M-tropic phenotype (primarily associated with the initial phases) into the T cell-tropic phenotype that predominates as the disease progresses and T cells undergo apoptosis. Further efforts are now required to evaluate whether different T-cell abnormalities seen in the lung of HIV-infected patients may be explained by the effects of the virus on the balance of proapoptotic/antiapoptotic signals provided by AMs during different inflammatory responses (31).

Furthermore, because it is known that highly active antiretroviral therapy (HAART) reverses apoptosis of T cells (32) and leads to a reduction of HIV load in the lung (33), conclusive studies should be performed to evaluate the ultimate effects of HAART on pulmonary immunocompetence. In this regard, our preliminary data indicate that therapeutic intervention with HAART is associated with a return to normal activation status of pulmonary CD8 T cells (34). However, subjects on HAART show a suppression of viral replication in the lung and a decrease in the incidence of T-cell alveolitis with respect to untreated patients. Because detectable concentrations of IL-15 and other cytokines may be measured in the fluid component of the BAL of patients with ILD, structured studies should be planned in a wide number of patients with HIV infection to evaluate whether antiretroviral therapy has an impact on the pulmonary expression of cytokines that regulate T-cell homeostasis. On clinical grounds, this information could have an impact on strategies for designing immune intervention at the pulmonary level.