We have previously developed a human macrophage hybridoma model system to study the effect of HIV-1 infection on monocytic function. Upon coculture of one chronically (35 days postinfection) HIV-1-infected human macrophage hybridoma cell line, 43HIV, there was a dose-dependent decrease in the viability of cocultured Ag-stimulated T cells associated with an increase in DNA strand breaks. Enhanced apoptosis was determined by labeling with biotinylated dUTP and propidium iodide, increased staining with annexin V, increased side light scatter and expression of CD95, and decreased forward light scatter and expression of Bcl-2. There was also increased DNA strand breaks as determined by propidium iodide staining in unstimulated T cells cocultured with 43HIV and in T cells stimulated with anti-CD3 mAb and PHA. Pretreatment with 5145, a human polyclonal anti-gp120 Ab that recognizes the CD4 binding region, as well as with an anti-Fas ligand mAb blocked apoptosis in CD4+ T cells but not in CD8+ T cells. A soluble factor with a Mr below 10,000 Da was defined that induced apoptosis in CD4+ and CD8+ T cells and B cells. SDS-PAGE analysis of the active fractions revealed a band of 6000 Da that, after electroelution, had proapoptotic activity. The pI of the activity was estimated to be between 6.5 and 7.0. In conclusion, chronically HIV-1-infected monocytic cells induce apoptosis in bystander-, Ag-, anti-CD3-, and mitogen-stimulated T cells by multiple factors, which may contribute to the depletion of lymphocytes induced by HIV-1.
Progressive depletion of CD4+ T cells is a characteristic feature of HIV-1 infection (1). Despite intense investigation, it is not entirely clear how this process occurs. Both virologic and immunologic mechanisms have been thought to play an important role in the loss of CD4+ T cells based on in vitro and in vivo observations. Investigators have estimated that 109 virions are produced in HIV-1-infected individuals, and 5% of the total CD4+ T cells are destroyed and replaced daily (2, 3). HIV-1 mRNA can also be detected in a large number of CD4+ T cells in lymphoid tissue, suggesting that viral cytopathic effects may account for much of the T cell loss (4, 5). Nevertheless, immunologic mechanisms contribute to the decline in number and function of T cells seen during the course of the disease (6). Apoptosis or programmed cell death has been proposed as an alternative explanation for T cell loss in HIV-1-infected individuals (6).
Spontaneous apoptosis of CD4+ and CD8+ T cells and activation-induced apoptosis have been reported in PBMC and lymph nodes during HIV-1 infection (7, 8, 9, 10, 11, 12, 13). The accelerated apoptosis may relate to cross-linking of CD4 by gp120, leading to aberrant T cell signaling (14, 15, 16), cytokines (6), Fas and FasL3 interactions (17, 18, 19), superantigen activity encoded by HIV-1 products (20, 21), or the involvement of accessory cells. Several lines of evidence implicate accessory cells, including monocytes and dendritic cells, in the induction of apoptosis during the course of HIV-1 infection. Monocytes and dendritic cells serve as reservoirs for HIV-1-providing virions and the envelope protein gp120 to target CD4+ T cells (22). APC dysfunction as a result of HIV-1 infection may cause defective T cell activation, resulting in apoptosis instead of cellular activation (23, 24, 25, 26, 27). HIV-1 infection or cross-linking of CD4 on monocytes results in the up-regulation of FasL expression, which could induce apoptosis in uninfected bystander CD4+ T cells (28, 29).
We have been interested in studying HIV-1-monocyte interactions using a series of human monocyte and macrophage hybridomas obtained by fusing monocytes and macrophages with a mutagenized U937 promonocytic cell line (30). We have demonstrated early defects in monocytic function occurring in this system, including the inability to present soluble Ag to MHC-matched responder T cells and impaired accessory cell function for mitogens and anti-CD3-induced T cell proliferation (31, 32). There was also aberrant cytokine production with loss of IL-1 and IL-12 production and induction of IL-6 and IL-10 (31, 32). We have defined a series of late defects that occurred after prolonged HIV-1 infection (>4 wk), including impaired Ag trafficking and a direct toxic effect of the human macrophage hydridoma cell lines on Ag-stimulated cocultured T cells (31, 33). There was a dose-dependent decrease in the viability of tetanus toxoid (TT)-stimulated MHC-matched T cells cocultured with the chronically HIV-1-infected human macrophage hybridomas. In situ hybridization studies looking for the presence of HIV-1-specific mRNA in the cocultured T cells failed to detect any active infection (31). Here, we demonstrate that the chronically HIV-1-infected human macrophage hybridomas (and primary HIV-1BaL-infected monocytes infected for 2 wk) induce apoptosis in Ag-, mitogen-, and anti-CD3-stimulated T cells as well as in bystander T cells by multiple mechanisms.
Materials and Methods
Human macrophage hybridomas
Human macrophage hybridomas were obtained by fusing macrophages (obtained by allowing monocytes to mature into macrophages in Teflon bag cultures) with a hypoxanthine-guanine phosphoribosyl transferase-deficient promonocytic line (U937) as previously described (30). We have uniformly infected and characterized one clone, 43, with HIV-1 (43HIV) (33).
Mononuclear cells were separated from buffy coats obtained from normal healthy volunteers by Ficoll-Hypaque (Pharmacia, Piscataway, NJ) density gradient centrifugation. The cells were washed three times with sterile PBS and resuspended in RPMI 1640 (Life Technologies, Grand Island, NY) supplemented with 10% FCS (Life Technologies), 2 mM l-glutamine, and 1% penicillin/streptomycin (Life Technologies), henceforth called complete medium (CM). Freshly isolated PBMC were incubated at 37°C in CM and allowed to adhere for 45 min. The nonadherent cells were removed, and adherent cells were washed with sterile PBS, harvested with a rubber policeman, and stained with monocyte-specific anti-CD14 mAbs to assess the purity of the preparation. Ninety percent of the isolated cells expressed CD14 (30).
Monocytes or 43 cells were infected with HIV-1IIIB (34), HIV-1ADA (35), HIV-187.9 (36), and HIV-1BaL (37) as previously described (31, 33). These reagents were obtained through the AIDS Research and Reference Reagent Program, Division of AIDS, National Institute of Arthritis and Infectious Diseases, National Institutes of Health (Bethesda, MD). Dilutions of HIV-1 containing supernatant standardized to contain reverse transcriptase activity of 80,000 cpm/ml were incubated for 90 min followed by three washes with PBS.
Ag-, mitogen-, and anti-CD3-induced apoptosis
Clone 43 cells, 43HIV cells, and HIV-1-infected and uninfected monocytes were used as accessory cells in Ag-, mitogen-, and anti-CD3-induced T cell apoptosis. For these experiments, T cells were obtained from MHC-matched PBMC from normal blood donors and were monocyte depleted using a nylon wool column. The PBMC were incubated on the column for 45 min at 37°C and eluted. T cells isolated in this manner failed to respond to PHA or anti-CD3 mAbs. Monocyte-depleted T cells (105) were cocultured with varying concentrations of irradiated (cesium source) 43 cells, 43HIV cells, HIV-1-infected or uninfected monocytes (103-105), and either PHA (0.01–1 μg/ml) (Sigma, St. Louis, MO) or the anti-CD3 mAb 446 (1 μg/ml) in 0.2 ml of CM at 37°C in a 5% CO2 incubator for 24, 48, and 72 h. The 446 mAb has been characterized (38). It is an Ab directed against the γ-chain of the CD3 complex and can stimulate T cells using the FcR expressed on macrophages to cross-link the CD3/TCR complex. The Ab was purified from culture supernatant on a protein G column. The cells were harvested, and the apoptosis assay was performed as described below (32).
Isolation of purified CD4+ and CD8+ T cell populations
Purified CD4+ and CD8+ populations were isolated using Ab-coated plates. Primary T cells were first obtained by rosetting with neuramidase-treated SRBC by previously established methods (38). Anti-mouse Ig (Accurate Antibodies, Westbury, NY) was diluted in PBS to a concentration of 100 μg/ml and coated onto a 100-mm plate (Nunc, Naperville, IL) at 4°C overnight followed by five PBS washes. T cells were incubated either with anti-CD4 or anti-CD8 mAb for 45 min at 4°C followed by three PBS washes and then allowed to settle onto the coated plates for 30 min at room temperature (39). The nonadherent cell population was removed, and the purity of the isolated population was assessed by flow cytometry as described below. Ninety percent of the isolated cells expressed either CD4 or CD8.
B cell isolation
PBMC were isolated from leukocyte concentrate packs obtained from normal blood donors at the Mount Sinai Blood Bank (New York, NY) as previously described (30). T and non-T cells were separated by SRBC rosetting, and non-T cells were depleted of monocytes by plastic adherence yielding B cell-enriched populations (38).
Detection of DNA strand breaks and side and forward light scatter associated with apoptosis
MHC-matched T cells were cocultured with HIV-1-infected and uninfected human macrophage hybridoma cell lines or with HIV-1BaL-infected and uninfected monocytes and then assessed for the induction of apoptosis. In other experiments, T cells were incubated with different concentrations of UV-irradiated supernatant (180 min in a 280-nm UV transilluminator) from HIV-1BaL-infected human monocytes or 43HIV and then assessed for the induction of apoptosis. The cell-free supernatants were collected from chronically infected cells (4 wk after infection) and used fresh at varying concentrations. The cells were washed in sterile PBS three times and suspended in cacodylate buffer consisting of 0.2 M potassium, 25 mM Tris-HCl buffer (pH 6.6), 2.5 mM cobalt chloride (CoCl2), 0.25 μg/ml BSA, 100 U terminal deoxynucleotide transferase, and 0.5 mM biotin for 30 min at 37°C. The cells were then fixed in 1% formaldehyde for 25 min at 4°C, washed in PBS, and resuspended in 100 μl of sodium citrate buffer consisting of 2.5 μg/ml fluorescein avidin, 0.1% Triton-X, and 5% (w/v) nonfat dry milk for 30 min at 25°C in the dark. The cells were then stained with propidium iodide (5 μg/ml propidium iodide and 0.1% RNase A) and analyzed by flow cytometry for DNA strand breaks and changes in cell cycle. In some experiments the HIV-1-infected human macrophage hybridomas were pretreated with different concentrations of 5145, an anti-gp120 Ab (supplied by A. Pinter), or an anti-FasL Ab (PharMingen, Burlingame, CA). In the light scatter method to determine apoptosis in the T cells cocultured with 43HIV, cells were gated and analyzed for forward and side light scatter (32).
FITC-labeled annexin V, a phospholipid binding protein of the annexin family, was used to measure apoptosis with a commercially available kit (Coulter, Hialeah, FL). After incubating the HIV-1-infected human macrophage hybridomas with T cells, the cell samples were washed with ice-cold PBS followed by centrifugation at 500 × g at 4°C. The cells were stained simultaneously with annexin V FITC/phycoerythrin-labeled anti-CD3 mAb and incubated at room temperature for 10 min in the dark. The cells were then analyzed by flow cytometry to measure costaining of the CD3+ and annexin V+ populations, gating on the live cells.
T cells cocultured with HIV-1-infected human macrophage hybridomas were stained by indirect methods as previously described. For Bcl-2, the T cells were first permeabilized with 70% ethanol, washed three times with PBS followed by staining with anti-Bcl-2 mAbs (Dako, Carpinteria, CA) followed by affinity-purified F(ab′)2 FITC-conjugated goat anti-mouse Ig (Tago, Burlingame, CA). The cells were concurrently stained with phycoerthyrin-labeled anti-CD3 mAbs (Leu 4, Becton Dickinson, Mountain View, CA) and analyzed by flow cytometry, gating on live cells. CD95 staining was performed with commercially available Abs (Coulter) followed by affinity-purified F(ab′)2 FITC-conjugated goat anti-mouse Ig Abs (Tago) (30).
Internal labeling with [35S]cysteine and [35S]methionine and immunoprecipitation
Metabolic labeling was performed as follows. 43HIV was starved in methionine- and cysteine-free medium (RPMI 1640 Selectamine kit, Life Technologies, Grand Island, NY) and pulsed for 4 h with [35S]cysteine and [35S]methionine each at 50 μCi/5 × 106 cells. The labeled cells were lysed and subjected to immunoprecipitation, as previously described, with mAbs against class I (W6/32), polyclonal anti-HIV-1 serum, and 5145. Isotype-matched control Abs were used in the control lanes (40).
Fractionation of the monocytic supernatant
Molecular sizing of the 43, 43HIV, monocyte, and HIV-1 infected monocyte supernatants was conducted on a Pharmacia Sephacryl S-100 column using 12-ml initial fractions of crude concentrate (Amicon, Danvers, MA; Mr cut-off, 500 Da) in 50 mM NaP04 and 0.15 M NaCl, pH 7.0, as a buffer. One-milliliter fractions were collected from the column, and protein content was determined by absorbance at A280. The fractions were tested in the apoptosis assay or were precipitated with 10% TCA in preparation for SDS-PAGE analysis and silver staining.
Western blot analysis
Proapoptotic fractions from the Sephacryl S-100 column were run on a 10% SDS-PAGE with 500 mA constant current for 4 h. The proteins were transferred onto a nitrocellulose membrane (Schleicher and Schuell, Keene, NH) in transfer buffer (20% methanol, 150 mM glycine, and 25 mM Tris, pH 8.3). After transfer, the nitrocellulose membrane was blocked with 5% milk in PBS, incubated for 2 h at room temperature with polyclonal anti-HIV serum (provided by Dr. A. Pinter) (41). After washing five times in washing buffer (0.05% Tween-20 in PBS), secondary Ab was added for 2 h (2 mg/ml; horseradish peroxidase-conjugated goat anti-human IgG, Cappel Laboratory, Durham, NC). After washing five times, the membrane was developed by chemiluminescence (ECL, DuPont, Wilmington, DE) (32).
Electroelution of proapoptotic activity
In some experiments proapoptotic fractions were run on a 12% nondenaturing polyacrylamide gel, cut into 3-mm sections, and elutroeluted for 30 min in 1 ml of buffer. The eluates were then used in the apoptosis assays described above (42).
Reverse phase HPLC analysis
Reverse phase HPLC was performed on the proapoptotic fractions obtained from the Sephacryl S-100 column using a C18 (4.6 × 250 mm) column. Elution of bound proteins was developed using a linear gradient of 0.1% (v/v) trifluoroacetic acid in water and 70% (v/v) acetonitrile in 1% trifluoroacetic acid. A gradient volume of 60 ml was developed at a flow rate of 1 ml/min. Elution profiles were monitored at an absorbance of 215 nm. Solvent in the protein-containing fractions was removed by vacuum centrifugation in a Speed-Vac (Savant, Piscataway, NJ) (43).
Acetone precipitation was conducted on crude supernatants from the 43 and 43HIV cell lines. Acetone was chilled in an ice-salt bath to attain a temperature below 0°C. Proteins were fractionated from 43 and 43HIV supernatants by precipitation in 50 and 95% (v/v) acetone sequentially. The precipitated proteins were collected by centrifugation, and the residual acetone in the precipitates was removed by vacuum centrifugation in a Speed-Vac (Savant).
Anion exchange chromatography
Anion exchange chromatography was performed with a Mono-Q HR 5/5 (5 × 50 mm) column on an FPLC system (Pharmacia). The elution gradient was developed using 20 mM Tris-HCl, pH 7.5 (buffer A), and 1 M NaCl in buffer A (buffer B) at a flow rate of 1 ml/min. Samples were prepared for anion exchange chromatography by exhaustive dialysis in buffer A. The protein elution profile was monitored by absorbance at 280 nm (44).
Chronically HIV-1-infected monocytic cells induce apoptosis in Ag-activated T cells
Our previous data demonstrated that chronically HIV-1-infected (>35 days) human macrophage hybridomas were directly toxic to MHC-matched Ag-stimulated T cells. Since in situ hybridization studies did not document HIV-1 mRNA in the cocultured T cells (31), we investigated whether these cells were undergoing apoptosis by measuring DNA fragmentation. TT-stimulated MHC-matched T cells (105) were cocultured with different concentrations (103-105) of 43 and 43HIV for varying periods (24, 48, and 72 h). There was a dose-dependent “left shoulder” in the DNA staining (hypodiploid cells; Fig. 1,A) and a right shift in the UDP-biotin-stained DNA consistent with apoptosis. This diminished when the T cells were cocultured with decreasing concentrations of 43HIV. There was no DNA fragmentation when the TT-stimulated monocyte-depleted T cells were cocultured with the uninfected 43 cells, and there was no DNA fragmentation in the 43HIV cells (data not shown). Optimal results were obtained when the T cells were cocultured with 43HIV for 48 h. We also assessed apoptosis by light scatter analysis and demonstrated a progressive increase in the side scatter and decreased forward scatter of TT-stimulated T cells cocultured with increasing concentrations of 43HIV (data not shown). To confirm the induction of apoptosis, we stained the T cells with an anti-CD3 mAb and annexin V that has selective affinity for phosphatidylserine (45). Phosphatidylserine is exposed at the cell surface during early apoptosis and can be measured by annexin V staining in a variety of cell types (46). Consistent with the propidium iodide (PI), dUDP-biotin, and light scatter analysis, the number of CD3+ T cells staining with annexin V did increase when greater numbers of 43HIV cells were added (Fig. 1,B). No annexin V staining was observed in either the 43 or 43HIV cells (data not shown). Increased surface expression of CD95 and reduced intracytoplasmic expression of Bcl-2 are also associated with apoptosis (6). To further assess the mechanism of induction of apoptosis in our system, we measured surface expression of CD95 and intracytoplasmic expression of Bcl-2 in the T cells cocultured with 43HIV cells. Consonant with the DNA fragmentation experiments, the light scatter analysis, and the annexin V staining, there was reduced intracytoplasmic Bcl-2 staining (Fig. 1 C). CD95 expression on the cocultured T cells was marginally increased. Multiple time points were determined, and reduced intracytoplasmic Bcl-2 expression was determined 16 h after coculture with 43HIV.
Chronically HIV-1-infected human macrophage hybridomas induce apoptosis in bystander-, mitogen-, and anti-CD3-activated T cells
Bystander- and activation-induced apoptosis has been described in PBMC and lymph nodes of HIV-1-infected patients (8, 12). Since the 43HIV cells induced apoptosis in the TT-stimulated T cells we next wanted to determine whether 43HIV cells could also induce apoptosis in unstimulated bystander T cells as well as T cells stimulated with mitogens and anti-CD3 mAbs. In these experiments, MHC-matched monocyte-depleted T cells were cocultured with 43 and 43HIV cells and left unstimulated, while others were stimulated with PHA (0.01–1 μg/ml) and the anti-CD3 mAb 446 (1 μg/ml). More DNA fragmentation, as determined by PI staining, was present in the unstimulated and PHA- and 446-stimulated MHC-matched T cells cocultured with the 43HIV cells compared with that in the same T cells cocultured with uninfected 43 cells (Fig. 2).
Since we observed apoptosis in Ag-, mitogen-, and anti-CD3-stimulated T cells as well as in bystander T cells cocultured with the HIV-1-infected human macrophage hybridomas, we sought to validate our results using primary HIV-1-infected monocytes. TT-, mitogen-, and anti-CD3-stimulated monocyte-depleted T cells were cocultured with autologous, chronically (14 days) HIV-1BaL-infected and uninfected monocytes. More DNA fragmentation was detected in the T cells cocultured with HIV-1BaL-infected monocytes than in T cells cocultured with the uninfected monocytes (Table I). The data, however, were more variable. In four of nine separate experiments using different monocyte and T cell preparations, there was more DNA fragmentation induced by the HIV-1BaL-infected monocytes compared with the uninfected monocytes. Similar results were obtained when we infected the monocytes with other monocytotropic strains, including HIV-1ADA and HIV-187.9 (data not shown). Since there are many potential explanations for the induction of apoptosis, we explored the various possibilities.
|Unstimulated Medium (%) .||.||Stimulus (%) .||.||.|
|T cells .||T + M/T + MHIV .||PHA T + M/T + MHIV .||446 T + M/T + MHIV .||TT T + M/T + MHIV .|
|Unstimulated Medium (%) .||.||Stimulus (%) .||.||.|
|T cells .||T + M/T + MHIV .||PHA T + M/T + MHIV .||446 T + M/T + MHIV .||TT T + M/T + MHIV .|
HIV-1BaL-infected and uninfected monocytes from nine separate monocyte preparations were cocultured with autologous monocyte-depleted PBMC and either left unstimulated or stimulated with PHA, 446, and TT and cultured for 5 days. Apoptosis was assessed by propidium iodide staining. Bystander and activation-induced apoptosis (*) was observed in four experiments. The values represent the percentages of subdiploid fraction (Ao).
The gp120 induces apoptosis in CD4+ T cells cocultured with 43HIV
Engagement of CD4 by gp120 resulting in aberrant T cell signaling has been implicated in the induction of apoptosis in CD4+ T cells (14, 15, 16, 47). Since gp120 produced by virus may be playing a role in the 43HIV cells, we first labeled the 43HIV cell line with [35S]cysteine and [35S]methionine and immunoprecipitated gp120 to determine the amount produced by the 43HIV cells compared with other HIV-1 proteins. The replication pattern of HIV-1 in macrophages is different from that in T cells. Consistent with previously reported data, there was reduced production of gp120 in the 43HIV cells compared with other HIV-1 proteins, including p24 (Fig. 3,A) (48). This is illustrated in the lane using the anti-HIV Ig, where only p24 is immunoprecipitated. We next attempted to block apoptosis in purified populations of CD4+ and CD8+ T cells by pretreating the 43HIV cells with 5145, a human anti-gp120 Ab that recognizes the CD4 binding site (49). Apoptosis was blocked with 5145 in the purified CD4+ T cell populations but not in the purified CD8+ T cell populations (Fig. 3 B). The induction of apoptosis in the CD8+ T cells cocultured with 43HIV was an unexpected finding in our system. However, in the lymph nodes of HIV-1-infected patients, apoptosis is not only restricted to CD4+ T cells but occurs in CD8+ T cells and in B cells as well (50, 51).
FasL expression induces apoptosis in the cocultured T cells
Fas-FasL interactions effectively induce apoptosis in T cells (52, 53). Up-regulation of FasL expression, which occurs in monocytes after HIV-1 infection and after cross-linking of CD4, has been proposed as a possible mechanism for bystander T cell death (28, 29). We performed surface immunofluorescence studies for FasL on the 43HIV cells at different time points after HIV-1 infection to determine whether there was any potential role for FasL in the induction of apoptosis in our system. There was a modest increase in FasL expression 3 wk after infection (8 vs 22%) in the 43HIV cells compared with the uninfected 43 cells (Fig. 4,A). We next pretreated the 43HIV cells with anti-FasL Ab, attempting to block apoptosis. The anti-FasL Ab blocked apoptosis in the purified CD4+ T cell population but not in the CD8+ T cells (Fig. 4 B). Accessory cell dysfunction may also lead to defective T cell activation, resulting in apoptosis rather than T cell proliferation. Abnormalities in expression of costimulatory molecules on accessory cells may impair T cell signaling (18, 54). To determine whether HIV-1 infection affects the expression of costimulatory molecules required in T cell activation, we performed surface immunofluorescence staining for LFA-1-α, LFA-3, CD80, and CD86 on 43 and 43HIV cells at different time points. There was no difference in the surface expression of LFA-1-α, LFA-3, CD80, or CD86 in the 43HIV cells compared with the uninfected 43 cells at all time points tested (data not shown).
Identification of a soluble proapoptotic factor
Since we identified apoptosis in CD8+ T cells that was not mediated by FasL or gp120, we looked for the production of a soluble proapoptotic factor. Using UV-irradiated supernatant from 43HIV cocultured with purified populations of CD4+ and CD8+ T cells, it was possible to demonstrate a dose-dependent increase in DNA fragmentation (as determined by propidium iodide staining) in both the CD4+ and CD8+ T cell populations (Fig. 5,A). The supernatant-induced apoptosis was not blocked by the 5145 and anti-FasL Abs. The UV-treated supernatant also induced apoptosis in mitogen- and anti-CD3-stimulated T cells, and the proapoptotic activity was lost after treatment with pronase (data not shown). No DNA fragmentation was observed in the supernatant from the uninfected 43 cells (data not shown). Apoptosis of B cells has also been reported to occur in the lymph nodes of HIV-1-infected patients (50, 51). To determine whether the 43HIV supernatant had proapoptotic activity in these cells as well, we cocultured dilutions of UV-treated 43HIV supernatant with target B cell populations and assessed apoptosis by staining the DNA with propidium iodide. Similar to the data in the T cells, there was dose-dependent DNA fragmentation with increasing concentrations of UV-treated 43HIV supernatant (Fig. 5,B). Consistent with the propidium iodide studies, there was also increased annexin V staining in the T cells cocultured with UV-treated 43HIV supernatant (data not shown). We also sought to validate our findings in UV-treated supernatant from chronically HIVBaL-infected (14 days) peripheral blood monocytes. Similar to the supernatant data from the 43HIV cells, there was a dose-dependent induction of DNA fragmentation in T cells cultured with different concentrations of UV-treated supernatants from four of the nine preparations of HIV-1BaL-infected monocytes (Table II).
|T Cells Alone (%) .||T Cells + UV Supernant from HIV-1BaL-Infected Monocytes (%) .|
|T Cells Alone (%) .||T Cells + UV Supernant from HIV-1BaL-Infected Monocytes (%) .|
UV-treated supernatants from nine different HIV-1BaL-infected peripheral blood monocyte preparations were cocultured with unstimulated target T cells and apoptosis determined by propidium iodide staining after 48 h. Apoptosis (*) was observed in four experiments. The values represent the percentages of cells in the subdiploid fraction (Ao).
Characterization of the proapoptotic activity
To more carefully characterize the proapoptotic activity in the 43HIV supernatant, we concentrated 1 l of UV-treated supernatant using Amicon ultrafiltration (Mr cut-off, 500 Da) and passed it over a Sephacryl S-100 sizing column (Fig. 6,A). Proapoptotic activity was then determined in the different fractions by measuring DNA fragmentation in unstimulated target T cell populations. Proapoptotic activity was maximal in those fractions (fractions 5 and 6) corresponding to a Mr of 6,000 Da (Fig. 6,B). When the active fractions were electrophoresed on a 10% SDS-PAGE gel, a band corresponding to a Mr of 6,000 Da was detected (Fig. 6,C). The active fractions from the 43HIV cell line were then electrophoresed on a nondenaturing SDS-PAGE gel, and proapoptotic activity was electroeluted from gel slices corresponding to a Mr <10,000 Da (Fig. 6 D). Since several HIV-1 proteins, including Tat and Vpu, have been associated with apoptosis (55, 56), we performed a Western blot analysis of fractions 5 and 6 with polyclonal anti-HIV-1 Abs known to contain anti-Tat and anti-Vpu Abs. No reactivity was observed between these Abs and the peptide species contained in fractions 5 and 6 by Western blot analysis (data not shown). Again to validate our findings in the 43HIV cell line, we pooled and concentrated the supernatants from the HIV-1BaL-infected primary monocytes with Amicon ultrafiltration (Mr cut-off, 500 Da) and passed it over the Sephacryl S-100 column sizing column. Proapoptotic activity was present in those fractions corresponding to a Mr <10,000 Da (data not shown). No proapoptotic activity was present in the concentrated supernatant from the uninfected monocytes. We further characterized fractions 5 and 6 from the 43HIV supernatant by reverse phase HPLC analysis. We compared HPLC elution profiles of fractions 5 and 6 and demonstrated that unique fractions that had proapoptotic activity were present in the 43HIV supernatant but not in the 43 supernatant. We then attempted to isolate proapoptotic activity by precipitation with acetone. Pro-apoptotic activity could not be precipitated with acetone at a concentration <80% saturation, which is a characteristic that is observed with smaller peptides. Results from peptide binding to anion exchange matrixes at different pH values indicated that the pI of the proapoptotic activity is estimated to be between 6.5 and 7.0.
Although apoptosis is an important factor in the T cell depletion that occurs during the course of HIV-1 infection, it is unclear how this process occurs. Cross-linking of CD4 by gp120, involvement of cytokines, superantigen activity encoded by HIV-1 proteins, and aberrant accessory cell function have all been proposed as potential mechanisms to account for the induction of apoptosis (6). Recent evidence, however, underscores the importance of accessory cells, including dendritic cells and monocytes, as effector cell populations in inducing lymphocyte apoptosis in AIDS (22, 23, 24, 25, 26, 27, 28, 29, 57, 58). In chimpanzees, in which HIV-1 is unable to infect monocytes, the persistent infection in T cells occurs without the development of T cell apoptosis (24, 25, 26, 27). In the hu-PBL-SCID mouse model, monocytotropic strains of HIV-1, but not T cell-tropic strains, can cause extensive T cell depletion (23). In human in vitro systems, Wu et al. demonstrated that both CD4+ and CD8+ T cells treated with anti-CD3 mAb or gp120 could be primed for apoptosis after coculture with PMA-treated monocytes (57). Munn et al. demonstrated that granulocyte-macrophage CSF-treated monocytes could induce apoptosis in anti-CD3-stimulated T cells (58), and Suiss et al. showed that dendritic cells could kill CD4+ T cells via Fas/Fas ligand interactions (59). Badley et al. (28) have reported that HIV-1 infection of monocytes results in the induction of FasL expression, and Oyaizu et al. (29) have demonstrated that cross-linking of CD4 on monocytes also induces FasL expression, which causes apoptosis of bystander T cells.
We have extended our previous observations on the toxic effect of chronically HIV-1-infected human macrophage hybridomas on Ag-activated T cells to determine whether they were capable of inducing apoptosis (31). Ag-, anti-CD3-, and mitogen-stimulated T cells as well as unstimulated bystander T cells underwent apoptosis when exposed to the chronically HIV-1-infected monocytic cells (Figs. 1 and 2 and Table I). The induction of apoptosis by the primary HIV-1BaL-infected monocytes was more variable than that observed with the 43HIV cells, which may relate to the different rates of monocytic infection with HIV-1 or the level of chronicity of infection. We identified three different mechanisms by which the 43HIV cells induced apoptosis, including expression of gp120 and FasL and production of a proapoptotic factor. Interestingly, there was no change in the surface expression of the costimulatory molecules CD80 and CD86 on 43HIV cells after infection. Defective interactions between costimulatory molecules on APCs, especially the B7 family of proteins, and T cells (CD28/CTLA-4) have been implicated in the induction of apoptosis (60). Anti-CD28 Ab can provide a rescue signal to block apoptosis in T cells of HIV-1-infected patients (8), while stimulation with anti-CTLA-4 Ab promotes apoptosis (54).
All these mechanisms may play a role in the induction of apoptosis in T cells in HIV-1-infected patients, although perhaps to varying degrees. The replication pattern of HIV-1 in monocytes is different from that in T cells (48). In monocytes the virus produces lesser amounts of gp120 and accumulates in intracytoplasmic vacuoles with relatively little shedding (61). The amount of gp120 exposed on the cell surface needed to engage CD4 and induce apoptosis is reduced compared with that in T cells (Fig. 3,A). Furthermore, in our system gp120 would only account for apoptosis in the CD4+ T cells, not in CD8+ T cells or B cells (Fig. 3,B). Increased surface expression of FasL on 43HIV cells induced apoptosis in the cocultured bystander T cells (Fig. 4,A). However, in line with the published results of others, FasL-mediated apoptosis occurred in CD4+ T cells but not in CD8+ T cells (62) (Fig. 4,B). In our system, apoptosis of CD4+ T cells could be explained by gp120 and FasL expression but not CD8+ T cells and B cells. Even though the anti-gp120 and anti-FasL Abs blocked apoptosis in CD4+ but not in CD8+ T cells, the proapoptotic factor induced apoptosis in both CD4+ and CD8+ T cells (Fig. 5 A). In these experiments the supernatant is from cells continuously growing in culture, so that the concentration of the proapoptotic factor is higher than that in the blocking studies. The interaction between gp120 and FasL, and the proapoptotic factor is uncertain. There may also be differences in signaling pathways. The anti-gp120 and anti-FasL Abs may suppress the production or release of the proapoptotic factor from 43HIV cells.
The proapoptotic factor produced by 43HIV induced apoptosis in CD4+ and CD8+ T cells as well as B cells (Fig. 5, A and B). We could also identify proapoptotic activity in the supernatants from four preparations of HIV-1BaL-infected primary monocytes 2 wk after infection (Table II). When we fractionated the pooled supernatant from 43HIV cells and primary HIV-1BaL-infected monocytes, we could identify proapoptotic activity in fractions corresponding to Mr <10,000 Da (Fig. 6). In other studies describing the effect of monocytes on the induction of apoptosis in cocultured T cells, no soluble factor was identified (29, 57, 58, 59). In these studies monocytes were treated with PMA, macrophage CSF, or anti-CD4 mAb, but not infected with HIV-1. Our preliminary analysis suggests that this factor is a peptide (since its activity is lost after pronase treatment) that does not appear to be HIV-1 derived and can induce apoptosis not only in bystander T cells but also in mitogen-, antigen-, and anti-CD3-activated T cells. Since the low Mr peptide induced apoptosis in a variety of cell types, the production of this factor by HIV-1-infected monocytes may contribute to the generalized state of apoptosis for CD4+, CD8+ T cells and B cells that has been described in HIV-1-infected patients (7, 12, 13). In lymph nodes, apoptosis is related to a general state of immune activation but not to viral load or stage of disease (52). It is uncertain what role this peptide has in the induction of apoptosis in HIV-1-infected patients. Studies are presently underway to define this factor biochemically and to measure levels in HIV-1-infected individuals in lymph nodes and PBMC at different stages of disease.
There is precedence for the concept of macrophage-derived proapoptotic factors. Macrophages have been reported to produce proapoptotic cytokines as well as apoptosis promoting low Mr molecules, such as reactive oxygen molecules, PGs, and nitric oxide (6). After HIV-1 infection, there is increased production of proinflammatory cytokines, including IL-6, IL-8, and TNF-α (63). In HIV-1-infected individuals, this cytokine imbalance may contribute to apoptosis. Both TNF-α as well as IFN-γ promote apoptosis (64, 65, 66, 67, 68, 69, 70). Recent evidence also suggests that two other predominantly macrophage-derived cytokines, IL-10 and IL-12, are involved in the regulation of apoptosis. IL-10 promotes apoptosis, while IL-12 prevents it (70). We have demonstrated that HIV-1 infection of the human macrophage hybridomas induces IL-6, IL-8, and IL-10 production along with loss of IL-1 and IL-12 production (31, 32). There may be synergy in the induction of apoptosis between the cytokines induced by HIV-1 infection and the low Mr proapoptotic factor.
The transduction of a proapoptotic signal at least through Fas involves the activity of proteases, including IL-1β-converting enzyme (71, 72). It has also been possible to block apoptosis in vitro in murine systems (73) and in PBMC of HIV-1-infected individuals by treatment with protease inhibitors (11). Similarly, IL-2 has been reported to inhibit apoptosis in PBMC from HIV-1-infected patients by increasing intracytoplasmic levels of Bcl-2 (74). Although T cells cocultured with the 43HIV cells had decreased Bcl-2 expression (Fig. 1 C) and some increase in FasL expression, there was no change in either CD95 or Bcl-2 expression in T cells cocultured with the low Mr factor. It remains to be determined whether apoptosis induced by this peptide uses pathways that can be blocked with IL-1β-converting enzyme protease inhibitors and IL-2.
In conclusion, the human macrophage hybridoma cell lines induce apoptosis by multiple mechanisms, including gp120, FasL expression, and the production of a low weight proapoptotic factor. Apoptosis induced by monocytes may play an important role in the lymphoid depletion seen in AIDS.
This work was supported by National Institutes of Health Grant R-29-256990 and the Irma T. Hirschl Career Development Trust (to K.S.), and in part by a Research Centers in Minority Institution Award (RR-03037) to Hunter College from the National Center for Research Resources, National Institutes of Health (to Y.K.Y.).
Abbreviations used in this paper: FasL, Fas ligand; TT, tetanus toxoid; CM, complete medium; PI, propidium iodide.