Activation-Induced T Cell Death Exacerbates Trypanosoma cruzi Replication in Macrophages Cocultured with CD4⁺ T Lymphocytes from Infected Hosts

Mice

Male BALB/c and C57.Bl.6 (B6) mice, 6 to 8 wk of age, were obtained from the Oswaldo Cruz Institute (IOC-FIOCRUZ, Rio de Janeiro, Brazil) animal facility. Mice were infected s.c. with 10⁵ chemically induced metacyclic trypomastigotes from the T. cruzi clone Dm28c in 0.1 ml, as previously described (17). Uninfected littermates were used as controls and as a source of normal peritoneal Mφ. Chemically induced metacyclic forms were obtained with triatomine artificial urine-proline medium as originally described (19). Parasitemia peaked between 18 and 26 days of infection. All T cell preparations used in this study were obtained between 18 and 26 days of infection.

Infection of Mφ monolayers with T. cruzi

Normal resident Mφ from uninfected BALB or B6 mice were obtained by peritoneal lavage with DMEM. Cells were washed and cultured at 37°C in 7% CO₂ in a humid atmosphere at 3 × 10⁵ cells/well in 24-well vessels (Corning Glass Works, Corning, NY) in 1 ml of complete culture medium containing DMEM (Sigma Chemical Co., St. Louis, MO), supplemented with 10% heat-inactivated FCS, 2-ME (5 × 10⁻⁵ M), MEM nonessential amino acids, 10 mM HEPES, 2 mM l-glutamine, 1 mM sodium pyruvate, and 10 μg/ml gentamicin sulfate. After overnight incubation, nonadherent cells were discarded, and adherent cells were washed, leaving approximately 2 × 10⁵ Mφ. These Mφ monolayers were infected by overnight incubation with Dm28c metacyclic trypomastigotes (19) at a 10:1 parasite: cell ratio in culture medium at 37°C in 7% CO₂. After 18 h, monolayers were extensively washed to remove extracellular parasites.

Coculture of Mφ monolayers with purified T cell subsets

Splenocytes obtained from either T. cruzi-infected or control mice syngeneic with the Mφ donors were depleted of erythrocytes by treatment with Tris-buffered ammonium chloride and filtered through nylon wool columns. Nonadherent cells were subjected to negative selection by magnetic cell sorting using a mixture of anti-CD8 (or anti-CD4, in the case of purified CD8⁺ T cells), anti-MHC class II, anti-B220, anti-MAC-1, and anti-CD16/CD32 mAbs as previously described (17) plus anti-TCRγδ to yield highly purified CD4⁺ or CD8⁺ T cells after three rounds of separation with anti-rat Ig-coated magnetic beads (Bio-Mag, Perseptive Diagnostics, Cambridge, MA). By flow cytometry, CD4⁺ T cells had <1% contaminating CD8⁺ and MHC class II⁺ cells, in both control and infected groups.

The Mφ monolayers provided an efficient source of accessory cells for anti-TCR activation of purified T cells. Since addition of phorbol ester does not interfere with CD4⁺ T cell AICD in cells from infected mice (14), and since T cell activation by anti-Ly-6 strictly requires phorbol ester as a cofactor (20), most experiments were conducted in the presence of phorbol ester (0.5 ng/ml PMA; Sigma). Anti-TCR activation of T cells in the absence of phorbol ester yielded similar results (see Table II⇓).

Two kinds of experiments required modifications of the above procedure. First, in cultures performed with membrane-separated culture inserts (MilliCell inserts for 24-well vessels; 0.4 μm; Sigma), endogenous splenic accessory cells were not depleted to ensure proper T cell activation in the compartment separated from Mφ. Therefore, CD4⁺ T cells were obtained by treatment of nylon wool-nonadherent cells with anti-CD8 mAb and C, followed by a second cycle of indirect cytotoxicity as previously described (17). Second, in all experiments involving anti-Ly-6 mAb, purified T cells were supplemented with 10% irradiated (3000 rad) normal syngeneic splenocytes regardless of the stimulus used. This treatment restored anti-Ly-6-mediated T cell activation, which was deficient with adherent Mφ.

Purified CD4⁺ T cells (2 × 10⁶) in 1 ml of complete culture medium, were added to prewashed, T. cruzi-infected syngeneic Mφ monolayers (2 × 10⁵) in the presence of PMA and in the presence or the absence of 10% (v/v) culture supernatant of anti-TCRαβ mAb H57.597 (21). Additional Abs added along with anti-TCR included mouse (gld strain) anti-mouse FasL (B6 strain) mAb K10 (10 μg/ml), a neutralizing IgG2b/k anti-FasL Ab that reacts with an allotypic determinant on mouse FasL that is present in B6- but not in BALB-activated T cells (22); and mouse isotype control mAb 156-8 (10 μg/ml), an IgG2b/k Ab specific for hepatitis virus B (AYW3 strain, produced at FIOCRUZ, Rio de Janeiro, Brazil). Cultures were set up in duplicate and maintained for up to 13 days. On the indicated days, individual wells were gently pipetted up and down, and 20-μl aliquots of the supernatant were removed for assessment of viable extracellular T. cruzi trypomastigote numbers. Motile trypomastigotes were easily identified in the counting chamber, based on established morphologic criteria (23). Results are presented as the mean and SE of trypomastigote number in the duplicate cultures.

In experiments comparing the effect of anti-Fas treatment (using the Fas agonist mAb Jo2) (24) on Ly-6 activation, a two-step culture procedure was used to avoid coincidence of both CD4⁺ T cell AICD and the onset of lymphokine secretion (25). Purified CD4⁺ T cells (2 × 10⁶, plus 10% irradiated splenocytes) were cultured alone in 1 ml of complete culture medium containing PMA plus 2 μg/ml hamster anti-mouse Fas mAb Jo2 or 2 μg/ml of control hamster IgG mAb. After overnight incubation, the entire content of individual wells was aspirated and transferred to wells containing decanted T. cruzi-infected Mφ monolayers. Cocultures were then initiated in the absence or the presence of newly added activating anti-TCR or 0.5% (v/v) ascites of anti-Ly-6 A/E mAb D7 (20). Extracellular trypomastigote accumulation was measured as described above. Controls included infected Mφ cultured with PMA in the absence of T cells and in the presence or the absence of anti-Fas or anti-Ly-6. No changes in Mφ viability or number of trypomastigotes recovered were noticed with any of these treatments. Suspensions of CD4⁺ T cells from infected mice contained a limited (<10²) number of motile trypomastigotes. Following activation in the absence of adherent Mφ, no detectable parasites were recovered after 5 days or more, even when partially purified T cells were strongly activated.

Quantitation of T cell viability

Either partially or highly purified CD4⁺ T cells were cultured alone or added to infected Mφ at 2 × 10⁶/ml in replicates in 24-well vessels in the absence (PMA alone) or the presence of anti-TCR (10% supernatant), anti-Ly-6 (0.5% ascites), anti-Fas, or control hamster IgG (2 μg/ml) and anti-FasL or control mouse IgG2b (10 μg/ml). After 20 h, cells were gently dispersed with a Pasteur pipette or detached with EDTA and washed, and viable cell counts for each individual well were determined by trypan blue exclusion. The mean viable cell recovery in unstimulated cultures was taken as a reference. The percent cell loss (the reciprocal of remaining viable cells) was calculated for each individual well according to the formula: % cell loss = 100 − (viable cell number in stimulated culture) × 100/(viable cell number with PMA alone). The mean and SE of duplicate or triplicate cultures are shown.

Assay for IFN-γ production

Purified CD4⁺ T cells were cultured overnight with anti-Fas or isotype control as described above and were stimulated, or not, with anti-TCR the next day. Supernatants were collected 48 h after TCR activation and stored frozen until use. The IFN-γ content in supernatants was evaluated by a sandwich ELISA technique, using two rat anti-mouse IFN-γ mAbs (PharMingen, San Diego, CA), one of which was biotinylated. The reaction was revealed with 2.5 μg/ml avidin-peroxidase conjugate (Sigma) using ABTS substrate, according to a protocol provided by the manufacturer (PharMingen). Dilutions of the supernatants were compared with a standard curve of murine rIFN-γ (PharMingen), ranging from 1.0 to 10.0 ng/ml.

Abs and cytokines

Anti-CD8 mAb 53-6.7, anti-B220 mAb 6B2, anti-MHC class II mAb AMS 32-1, anti-MAC-1 mAb M1/70, anti-CD16/CD32 mAb 2.4G2, anti-Fas mAb Jo2, anti-FasL mAb K10 (22), and control hamster IgG mAb were purchased from PharMingen. Anti-CD4 mAb GK1.5, anti-TCRγδ mAb 13D5, anti-rat κ chain mAb MAR 18.5, and anti-Ly-6 A.2/E.1 mAb D7 (20) were gifts from Dr. Ethan Shevach, National Institutes of Health (Bethesda, MD). Anti-TCRαβ mAb H57.597 (21) was a gift from Dr. Maria Bellio, Institute Pasteur (Paris, France). A control mouse IgG2b mAb (mAb 156-8 against HVB) was obtained from Bio-Manguinhos, FIOCRUZ. Anti-IFN-γ mAb R4-6A2 and biotinylated anti-IFN-γ mAb XMG1.2 were obtained from PharMingen. The murine recombinant cytokines used were rIFN-γ (20 ng/ml), rGM-CSF (1 ng/ml), rTNF-α (40 ng/ml), rIL-1β (20 ng/ml), rIL-2 (50 U/ml), rIL-4 (10 ng/ml), and rIL-10 (40 ng/ml) and were obtained from PharMingen.

Statistical analysis

Comparisons between control and experimental groups were made using Student’s t test. p < 0.05 or lower was considered significant. Since trypomastigote accumulation in culture is exponential, data were normalized by taking the logarithmic transformation before the t test was applied.

Decrease in T. cruzi replication in macrophage monolayers induced by rIFN-γ and by supernatants from activated CD4⁺ T cells

Peritoneal resident Mφ monolayers from BALB/c mice were infected by 18-h exposure to T. cruzi metacyclic trypomastigotes, washed, and recultured for up to 13 days. Extracellular motile trypomastigote (infective) forms were first observed after 3 days in culture, and their number continuously increased thereafter. At the Mφ density used, trypomastigotes comprised the predominant parasite form in culture. There was variability in the number of trypomastigotes produced between different experiments, probably related to the period of time that epimastigotes are maintained in the laboratory before induction of metacyclogenesis. However, trypomastigote growth in replicate cultures within the same experiment was quite similar. Several recombinant murine cytokines were tested to determine the kinetics of extracellular trypomastigote accumulation in Mφ cultures (Fig. 1⇓). Among all cytokines tested (rIL-1β, rIL-2, rIL-4, rIL-10, rTNF-α, rGM-CSF, and rIFN-γ), only rIFN-γ had a clear protective and long term effect against replication of T. cruzi in Mφ monolayers, as detected after 5 to 8 days in culture (Fig. 1⇓). On the other hand, rGM-CSF consistently increased the number of accumulated T. cruzi trypomastigotes (Fig. 1⇓). It should be noted that at earlier time points, such as 3 days in culture, IFN-γ completely blocked the emergence of motile parasites in the medium (not shown). We also tested the effect of supernatants from anti-TCR-activated CD4⁺ T cells from either control or T. cruzi-infected mice (Fig. 1⇓). Only CD4⁺ T cells from infected hosts produced supernatants active against T. cruzi replication. In agreement with previous studies (12), we found that CD4⁺ T cells from acutely infected mice produced 18 times more IFN-γ than control CD4⁺ T cells following activation with anti-TCR mAb (not shown).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Blockade of trypomastigote growth in Mφ by IFN-γ and supernatants from activated CD4⁺ T cells. BALB peritoneal Mφ, infected with metacyclic T. cruzi trypomastigotes (Dm28c), were treated with an excess of the indicated recombinant murine cytokines or with 10% (v/v) anti-TCR-activated T cell supernatants in a single addition on day 0. Cultures were maintained for 8 days (Expt. 1) or 5 days (Expt. 2), and viable trypomastigote number was determined. Results are the mean and SE of duplicate cultures. Control indicates infected Mφ alone. Significant differences were obtained with IFN-γ, GM-CSF, and supernatant from infected CD4⁺ T cells only.

Direct CD4⁺ T cell activation exacerbates T. cruzi replication in cocultured Mφ monolayers

Given the above results, we expected to find significant containment of parasite replication in Mφ following activation of cocultured CD4⁺ T cells with anti-TCR mAb. However, coculture of purified CD4⁺ T cells from infected mice by itself increased trypomastigote replication 2-fold in the absence of exogenous stimulation (Table I⇓). Moreover, anti-TCR activation of these T cells resulted in a marked increase (12-fold in this experiment) in the number of T. cruzi trypomastigote forms recovered (Table I⇓). Enhancement of parasite replication was observed at all time points investigated throughout a 13-day period. Growing parasites were derived from infected Mφ, since activation of CD4⁺ T cells alone did not result in detectable parasite growth (data not shown). The ability to potentiate parasite replication was a characteristic of CD4⁺ T cells from T. cruzi-infected donors, since activation of CD4⁺ T cells from control mice did not result in enhanced parasite replication or resulted in much less replication compared with that in T cells from infected mice (data not shown). Exacerbation of parasite replication required CD4⁺ T cell activation by anti-TCR, but addition of PMA was not necessary even though the number of trypomastigotes increased (Table II⇓). However, since we studied accessory activation pathways that required phorbol ester, addition of PMA was maintained in the remaining experiments.

We compared CD4⁺ and CD8⁺ T cell subsets from infected mice for their ability to potentiate T. cruzi replication in Mφ monolayers (Fig. 2⇓A). Activation of CD4⁺ T cells markedly potentiated trypomastigote growth, while activation of CD8⁺ T cells resulted in much less trypomastigote replication in Mφ (Fig. 2⇓A). However, while CD4⁺ T cells invariably exacerbated trypomastigote growth, use of activated CD8⁺ T cells gave variable results. In some experiments, CD8⁺ T cells had little effect on trypomastigotes and increased the number of noninfective epimastigotes, while in other experiments, CD8⁺ T cells actually suppressed trypomastigote growth. The reason for this variable behavior of CD8⁺ T cells is being investigated. In any case, CD8⁺ T cells were clearly distinct from CD4⁺ T cells in the ability to help trypomastigote growth.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

Characterization of the CD4⁺ T cell-mediated deleterious effect on parasite replication. A, Differential ability of CD4⁺ and CD8⁺ T cells to increase parasite replication. Infected BALB Mφ were cultured with CD4⁺ or CD8⁺ T cells purified from the same pool of splenocytes from T. cruzi-infected mice with or without anti-TCR stimulation. Viable trypomastigote number was determined after 11 days. Results are the mean and SE of duplicate cultures. The difference between CD4⁺ and CD8⁺ T cell addition is significant (p < 0.001). B, Cell contact between CD4⁺ T cells and Mφ is required for increased parasite replication. Infected Mφ were cultured with CD4⁺ T cells as described in the text, either unseparated or separated in 24-well culture inserts. Viable trypomastigote number was assessed after 7 or 9 days. Results are the mean and SE of duplicate cultures. Differences between separated and unseparated cultures are significant (p < 0.001 on day 7; p < 0.002 on day 9).

Since activated CD4⁺ T cells clustered around infected Mφ in culture, we tested the need for direct T/Mφ cell contact for exacerbation of parasite replication (Fig. 2⇑B). Infected Mφ monolayers and CD4⁺ T cells were cultured in the same compartment or separated by a cell-impermeable membrane in culture inserts. Splenic CD4⁺ T cells contained endogenous accessory cells to allow full activation when separated from Mφ. Separation of CD4⁺ T cells and Mφ markedly reduced parasite growth compared with that in unseparated cocultures (Fig. 2⇑B). These results demonstrate the need for direct cell contact between activated CD4⁺ T cells and Mφ to potentiate the growth of T. cruzi.

Modulation of CD4⁺ T cell activation by Fas-induced death regulates parasite replication in Mφ monolayers

Activation of CD4⁺ T cells from T. cruzi-infected mice with anti-TCR, but not with anti-Ly-6, results in both AICD by apoptosis and down-regulated proliferation (14, 17). We then investigated the role of CD4⁺ T cell AICD in exacerbation of parasite growth. Following coculture with infected Mφ monolayers in the presence of different stimuli, AICD was evaluated in recovered CD4⁺ T cells. Addition of anti-TCR killed a substantial proportion of the T cells, while addition of anti-Ly-6 did not result in T cell death and, instead, increased T cell viability after 1 day in culture (Fig. 3⇓A), presumably because of lymphokine-mediated rescue from spontaneous cell death. The Fas/FasL death pathway has been implicated in mediating CD4⁺ T cell AICD (16). Addition of an agonist anti-Fas mAb killed CD4⁺ T cells, and when combined with anti-Ly-6, anti-Fas abrogated the protective effect of anti-Ly-6, leading to cell death (Fig. 3⇓A). Susceptibility to both anti-TCR- and anti-Fas-mediated death was restricted to CD4⁺ T cells from infected, but not from control, mice (not shown). The same results were observed in a repeat experiment. The data indicate that AICD is preserved in CD4⁺ T cell interaction with infected Mφ and confirm that the Ly-6 activation pathway is deficient in inducing AICD. We then explored this differential behavior by evaluating the role of Fas-induced death in the control of parasite replication (Fig. 3⇓, B and C). Contrasting with the exacerbated trypomastigote growth elicited by TCR activation, Ly-6 activation of CD4⁺ T cells from infected mice did not result in appreciable enhancement of parasite replication in Mφ (Fig. 3⇓, B and C) compared with that in unstimulated T cells. Strikingly, when Ly-6 activation was combined with prior treatment with anti-Fas, a marked enhancement of parasite replication was seen, similar to that with TCR activation (Fig. 3⇓, B and C). The experiment was repeated with identical results. Treatment with a control hamster IgG mAb instead of anti-Fas did not result in enhanced parasite growth (not shown). In the absence of T cells, neither anti-Fas nor anti-Ly-6 mAbs had any effect on Mφ viability or parasite replication. Both killing and activation of host CD4⁺ T cells are required for exacerbating parasite growth, since treatment with anti-Fas alone was far less effective (Fig. 3⇓, B and C) despite inducing comparable T cell killing in the presence or the absence of anti-Ly-6 (see Fig. 3⇓A).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

Activation-induced CD4⁺ T cell death triggered by TCR or Fas, but not by Ly-6 activation, exacerbates parasite replication. A, Differential triggering of AICD by TCR and Ly-6 pathways. T. cruzi-infected Mφ were cultured with purified CD4⁺ T cells from T. cruzi-infected mice and with the indicated stimuli. The percent cell loss was measured after 20 h as indicated in the text. Results are the mean and SE of triplicate cultures. The difference between anti-Ly-6 and anti-Ly-6 plus anti-Fas groups is significant (p < 0.001). The difference between anti-Fas and anti-Ly-6 plus anti-Fas groups is not significant. B and C, Enforced Fas-mediated T cell killing exacerbates parasite replication. Infected Mφ were cultured with purified CD4⁺ T cells from T. cruzi-infected mice, treated or not with anti-Fas, in the presence or the absence of anti-Ly-6 or anti-TCR. Viable trypomastigote number was determined after 8 days (B) or 12 days (C). Results are the mean and SE of duplicate cultures. Differences between anti-Ly-6 and anti-Ly-6 plus anti-Fas groups are significant (p < 0.001). The difference between anti-Ly-6 plus anti-Fas and anti-TCR groups on day 12 is not significant.

To confirm that Fas-mediated AICD is involved in exacerbation of T. cruzi growth, a neutralizing anti-FasL mAb, specific for B6 mouse FasL allotype (22), was used to block CD4⁺ T cell AICD. In several experiments, anti-FasL mAb consistently blocked 50 to 60% of AICD in CD4⁺ T cells from T. cruzi-infected B6 mice, while an isotype control mAb had no effect (Fig. 4⇓A). Anti-FasL Ab had little or no effect on proliferation or viability of CD4⁺ T cells from control B6 mice, but consistently increased the otherwise depressed anti-TCR;CD3 proliferative responses of CD4⁺ T cells from infected mice (not shown). When added to cocultures of infected B6 Mφ and TCR-activated CD4⁺ T cells from infected B6 mice, anti-FasL mAb, but not an isotype control, markedly reduced the growth of T. cruzi trypomastigotes in culture (Fig. 4⇓, B and C) in a manner consistent with the blockade of AICD achieved in CD4⁺ T cells (Fig. 4⇓A). Blockade of parasite replication by anti-FasL was confirmed in two additional experiments. On the other hand, anti-FasL mAb had no effect on trypomastigote replication in Mφ cultured in the absence of T cells (data not shown). Together, these results indicate the involvement of CD4⁺ T cell AICD in up-regulation of parasite replication in Mφ monolayers.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

Blockade of CD4⁺ T cell AICD by anti-FasL reduces parasite replication. A, Blockade of T cell AICD by anti-FasL. Partially purified CD4⁺ T cells from T. cruzi-infected B6 mice were cultured with anti-TCR in the presence of either anti-FasL or an isotype control. After 20 h, the percent cell loss was determined. Results are the mean and SE of duplicate cultures. B and C, Anti-FasL reduces trypomastigote growth in T cell/Mφ cocultures. T. cruzi-infected B6 Mφ were cultured with highly purified CD4⁺ T cells from T. cruzi-infected B6 mice in the presence of anti-TCR and with anti-FasL or an isotype control. Viable trypomastigote number was determined after 6 days (B) or 8 days (C). Results are the mean and SE of duplicate cultures. Differences between anti-FasL and the isotype control are significant (p < 0.001).

Finally, since IFN-γ is critically involved in the control of parasite replication in Mφ, it was of interest to investigate the effect of Fas-mediated death in IFN-γ production by CD4⁺ T cells from T. cruzi-infected mice. To avoid the coincidence of IFN-γ secretion and the onset of AICD (25), T cells were first treated overnight with either anti-Fas or a control hamster mAb and subsequently stimulated with anti-TCR, and resulting supernatants were assayed for IFN-γ content (Fig. 5⇓). Prior overnight treatment with anti-Fas results in killing of 25 to 30% of CD4⁺ T cells from infected hosts (see Fig. 3⇑A), but eliminated 70.6% of the IFN-γ produced upon subsequent TCR activation (Fig. 5⇓). Killing with anti-Fas also eliminated 67.0% of IFN-γ production upon Ly-6 activation (not shown). These results indicate that although they are a minor population, Fas-susceptible CD4⁺ T cells comprise the majority of IFN-γ-producing cells or contain most of IFN-γ high producer CD4⁺ T cells in T. cruzi-infected hosts.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 5.

Fas-mediated killing ablates most of IFN-γ production upon activation. Partially purified CD4⁺ T cells from T. cruzi-infected BALB mice were cultured overnight with either anti-Fas or a hamster isotype control. After 20 h, cells were either left unchanged or were stimulated with anti-TCR mAb as indicated. Supernatants were collected 48 h after TCR stimulation and assayed for IFN-γ content by sandwich ELISA. Results are the mean and SE of duplicate cultures. The difference between anti-Fas and isotype control is significant (p < 0.005) for TCR-activated groups.