CD40 Ligand Is Not Essential for Induction of Type 1 Cytokine Responses or Protective Immunity after Primary or Secondary Infection With Histoplasma capsulatum

CD40L^−/− mice generated on C57BL/6J × 129/J background (27) were obtained from The Jackson Laboratory (Bar Harbor, ME). As a control, C57BL/6 × 129(F2), also purchased from The Jackson Laboratory, were used and designated as CD40L^+/+ mice. In some experiments, CD40L^−/− mice obtained from Immunex (Seattle, WA) were bred on a C57BL/6 background (greater than six generations). Virus-free female C57BL/6 mice purchased from Division of Cancer Treatment, National Cancer Institute (Frederick, MD) were used as controls for these experiments. In all experiments, mice were between 5 and 10 wk of age. Mice were inoculated intravenously in 0.5 ml sterile PBS with varying doses of H. capsulatum yeast cells. In one experiment, CD40L^−/− mice were challenged with 10⁵ metacyclic promastigotes L. major (WHOM/IR/-/173) in their hind footpads as previously described (28). Weekly footpad swelling measurements were recorded using a caliper.

HBSS (Biofluids, Inc., Rockville, MD) was used as a wash medium. Complete medium: RPMI 1640 (Biofluids, Inc.) supplemented with 10% fetal bovine serum (Biofluids, Inc.), penicillin (100 U/ml), streptomycin (100 μg/ml), l-glutamine (2 mM), sodium pyruvate (1 mM), and 2-ME (0.05 mM) was used for culturing spleen cells.

The yeast phase of H. capsulatum (strain GS-57) was used in all experiments, and quantitation of H. capsulatum was performed as previously described (26). In brief, spleens from mice infected with H. capsulatum and/or treated with various cytokine antagonists were killed at various times after infection. In most experiments, three individual spleens from each group were quantitated for CFUs. In some experiments, one-third of each spleen from two or three individual animals in each group were combined and homogenized in a sterile mortar using PBS to prepare a 1:10 wt/vol suspension. 10-fold dilutions in PBS were plated in duplicate at 0.05 ml/plate on BHI-SAGC medium and incubated for 7 d at 30°C. Colonies were enumerated and the counts were recorded as CFUs.

Most antibodies were purified from ascites by ammonium sulfate precipitation. Rat anti–mouse IFN-γ mAb (XMG1.2; reference 29), anti-CD4 (GK1.5; reference 30), and anti-CD8 (2.43; reference 31) were used to neutralize IFN-γ and to deplete CD4⁺ and CD8⁺ T cells, respectively. Purified mAbs against murine IL-12 (C17.8; reference 32) were obtained from Dr. Giorgio Trinchieri (Wistar Institute, Philadelphia, PA) and have been shown to be effective in neutralizing IL-12 in vivo. Mice were injected with anti–IFN-γ (1 mg) intraperitoneally at the time of primary infection. Anti-CD4 and anti-CD8 antibodies were injected 3 d before, at the time of, and 5–7 d after infection. This treatment resulted in a >95% depletion of CD4⁺ or CD8⁺ T cells from spleen at 1 wk after infection, as assessed by FACS^®.

Cytokine mRNA levels were determined by semiquantitative reverse transcription PCR techniques as previously described (26). In brief, total RNA was isolated from spleen cells by resuspending in RNAzol B (Tel-Test, Friendswood, TX) and recovering the aqueous phage after addition of chloroform. RNA was precipitated with alcohol and resuspended in RNase-free H₂O. Total RNA (1 μg) was reverse transcribed by Moloney murine leukemia virus reverse transcriptase (GIBCO BRL, Gaithersburg, MD). The reaction mixture was then diluted 1:8, and 10 μl was used for specific semiquantitative amplification of cytokine mRNA with Taq DNA polymerase (Promega, Madison, WI) and specific cytokine sense and antisense primers. The number of amplification cycles was as follows: 24 (hypoxanthine phosphoribosyltransferase), 25 (IFN-γ), 29 (IL-12 p40), and 33 (NO and TNF-α). Southern transfers of PCR products were subsequently probed with internal cytokine-specific oligonucleotides and visualized using the ECL chemiluminescent detection system (Amersham Corp., Arlington Heights, IL).

Statistical evaluation of differences between means of experimental groups was done by analyses of variance and multiple Student's t tests. The log-rank was used for statistical analysis of mortality. A value of P <0.05 was considered to be significant.

Recent work by several groups has shown that CD40^−/− or CD40L^−/− mice were susceptible to Leishmania infection due to diminished production of type 1 cytokines (17–19). In other studies, mice infected with H. capsulatum (6 × 10⁵) and treated with anti–IL-12 or anti– IFN-γ have accelerated mortality, demonstrating a critical role for these cytokines in primary infection to H. capsulatum (24, 25). Thus, the role of CD40L in mediating protective immunity and the generation of type 1 cytokine production in response to H. capsulatum infection was evaluated. As shown in Fig. 1,A, the rate at which CD40L^−/− mice succumbed to infection (14.67 ± 3.82 d) was not different than that for CD40L^+/+ mice (14.75 ± 3.84 d). By contrast, in the same experiment, CD40L^−/− mice infected with L. major were found to be susceptible to infection compared with control CD40L^+/+ mice (Fig. 1 B). These data suggest that CD40–CD40L stimulation, although essential for induction of type 1 responses to L. major (17–19), may not be required for eliciting type 1 cytokine responses to H. capsulatum.

In addition to the experiments shown above using CD40L^−/− mice bred several generations (greater than six) onto a C57BL/6 background, a series of experiments using CD40L^−/− mice on a C57BL/6 × 129 background were initiated. It should be noted that in these experiments, the control CD40L^+/+ mice are C57BL/6 × 129(F2), approximating the background of the CD40L^−/− mice. As shown in Fig. 2 in data combined from two independent experiments, infection of CD40L^−/− mice with H. capsulatum (6 × 10⁵) showed a modest but not statistically significant increase in the rate of mortality (35.6 ± 20.4 d) compared with the control CD40L^+/+ mice (41.5 ± 21.3 d). In addition, all mice infected with a sublethal dose (6 × 10⁴ or 6 × 10³) survived infection. Thus, these data support a CD40L-independent role in protective immunity.

To correlate whether the fatal outcome shown above was due to an increase in the infectious burden of H. capsulatum, quantitative cultures were set up from spleen cells at various times after infection. In the same experiment as shown in Fig. 2, spleen cells of mice were harvested 7 d after infection and CFUs for H. capsulatum were determined (Table 1). CD40L^−/− mice infected with 6 × 10⁵ yeast cells had a twofold increase in infectious burden at 7 d after infection compared with control mice. Furthermore, CD40L^−/− mice infected with lower doses (6 × 10⁴, 6 × 10³) had almost identical CFUs for H. capsulatum compared with control mice. These data further support the contention that CD40L is dispensable for this infection.

In the studies examining cytokine production in CD40^−/− or CD40L^−/− mice in response to leishmanial infection, it was clear that there was a marked deficiency in production of IFN-γ (17–19), IL-12 (17, 19), and nitric oxide (NO) (18). To determine whether any qualitative or quantitative changes in cytokine production occurred in CD40L^−/− mice compared with control CD40L^+/+ mice after H. capsulatum infection, mRNA expression for several cytokines previously found to be important in regulating primary immunity to H. capsulatum was assessed by semiquantitative PCR at various time points after infection. As shown in Fig. 3, mRNA for IFN-γ, IL-12 p40, and TNF-α were expressed from spleen cells of CD40L^−/− and CD40L^+/+ mice 7 and 20 d after infection with any of the doses tested. In addition, mRNA for NO was expressed from both types of mice 7 d after infection. As a control, there was minimal mRNA expression for any of the cytokines from spleen cells of uninfected mice prepared at day 7 after infection. Thus, these data confirm that type 1 cytokines are induced in CD40L^−/− mice after infection with H. capsulatum. Finally, it should be noted that similar amounts (4–7 ng/ml) of IL-12 protein (p40 + p70) were detected from serum 3 and 7 d after infection from both CD40L^−/− and CD40L^+/+ mice (data not shown).

The data shown above suggest that CD40L^−/− and CD40L^+/+ mice induce comparable type 1 immune responses that might provide some immunity against a lethal challenge with H. capsulatum. Functional evidence showing that CD40L^−/− mice are capable of producing IL-12 and IFN-γ in vivo was demonstrated by treating mice with neutralizing antibodies against IL-12 or IFN-γ at the time of infection and assessing their outcome. As shown in Fig. 4, CD40L^−/− mice treated with either anti–IL-12 (8.6 ± 1.3 d) or anti–IFN-γ (8.8 ± 1.1 d) had accelerated mortality compared with mice infected with only CD40L^−/− (31.0 ± 15.1 d; P <0.05). Similar results were seen from CD40L^+/+ mice after infection. Neutralization of IL-12 or IFN-γ also resulted in a 3–10-fold increase in infectious burden of H. capsulatum compared with infected-only mice (P <0.001), providing further evidence that CD40L^−/− mice are capable of making IL-12 and IFN-γ in vivo.

Although CD40L^−/− mice had similar outcomes to control mice infected with a lethal dose of H. capsulatum, most of them still succumbed to infection. By contrast, as demonstrated in Fig. 2, CD40L^−/− mice infected with 6 × 10⁴ yeast cells were able to control infection, suggesting that CD40L is not essential for protective immunity at this infective dose. Since we had shown in a previous study that IFN-γ^−/− mice infected with 6 × 10⁴ or 6 × 10³ yeast cells succumbed to infection (26), the relative dispensability of CD40L compared with IFN-γ in developing effective immunity against primary infection with a sublethal dose of H. capsulatum (6 × 10⁴) was assessed. As shown in Fig. 5, all CD40L^−/− and CD40L^+/+ mice survived infection with 6 × 10⁴ yeast cells, whereas all mice treated with anti–IFN-γ had a fatal outcome with accelerated mortality. These experiments show conclusively that effective primary immunity to H. capsulatum occurs in an IFN-γ–dependent, CD40L-independent manner.

Both CD4⁺ and CD8⁺ T cells have been shown to be important in primary immunity against H. capsulatum (33, 34). To assess the role of CD4⁺ and CD8⁺ T cells in CD40L^−/− mice infected with H. capsulatum, animals were treated with antibodies against CD4⁺ and CD8⁺ T cells before, at the time of, and after infection to ensure depletion (>95% by FACS^®). As shown in Fig. 6 in data combined from two independent experiments, depletion of either CD4⁺ (12.8 ± 0.9 d) or CD8⁺ T cells (12.3 ± 0.05 d) caused accelerated mortality compared with infected controls (34.8 ± 19.6 d). This increase in mortality was associated with a three- to fourfold increase in CFUs for H. capsulatum from spleen cells of mice depleted of either CD4⁺ or CD8⁺ T cells (P <0.001). Thus, in the absence of CD40L, CD4⁺ or CD8⁺ T cells play a critical role in protective immunity.

One final aspect of these studies was to examine the requirement for CD40L in a memory or secondary immune response to H. capsulatum. To assess the role of CD40L^−/− in the maintenance of immunity after secondary infection, CD40L^−/− mice were initially infected with a sublethal dose of H. capsulatum (6 × 10⁴) and then reinfected 3–4 wk later with a lethal dose (6 × 10⁵). As shown in Fig. 7 in data combined from two independent experiments, all CD40L^−/− mice reinfected with a lethal dose (6 × 10⁵) survived and remained healthy up to 120 d after infection. As a control, a majority of the CD40L^−/−mice undergoing primary infection (6 × 10⁵) at the same time succumbed within 20–30 d, consistent with the data shown in the previous figures.

To determine whether effective immunity to a secondary challenge was due to control of H. capsulatum in vivo, the infectious burden of H. capsulatum was assessed from spleen cells at various time points after reinfection. Mice originally infected with 6 × 10⁴ yeast cells had low but detectable amounts of H. capsulatum at the time of reinfection (day 0) in one of the experiments (Table 2, Exp. 1) and no detectable amounts in the other experiment (Table 2, Exp. 2). When assessed 7 d after reinfection, previously infected CD40L^−/− mice had a 3-log reduction (P <0.001) in H. capsulatum compared with CD40L^−/− mice undergoing primary infection (Table 2, Exp. 2) at the same time. Moreover, in neither experiment was there any H. capsulatum detected from spleen cells at 60 or 90 d after reinfection, demonstrating that CD40L^−/− is not essential for the development of sterilizing immunity after reinfection.

CD40L–CD40 interactions have a central role in the induction of humoral and cellular immune responses (35, 36). With regard to its effects on the cellular immune response, CD40L–CD40 induction of inflammatory cytokines and costimulatory molecules from macrophages and dendritic cells has been shown to enhance the cellular immune response with production of type 1 cytokines (35). Since IL-12–dependent production of IFN-γ was shown to be essential for protective immunity after primary infection to the intracellular fungi H. capsulatum (24), it was of interest to determine the requirement for CD40L in the generation of type 1 cytokine responses after primary and secondary infection with H. capsulatum. The initial experiments using CD40L^−/− mice bred onto a C57BL/6 background (>6 generations) showed that they did not have accelerated mortality or increased fungal burden (data not shown) compared with CD40L^+/+ mice (Fig. 1). In several additional experiments using CD40L^−/− mice on a C57BL/6 × 129 background, there did appear to be a modest increase in the rate of mortality (Figs. 2 and 4) and the fungal burden (Table 1) compared with the control CD40L^+/+ mice in response to a lethal dose (6 × 10⁵ yeast) of H. capsulatum. However, it should be noted that CD40L^−/− or CD40L^+/+ mice infected with a sublethal amount of H. capsulatum (6 × 10⁴) had identical outcomes. Furthermore, the data showing that mRNA for IL-12, IFN-γ, TNF-α, and NO are similar between the CD40L^−/− and CD40L^+/+ mice, combined with the fact that in vivo neutralization with anti–IFN-γ or anti–IL-12 caused accelerated mortality and increased fungal burden, provides strong evidence that CD40L is not required for the induction of type 1 responses after infection with H. capsulatum. Finally, the demonstration that all CD40L^−/− mice survived infection with a sublethal dose of H. capsulatum (Fig. 5) but all mice treated with an anti–IFN-γ antibody had accelerated mortality and increased fungal burden provided conclusive evidence that there is a CD40L-independent pathway for IFN-γ production.

The ability of CD40L^−/− mice to generate functional Th1 responses in vivo is in contrast to the studies previously alluded to in which CD40^−/− or CD40L^−/− mice had a striking deficiency in production of IL-12 or IFN-γ after infection with L. major or Leishmania amazonensis (17–19). Several potential mechanisms could account for this failure of these deficient mice to develop functional type 1 responses in response to leishmanial infection. One is that L. major is a relatively poor direct inducer of IL-12 production or other inflammatory mediators compared with other infectious pathogens. This is supported by previous studies showing that Leishmania promastigotes are able to evade or inhibit IL-12 production (37, 38). In data not shown, we found that peritoneal macrophages from CD40L^−/− mice produced comparable amounts of IL-12, TNF-α, and NO in vitro as CD40L^+/+ mice in response to Staphylococcus aureus Cowan strain, Toxoplasma gondii, and H. capsulatum; however, there was no induction of these cytokines in response to L. major promastigotes (data not shown). These data are consistent with a recent report by DeKruyff et al. showing that heat-killed Listeria monocytogenes induces IL-12 production in a CD40L-independent manner in vitro (20). In addition, the demonstration that SCID mice infected with H. capsulatum (39), L. monocytogenes (40), or T. gondii (32) all had accelerated mortality when treated at the time of infection with anti–IL-12 provides additional evidence that these pathogens can induce IL-12 directly or at least in the absence of T cells. Taking these data together, we would conclude that, in addition to H. capsulatum, L. monocytogenes and T. gondii would also induce IL-12 in vivo independently of CD40L. Finally, it should be noted that although L. major may be a relatively poor direct inducer of IL-12, mice on a resistant C57BL/6 background develop effective immunity after infection with L. major in an IL-12–dependent manner, suggesting additional factors to explain the enhanced susceptibility of CD40L^−/− mice to L. major (see below).

A second mechanism to explain why CD40L^−/− mice fail to induce IL-12 in response to L. major infection is that this experimental model is highly reliant on MHC class II– dependent CD4 responses for primary immunity (41–43). Furthermore, based on in vivo (1) and in vitro (6, 20) evidence showing that CD4⁺ T cell–mediated production of type 1 cytokines is CD40L dependent in response to protein antigens, it is possible that antigens that require CD4 and/or are relatively poor direct inducers of IL-12 (i.e., L. major) would require CD40L for a functional IL-12 response. This relative CD4 dependence in the Leishmania model would be contrasted to experimental models of listeriosis, toxoplasmosis, and histoplasmosis in which both CD4⁺ and CD8⁺ T cells have a role in primary immunity (see below). Moreover, L. monocytogenes, T. gondii, and H. capsulatum are all potent inducers of IFN-γ from NK cells through induction of IL-12. Thus, in these latter models, CD8⁺ T cells or NK cells may provide a sufficient amount of IFN-γ or other proinflammatory mediators that could help control infection directly and/or potentiate further production of IL-12 by a positive feedback mechanism (44, 45).

As noted above, there is evidence that CD8⁺ T cells also play an important role in primary immunity against a variety of intracellular pathogens including T. gondii (46), Mycobacterium bovis (47), Mycobacterium tuberculosis (48), and H. capsulatum (34). It has been shown that the ability of CD8⁺ T cells to influence immunity is through at least two independent mechanisms. For secondary Listeria infection, CD8⁺ T cells have been shown to protect mice through a cytolytic mechanism independent of IFN-γ (49, 50). By contrast, in murine models of L. monocytogenes (50), M. tuberculosis (51, 52), T. gondii (53), or H. capsulatum infection (Zhou, P., manuscript in preparation), primary immunity is maintained in the absence of perforin- or granzyme-mediated cytolysis. Since IFN-γ is required for effective primary immunity to all of these infections, these data suggest that IFN-γ produced from CD8⁺ T cells is responsible for the effector function at this phase of the response.

In the experiments reported here, CD40L^−/− mice depleted of either CD4⁺ or CD8⁺ T cells had similar outcomes in terms of accelerated mortality and increased fungal burden after primary infection with H. capsulatum. Explanations for these data follow. (a) There is a quantitative threshold for cytokine (i.e., IFN-γ) production necessary for effective primary immunity in the absence of CD40L requiring both CD4⁺ and CD8⁺ T cells. In preliminary data, CD40L^−/− mice depleted of CD4⁺ T cells at the time of infection had markedly diminished production of IFN-γ (data not shown). (b) Depletion of either CD4⁺ or CD8⁺ T cells changes the qualitative cytokine response. This is consistent with a striking enhancement of the Th1 counterregulatory cytokine IL-10 noted from mice depleted of CD4⁺ or CD8⁺ T cells (data not shown). (c) There are recent reports showing that CD4⁺ T cells are required for CD8⁺ T cell function (54, 55). This dependence on CD4⁺ T cells could occur through modification of the antigen presenting cell (i.e., CD40L-inducing activation and/or cytokine production) or to provide a cytokine-rich environment (i.e., IL-2) for the CD8⁺ T cells. These explanations may relate to the studies reported here in which the absence of both CD4⁺ T cells and CD40L may not provide help sufficient for CD8⁺ T cell activation. Experiments are now in progress to determine the immune mechanism by which CD4 and CD8 depletion worsens the course of infection in these mice.

In a previous study, we showed that effective primary immunity to systemic infection requires a coordinated immune response requiring many factors including IL-12, IFN-γ, TNF-α, or NO. By contrast, none of these factors alone, including IFN-γ, were required for the maintenance of immunity after secondary challenge to H. capsulatum (26); however, treatment of mice with both anti–IFN-γ and anti–TNF-α at the time of reinfection did result in a fatal outcome. The only studies of the role of CD40L in memory immunity were done in experimental murine models of viral infection. In these studies, CD40L^−/− mice had relatively normal primary CTL responses but had markedly diminished memory responses (14). In this report, CD40L^−/− mice initially infected with a sublethal dose of H. capsulatum and then infected 3 wk later with a lethal dose remain alive more than 90 d after infection with no detectable H. capsulatum from spleen cells. These results demonstrate that CD40L is not required for effective secondary immunity.

To summarize, L. major may in fact be one of the few intracellular infections requiring CD40–CD40L stimulation for the induction of functional type 1 cytokine responses. Infection with other pathogens such as L. monocytogenes, T. gondii, and M. tuberculosis, which directly induce IL-12 and/or have a role for CD8⁺ T cells in immunity, would not show increased susceptibility in the absence of CD40L. This is supported by the observations in studying patients with hyper-IgM syndrome (56). In a large review of such patients, a minority of patients were reported to have increased susceptibility to opportunistic infections associated with T cell deficiencies such as Pneumocystis carinii (<10%) and Cryptosporidium (<2%), whereas there were no cases described for other infections mediated by type 1 cellular immune responses such as T. gondii, M. tuberculosis, or H. capsulatum. These latter findings would be consistent with the ability of these infectious pathogens to induce type 1 cytokine responses by the mechanisms listed above. In addition, since exogenous B7 costimulation has been shown to restore IFN-γ production in CD40L^−/− mice (12, 13), it is possible that these various pathogens may differ in their capacity to activate accessory cells in vivo, providing a mechanism to activate T cells independently of CD40L–CD40.

Finally, with regard to autoimmune disease, since CD40L– CD40 stimulation has been shown to be critical for both IL-12 production and T cell activation in response to protein antigens in vivo, inhibition of CD40L–CD40 may be a potent treatment for certain organ-specific autoimmune diseases. This is supported by murine models of experimental autoimmune encephalitis and inflammatory colitis in which treatment with anti–IL-12 (57, 58) or anti-CD40L (59, 60) was shown to ameliorate disease. One cause of concern from chronic treatment with either of these inhibitors would be increased susceptibility to intracellular infection. The ability to sustain and generate an effective immune response to certain intracellular pathogens independently of CD40L may lessen this potential increase in susceptibility to these particular infections by anti–CD40L–CD40 treatment. By contrast, treatments directed at inhibiting IL-12 while ameliorating autoimmune disease could still increase vulnerability to the aforementioned intracellular infections.

We thank Brenda Preuninger for technical support and Brenda Rae Marshall for editorial assistance.