IFN-γ Is Necessary But Not Sufficient for Anti-CD40 Antibody-Mediated Inhibition of the Th2 Response to Schistosoma mansoni Eggs¹

Chronic infection with Schistosoma mansoni results in granulomatous disease that may eventually lead to fibrosis, occlusion of the presinusoidal vascular bed, and fatal portal hypertension. Granulomas form around viable eggs trapped in the small vessels of the liver and intestine of their mammalian host. Granuloma development follows the appearance of Th2-dominant immune responses characterized by high levels of IL-4, IL-5, and IL-13 production (1). There is increasing evidence that Th2 cytokines are necessary for granuloma development and that immune deviation of the egg-induced immune response toward a Th1-polarized phenotype attenuates disease. Specifically, vaccination of mice with rIL-12 and eggs prevented the subsequent development of liver pathology during murine schistosomiasis by IFN-γ-dependent mechanisms (2, 3, 4). Consequently, the characterization of other immune interventions capable of abrogating S. mansoni egg-induced Th2 development may indicate potential therapeutic approaches while providing mechanistic insights into this important disease.

Because egg laying is delayed by several weeks in the mouse model of natural infection, the early immunologic events associated with Th2 cell differentiation to egg Ags are examined more conveniently by injecting S. mansoni eggs into the footpads of mice (5). This reproducibly triggers Ag-specific Th2 responses within the popliteal lymph node that peak within 1 wk of injection and that can be modulated by simultaneous administration of recombinant cytokines or neutralizing mAbs specific for cytokines or costimulatory molecules (6, 7). This model is therefore well suited for the analysis of the early events associated with induction and regulation of Th2 subset differentiation, as well as for examining potential therapeutic interventions. For example, administration of rIL-12 decreased egg Ag-induced IL-4 and IL-5 production (6), consistent with the well-characterized Th1-polarizing effects of IL-12 in models of cellular immunity (8). In the following studies, we test whether Ab-mediated activation of CD40 on accessory cells at the time of egg injection would similarly redirect the emerging immune response away from the Th2 cytokine phenotype normally induced in this model.

Although CD40 was originally described as a receptor protein on B cells that activates B cell growth and differentiation in support of Ab production (9), the engagement of macrophage and dendritic cell CD40 with the CD154 counterligand on T cells induces a coordinated set of responses important in T cell differentiation. These include the production of immunoregulatory cytokines, such as IL-12, and the up-regulation of costimulatory proteins and Ag-presenting capability (10, 11, 12). The complexity of these responses suggests that experimental activation of CD40 by agonistic Abs or soluble CD154 may affect T cell differentiation by mechanisms that extend beyond those mediated by IL-12 alone. The immunoregulatory properties of in vivo CD40 activation have been demonstrated by the Th1-promoting effects of anti-CD40 mAb during murine leishmaniasis in Th2-predisposed, susceptible BALB/c mice and in the disruption of experimentally induced Th2-dependent neonatal transplantation tolerance to alloantigen (13, 14). Similarly, mice treated with DNA encoding the soluble CD40L trimer become highly resistant to tumor metastasis and to Leishmania major infection (15).

On the basis of these studies, we hypothesized that experimental activation of CD40 at the time of S. mansoni egg injection would switch the developing Ag-specific immune response from a Th2 to a Th1 cytokine phenotype. To test this hypothesis, eggs were injected into the footpad of mice concurrent with administration of agonistic anti-CD40 mAb FGK45, which activates CD40 in vitro in a manner similar to that observed with the natural ligand CD154 (16, 17, 18). In this well-characterized model for Th2 CD4⁺ T cell development, anti-CD40 treatment strongly suppressed the development of IL-4-, IL-5-, and IL-13-secreting immune responses and instead promoted the development of CD4⁺ T cell-dependent IFN-γ synthesis in vivo. In parallel experiments using IFN-γ and IL-12 (p40) knockout mice, anti-CD40 was unable to suppress IL-4, IL-5, and IL-13 production in response to S. mansoni egg Ags, despite the presence of strong, anti-CD40-inducible IFN-γ responses in IL-12 knockout mice.

Six- to 8-wk-old C57BL/6J and IL-12 p40^−/− (C57BL/6J IL-12b^tm/Jm) (19) mice were obtained from The Jackson Laboratory (Bar Harbor, ME). Six- to 8-wk-old IFN-γ-deficient mice, C57BL/6 IFN-γ^tm/tls (20), were originally obtained from Genentech (South San Francisco, CA) and bred locally. All mice were housed in microisolator cages at Case Western Reserve University in the animal resource center.

Six- to 8-wk-old female C57BL/6J, IFN-γ^−/−, or IL-12 p40^−/− mice were injected i.p. with 250 μg of an activating CD40 mAb or with a control rat IgG (Sigma, St. Louis, MO), followed by inoculation with 5000 S. mansoni eggs in each hind footpad, as described (7). Mice were sacrificed on day 7 by cervical dislocation. FGK45 and soluble recombinant CD40 ligand have similar in vitro abilities to stimulate IL-12 p40 production from cultured LNC³ (16), and FGK45 stimulates CD40 in vivo (17).

Mice were bled and sacrificed 7 days after egg injection. Single cell suspensions of popliteal LNC were made by forcing the lymph nodes through a mesh screen. The cells were washed extensively, then cultured at 5 × 10⁶/ml in DMEM (BioWhittaker, Walkersville, MD) containing 10% FBS (HyClone, Logan, UT), 2 mM glutamine (BioWhittaker), 10 mM HEPES (Sigma), 2-ME (Life Technologies, Grand Island, NY), and antibiotics (BioWhittaker) in 96-well tissue culture plates (Corning, Corning, NY). The cells were incubated with 10 μg/ml αIL-4R mAb (Genzyme, Cambridge, MA) to prevent sequestration of IL-4 (21), along with medium alone or 25 μg/ml soluble egg Ag, prepared as previously described (22), in a 37°C incubator with 5% CO₂. Culture supernatants were removed 72 h later for analysis of cytokine content by ELISA. Remaining cells from LNC preparations were resuspended in STAT60 (TelTest B, Friendswood, TX) and frozen at −70°C for later RNA extraction.

RNA was extracted from STAT60 tissue homogenates with chloroform, precipitated with isopropanol, and rinsed with ethanol, as per manufacturer’s protocol. cDNA was made, and cytokine-specific DNA products were amplified using Taq polymerase, as previously described (7). The PCR product was transferred to positively charged nylon membranes (Boehringer Mannheim, Indianapolis, IN), prehybridized, then probed with digoxigenin-labeled cytokine-specific oligonucleotides. After incubating with antidigoxigenin-conjugated alkaline phosphatase, the blots were developed with Lumi-Phos (Boehringer Mannheim) and exposed to Kodak Biomax MR film (Eastman Kodak, Rochester, NY). Thirty-minute exposures are shown for all blots.

Cytokine levels in culture supernatants and in serum were determined by two-site ELISA, as previously described (7). IL-13 levels were determined using the Quantikine kit (R&D Systems, Minneapolis, MN). IL-12 p40 and p70 levels were measured by ELISA, using previously reported reagents and techniques (23). Concentrations were calculated using recombinant cytokines as standards.

CD4⁺ cells were selected with immunomagnetic beads coated with anti-CD4, according to the manufacturer’s instructions (Miltenyi Biotec, Auburn, CA). Total LNC, CD4⁺ LNC, and CD4⁻ LNC populations were collected, and RNA obtained by phenol/chloroform extraction. RT-PCR analysis of HPRT and IFN-γ levels was performed. Densitometry data were obtained using the ScanAnalysis program (Biosoft, Ferguson, MO). Frequencies of IFN-γ-secreting cells from total LNC and CD4-depleted LNC of anti-CD40-treated mice were determined by enzyme-linked immunospot, as previously described (7).

Eosinophils were detected according to the method described by Pearlman et al. (24, 25). Briefly, footpads were fixed in 10% Formalin for 24 h and processed in a Tissue-Tek VIP tissue processor. Five-micrometer sections were immunostained with rabbit antisera to murine eosinophil MBP 18 kindly provided by Drs. Kirsten Larson and Jamie Lee at the Mayo Clinic (Scottsdale, AZ). Sections were incubated at room temperature in a humid chamber for 2 h with αMBP (diluted 1/5000) in 0.05 M TBS (pH 7.6) containing 1% FCS. After washing, sections were incubated for 30 min with biotinylated goat anti-rabbit Ig (Dako, Carpenteria, CA) diluted 1/200 or prediluted Rat Link (BioGenex, San Ramon, CA), washed, and incubated with prediluted alkaline phosphatase-conjugated streptavidin (BioGenex). Positive reactivity was detected using Vector Red Substrate containing 12 mg levamisole (Sigma), followed by counterstaining with modified Harris’ hematoxylin (Richard-Allen, Kalamazoo, MI).

Cytokine values from experimental vs control groups were compared using a Student’s t test, with p < 0.05 being considered statistically significant.

To test whether anti-CD40 decreases the Th2 response to S. mansoni eggs, we injected anti-CD40 i.p. into C57BL/6J mice at the same time they were inoculated in the footpad with 5000 eggs. Mice were sacrificed 7 days later, and cytokine levels were measured in cultures of popliteal LNC and in the serum. LNC of mice injected with eggs alone generated abundant IL-4, IL-5, and IL-13 in response to egg Ag. With the exception of IL-13, these cytokines were also secreted spontaneously. In contrast, mice treated with anti-CD40 mAb at the time of egg inoculation were unable to generate detectable levels of either spontaneous or Ag-inducible IL-4, IL-5, and IL-13 in culture (Fig. 1,A). Conversely, the production of IFN-γ, which was undetectable in egg-injected control mice, increased dramatically in anti-CD40-treated mice (Fig. 1,A). IL-12 p40 levels were also increased more than 5-fold by prior CD40 activation. Anti-CD40-treated mice similarly demonstrated increased levels of IL-12 p40, IL-12 p70, and IFN-γ in the serum on day 7 (Fig. 1 B).

To determine whether these anti-CD40-dependent effects on Th2 and Th1 cytokine production were present in vivo, we examined steady state levels of cytokine mRNA present in the lymph node at 7 days using comparative RT-PCR analysis. As shown in Fig. 2, anti-CD40-treated mice expressed markedly reduced levels of IL-4 mRNA and reciprocally increased levels of IFN-γ mRNA. These data demonstrate that anti-CD40 differentially regulates the development of in vivo Th2 and Th1 cytokine responses to S. mansoni egg injection. Although inhibitory effects on IL-4-producing responses had been described in other models (13, 14, 15), the current findings extend the regulatory effects of anti-CD40 to the inhibition of IL-5 and IL-13 synthesis.

To determine whether anti-CD40-induced cytokine modulation was associated with changes in the eosinophilic predominance of the subdermal cellular response elicited by S. mansoni eggs, we immunostained paraffin-embedded footpad sections with Abs specific for the eosinophil granule protein, MBP. Treatment of mice with anti-CD40 markedly decreased the extent of tissue eosinophilia at the site of egg injection compared with control mice (Fig. 3). The average number of eosinophils per high power field decreased from 124 ± 19 in control mice to 4.56 ± 0.62 in mice treated with anti-CD40 (p = 0.003). These findings are consistent with the generalized inhibition of Th2 activities resulting from anti-CD40 treatment, including suppressed synthesis of the eosinophil growth factor IL-5 (26).

The role of IFN-γ in the anti-CD40-induced down-regulation of S. mansoni egg-induced Th2 response was examined in C57BL/6 IFN-γ^−/− mice. Compared with the strong suppression of Th2 cytokine responses present 7 days after anti-CD40 treatment of egg-injected wild-type C57BL/6 mice, Ag-induced IL-4 and IL-5 levels were unaffected by anti-CD40 in IFN-γ^−/− mice (Fig. 4,A). Although treatment of IFN-γ^−/− mice with anti-CD40 induced spontaneous IL-12 p40 production from cultured popliteal LNC (Fig. 4,A), the addition of egg Ag repeatedly suppressed IL-12 p40 production in lymph node cultures from both IFN-γ^−/− and wild-type mice previously treated with anti-CD40. The serum of IFN-γ^−/− mice showed small, but significant increases in circulating IL-12 p40 following anti-CD40 treatment that was 3- to 4-fold reduced relative to anti-CD40-treated wild-type mice (Fig. 4 B). These data show that the presence of IFN-γ is necessary for anti-CD40-induced Th2 cytokine inhibition and critical for the optimal synthesis of IL-12 p40 in response to CD40 stimulation.

We next examined whether IL-12 mediates anti-CD40-induced suppression of egg Ag-specific Th2 responses using C57BL/6 mice with a targeted mutation of the IL-12 p40 subunit gene. These mice are consequently unable to produce bioactive IL-12 p35/p40 heterodimer. Compared with inoculation with eggs alone, the administration of anti-CD40 in IL-12^−/− mice failed to suppress IL-4 and IL-5 production (Fig. 5,A). However, treatment with anti-CD40 induced levels of IFN-γ in IL-12^−/− mice that were comparable with those observed in wild-type mice (Fig. 5 B). These results indicate that IFN-γ in the absence of IL-12 is incapable of mediating anti-CD40-induced suppression of Th2 production.

To determine the cellular source of anti-CD40-induced IFN-γ, we positively selected CD4⁺ cells from anti-CD40-treated wild-type and IL-12^−/− mice and examined IFN-γ mRNA levels. Densitometry analysis of RT-PCR data showed expression of IFN-γ mRNA in the CD4⁺-enriched population of LNC from anti-CD40-treated C57BL/6 mice, while IFN-γ mRNA expression was decreased 3-fold in the CD4-depleted population (Fig. 6). Furthermore, depletion of CD4⁺ cells from LNC of anti-CD40-treated mice resulted in a 90% decrease in the frequency of IFN-γ-secreting cells (360 ± 28 vs 15 ± 8.6, data not shown). Similarly, CD4⁺ cells were the primary source of IFN-γ production by LNC of anti-CD40-treated IL-12^−/− mice. Depletion of CD4⁺ cells decreased IFN-γ message levels by 10-fold (Fig. 6). These data show that anti-CD40 promotes the development of CD4⁺ cells producing IFN-γ in the absence of endogenous IL-12.

The central finding of these studies is that anti-CD40 treatment blocks the ability of draining LNC to produce IL-4, IL-5, and IL-13 following footpad injection with S. mansoni eggs. Previous studies in mouse models of L. major infection and transplantation tolerance also demonstrated CD40-induced suppression of IL-4 production (13, 14). The wider range of Th2-associated cytokines affected by anti-CD40 in our studies indicates a generalized disruption of Th2 immune differentiation and not the selective suppression of individual cytokine responses. This is notable given the mutual independence of IL-4, IL-13, and IL-5 development and function in the setting of several well-defined Th2 immunopathologies. For instance, IL-13 and IL-4 are separately capable of inducing pulmonary granuloma formation around S. mansoni eggs (27), and both cytokines are critical in the induction of experimental murine asthma (28). A similar IL-4 independence of IL-5 production and function in helminth disease models has been reported (29, 30). Finally, the decreased production of IL-5 in the draining lymph node corresponded with a marked decrease in footpad eosinophilia, showing that anti-CD40 treatment had wide-reaching effects on both Th2 cytokine production and Th2-mediated cellular responses characteristic of helminth-induced tissue pathology.

Treatment with anti-CD40 mAb reciprocally induced 3- to 10-fold increases in IL-12 and IFN-γ levels in vivo in this and other models of infection (13, 14, 15). We observed significant increases in circulating IL-12 that included both inactive free p40 and bioactive p70 forms. Whether there are distinct, additional cellular sources of local and systemic IFN-γ production remains to be determined. Consistent with a CD4⁺ T cell source for production of this cytokine at the site of the Ag-specific immune response, IFN-γ mRNA expression was enriched in CD4⁺ cells isolated from the LNC of anti-CD40-treated mice, whereas depletion of lymph node CD4⁺ cells reduced expression of IFN-γ mRNA. Additionally, depletion of lymph node CD4⁺ cells decreased the frequency of IFN-γ-secreting cells by 90%. These data are consistent with previous observations indicating Th1 cell differentiation was necessary for anti-CD40-induced cure of L. major infection in susceptible BALB/c mice (14). However, levels of spontaneous IFN-γ production in our studies were high and minimally increased upon Ag stimulation. These findings therefore do not differentiate between cytokine synthesis by T cells induced by residual egg Ag in the harvested tissues or the additional presence of nonspecific bystander T cell and NK cell responses (31, 32, 33). Although CD4⁺ cellular immune responses were shown to contribute to the observed increase in Th1 cytokine activity after anti-CD40 treatment, further studies are needed to exclude an activated innate cellular immune response as an additional source for local or systemic IFN-γ synthesis.

We initially hypothesized that the inhibition of developing Th2 responses by anti-CD40 mAb would be dependent on IL-12-elicited IFN-γ. Consistent with this, C57BL/6 mice lacking either endogenous IFN-γ or IL-12 were unable to down-regulate the egg-induced Th2 response following treatment with anti-CD40 mAb. However, anti-CD40 induced both local and systemic production of IFN-γ in IL-12 p40^−/− mice, yet did not suppress the S. mansoni egg-induced Th2 response. This observation is significant in two respects. First, the ability of CD40-induced responses to generate IL-12-independent IFN-γ synthesis by CD4⁺ cells in vivo is in itself a novel finding. IL-12-independent Th1 cytokine responses have been observed previously in murine models of hepatitis coronavirus or Toxoplasma gondii infection, or in response to cardiac allograft (34, 35, 36). Additionally, McDyer et al. demonstrated that CD40 activation could enhance IFN-γ production from PHA-stimulated human PBMCs cultured with αIL-12 (37). Our findings extend these observations to implicate CD40-activated responses in the synthesis of IL-12-independent IFN-γ by CD4⁺ LNC. Possible pathways of action may include the induction of alternative IFN-γ-inducing cytokines, such as IL-18, which may promote IFN-γ production in the absence of IL-12 (38). Alternatively, IFN-γ may be induced by enhanced IL-12-independent costimulatory interactions between T cells and CD40-activated APC (38).

Second, the induction of IFN-γ synthesis by anti-CD40 treatment in the absence of IL-12 provides new insights into whether IFN-γ alone is capable of mediating Th2-regulatory effects. The experimental inhibition of IL-12 activity in vivo is usually tightly associated with a concomitant down-regulation of IFN-γ synthesis (39, 40). Consequently, immunomodulatory effects specifically attributable to IL-12 or IFN-γ alone cannot be easily discerned. In contrast, treatment with anti-CD40 mAb dissociates the mutual regulatory interdependence of IL-12 and IFN-γ and permits us to conclude that IFN-γ alone is insufficient to down-regulate S. mansoni egg-induced Th2 responses. Further studies are needed to determine the molecular and cellular mechanisms responsible for anti-CD40-induced IFN-γ synthesis in IL-12-deficient mice and to better define the joint requirement for IL-12 and IFN-γ in the down-regulation of S. mansoni egg-induced immunity.

These findings show that CD40 activation in vivo inhibits the development and subsequent expression of IL-4-, IL-5-, and IL-13-producing cellular immune responses to a potent and well-characterized Th2-biasing stimulus. Our findings further suggest that the mechanism for Th2 inhibition is both IFN-γ and IL-12 dependent. Finally, we demonstrate the capacity of CD40 activation to stimulate IFN-γ synthesis in vivo in the absence of IL-12, although the development of Th2 cytokine production was not inhibited under these conditions. Because Th2 cytokine responses are a common pathogenic mechanism in a spectrum of clinical diseases, including atopy and helminth-induced tissue eosinophilia, these studies are significant in that they define a potent immunotherapy for inhibiting both Th2 cytokine production and Th2-dependent tissue responses. However, schistosome-infected mice treated with anti-CD40 do not demonstrate altered cytokine or Ab isotype production in ongoing preliminary studies. Further studies are needed to dissect the therapeutic effect of anti-CD40 in the chronic immune response of schistosome-infected mice. Our studies are also important in that they introduce a unique in vivo model of IL-12-independent IFN-γ production that may provide new insights into alternative mechanisms for activation of Th1-type cellular immunity. We conclude that CD40 activation inhibits Th2 cytokine responses and promotes Th1 cellular immunity, including effects extending beyond the currently understood boundaries of the IL-12/IFN-γ cascade.

We thank Drs. Laurie Hall, Alan Levine, and Lloyd Culp for their many helpful reviews and suggestions in the writing of this manuscript.