IL-18 Regulates Intestinal Mastocytosis and Th2 Cytokine Production Independently of IFN-γ During Trichinella spiralis Infection¹

Gastrointestinal nematodes cause some of the most prevalent and chronic human diseases worldwide. CD4⁺ Th2 cells generated in the mesenteric lymph nodes (MLN)³ during the course of infection are critical in host-protective immunity to many intestinal nematodes, including Trichinella spiralis, a parasitic nematode that can infect almost all mammalian species (1). T. spiralis infects the small intestine, and adults release newborn larvae that penetrate the intestinal wall and subsequently encyst in skeletal muscle. The expulsion mechanism of adult T. spiralis worms from the small intestine is a complex immune-mediated process that involves the activation of Th2 cells and mucosal mast cells (MMC). A significant role for both MMC (2, 3, 4, 5) and Th2-type cytokines such as IL-9, IL-4, and IL-13 (6, 7, 8, 9) in the expulsion process of T. spiralis has been demonstrated in vivo.

It is well established that cytokines produced by Th1 and Th2 cells act antagonistically and mutually regulate each other. However, the critical question of which factors regulate the balance between Th1 and Th2 cytokines during the course of an immune response is still largely unresolved. IL-18 was originally named IFN-γ-inducing factor (10) and is a potent inducer of IFN-γ, particularly when acting in concert with IL-12 (10, 11, 12, 13). In mucosal defenses, IL-18 is believed to be proinflammatory and has been shown to be up-regulated in the intestinal mucosa of patients with inflammatory bowel disorders such as Crohn’s disease (14, 15) and in patients with Helicobacter infection (16). The main body of work regarding IL-18 has thus far been focused on the ability of this cytokine to induce IFN-γ secretion, particularly together with IL-12. We have recently reported that IL-18 is essential for the development of chronic infection of the large intestine with the nematode Trichuris muris. Importantly, the critical effects of IL-18 in this infection are independent of IFN-γ and are mediated by the direct down-regulatory effect of IL-18 on IL-13 (17), demonstrating that IL-18 exerts immunomodulatory functions that extend beyond the IL-12/IFN-γ axis.

In this report, we provide new information on IL-18 as a key regulator of MMC development and Th2 responses in the small intestine. This study provides, for the first time, conclusive evidence that IL-18, without the help of IFN-γ, has a direct effect on MMC responses and that IL-18 plays a significant role in the development of pathology caused by gastrointestinal nematode infections.

Male NIH and C57BL/6 mice, 6–8 wk old, were purchased from Harlan Olac (Bicester, U.K.). Mice in which the IL-18 gene is disrupted (IL-18 knockout (KO) mice) were kindly provided by K. Takeda and S. Akira (18). IL-12 KO mice were originally provided by J. Magram (Hoffman-LaRoche, Nutley, NJ) (19), and IFN-γ KO mice were originally purchased from The Jackson Laboratory (Bar Harbor, ME). All gene-deficient mice were on a C57BL/6 background. All experiments were performed under the regulations of the Home Office Scientific Procedures Act (1986).

Maintenance, infection, and recovery of T. spiralis were as described previously (20). Experimental mice were infected with 300 infective T. spiralis larvae by oral gavage on day 0, and the numbers of adult worms in the small intestine were assessed at various time points postinfection (p.i.) as detailed in the text. Muscle larvae burden were determined on day 30 p.i. T. spiralis Ag was prepared as described previously (1).

In vivo treatment with rIL-18 was performed by i.p. injections of 200 ng rIL-18 (PeproTech, London, U.K.) per mouse daily from day 0 to day 10 post-T. spiralis infection. Control mice received i.p. injections of PBS.

MLN were removed from uninfected and infected animals and resuspended in RPMI 1640 supplemented with 10% heat-inactivated FCS, 2 mM l-glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin, and 0.05 mM 2-ME (all from Invitrogen, Paisley, U.K.). MLN were cultured at 37°C and 5% CO₂ in flat-bottom 96-well plates (Nunc, Roskilde, Denmark) at a final concentration of 5 × 10⁶/ml in a final volume of 0.2 ml/well. Cells were stimulated with T. spiralis Ag (50 μg/ml) or plate-bound anti-CD3 Ab (mAb 145-2C11, 10 μg/ml; American Type Culture Collection, Manassas, VA). Anti-IL-4R mAb (M1, 5 μg/ml; from Dr. C. Maliszewski, Immunex, Seattle, WA) was added to cultures to increase detection of IL-4. Cell-free supernatants were harvested after 48 h and stored at −80°C.

Bone marrow-derived mast cells (BMMC) were generated by culturing bone marrow cells (0.5 × 10⁶/ml) from femurs of naive 8-wk-old C57BL/6 mice for 25 days in culture medium supplemented with 20% WEHI-3-conditioned medium and various doses of rIL-18 (PeproTech). Nonadherent cells were transferred into new flasks every 7 days and suspended in fresh medium with or without added rIL-18. After 25 days in culture, the cells were 95% BMMC as determined by toluidine blue staining of cytospins (data not shown). The cells were carefully resuspended, washed three times, and counted using an automated cell counter (Casy 1 TT; Scharfe System, Reutlingen, Germany). The cells were then replated in either normal tissue culture plates (0.5 × 10⁶/ml) or ELISPOT plates (10⁴ cells/well) and cultured for an additional 24–48 h in the presence of 5 ng/ml PMA and 500 ng/ml ionomycin (both from Sigma-Aldrich, Gillingham, U.K.). Cell-free supernatants for the detection of MMC protease-1 (MMCP-1) levels were harvested after 48 h and stored at −80°C.

Cytokine analyses were conducted using sandwich ELISAs for IL-4 (mAb BVC4-1D11 and BVD6-24G2.3; BD PharMingen, San Diego, CA) and IFN-γ (R46A2 and XMG1.2; BD PharMingen). IL-13 and IL-10 were analyzed using Ab pairs from R&D Systems (Abingdon, U.K.). IL-4-producing BBMC cells were enumerated using an IL-4-ELISPOT kit (Mabtech, Nacka, Sweden).

Serum and supernatant levels of MMCP-1 were determined using a commercially available kit (Moredun Animal Health, Penicuik, U.K.).

Consecutive lengths of small intestine taken 10 cm from the pyloric sphincter were fixed in Carnoy’s fluid or neutral buffered formalin and histologically processed using standard methods; 5-μm sections were stained for MMC (0.5% toluidine blue), goblet cells (periodic acid-Schiff), and eosinophils (H&E). The number of cells per 20 randomly selected villus crypt units were determined under light microscopy from at least two sections per animal.

Total RNA was extracted from tissue specimens taken from the small intestine using Trizol (Invitrogen) according to the manufacturer’s instructions. A custom-made Riboquant template (BD PharMingen) was used to assay mRNA levels of IL-4, IL-13, IL-10, IL-9, IFN-γ, and GAPDH. In vitro transcription with [³²P]UTP (Amersham, Little Chalfont, U.K.) was performed using a Riboprobe kit (Promega, Southampton, U.K.) and T7 polymerase (Promega). From each sample, 10 μg of RNA were hybridized with the radiolabeled antisense RNA probe set, digested with RNases, and purified, and the protected probes were resolved on denaturing sequencing gels. Dried gels were exposed to phosphorimaging screens, and protected fragments were visualized using a Molecular Imager FX System (Bio-Rad Laboratories, Hertfordshire, U.K.). All samples were normalized with respect to the housekeeping gene GAPDH to ensure equal input of RNA.

Significant differences (p < 0.05) between experimental groups were determined using the Mann-Whitney U test.

The response of IL-18 KO mice to primary infections with T. spiralis was examined. KO mice and C57BL/6 WT mice were infected with 300 T. spiralis larvae, and worm burdens were assessed at days 8, 12, and 16 p.i. (n = 5). IL-18 KO mice had significantly reduced worm burdens by day 8 as compared with WT mice (p < 0.05) (Fig. 1,A). The difference in expulsion kinetics was still apparent at days 12 and 16 p.i. when WT mice exhibited significant numbers of worms in the intestine while IL-18 KO mice had almost completed the expulsion process (p < 0.01) (Fig. 1,A). IL-12 p40 KO mice were also infected in a parallel experiment, and these mice did not differ in their worm expulsion kinetics from the WT mice at any time point (Fig. 1 A).

Skeletal muscle larvae burdens were assessed in the infected groups at 30 days p.i. The results revealed that IL-18 KO mice had a 4-fold decrease in muscle larvae deposition as compared with WT mice (p < 0.05) (Fig. 1 B) correlating with the increased worm expulsion rate seen in these animals.

Because MMC play an important role in the expulsion of T. spiralis (3, 7, 21) we investigated mast cell recruitment and activity in infected mice from the two groups. The number of MMC in the jejunum of IL-18 KO mice were significantly increased over WT controls at all time points during infection (p < 0.05) (Fig. 2,A). There was no significant difference in the number of MMC in the uninfected animals (Fig. 2 A).

To determine whether the increased recruitment of mast cells to the small intestine in IL-18 KO mice was also reflected in increased mast cell degranulation, we analyzed the levels of MMCP-1 in serum. IL-18 KO mice expressed high levels of serum MMCP-1 throughout the course of T. spiralis infection (p < 0.05 at all time points p.i. compared with WT mice), indicating that the mast cell hyperplasia seen in IL-18 KO mice consisted of mature and active mast cells. Naive IL-18 KO mice did not express elevated levels of MMCP-1 in their sera (Fig. 2 B).

We also investigated the number of goblet cells and eosinophils in the small intestine during infection and found no significant difference between the groups (Fig. 2, C and D), demonstrating that the absence of IL-18 preferentially stimulates a MMC response.

The Th2 cytokines IL-4, IL-13, and IL-10 are important in the development and recruitment of MMC (8, 22, 23). To investigate the cytokine response in T. spiralis-infected IL-18 KO and WT mice, MLN cells were harvested at various time points p.i. and restimulated in vitro with T. spiralis Ag. The results in Fig. 3 demonstrate that IL-18 KO mice develop strong Th2 responses as compared with WT mice. IL-18 KO mice displayed a 30-fold increase in IL-4 secretion as compared with WT mice on day 8 p.i. (IL-18 KO, 7416 ± 435 pg/ml; WT mice, 267 ± 158 pg/ml; p < 0.03; Fig. 3,A), and a >100-fold increase in IL-13 secretion (IL-18 KO, 84.29 ± 16.22 ng/ml; WT mice, 0.73 ± 0.16 ng/ml; p < 0.03; Fig. 3,B). The levels of IL-10 were also significantly increased at day 8 p.i. in IL-18 KO mice as compared with WT mice (IL-18 KO, 2427 ± 302 pg/ml; WT mice, 188 ± 126 pg/ml; p < 0.03; Fig. 3 C).

To investigate whether the strong increase in Th2 responses seen in the IL-18 KO mice was reflected in a similar decrease in Th1 response, we measured the levels of IFN-γ. The secretion of IFN-γ was significantly reduced in IL-18 KO as compared with WT mice on day 8 p.i. (IL-18 KO, 30.4 ± 16 ng/ml; WT mice, 439 ± 165 ng/ml; p < 0.03; Fig. 3 D)

The effect of exogenous IL-18 was investigated by treating T. spiralis-infected NIH mice with daily i.p. injections of 200 ng rIL-18 from day 0 to day 10 p.i. NIH mice are fast responders to T. spiralis infection and have normally completed worm expulsion around day 10–14 p.i. Control NIH mice treated with PBS started to expel the worms at day 7 p.i., whereas the rIL-18-treated group still exhibited full worm burden at this time point (Fig. 4,A; p < 0.02). Ten days p.i., the PBS-treated controls had almost completed expulsion, whereas the rIL-18-treated animals still had significantly higher worm burdens (Fig. 4,A; p < 0.02). The expulsion process was completed in both groups by day 13 p.i. (Fig. 4 A).

To investigate the biological significance of the delayed expulsion observed in the rIL-18-treated animals, skeletal muscle larvae burdens were assessed at 30 days p.i. The group that received rIL-18 treatment had a 2-fold higher number of muscle larvae than did the PBS-treated control group (p < 0.02), demonstrating that IL-18 plays an important biological role in the outcome of T. spiralis infection in vivo (Fig. 4 B).

The numbers of MMC in the jejunum of rIL-18-treated mice were significantly decreased at day 7 p.i. as compared with PBS-treated controls (p < 0.05; Fig. 5,A). There was no significant difference in mast cell numbers at any other time point. Furthermore, the levels of MMCP-1 in serum were also significantly decreased in the rIL-18-treated group as compared with PBS-treated controls on days 4, 7, and 13 (p < 0.05) (Fig. 5 B). Taken together, these data demonstrate that in vivo treatment with rIL-18 inhibits MMC recruitment as well as maturation and/or activation.

We also investigated the effect of in vivo rIL-18 treatment on the number of goblet cells and eosinophils during the infection. The in vivo treatment with rIL-18 had no significant effect on the number of goblet cells (Fig. 5,C), but a significant increase in the number of intestinal eosinophils were detected in rIL-18-treated animals on day 4 p.i. as compared with PBS-treated controls (Fig. 5,D; p < 0.02). There was no significant difference in the number of eosinophils at any other time point during the infection (Fig. 5 D).

When cytokine secretion from Ag-stimulated MLN cultures were examined, the results show that the PBS-treated NIH mice develop strong Th2 responses during the course of T. spiralis infection (Fig. 6). The rIL-18-treated NIH mice, however, had significantly reduced secretion of IL-13 on days 4 and 7 p.i. (p < 0.02 on day 4 and p < 0.04 on day 7; Fig. 6,B) and IL-10 at day 4 p.i. (p < 0.04; Fig. 6,C) as compared with the PBS-treated controls. There was no significant decrease in the amount of secreted IL-4 at any time point (Fig. 6,A). Ag-specific IFN-γ secretion was detected in both groups at day 4 p.i., but thereafter the levels of IFN-γ decreased to low levels. The levels of IFN-γ were not significantly different between the groups at any time point (Fig. 6 D).

To investigate the kinetics of the cytokine response at the site of infection, the cytokine mRNA levels in the small intestine were analyzed by RPA. IL-18-treated mice had significantly reduced expression of intestinal IL-13 mRNA on both days 4 and 7 p.i. (p < 0.04 for both time points; Fig. 7,B) and reduced expression of IL-9 and IL-10 at day 7 p.i. (p < 0.04 for both IL-9 and IL-10; Fig. 7, C and D). The levels of IFN-γ mRNA never increased significantly over the levels seen in uninfected controls in either rIL-18-treated or PBS-treated mice (Fig. 7 E).

To confirm that the effects of IL-18 seen in vivo during T. spiralis infection were not mediated through the induction of IFN-γ, we treated infected IFN-γ KO mice with daily injections of rIL-18 (200 ng/mouse). The control group received PBS injections. Worm burdens were assessed and found to be significantly higher at days 8 and 12 p.i. in the rIL-18-treated group (p < 0.04 for day 8 and p < 0.03 for day 12 p.i.; Fig. 8,A). This was also reflected in the increased number of encysted muscle larvae detected at day 30 p.i. (p < 0.01; Fig. 8,B) as well as in a significant decrease in MMC numbers at days 12 and 15 p.i. (p < 0.04 for days 12 and 15 p.i.; Fig. 8 C). Thus, the data confirm that the effects of IL-18 in promoting worm survival and fecundity is IFN-γ independent in vivo.

When the Ag-specific cytokine responses were analyzed, the results demonstrated that the rIL-18-treated mice secreted significantly reduced levels of Ag-specific IL-13 on days 8 and 12 p.i. as compared with the PBS-treated controls (p < 0.04 on days 8 and 12; Fig. 9,B). Ag-specific IL-4 and IL-10 secretion was significantly reduced in the rIL-18 treated group at day 12 p.i. (p < 0.04; Fig. 9, A and C). These results clearly demonstrate that IL-18 can inhibit Th2 cytokine responses in the absence of IFN-γ.

To further investigate whether IL-18 has a direct effect on mast cell differentiation, we cultured BMMC with varying doses of rIL-18 added to the culture medium. After 25 days in culture, the cells were determined to be 95% mast cells by toluidine blue staining of cytospins for all culture conditions used (data not shown).

When the cell yield for each condition was determined, the data revealed that the addition of rIL-18 resulted in a dose-dependent decrease in the number of cells per well (Fig. 10 A). There was no difference in the morphology of the mast cells, and there was no increase in any other cell type in the wells (data not shown).

We also analyzed the level of maturation of the different BMMC. The mast cells from the different conditions were replated at equal numbers per well and stimulated with PMA-ionomycin for 48 h and the supernatants were analyzed for MMCP-1 secretion. As shown in Fig. 10,B, there was no difference in the levels of MMCP-1 secreted by the BMMC, indicating that all the mast cells were mature regardless of whether or not they were grown with IL-18. However, when we analyzed the number of IL-4-producing mast cells by ELISPOT, there was a significant dose-dependent decrease in the number of IL-4-producing cells (Fig. 10 C). When rIL-18 was added to already mature BMMC (after 25 days of culture) for 48 h, no effect was seen on cell numbers, MMCP-1 secretion or IL-4 production (data not shown). Taken together, these results demonstrate that IL-18 is effective in inhibiting mast cell responses in several ways, including reduced proliferation and the inhibition of cytokine secretion, whereas other responses, such as protease secretion, remains unaffected and that the effect of IL-18 appears to be on mast cell precursors rather than on mature mast cells.

IL-18 is an important component in the development of a Th1 response but is unable to initiate a Th1 response without the presence of IL-12 (10, 11, 12). However, recent findings indicate that the in vivo effects of IL-18 now extend to include direct antimicrobial and immunomodulatory effects that are IFN-γ and/or IL-12 independent (17, 24). The data presented here provide the first report that IL-18 regulates the development of MMC and Th2 responses in the small intestine. Furthermore, we demonstrate that this effect is independent of IFN-γ and that it has a significant impact on the development of intestinal immunity against T. spiralis infection.

Mouse strains can be divided into slow or fast responders according to their ability to expel T. spiralis from the small intestine (21, 25). The speed of expulsion is clearly correlated with the capability to develop MMC hyperplasia (21) and blocking mast cell development by anti-stem cell factor treatment or by blocking the c-kit receptor in vivo significantly delays worm expulsion (2, 3). Furthermore, mast cell-deficient W/W^v mice (4) as well as MMCP-1 KO mice (5) display delayed expulsion, demonstrating that MMC are actively involved in the expulsion of adult T. spiralis worms. The exact mechanism of expulsion is not clear but is known to be dependent on a Th2 type of response (involving IL-4, IL-13, and IL-9) that leads to the activation of MMC (1, 2, 3, 7, 8, 9).

Because IL-18 is known as a potent inducer of IFN-γ (10, 11, 12, 13), we wanted to investigate the importance of this cytokine during a T. spiralis infection in vivo. When IL-18 KO and WT (C57BL/6) mice were infected with T. spiralis, the IL-18 KO mice rapidly expelled the worms; as a result, the muscle larvae burdens were significantly decreased in IL-18 KO mice as compared with WT mice. Thus, IL-18 may promote the survival of T. spiralis in vivo. Interestingly, when IL-12KO mice were infected in parallel, these mice were no different from WT mice in terms of speed of worm expulsion or in the number of encysted muscle larvae, indicating that endogenous IL-12 does not play a major role in delaying T. spiralis expulsion. Previous work has shown that vivo administration of rIL-12 to Nippostrongylus brasiliensis-infected mice delay worm expulsion and inhibit mastocytosis in an IFN-γ-dependent manner (26). Thus, exogenous IL-12 may inhibit mast cell development through its capacity to induce IFN-γ responses rather than acting directly on mast cell precursors.

When we investigated the cytokine basis for the observed differences in the kinetics of worm expulsion between IL-18 KO and WT mice, the results showed that the rapid expulsion seen in IL-18 KO mice correlated well with high levels of Ag-specific Th2 cytokine secretion (IL-4, IL-13, and IL-10). The levels of IFN-γ secretion from IL-18 KO mice were significantly lower than those from WT mice in agreement with other studies (18).

NIH mice are fast responders in terms of T. spiralis expulsion (the expulsion process is usually completed around day 10–12 p.i.) (21, 25), whereas the slow responder strain C57BL/6 does not complete expulsion until around day 15–18 p.i. However, when NIH mice were treated with daily injections of rIL-18, they showed significantly delayed expulsion and increased numbers of muscle larvae. This finding, together with the IL-18 KO data, demonstrates that administration of rIL-18 to a fast responder strain, such as NIH, can turn it into a slow responder strain, whereas removing IL-18 from a slow responder strain (C57BL/6) will turn it into a fast responder (IL-18 KO). When NIH mice were treated with rIL-18 in vivo during T. spiralis infection, a significant decrease in IL-13 and IL-10 secretion was detected during early time points of the infection in the MLN. When the local cytokine mRNA expression of the small intestine was analyzed by RPA, a similar decrease could be detected for IL-13, IL-10, and IL-9 in situ. The levels of IFN-γ secretion in MLN cultures were no different between the rIL-18-treated group and the PBS-treated controls, and IFN-γ mRNA did not increase in the small intestine at any time point p.i. in either group. Taken together, the data show that the IL-18-induced suppression of Th2 responses was not linked to an increase in IFN-γ, either in the MLN or locally at the site of infection, indicating that the effect might be IFN-γ independent. To confirm this, we infected IFN-γ KO mice and treated them with rIL-18 or PBS injections. IFN-γ KO mice are highly resistant to T. spiralis infection and expel the parasites quickly under normal conditions. In vivo treatment with rIL-18, however, delayed expulsion of the worms, increased muscle larvae burdens, and suppressed Th2 cytokine secretion significantly, clearly demonstrating that the in vivo effects of IL-18 on gastrointestinal nematode infections is IFN-γ independent, in agreement with our previous studies (17).

The maturation and development of MMC are largely dependent on Th2 cytokines such as IL-3, IL-4, IL-13, IL-9, and, in particular, IL-10 (8, 22, 23, 27). T. spiralis-infected IL-18 KO mice developed a high level of mastocytosis in the small intestine during infection, correlating with the enhanced expulsion rate of the parasites and the increase in Th2 cytokine secretion. Injections of rIL-18 into normal mice during infection significantly decreased the mastocytosis as well as the levels of Th2 cytokines. This was also true in rIL-18-treated IFN-γ KO mice. Goblet cell hyperplasia was not affected by IL-18 treatment, and the effect on eosinophil recruitment was only marginal, demonstrating that the effect of IL-18 appears to be specific for MMC in vivo. Both mast cells and basophils express IL-18R (28), and human intestinal mast cells have been shown to express IL-18 (29), which may possibly function as a negative feedback mechanism. To confirm that IL-18 has a direct regulatory effect on MMC, we cultured BMMC with rIL-18 in vitro. Bone marrow cells from normal mice were cultured to BMMCs according to standard methodology with varying doses of rIL-18 present in the cultures throughout the experiment. The effect of adding rIL-18 to the developing BMMC was pronounced. IL-18 inhibited cell proliferation and cytokine production in a dose-dependent manner but did not affect the production of MMCP-1. These results demonstrate that IL-18 is effective in inhibiting mast cell responses in several ways, including reduced proliferation and the inhibition of cytokine secretion, whereas other responses, such as protease secretion, are unaffected. IL-4 enhances mast cell proliferation (30), and because we detected a concomitant decrease in the number of IL-4-producing mast cells and cell yield in the rIL-18-treated BBMC cultures, this may represent a possible inhibitory pathway of IL-18 on mast cell development. When we added rIL-18 to already mature BMMC cultures, there was no detectable effect on cell yield, MMCP-1 secretion, or IL-4 production (data not shown), indicating that IL-18 acts on precursors rather than on mature mast cells. MMC are believed to be immature when they arrive in the intestine during the first week of T. spiralis infection. In the intestine, they then become mature and express MMCP-1 (31, 32). Thus, IL-18 may act locally in the intestine to inhibit mast cell maturation but may also already act at the level of mast cell precursors in the bone marrow.

In this report, we provide new information on IL-18 as a key regulator of MMC development and Th2 responses in the small intestine. This study provides, for the first time, conclusive evidence that IL-18 has a direct effect on MMC responses and that IL-18 plays a significant role in the development of pathology caused by gastrointestinal nematode infections. This is the first report showing the importance of IL-18 in regulating intestinal mast cell functions in vivo, and these results extend our knowledge of the cytokine-mediated regulation of intestinal inflammation. The data may thus provide important information for the design of rational therapies against parasitic infections, allergic reactions, and inflammation of the intestine and other mucosal sites.

We thank Neil Humphreys and Ann Lowe for excellent technical assistance and Prof. Nancy Rothwell for kindly providing animals.