|Home | About | Journals | Submit | Contact Us | Français|
IL 4 receptor α (IL-4Rα) expression by non-bone marrow (BM)-derived cells is required to protect hosts against several parasitic helminth species. In contrast, we demonstrate that IL-4Rα expression by BM-derived cells is both necessary and sufficient to prevent Schistosoma mansoni-infected mice from developing severe inflammation directed against parasite ova, whereas IL-4Rα expression by non-BM-derived cells is neither necessary nor sufficient. Chimeras that express IL-4Rα only on non-BM-derived cells still produce Th2 cytokines, but overproduce IL-12p40, TNF, and IFN-γ, fail to generate alternatively activated macrophages, and develop endotoxemia and severe hepatic and intestinal pathology. In contrast, chimeras that express IL-4Rα only on BM-derived cells have extended survival, even though the granulomas that they develop around parasite eggs are small and devoid of collagen. These observations identify distinct roles for IL-4/IL-13 responsive cell lineages during schistosomiasis: IL-4Rα-mediated signaling in non-BM-derived cells regulates granuloma size and fibrosis, whereas signaling in BM-derived cells suppresses parasite egg-driven inflammation within the liver and intestine.
Interleukins 4 and 13 drive many immune and inflammatory responses that are associated with Th2 cytokines (1). IL-4 receptor α-chain (IL-4Rα)3 is essential for signaling by both cytokines and pairs with cytokine receptor common γ-chain to form the type 1 IL-4R, which is activated uniquely by IL-4, or with IL-13 Rα1 to form the type 2 IL-4 receptor, which is activated by both IL-4 and IL-13 (2). Although IL-4Rα is required for protective immunity against most intestinal and some nonintestinal parasitic helminths, it does not protect against all helminth parasites through a single mechanism (3). For example, expulsion of the intestinal nematodes Nippostrongylus brasiliensis and Strongyloides venenzuelensis only requires IL-4Rα expression by non-bone marrow (BM)-derived cells, while Trichinella spiralis expulsion requires IL-4Rα expression by both BM- and non-BM-derived cells (4, 5). No situation has been described to date in which IL-4Rα expression by non-BM-derived cells is not required for host protection against a nematode parasite.
Infection of mice with Schistosoma mansoni, a parasitic blood fluke that infects 250 million persons worldwide, is characterized by markedly increased IL-4 and IL-13 production that drives extensive fibrosis and granulomatous immunopathology directed against parasite eggs that lodge in the liver and intestine (6, 7). from Parasite ova cause the recruitment of lymphocytes, eosinophils, and macrophages to the local microenvironment and this inflammatory response may damage the entire organ if left unchecked. In this model, IL-4Rα expression on macrophages (M) is required to limit the severity of egg-induced inflammation (8), whereas T cell responsiveness to IL-4 is nonessential for host survival or www.jimmunol.org collagen deposition (9). M IL-4/IL-13 responsiveness is sufficient to induce myeloid precursors to differentiate into alternatively activated macrophages (AAM) (10). These cells secrete several proteins involved in the wound healing response that are not produced by classically activated macrophages (CAM), including arginase 1, relm α (FIZZ-1), and CCL17 (11). Conversion on July 28, 2010 of the common substrate L-arginine to polyamines and proline via arginase 1 instead of NO via inducible NO synthetase (NOS-2) is one mechanism that may allow AAM to block the effects of CAM. Although our previous work supports a hypothesis that BM-derived cells are required to protect against severe inflammation during S. mansoni infection, it was not addressed whether host protection against this parasite shares the requirement for IL-4Rα expression by non-BM-derived cells that has been identified in all other investigations of nematode parasite infection.
We have now used irradiation BM chimeras to distinguish between the contributions of IL-4Rα on BM-derived vs non-BM-derived cells to host protection against S. mansoni egg-induced immunopathology. Our results demonstrate two surprising observations: 1) IL-4Rα expression by non-BM-derived cells is not necessary to protect against lethality; and 2) granulomas that develop around parasite eggs are very small and devoid of fibrosis in mice in which non-BM-derived cells lack IL-4Rα. Thus, contrary to expectations, the development of large, fibrotic egg granulomas during the initial phase of infection does not reduce hepatotoxicity or promote host protection during acute schistosomiasis.
Mice were purchased from Taconic Farms. All mouse experiments were approved by the Institutional Animal Care and Use Committee at the Cincinnati Veterans Affairs Medical Center.
Six- to 8-wk-old BALB/c male wild-type (WT) and IL-4Rα-deficient mice were used. Mice received two doses of 475 Rads 3 h apart with a 154Cs irradiator, and 4 h later were administered 3–5 × 106 BM cells i.v. Mice were subsequently fed doxycycline-containing food for 3–4 wk. Mice were inoculated with S. mansoni 4–5 wk post-BM reconstitution.
Chimeric mice were anesthetized and percutaneously infected with 50 – 60 S. mansoni cercariae as previously described (8). Parasites were provided by the National Institute of Allergy and Infectious Diseases Schistosomiasis Resource Center at the Biomedical Research Institute, through National Institute of Allergy and Infectious Diseases Contract N0-AI-30026.
PBMC, isolated by Ficoll gradient, and splenocytes were washed in FACS buffer (HBSS, 1% FBS, and 0.2% sodium azide) and incubated with anti-FcγRII/RIII mAb (2.4G2). Cells were then stained with the mAbs described and analyzed with a BD FACSCalibur and CellQuest software. The percent of donor cells in chimeric mice was determined as percentage of cells that appeared IL-4Rα+ or IL-4Rα− in the chimeric mouse, divided by the percentage of cells that appeared IL-4Rα+ in WT mice or the percentage of cells that appeared IL-4Rα− in IL-4Rα-deficient mice.
IL-12 p40 levels were measured by ELISA (eBioscience). IL-4, TNF, and IFN-γ were measured by in vivo cytokine capture assay (18). Serum AST concentration was measured by the Cincinnati Veterans Affairs Medical Center clinical pathology laboratory. Serum LPS concentration was measured as described (8, 19).
Biopsies of liver or intestine (ileum) were collected from each mouse, weighed, and digested in 5% KOH at 37°C for 16 h, and eggs were counted at 40× magnification. Data were expressed as eggs per gram. Histologically processed sections of liver tissue (5 μm) were stained with H&E or Masson’s trichrome. Individual granuloma areas were measured on coded slides; only granulomas that possessed a central egg were evaluated. Quantitation of area was performed using a SPT Diagnostics imaging system and Simple PCI C-Imaging systems software. Data shown are the mean ± SE of 150 granulomas per group from two independent experiments.
RNA was obtained from hepatic tissue, DNase I-treated, and cDNA was generated using SuperScript II Reverse Transcriptase (Invitrogen Life Technologies). Real-time PCR was conducted on a Gene Amp 7500 instrument (PE Biosystems) with the Syber Green detection reagent. Cycle threshold values for NOS-2 and FIZZ-1 were determined and fold induction was compared with vimentin using the 1/ΔCT method (8) as follows: Vimentin Forward: 5′-TGA CCG GCT TGT ATG CTA TC-3′; Vimentin Reverse: 5′-CAG TGT GAG CCA GGA TAT AG-3′; NOS-2 Forward: 5′-CAG AAG AAT GGA AGA GTC AG-3′; NOS-2 Reverse: 5′-CAG ATA TGC AGG GAG TCA CC-3′; Relm-α/FIZZ-1 Forward: 5′-AGA TGG GCC TCC TGC CCT GCT GGG-3′; Relm-α/FIZZ-1 Reverse: 5′-ACC TGG TGA CGG GCG ACG ACG GTT-3′.
Statistical significance was assessed by either one-tailed Student’s t test (two groups) or ANOVA for multiple groups and a post hoc Bonferroni’s test to determine significance, all performed using Prism Graph Pad software.
IL-4Rα-deficient mice, but not WT mice, usually die 6 – 8 wk after natural infection with S. mansoni (8). To determine whether IL-4Rα expression on BM-derived and/or non-BM-derived cells is required to protect against S. mansoni-induced pathology, irradiation BM chimeras were generated by reconstituting WT mice with IL-4Rα−/− BM (BMIL-4Rα−/− mice) and IL-4Rα−/− mice with WT BM (non-BMIL-4Rα−/− mice). These mice and control mice produced by reconstituting WT mice with WT BM, and IL-4Rα−/− mice with IL-4Rα−/− BM were infected with 50 – 60 S. mansoni cercariae 5 wk posttransplant and evaluated 1 wk later by flow cytometry for BM engraftment by determining IL-4Rα expression. Lymphocytes and monocyte/macrophages showed highly efficient donor BM reconstitution, while neutrophil engraftment was less complete (Fig. 1). Of note, neutrophils (CD11b+/Ly6Ghigh cells) express relatively low levels of IL-4Rα, which makes it difficult to distinguish between IL-4Rα+ and IL-4Rα− cells (Fig. 1). Evaluation of IL-4Rα levels on CD45+ (BM-derived) splenocytes 7 wk postinfection gave consistent results, with only a small residual population of host cells in chimeric mice (Fig. 2A). S. mansoni infection induced striking differences in the responses of the two chimeric strains: body weights of BMIL-4Rα −/− and IL-4Rα−/− mice rapidly declined starting 6 wk postinfection (Fig. 2B) that preceded death of these animals (Fig. 2C). In contrast, non-BMIL-4Rα−/− mice lost only slightly more weight than WT mice, and had a significantly extended rate of survival compared with WT controls at 13 wk postinfection (Fig. 2C). Thus, protection against cachexia and lethality during murine schistosomiasis requires an IL-4 and/or IL-13 effect on a BM-derived cell, but not on a non-BM-derived cell.
Because cytokine abnormalities are associated with increased inflammation during schistosomiasis (20), we determined whether BM IL-4Rα-deficiency causes proinflammatory cytokine overproduction during S. mansoni infection. S. mansoni-infected IL-4Rα−/− and BMIL-4Rα−/− mice produced significantly more IL-12p40, IFN-γ, and TNF and less but still considerable IL-4 as compared with WT mice (Fig. 3, A–D), as has been described (21). In contrast, even less TNF and IFN-γ were produced in infected non-BMIL-4Rα−/− mice than in WT mice, while IL-10 production was similarly elevated in all strains (Fig. 3, B, C, and E).
Studies also evaluated whether weight loss and mortality in S. mansoni-infected IL-4Rα−/− and BMIL-4Rα−/− mice was associated with liver damage (measured as increased serum aspartate transaminase (AST), an enzyme released by damaged hepatocytes (7) and/or a leak of intestinal bacteria or LPS into the systemic circulation (as reflected by increased serum LPS). AST levels were ~4 – 6-fold higher and serum LPS levels >100-fold higher in S. mansoni-infected IL-4Rα−/− and BMIL-4Rα −/− mice than in WT mice and non-BMIL-4Rα−/− mice at 7 wk postinfection (Fig. 4, A and B). Quantitation of liver mRNA transcript levels for NOS-2 and FIZZ-1 was used to investigate whether hepatic injury was associated with CAM (NOS-2) or AAM (FIZZ-1). S. mansoni-infected IL-4Rα−/− and BMIL-4Rα−/− mice produced 1,000 to >2,000-fold higher levels of NOS-2 compared with WT mice, while transcripts in non-BMIL-4Rα−/− were much less elevated (Fig. 4C). In contrast, FIZZ-1 gene expression was 50-fold higher in infected non-BMIL-4Rα−/− chimeras than in IL-4Rα−/− or BMIL-4Rα−/− mice and 8-fold higher than in WT mice (Fig. 4D). These observations are consistent with the observation that S. mansoni-infected mice that selectively lack IL-4Rα on M (MIL-4Rα−/− mice) develop a dramatic increase in serum LPS (8). Both MIL-4Rα−/− mice and BMIL-4Rα−/− mice, like IL-4Rα−/− and IL-4/IL-13−/− mice, develop obvious gut hemorrhage during S. mansoni infection (Fig. 5) (8, 23).
The increased parasite ova-induced intestinal and liver damage in S. mansoni-infected BMIL-4Rα−/− mice was not due to increased egg production, inasmuch as liver egg burden was similar in WT, IL-4Rα−/−, non-BM-IL-4Rα−/−, and BM-IL-4Rα−/− mice (Fig. 6A). Furthermore, although intestinal egg burden was similarly increased in both chimeric mouse strains (Fig. 6B), the failure of S. mansoni-infected non-BMIL-4Rα−/− to develop increased immunopathology indicates that increased intestinal egg burden is not sufficient to induce severe immunopathology and that IL-4Rα-expressing BM-derived cells exclusively suppress egg-induced intestinal inflammation and breakdown of mucosal barrier function.
Both M (BM-derived) and fibroblasts (non-BM-derived) have been shown to respond to IL-13 and participate in collagen biosynthesis (24). Studies were performed to distinguish whether BM-derived or non-BM-derived cells were essential for granuloma size and hepatic collagen (measured by hydroxyproline concentration) (22). Results demonstrate that liver granuloma size and fibrosis were suppressed in non-BMIL-4Rα−/− mice and IL-4Rα−/− mice to nearly the same extent and were increased in BMIL-4Rα−/− mice compared with WT mice (Figs. 6, C and D and and7).7). These results dissociate IL-4Rα-dependent hepatic granuloma formation and collagen deposition from IL-4Rα-dependent control of inflammation and mortality during S. mansoni infection.
Our observations establish that IL-4Rα expression by non-BM-derived cells is neither necessary nor sufficient for survival of mice during acute infection with S. mansoni, while IL-4Rα expression by BM-derived cells is both necessary and sufficient. Although a requirement for IL-4Rα-expressing BM-derived cells for host protection against this parasite was obvious from our previously demonstrated requirement for IL-4Rα-expressing macrophages (8), the lack of a requirement for IL-4Rα-expressing non-BM-derived cells was surprising. IL-4 and IL-13, signaling through IL-4Rα and STAT6, have previously been shown to protect mice from infection against several different nematode parasites, including Nippostrongylus brasiliensis, Trichuris muris, and Trichinella spiralis (5, 25–27). IL-4Rα expression by non-BM-derived cells was required for host protection against each of these parasites, while the requirement for IL-4Rα expression by BM-derived cells was variable. Thus, our studies with S. mansoni provide the first example where IL-4Rα-expression by BM-derived cells is sufficient to protect mice against a parasitic helminth.
One reason for this difference between S. mansoni and all other parasitic worms studied to date may be that studies of host protection against other worms used worm expulsion as a readout, while the immune system protects hosts against S. mansoni by limiting inflammation against worm ova rather than by hastening parasite death or expulsion. Adult S. mansoni worm number is no greater in IL-4Rα-deficient mice than in WT mice (28), but IL-4Rα-deficient mice develop lethal inflammation that primarily affects the intestines and liver, while inflammation is controlled in wild-type mice. IL-4Rα-mediated STAT-6 activation is not required for Th2 polarization in S. mansoni-infected mice (29); neither does it promote the development of T cells that secrete the immunosuppressive cytokine IL-10. Thus, although mice doubly deficient in IL-4 and IL-10 develop more severe inflammation than mice lacking either cytokine alone (20), the IL-4/IL-13 limitation of inflammation by activation of AAM during acute S. monsoni infection does not appear to involve increased secretion of IL-10.
During the course of infection, worm ova are first produced 5– 6 wk after mice are inoculated with cercariae (30). Ova must pass from the host veins, where the adult worms reside, through vein walls and intestinal mucosa to be expelled in host feces and complete the worm life cycle (30). In this process, eggs have the potential to induce a severe intestinal inflammatory response. In normal mice, this inflammatory response is kept in check by macrophages that are responsive to IL-4 and IL-13, as shown by the development of intense, hemorrhagic intestinal inflammation in S. mansoni-infected mice that lack IL-4Rα on all cells, all BM-derived cells, or selectively on macrophages (8). This intense intestinal inflammation is accompanied by an increase in intestinal permeability that allows bacteria and/or bacterial products to enter the systemic circulation, as shown by a large increase in serum levels of LPS. The increased inflammatory response is also accompanied by increased production of inflammatory cytokines, including TNF, IFN-γ, and IL-12 p40 (a component of both IL-12, which stimulates IFN-γ production, and IL-23, which promotes the production of IL-17) (31). Increased production of these inflammatory cytokines probably contributes to the development of lethal inflammation in mice that lack IL-4Rα on BM-derived cells because anti-TNF mAb can prolong survival in these mice (32) and excessive IL-12 p40 production exacerbates hepatic immunopathology in S. mansoni-infected mice in association with expansion of IL-17 producing cells (33).
Although an increased number of worm eggs are trapped in the intestines of S. mansoni-infected IL-4Rα-deficient mice, our observations demonstrate that this abnormality is not sufficient to cause lethal inflammation; increased trapping of worm eggs also is observed in non-BMIL-4Rα−/− chimeras, which are protected against lethal inflammation. Thus, although an IL-4/IL-13-responsive non-BM-derived cell must contribute to the passage of eggs from veins into the gut lumen, an anti-inflammatory mechanism mediated by IL-4/IL-13-responsive BM-derived cells can compensate for the increased intestinal egg burden. Interestingly, chimeras that expressed IL-4Rα only on BM derived cells (non-BMIL-4Rα−/−) generated FIZZ-1 mRNA transcripts that were 8-fold higher than WT mice. This suggests that M activated by production of IL-4/IL-13 by cells of the adaptive immune system are involved in the compensatory mechanism that protects non-BMIL-4Rα−/− chimeras. Consistent with this, Loke et al. recently demonstrated that MHC class II restricted CD40T cell help is required to maintain AAM following helminth infection (34) and S. mansoni-infected RAG2-deficient mice develop severe tissue injury and significant mortality by 9 wk postinfection, despite the presence of IL-4Rα+ macrophages in these mice (35).
In addition to their increased intestinal pathology, BMIL-4Rα−/− chimeras, like fully IL-4Rα-deficient mice, develop severe liver disease, with substantial increases in serum levels of AST, which is released by damaged and dying hepatocytes (7). Previous studies with T cell-deficient and globally IL-4Rα-deficient mice were consistent with the possibility that liver damage resulted from the failure to develop perioval granulomas that might contain toxic products of the worm egg (36). In contrast, other studies demonstrated that mouse strains that produce liver egg granulomas of excessive size also suffer increased morbidity (37). Our new observations are consistent with the latter observations and indicate that large, fibrotic liver granulomas do not necessarily protect against host morbidity. Fully formed hepatic granulomas and increased serum levels of AST coexist in S. mansoni-infected BMIL-4Rα−/− chimeras, while infected non-BMIL-4Rα−/− chimeras have only slightly elevated serum AST levels despite their development of only rudimentary liver granulomas and their failure to develop egg-induced hepatic fibrosis. Thus, although hepatic granuloma formation in S. mansoni-infected mice depends on an IL-4/IL-13 effect on a non-BM-derived liver cell, hepatic protection against worm egg-induced toxicity during acute infection does not result from granuloma- or fibrosis-mediated containment and may be mediated by IL-4/IL-13-induced macrophage production of anti-inflammatory products. Indeed, the locus of protection may not even be the liver; we cannot currently rule out the possibility that hepatocyte toxicity results from the elevated levels of LPS or enteric bacteria that are caused by the increase in intestinal permeability.
Demonstration that macrophage-specific IL-4Rα-deficient mice (8) and BMIL-4Rα−/− chimeras produce as much liver collagen as WT mice argues against a hypothesis that AAM are required to promote egg-induced liver fibrosis during acute schistosomiasis (24). Instead, our studies are consistent with a hypothesis that IL-4/IL-13 act directly upon liver fibroblasts (hepatic stellate cells) to drive collagen synthesis (19, 38). Fibroblasts express the IL-13Rα1 chain and both IL-4 and IL-13 can induce type 1 collagen synthesis in murine fibroblast cell lines (39). This issue has considerable practical importance, because hepatic fibrosis (cirrhosis) is an important cause of morbidity and mortality during chronic murine schistosomiasis and in humans infected with S. mansoni (7). Our dissociation of fibrosis from host protection against lethal worm-induced inflammation suggests that it may be possible to inhibit hepatic cirrhosis without inducing the inflammatory effects that are normally prevented by IL-4Rα signaling.
Although our current and previous observations promote our understanding of how hosts respond to worm infection by identifying responses to worm eggs that are mediated by both BM-derived cells (inhibition of lethal inflammation) and non-BM-derived cells (granuloma formation, fibrosis, intestinal egg expulsion), they also raise several important issues: 1) what cell types are directly responsible for IL-4/IL-13-induced granuloma formation and fibrosis; 2) are macrophages the only IL-4/IL-13-responsive BM-derived cell that is required to inhibit lethal inflammation (our previous studies demonstrate that IL-4/IL-13-responsive T cells are not required); and 3) how do IL-4/IL-13-stimulated, “alternatively activated” macrophages protect against lethal inflammation. Each of these issues is currently under study.
We thank J. Bailey for assistance with bone marrow chimera generation, M. Chiaramonte for help with hydroxyproline measurements, and M. Khodoun for critical reading of this manuscript.
1This work was supported by the U.S. Department of Veterans Affairs and National Institutes of Health Grants R01 GM49758 and R01 AI052099, which included a diversity supplement.
3Abbreviations used in this paper: IL-4Rα, IL-4 receptor α-chain; BM, bone marrow; M, macrophage; AAM, alternatively activated macrophage; CAM, classically activated macrophage; NOS-2, inducible NO synthase; WT, wild type; FIZZ-1, relm-α; AST, aspartate transaminase.
The authors have no financial conflict of interest.