Human coinfection with the helminth parasite Schistosoma mansoni and hepatitis B and hepatitis C viruses is associated with increased hepatic viral burdens and severe liver pathology. In this study we developed a murine S. mansoni/lymphocytic choriomeningitis virus (LCMV) coinfection model that reproduces the enhanced viral replication and liver pathology observed in human coinfections, and used this model to explore the mechanisms involved. Viral coinfection during the Th2-dominated granulomatous phase of the schistosome infection resulted in induction of a strong LCMV-specific T cell response, with infiltration of high numbers of LCMV-specific IFN-γ-producing CD8+ cells into the liver. This was associated with suppression of production of the Th2 cytokines dominant during S. mansoni infection and a rapid increase in morbidity, linked to hepatotoxicity. Interestingly, the liver of coinfected mice was extremely susceptible to viral replication. This correlated with a reduced intrahepatic type I IFN response following virus infection. Schistosome egg Ags were found to suppress the type I IFN response induced in murine bone marrow-derived dendritic cells by polyinosinic-polycytidylic acid. These results suggest that suppression of the antiviral type I IFN response by schistosome egg Ags in vivo predisposes the liver to enhanced viral replication with ensuing immunopathological consequences, findings that may be paralleled in human schistosome/hepatotropic virus coinfections.
Schistosomiasis is a chronic and debilitating disease resulting from infection with trematode helminth parasites of the genus Schistosoma, which affects ∼200 million people worldwide, particularly in tropical countries (1). Schistosoma mansoni is the major cause of intestinal schistosomiasis, and S. mansoni infection of mice is a well-characterized model for the human disease. Following infection with schistosome cerceriae, worms mature in the portal venous system, where after ∼5 wk, egg laying occurs. The pathology associated with the infection is predominantly caused by the host immune response to eggs that are subsequently trapped in the liver and intestines (2, 3). Deposition of eggs in the liver precipitates a dramatic circumoval granulomatous inflammation, which is composed primarily of macrophages, eosinophils, lymphocytes, and fibroblasts (4). The granulomatous pathology is orchestrated by CD4 T cells, which possess a strong Th2 cytokine bias characterized by production of high levels of IL-4, IL-5, IL-10, and IL-13 (5, 6, 7, 8). However a number of studies have also indicated that the Th2 cytokine environment is required for survival. This finding is exemplified by mice deficient in both IL-4 and IL-10, which show a marked Th1 cytokine response to schistosome egg Ags (SEA)3, resulting in diminished granuloma formation but markedly increased hepatotoxicity and mortality (9, 10, 11, 12). Interestingly, dispersed isolated granulomas appear to be composed of both Th1 and Th2 cytokine-producing cells, although in vivo the egg Ag response is Th2 cytokine biased, indicating that the Th1 cells are under suppression (13, 14). Collectively these studies suggest that granuloma formation is under tight immunoregulatory control, with any dramatic polarization of the response being detrimental.
It would thus seem likely that immune responses to concomitant infections could influence the course of the schistosome infection. In addition, the Th2 bias in the cytokine profile during schistosome infections may influence the immune response to unrelated Ags/pathogens, a phenomenon documented in both human and murine schistosome infections. Perhaps the most striking example of the potential consequences of coinfection comes from studies of patients coinfected with S. mansoni and hepatitis B virus (HBV) or hepatitis C virus (HCV) (1, 15, 16). These coinfected individuals show a higher incidence of HBV or HCV chronicity with a significantly increased hepatic viral burden. They also demonstrate a rapid progression to liver pathology, which correlates with high serum transaminase and NO levels, and an increased mortality. The mechanisms responsible for the increased viral replication, liver damage and mortality, are not well understood.
We aimed to establish a murine model of schistosome and hepatotropic virus coinfection that could be used to provide insight into the pathogenesis of human coinfection with S. mansoni and HBV or HCV. As HBV and HCV do not infect mice, we used a hepatotropic strain of lymphocytic choriomeningitis virus (LCMV) (17), which is a natural pathogen of mice, as our coinfection model. Murine LCMV infection has many features in common with human HBV and HCV infections. LCMV is a noncytopathic virus that replicates systemically in tissues including the liver, where infection can be associated with an immunopathologically mediated hepatitis (18). Like HBV and HCV, LCMV can cause either acute or persistent infections depending on how well its replication is controlled by the immune response. The strong CD8+ CTL response mounted to the virus plays an essential role in control of acute LCMV infection (19, 20, 21), although long-term containment of virus replication also requires help from other arms of the immune response. Type I IFNs make a crucial contribution to the control of virus replication at early times postinfection (p.i.), and type I IFN receptor-deficient mice show enhanced viral replication and suffer a prolonged viremia (22, 23).
Previous studies in murine models of schistosome coinfection with viruses infecting the liver have demonstrated a dominance of the schistosome-induced Th2-type cytokine response, and have not reproduced the severe liver pathology and associated morbidity and mortality observed in human schistosome/hepatotropic virus coinfections (24, 25). In contrast, we show that our murine S. mansoni/LCMV coinfection model parallels many aspects of human schistosome/HBV and HCV coinfection, including enhanced intrahepatic viral replication, and a dramatic increase in morbidity and mortality, which correlated with exacerbated liver pathology. We also show that the type I IFN response induced in the liver following LCMV infection was markedly reduced in S. mansoni-infected mice and that SEA was able to impair the type I IFN response induced by polyinosinic-polycytidylic acid (poly(I:C)) in bone marrow-derived dendritic cells (BMDC) in vitro, suggesting a potential mechanistic basis for the observed in vivo impairment of the type I IFN response in liver tissue with a high egg burden. These results give insight into mechanisms that may contribute to the enhanced viral replication and severe liver pathology in human schistosome/hepatotropic virus coinfections (15, 26, 27).
Materials and Methods
Mice and schistosome and LCMV infections
C57BL/6 mice obtained from Charles River Breeding Laboratories were infected at 8–12 wk of age with a Puerto Rican strain of S. mansoni. The parasite lifecycle was maintained in laboratory reared Biomphalaria glabrata snails and Swiss Albino (CD1) strain mice (Charles River Breeding Laboratories). Mice were infected percutaneously with either 25 or 50 S. mansoni cerceriae. Coinfections were conducted at 10 wk postschistosome infection, by i.v. inoculation of 105 PFU of a hepatotropic strain of LCMV, WE2.2, which is a clone plaque-purified from the WE total population (17). Stocks of LCMV WE2.2 were prepared on baby hamster kidney cells, and viral titers of these stocks and experimental samples were determined by plaque assay on Vero cells (28). All animal studies were approved by the site ethical review committee and were conducted in accordance with U.K. Home Office regulations.
Hepatic schistosome egg burden analysis
Schistosome-infected liver tissue was weighed and digested by overnight incubation with a 10× volume of 4% KOH at 37°C. Egg counts were performed with a light microscope on a fixed volume of digested liver material. Counts are expressed as egg number per gram of liver tissue.
Liver pathology was assessed by histologic examination of formalin-fixed, H&E-stained tissue sections (prepared by the histology service, Institute for Animal Health, Compton, U.K.). Aspartate transaminase (AST) levels in murine serum were determined using a kit (Sigma-Aldrich), and are expressed as Sigma-Frankel units per milliliter derived from a standard curve. Alanine transaminase (ALT) levels in murine serum were also determined using a kit (Fortress Diagnostic). NO levels were determined with a total NO detection kit, based on the Greiss reaction (StressGen Biotechnologies) and are expressed as micromoles per milliliter.
Splenocyte and intrahepatic lymphocyte (IHL) preparation
RBC-depleted splenocyte suspensions were prepared in RPMI 1640 medium (Sigma-Aldrich) supplemented with 10% FCS, 100 U/ml penicillin, and 100 μg/ml streptomycin (complete RPMI). IHL were prepared by initially perfusing the liver in vivo via the portal vein with a PBS/heparin (75 U/ml) mixture. The excised liver was cut into small pieces and incubated in 10 ml of liver digest buffer (collagenase IV/dispase mix; Invitrogen Life Technologies) for 30 min at 37°C. The liver suspension was then mashed through a cell sieve, suspended in 25 ml of complete RPMI 1640 with 0.1 M EDTA, and centrifuged at low speed to sediment the hepatocytes. The remaining cells were separated on a 35% Percoll gradient by centrifuging at 600 × g. The lymphocyte fraction was resuspended in 2 ml of red cell lysis buffer, then washed in 10 ml of complete RPMI 1640 with 0.1 M EDTA.
Splenocytes or IHL were cultured at 5 × 105 cells/well in 96-well plates in complete RPMI medium for 3 days at 37°C with either 2 μg/ml Con A (Sigma-Aldrich) or 25 μg/ml SEA obtained from Prof. M. Doenhoff (University of Wales, Bangor, U.K.). Culture supernatants were tested for IL-4 or IL-5 using a cytometric bead array kit from BD Biosciences. Samples were run on a FACSCalibur (BD Biosciences) using CellQuest software (BD Biosciences) and the results were analyzed using the CBA analysis software. Culture supernatants were tested for IFN-γ, IL-10, and IL-13 levels using R&D Systems mouse Duoset kits. ELISA plates were read on a spectrophotometer (Molecular Devices) and data were analyzed with Softmax-Pro software (Molecular Devices).
Intracellular cytokine staining
A total of 5 × 105 splenocytes or IHLs were cultured for 5 h in 24-well plates in complete RPMI medium containing 10 U/ml IL-2, 1 μl/ml Golgi-Plug (BD Biosciences), and peptides (1 μM), synthesized by Dr. L. Hunt, Institute for Animal Health (Compton, Newbury, Berkshire, U.K.), corresponding to the LCMV NP396 and GP33 H-2b-restricted CD8 T cell epitopes. Cells were stained with an anti-mouse CD8 Ab (Ly-2) conjugated to PE (BD Biosciences) in staining buffer (PBS, 2% FCS, 0.1% NaN3). Following fixation, cells were permeabilized in 0.5% saponin (Sigma-Aldrich) in PBS and stained with anti-mouse IFN-γ Ab (XMG1.2) conjugated to allophycocyanin (BD Biosciences). Samples were analyzed on a FACSCalibur using CellQuest software.
Liver tissue was snap frozen in OCT (BDH Laboratory Supplies) by immersion in liquid nitrogen. The 10-μm tissue sections were cut on a cryostat and fixed in acetone/ethanol (1:1). Sections were treated with an avidin/biotin blocking reagent (Vector Laboratories) and then with the M.O.M. Mouse Ig blocking reagent (Vector Laboratories). They were then stained with purified anti-LCMV nucleoprotein Ab (NP113) prepared from a hybridoma from Dr. M. Buchmeier (The Scripps Research Institute, La Jolla, CA; the Ab was conjugated to Cy5 using a FluoroLink-Ab Cy5 labeling kit (Amersham Biosciences)), and either anti-mouse CD11b (BD Biosciences) or anti-mouse F4/80 (Caltag Laboratories) conjugated to FITC. After washing, slides were treated with Vectashield (Vector Laboratories) and analyzed on a Leica TCS-NT confocal microscope.
Stimulation of BMDC with poly(I:C)
Murine BMDC were prepared as previously described (29). Cells were cultured overnight at 37°C in 24-well plates at 106 cells/well in the presence or absence of 50 μg/ml SEA. SEA-treated and untreated BMDC were then stimulated by addition of 50 μg/ml poly(I:C) (Sigma-Aldrich). Cells were harvested 24 and 48 h later for analysis of IFN and IFN-response gene mRNA expression.
RT-PCR analysis of type I IFN and 2′5′-oligoadenylate synthetase (OAS) mRNA expression
Pieces of liver tissue were snap frozen by immersion in liquid nitrogen. Total liver RNA was purified using the Totally RNA kit (Ambion). Genomic DNA contamination was removed by treatment with 50 U of RNase-free DNase (Promega), then RNA was purified by phenol/chloroform extraction and precipitation in acidified sodium acetate and ethanol. RNA was extracted from BMDC using a RNeasy mini kit (Qiagen) with on-column genomic DNA clean-up. cDNA synthesis was performed using the first strand cDNA synthesis kit for RT-PCR (Roche). Two microliters of the reverse transcriptase product was used for PCR with primers specific for IFN-β or 2′5′-OAS (30) and β-actin sense, (5′-TGACGGGGTCACCCACACTGTGCCCATCTA-3′) and antisense, (5′-TAGAAGCATTGCGGTGGAGCATGGAGGG-3′) (MWG Biotech) using TaqDNA polymerase (Promega). PCR products were separated and analyzed on a 2% agarose gel. IFN and 2′5′-OAS mRNA levels were calculated as a percentage of β-actin gene expression, which was determined using Quantity-One software (Bio-Rad).
The statistical significance of differences in the responses of groups of mice was assessed using an unpaired Student’s t test.
Coinfection of S. mansoni-infected mice with LCMV leads to a dramatic increase in morbidity, development of which correlates with hepatic egg burden.
Human schistosome and hepatitis virus coinfections are often associated with an increase in morbidity and mortality (15, 31). To address whether this would be reproduced in a murine model of S. mansoni and LCMV coinfection, mice were infected with either 25 or 50 S. mansoni cerceriae; then 10 wk later, after egg laying and induction of Th2-type cytokine-driven granuloma formation, mice were coinfected with the hepatotropic LCMV isolate WE2.2. Mice were then monitored over a 2-wk period for development of clinical signs. Animals that had been infected with 25 S. mansoni cerceriae showed no overt clinical signs at 10 wk p.i., and did not exhibit any change in clinical status over the next fortnight. Mice infected with LCMV only showed mild clinical signs (slight piloerection and hunching) 7–9 days p.i., when the virus-specific CD8 T cell response is known to reach its peak. No singly infected animals required humane sacrifice before the termination of the experiment. In contrast, S. mansoni-infected mice coinfected at 10 wk p.i. with LCMV started to display clinical signs including hunching and piloerection at ∼4 days post-LCMV infection, and increasing numbers of mice needed to be humanely sacrificed due to their severity of disease from day 5, reaching a peak at day 7 p.i. (Fig. 1,A). Analysis of the schistosome egg burden in the liver of coinfected mice requiring sacrifice at different times post-LCMV infection revealed a striking correlation between the hepatic egg burden and development of morbidity (Fig. 1 B). Mice that succumbed to the coinfection in the first 5–7 days had a significantly higher egg burden (p = < 0.0052) than mice that survived to day 14 post-LCMV infection.
Morbidity in coinfected mice is associated with hepatotoxicity
Serum AST levels were found to be significantly higher in the coinfected group of animals than in the schistosome-only and LCMV-only infected groups (Fig. 2,A), and serum ALT levels in the former animals were also higher (data not shown). The high serum AST and ALT levels in coinfected mice indicated a substantial degree of liver damage. Notably, the nitrite ion content (a marker of NO production) of the serum of the coinfected group of mice was also elevated compared with that in the schistosome-only and LCMV-only infected groups of mice (Fig. 2 B), suggesting that the production of active NO radicals may have played a role in mediating the liver pathology observed in the coinfected mice.
Histological analysis of liver sections from coinfected and singly infected mice (Fig. 2, C–J) confirmed the existence of substantial liver pathology in coinfected animals. In mice infected with LCMV only, mild hepatitis was evident on day 7 p.i., with small perivascular cuffs and scattered patches of inflammatory cell infiltration within the liver parenchyma (Fig. 2,D). Analysis of liver sections from schistosome-only infected mice revealed characteristic well-organized granulomas surrounding schistosome eggs, with little inflammation elsewhere in the liver parenchyma (Fig. 2, C, G, and H). In coinfected mice with a low schistosome egg burden (∼5000–12,000 eggs/g), which survived until the end of the 14-day experiment, the granuloma architecture was generally unaffected, although there was more pronounced inflammatory cell infiltration within the parenchyma than in schistosome-only infected mice. Strikingly however, in coinfected mice with a high hepatic egg burden (>25,000 eggs/g), there was evidence of granuloma breakdown, with some granulomas being replaced by areas of necrosis and extensive inflammatory cell infiltration being present throughout the liver parenchyma (Fig. 2, E, F, I, and J).
Collectively, the biochemical data and histological data thus indicated that coinfected mice exhibited a much more severe liver pathology than was observed in animals infected with either pathogen alone, and that the level of hepatotoxicity in coinfected animals was dependent on the liver egg burden.
Th2-type cytokine secretion by lymphocytes from S. mansoni-infected mice is suppressed following coinfection with LCMV
Analysis of the secretion of IL-4, IL-5, IL-10, and IL-13 by splenocyte and IHL cultures from schistosome-only, LCMV-only, and coinfected groups of mice in response to stimulation with either Con A or SEA showed a significant degree of suppression of secretion of all of the Th2 cytokines tested in the coinfected mice (Fig. 3). The differences between cytokine secretion in LCMV-only and coinfected mice were significant (p < 0.05) in both Con A- and SEA-stimulated cultures and for every Th2 cytokine tested. IFN-γ levels in supernatants from Con A- and SEA-stimulated cultures were also evaluated. As expected, cells from LCMV-infected mice produced high levels of IFN-γ when stimulated with Con A, whereas cells from schistosome-only infected mice produced minimal quantities of IFN-γ on stimulation with either Con A or SEA. Cells from coinfected mice produced somewhat more IFN-γ when stimulated with SEA than cells from schistosome-only infected mice, but we did not observe high-level IFN-γ production by SEA-specific T cells from the former animals.
The LCMV-specific T cell response is unaffected by the Th2 cytokine environment
Splenocytes and IHLs from singly and coinfected mice were stimulated ex vivo with peptides corresponding to LCMV CD8 T cell epitopes and IFN-γ production by responding LCMV-specific T cells was assessed by intracellular cytokine staining (Fig. 4). Mice infected with LCMV only were found to have mounted a strong virus-specific CD8 T cell response, with 15–20% of splenic CD8 T cells and 20–30% of liver CD8 T cells producing IFN-γ in response to stimulation with the two immunodominant epitope peptides on day 7 post-LCMV infection. Interestingly, a vigorous LCMV-specific CD8 T cell response was also found to have been primed in coinfected mice, despite the strong Th2 cytokine bias present at the time of infection. In the spleen, the percentage of LCMV-specific IFN-γ-producing cells was approximately equal to that in mice infected with LCMV only (p = 0.147). In contrast, the percentage of LCMV-specific IFN-γ-producing CD8 T cells in the liver of coinfected mice was significantly higher (p = 0.008) than that in mice infected with LCMV only (Fig. 4,B). Analysis of the absolute number of LCMV-specific IFN-γ-producing CD8 T cells in the spleen and liver of these mice verified that there was a higher magnitude virus-specific T cell response in the liver (but not the spleen) of coinfected animals (Fig. 4 C). The magnitude of the LCMV-specific T cell response in the liver and spleen was also analyzed by tetramer staining (data not shown), which confirmed the unexpected observation of an elevated frequency of LCMV-specific T cells in the liver of coinfected mice.
Enhanced early virus replication and impaired clearance of virus from the liver of coinfected mice
To determine whether the magnitude and/or kinetics of LCMV replication were altered in S. mansoni-infected animals, we analyzed viral titers in the spleen and liver over time following LCMV infection of schistosome-infected and control mice. LCMV titers in the spleen of coinfected mice did not differ significantly from those in mice infected with LCMV only at any of the time points tested (Fig. 5). In both groups of mice, high viral titers were present in the spleen on day 2 p.i., and there was evidence of viral clearance occurring between days 5 and 7 p.i., indicative of effective immune control of virus replication at this site. The kinetics of LCMV replication in the liver are delayed compared with those in the spleen. In mice infected with LCMV only, viral titers in the liver were below the level of detection in the plaque assay on day 2 p.i. Viral replication in the liver was apparent by day 5 p.i. and a subsequent decrease in hepatic viral titers was observed by day 7 p.i. (Fig. 5). By contrast, there was enhanced early virus replication in the liver of coinfected mice, with infectious virus being detected on day 2 post-LCMV infection. In addition, containment of viral replication in the liver was impaired at subsequent time points, with liver viral titers in schistosome-infected animals increasing rather than declining between days 5 and 7 p.i., so that the mean hepatic LCMV titer in coinfected animals was on average 3 logs higher than that in mice infected with LCMV by day 7 p.i. There was thus a compartmentalized effect on the early kinetics and subsequent containment of LCMV replication in S. mansoni-infected mice, with the liver but not the spleen being highly susceptible to early virus replication and exhibiting impaired control of subsequent viral replication.
Sites of LCMV replication in the liver
To analyze the location of the sites of viral replication within the liver, confocal microscopy was used (Fig. 6). LCMV-infected cells were identified by staining with a LCMV nucleoprotein-specific mAb (Fig. 6, red). Costaining with an Ab to CD11b, a marker for macrophages and myeloid lineage cells (Fig. 6, green) was used to highlight the location of granulomatous circumoval inflammation in the schistosome-only and coinfected liver sections. In mice infected with LCMV only, very few virus-infected cells were observed in the liver on day 2 p.i. (when infectious viral titers were below the limit of detection by plaque assay). The infected cells were mainly liver resident macrophages, Kupffer cells. In contrast, cells staining for viral Ag were very evident in the liver of coinfected mice at this time point, and were preferentially localized around the outer rim of the schistosome granulomas. Some of the cells infected at this early time point were macrophages (Fig. 6, yellow, overlay confocal image), but many were other cell types. These did not stain for either F4/80 or the eotaxin receptor, CCR3 (data not shown); they may have been hepatocytes and/or fibroblasts in the outer rim of the granuloma. In coinfected mice, LCMV replication was thus promoted in the vicinity of schistosome granulomas in the liver.
Virus was readily detected in the liver of mice infected with LCMV only on day 5 p.i. (data not shown), but by day 7 p.i. it was being cleared. Its presence was only indicated by discrete foci of staining, often at the edge of portal tracts (Fig. 6 B). However, in agreement with the plaque assay data, large amounts of viral Ag were detected in the liver of coinfected mice at day 7 post-LCMV infection. At this late stage of infection there was extensive viral replication throughout the liver parenchyma, the main cellular target probably being hepatocytes. There was also evidence of viral replication deeper within the granulomas, not seen at earlier time points. Notably, the CD11b staining in the granulomas of coinfected livers appeared disorganized, unlike the characteristic circumoval staining seen in the schistosome-only sections, indicative of disruption of the tightly regulated granuloma architecture.
Reduced type I IFN response in the liver of coinfected mice during early LCMV infection
Type I IFN production constitutes a critical component of the early immune response to LCMV infection. These innate cytokines induce many downstream effector genes, such as 2′5′-OAS, which contribute to the establishment of a cellular antiviral state. We thus addressed whether the enhanced early virus replication observed in the livers of coinfected mice was associated with impairment of the type I IFN response at this site. We examined expression of IFN-β mRNA, which is transiently up-regulated following LCMV infection, and the IFN-responsive antiviral gene 2′5′-OAS, a relatively stable marker of type I IFN activity. In mice infected with LCMV only, IFN-β mRNA was up-regulated in the spleen and liver within 24 h of infection (Fig. 7,A). Elevated 2′5′-OAS mRNA levels were observed in the liver on day 2 p.i., and had increased further by day 5 p.i. (Fig. 7,B). By contrast, in coinfected mice there was minimal induction of IFN-β mRNA in the liver on day 1 post-LCMV infection, and intrahepatic 2′5′-OAS mRNA levels were only slightly elevated on day 2, with no further increase being observed by day 5 p.i. At the latter time there was thus a dramatic difference between the levels of 2′5′-OAS mRNA expression in the livers of mice infected with LCMV only and those of coinfected mice (Fig. 7 B). This suggested that the critical innate antiviral response induced by type I IFNs in the liver was suppressed in the coinfected mice, which likely contributed to the enhanced intrahepatic LCMV replication.
Egg Ags suppress type I IFN induction by dendritic cells
As the liver was the only organ that displayed increased viral replication, and there was little impairment of type I IFN mRNA up-regulation in the spleen of coinfected mice (Fig. 7,A), we hypothesized that egg components may be responsible for the observed suppression of innate antiviral responses. To test this hypothesis, we used an in vitro system in which BMDC treated with or without SEA were stimulated with poly(I:C) and the type I IFN response induced was evaluated (Fig. 7 C). Poly(I:C) stimulation of control BMDC resulted in a strong up-regulation of IFN-β and 2′5′-OAS mRNA expression compared with that in unstimulated cultures. Treatment of BMDC with SEA alone also produced a modest increase in both IFN-β and 2′5′-OAS expression. Notably, however, when BMDC were treated with SEA and poly(I:C), induction of both IFN-β and 2′5′-OAS expression was markedly reduced compared with that in cells treated with poly(I:C) alone. Schistosome egg components were thus able to impair poly(I:C)-stimulated induction of the type I IFN pathway.
A number of studies in humans have shown that underlying schistosome infection predisposes to the establishment of persistent infection with hepatotropic viruses, and that the clinical prognosis of coinfected individuals is poorer than that of patients with single infections (15, 31). The murine schistosome and LCMV coinfection model established in this study shows a number of similarities to human schistosome and HCV or HBV coinfections. Notably, coinfected mice displayed a rapid increase in morbidity associated with hepatotoxicity, the level of which was related to the hepatic schistosome egg burden; and hepatic viral replication was markedly enhanced in these animals. Investigation of the underlying mechanisms revealed that the Th2-biased SEA-specific cytokine response characteristic of S. mansoni infection was suppressed in mice coinfected with LCMV, and a strong LCMV-specific CD8 T cell IFN-γ response was induced. The associated disruption of granuloma architecture and increase in levels of toxic mediators such as NO within the liver was likely the cause of the observed hepatotoxicity and morbidity. The enhanced early intrahepatic LCMV replication that occurred in coinfected mice was shown to be associated with suppression of the type I IFN response in the liver. Importantly, SEA was found to suppress poly(I:C)-stimulated up-regulation of the type I IFN pathway in BMDC in vitro, suggesting that the in vivo susceptibility of granulomatous liver tissue to LCMV infection may have been attributable to impaired establishment of the type I IFN-induced antiviral state by schistosome egg products.
Th1- and Th2-type cytokines are known to reciprocally down-regulate the production of one another, with IL-12 and IFN-γ reducing Th2 responses and IL-4 and IL-10 reducing Th1 responses (32). In line with this, most previous studies of the immune response induced to unrelated Ags or pathogens in the context of the Th2-biased cytokine response stimulated by schistosome eggs have reported a domination of the Th2 environment, with suppression of Th1 responses to the incoming agents (24, 33, 34). In contrast, in our coinfection model the response to LCMV appeared to dominate the established Th2-type SEA-specific response, with dramatic suppression of splenocyte production of IL-4, IL-5, IL-10, and IL-13 and the expansion of a strong Th1-biased LCMV-specific T cell response. Notably, large numbers of LCMV-specific IFN-γ-producing T cells were found to be present in the liver, despite the fact that IHL have the highest exposure to SEA. As it has previously been shown that LCMV-specific T cells are recruited into schistosome egg granulomas (35), it is likely that the LCMV-specific IFN-γ-producing T cells had a major influence on the cytokine balance in the liver, particularly within the granulomas. This may have contributed to the breakdown of granuloma organization observed in the livers of coinfected mice.
It was not clear why a strong Th1-type immune response was expanded following LCMV infection of schistosome-infected mice. This contrasts with the outcome of coinfection of schistosome-infected mice with vaccinia virus, which was found to be associated with inhibition of virus-specific T cell responses (24, 25). LCMV replicates to much higher titers than vaccinia virus, particularly within lymphoid tissues; but as Th2 responses are reportedly promoted at high Ag doses, the difference in antigenic load is unlikely to account for the Th1-dominance of the response induced to LCMV. Instead, it is more likely that differences in the innate response stimulated in the two infections, particularly in the interaction of the two viruses with APCs, were critical. Notably, a HIV-1 immunogen administered with IFA to schistosome-infected mice elicited Th2 anti-HIV responses, whereas coadministration of the immunogen with CpG immunostimulatory sequences induced a potent Th1-biased anti-HIV immune response in schistosome-infected mice, suggesting that the mode of activation of APCs through particular TLR may be important in overriding the Th2 environment (36).
Concomitant with the dramatic decrease in Th2 cytokines was a significant increase in morbidity in coinfected mice. Studies in cytokine knockout mice have indicated a protective role for the Th2-type response in schistosome infections, with excessive biasing toward Th1 cytokine responses to SEA correlating with immunopathologically mediated hepatotoxicity (9, 10, 11). In support of these findings, recent human studies have suggested that severe hepatosplenic disease correlates with a Th1-type rather than Th2-type cytokine response to SEA (37). The Th2 cytokines IL-4 and IL-13 play a crucial role in controlling Th1 immunopathology in S. mansoni infections via the induction of alternatively activated macrophages, which have anti-inflammatory properties (38, 39). The Th2 cytokine IL-10 also plays an important role in dampening immunopathologic cytokine production in schistosome-infected mice, controlling both acute Th1-mediated hepatotoxicity and the subsequent fibrotic disease that can be associated with chronic production of high levels of Th2 cytokines (11, 40, 41). It thus seems likely that the suppression of Th2-type cytokine production, disruption of protective immunomodulatory mechanisms, and excessive production of proinflammatory Th1 cytokines contributed to the severe liver pathology observed in our coinfection model.
As in many of the studies described, the acute morbidity in our coinfected mice was associated with increased serum AST and ALT levels, which is indicative of hepatotoxicity. Histological analysis revealed widespread inflammatory cell infiltration throughout the liver parenchyma, and in severe cases (corresponding to high schistosome egg burdens), disruption of the granuloma architecture, with necrosis of surrounding liver tissue. Factors contributing to liver damage may have included activation of innate cells such as macrophages to produce toxic mediators such as NO, serum levels of which were found to be elevated in coinfected mice, and release of hepatotoxic egg proteins due to granuloma breakdown. Notably, correlating with the increased viral burden in the liver, an increased intrahepatic LCMV-specific CD8 T cell response was also evident. Because serum AST levels increased dramatically from day 5 to 7 p.i. (data not shown), coinciding with the peak of the antiviral CD8 T cell response, the evidence strongly implicates immunopathological damage (particularly to infected hepatocytes) as a major contributor to the liver pathology.
Enhanced intrahepatic viral replication is a common feature of schistosome/viral coinfections in both humans and murine models. In one schistosome/HCV coinfection study in Egypt, the hepatic viral burden in coinfected individuals was shown to be 10-fold higher than in individuals infected with HCV only; likewise other studies also suggest that schistosome infection impairs control of HBV and HCV (26, 27, 42, 43, 44). Similarly, in mice coinfected with schistosomes and vaccinia virus, liver viral titers were increased and clearance was significantly delayed, with virus replicating preferentially in granulomas (25). We also observed enhanced LCMV replication in the liver of S. mansoni-infected mice, with confocal microscopy revealing that the early virus replication occurred in regions of liver tissue surrounding the granulomas. A minority of the infected cells surrounding the granulomas were macrophages, costaining with F4/80; but the vast majority of the cells were either fibroblasts or hepatocyte-like cells. This suggests that the inflammatory cells in the granulomas did not facilitate early virus replication in the liver by comprising an infection-susceptible cell population for the virus to target; but rather that the hepatocytes surrounding the granulomas appeared to have become more susceptible to virus infection than hepatocytes in singly infected mice. This mirrors findings made in schistosome/HBV coinfection, in which examination of liver wedge biopsy specimens showed HBV replication to be localized to hepatocytes outside the granulomas (44). Notably, we did not observe any enhancement of LCMV replication in the context of S. mansoni infection in areas of liver tissue devoid of egg deposition; likewise although viral titers in the liver of coinfected mice were 3 logs higher than those in mice infected with LCMV only on day 7 p.i., viral clearance from the spleen was only slightly delayed. There was thus a correlation between tissue egg deposition and susceptibility to enhanced viral replication.
Innate responses, particularly production of type I IFNs, whose functions include up-regulation of genes with antiviral activity, play an important role in limiting LCMV replication at early times p.i. (22, 23). Therefore, the striking impairment we found in the up-regulation of the type I IFN response in coinfected mice likely facilitated rapid LCMV replication and enhanced dissemination throughout the liver tissue. As LCMV titers were only enhanced in the liver in this coinfection model, and early viral replication occurred preferentially in the vicinity of granulomas, it seemed likely that soluble factors released from granulomas may be responsible for the observed suppression of the early type I IFN response in the liver. These may include cytokines, or SEA. Interestingly, we demonstrated a marked suppression of both IFN-β and 2′5′-OAS expression in SEA-treated poly(I:C)-stimulated BMDC. This indicates that schistosome egg components can modulate the type I IFN response. The point in the type I IFN pathway that SEA affects is currently unknown and is under investigation. Interestingly, a recent study on inflammatory gene induction in a dendritic cell-like line after SEA treatment showed that IFN-β was up-regulated, along with the type I IFN-induced effector gene Mx (45). These findings are not at odds with our data, as we also observed some up-regulation of IFN-β mRNA in response to SEA, although this was small compared with the response stimulated by poly(I:C) or viral infection. Our in vitro results thus suggest a potential mechanism for the observed in vivo suppression of the type I IFN response in the granulomatous liver. Studies in HCV-infected chimpanzees have shown that multiple IFN-inducible genes start to be up-regulated in the liver within days of infection (46, 47), and there is direct evidence from a transgenic mouse model of HBV infection to support the importance of type I IFN-induced antiviral effector mechanisms in containment of intrahepatic viral replication (48). This suggests the potential for SEA-mediated suppression of the intrahepatic type I IFN response to have a significant effect on the containment of HCV/HBV replication during human coinfection with schistosomes and these hepatotropic viruses. The observations made in our murine schistosome-LCMV coinfection model thus suggest a novel mechanism that may help to explain the enhancement of viral replication observed during human coinfection with HBV or HCV.
In conclusion, this murine model of schistosomiasis and hepatotropic virus coinfection has many similarities to human schistosome and HCV/HBV coinfections (15, 26, 27, 31, 42, 43). Humans and mice display dramatic increases in morbidity and mortality, associated with rapid progression to liver disease, correlating with increased serum AST and NO levels. We show modulation of the schistosome Ag-induced Th2 response by viral coinfection, with suppressed Th2-like cytokine responses correlating with pathology, in line with many studies indicating the protective nature of Th2 responses in this infection (37). Humans and mice also show a marked increase in hepatic viral burdens, with studies in humans indicating schistosomiasis to be a predisposing factor in establishment of chronic infection with hepatotropic viruses. An important finding from this study is that we have correlated the dramatic increase in hepatic viral burden with suppression of the type I IFN response by SEA. We propose that suppression of the type I IFN pathway sensitizes the liver to viral replication, with ensuing consequences for the efficiency of containment of hepatic viral replication by the adaptive response and associated immunopathologic hepatitis.
We thank Professor M. Doenhoff for provision of schistosome egg Ags.
The authors have no financial conflict of interest.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
This work was supported by core funding from The Edward Jenner Institute. This is The Edward Jenner Institute for Vaccine Research publication number 93.
Abbreviations used in this paper: SEA, schistosome egg Ag; BMDC, bone marrow-derived dendritic cell; AST, aspartate transaminase; HBV, hepatitis B virus; HCV, hepatitis C virus; ALT, alanine transaminase; IHL, intrahepatic lymphocyte; LCMV, lymphocytic choriomeningitis virus; OAS, oligoadenylate synthetase; p.i., postinfection.