Tissue macrophages suppress viral replication and prevent severe immunopathology in an interferon-I-dependent manner in mice

Authors


  • Potential conflict of interest: Nothing to report.

Abstract

The innate immune response plays an essential role in the prevention of early viral dissemination. We used the lymphocytic choriomeningitis virus model system to analyze the role of tissue macrophages/Kupffer cells in this process. Our findings demonstrated that Kupffer cells are essential for the efficient capture of infectious virus and for preventing viral replication. The latter process involved activation of Kupffer cells by interferon (IFN)-I and prevented viral spread to neighboring hepatocytes. In the absence of Kupffer cells, hepatocytes were not able to suppress virus replication, even in the presence of IFN-I, leading to prolonged viral replication and severe T cell-dependent immunopathology. Conclusion: Tissue-resident macrophages play a crucial role in early viral capture and represent the major liver cell type exhibiting responsiveness to IFN-I and providing control of viral replication. (HEPATOLOGY 2010)

Persistent infection with hepatitis B or hepatitis C virus is one of the leading causes of lethal liver disease resulting from the development of liver cirrhosis and/or hepatocellular cancer.1 Because both viruses are poorly cytopathic, most of the ensuing liver destruction results from CD8+ T cell responses directed against virus-infected hepatocytes. These cells prevent rapid dissemination of the virus and control viral replication by secreting interferon (IFN)-γ and perforin. However, in the event of viral persistence, exaggerated and prolonged T cell activation results in severe liver pathology which can eventually be fatal.2, 3 Thus CD8+ T cells play a crucial role in both virus control and immunopathology.4-8

Macrophages are resident in every organ of the body.9-11 With their potent phagocytic capacity, they act as a first line of defense between pathogens entering tissues through the bloodstream and parenchymal cells. Kupffer cells in the liver and splenic marginal zone macrophages share a common feature in that they are both associated with the endothelium and can capture antigens from the blood vessel lumen.9, 12 Thus, these cells are ideally situated to clear immune complexes and high molecular and complex particles, including pathogens and apoptotic cells from the circulation,13-18 and constitute part of the reticulo-endothelial system.11, 19 Recent data have indicated that tissue macrophages play an essential role in virus control13; however, the role of these cells in modulating immunopathology and liver disease remains unclear.

In this study, we analyzed the role of tissue macrophages during systemic virus infection, using the well-characterized model system of murine lymphocytic choriomeningitis virus (LCMV) infection. Our data demonstrated that an absence of Kupffer cells results in rapid viral dissemination and replication within hepatocytes leading to severe CD8+ T cell-mediated immunopathology and liver damage. The crucial role of Kupffer cells during viral control correlated with an enhanced ability of these cells to respond to IFN-I. Lack of IFN-I receptors on Kupffer cells mimicked macrophage depletion and resulted in rapid viral spread to hepatocytes and impaired control of viral replication, which was followed by a severe CD8 T cell-dependent liver cell damage.

Abbreviations:

ALT, alanine aminotransferase; IFN, interferon; IFNAR, IFN-α receptor; LCMV, lymphocytic choriomeningitis virus; LCMV-NP, lymphocytic choriomeningitis virus nucleoprotein.

Materials and Methods

Mice and Viruses.

The LCMV strain WE was originally obtained from F. Lehmann-Grube (Heinrich Pette Institute, Hamburg, Germany) and was propagated in L929 cells. LCMV Armstrong was obtained from M. Buchmeier (Scripps Clinic and Research Center, La Jolla, CA). Virus titers were measured using a focus-forming assay as described.20 Mice were infected with LCMV doses as indicated. All mice used in this study were maintained on the C57BL/6 genetic background. All experiments were performed in single ventilated cages. If mice showed symptoms indicating undue stress or likely death, they were euthanized. Animal experiments were performed with authorization of the Veterinäramt of the Kanton Zurich and in accordance with the Swiss laws for animal protection and according to institutional guidelines at the Ontario Cancer Institute.

Depletion of CD8+ T Cells and Macrophages.

CD8+ T cells were depleted by intraperitoneal or intravenous injection of a rat monoclonal antibody specific for mouse CD8 (clone YTS169.4) on days −2 and −1 before infection. For depletion of macrophages mice were treated with 200 μL of clodronate-liposomes. Control mice were treated with 200 μL of empty control-liposomes. Clodronate was a gift of Roche Diagnostics, Mannheim, Germany. It was encapsulated in liposomes as described.21

Isolation of Hepatocytes and Macrophages.

Hepatocytes were isolated according to the protocol of Graf et al.22 In brief, mice were anesthetized and a vein catheter was laid into the portal vein. The heart and aorta were opened, and the liver was perfused for 2 minutes with phosphate-buffered saline, followed by collagenase A. After 2 minutes of digestion, hepatocytes were isolated. Isolated hepatocytes were cultured for 36 hours and then analyzed by way of reverse-transcription polymerase chain reaction. Macrophages were isolated by flushing the peritoneum with 10 mL of cell culture medium. Macrophages were cultured for 36 hours and then used for reverse-transcription polymerase chain reaction analysis.

Histological Analysis.

Histological analysis were performed as described23 using monoclonal antibodies against mouse macrophages (F4/80, BD Biosciences, eBioscience), mouse hepatocytes (anti-insulin receptor-β, BD Biosciences), and LCMV nucleoprotein (LCMV-NP) (clone VL4, clone KL53). Staining was developed using alkaline phosphatase, peroxidase or fluorescent dyes (secondary antibodies or direct labeling of antibodies).24 Scale bars represent 25 μm unless mentioned otherwise.

Bilirubin, Alkaline Phosphatase and Alanine Aminotransferase.

Bilirubin, alkaline phosphatase, and alanine aminotransferase (ALT) levels were measured using a serum multiple biochemical analyzer (Ektachem DTSCII, Johnson & Johnson).

Enzyme-Linked Immunosorbent Assay.

IFN-α enzyme-linked immunosorbent assays were performed according to the manufacturer's instructions (Research Diagnostics).

Messenger RNA Gene Profiling by Way of Quantitative Reverse-Transcription Polymerase Chain Reaction.

RNA extraction and complementary DNA synthesis was performed using Trizol. Gene expression analysis was performed using kits from Applied Biosystems. For analysis, the expression levels of all target genes were normalized against glyceraldehyde 3-phosphate dehydrogenase (ΔCt). Gene expression values were then calculated based on the ΔΔCt method, using the mean of control group as calibrator to which all other samples were compared. Relative quantities (RQ) were determined using the equation RQ = 2−ΔΔCt.

Fluorescence-Activated Cell Sorting Analysis.

Tetramer staining, surface staining, and intracellular fluorescence-activated cell sorting were performed as described.20 Tetramers were house-made or provided by the National Institutes of Health tetramer facility.

Statistical Analysis.

Data are expressed as the mean ± standard error of the mean. Statistically significant differences between two different groups were analyzed using the Student t test. Analysis including several groups were performed using one-way analysis of variance with additional Bonferroni or Dunnett test. Statistically significant differences between experimental groups over multiple time points were calculated using two-way analysis of variance (repeated measurements). P values <0.05 were considered statistically significant.

Results

Macrophages Play a Crucial Role in Preventing Viral-Induced Liver Pathology.

To investigate the role of macrophages during noncytopathic viral infection, we depleted macrophages by treating mice with clodronate-filled liposomes21 1 day before infection with LCMV strain WE. Control mice, treated with empty liposomes, failed to show any signs of acute liver disease during LCMV infection, whereas macrophage-depleted mice exhibited severe liver cell damage as evidenced by highly elevated serum ALT and bilirubin levels (Fig. 1A,B). Liver damage was a direct result of the viral infection rather than reflecting possible hepatotoxic effects of clodronate, as mice receiving clodronate alone did not exhibit elevated ALT (Fig. 1A).

Figure 1.

Macrophage depletion results in severe hepatitis. C57BL/6 mice were treated with clodronate-filled liposomes to deplete macrophages. Control mice were treated with empty control liposomes (L) on day −1. (A) On day 0, mice were infected with 2 × 104 PFU of LCMV-WE. One group of clodronate-treated mice remained noninfected. Sera from all mice were analyzed for ALT activity (n = 4, one of two representative experiments is shown). Two out of four mice from the clodronate-treated, LCMV-infected group died after day 10. (B) In a separate experiment, sera from LCMV-infected clodronate-filled mice (CL-L) and control mice were analyzed for the level of bilirubin (n = 3, one of two representative experiments is shown).

Kupffer Cells Capture Viral Particles at Early Time Points Following Infection.

To further elucidate the mechanisms by which Kupffer cells prevent viral-induced liver pathology, we analyzed viral capture and dissemination at early time points following infection. Mice were treated with empty or clodronate-filled liposomes, then given an intravenous infection with 5 × 106 PFU LCMV 1 day later. Both control- and clodronate-treated mice exhibited peak viral titers in the blood 1 minute after infection (distribution phase) (Fig. 2A). After 1 hour, viral titers decreased to below the detection limit in control mice but remained high in clodronate-treated animals, indicating that macrophages normally function to capture the virus at early time points following infection (Fig. 2A). Similar to the lack of virus uptake in the blood, the spleen, liver, lung, and kidneys all exhibited increased levels of infectious viral particles in the absence of macrophages (Fig. 2B). To determine the location of the macrophages responsible for viral uptake, we used immunohistochemistry to analyze tissue sections of the spleen, liver, lung, and kidney for presence of LCMV-NP and macrophages. LCMV-NP was determined to colocalize exclusively with F4/80+ (macrophage marker) cells within the spleen and liver (Fig. 2C,D). These data suggest that Kupffer cells and splenic macrophages were largely responsible for viral uptake, no doubt reflecting their location within the bloodstream. Most of the Kupffer cells were sinusoidal, and a subpopulation of these stained positively for LCMV-NP with a random distribution pattern (Supporting Fig. 1). To determine the relative contribution of splenic macrophages versus Kupffer cells in viral capture, we next infected splenectomized mice. Splenectomized mice exhibited normal viral clearance (Fig. 2E), indicating that macrophages within other tissues could compensate for the lack of splenic phagocytes. We next determined the role of lymph node-resident macrophages by infecting splenectomized aly/aly mice, which lack peripheral lymph nodes.25 aly/aly Mice cleared the virus at the same rate as normal or splenectomized C57BL/6 mice (Fig. 2F). Kupffer cells represent the dominant phagocyte remaining within the reticulo-endothelial system of splenectomized aly/aly mice. Our data therefore indicate that Kupffer cells are both necessary and sufficient for the capture of infectious viral particles at early time points following intravenous LCMV infection.

Figure 2.

Kupffer cells capture virus at early timepoints following infection. C57BL/6 mice were treated with clodronate or control-liposomes on day -1. After 24 hours mice were infected intravenously with 5 × 106 PFU of LCMV-WE. (A) Blood was collected 0.5, 10 and 60 minutes following infection and analyzed for infectious virus using a virus plaque forming assay (n = 4-8). (B) Virus PFUs were analyzed in the spleen, liver, lung and kidney after 0.5, 10, and 60 minutes (n = 8). (C) C57BL/6 mice were infected intravenously with 2 × 104 PFU of LCMV-WE or were left untreated. Tissues were sectioned and stained for F4/80 (a macrophage marker) and LCMV-NP at day 4 following infection (n = 3). (D) C57BL/6 mice were infected intravenously with 2 × 106 PFU of LCMV-WE. On day 2 after infection, frozen liver sections were costained for F4/80 (blue) and LCMV-NP (red). (E) C57BL/6 mice were splenectomized. Seven days later, splenectomized mice were treated with clodronate-filled liposomes and infected intravenously with 5 × 106 PFU of LCMV-WE 24 hours later. Organs were analyzed for infectious virus 60 minutes later (n = 3). (F) C57BL/6 mice and aly/aly mice were splenectomized. aly/aly Mice have no peripheral lymph nodes but do have Kupffer cells (data not shown). Seven days after splenectomy, splenectomized mice, clodronate-treated C57BL/6 mice and untreated C57BL/6 mice were infected intravenously with 5 × 106 PFU of LCMV-WE. After 1.5, 10, and 60 minutes, blood viral titers were analyzed using a plaque-forming assay (n = 4).

Macrophages Play a Crucial Role in Controlling Viral Replication During LCMV Infection.

Because Kupffer cells were found to play a key role during early viral capture, we next investigated the ability of these cells to control viral replication. For this purpose we infected clodronate-treated C57BL/6 mice with 2 × 104 PFU of LCMV and analyzed liver histology 4 days later. In control liposome-treated mice, LCMV-NP was almost exclusively localized within cells exhibiting Kupffer cell morphology (Fig. 3A). In contrast, clodronate-treated mice lacking Kupffer cells exhibited widespread viral dissemination and viral localization within hepatocytes (Fig. 3A). These data were confirmed by way of immunohistological studies showing colocalization of LCMV-NP within F4/80 (Kupffer cell marker) or insulin receptor-β (hepatocyte marker)-positive cells of control of clodronate-treated mice, respectively (Supporting Fig. 2). In the presence of macrophages, infectious virus could not be detected within the liver after 2 days (Fig. 3B). In contrast, macrophage depletion by clodronate-filled liposomes resulted in prolonged and significantly enhanced viral replication at this site (Fig. 3B). Thus, Kupffer cells functioned not only to capture blood-borne viral particles but additionally prevented viral replication and dissemination to neighboring hepatocytes.

Figure 3.

Enhanced LCMV replication in livers of macrophage-depleted mice. C57BL/6 mice were treated with clodronate-filled liposomes (CL-L) to deplete macrophages. Control mice were treated with empty control liposomes (L) on day −1. (A) On day 0, mice were infected intravenously with 2 × 104 PFU LCMV-WE and 4 days later livers were stained for LCMV-NP by immunohistological analysis (n = 6). (B) Mice were infected intravenously with 200 PFU of LCMV-WE, and viral titers were analyzed in the spleen and liver (n = 6).

Kupffer Cells, but Not Hepatocytes, Can Control Viral Replication in an IFN-I-Dependent Manner.

The finding that macrophage-depleted mice exhibited increased viral replication within the liver indicated that hepatocytes are less able to control viral replication than Kupffer cells. To further investigate this hypothesis, we treated IFN-α receptor (IFNAR)-deficient mice (ifnar−/− mice) or wild-type mice with macrophage-depleting clodronate-filled liposomes prior to infection with LCMV. Wild-type mice were able to effectively contain viral replication in the liver, lung, and kidney, whereas both macrophage-depleted and IFNAR-deficient mice exhibited strongly enhanced viral titers within all organs (Fig. 4A). These data suggest that macrophages were either the predominate cell type responsible for IFN-I production or were the only cell type responsive to this cytokine. Clodronate treatment was observed to have little impact on IFN-I serum levels (Fig. 4B), and infection of IFN-β/YFP reporter (IFN-βmob/mob) mice26 showed absence of IFN-I production in macrophages (Supporting Fig. 3). To further investigate the IFN-I responsiveness of macrophages versus hepatocytes, we infected mice lacking IFNAR specifically on macrophages (Ifnar1fl/fl LysMCre mice)27, 28 and compared their ability to control viral replication to mice lacking IFNAR on all cell types.29Ifnar1fl/fl LysMCre mice exhibited viral titers comparable with mice lacking global IFNAR expression (Fig. 4C), confirming the hypothesis that macrophages represent the major cell present within the liver, which can respond to IFN-I and limit viral replication. To further elucidate the mechanisms for the selective responsiveness of macrophages to IFN-I, we compared the gene expression of interferon signaling components in macrophages versus hepatocytes isolated from the liver. Interestingly, hepatocytes expressed limited levels of messenger RNA coding for the IFN-I receptor, Stat1, and Stat3 when compared with macrophages. Because these gene products are all involved in IFN responsiveness and signal transduction, their limited expression within hepatocytes likely explains the crucial role of macrophages during IFN-I-mediated control of viral replication within the liver.

Figure 4.

Kupffer cells control viral replication in an IFN-I-dependent manner. (A) Wild-type mice and Ifnar−/− mice were treated with clodronate-liposomes to deplete macrophages. Control mice were treated with control liposomes. All mice were infected with 2 × 106 PFU of LCMV-WE 24 hours later, and organ viral titers were analyzed on day 2 following infection (n = 3). (B) C57BL/6 mice were treated with clodronate-filled or control liposomes on day −1. On day 0, mice were infected intravenously with 200 PFU of LCMV-WE, and serum IFN-α was analyzed 2 days later (n = 8). (C) Mice lacking IFNAR expression exclusively on macrophages (Ifnar1fl/fl LysMCre), wild-type mice, and Ifnar−/− mice were infected with 2 × 104 PFU of LCMV-WE. Virus titers were analyzed on day 3 after infection for the indicated organs (n = 3). (D) Macrophages and hepatocytes were isolated from C57BL/6 mice as described in Materials and Methods. Expression of IFN signaling components were analyzed (n = 6).

Enhanced Viral Replication in the Absence of Macrophages Results in Immunopathology Mediated by CD8 T Cells.

We next wanted to determine the exact cause of liver damage following LCMV infection of macrophage-depleted mice. Because liver pathology during LCMV infection is generally recognized to require CD8 T cells, we next investigated the impact of macrophage depletion on LCMV-induced CD8 T cell responses. Clodronate treatment had little impact on either the expansion of virus-specific CD8+ T cells (Fig. 5A) or their ability to produce IFN-γ (Fig. 5B). CD8 T cells primed in the absence of macrophages were also able to normally migrate to livers of virus-infected mice (Fig. 5C). Moreover, CD8 T cells appeared to be largely responsible for mediating liver immunopathology in macrophage-depleted mice, as treatment of such mice with a depleting anti-CD8 monoclonal antibody was able to prevent liver pathology (Fig. 5D). Natural killer and natural killer T cells may also be expected to contribute to liver pathology30-33; however, depletion of these cells in clodronate-treated mice with the PK136 monoclonal antibody did not prevent liver cell damage (Supporting Fig. 4). Taken together, these data indicate that Kupffer cells are the predominant cell type present within the liver responsible for viral control, and that in their absence enhanced viral dissemination and replication results in severe CD8 T cell-mediated immunopathology and liver damage.

Figure 5.

Enhanced virus replication in the absence of macrophages leads to increased liver pathology mediated by CD8 T cells. C57BL/6 mice were treated with clodronate or empty liposomes on day −1. (A) On day 0, mice were infected intravenously with 2 × 104 PFU LCMV-WE. LCMV-specific CD8+ T cell priming in the spleen was analyzed using tetramers (n = 3). (B) Splenocytes were restimulated with LCMV-derived peptides (GP33 and NP396), and intracellular IFN-γ was analyzed 6 hours later (n = 3). (C) On day 0, mice were infected intravenously with 2 × 104 PFU of LCMV-WE. Virus-specific CD8+ T cells were quantified by way of tetramer staining 8 days after infection in the livers of macrophage-depleted and splenectomized macrophage-depleted mice (n = 3). (D) One group of clodronate-treated mice was additionally depleted of CD8+ T cells through administration of anti-CD8 monoclonal antibody. Mice were then infected intravenously with 2 × 104 PFU LCMV-WE, and ALT activity and bilirubin concentration were analyzed in the serum (n = 3). Significance was determined between total curves using two-way analysis of variance.

Discussion

Our data have revealed a crucial role for Kupffer cells in the early capture of blood-borne viruses and the later control of viral replication within the liver. Virus capture by splenic macrophages and Kupffer cells occurred very rapidly (10–60 minutes) after infection, before IFN-I could be expected to have exerted its function. However, viral capture alone was not sufficient for viral control, as evidenced by enhanced viral replication in IFNAR-deficient mice. IFN-I-dependent control of viral replication in the liver occurred predominately within Kupffer cells, because mice lacking IFNAR exclusively on Kupffer cells exhibited the same lack of viral control as mice exhibiting a global defect in IFNAR expression. Importantly, induction of IFN-I did not require Kupffer cells. Instead, Kupffer cells represented the sole cell type present in the liver, which can respond effectively to IFN-I. The absence of macrophages led to viral spread to hepatocytes and resulted in a lack of viral control as a consequence of an inability of these cells to respond to IFN-I. The inability of hepatocytes to respond to IFN-I could be explained by the finding that hepatocytes failed to express significant levels of messenger RNA encoding IFNAR and downstream signaling molecules compared with macrophages.

Liver endothelial cells are equipped with Toll-like receptors and are involved in the antiviral innate immune response,34, 35 including uptake of serum particles.15 In addition to cytokine production, liver sinusoid endothelial cells show a strong antiviral response to IFN-I,34, 36 which could suppress virus replication37; however, whether this is of benefit for a systemic virus infection remains to be studied. We found that most of the LCMV was captured by Kupffer cells. Lack of those cells led to infection of hepatocytes.

We found that Kupffer cells take up virus and limit virus replication. This protects from fast virus propagation and immunopathology. The role of Kupffer cells during chronic virus infection in humans needs to be further analyzed. Kupffer cells are virtually uninfected by human hepatitis C virus38, 39; therefore, their role during HCV infection remains unclear. Hepatitis B virus and human immunodeficiency virus can be taken up by macrophages or Kupffer cells.40-43 Although macrophages are involved in the progression of acquired immune deficiency syndrome, their early role in limiting virus propagation remains unclear.

In conclusion, tissue Kupffer cells are both sufficient and necessary for the capture of blood-borne virus particles at early time points following systemic infection. These cells exhibit an exclusive responsiveness to IFN-I leading to improved control of viral replication. This dual antiviral strategy of “capture and famish” controls viral dissemination within the liver, thereby preventing excessive immunopathology mediated by the adaptive immune system.

Acknowledgements

We thank Alisha Elford for technical support. We also thank the National Institutes of Health tetramer facility for providing tetramers.

Ancillary