IFN-α/β-mediated functions promote production of MIP-1α (or CCL3) by mediating the recruitment of MIP-1α-producing macrophages to the liver during early infection with murine CMV. These responses are essential for induction of NK cell inflammation and IFN-γ delivery to support effective control of local infection. Nevertheless, it remains to be established if additional chemokine functions are regulated by IFN-α/β and/or play intermediary roles in supporting macrophage trafficking. The chemokine MCP-1 (or CCL2) plays a distinctive role in the recruitment of macrophages by predominantly stimulating the CCR2 chemokine receptor. Here, we examine the roles of MCP-1 and CCR2 during murine CMV infection in liver. MCP-1 production preceded that of MIP-1α during infection and was dependent on IFN-α/β effects for induction. Resident F4/80+ liver leukocytes were identified as primary IFN-α/β responders and major producers of MCP-1. Moreover, MCP-1 deficiency was associated with a dramatic reduction in the accumulation of macrophages and NK cells, as well as decreased production of MIP-1α and IFN-γ in liver. These responses were also markedly impaired in mice with a targeted disruption of CCR2. Furthermore, MCP-1- and CCR2-deficient mice exhibited increased viral titers and elevated expression of the liver enzyme alanine aminotransferase in serum. These mice also had widespread virus-induced liver pathology and succumbed to infection. Collectively, these results establish MCP-1 and CCR2 interactions as factors promoting early liver inflammatory responses and define a mechanism for innate cytokines in regulation of chemokine functions critical for effective localized antiviral defenses.

Murine CMV (MCMV)4 is a hepatotropic herpesvirus that induces innate cytokine and chemokine responses critical for survival against a viral infection establishing itself in tissues (1, 2). Specifically, the chemokine MIP-1α (or CCL3) mediates the recruitment of inflammatory NK cells for effective delivery of IFN-γ responses, which are critical in early defense against MCMV infection in the liver (2, 3, 4, 5, 6, 7, 8). Previous studies demonstrated that the type 1 IFNs, IFN-α/β, were produced in the liver and contributed significantly to MCMV defense (9). Furthermore, IFN-α/β immunoregulatory responses elicited the initiation of local MIP-1α production that is essential for NK cell inflammation. The cytokines mediated their effects by triggering the recruitment of MIP-1α-producing macrophages to the liver (9). The mechanisms promoting IFN-α/β-mediated migration of inflammatory macrophages under the conditions of MCMV infection have not been established. It is uncertain if additional chemokine responses are regulated by IFN-α/β and/or play an intermediary role in macrophage recruitment.

MCP-1 (or CCL2) has been shown to be an effective promoter of macrophage infiltration in various inflammatory models of disease and host defense against several pathogens (10, 11, 12, 13, 14, 15, 16, 17, 18, 19). These responses are mediated by the exclusive binding of MCP-1 to the chemokine receptor CCR2 (20, 21, 22, 23, 24, 25). The contributions of MCP-1 and/or CCR2 functions in macrophage recruitment to liver during MCMV infection have not been evaluated. Moreover, the in vivo role of innate cytokines on MCP-1 induction in liver remains to be determined.

The studies presented here were undertaken to define a role for IFN-α/β in regulation of MCP-1 production and to establish the contribution of MCP-1 and CCR2 functions on macrophage and NK cell accumulation and antiviral defense. The results indicate that MCP-1 is produced upstream of MIP-1α in the liver during MCMV infection and that IFN-α/β is a prominent factor eliciting induction of MCP-1. Resident F4/80+ liver leukocytes are shown to respond to IFN-α/β stimulation for production of MCP-1. Furthermore, MCP-1 and CCR2 effects are required for optimal macrophage recruitment and MIP-1α production, as well as NK cell accumulation, IFN-γ expression, and increased resistance to MCMV-induced liver disease and death. Taken together, these results establish an immunoregulatory function for IFN-α/β on chemokine responses that are critical for inflammation and promotion of effective antiviral defenses in infected tissues.

Specific pathogen-free C57BL/6 mice were purchased from Taconic Laboratory Animals and Services or The Jackson Laboratory. 129/Sv mice were purchased from Taconic Laboratory Animals and Services. 129-IFN-α/βR-deficient mice (26) were originally obtained from B&K Universal and maintained at Brown University. C57BL/6-CCR2-deficient (22) breeding pairs were obtained from Dr. W. Kuziel (University of Texas, Austin, TX) and used to establish colonies at Brown University. C57BL/6-MCP-1-deficient breeding pairs were purchased from The Jackson Laboratory. Male and female mice were used between 5 and 6 wk of age. Mouse handling and experimental procedures were conducted in accordance with institutional guidelines.

Infections were initiated on day 0 by i.p. injections of 5 × 104 PFU of MCMV v70 Smith strain (9). In survival experiments, mice were infected with 105 PFU of MCMV and assessed for mortality twice daily. Animals treated in vivo with recombinant human IFN-αA/D (rIFN-α; Pestka Biomedical Laboratories), biologically active in mice, were given three i.p. daily injections of 1 × 105 U in PBS containing 0.1% BSA. Control animals were given vehicle injections. Responses were evaluated 24 h after the third treatment.

Liver leukocytes were prepared as described previously (7, 9, 27). Viable cell yields were enumerated by trypan blue exclusion (Life Technologies). Proportions and numbers of cells were evaluated as reported previously (7, 9). Cell surface staining was performed as previously described (7, 9) using the following mAbs: F4/80-PE (Serotec) and CD11b-FITC (BD Biosciences) to distinguish migrating from resident macrophages (28) and NK1.1-PE and TCRβ-FITC (BD Biosciences). All Abs were diluted in the presence of mAb 2.4G2 (FcγII/IIIR; BD Biosciences) to block nonspecific binding of Abs to FcRs. Control Abs not recognizing specific murine determinants (BD Biosciences) were used to correct for background fluorescence and set analyses gates. Cells were acquired and analyzed with a FACSCalibur and CellQuest software (BD Biosciences).

Pooled liver leukocytes were resuspended at ≤1 × 107 cells/ml and blocked with 2.4G2 Ab. Cells were incubated additionally with PE-labeled F4/80 (Serotec) and positively selected using anti-PE magnetic beads (Miltenyi Biotec) and an Automacs machine according to the manufacturer’s protocol (Miltenyi Biotec). Where indicated, macrophage populations were purified from total PE-labeled F4/80+ cells by FACS using a FACSVantage (BD Biosciences) with a 70-μm nozzle. Cells negative for the selection marker were also retrieved. Purity and viability were analyzed immediately after sorting. F4/80+ cells were usually enriched by at least 70%.

Total liver leukocytes from uninfected mice or from sorted cells were cultured in 96-well microtiter plates with or without various amounts of rIFN-αA/D (Pestka Biomedical Laboratories) as described before (29). After 24 h of incubation, microtiter plates were centrifuged, and cell supernatants were collected.

Culture supernatants after in vitro stimulation, liver homogenates, or liver leukocyte-conditioned media, prepared by previously described methods (8, 9), were tested for MCP-1, MIP-1α, or IFN-γ using commercial sandwich ELISA (R&D Systems).

Median liver lobes were fixed in 10% buffered formalin and paraffin embedded for analyses. Tissue sections were cut (5 μm) and stained with H&E for microscopic observation. Overall liver damage was determined by measuring alanine aminotransferase (ALT) levels in serum samples using an ALT colorimetric kit according to manufacturer’s instructions (Biotron Diagnostics).

Organs were weighed, homogenized in supplemented RPMI 1640 media (Life Technologies), and supernatants were collected after centrifugation. Duplicate samples of serially diluted supernatants were incubated on NIH 3T3 cells (American Type Culture Collection), and viral titers were determined by standard plaque assay as described previously (6, 7).

Statistical significance of experimental results was analyzed by two-tailed Student’s t test where indicated.

The kinetics of MCP-1 and MIP-1α production were examined after MCMV infection of 129 mice. As measured by ELISA, uninfected mice did not have detectable levels of MCP-1 and MIP-1α in liver. During infection, MCP-1 induction became evident as early as 24 h and rose sharply in a time-dependent manner (Fig. 1,A). Maximal values of MCP-1 protein were observed between 32 and 40 h and declined 70% by 48 h of infection. In contrast, MIP-1α levels were below the limits of detection before 32 h and gradually became elevated, reaching peak production by 40 h of infection. Thereafter, the magnitude of MIP-1α protein declined but remained extended through 48 h of infection (Fig. 1 B). These results demonstrate that MCMV induces significant levels of MCP-1 and MIP-1α in a temporal fashion in liver, whereby MCP-1 production is before that of MIP-1α at this site.

FIGURE 1.

Kinetics of MCP-1 and MIP-1α during MCMV infection in liver. Liver homogenates were prepared from 129 mice that were uninfected or infected with MCMV at the indicated time points. MCP-1 (A) and MIP-1α (B) protein levels were measured by standard sandwich ELISA. The levels of detection for MCP-1 and MIP-1α were 0.08 and 0.02 ng/g of liver, respectively. Data are the means ± SE (n = 3–4 mice tested individually for each time point). Results are representative of one of two experiments with similar results.

FIGURE 1.

Kinetics of MCP-1 and MIP-1α during MCMV infection in liver. Liver homogenates were prepared from 129 mice that were uninfected or infected with MCMV at the indicated time points. MCP-1 (A) and MIP-1α (B) protein levels were measured by standard sandwich ELISA. The levels of detection for MCP-1 and MIP-1α were 0.08 and 0.02 ng/g of liver, respectively. Data are the means ± SE (n = 3–4 mice tested individually for each time point). Results are representative of one of two experiments with similar results.

Close modal

During early MCMV infection, IFN-α/β protein can be induced in liver (9). Moreover, the cytokines have been shown to indirectly influence MIP-1α production through promotion of macrophage trafficking events (9). To assess the contribution of IFN-α/β-dependent functions on MCP-1 production, a known chemoattractant for macrophages, ELISAs were performed with liver homogenates prepared from uninfected or day 1.5 MCMV-infected 129 mice that were immunocompetent (wild type (WT)) or deficient in IFN-α/β functions (IFN-α/βR) as a result of gene disruption. The results show induction of MCP-1 protein in livers of WT and IFN-α/βR mice during infection (Fig. 2 A). However, the induced responses observed in IFN-α/βR mice were considerably reduced by 5-fold when compared with those in infected immunocompetent mice. In the absence of infection, MCP-1 levels were undetectable in both groups of mice. Thus, IFN-α/β-mediated events appear to promote the expression of MCP-1 in liver during MCMV infection.

FIGURE 2.

IFN-α/β-dependent effects on MCP-1 production. Liver homogenates were prepared from 129 (WT) or IFN-α/βR mice (A) that were uninfected (day 0) or infected with MCMV for 1.5 days or 129 mice (B) treated with three daily i.p injections of 1 × 105 U of rIFN-α or vehicle as described in Materials and Methods. MCP-1 protein was measured by ELISA. The limits of detection were 0.08 ng/g of liver. Data represent the means ± SE (n = 4–5 mice tested individually per group). Statistically significant differences were observed between uninfected and infected mice (∗, p ≤ 0.03) and between vehicle- and rIFN-α-treated mice (∗∗, p ≤ 0.001). C, Total liver leukocytes from naive 129 (WT) or IFN-α/βR mice were stimulated with rIFN-α at the doses indicated. Leukocytes were cultured overnight, and collected supernatant fluids were evaluated for production of MCP-1 by sandwich ELISA. The limits of detection were 4 pg/ml. Data represent the means ± SE (n = 3 mice per indicated dose). The results are representative of at least two independent experiments.

FIGURE 2.

IFN-α/β-dependent effects on MCP-1 production. Liver homogenates were prepared from 129 (WT) or IFN-α/βR mice (A) that were uninfected (day 0) or infected with MCMV for 1.5 days or 129 mice (B) treated with three daily i.p injections of 1 × 105 U of rIFN-α or vehicle as described in Materials and Methods. MCP-1 protein was measured by ELISA. The limits of detection were 0.08 ng/g of liver. Data represent the means ± SE (n = 4–5 mice tested individually per group). Statistically significant differences were observed between uninfected and infected mice (∗, p ≤ 0.03) and between vehicle- and rIFN-α-treated mice (∗∗, p ≤ 0.001). C, Total liver leukocytes from naive 129 (WT) or IFN-α/βR mice were stimulated with rIFN-α at the doses indicated. Leukocytes were cultured overnight, and collected supernatant fluids were evaluated for production of MCP-1 by sandwich ELISA. The limits of detection were 4 pg/ml. Data represent the means ± SE (n = 3 mice per indicated dose). The results are representative of at least two independent experiments.

Close modal

Because IFN-α/β have been shown to be required for antiviral defenses in liver, the in vivo effects of the cytokines on MCP-1 protein expression were examined in the absence of potential secondary insults resulting from uncontrolled viral burdens. Studies were conducted using liver homogenates prepared from naive 129 mice that were administered vehicle or rIFN-α. The levels of detected MCP-1 protein were low in mice treated with vehicle (0.04 ± 0.005 ng/g of liver; Fig. 2,B). In contrast to this minimal induction, mice treated with rIFN-α produced significant levels of MCP-1 protein with values of 0.24 ± 0.04 ng/g of liver (Fig. 2 B). Hence, the introduction of rIFN-α results in a 6-fold elevation of MCP-1 in liver.

Additionally, to define and specifically evaluate the contribution of IFN-α/β responses in regulation of MCP-1 induction, primary liver leukocytes were stimulated in culture with rIFN-α. Cells were prepared from naive WT or IFN-α/βR mice and incubated with increasing doses of the cytokine. As shown in Fig. 2 C, MCP-1 protein was induced in a dose-dependent manner in samples from immunocompetent (WT) mice. However, MCP-1 was not detected in liver leukocyte samples from stimulated IFN-α/βR mice. Similar results were evident when naive liver leukocytes from C57BL/6 mice were stimulated with rIFN-α (data not shown). Taken together, these results establish that the induction of MCP-1 protein in liver is dependent on IFN-α/β-mediated functions.

As demonstrated above, naive liver leukocytes can be induced to produce MCP-1 upon stimulation with rIFN-α. To determine whether resident macrophages, a known source of MCP-1 (13, 30, 31), could respond to the effects of IFN-α/β for MCP-1 production, total liver leukocytes, F4/80, or enriched F4/80+ cells were prepared from uninfected C57BL/6 mice. MCP-1 induction was then examined after incubation of cells alone or with rIFN-α. As shown in Fig. 3 A, total populations were induced to produce MCP-1 upon treatment with rIFN-α. However, the levels of MCP-1 produced by enriched F4/80+ cells were profoundly elevated over the levels produced by the total leukocytes. In contrast, the levels of MCP-1 protein produced from the F4/80 sorted population were only marginally detected and below the level of induction from total liver leukocytes. All untreated cells exhibited MCP-1 production below the level of detection. Thus, resident F4/80+ cells can respond to rIFN-α and are a major source of MCP-1 production in the liver.

FIGURE 3.

Characterization of IFN-α/β-responding cells. A, Liver leukocytes were prepared from uninfected C57BL/6 mice and enriched for F4/80 using magnetic separation as described in Materials and Methods. Leukocyte-conditioned media were generated from total, F4/80-enriched, and F4/80-depleted cellular fractions after 24 h of incubation with or without rIFN-α (625 U/ml). Collected supernatants were evaluated for production of MCP-1 by ELISA. Statistically significant differences were observed between the F4/80-enriched and total populations (∗, p ≤ 0.009). B, F4/80-enriched liver leukocytes were obtained from C57BL/6 mice infected with MCMV for 24 h using FACS. Leukocyte-conditioned media were generated from total, F4/80-enriched, and F4/80-depleted cells after 24 h of incubation. Supernatants were tested for MCP-1 using ELISA. Statistically significant differences were observed between the F4/80-enriched and total populations (∗, p ≤ 0.03). For A and B, the limits of detection were 16 pg/ml. Data represent the means ± SE (n = 5–6 pooled mice per experiment). Results of at least three independent experiments are shown.

FIGURE 3.

Characterization of IFN-α/β-responding cells. A, Liver leukocytes were prepared from uninfected C57BL/6 mice and enriched for F4/80 using magnetic separation as described in Materials and Methods. Leukocyte-conditioned media were generated from total, F4/80-enriched, and F4/80-depleted cellular fractions after 24 h of incubation with or without rIFN-α (625 U/ml). Collected supernatants were evaluated for production of MCP-1 by ELISA. Statistically significant differences were observed between the F4/80-enriched and total populations (∗, p ≤ 0.009). B, F4/80-enriched liver leukocytes were obtained from C57BL/6 mice infected with MCMV for 24 h using FACS. Leukocyte-conditioned media were generated from total, F4/80-enriched, and F4/80-depleted cells after 24 h of incubation. Supernatants were tested for MCP-1 using ELISA. Statistically significant differences were observed between the F4/80-enriched and total populations (∗, p ≤ 0.03). For A and B, the limits of detection were 16 pg/ml. Data represent the means ± SE (n = 5–6 pooled mice per experiment). Results of at least three independent experiments are shown.

Close modal

To determine whether resident F4/80+ cells contribute to MCP-1 production under the conditions of MCMV infection, conditioned media were prepared from total liver leukocytes, F4/80, or enriched F4/80+ cells isolated from C57BL/6 mice that were uninfected or infected with MCMV for 24 h. They were then evaluated for their ability to make MCP-1. Total liver leukocytes demonstrated production of the chemokine after infection (Fig. 3 B). Nevertheless, the levels of MCP-1 produced by enriched F4/80+ cells were greatly elevated over the levels produced by the infected total leukocytes. On the contrary, the F4/80 sorted population had levels substantially below those evident in total liver cells. MCP-1 production was below the level of detection in cells from uninfected mice. Taken together, these results identify resident F4/80+ cells as initial responders to IFN-α/β for early production of MCP-1 in the liver.

It has been clearly established that MCP-1 is a potent factor involved in the directed migration of inflammatory macrophages (10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25). Moreover, MCP-1 responses are dependent on CCR2-mediated signaling pathways (15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 32). To examine the contribution of MCP-1 and CCR2 on macrophage accumulation under the conditions of MCMV infection, liver leukocytes were prepared from uninfected or day 2 MCMV-infected C57BL/6 mice that were immunocompetent (WT) or genetically deficient in MCP-1 (MCP-1) or CCR2 (CCR2). The proportions and numbers of migrating macrophages, identified as F4/80+CD11b+ (28), were determined using flow cytometric and cell yield analyses. Uninfected WT and MCP-1 mice had comparable macrophage proportions and absolute numbers (Fig. 4, A and B). Results were similar in uninfected mice lacking CCR2. However, at day 2 after MCMV infection, WT mice had 4-fold increases in both the percentages (Fig. 4,A) and absolute F4/80+CD11b+ numbers when compared with uninfected WT mice (Fig. 4,B). In contrast, MCP-1 mice exhibited a significant 40 and 50% reduction in percentages and absolute numbers of F4/80+CD11b+ cells, respectively, when compared with infected WT mice (Fig. 4, A and B). Likewise, CCR2 mice showed a similar reduction in the frequency (60%) and absolute macrophage numbers (55%) when compared with infected WT mice (Fig. 4, A and B). Thus, MCP-1- and CCR2-mediated functions promote the accumulation of macrophages in liver during MCMV infection.

FIGURE 4.

MCP-1 and CCR2 requirements for accumulation of macrophages and NK cells. Samples were prepared from C57BL/6 (WT) or mice genetically deficient in MCP-1 (MCP-1) or CCR2 (CCR2) that were uninfected (day 0) or infected with MCMV for 2 days. Total liver leukocytes were harvested and analyzed by flow cytometry. The percentage (A) and number (B) of F4/80+CD11b+ cells and the percentage (C) and number (D) of NK1.1+TCRβ cells per liver are shown. Data are the means ± SE (n = 6–9 mice). Differences between infected WT and MCP-1 or CCR2-deficient mice are significant at ∗, p ≤ 0.001 and ∗∗, p ≤ 0.0001. Data shown are representative of at least three experiments.

FIGURE 4.

MCP-1 and CCR2 requirements for accumulation of macrophages and NK cells. Samples were prepared from C57BL/6 (WT) or mice genetically deficient in MCP-1 (MCP-1) or CCR2 (CCR2) that were uninfected (day 0) or infected with MCMV for 2 days. Total liver leukocytes were harvested and analyzed by flow cytometry. The percentage (A) and number (B) of F4/80+CD11b+ cells and the percentage (C) and number (D) of NK1.1+TCRβ cells per liver are shown. Data are the means ± SE (n = 6–9 mice). Differences between infected WT and MCP-1 or CCR2-deficient mice are significant at ∗, p ≤ 0.001 and ∗∗, p ≤ 0.0001. Data shown are representative of at least three experiments.

Close modal

As macrophages contribute to the early production of MIP-1α, which mediates NK cell inflammation in liver during MCMV infection (2, 6, 9), the effects of MCP-1 and CCR2 on NK cell accumulation were subsequently determined. Uninfected and day 2 MCMV-infected C57BL/6 (WT), MCP-1, and CCR2 mice were evaluated. Liver leukocytes were prepared, and proportions and absolute numbers of NK1.1+TCRβ cells were determined using flow cytometric and cell yield analyses. During MCMV infection, WT mice demonstrated a 2-fold elevation in frequency (Fig. 4,C) and a 4-fold increase in absolute numbers of NK cells when compared with uninfected WT mice (Fig. 4,D). Conversely, under the conditions of MCMV infection, the percentages of NK cells in MCP-1 and CCR2 mice did not change significantly, but the absolute numbers of NK cells were modestly increased in samples from both groups, when respectively compared with uninfected MCP-1 or CCR2 mice (Fig. 4, C and D). Nevertheless, MCP-1 and CCR2 mice exhibited a 40 and 56% reduction, respectively, in the total number of accumulated NK cells. Collectively, these results identify a significant role for MCP-1 and CCR2 interactions in mediating the recruitment of macrophages and NK cells to liver during MCMV infection.

Previous studies have established that production of MIP-1α is required for promoting early NK cell inflammatory responses and delivery of IFN-γ during MCMV infection in the liver (2, 6). Furthermore, during infection, trafficking macrophages are greatly enriched for localized production of MIP-1α (2, 9). To examine whether MCP-1 and CCR2 functions promote regulation of MIP-1α and/or IFN-γ responses, liver leukocyte-conditioned media prepared from C57BL/6 (WT), MCP-1, or CCR2 mice that were uninfected or infected with MCMV for 2 days were evaluated for cytokine production using ELISA. The results show induction of MIP-1α and IFN-γ protein upon infection in samples from all mice tested (Fig. 5, A and B). However, MIP-1α responses in infected WT mice were profoundly elevated from the uninfected values of 38 ± 2 pg/ml to values of 358 ± 65 pg/ml during infection (Fig. 5,A). In contrast, the induction of MIP-1α in MCP-1 and CCR2 mice was significantly affected such that cytokine production was reduced by 62% to values of 136 ± 40 pg/ml in MCP-1 and by 78% to values of 80 ± 21 pg/ml in CCR2, when compared with infected WT mice (Fig. 5,A). Similar results were evident for IFN-γ protein production. In WT mice, the levels of IFN-γ were greatly enhanced from the uninfected values of 9 ± 1 pg/ml to the infection-induced values of 603 ± 41 pg/ml (Fig. 5,B). By comparison, IFN-γ levels were substantially reduced by 32% to values of 408 ± 49 pg/ml and by 55% to values of 272 ± 20 pg/ml in MCP-1 and CCR2 mice, respectively, when compared with infected WT mice (Fig. 5 B). Thus, during MCMV infection, MCP-1 and CCR2 functions are necessary for optimal induction of MIP-1α and IFN-γ responses in liver.

FIGURE 5.

Effects of MCP-1 and CCR2 on cytokine induction. Liver leukocyte-conditioned media were prepared from C57BL/6 (WT) or mice genetically deficient in MCP-1 (MCP-1) or CCR2 (CCR2) that were uninfected (day 0) or infected i.p. with MCMV for 2 days. The levels of MIP-1α (A) or IFN-γ (B) were measured using ELISA. The levels of detection were 30 and 18 pg/ml for MIP-1α and IFN-γ, respectively. Data represent the means ± SE (n = 3–6 mice tested individually). Differences between infected WT and MCP-1- or CCR2-deficient mice are significant at ∗, p ≤ 0.05 and ∗∗, p ≤ 0.003. Results are representative of at least two experiments.

FIGURE 5.

Effects of MCP-1 and CCR2 on cytokine induction. Liver leukocyte-conditioned media were prepared from C57BL/6 (WT) or mice genetically deficient in MCP-1 (MCP-1) or CCR2 (CCR2) that were uninfected (day 0) or infected i.p. with MCMV for 2 days. The levels of MIP-1α (A) or IFN-γ (B) were measured using ELISA. The levels of detection were 30 and 18 pg/ml for MIP-1α and IFN-γ, respectively. Data represent the means ± SE (n = 3–6 mice tested individually). Differences between infected WT and MCP-1- or CCR2-deficient mice are significant at ∗, p ≤ 0.05 and ∗∗, p ≤ 0.003. Results are representative of at least two experiments.

Close modal

Previous studies have demonstrated that immunocompetent C57BL/6 mice establish viral clearance and resolution of virus-induced liver disease after 5 days of MCMV infection (3, 4, 7, 33). To assess the contribution of MCP-1 and CCR2 responses to antiviral defense, spleen and liver samples were prepared from MCMV-infected C57Bl/6 (WT), MCP-1, or CCR2 mice. Mice were infected with 5 × 104 PFU MCMV for 4 or 5 days, and viral burdens were evaluated. As compared with WT mice, viral titers in both compartments were increased by 1 log on day 4 and 2 logs on day 5 after infection in both MCP-1- and CCR2-deficient mice (Fig. 6, A and B). The results demonstrate that MCP-1 and CCR2 functions contribute to MCMV resistance.

FIGURE 6.

Effects of MCP-1 and CCR2 on susceptibility to MCMV infection. C57BL/6 (WT) or mice genetically deficient in MCP-1 (MCP-1) or CCR2 (CCR2) were infected with 5 × 104 PFU MCMV for 4 or 5 days. Livers (A) and spleens (B) were harvested and homogenized for viral titer determination as described in Materials and Methods. Limit of detection for the assay was 2 log PFU/g of tissue. Means ± SE (n = 3–6) are shown. Differences between WT and MCP-1- or CCR2-deficient mice are significant at ∗, p ≤ 0.0003 and †, p ≤ 0.005. Data are representative of two experiments. C, WT, MCP-1, and CCR2 mice were uninfected or infected with 5 × 104 PFU MCMV for 4 or 5 days. Serum was collected and used to measure ALT levels as described in Materials and Methods. Means ± SE (n = 3–6) are shown. Differences between WT and MCP-1- or CCR2-deficient mice are significant at ∗, p ≤ 0.0005 and ∗∗, p ≤ 0.05. Data are representative of two experiments.

FIGURE 6.

Effects of MCP-1 and CCR2 on susceptibility to MCMV infection. C57BL/6 (WT) or mice genetically deficient in MCP-1 (MCP-1) or CCR2 (CCR2) were infected with 5 × 104 PFU MCMV for 4 or 5 days. Livers (A) and spleens (B) were harvested and homogenized for viral titer determination as described in Materials and Methods. Limit of detection for the assay was 2 log PFU/g of tissue. Means ± SE (n = 3–6) are shown. Differences between WT and MCP-1- or CCR2-deficient mice are significant at ∗, p ≤ 0.0003 and †, p ≤ 0.005. Data are representative of two experiments. C, WT, MCP-1, and CCR2 mice were uninfected or infected with 5 × 104 PFU MCMV for 4 or 5 days. Serum was collected and used to measure ALT levels as described in Materials and Methods. Means ± SE (n = 3–6) are shown. Differences between WT and MCP-1- or CCR2-deficient mice are significant at ∗, p ≤ 0.0005 and ∗∗, p ≤ 0.05. Data are representative of two experiments.

Close modal

To specifically evaluate overall liver damage during MCMV infection, expression of the liver enzyme ALT was measured in serum of WT, MCP-1, or CCR2 mice that were uninfected or infected with MCMV for 4 or 5 days. Uninfected mice had comparable baseline levels of the circulating enzyme (Fig. 6,C). By day 4 after infection, 2-fold increases in ALT levels were detected above baseline in all groups of mice (Fig. 6 C). On day 5 of infection, ALT levels continued to rise reaching values of 150 ± 10 and 150 ± 55 U/L in MCP-1 and CCR2 mice, respectively. Thus, induced ALT levels were up to 3- and 5-fold higher than baseline levels detected in uninfected MCP-1 and CCR2 mice. In contrast, the ALT levels in WT mice returned to uninfected baseline values of 39 ± 5 U/L by day 5 after infection. Hence, MCP-1- and CCR2-mediated functions promote resolution of hepatic damage.

To additionally define the effects of MCP-1 and CCR2 responses on antiviral defense, H&E-stained liver sections were prepared from WT, MCP-1, or CCR2 mice that were uninfected or infected with MCMV for 5 days. The histological appearance of all uninfected mice was identical (Fig. 7, A–C). Upon infection, WT mice did not exhibit profound pathology (Fig. 7,D) and were instead histologically similar to uninfected mice (Fig. 7,A), with the exception of intermittent small clusters of inflammatory foci. In contrast, large areas of necrotic lesions were evident in MCMV-infected MCP-1 and CCR2 mice (Fig. 7, E and F). Furthermore, cytomegalic inclusion bodies were readily visible in viable hepatocytes. The results show that MCP-1 and CCR2 effects limit the development of significant liver pathology.

FIGURE 7.

Characterization of MCMV-induced liver damage in MCP-1- and CCR2-deficient mice. A and D, C57BL/6 (WT) or mice genetically deficient in (B and E) MCP-1 (MCP-1) or (C and F) CCR2 (CCR2) were either uninfected (A–C) or infected with 5 × 104 PFU MCMV for 5 days (D–F). Livers were harvested, paraffin embedded, and sectioned for H&E staining. Bar represents 100 μm. Arrows indicate necrotic lesions. Arrows within insets indicate cytomegalic inclusion bodies.

FIGURE 7.

Characterization of MCMV-induced liver damage in MCP-1- and CCR2-deficient mice. A and D, C57BL/6 (WT) or mice genetically deficient in (B and E) MCP-1 (MCP-1) or (C and F) CCR2 (CCR2) were either uninfected (A–C) or infected with 5 × 104 PFU MCMV for 5 days (D–F). Livers were harvested, paraffin embedded, and sectioned for H&E staining. Bar represents 100 μm. Arrows indicate necrotic lesions. Arrows within insets indicate cytomegalic inclusion bodies.

Close modal

In host survival studies challenging mice with 105 PFU of MCMV, all WT mice survived beyond 15 days of infection. Conversely, MCP-1 and CCR2 mice succumbed to the lethal effects of MCMV and exhibited 100% mortality by day 5 of infection (Fig. 8). Taken together, these results demonstrate that MCP-1 and CCR2 are critical factors promoting effective control of virus replication and the fatal consequences of virus-induced liver disease.

FIGURE 8.

MCP-1 and CCR2 effects on host survival. C57BL/6 (WT) or mice genetically deficient in MCP-1 (MCP-1) or CCR2 (CCR2) were uninfected or infected with 105 PFU MCMV and monitored twice daily for survival (n = 6–9).

FIGURE 8.

MCP-1 and CCR2 effects on host survival. C57BL/6 (WT) or mice genetically deficient in MCP-1 (MCP-1) or CCR2 (CCR2) were uninfected or infected with 105 PFU MCMV and monitored twice daily for survival (n = 6–9).

Close modal

These studies evaluate the immediate effects of IFN-α/β on production of MCP-1 and characterize a significant role for MCP-1- and CCR2-mediated responses during MCMV infection in the liver. The results show that MCP-1 production precedes that of MIP-1α and that IFN-α/β effects can promote MCP-1 expression from resident F4/80+ cells. The studies also identify essential functions for MCP-1 and CCR2 in mediating the recruitment of macrophage and NK cell populations. Furthermore, the results indicate that MCP-1- and CCR2-mediated effects promote hepatic expression of MIP-1α and IFN-γ, as well as the development of protective responses against MCMV infection. Taken together, the data clearly show that immunoregulatory effects mediated by IFN-α/β are required for MCP-1 induction and that MCP-1 is a prominent mediator between IFN-α/β and MIP-1α in liver. Collectively, these cytokine and chemokine responses are required for liver inflammation and innate host defense against viral pathogens.

The central roles played by IFN-α/β in innate immune responses and antiviral defense have been extensively characterized (34, 35). Moreover, IFN-α/β also promotes leukocyte-trafficking events (36, 37, 38, 39). Under the conditions of early MCMV infection, IFN-α/β functions have been found to trigger the recruitment of macrophages enriched for production of MIP-1α and, as a result, the accumulation of NK cells in the liver (6, 9). The biological effects of MCP-1 are equally critical to innate immunity because they are uniquely necessary for monocyte/macrophage recruitment and the establishment of host defense against multiple pathogens (10, 15, 16, 17, 18, 19, 20). Recent studies have suggested that induction of IFN-α/β during infection with intracellular bacteria may promote MCP-1 production (19, 40). Moreover, IFN-β treatment of cultured human microglia and peripheral blood monocytes has been shown to induce expression of MCP-1 (41, 42). To our knowledge, the studies presented here are the first to conclusively document that MCP-1 induction is dependent upon immediate IFN-α/β effects during a viral infection established in tissue sites. Additionally, the response appears to be unique to MCP-1 expression because IFN-α/βR deficiencies did not significantly affect production of MCP-3 and MCP-5, which are other members of the MCP subfamily (Refs.43, 44, 45 ; data not shown).

Our results identify a resident F4/80+ cell, most likely a Kupffer cell, as an early responder to the immunoregulatory effects of IFN-α/β for production of MCP-1 in liver. These observations define a potential mechanism whereby IFN-α/β induced in response to MCMV infection as early as 24 h stimulates the initial peak of MCP-1 from Kupffer cells. To sustain enhanced levels of MCP-1, trafficking cell populations may contribute to production of the chemokine to ensure the activation of downstream inflammatory responses necessary for antiviral defense.

MCP-1 functions are exclusively dependent on binding to the CCR2 receptor (20, 21, 22, 23, 24, 25). During MCMV infection, mice genetically deficient in either MCP-1 or CCR2 were shown to be impaired in macrophage accumulation in liver, indicating that this response is largely dependent on MCP-1 and CCR2 interactions. This is consistent with reports that macrophages effectively express CCR2 (22, 30, 46), and MCP-1 is a potent chemoattractant for these cells (10, 11, 12, 13, 14, 15, 16, 17). MCP-1- and CCR2-deficient mice also had pronounced reductions in expression of MIP-1α in liver. These results concur with previous observations that trafficking macrophages contribute to the initial delivery of MIP-1α (9). The studies presented also demonstrate that MCP-1 and CCR2 deficiencies significantly diminished NK cell recruitment and IFN-γ production. Collectively, the results strongly suggest that a major role for MCP-1 is to directly promote the migration of a population of inflammatory macrophages that appreciably contribute to MIP-1α production and, as a consequence, initiate NK cell recruitment and delivery of IFN-γ in infected tissues. As CCR2-deficient mice exhibited similar responses, MCP-1 is primarily responsible for initiating these events. Moreover, MCP-1- and CCR2-deficient mice demonstrated increased sensitivity to the consequences of MCMV infection. Increased viral burden was associated with high levels of the circulating liver enzyme ALT, indicative of general manifestations of liver damage (47), and extensive gross liver pathology. Mortality in these mice coincided with virus-induced liver disease. Taken together, these results are consistent with studies using mice that are genetically deficient in IFN-α/βR, MIP-1α, IFN-γ (2, 6, 7, 9), or are NK cell deficient (33). Thus, the mechanisms promoting NK cell inflammation and protective immune responses in liver do include MCP-1- and CCR2-mediated interactions.

We cannot exclude the possibility that MCP-1 and CCR2 effects are directly acting upon NK cells for accumulation. This seems unlikely, however, because MIP-1α-deficient mice exhibit macrophage infiltration (6, 9) and produce significant levels of MCP-1 in liver during MCMV infection (K. L. Hokeness and T. P. Salazar-Mather, unpublished data). Nevertheless, MIP-1α-deficient mice are still impaired in their ability to effectively promote NK cell inflammatory responses and sustain production of IFN-γ in liver (6, 7). Furthermore, murine NK cells have been shown to preferentially express the chemokine receptor CCR5 (46).

In summary, our results demonstrate that IFN-α/β-mediated functions are key in the initiation of MCP-1 production through stimulation of a resident F4/80+ cell. Furthermore, this first response is required for macrophage accumulation and activation of downstream inflammatory events in liver. Taken together, our findings establish additional insight into the complex cytokine and chemokine networks necessary for efficient innate immunity against viral infections.

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.

1

This work was supported by National Institutes of Health Grants CA-102708, CA-41268, and P20RR15578 and by Department of Education Predoctoral Training Grant P200A000117 (to K.L.H.).

4

Abbreviations used in this paper used in this paper: MCMV, murine CMV; ALT, alanine aminotransferase; WT, wild type.

1
Biron, C. A., K. B. Nguyen, G. C. Pien, L. P. Cousens, T. P. Salazar-Mather.
1999
. Natural killer cells in antiviral defense: function and regulation by innate cytokines.
Annu. Rev. Immunol.
17
:
189
.
2
Salazar-Mather, T. P., K. L. Hokeness.
2003
. Calling in the troops: regulation of inflammatory cell trafficking through innate cytokine/chemokine networks.
Viral Immunol.
16
:
291
.
3
Orange, J. S., B. Wang, C. Terhorst, C. A. Biron.
1995
. Requirement for natural killer cell-produced interferon γ in defense against murine cytomegalovirus infection and enhancement of this defense pathway by interleukin 12 administration.
J. Exp. Med.
182
:
1045
.
4
Orange, J. S., C. A. Biron.
1996
. An absolute and restricted requirement for IL-12 in natural killer cell IFN-γ production and antiviral defense.
J. Immunol.
156
:
1138
.
5
Tay, C. H., R. M. Welsh.
1997
. Distinct organ-dependent mechanisms for the control of murine cytomegalovirus infection by natural killer cells.
J. Virol.
71
:
267
.
6
Salazar-Mather, T. P., J. S. Orange, C. A. Biron.
1998
. Early murine cytomegalovirus (MCMV) infection induces liver natural killer (NK) cell inflammation and protection through macrophage inflammatory protein 1α (MIP-1α)-dependent pathways.
J. Exp. Med.
187
:
1
.
7
Salazar-Mather, T. P., T. A. Hamilton, C. A. Biron.
2000
. A chemokine-to-cytokine-to-chemokine cascade critical in antiviral defense.
J. Clin. Invest.
105
:
985
.
8
Pien, G. C., A. R. Satoskar, K. Takeda, S. Akira, C. A. Biron.
2000
. Cutting edge: selective IL-18 requirements for induction of compartmental IFN-γ responses during viral infection.
J. Immunol.
165
:
4787
.
9
Salazar-Mather, T. P., C. A. Lewis, C. A. Biron.
2002
. Type I interferons regulate inflammatory cell trafficking and macrophage inflammatory protein 1α delivery to the liver.
J. Clin. Invest.
110
:
321
.
10
Huffnagle, G. B., R. M. Streiter, T. J. Standiford, R. A. McDonald, M. D. Burdick, S. L. Kunkel, G. B. Toews.
1995
. The role of monocyte chemoattractant protein-1 (MCP-1) in the recruitment of monocytes and CD4+ T cells during pulmonary cryptococcus infection.
J. Immunol.
155
:
4790
.
11
Fuentes, M. E., S. K. Durham, M. R. Swerdel, A. C. Lewin, D. S. Barton, J. R. Megill, R. Bravo, S. Lira.
1995
. Controlled recruitment of monocytes and macrophages to specific organs through transgenic expression of monocyte chemoattractant protein-1.
J. Immunol.
155
:
5769
.
12
Rollins, B. J..
1996
. Monocyte chemoattractant protein 1: a potential regulator of monocyte recruitment in inflammatory disease.
Mol. Med. Today
2
:
198
.
13
Rollins, B. J..
1997
. Chemokines.
Blood
90
:
909
.
14
Gu, L., B. J. Rutledge, J. Fiorillo, C. Ernst, I. S. Grewal, R. A. Flavell, R. P. Gladue, B. J. Rollins.
1997
. In vivo properties of monocyte chemoattractant protein-1.
J. Leukocyte Biol.
62
:
577
.
15
Lu, B., B. J. Rutledge, L. Gu, J. Fiorillo, N. W. Lukacs, S. L. Kunkel, R. North, C. Gerad, B. J. Rollins.
1998
. Abnormalities in monocyte recruitment and cytokine expression in monocyte chemoattractant protein 1-deficient mice.
J. Exp. Med.
187
:
601
.
16
Tesch, G. H., S. Maifert, A. Schwarting, B. J. Rollins, V. R. Kelley.
1999
. Monocyte chemoattractant protein 1-dependent leukocytic infiltrates are responsible for autoimmune disease in MRL-Faslpr mice.
J. Exp. Med.
190
:
1813
.
17
Huang, D. R., J. Wang, P. Kivisakk, B. J. Rollins, R. M. Ransohoff.
2001
. Absence of monocyte chemoattractant protein 1 in mice leads to decreased local macrophage recruitment and antigen-specific T helper cell type 1 immune response in experimental autoimmune encephalomyelitis.
J. Exp. Med.
193
:
713
.
18
Rutledge, B. J., H. Rayburn, R. Rosenberg, R. J. North, R. P. Gladue, C. L. Corless, B. J. Rollins.
1995
. High level monocyte chemoattractant protein-1 expression in transgenic mice increases their susceptibility to intracellular pathogens.
J. Immunol.
155
:
4838
.
19
Serbina, N. V., W. Kuziel, R. Flavell, S. Akira, B. Rollins, E. G. Pamer.
2003
. Sequential MyD88-independent and -dependent activation of innate immune responses to intracellular bacterial infection.
Immunity
19
:
891
.
20
Boring, L., J. Gosling, S. W. Chensue, S. L. Kunkel, R. V. Farese, Jr, H. E. Broxmeyer, I. F. Charo.
1997
. Impaired monocyte migration and reduced type 1 (Th1) cytokine responses in C-C chemokine receptor 2 knockout mice.
J. Clin. Invest.
100
:
2552
.
21
Kurihara, T., G. Warr, J. Loy, R. Bravo.
1997
. Defects in macrophage recruitment and host defense in mice lacking the CCR2 chemokine receptor.
J. Exp. Med.
186
:
1757
.
22
Kuziel, W. A., S. J. Morgan, T. C. Dawson, S. Griffin, O. Smithies, K. Ley, N. Maeda.
1997
. Severe reduction in leukocyte adhesion and monocyte extravasation in mice deficient in CC chemokine receptor 2.
Proc. Natl. Acad. Sci. USA
94
:
12053
.
23
Sato, N., W. A. Kuziel, P. C. Melby, R. L. Reddick, V. Kostecki, W. Zhao, N. Maeda, S. K. Ahuja, S. S. Ahuja.
1999
. Defects in the generation of IFN-γ are overcome to control infection with Leishmania donovani in CC chemokine receptor (CCR) 5-, macrophage inflammatory protein-1α, or CCR2-deficient mice.
J. Immunol.
163
:
5519
.
24
Izikson, L., R. S. Klein, I. F. Charo, H. L. Weiner, A. D. Luster.
2000
. Resistance to experimental autoimmune encephalomyelitis in mice lacking the CC chemokine receptor (CCR)2.
J. Exp. Med.
192
:
1075
.
25
Traynor, T. R., W. A. Kuziel, G. B. Toews, G. B. Huffnagle.
2000
. CCR2 expression determines T1 versus T2 polarization during pulmonary Cryptococcus neoformans infection.
J. Immunol.
164
:
2021
.
26
Muller, U., U. Steinhoff, L. F. Reiss, S. Hemmi, J. Pavlovic, R. M. Zinkernagel, M. Auget.
1994
. Functional role of type I and type II interferons in antiviral defense.
Science
264
:
1918
.
27
Pien, G. C., C. A. Biron.
2000
. Compartmental differences in NK cell responsiveness to IL-12 during lymphocytic choriomeningitis virus infection.
J. Immunol.
164
:
994
.
28
Lepay, D. A., R. M. Steinman, C. F. Nathan, H. W. Murray, Z. A. Cohn.
1985
. Liver macrophages in murine listeriosis.
J. Exp. Med.
161
:
1503
.
29
Nguyen, K. B., W. T. Watford, R. Salomon, S. R. Hofmann, G. C. Pien, A. Morinobu, M. Gadina, J. J. O’Shea, C. A. Biron.
2002
. Critical role for STAT4 activation by type 1 interferons in the interferon-γ response to viral infection.
Science
297
:
2063
.
30
Bautista, A. P..
2002
. Chronic alcohol intoxication primes Kupffer cells and endothelial cells for enhanced CC-chemokine production and concomitantly suppresses phagocytosis and chemotaxis.
Front. Biosci.
7
:
117
.
31
Leifeld, L., F. L. Dumoulin, I. Purr, K. Janberg, C. Trautwein, M. Wolff, M. P. Manns, T. Sauerbruch, U. Spengler.
2003
. Early up-regulation of chemokine expression in fulminant hepatic failure.
J. Pathol.
199
:
335
.
32
Charo, I. F..
1999
. CCR2: from cloning to the creation of knockout mice.
Chem. Immunol.
72
:
30
.
33
Orange, J. S., T. P. Salazar-Mather, S. M. Opal, C. A. Biron.
1997
. Mechanisms for virus-induced liver disease: tumor necrosis factor-mediated pathology independent of natural killer and T cells during murine cytomegalovirus infection.
J. Virol.
71
:
9248
.
34
Biron, C. A..
2001
. Interferons α and β as immune regulators: a new look.
Immunity
14
:
661
.
35
Biron, C. A., M. Dalod, T. P. Salazar-Mather.
2002
. Innate immunity and viral infections. S. H. E. Kaufmann, Jr, and A. Sher, Jr, and R. Ahmed, Jr, eds.
Immunology of Infectious Diseases
139
. ASM Press, Washington, DC.
36
Korngold, R., K. J. Blank, D. M. Murasko.
1983
. Effect of interferon on thoracic duct lymphocyte output: induction with either polyI:C or vaccinia virus.
J. Immunol.
130
:
2236
.
37
Wiltrout, R. H., A. M. Pilaro, M. E. Gruys, J. E. Talmadge, D. L. Longo, J. R. Ortaldo, C. W. Reynolds.
1989
. Augmentation of mouse liver-associated natural killer cell activity by biological response modifiers occurs largely via rapid recruitment of large granular lymphocytes from the bone marrow.
J. Immunol.
143
:
372
.
38
Ishikawa, R., C. A. Biron.
1993
. IFN induction and associated changes in splenic leukocyte distribution.
J. Immunol.
150
:
3713
.
39
Salazar-Mather, T. P., C. A. Biron.
1996
. NK cell trafficking and cytokine expression in splenic compartments after IFN induction and viral infection.
J. Immunol.
157
:
3054
.
40
O’Riordan, M., C. H. Yi, R. Gonzales, K. D. Lee, D. A. Portnoy.
2002
. Innate recognition of bacteria by a macrophage cytosolic surveillance pathway.
Proc. Natl. Acad. Sci. USA
99
:
13861
.
41
McManus, C. M., J. S. H. Liu, M. T. Hahn, L. L. Hua, C. F. Brosnan, J. W. Berman, S. C. Lee.
2000
. Differential induction of chemokines in human microglia by type I and II interferons.
Glia
29
:
273
.-280.
42
Fantuzzi, L., I. Canini, F. Belardelli, S. Gessani.
2003
. IFN-β stimulates the production of β-chemokines in human peripheral blood monocytes: importance of macrophage differentiation.
Eur. Cytokine Network
4
:
597
.-603.
43
Combadiere, C., S. K. Ahuja, J. Van Damme, H. L. Tiffany, J. L. Gao, P. M. Murphy.
1995
. Monocyte chemoattractant protein-3 is a functional ligand for CC chemokine receptors 1 and 2B.
J. Biol. Chem.
270
:
29671
.
44
Franci, C., L. M. Wong, D. J. Van, P. Proost, I. F. Charo.
1995
. Monocyte chemoattractant protein-3, but not monocyte chemoattractant protein-2, is a functional ligand of the human monocyte chemoattractant protein-1 receptor.
J. Immunol.
154
:
6511
.
45
Sarafi, M. N., E. A. Garcia-Zepeda, J. A. Maclean, I. F. Charo, A. D. Luster.
1997
. Murine monocyte chemoattractant protein (MCP)-5: a novel CC chemokine that is a structural and functional homologue of human MCP-1.
J. Exp. Med.
185
:
99
.
46
Mack, M., J. Cihak, C. Simonis, B. Luckow, A. E. I. Proudfoot, J. Plachy, H. Bruhl, M. Frink, H.-J. Anders, V. Vielhauer, et al
2001
. Expression and characterization of the chemokine receptors CCR2 and CCR5 in mice.
J. Immunol.
166
:
4697
.
47
Shanley, J. D., L. Biczak, S. J. Forman.
1993
. Acute murine cytomegalovirus infection induces lethal hepatitis.
J. Infect. Dis.
167
:
264
.