Chemokine Receptor Ccr2 Is Critical for Monocyte Accumulation and Survival in West Nile Virus Encephalitis

West Nile virus (WNV) is a re-emerging flavivirus that cycles naturally between mosquitoes and birds, with humans and other vertebrates serving as accidental hosts. WNV did not appear in the United States until 1999, when it was identified as the cause of a meningoencephalitis outbreak in the New York City area. Since then, the virus has spread rapidly across North America, causing severe neurologic illness and death in humans and decimating bird populations. There were 29,625 cases of WNV reported to the Centers for Disease Control and Prevention (Atlanta, GA) as of November 2009, including 1173 deaths (4.0%). Neither vaccines nor antiviral therapies are available, and new concepts based on precise understanding of the molecular pathogenesis are needed.

WNV infection remains subclinical in most humans, but ∼20–30% may develop symptoms of WNV disease ranging from fever to meningoencephalitis, leading to paralysis or death (1). In both mice and humans, WNV encephalitis is characterized by virus-associated pathological processes, including reaction of the CNS resident cells and infiltration of inflammatory leukocytes in the perivascular space and parenchyma, primarily monocytes and T cells (2–5). At the molecular level, leukocyte trafficking is mediated through a multistep process involving chemokines and chemokine receptors, which guide leukocytes to infected tissues. Recently, we identified chemokine receptor CCR5 as a critical host factor for protection against WNV infection in both mice and humans (6). Infection of Ccr5-deficient mice was uniformly fatal with significantly fewer T cells and monocytes accumulating in the CNS (2). A similar increase in susceptibility was observed in individuals homozygous for the complete loss-of-function mutation CCR5Δ32; conversely CCR5Δ32 homozygosity was absent in asymptomatic WNV-infected individuals (7–9).

The importance of T cell recruitment and function in the CNS in WNV infection has been demonstrated in studies of mice lacking the Th1-specific chemokine Cxcl10 and its receptor Cxcr3 (4, 10). Loss of either resulted in decreased CD4⁺ and CD8⁺ T cell accumulation in the CNS, increased viral burden, and death. Perforin-mediated CD8⁺ T cell responses are also critical for effective clearance of WNV from infected neurons (11). Consistent with this, enhancing migration of CD8⁺ T cells into the parenchyma of the WNV-infected CNS has been reported to increase viral clearance and survival (12).

In general, monocytes play an important role during CNS injury as precursors of macrophages and microglia (13). However, their role in WNV infection has not been clearly delineated. In a recent study, lethal intranasal delivery of WNV using a non-neurotropic strain caused monocytes to accumulate in the CNS, >90% of which were Ly6c^hi and capable of differentiating into both microglia and macrophages (14). Administration of neutralizing anti-Ccl2 Abs delayed migration of these cells into the brain and prolonged survival, suggesting a Ccr2-dependent, pathogenic role for monocytes in the brain in the model. In contrast, monocyte depletion in vivo using clodronate-loaded liposomes resulted in increased mortality of mice challenged i.p. with a neurotropic strain of WNV, supporting a protective role for these cells (15). Thus, monocytes may be beneficial or harmful depending on the context and details of the model.

Two major monocyte subsets have been identified in both humans and mice: “resident” monocytes, which traffic into peripheral tissues under noninflammatory conditions, and “inflammatory” monocytes, which rapidly respond to tissue injury and inflammation (16). High expression of Ccr2 is associated with the “inflammatory” (Ly6c^hi) monocyte subset and correlates with its function of rapid recruitment into inflamed tissues. Conversely, “resident” (Ly6c^lo) monocytes express high levels of Cx₃cr1, not Ccr2. Ccr2 is expressed primarily on monocytes and monocyte-derived cells but is also expressed on a subset of T cells. Ccr2 is activated by the selective ligand Ccl2 and the nonselective ligands Ccl7 and Ccl12. Ccr2-deficient mice are monocytopenic, and tissue recruitment of monocytes is impaired in these mice in infectious disease models, including Listeria monocytogenes (17, 18), Cryptococcus neoformans (19, 20), and tuberculosis (21, 22). In the current study, we have addressed the role of Ccr2 in a mouse model of WNV infection.

Mouse studies were carried out in an animal biosafety level 3 facility under protocols approved by the National Institute of Allergy and Infectious Diseases/National Institutes of Health (Bethesda, MD) Animal Care and Use Committee. Ccr2^−/− mice and wild-type C57BL6/J control mice were purchased from The Jackson Laboratory (Bar Harbor, ME). Ccr2^−/− mice have been back-crossed onto wild-type C57BL/6J mice from The Jackson Laboratory for 8 generations. All experiments were initiated using female mice 8–12 wk of age. WNV strain NY99-35262 was kindly provided by Dr. R. Lanciotti (Centers for Disease Control and Prevention, Fort Collins, CO). Mice were injected s.c. in the scruff of the neck with 10² focus-forming units (FFU) WNV-NY99-35262 suspended in 50 μl HBSS (Invitrogen, Carlsbad, CA) or with HBSS alone (mock-infected). Mice were monitored visually and weighed daily. Mice were perfused prior to aseptic removal of organs, which were weighed and placed directly in either 2 ml HBSS (for ELISA and viral titer assay) or 5 ml FACS buffer (PBS supplemented with 1% BSA and 0.1% NaN₃).

Total RNA was isolated from brain homogenates using Qiagen Lipid RNeasy kit (Qiagen, Valencia, CA). After DNase I treatment (Qiagen), total RNA was converted to cDNA using a Superscript III Supermix kit (Invitrogen) with random hexamers. Real-time PCR was performed in a total reaction volume of 25 μl using 2 μl cDNA, 12.5 μl 2× TaqMan PCR Master mix, and 1.25 μl primer–probe mix. All primers and probes used for quantitative PCR were predesigned by the manufacturer (Applied Biosystems, Carlsbad, CA) and run using the ABI 7900 Fast Real Time PCR System (Applied Biosystems). Calculated copies were normalized against copies of housekeeping gene GAPDH, and fold increase was determined using the ΔΔCT method.

Brains were homogenized into HBSS and centrifuged at 1500 × g for 25 min at 4°C. The supernatants were aliquoted and frozen at –80°C until further use. Plasma from uninfected or infected mice was collected and frozen at –80°C until further use. Mouse Ccl2 and Ccl12 were quantified using Quantikine Immunoassay kits (R&D Systems, Minneapolis, MN). Mouse Ccl7 was measured using CCL7 Instant ELISA kit (BenderMed Systems, Vienna, Austria). Mouse Ccl8 was quantified by sandwich ELISA using a kit according to the manufacturer’s instructions (Antigenix America, Huntington Station, NY).

Vero cells were maintained in OptiPro SFM (Invitrogen). For the viral titer assay, the medium was supplemented with 2% FBS and 50 μg/ml gentamicin. Focus-forming units were assayed on confluent Vero cell monolayers in 12-well plates. Two hundred microliters of virus-containing material was diluted and placed on the cells in duplicate. Virus was adsorbed for 1 h at 37°C, then cells were overlayed with 2 ml Opti-MEM (Invitrogen) plus 8 g/l methyl cellulose containing 2% FBS and 50 μg/ml gentamicin. Cells were incubated for 2 d at 37°C then fixed for 20 min with 100% methanol, washed three times with PBS, and incubated for 1 h at 37°C with 500 μl 1:1000 diluted WNV-specific Abs in hyperimmune mouse ascetic fluid (ATCC no. VR-82). Samples were then washed three times, and 500 μl anti-mouse HRP-labeled polymer (DakoCytomation, Carpinteria, CA) at a concentration of 1:10 was added. Cells were incubated for 1 h at 37°C, washed three times, and WNV FFU were visualized by addition of 1 ml diaminobenzidine chromogen (4.5 mg diaminobenzidine/10 ml PBS containing 4.5 μl 30% H₂O₂ in 10 ml PBS). Viral titers are expressed as FFU/g tissue.

Mice were anesthetized prior to perfusion or sacrificed via cervical dislocation prior to aseptic removal of organs. Brains were collected in FACS buffer and homogenized using the back of a 6-cc syringe on ice in a tissue culture plate. The suspension was brought to 7 ml with FACS buffer, and 3 ml 90% Percoll (in PBS) was added. After thorough mixing, the solution was underlayed with 1 ml 70% Percoll (in PBS). This was centrifuged at 2470 rpm for 30 min at 4°C. Leukocytes at the interphase were isolated and washed three times in FACS buffer. Cells were then filtered though a 70-μm mesh screen. Three hundred microliters of peripheral blood was collected into EDTA tubes by mandibular vein puncture or cardiac puncture. Erythrocytes were removed by incubation with 1× PharmLyse buffer (BD Biosciences) according to the manufacturer’s protocol. Cells were washed three times with FACS buffer and blocked in 100 μl Fc block (1:200 dilution of CD16/32 Ab [BD Biosciences, San Diego, CA] in FACS buffer). Cells were then incubated with specific Abs for 30 min at 4°C, washed three times, and suspended in 400 μl 2% paraformaldehyde buffer. Rat anti-mouse monoclonal Abs coupled to the following fluorophores and directed against the following surface molecules were used for flow cytometry: FITC-conjugated CD19 and Ly6c (BD Biosciences); PE-conjugated CD115 and CD3; PE-Cy7–conjugated F4/80 and CD8α; allophycocyanin-conjugated CD45 and NK1.1; allophycocyanin-Cy7–conjugated CD11b and CD4. All Abs were purchased from eBioscience (San Diego, CA) unless otherwise indicated. UV live/dead cell staining was used to identify cell death (Invitrogen, Carlsbad, CA). Cells were fixed using PBS containing 2% paraformaldehyde (USB Corporation, Santa Clara, CA). FACS data were collected on an LSR II instrument (BD Biosciences), and analysis was performed using FlowJo version 8.5.3 (Tree Star, Ashland, OR). Cell numbers were quantified using counting beads (Spherotech, Lake Forest, IL).

Brains were aseptically removed from mice after cervical dislocation and placed directly into 10% normal buffered formalin (Fisher Scientific, Pittsburgh, PA) for 24 h. Four- to five-micrometer sections of paraffin-embedded slides were pretreated using Diva Decloaking solution (Biocare Medical, Concord, CA) and blocked with 10% normal goat serum (Vector Laboratories, Burlingame, CA). Slides were then incubated with primary rat anti-mouse Mac-2 hybridoma (American Type Culture Collection, Manassas, VA) or with anti-CD3 Ab (DakoCytomation) in the presence of background reducing agent (DakoCytomation). After incubation with anti-mouse polymer-HRP (DakoCytomation), slides were developed with the streptavidin ABC system using biotinylated goat anti-rat IgG (mouse adsorbed; Biocare Medical) and a diaminobenzidine chromogen (DakoCytomation) followed by counterstaining with hematoxylin.

In mice, systemic administration of recombinant Flt3L or injection of B16 melanoma cells expressing Flt3L have both previously been shown to expand markedly dendritic cells in many tissues, including bone marrow (23, 24). We found that injection of mice with Flt3L-expressing B16 cells (24) also markedly increased the percentage of monocytes in bone marrow, and we used this observation to generate sufficient monocytes for adoptive transfer studies (A. Rivollier, unpublished observations). Eight- to twelve-week-old wild-type and Ccr2^−/− C57BL/6J mice were injected s.c. in the axilla with 2 × 10⁶ B16 cells expressing Flt3L. Tumors grew at the site of cell injection to variable sizes but were not allowed to grow beyond 2 cm, per protocol. By 2 wk of growth, monocyte yields typically were increased 3- to 4-fold over yields from uninjected mice. Bone marrow cells were harvested and enriched for monocytes using positive selection. Briefly, rat anti-mouse CD115-biotinylated mAb (eBioscience) was incubated with bone marrow cells harvested from C57BL/6J or Ccr2^−/− mice. Using anti-biotin microbeads (Miltenyi Biotec, Auburn, CA), cells positive for CD115 were separated using the AutoMACS automatic magnetic cell sorting system. CD115-enriched cells were further purified for the classical monocyte subset by FACS using rat anti-mouse monoclonal Abs for the following markers: CD4 (allophycocyanin), CD8α (allophycocyanin), CD19 (allophycocyanin), CD11c (allophycocyanin), Ly6c (FITC; BD Biosciences), CD11b (allophycocyanin-Cy7), CD117 (allophycocyanin), TCRβ (allophycocyanin); and streptavidin-PE were also used. All Abs were purchased from eBioscience unless otherwise indicated. All cells staining for allophycocyanin were excluded, and cells positive for PE (CD115), high expression of FITC (Ly6c), and positive for allophycocyanin-Cy7 (CD11b) were collected and used for monocyte labeling and injection. Thus, FACS-purified monocytes from bone marrow of tumor cell-injected mice had the following immunophenotype: CD115⁺ Ly6C^hi CD11b^hi CD4⁻ CD8α⁻ CD11c⁻ CD19⁻ NK1.1⁻ CD117⁻. Tumor cell injection did not affect activation markers on harvested monocytes (no upregulation of MHC class II, CD40, CD80, CD86, or CD69). Monocytes were resuspended at 5 × 10⁶ cells/ml in RPMI 1640 complete medium (containing 10% FBS, 1% l-glutamine, 1% nonessential amino acids, 1% penicillin/streptomycin). CMFDA Cell Tracker Green (Invitrogen) or CMTMR Cell Tracker Orange (Invitrogen) were added at a final concentration of 5 μM and incubated in a water bath at 37°C for 7 min. Cells were washed with cold RPMI 1640 complete medium three times and resuspended in sterile PBS at a 1:1 ratio. A total of 2 × 10⁶ to 2.5 × 10⁶ monocytes were injected intravenously into Ccr2^−/− mice on day 7 postinfection. Cell migration was calculated based on the ratio detected in the output organ (CNS or blood) compared with the initial input ratio of labeled cells at the time of cell transfer (Ccr2^+/+:Ccr2^−/− cells).

We and others (2, 4) have previously reported that mRNA for most inflammatory chemokines is absent or expressed at very low levels in the brains of wild-type C57BL/6 mice but can be induced by infection with WNV soon after the virus enters the brain. In this study, we extended these results by using quantitative PCR and found that mRNA for all of the Ccr2 ligands was induced in the CNS. At day 7, the rank order was Ccl2 > Ccl7 > Ccl12 > Ccl8. At day 12, the induction was greatly reduced compared with that at day 7, but mRNA for Ccl2 remained elevated compared with that for the other ligands, which appeared to be induced equivalently (Fig. 1A, 1B). We also tested several chemokine receptors, including Ccr5 and Cxcr3, both of which have previously been demonstrated to be important for control of WNV in the CNS and their respective ligands, Ccl3, Ccl4, Ccl5 (Ccr5 ligands), Cxcl9, Cxcl10, and Cxcl11 (Cxcr3 ligands). Relative to these, mRNA for Ccr2 and its ligands were induced by WNV to a similar level in the CNS (2, 10).

Because Ccr2 and three of its ligands were induced in the CNS after WNV infection, we hypothesized that Ccr2 may be critical for pathogenesis. To test this, we compared survival in Ccr2-deficient (Ccr2^−/−) mice with that in wild-type controls (Ccr2^+/+ mice). Compared with control mice, where survival was ∼65%, Ccr2 deficiency resulted in markedly increased mortality (∼20% survival), as shown in Fig. 1D. This was associated with increased viral load in the CNS of Ccr2-deficient mice on day 12 postinfection (Fig. 1E). This appeared to be specific to the brain, as viral load in the blood was not elevated in Ccr2^−/− mice compared with that in controls (Fig. 1F).

To characterize further the involvement of Ccr2 in the CNS and to test whether differences in the Ccr2^−/− mice may exist in their intrinsic capacity to produce Ccr2 ligands, we compared the induction of a panel of inflammatory chemokines at the RNA level between the Ccr2^+/+ and Ccr2^−/− mice, including Ccl2, Ccl7, Ccl8, and Ccl12, on days 7 and 12 postinfection (Fig. 2A, 2B). No significant differences were observed at either time point, suggesting that the capacity to broadcast proinflammatory signals in response to WNV infection in the CNS was not significantly altered in the Ccr2^−/− mice. We also tested the accumulation of the Ccr2 ligands at the protein level (Fig. 2C). Ccl2, Ccl7, and Ccl12 but not Ccl8 protein was induced by day 7 postinfection, and this induction was similar for wild-type and knockout mice. However, Ccl2 and Ccl7 protein levels were significantly elevated in knockout mouse brain over those in wild-type mouse brain on day 12 postinfection, suggesting that Ccr2 may be involved in protein clearance.

To examine further the factors in the CNS that may be involved in viral clearance, we conducted a thorough analysis of leukocyte accumulation after WNV infection. As previously reported (2), uninfected wild-type mouse brain contained few leukocytes other than microglia. Postinfection, no significant leukocyte accumulation was observed in the CNS until day 7, when monocytes, neutrophils, CD4⁺ and CD8⁺ T cells, and NK cells began to accumulate in the CNS of Ccr2^+/+ mice (Fig. 3). In Ccr2^−/− mice, similar levels of CD4⁺ and CD8⁺ T cells and NK1.1⁺ cells were observed as those in control mice (Fig. 3A–C). The one major exception was the inflammatory monocyte population (Ly6c^hiCD11b⁺), which was severely deficient at all time points tested in Ccr2^−/− mice relative to wild-type controls (Fig. 3D). A representative FACS plot at day 11 postinfection is shown in Fig. 3F (upper gate). Compared with infected wild-type mice, infected Ccr2^−/− mice also had a small increase in the number of neutrophils accumulating in the CNS, although the difference did not reach statistical significance (p > 0.05; Fig. 3E, 3F [lower gate]). The selective deficiency in monocyte accumulation in knockout mice was further validated by histopathological examination of WNV-infected Ccr2^+/+ and Ccr2^−/− mice stained for Mac-2 and CD3. Infected Ccr2^+/+ mice showed immunoreactivity for Mac-2 throughout the brain, with intense staining in the hippocampus and thalamus, whereas brain from infected Ccr2^−/− mice showed markedly fewer immunoreactive cells (Fig. 4). In contrast, the density of CD3-immunostaining cells was similar for CD3 in brain for the infected control and knockout mice, consistent with the FACS data shown in Fig. 3. The data indicate that Ccr2 may be specifically involved in the accumulation of inflammatory monocytes in the CNS and are consistent with the known role of Ccr2 as major chemokine receptor on inflammatory monocytes.

Recent studies have identified a function for Ccr2 in the mobilization of inflammatory monocytes from the bone marrow to the blood under homeostatic conditions (18, 25). Unstressed Ccr2^−/− mice are severely monocytopenic, and this is dependent on both Ccl2 and Ccl7 as both Ccl2^−/− and Ccl7^−/− mice are also monocytopenic (17). Thus, the deficiency of inflammatory monocytes in the CNS of WNV-infected Ccr2^−/− mice may be due to monocytopenia or the absence of Ccr2-dependent monocyte trafficking from blood to brain, or a combination of the two. To address these possibilities, we first measured the effect of WNV infection on the levels of monocytes and other leukocyte subsets in the blood over time. In Ccr2^+/+ mice, infection resulted in ∼5-fold induction of monocyte numbers by day 5 postinfection (Fig. 5A). By days 7, 9, and 11, the level of monocytes decreased in the blood of these mice, coincident with their accumulation in the CNS (Fig. 3D). In contrast, before infection Ccr2^−/− mice were severely monocytopenic, as has been previously described, and WNV infection failed to induce monocytosis. Representative FACS plots are shown for day 0 (uninfected) and day 7 postinfection for both Ccr2^+/+ and Ccr2^−/− mice (Fig. 5C). Other leukocyte subsets in the blood remained relatively unaffected by WNV throughout the time course of the infection, and levels were similar for Ccr2^+/+ and Ccr2^−/− mice (Fig. 5B, 5D). These data indicate that WNV infection induces an early and highly selective monocytosis that is entirely Ccr2 dependent.

Because Ccr2-dependent monocytosis is also dependent on Ccl2 and Ccl7, but not Ccl12, we measured the induction and accumulation of these ligands in the blood during the course of infection as shown in Fig. 5E (17, 25). As measured by ELISA, we found that Ccl2 levels in the blood of Ccr2^+/+ mice were low in steady-state condition (0.37 ± 0.05 ng/ml), and these levels increased 6-fold at day 7. In Ccr2^−/− mice, a higher steady-state level was observed (0.69 ± 0.10 ng/ml). However, during infection, Ccl2 increased ∼11-fold, peaking at day 9 postinfection. An even more profound increase was observed for Ccl7, where protein was undetectable at the steady state of both Ccr2^+/+ and Ccr2^−/− mice. However, Ccl7 accumulation in the Ccr2^−/− mice after WNV infection was measured at 28.10 ± 0.70 ng/ml, which was ∼10-fold higher than what we observed in Ccr2^+/+ mice (2.9 ± 1.65 ng/ml). Analysis of Ccl12 showed no significant differences at any time point between the two strains. Our data are consistent with the idea that these ligands are important for monocytosis in the blood, and without these cells, the ligands may accumulate.

To test the role of Ccr2 in monocyte trafficking from blood to CNS directly, we conducted a competitive repopulation study where equal numbers of Ccr2^+/+ and Ccr2^−/− Ly6c^hi monocytes were sorted by FACS from uninfected mice, then differentially labeled, mixed 1:1, and injected into WNV-infected Ccr2^−/− recipient mice on day 7 postinfection. This time point is after virus has entered the CNS and when chemokine production is highest (Fig. 2). We used Ccr2^−/− mice as recipients because they are monocytopenic, which would facilitate analysis of donor cell trafficking to brain. Twelve hours after transfer, mice were euthanized, and the CNS was analyzed for the presence of the labeled cells. As shown in Fig. 6A, the ratio of labeled Ccr2^+/+ and Ccr2^−/− monocytes prior to injection was ∼1:1, as intended (left panel). Twelve hours postinjection, the ratio of Ccr2^+/+:Ccr2^−/− monocytes in the CNS was strongly skewed toward Ccr2^+/+ monocytes at approximately a 4:1 ratio. However, this preponderance of Ccr2-sufficient cells in the CNS was also observed in blood collected from the same mouse at the same time. This pattern held up for all animals tested (n = 9; Fig. 6B), suggesting that the increased proportion of Ccr2⁺ monocytes in the CNS was not due to a preferential trafficking of Ccr2-expressing monocytes but rather to the disappearance of the Ccr2-deficient monocytes from the blood.

To test whether the unequal distribution of Ccr2^+/+ versus Ccr2^−/− donor monocytes we observed was caused by WNV infection, we performed the same experiment using uninfected Ccr2^−/− mice as recipients (Fig. 6C, 6D). Blood was collected at 3 and 18 h postinjection and analyzed for the ratio of differentially labeled Ccr2^+/+:Ccr2^−/− monocytes by FACS. As shown in Fig. 6C (left-most panel), the ratio of Ccr2^+/+:Ccr2^−/− monocytes prior to injection was ∼1:1, as intended. At 3 h postinjection, the ratio of Ccr2^+/+:Ccr2^−/− cells had increased, indicating an advantage of Ccr2^+/+ monocytes for populating the blood. At 18 h, the ratio had increased further and was similar to that observed when the cells were injected into WNV-infected mice (Fig. 6A). Thus, the skewed repopulation is not due to WNV infection per se. Analysis of spleen tissue from uninfected mice 18 h after injection showed a skewed distribution of donor cells similar to that found in the blood (data not shown). However, analysis of bone marrow from uninfected Ccr2^−/− mice 18 h after injection of donor cells revealed that Ccr2^+/+ and Ccr2^−/− monocytes were present at nearly a 1:1 ratio (data not shown). Together, these data indicate that monocytes may traffic back to the bone marrow and again accumulate, unable to egress in the absence of Ccr2.

The current study demonstrates a critical role for chemokine receptor Ccr2 on Ly6c^hi inflammatory monocytes for accumulation of these cells in the brain in a mouse model of WNV encephalitis. This extends to the molecular level previous work showing that monocytes accumulate in the brains of WNV-infected mice and humans and are important for survival in the mouse model (2, 3, 5, 14). Consistent with this, we found that Ccr2-deficiency resulted in markedly decreased survival and increased viral load in the brains after WNV infection of mice.

The detailed leukocyte dynamics that we defined in the blood and brains of WNV-infected mice allow important new insights into the role of Ccr2 in WNV pathogenesis, based on several unexpected findings. First, although WNV induces a multifocal mixed leukocyte encephalitis, leukocyte deficiency in the brain in WNV-infected Ccr2 knockout mice appears to be restricted exclusively to monocytes (∼90% reduction). This suggests that Ccr2 is most important in monocyte trafficking in mouse brain, however additional work will be needed to define whether there is also a deficiency of T cell subsets and other leukocyte subsets known to express Ccr2, such as highly differentiated Th1 cells that are compensated by subsets that do not express the receptor.

Second, although Ccr2 ligands Ccl2, Ccl7, and Ccl12 are strongly induced by WNV in the brain at the RNA and protein levels, Ccr2 appears to be dispensable for trafficking of monocytes from the blood to the brain in WNV-infected mice. Instead, the driving force appears to be a highly selective, WNV-induced, and entirely Ccr2-dependent monocytosis that precedes accumulation of monocytes in the brain. This conclusion is based in part on competitive repopulation experiments performed using differentially labeled Ccr2^+/+ and Ccr2^−/− donor monocytes transferred to monocytopenic WNV-infected Ccr2^−/− mice. These studies also suggest that monocytes lacking Ccr2 may traffic back to the bone marrow and preferentially accumulate, unable to egress. Thus the mechanism of monocytopenia in uninfected Ccr2^−/− mice may involve both decreased monocyte egress from bone marrow to blood and increased monocyte return from blood to bone marrow, among other possibilities.

Because Ccr2 appears not to be involved in the migration of monocytes from blood into the CNS, this suggests that another chemokine receptor may be involved at this level. Our previous studies have shown a dominant role of chemokine receptor Ccr5 in migration of monocytes, T cells, and NK cells into the CNS (2). However, our data suggest that Ccr2 is involved very early postinfection with WNV, increasing monocyte numbers in the periphery. Once monocytes are in circulation, it appears that another receptor is critical for their migration into the infected CNS, perhaps Ccr5. To address this question, a similar competitive repopulation analysis to the one we conducted in Fig. 6 could be performed using monocytes derived from Ccr5^+/+ and Ccr5^−/− mice.

We are unaware of any other infectious agent that induces a selective monocytosis. As with WNV, other infectious agents have been reported that fail to overcome the monocytopenia in Ccr2-deficient mice, but in these diseases a specific monocytosis has not been described as a typical feature. Additional work will be needed to define whether Ccr2 and its ligands are needed for trafficking, organization, and/or activation of monocytes within the brain, once the cells have crossed the blood–brain barrier. Although previous work has shown that both Ccl2 and Ccl7 are important determinants of the normal monocyte count in unstressed wild-type mice, additional work will be needed to define which Ccr2 ligands induce monocytosis in WNV-infected mice.

Although previous studies have shown that monocytes play an important role in controlling WNV infection in mice, their precise mechanism of action has not been defined and may vary depending on the model. One study using a lethal intranasal model has suggested that monocytes may play a pathogenic role, as delaying migration of these cells prolonged survival. The mechanism may involve macrophage/microglial cell production of substances potentially toxic to tissue, such as superoxide anion, NO, and proinflammatory cytokines (26, 27). In contrast, a second study using a model very similar to ours showed that monocytes may be protective, as depletion using clodronate-loaded liposomes decreased survival of the infected mice (15). This is also plausible, as a beneficial role of macrophages/microglia in CNS recovery has also been demonstrated through the production of neurotrophic factors, removal of debris, and axonal regeneration (28–31). Future approaches for defining functional roles of monocytes in controlling WNV infection, once they reach the CNS, may include 1) histopathologic studies of the correspondence of virus, macrophages, and neurons in vivo; 2) ex vivo studies of expression of proinflammatory cytokines and effector molecules in monocytes and monocyte-derived cells (macrophages and microglia) purified from infected brain; 3) immunophenotyping the cells that produce Ccr2 ligands after viral entry into the CNS; and 4) in vitro studies of monocyte/macrophage interaction with virally infected neurons.

At present, there are no data delineating leukocyte dynamics in patients with WNV infection. Our data suggest that factors limiting monocytosis could be associated with poor outcome and that novel therapeutics able to induce monocytosis, such as GM-CSF and AMD3100 (a CXCR4 antagonist), could be useful in such patients. AMD3100 is currently approved for stem cell mobilization and can also mobilize monocytes in both humans (32) and in mice (33). In this regard, a study by Klein and colleagues (12) showed that AMD3100 can significantly improve survival from WNV infection in mice, although the mechanism of leukocytosis has not been defined.

Monocytosis could also potentially be induced in humans using agonists to CCR2, as the “inflammatory” monocyte subset in humans (CD14⁺CD16⁻ monocytes) that is analogous to the Ccr2⁺ Ly6C^hi mouse monocyte subset that we studied also uniformly expresses CCR2 (16). At present, CCR2 agonists are not available clinically and are in general limited to chemokine ligands, which have very short half-lives. Unless modified, CCR2 ligands may not be ideal for such an application. Conversely, CCR2 antagonists, which are currently being developed by the pharmaceutical industry with anticipated long-term administration for chronic inflammatory diseases, could potentially predispose to poor outcomes in WNV-infected patients.

In previous work, we have identified two genetic determinants of WNV susceptibility in humans, homozygous CCR5Δ32 (7, 8) and OAS1 rs10774671 (34). These mutations were tested as candidates based on work we performed with the corresponding mouse genes in mouse models of WNV (35, 36). A similar genetic approach may also be available for translating our findings on Ccr2 to human WNV disease, as several common polymorphisms have been described for CCR2, CCL2, and CCL7 (37, 38). Understanding the mechanisms of Ccr2 action in monocyte migration during WNV infection in mice is an important first step toward these goals.

In conclusion, our data identify Ccr2 as a critical protective factor in encephalitis caused by WNV in a mouse model. To our knowledge, our data show for the first time that chemokine receptors may function at an entirely different migratory step than previously shown for WNV infection and that early events after infection may affect outcome of WNV encephalitis. Further studies are needed to investigate the role of Ccr2 ligands and to understand further the functional role of monocytes in the CNS.

We thank Dr. Jerrold Ward and Lawrence Faucette for assistance with immunohistochemistry and pathology.

Disclosures The authors have no financial conflicts of interest.