|Home | About | Journals | Submit | Contact Us | Français|
Macrophages are an important cell type for regulation of immunity, and can play key roles in virus pathogenesis. Here we address the effect of infection of primary human macrophages with the related paramyxoviruses Parainfluenza virus 5 (PIV5) and Mumps virus (MuV). Monocyte-derived macrophages infected with PIV5 or MuV showed very little cytopathic effect, but were found to be defective in migration toward a gradient of chemokines such as macrophage colony stimulating factor (MCSF) and vascular endothelial growth factor (VEGF). For MuV infection, the inhibition of migration required live virus infection, but was not caused by a loss of chemokine receptors on the surface of infected cells. MuV-mediated inhibition of macrophage chemotaxis was through a soluble factor released from infected cells. MuV infection enhanced secretion of TNF-α, but not macrophage inhibitory factor (MIF). Antibody inhibition and add-back experiments demonstrated that TNF-α was both necessary and sufficient for MuV-mediate chemotaxis inhibition.
Macrophages are an important cell type that plays a crucial role in modulating both innate and adaptive immune responses to viral infection (reviewed in Mosser and Edwards 2008; Murray and Wynn, 2011). Upon detection of a virus, macrophages are recruited to the site of infection through the release of chemokines by epithelial cells and innate immune cells. Once localized to the site of infection, macrophages are able to phagocytose viruses or infected cell debris, process and present antigen, and migrate to lymph nodes to activate T cells and stimulate the adaptive immune response (Lehtonen et al 2007). However, a fine regulation of macrophage trafficking is needed to avoid chronic inflammation, impaired tissue healing, and recurrent infection in a host (Frascaroli et. al. 2009). Therefore, it is important that macrophages are readily able to move to the site of an infection, be retained in those tissues to combat a pathogen, and then egress out of the tissues to avoid overt damaging the host. In this study, we address the question of the impact of two related paramyxoviruses, Mumps Virus (MuV) and Parainfluenza virus 5 (PIV5) on macrophage migration.
Macrophages are specialized lymphocytes whose functions include a number of important aspects in innate and adaptive responses to virus infection (Lehtonen et al 2007). In some cases such as with influenza virus, the ability of macrophages to respond to virus has been shown to be crucial in determining the outcome of respiratory tract infection (Kim et. al. 2008, Wijburg et. al. 1997). Macrophages have been shown to express high levels of the co-stimulatory molecules CD80 and CD86 on their cell surface, a property which allows these cells to present antigen to lymphocytes (Mosser and Edwards 2008, Murrah and Wynn, 2011; Wijburg et. al. 1997). In addition, macrophages respond to pathogens by secreting pro-inflammatory cytokines such as IL-6, IFN-β, RANTES, and TNF-α (Assuncao-Miranda et al., 2010; Mosser and Edwards 2008), and these cytokines are able to act in both an autocrine and paracrine fashion to stimulate additional innate and adaptive immune responses. Here, we demonstrate that MuV infection of primary human macrophages induces the production of TNF-α, and this MuV-mediated production of TNF-α has a profound effect on macrophage function.
PIV5 is a prototype virus which has served as a model parainfluenza for the study of paramyxoviruses in general and the Rubulavirus family in particular (reviewed in Lamb and Parks, 2007). PIV5 is closely related to two other Rubulaviruses: MuV and Human Parainfluenza virus type 2. MuV is the causative agent of mumps in humans, a viral infection of children and adolescents characterized by swelling of the parotid glands (Hviid et. al. 2008). In addition, MuV infection often results in secondary complications, including aseptic meningitis, deafness, sterility in males, and can be highly neurotropic (Rubin and Afzal, 2011). Although the introduction of the MuV vaccine has been successful in reducing infection in the general population, there have been increasing numbers of MuV infections reported in vaccinated populations. Available data support waning immunity as the mechanism behind these outbreaks, as opposed to immune escape (Rubin et al., 2011).
Given that macrophages rely heavily on chemotaxis to perform their functions for both innate and adaptive immune responses, we have tested the effects of PIV5 and MuV infection on macrophage motility. Using primary human macrophages, we demonstrate that infections with both PIV5 and MuV show minimal cytopathic effect (CPE), but infected macrophages have a significantly reduced ability to migrate toward chemoattractants. For MuV, TNF-α secreted during viral infection was found to be both necessary and sufficient to account for the defect in macrophage migration. Our results have implications for potential treatment of viral infections that are capable of exploiting normal host cell responses for their own benefit.
Immature peripheral blood mononuclear cells (PBMCs) were derived from whole human blood as described previously (Arimilli et. al. 2006, Briggs et al. 2011). Briefly, PBMCs were isolated from randomly selected donors by standard density gradient centrifugation on Ficoll-Hypaque. CD14+ monocytes were isolated by positive selection using CD14+ microbeads (Miltenyi-Biotec). Generally, the isolated cells were found to be ≥95% CD14+ as assayed by flow cytometry. The resulting CD14+ monocytes were cultured for 6 d in RPMI media (Lonza) containing 10 ng/mL macrophage colony stimulating factor (MCSF) supplemented with 10% FBS, and 1% each of L-glutamine, penicillian/streptomycin, NEAA, 1M HEPES, and sodium pyruvate (Lonza). After 3 d in culture, half of the media was replaced, and additional MCSF was added at the concentrations listed above. At day 6, ≥90% of cells were CD64+, CD14+ (data not shown). Data are representative of the results obtained from multiple experiments with individual cells from different donors. Donor cells were not pooled for experiments. WT PIV5-GFP and the P/V-CPI- mutant have been described previously (Wansley and Parks, 2002). MuV (Enders Strain, ATTC VR-1379) was grown in CV-1 cells as described (Johnson et. al. 2008).
In radiolabel experiments, macrophages were mock infected or infected at a high moi with virus as indicated in figure legends. At the time pi indicated in figure legends, cells were starved for 25 min with DMEM lacking cysteine and methionine and then radiolabeled for 2 hrs with 100 μCi/ml Tran[35S]-label. Equal amounts of protein were immunoprecipitated using rabbit polyclonal antisera to PIV5 P proteins as described previously (Dillon et. al. 2006), before SDS-PAGE and autoradiography.
Human recombinant MCSF, Regulated upon Activation, Normal T Cell Expressed and Secreted (RANTES/CCL5), vascular endothelial growth factor (VEGF), or monocyte chemotactic protein-1 (MCP-1/CCL2) were from Invitrogen and were each used at the final concentration of 100 ng/ml in migration media (RPMI 1640 supplemented with 1% FBS). Macrophages were removed from culture dishes using Accutase (Millipore, Inc), pelleted and resuspended in migration media. Cells were seeded at a density of 2×105 cells/well on one side of the culture dish. Chemotaxis was evaluated using 24-well Boyden chambers (BD Biosciences) containing a 3–μm pore polycarbonate filters as previously described (Frascaroli et. al. 2009). Cells were allowed to migrate across the filter for 5 h at 37°C in a tissue culture incubator. Filters were then removed, fixed and stained using Diff QUICK Stain kit (IMEB Inc.) Using a microscope, the number of cells that had migrated to the opposing side of the filter was calculated as the number of cells counted in ten consecutive high power fields. All stimuli were assayed in triplicate, and results were expressed as mean number of cells ± SD. To determine the effect of TNF-α on migration, a neutralizing antibody against TNF-α (Biosource Inc.) or IFN-β (PBL International) was added to cells immediately following infection at a concentration of 10 ug/mL. Exogenous TNF-α (Promega) was added to naïve macrophages at a concentration of 100 pg/mL.
Immunoreactive IL-6, TNF-α (BD Opt EIA; BD Biosciences), or MIF (Ray Biotech) was quantified by ELISA according to manufacturer’s protocols, and levels were normalized to 106 cells. Cell viability was measured at the indicated times pi by using a CellTiter 96 AQueaous One Solution cell proliferation assay (Promega) according to the manufacturer’s instructions. Data were from quadruplicate samples and are expressed as fold increase in cell viability relative to mock infected cells.
For detection of surface markers, cells were stained with antibodies specific for the CCR1 and CCR5 receptors according to manufacturer’s recommendations (BD Pharmigen). Samples were analyzed on a FACScan Caliber instrument using Cell Quest software (Becton-Dickinson).
In immunofluorescence experiments, cells were infected at high moi and at 24 h pi were fixed at 4°C in ice-cold 100% methanol. After washing, cells were incubated with a primary antibody against the MuV NP protein (gift of Tony Schmidt, Penn State University), and a secondary antibody conjugated to a 488 fluorophore (Alexafluor). Phase and fluorescent microscopy were carried out as described previously using a Nikon Eclipse microscope and a 20x lens (Wansley and Parks, 2002). Images were captured under phase contrast or visualized for GFP or DAPI expression using Qimaging digital camera and processed using Q-capture software. Exposure times were manually set to be constant among samples.
To determine the outcome of infection of primary human macrophages with PIV5, CD14+ monocytes were isolated from human PBMCs and cultured for six days in the presence of MCSF. At that time, >90% of cells were found to be CD14+, CD11b+, and CD64+, a characteristic marker profile of monocyte-derived macrophages (MDMs, Briggs et. al.. 2011). Macrophages were mock infected or infected with WT PIV5 that encoded GFP at an moi of 10 and then examined by microscopy at 24 h pi. As shown in Fig 1A, PIV5 was capable of initiating a productive infection of human macrophages as evidenced by new gene expression. However, no progeny virus was detected (data not shown). Importantly, PIV5 infection did not show overt CPE compared to mock infected macrophages. To confirm the lack of CPE, macrophages from three donors were infected with WT PIV5-GFP at an moi of 10. At 24 h pi, an MTT assay was done to evaluate cell viability. As shown in Fig 1B, infected macrophages from all three donors showed a slight reduction in cell proliferation, similar to that seen during infections of human epithelial cells in culture (Manuse and Parks, 2010; Wansley and Parks, 2002) and consistent with the finding that PIV5 slows the cell cycle (Lin and Lamb, 2000). Thus, in contrast to the dramatic cell death seen with WT PIV5 infection of human DC (Arimilli et al, 2006), infected primary human macrophages display little CPE.
We have previously shown that WT PIV5 is a poor inducer of proinflammatory cytokines in human DC and epithelial cells. To determine if cytokines were induced in infected macrophages, human MDMs were mock infected or infected with WT PIV5 at an moi of 10, and levels of IL-6 were assayed by ELISA at days 1, 2, and 3 pi. As shown in Fig. 1C, PIV5-infected macrophages produced only low levels of IL-6, in contrast to cells treated with the positive control stimulus LPS or infected with the cytokine-inducing PIV5 P/V mutant (Wansley and Parks, 2002). PIV5 infection also resulted in minimal increases in MIF and TNF-α (not shown). Thus, cytokine induction by primary human MDMs infected with PIV5 is similar to that seen with human epithelial cells.
To determine if macrophage chemotaxis was altered by PIV5 infection, primary human macrophages were infected with WT PIV5-GFP at an moi of 10. At 24 h pi, cells were collected, plated in a transwell culture dish, and allowed to undergo chemotaxis for 4 h toward the chemokines MCP-1 or MCSF, or toward PBS as a negative control. Filters were fixed and stained, and the average number of cells per 10 microscopic fields was calculated. As shown in Fig 1D, mock infected cells (hatched bars) showed substantial migration toward both MCP-1 and MCSF. In contrast, PIV5-infected macrophages displayed significantly reduced migration toward these two chemokines. Taken together, these results indicate that while WT PIV5 infection of macrophages appears to have little overt effect on cell viability or cytokine production, infected cells have a reduced capacity to migrate toward chemoattractants.
Based on the close relatedness of MuV and PIV5, we hypothesized that MuV infection of primary human macrophages would also result in restricted macrophage chemotaxis. As a first step, we determined if macrophages were susceptible to MuV infection and the relative CPE caused by infection. Primary human macrophages were mock infected or infected with PIV5 as a positive control or MuV at an moi of 10. At 24 h pi, cells were radiolabled with 35S-labelled methionine and cysteine, and the P protein was immunopreciptated from cell lysates using anti-P antibody and SDS-PAGE. As shown in Fig 2A, MuV-infected cells produced P protein, similar to PIV5 infected cells. To confirm that most cells in the culture were susceptible to infection, macrophages were infected at an moi of 10 with MuV and examined at 24 h pi by immunofluorescence with an antibody against the MuV NP protein. As shown in Fig 2B, the majority of the cells in the population showed strong labeling for NP protein, which was not seen in mock infected cells. However, similar to that seen with PIV5, no progeny MuV was detected (data not shown). For unknown reasons, many cells showed strong nuclear staining for NP. This was not limited to NP, since nuclear staining was also seen for both GFP (see Fig. 1A) and for P protein (not shown). The role of nuclear accumulation of viral proteins is unknown.
Similar to the finding with PIV5, MuV infection did not result in a large loss of macrophage viability. This is evident in the viability assay for cells from three donors shown in Fig. 2C, where the cell viability in MuV-infected cells was ~100% of that seen with mock infected cells. Similarly, analysis of MuV-infected macrophages for apoptotic markers such as AnnexinV staining showed no significant difference compared to mock infected cells (Fig. 2D). Taken together, these data indicate that primary human macrophages are susceptible to infection with MuV, and that this infection does not have a significant impact on cell viability at 24 h pi.
To determine if MuV infection reduced the ability of macrophages to migrate toward chemoattractants, MDMs were mock infected or infected with MuV at an moi of 10. At 24 h pi, macrophages were analyzed for migration toward MCSF. As shown in Fig 3A, mock infected macrophages displayed low levels of spontaneous migration toward the control PBS, but high levels of migration toward MCSF. By contrast, macrophages infected with MuV were blocked in migration toward MCSF. Decreased migration by MuV-infected macrophages was independent of which chemokine was used as an attractant (Fig. 3B). Mock infected macrophages migrated toward MCSF, RANTES and VEGF, while migration was inhibited toward all of these chemoattractants in the case of MuV infected cells. As shown in Fig. 3C, the MuV-induced inhibition of migration toward MCSF was dependent on the moi, with minimal effect seen at an moi of 1, but strong inhibition seen when cells were infected at mois of 5 and 25. Treatment of macrophages with UV-inactivated MuV virions did not result in a significant decrease in macrophage chemotaxis (data not shown). It is noteworthy that the number of cells which migrated toward a chemokine and the level of spontaneous migration differed between individual cell preparations (e.g., compare the y axis in Fig. 3 panels), as would be expected when assaying primary cells from a range of donors. However, in more than 10 assays MuV infection always significantly inhibited migration compared to mock infected samples.
One possible explanation for decrease chemotaxis of MuV-infected macrophages could be that chemokine receptors are downregulated. To test this hypothesis, primary human macrophages were mock infected or infected with MuV at an moi of 10, and MDMs were surface stained at 24 h pi with antibodies against either CCR1 or CCR5, two known receptors for the chemokine RANTES but not for MCSF or VEGF. As shown in Fig. 3D, no significant decrease in CCR1 or CCR5 expression was seen in MuV-infected macrophages when compared to mock infected macrophages.
As an alternative, we tested the hypothesis that a soluble factor produced during MuV infection is responsible for the inhibition of chemotaxis. Primary human macrophages were mock infected or infected with MuV or PIV5 at an moi of 10. At 24 h pi, media from these cells was removed and treated with UV light to inactivate any live virus. The UV-inactivated media was added to naive uninfected macrophages overnight, and the macrophages were then used in a migration assay toward MCSF. As shown in Fig 4A, cells treated with media from mock infected cells were capable of migration toward MCSF. In contrast, UV-inactivated media from either PIV5- or MuV-infected cells had a significant inhibiting effect on migration of uninfected macrophages. These data demonstrate that a soluble factor released from virus-infected macrophages is capable of reducing chemotaxis in the absence of a live virus infection.
Macrophage inhibitory factor (MIF) is produced in response to pathogens as a mechanism to retain macrophages at the site of infection (Baugh and Bucala, 2002; Calandra and Rogers 2003). To determine if MIF was produced by MuV infected macrophages, the amount of MIF released from infected cells was assayed by ELISA. Since an moi of 25 was maximal at inhibition of migration (see Fig. 3C), this moi was use in order to maximize differences in cytokine induction. As shown in Fig 4B, mock infected macrophages displayed low levels of MIF, which is consistent with reports that low levels of MIF are constitutively produced by some cells such as macrophages (Calandra and Rogers 2003). The positive control sample from A549 cells infected with the cytokine-inducing PIV5 mutant P/V-CPI-, showed high levels of MIF. Importantly, MuV-infected macrophages produce the same levels of MIF as that seen with mock infected macrophages, a finding that is inconsistent with an MIF-mediated inhibition of macrophage migration.
TNF-α has been shown to inhibit chemotaxis (Grimshaw and Balkwill, 2001; Herriott et. al. 1993). To determine if TNF-α was being produced by MuV-infected macrophages, media was collected from infected MDMs and analyzed by an ELISA specific for TNF-α. As shown in Fig 4C for three experiments with three different donors, MuV-infected macrophages had elevated levels of TNF-α compared to mock infected macrophages.
To test the role of TNF-α in MuV-mediated inhibition of chemotaxis, MDMs were mock infected or infected with MuV at an moi of 10. An excess of a neutralizing antibody against TNF-α, or IFN-β as a control, was added to media immediately after infection, and cell migration was assayed as described above. As shown in Fig 5A, MuV infected macrophages had significantly reduced migration toward MCSF, but this inhibition was relieved in samples treated with the anti-TNF neutralizing antibody. It is noteworthy that the neutralizing anti-TNF-α antibody also enhanced the migration of mock infected cells (hatched bars, Fig. 5A). This is consistent with the finding of low levels of TNF-α that are release constitutively from macrophages from individual donors (see Fig. 4C). This enhancing effect of neutralizing antibody on cell migration was not seen with a control IFN-β antibody (Fig. 5B). MuV-infected macrophages treated the neutralizing antibody against IFN-β displayed a reduced level of migration, similar to that seen in non-treated, infected macrophages.
To determine if TNF-α alone was sufficient to inhibit macrophage chemotaxis, primary human macrophages were treated overnight with exogenous TNF-α, at levels similar to the amount of TNF-α produced during MuV infection seen in Fig 4C. Treated macrophages were used in a migration assay as described above but without prior MuV infection. As shown in Fig 6, uninfected macrophages treated with PBS as a control were able to migrate toward MCSF. Most importantly, uninfected macrophages treated with exogenous TNF-α were deficient in their ability to migrate toward MCSF. Together, these data demonstrate that TNF-α is both necessary (Fig. 5A) and sufficient (Fig 6) for MuV-mediated inhibition of macrophage chemotaxis.
Macrophages are important cells of both innate and adaptive immunity that play a crucial role in controlling viral infection. Our results indicate that primary human macrophages are susceptible to infection with both PIV5 and MuV, and that these infections do not appear to induce overt CPE. We have shown that macrophages infected with either PIV5 or MuV are inhibited from migration toward a range of chemoattractants. For MuV, this inhibition is caused by a soluble factor secreted during MuV infection. Our most striking finding came from our work demonstrating that the production of TNF-α by MuV-infected macrophages is both necessary and sufficient to inhibit macrophage migration. As detailed below, these data suggest a model whereby production of TNF-α during MuV infection acts in an autocrine and paracrine fashion to inhibit migration of naïve and infected macrophages alike.
We have previously shown that WT PIV5 induces massive CPE and apoptosis in primary human monocyte-derived dendritic cells (Arimilli et al., 2006). This contrasts with the findings here that WT PIV5 infection of MDMs resulted in a productive infection with no evidence of overt CPE or apoptosis. Thus, PIV5 infected macrophages show properties similar to that of infected epithelial or fibroblast cells: low CPE and poor induction of cytokines such as IL-6 and IFN-beta (He et al., 2002; Wansley and Parks, 2002). The mechanism for these widely differing outcomes of PIV5 infection of dendritic cells and macrophages is currently under investigation. Most importantly for this study, PIV5 infection of MDMs did not induce substantial levels of MIF-1 or TNF-α, consistent with the low induction of all cytokines by this virus. Thus, while PIV5 infection inhibited macrophage migration, it appears to be through a different mechanism from that of MuV.
MuV infected macrophages were strongly inhibited in chemotaxis toward a range of chemokines which utilize different receptors to induce migration. For example, the receptor which detects MCSF and VEGF is a homodimeric type III receptor tyrosine kinase (Hamilton, 2008; Holmes and Zachary 2007), whereas RANTES and MCP-1 are detected by G-protein coupled receptors (GPCRs) (Neote et. al. 1993; Monteclaro and Charo 1996). These findings raise the hypothesis that MuV infection might induce a global reduction in levels of cell surface receptors. Flow cytometric analysis indicated that MuV infection did not alter levels of MDM cell surface CCR1 and CCR5. Thus, while we cannot eliminate the possibility of changes at later times pi, our results do not support a model whereby the inhibition of migration is due to altered cell surface receptor expression. In addition, analysis of the macrophage actin cytoskeleton using phalloidin showed no obvious differences in morphology of mock infected versus MuV infected macrophages (data not shown). Thus, the inhibition of migration appears to occur after receptor-ligand binding, but before changes in the macrophage cytoskeleton have occurred.
Our data indicate that a soluble factor produced during MuV infection is responsible for inhibiting the migration of macrophages, consistent with previous results from the 1970s which showed that MuV infection leads to the secretion of migration-inhibitory factors. Parotid gland fluid recovered from chimpanzees with clinical signs of MuV infection was found to decrease macrophage migration toward chemoattractants (Flanagan et. al. 1973; Yoshida et. al. 1974). Due to the discovery of the chemokine MIF, the fluid was described as having “MIF-like” activity, although the identity of the secreted factor was never determined. Our results indicate that the inhibitory factor secreted from MuV-infected macrophages is not MIF, as MuV infection did not elicit levels of MIF above those produced by mock infected cells. This result differs from those obtained using human cytomegalovirus (HCMV) (Frascaroli et. al. 2009), where elevated levels of MIF produced during viral infection were shown to be responsible for the loss of chemotaxis (Frascaroli et. al. 2009).
Why is MIF induced by HCMV infection but not by MuV? The double stranded DNA virus HCMV is most likely detected by different host cell receptors compared to the negative strand RNA virus MuV. While cells such as macrophages have been shown to express MIF constitutively, MIF is significantly upregulated following TLR stimulation (Calandra and Rogers 2003). MIF is induced through TLR2 and TLR4 but not through TLR3/7 (Popa et al., 2006), and HCMV is capable of signaling through TLR2 (Compton et. al. 2003). Conversely, Rubulaviruses such as PIV5 are detected by cytoplasmic RIG-I, and not MDA-5 or TLR3 in epithelial cells (Manuse and Parks 2009). As macrophages express cytoplasmic sensors such as RIG-I (Manuse and Parks, 2009), we hypothesize that differential activation of pathogen sensing pathways results in the induction of MIF by HCMV but not by MuV. Why is PIV5 a much poorer inducer of TNF-alpha compared to MuV? While it is clear that multiple strain differences (both in PIV5 and MuV) can account for differential cytokine induction (e.g., Young and Parks, 2003), we have recently shown that the paramyxovirus polymerase components are an important factor in modulating host cell responses (Dillon and Parks, 2007). Thus, we hypothesize that PIV5 replication produces fewer inducers of host responses compared to MuV replication, and thus induces lower levels of TNF-alpha and other proinflammatory cytokines.
Our most striking result was the contribution of TNF-α to the inhibition of macrophage migration. MuV was shown to be a strong inducer of TNF-α from infected macrophages, and secreted TNF-α was necessary for MuV-mediated inhibition of macrophage migration. Importantly, the inhibition of migration could be reproduced by treatment of uninfected macrophages with TNF-α, indicating that TNF-α was also sufficient for inhibitor effects. This is consistent with prior studies demonstrating that treatment of naïve macrophages with TNF-α is sufficient to inhibit migration (Grimshaw and Balkwill 2001; Herriott et. al. 1993). TNF-α can clearly play important roles in promoting innate and adaptive immunity, but can also be problematic when expression is deregulated. In out model, MuV takes advantage of the induction of TNF-α during replication to prevent macrophage migration and thus, limit the pro-immune functions of macrophages in antiviral responses. The migration of many cell types requires signaling through the mitogen-activated protein kinase (MAPK) pathways, which act through a phosphorylation cascade (Pearson et. al. 2001). Whereas increased levels of MAPK phosphorylation is seen in macrophages exposed to the chemoattractant MCP-1 (Grimshaw and Balkwill 2001), exposure to TNF-α induced increased levels of MAPK phosphatase 1 (MKP-1) (Grimshaw and Balkwill 2001). MKP-1 has been shown to be a negative regulator of MAPK signaling, functioning by de-phosphorylating MAPK (Chi et. al. 2006; Salojin et. al. 2006). We hypothesize that secretion of TNF-α by MuV-infected macrophages is able to signal on cells in an autocrine or paracrine fashion, and upregulate the production of MKP-1. Future work will delineate the effect of MuV on MAPK signaling pathways, including the role of MKP-1 in macrophage migration.
We thank Ellen Young for excellent technical help and Darren Seals for help with the macrophage migration assay. This work was supported by NIH grant DC009619. AEM is supported by NIH Training Award Grant T32 OD010957.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.