Vesicular Stomatitis Virus Matrix Protein Impairs CD1d-Mediated Antigen Presentation through Activation of the p38 MAPK Pathway

Abstract

Natural killer T (NKT) cells are unique T lymphocytes that recognize CD1d-bound lipid antigens and play an important role in both innate and acquired immune responses against infectious diseases and tumors. We have already shown that a vesicular stomatitis virus (VSV) infection results in the rapid inhibition of murine CD1d-mediated antigen presentation to NKT cells. In the present study, it was found that the VSV matrix (VSV-M) protein is an important element in this decrease in antigen presentation postinfection. The VSV-M protein altered the intracellular distribution of murine CD1d molecules, resulting in qualitative (but not quantitative) changes in cell surface CD1d expression. The M protein was distributed throughout the infected cell, and it was found to activate the mitogen-activated protein kinase (MAPK) p38 very early postinfection. Infection of CD1d⁺ cells with a temperature-sensitive VSV-M mutant at the nonpermissive temperature both substantially reversed the inhibition of antigen presentation by CD1d and delayed the activation of p38. Thus, the VSV-M protein plays an important role in permitting the virus to evade important components of the innate immune response by regulating specific MAPK pathways.

Viruses are obligate parasites that rely heavily on the host cell for their survival. The protective immune response to viral infections is dependent on both the innate immune response and the development of antigen-specific lymphocytes that comprise the adaptive immune response. Virus infections usually trigger antiviral responses in host cells, but most viruses have evolved mechanisms to counter the antiviral responses of the host (29, 45). Vesicular stomatitis virus (VSV), the prototypic member of the family Rhabdoviridae, is an enveloped virus with a single-stranded nonsegmented RNA genome (11 kb) of negative polarity encoding five proteins: nucleoprotein (N), phosphoprotein (P), matrix (M), glycoprotein (G), and the large protein (L) (30). The M protein is a structural component of the virion and performs several important functions in virus assembly and budding and inhibits transcription by all three host RNA polymerases (2). Although wild-type (WT) VSV can suppress host antiviral immune responses by inhibiting the expression of interferons (IFNs) and other antiviral gene products, VSV M mutant viruses are defective in their ability to inhibit host gene expression, including IFNs (1, 60).

Natural killer T (NKT) cells are a distinct group of T lymphocytes that recognize CD1d molecules and play an important role in regulating immune responses to infectious diseases, tumors, and autoimmune diseases (13, 14, 49). CD1d molecules are expressed in virtually all mammals examined and are restricted mainly to hematopoietic cells (8, 9, 16). CD1d molecules are structurally similar to major histocompatibility complex class I molecules in their domain organization and noncovalent association with β₂-microglobulin for stable cell surface expression (12, 16). Unlike major histocompatibility complex class I and class II molecules that present peptide antigen to T cells, CD1d molecules present lipid antigens to NKT cells (11). Upon activation, NKT cells rapidly produce large amounts of both Th1 and Th2 cytokines and, as a result, activated NKT cells can confer adjuvant effects (13, 28, 61). The IFN-γ produced by NKT cells activates and induces the expansion of NK cells, neutrophils, dendritic cells (DC), and macrophages in the innate immune system and CD4⁺ Th1 or CD8⁺ T cells in the adaptive response (36). It has been reported that NKT cells have protective function against malaria, Streptococcus, and Pseudomonas infections (31, 36, 48). There are many reports on the role of NKT cells and CD1d molecules in infections with viruses and other pathogens. For example, the picornavirus encephalomyocarditis virus replicates to higher levels in CD1d- and/or NKT-cell-deficient mice (25, 26, 33). Blockade of NKG2D on nonclassical NKT cells prevented hepatitis B virus-induced acute hepatitis in a CD1d-dependent manner (63). Upon infection of human DC, low titers of herpes simplex virus (HSV) upregulated cell surface CD1b and CD1d expression, whereas high titers did the opposite, suggesting that CD1-mediated antigen presentation by DC is a target for HSV immune evasion (50, 51). This idea is supported by effects shown on CD1d recycling by HSV (65) or Kaposi's sarcoma-associated herpesvirus (56), and HSV-2 virus titers in NKT-cell-deficient mice are elevated in a genital herpes mouse model (4). NKT cells also contribute directly in acquired immune responses especially in the expansion of CD8⁺ T cells, as demonstrated in a study with respiratory syncytial virus infection (35). The expression of CD1d is essential for pathogenicity of coxsackievirus B3-induced myocarditis (32), and we (and others) have demonstrated the impaired cell surface expression of human CD1d by human immunodeficiency virus type 1 (HIV-1) Nef protein (17, 18). In nonvirus systems, NKT cells are important in various malaria models (57) and in Trypanosoma infections of mice (22-24, 47). The ability of NKT cells to rapidly secrete cytokines, activate cells of both the innate and adaptive immune responses, and recognize antigen in the context of CD1d molecules strongly suggests that NKT cells, via their interaction with CD1d, play a pivotal role in many antiviral immunity.

Mitogen-activated protein kinase (MAPK) signaling pathways serve as transducers of extracellular stimuli that allow cellular adaptation to changes in the environment. Recent reports have shown that several viruses can induce the activation of MAPK pathways in infected cells. Two classes of MAPKs, the c-Jun N-terminal kinase (JNK) and p38, function as key mediators of stress and immune signaling in mammals (21, 37). We reported earlier that VSV and vaccinia virus (VV) inhibit antigen presentation by CD1d molecules by affecting their intracellular trafficking, and this occurs in a p38-dependent manner (53). The present study was aimed at identifying the specific protein of VSV, which is responsible for the inhibition of CD1d-mediated antigen presentation to NKT cells and to understand the molecular mechanism(s) involved. Our findings strongly suggest that the VSV-M protein plays a major role in inhibiting CD1d-mediated antigen presentation to NKT cells, and this occurs in a p38-dependent manner.

MATERIALS AND METHODS

Cell lines, antibodies, viruses, and inhibitors.

LMTK-vector and LMTK-CD1d1 cells have been previously described (59) and were cultured in Dulbecco modified Eagle medium supplemented with 10% fetal bovine serum (FBS), 2 mM l-glutamine, and 500 μg of G418/ml. The mouse CD1d-specific NKT-cell hybridomas DN32.D3, N37-1A12, and N38-2C12 have all been described (15, 38, 55) and were cultured in Iscove's modified Dulbecco's medium supplemented with 5% FBS, 2 mM l-glutamine, and antibiotics. The anti-mouse CD1d monoclonal antibody (MAb) 1H6 was generated in our laboratory (55). Purified and biotinylated MAb specific for mouse interleukin-2 (IL-2) and R-phycoerythrin conjugated rat anti-mouse CD1d MAb (1B1) or isotype control (rat immunoglobulin G2b) were purchased from BD Pharmingen (San Diego, CA). Murine-specific anti-FcRγII antibody (2.4G2) was obtained from the American Type Culture Collection, Richmond, VA. Recombinant mouse IL-2 used as standards in the NKT-cell assays described below was obtained from PeproTech (Rocky Hill, NJ). Texas Red-labeled donkey anti-rat immunoglobulin antiserum and fluorescein isothiocyanate (FITC)-conjugated donkey anti-mouse immunoglobulin antiserum were purchased from Jackson Immunoresearch (West Grove, PA). Rat serum was purchased from Sigma-Aldrich (St. Louis, MO). The rat anti-mouse LAMP-1 hybridoma, 1D4B, was purchased from the American Type Culture Collection. The Indiana strain of VSV was kindly provided by J. Yewdell and J. Bennink (Laboratory of Viral Diseases, NIAID, NIH, Bethesda, MD). VSV-M and -G temperature-sensitive (ts) mutants and VSV-M- and -G-specific MAbs (39) were kindly provided by D. Lyles (Wake Forest University School of Medicine, Winston-Salem, NC). Virus stocks were propagated and titrated in LMTK cells as previously described (5, 43). The p38-specific inhibitor SB203580 was purchased from Promega (Madison, WI), whereas the MEK (ERK pathway)-specific inhibitor UO126 and antibodies specific for individual phosphorylated or total MAPK were purchased from Cell Signaling Technology, Inc., (Beverly, MA).

Generation of BMDC.

Bone marrow cells were isolated from the femurs and tibias of C57BL/6 mice and were cultured in RPMI 1640 supplemented with 2 mM l-glutamine, 50 μM 2-mercaptoethanol, 5% FBS, and antibiotics, as well as 10 ng each of murine granulocyte-macrophage colony-stimulating factor and IL-4/ml. On day 6, 1 μg of lipopolysaccharide/ml was added. On day 7, the plates were gently flushed (three to four times) to remove the loosely adherent cells, which were subsequently used in analyses as bone marrow-derived DC (BMDC).

T-cell hybridoma assays.

To measure endogenous antigen presentation by CD1d molecules before and after a virus infection, LMTK-CD1d1 cells or BMDC were mock infected or infected with WT VSV, the VSV-M ts mutants tsM301 or ts023, or the VSV-G ts mutant 045 at a multiplicity of infection (MOI) of 5 for 2 or 4 h at both permissive (31°C) and nonpermissive (39°C) temperatures. The cells were then fixed in 0.05% paraformaldehyde, washed two times in phosphate-buffered saline (PBS), and cocultured (5 × 10⁵ cells/well) with the NKT-cell hybridomas DN32.D3 (6), N37-1A12, or N38-2C12 (15) (all 5 × 10⁴ cells/well) in triplicate wells of 96-well microtiter plates. After a 24-h coculture, supernatants were harvested, and IL-2 production was measured by enzyme-linked immunosorbent assay (ELISA) as previously described (15, 55). To examine the role of MAPK in CD1d-mediated antigen presentation after infection, LMTK-CD1d1 cells were pretreated with the p38-specific inhibitor SB203580 or ERK1/2 pathway-specific inhibitor UO126 for 1 h at 37°C at either 31 or 39°C. The cells were then washed in ice-cold PBS, infected with WT VSV (MOI = 1) for 30 min or the VSV ts mutants (MOI = 5) for 2.5 h in the presence or absence of the same drug at both 31 and 39°C for comparable expression in this system, washed in ice-cold PBS, fixed in 0.05% paraformaldehyde, washed two additional times, and then cocultured with the NKT-cell hybridomas, and then IL-2 release was measured by ELISA as described above.

Immunoblotting.

LMTK-CD1d1 cells were infected with WT VSV, VSV-M tsM301, or VSV-G ts045 at an MOI of 5 at 31 or 39°C for different periods of time as indicated. The cells were washed in PBS and were resuspended in ice-cold lysis buffer (58 mM Tris [pH 8], 173 mM NaCl, 10 mM EDTA, 1.0% Nonidet P-40, 1 mM sodium orthovanadate, and 1 mM sodium fluoride [phosphatase inhibitors]) containing a protease inhibitor mix (Complete Mini; Roche, Mannheim, Germany). Cell lyates (200 μg of total protein) were dissolved in 2× sodium dodecyl sulfate (SDS) sample buffer (4% SDS, 100 mM Tris-HCl [pH 6.8], 20% glycerol, 2% [wt/vol] 2-mercaptoethanol, 0.1% bromophenol blue), resolved by SDS-10% polyacrylamide gel electrophoresis, and then transferred to a polyvinylidene difluoride membrane (Millipore, Bedford, MA). The blot was processed by using anti-p38, anti-ERK1/2, or anti-JNK1/2 antibodies specific for the phosphorylated forms and developed by using chemiluminescence before exposure on film. The blot was then stripped and reprobed with antibodies for the detection of total p38, ERK1/2, or JNK1/2. Quantitation of relative band intensity was determined by using ImageJ software (National Institutes of Health, Bethesda, MD).

Flow cytometry.

Aliquots of LMTK-CD1d1 cells mock infected or infected with the indicated viruses used for the NKT-cell coculture experiments were fixed in 1% paraformaldehyde in PBS for 10 min at room temperature, washed twice in PBS, and stained with the anti-mouse CD1d MAb (1H6), followed by a phycoerythrin (PE)-conjugated rabbit anti-mouse immunoglobulin antiserum (DakoCytomation). Mock-infected or infected BMDC were first treated with an FcRγII blocking MAb (2.4G2) for 15 min on ice and then stained with a PE-conjugated rat anti-mouse CD1d MAb. Analysis was performed by flow cytometry as previously described (55).

Confocal microscopy.

LMTK-CD1d1 cells were plated in sterile glass-bottom 35-mm dishes coated with collagen (MatTek, Ashland, MA) at a density of 5 × 10⁵ cells per dish. After overnight adherence, the cells were infected with WT or VSV ts mutants (MOI = 5) at either 31 or 39°C for 4 h. The cells were washed twice in PBS and fixed in 4% paraformaldehyde for 20 min at 4°C. Excess paraformaldehyde was quenched by using a 10 mM PBS-glycine solution for 10 min at 4°C. LAMP-1 staining was performed by incubating the cells with supernatant from the rat anti-mouse LAMP-1-secreting hybridoma, 1D4B, followed by a Texas Red-conjugated donkey anti-rat immunoglobulin antiserum. After blocking the free antibody reactive sites with rat serum (Sigma-Aldrich), immunofluorescent localization of CD1d molecules was then performed by incubating the cells with the anti-CD1d mouse MAb, 1H6 (55), followed by an FITC-conjugated donkey anti-mouse immunoglobulin antiserum. All antibodies were diluted and incubated in permeabilizing buffer (Hanks balanced salt solution-bovine serum albumin with 0.1% saponin) for 45 min at room temperature. After each step of antibody incubation, the dishes were washed three times in permeabilizing buffer. The cells were stored in the dark at 4°C in Hanks balanced salt solution-bovine serum albumin plus 0.02% azide until confocal analysis. For analysis, the cells were placed in mounting medium (10 mM Tris [pH 8.5], 2% DABCO) and viewed on a Bio-Rad MRC-1024 confocal laser-scanning microscope (Bio-Rad, Hercules, CA) equipped with a krypton-argon laser that has been modified for one-photon microscopy. The FITC and Texas Red emissions were recorded by using an oil immersion lens at ×60. Analysis of the relative level of CD1d colocalization with LAMP-1 was performed by using MetaMorph software (version 5; Molecular Devices, Sunnyvale, CA).

For staining of VSV-M and -G proteins, WT VSV-infected LMTK-CD1d1 cells were treated with VSV-M- and -G-specific MAbs, followed by a Texas Red-conjugated donkey anti-mouse immunoglobulin antiserum. The nucleus was stained using Hoechst (Molecular Probes, Eugene, OR) at 2 μg/ml in permeabilizing buffer for 5 min. Analysis was performed after the cells were resuspended in mounting medium as described above and viewed with a Zeiss LSM-510 laser scanning confocal microscope modified for one-photon microscopy. The Hoechst and Texas Red emissions were recorded by using a water immersion lens at ×63. More than 150 Z-stack sections were taken from images of mock- or WT VSV-infected LMTK-CD1d1 cells, and a movie was made to observe the three-dimensional distribution of VSV-M and -G proteins in infected cells.

Statistical analysis.

An unpaired two-tailed Student t test was performed by using GraphPad Prism software (version 5.00 for Windows; GraphPad, San Diego, CA). A P value of <0.05 was considered significant. The error bars in the bar graphs show the standard deviation (SD) from the mean.

RESULTS AND DISCUSSION

The VSV matrix protein impairs CD1d-mediated antigen presentation.

We have demonstrated that after an infection with VV or VSV, CD1d levels on splenic DC and macrophages are reduced within 1 week (40), suggesting that this quantitative change in cell surface CD1d may also contribute to the loss of NKT cells in vivo. We also observed a loss of NKT cells by activation-induced cell death in mice infected with lymphocytic choriomeningitis virus from that lasted 3 days to 3 months postinfection, although the decrease in NKT cells caused by this virus is CD1d independent (41). Recently, we reported a rapid reduction in CD1d-mediated antigen presentation to NKT cells after a VSV infection (53); this result correlated with an earlier report showing that VSV inhibits host protein synthesis and induces membrane alterations at a time when cells are actively engaged in viral protein synthesis (20). In VV, we found that the virus-encoded B1R and H5R proteins play an important role in inhibiting antigen presentation by CD1d molecules (64). The VSV-M protein is essential for viral replication and is solely responsible for inhibiting host cell transcription during infection (7) by inactivating the TFIID transcription factor (42).

In order to determine whether the M and/or G proteins of VSV were able to mediate the inhibition of antigen presentation by CD1d, LMTK-CD1d1 cells were infected with WT VSV, or VSV-M and -G ts mutants at the permissive (31°C) and nonpermissive (39°C) temperatures. At nonpermissive temperatures, ts mutants do not express a completely functional protein encoded by the mutated gene. As expected in LMTK-CD1d1 cells infected with WT VSV, antigen presentation to NKT cells was inhibited at both 31 and 39°C (Fig. 1A and B). Likewise, at 31°C, a substantial decrease in antigen presentation was caused by the M and G ts mutants. In contrast, infection with the VSV-M ts mutants at 39°C significantly rescued the ability of LMTK-CD1d1 cells to stimulate NKT cells. The rescue was not quite 100%, but this is consistent with the very low amount of functional M protein in the ts mutants (44), although the role of other viral protein(s) cannot be ruled out. In fact, we did observe a rescue in antigen presentation by CD1d using a VSV-L (polymerase) ts mutant (tsG011), although this was not detectable until 4 h postinfection (data not shown). In LMTK-CD1d1 cells infected with the VSV-G ts mutant at both 31 and 39°C, comparable levels of CD1d-mediated antigen presentation inhibition were observed (Fig. 1A and B). The cell surface expression of CD1d on LMTK-CD1d1 cells infected with the WT or any of the VSV ts mutants at either 31 or 39°C was not altered (Fig. 1C). Also, no effect of VSV-M on CD1d1 recycling could be detected (data not shown). It is important to note here that UV-inactivated VSV did not inhibit CD1d-mediated antigen presentation to NKT cells (data not shown), suggesting that viral gene expression is essential for the observed effects. Furthermore, VSV-M and -G protein expression could be detected in both VSV WT- and ts mutant-infected cells at 31 and 39°C as early as 15 min postinfection, which gradually increased over time (not shown).

FIG. 1.

The VSV-M (but not VSV-G) protein inhibits CD1d-mediated antigen presentation to NKT cells. LMTK-CD1d1 cells were mock infected or infected with WT VSV, VSV-M ts mutants (tsM301 and ts023), or a VSV-G ts mutant (ts045) at both permissive (31°C) and nonpermissive (39°C) temperatures for 2 h (A) or 4 h (B). After infection, the cells were washed, fixed, and cocultured with the indicated NKT-cell hybridomas for 24 h at 37°C. The culture supernatants were then harvested for an IL-2 ELISA. (C) Cell surface CD1d expression in mock- or virus-infected LMTK-CD1d1 cells for 4 h. Aliquots of the cells treated in panel B were stained with the anti-mouse CD1d MAb 1H6, followed by treatment with a PE-conjugated rabbit anti-mouse immunoglobulin antiserum. Cell surface CD1d expression was analyzed by flow cytometry. Open histogram, isotype control; filled histogram, anti-CD1d. (D) BMDC from C57BL/6 mice were mock infected or infected with the indicated viruses for 4 h, washed, fixed, and cocultured with the NKT cell hybridoma, DN32.D3. Aliquots of BMDC used for the NKT assay were treated with 2.4G2 before staining with a PE-conjugated anti-mouse CD1d MAb. Cells were analyzed by flow cytometry as described above. The numbers above each histogram correspond to the mean fluorescence intensity. The data shown are representative of three independent experiments.

To confirm the effect observed in LMTK-CD1d1 cells following VSV infection, we used BMDC as antigen-presenting cells. Similar to findings demonstrated with LMTK-CD1d1 cells, both WT and the VSV ts mutants could cause a substantial decrease in antigen presentation by CD1d without a quantitative change in the cell surface expression of CD1d molecules (Fig. 1D). Therefore, these results suggest that the VSV-M (but not VSV-G) protein is responsible for the reduction in CD1d-mediated antigen presentation after a VSV infection, and this was likely due to qualitative (but not quantitative) changes in cell surface CD1d molecules. It is important to note that the qualitative changes in CD1d observed after an in vitro infection in the present study are distinct from what occurs late after a VSV infection in vivo and what we previously reported (40). Both could contribute to impairments in antigen presentation by CD1d postinfection.

The VSV-M protein alters the intracellular distribution of CD1d.

Newly synthesized CD1d molecules travel to the cell surface, reenter the cell, and traffic to intracellular vesicular compartments, where they acquire endogenous ligands. These newly loaded CD1d molecules are reexpressed on the cell surface to be available for recognition by NKT cells (13, 34, 55). We and others have shown that CD1d traffics to late endocytic compartments for antigen loading before being able to stimulate NKT cells (10, 34, 55). We demonstrated earlier that VSV alters the intracellular localization of CD1d molecules after infection (53). Inhibition of p38 MAPK activity by a specific inhibitor enhanced CD1d colocalization with the late endosomal/lysosomal (LAMP-1⁺) compartment, with a concomitant increase in antigen presentation by CD1d molecules, whereas the opposite was observed with an ERK1/2 pathway inhibitor (53). In order to understand the molecular mechanism(s) used by VSV-M in the reduction of CD1d-mediated antigen presentation to NKT cells, LMTK-CD1d1 cells were mock infected or infected with WT VSV or the VSV-M and -G ts mutants for 4 h at 31 and 39°C. The intracellular distribution of CD1d molecules compared to LAMP-1 was then analyzed by confocal microscopy. In contrast to mock-infected LMTK-CD1d1 cells, an infection with WT VSV displayed a significant reduction in the colocalization of CD1d and LAMP-1, and these molecules were segregated to one side of the cell at either temperature (Fig. 2A). Similarly, infection with either of the VSV-M ts mutants (tsM301 and ts023) at the permissive temperature caused a comparable change in the intracellular distribution of CD1d molecules. In contrast, at 39°C, an infection with the VSV-M mutants resulted in significantly more CD1d colocalization with LAMP-1 (Fig. 2B). Infection with the VSV-G ts mutant was comparable to that observed with the WT virus at either temperature (Fig. 2A and B). These results suggest that the VSV-M protein plays an important role in inhibiting CD1d-mediated antigen presentation to NKT cells by altering the amount of CD1d molecules in the appropriate intracellular compartment.

FIG. 2.

VSV-M alters the intracellular distribution and/or trafficking of CD1d. (A) Confocal analysis of mock- or VSV-infected LMTK-CD1d1 cells. LMTK-CD1d1 cells were mock infected or infected with the indicated viruses for 4 h. The cells were fixed, permeabilized, and stained for LAMP-1 (red) and CD1d1 (green). Analysis was performed by confocal microscopy. (B) CD1d colocalization with LAMP-1. The percent colocalization of CD1d molecules with LAMP1 in mock- and VSV-infected LMTK-CD1d1 cells was analyzed by using Metamorph software. CD1d colocalization with LAMP-1 in uninfected cells was set at 100%, and the percent colocalization in infected LMTK-CD1d1 cells was calculated in relation to mock-infected control cells. Each bar is the mean of six random fields ± the SD. The data shown are representative of two independent experiments. **, P < 0.001. (C) Confocal microscopic analyses for the VSV-M and -G in LMTK-CD1d1 cells. LMTK-CD1d1 cells were infected with WT VSV (MOI = 5) for 2 h, washed, fixed, and permeabilized. Cells were treated with VSV-M and -G specific MAb, followed by a Texas Red-conjugated goat anti-mouse immunoglobulin antiserum. The nucleus was identified by Hoechst staining. Single optical sections of VSV-M- or VSV-G-infected cells with nuclear staining and overlay images are shown. The data shown are representative of three independent experiments.

After a 2-h infection with the WT virus, the VSV-M protein was distributed diffusely throughout the cell (Fig. 2C). VSV-M exists as a soluble molecule throughout the cytoplasm and is also found in lysosomes and endosomes (54). Since the M protein inhibits host gene translation very early postinfection due to alterations in eIF-4F (19), not surprisingly, the M protein is also found in the nucleus of infected cells (Fig. 2C), where it binds to mRNA export factor Rae1/mrnp41 (27). Furthermore, the M protein interacts (and forms a complex) with tubulin in infected cells (46), which could account for the intracellular segregation of CD1d observed in confocal microscopy analyses (Fig. 2). We did not detect any interaction between VSV-L, -M, and -G proteins with the mouse CD1d cytoplasmic tail in a yeast two-hybrid system (data not shown), suggesting that other (e.g., indirect) molecular mechanisms may allow the VSV-M protein to mediate the qualitative changes in CD1d molecules observed postinfection.

The VSV-M protein impairs CD1d-mediated antigen presentation via activation of the p38 MAPK.

BHK21 cells infected at 39°C with ts mutants of VSV in the NS (tsG22), N (tsG41), M (ts023), and G (ts045) proteins showed similar changes in the cytoskeleton compared to those after infection with WT VSV (58). Due to the lack of a quantitative change in the cell surface CD1d molecules after a VSV infection, it was unclear how VSV-M (but not VSV-G) inhibited antigen presentation to NKT cells. Viruses reorganize or utilize various cellular functions; it seems likely that viruses would take advantage of preexisting signaling pathways to induce cellular and/or viral gene expression to promote virus replication (52). The p38, ERK1/2, and JNK MAPK pathways are modulated by stress, such as that caused by viruses (21, 37). In fact, a number of viruses have been known to induce p38 MAPK activity (3), including VSV (53). The replication of varicella-zoster virus is positively regulated by activated p38, and the activation of this stress pathway activates cellular transcription factors but prevents the activation of cellular defense mechanisms (52). However, these studies restricted their focus to the effector cells (CD4⁺ and CD8⁺ T cells) rather than on the antigen-presenting cell itself. We reported the reciprocal regulation of CD1d-mediated antigen presentation to NKT cells by the p38 and ERK1/2 MAPK after infection (53) with VV or WT VSV. Thus, it made sense to determine whether there were differences in the activation of specific MAPK by the VSV ts mutant viruses. Thus, LMTK-CD1d1 cells were treated with or without a p38 inhibitor (SB203580) or ERK1/2 pathway inhibitor (UO126) at 31 and 39°C, washed, fixed in 0.05% paraformaldehyde, and cocultured with NKT cells. As shown in Fig. 3, treatment of LMTK-CD1d1 cells with the p38 inhibitor resulted in a significant increase in antigen presentation to NKT cells. In contrast, inhibition of the ERK1/2 pathway caused a significant decrease in antigen presentation; this response was observed at both 31 and 39°C and is in line with what we reported at the physiologic temperature (53). In order to assess whether the inhibition of antigen presentation by VSV-M was related to MAPK activation, LMTK-CD1d1 cells were pretreated with either the p38 or ERK1/2 pathway inhibitor and then incubated with VSV-M (tsM301) and -G (ts045) mutant viruses in the presence or absence of the respective MAPK inhibitors, and the cells were then fixed and used as targets with NKT cells. It should be noted that the inhibitors had no effect on VSV replication under the conditions used (data not shown). The results suggested that the p38 inhibitor significantly rescued the CD1d-mediated antigen presentation at both 31 and 39°C, whereas the ERK1/2 pathway inhibitor did not (Fig. 3).

FIG. 3.

The VSV-M protein regulates CD1d-mediated antigen presentation to NKT cells through the p38 MAPK. LMTK-CD1d1 cells were pretreated with a 20 μM concentration of a p38 or MEK (ERK1/2 pathway) inhibitor or with vehicle (dimethyl sulfoxide) for 1 h at both 31 and 39°C. The cells were then washed and infected with WT VSV or the M and G ts mutant viruses at both 31 and 39°C in the presence or absence of the same drug. The cells were washed, fixed, and used as targets in NKT cell cocultures as described above. The data shown are the percent control values, and each bar is the mean of triplicate cultures ± the SD. *, P < 0.05; **, P < 0.001; ***, P < 0.0001. V, vehicle (dimethyl sulfoxide); p38i, p38 inhibitor (SB203580); ERKi, MEK (ERK1/2 pathway) inhibitor (UO126). The results are representative of three independent experiments.

To investigate activated p38 and ERK1/2 MAPK levels after infection, LMTK-CD1d1 cells were infected with WT VSV or the VSV-M and -G ts mutant viruses for various lengths of time at 31 and 39°C. The cells were then lysed, blotted, and probed for activated and total p38, ERK1/2, and JNK1/2 by Western blot analysis. The relative density of activated versus total p38, ERK1/2, and JNK1/2 levels with both VSV ts mutants and the WT virus resulted in the activation of p38 at 31°C by all of the viruses (Fig. 4A). Interestingly, the VSV-M mutant virus failed to induce the significant phosphorylation of p38 at 39°C until 1 h postinfection, whereas the VSV-G ts mutant induced levels of p38 activation comparable to the WT virus at both temperatures (Fig. 4A). When ERK1/2 levels were analyzed, none of the viruses activated ERK1/2 activity; rather, we observed a downregulation of ERK1/2 (Fig. 4B). Activation of JNK1/2 by WT or mutant VSV was at best low and mostly reflected that which was observed with ERK1/2 (Fig. 4C). We have previously reported a reduction in phosphorylated ERK1/2 in WT VSV-infected LMTK-CD1d1 cells at 37°C (53). In the present study, the data suggest that the reciprocal regulation of CD1d-mediated antigen presentation by p38 and ERK1/2 is operative even at nonpermissive temperatures postinfection. Thus, the VSV-M protein may be responsible for inhibiting CD1d-mediated antigen presentation by specifically activating p38.

FIG. 4.

The VSV matrix protein rapidly activates p38. LMTK-CD1d1 cells were infected with WT VSV or the VSV-M (tsM301) or -G (ts045) ts mutants at either 31 or 39°C for the indicated periods of time. The cells were then lysed, and equal amounts of protein were used for the detection of phosphorylated and total p38 (A), ERK1/2 (B), and JNK1/2 (C) by Western blot analysis. The relative amount of phosphorylated versus total p38, ERK1/2, or JNK1/2 was quantitated by densitometry. Each bar corresponds to the mean density from two independent experiments ± the SD. The results are representative of two to four independent experiments.

In summary, the VSV-M protein inhibited CD1d-mediated antigen presentation to NKT cells by altering intracellular CD1d trafficking in a p38 MAPK-dependent manner. Many viruses such as rabies virus, paramyxoviruses, etc., express a matrix protein with comparable function, as with VSV, Ebola virus, influenza virus, Sendai virus, etc. (42, 62). Thus, understanding the function of the M protein in cells infected with any of these viruses may lead to the development of novel antiviral therapeutic options, including new targeted vaccine strategies.