ERK Activation Is Required for Double-stranded RNA- and Virus-induced Interleukin-1 Expression by Macrophages*

Double-stranded (ds) RNA, which accumulates during viral replication, activates the antiviral response of infected cells. In this study, we have identified a require-ment for extracellular signal-regulated kinase (ERK) in the regulation of interleukin 1 (IL-1) expression by macrophages in response to dsRNA and viral infection. a novel PKR-independent role for in the regulation of the

Double-stranded (ds) 1 RNA, which accumulates at various stages of viral replication, plays a primary role in the activation of antiviral responses in virally infected cells (1). Prominent sources of dsRNA include viral RNAs containing ds secondary structure and dsRNA formed during viral replication (1). The accumulation of these dsRNAs activates several antiviral responses including the dsRNA-dependent protein kinase, PKR (2,3). PKR has been implicated in both the transcriptional activation of antiviral genes (4 -7), and in the inhibition of protein translation mediated by the phosphorylation of the initiation factor eIF2␣ (8). In macrophages, dsRNA and viral infection stimulate the expression of proinflammatory cytokines such as IL-1␣, IL-1␤, tumor necrosis factor, and IL-6 as well as the inducible isoform of nitric-oxide synthase (4, 9 -12). Nitric oxide appears to play a primary role in host defense against a viral infection as the inhibitory actions of IFN-␥activated macrophages on viral replication are mediated by nitric oxide (13)(14)(15)(16). Also, iNOS-deficient mice have higher mortality rates and reduced viral clearance as compared with wild-type mice following virus infection (17)(18)(19). These findings support a role for nitric oxide, produced following iNOS expression, in the antiviral response.
The molecular mechanisms by which viral infection or dsRNA stimulates iNOS and IL-1 expression by macrophages are not well-defined. We have shown that dsRNA-induced iNOS and IL-1 expression by RAW 264.7 cells is prevented by the expression of dominant negative (dn) PKR, and that this PKR dependence can be overcome by IFN-␥ (4). In addition, dsRNA ϩ IFN-␥ stimulate iNOS expression to similar levels in peritoneal macrophages (PEC) isolated from PKR Ϫ/Ϫ and PKR ϩ/ϩ mice (4). These findings suggest that PKR may participate in, but is not essential for, dsRNA-induced iNOS expression. In contrast, nuclear factor B (NF-B) activation appears to be required for iNOS and IL-1 expression by macrophages. Inhibition of NF-B activation prevents iNOS and IL-1 expression stimulated by dsRNA in RAW 264.7 cells and dsRNA ϩ IFN-␥ in primary macrophages (9). NF-B is comprised of heterodimer or homodimers of the p50/NF-B1 and p65/RelA subunits and is sequestered in the cytoplasm of unstimulated cells in complex with inhibitor protein B (IB, Ref. 20). Pathways that stimulate the activation of NF-B are characterized by IB phosphorylation, polyubiquitination, and degradation by the 26 S proteasome complex (21). This allows for the release of NF-B and its translocation to the nucleus. PKR has been implicated in the activation of NF-B, either by directly stimulating IB phosphorylation or by activating IB kinase (IKK) through physical interaction (22)(23)(24). Recent studies have identified a PKR-independent pathway that results in the activation of NF-B by dsRNA. Magun and co-workers (25) have shown that dsRNA and encephalomyocarditis virus (EMCV) infection stimulates NF-B activation to similar levels in mouse embryonic fibroblasts (MEF) isolated from PKR Ϫ/Ϫ and PKR ϩ/ϩ mice. In addition, dsRNA stimulates IB degradation and NF-B nuclear localization to similar levels in macrophages and islets of Langerhans isolated from PKR Ϫ/Ϫ and PKR ϩ/ϩ mice (4,26).
Mitogen-activated protein kinase (MAPK) signaling pathways are activated by dsRNA (24,27,28) and viral infection (25), and MAPK have been implicated in the regulation of iNOS and IL-1 expression (29 -31). The role of PKR in dsRNA-in-duced MAPK kinase activation is less clear. It was first shown that dsRNA stimulates c-Jun N-terminal kinase (JNK) and p38 activation in MEF. In these studies, dsRNA-induced p38 activation was attenuated in PKR Ϫ/Ϫ MEF while JNK appeared to be activated by a PKR-independent mechanism (24). More recently, dsRNA and viral infection have been shown to stimulate p38 and JNK activation in MEF isolated from PKR Ϫ/Ϫ mice. While JNK activation appears to require a PKR-dependent inhibition of protein synthesis (25), p38 activation by dsRNA is not dependent on PKR or PKR-mediated inhibition of protein synthesis (25). While these findings indicate that MAPK are activated by dsRNA, the role of PKR and the mechanisms by which PKR may mediate MAPK activation have yet to be fully elucidated.
In this study, the role of the MAPK pathway, specifically ERK, in dsRNA-induced macrophage activation has been examined. We show that the MAP/ERK kinase (MEK) selective inhibitors U0126 and PD98059 prevent dsRNA-and virusinduced ERK phosphorylation, IL-1 expression, IL-1␤ reporter activity, and IL-1 release by macrophages. The role of ERK in the regulation of inflammatory gene expression in response to virus infection appears to be selective for the IL-1 pathway, as ERK inhibition does not affect the ability of dsRNA to stimulate iNOS expression or nitrite production by macrophages. Using peritoneal macrophages isolated from PKR ϩ/ϩ and PKR Ϫ/Ϫ mice we show that the genetic absence of PKR does not modulate the ability of dsRNA to induce IL-1 expression and release or to stimulate ERK phosphorylation. These findings support a role for ERK activation in the PKR-independent regulation of the antiviral response of IL-1 expression and release by macrophages.

EXPERIMENTAL PROCEDURES
Materials and Animals-Defined fetal bovine serum was purchased from Hyclone (Logan, UT). Poly IC, myelin basic protein, and protein A-Sepharose were purchased from Sigma Chemical Co. Poly IC was prepared as previously described (4). U0126 was from CalBiochem (San Diego, CA), and PD98059 was purchased from Alexis Biochemicals (San Diego, CA). The human IL-1␤ promoter luciferase reporter plasmid (XL-Luc) has been previously described (32) and was provided by Dr. Matthew Fenton (Boston University, Boston, MA). Rabbit anti-phospho-ERK was purchased from Promega (Madison, WI). Rabbit anti-ERK1 was a gift from Dr. John C. Lawrence, Jr. (University of Virginia, Charlottesville, VA). Rabbit anti-mouse macrophage iNOS was a gift from Dr. Tom Misko (Amersham Biosciences). 3ZD monoclonal mouse anti-IL-1␤ was from Biological Resources Branch, DCTD at the NCI. PKR Ϫ/Ϫ mice (C57BL/6(J)ϫSV129 background) were the generous gift of Dr. Randal Kaufman (University of Michigan, Ann Arbor, MI) and have been previously described (33)(34)(35). C57BL/6(J)ϫSV129 (PKR ϩ/ϩ ) mice were obtained from Jackson Laboratories (Bar Harbor, ME). iNOS and cyclophilin cDNA were the generous gift of Dr. Charles Rodi (Amersham Biosciences) and Dr. Steve Carroll (University of Alabama-Birmingham, Birmingham, AL), respectively. IL-1␣ and IL-1␤ cDNAs were gifts from Dr. Cliff Bellone (Saint Louis University, St. Louis, MO) and have been previously described (36). The wild-type ERK1 plasmid was the gift of Dr. Joseph Baldassare (Saint Louis University). The B-strain of encephalomyocarditis virus (EMCV) was the generous gift of Dr. J. W. Yoon (University of Calgary, Calgary, Alberta, Canada). All other reagents were obtained from commercially available sources.
Cell Culture, Peritoneal Macrophage Isolation, and EMCV Infection-RAW 264.7 and RINm5F cells were removed from growth flasks by treatment with 0.05% trypsin, 0.02% EDTA for 5 min at 37°C. Cells were washed twice with media and plated at the indicated concentrations. The cells were allowed to adhere for 2 h under an atmosphere of 5% CO 2 , 95% air before the initiation of experiments. Peritoneal exudate cells (PEC) were isolated from PKR Ϫ/Ϫ and PKR ϩ/ϩ mice by lavage as previously described (37). After isolation, 4 ϫ 10 5 cells/condition were incubated in 400 l of complete CMRL-1066 (CMRL-1066 containing 10% heat-inactivated fetal bovine serum, 1ϫ L-glutamine, 100 units/ml penicillin, and 100 g/ml streptomycin) for 2 h under an atmosphere of 5% CO 2 , 95% air. The cells were washed three times with 400 l of complete CMRL-1066 to remove non-adherent cells before initiation of experiments. RAW 264.7 cells (4 ϫ 10 5 /400 l of DME) were infected with the indicated amounts of EMCV virus in 100 l of DME for 30 min, 300 l of DME was added, and the cells were incubated for 24 additional hours. EMCV was propagated as previously described (15).
Nitrite and IL-1 Determinations-Nitrite production was determined by the Greiss assay, and IL-1 release was measured using the RINm5F cell bioassay as previously described (38,39). The RINm5F cell IL-1 bioassay assay measures cumulative biological activity of both IL-1␣ and IL-1␤.
Northern Blot and RT-PCR Analysis-Total RNA was isolated from RAW 264.7 cells using the RNeasy kit (Qiagen, Santa Clarita, CA) according to the manufacturer's instructions. Northern blot analysis and RT-PCR were performed using iNOS, IL-1␣, IL-1␤, cyclophilin, and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) probes as previously described (9,40). Cyclophilin was used as an internal RNA loading control for Northern blots, and GAPDH used as a control for the RT reaction.
Transient Transfections and IL-1␤ Reporter Assays-RAW 264.7 cells were transiently transfected with 1 g of the human IL-1␤ luciferase reporter (positions Ϫ3757 to ϩ11; XL-LUC) and 1 g of the pCMV-SPORT-␤-galactosidase (Invitrogen) control plasmid using the Qiagen Superfect reagent according to the manufacturer's instructions. After an 8-h incubation at 37°C experiments were initiated by addition

FIG. 1. Effects of MEK inhibition on dsRNA-induced iNOS expression and nitrite production by RAW 264.7 macrophages.
RAW 264.7 cells (4 ϫ 10 5 /400 l of DME) were pretreated for 30 min with the indicated concentrations of U0126, poly(IC) (50 g/ml) was added, and the cells were cultured for 24 additional hours. Nitrite production was determined on culture supernatants (A) and the expression of iNOS at the protein level was determined by Western blot analysis of the isolated cells (B). C, RAW 264.7 cells (5 ϫ 10 6 /2 ml of DME) were pretreated with U0126 or Me 2 SO (vehicle control, DMSO) for 30 min at 37°C, poly(IC) (50 g/ml) was added, and the cells were incubated for 6 additional hours. Total RNA was isolated and used to determine iNOS mRNA accumulation by Northern blot analysis. Cyclophilin mRNA accumulation is shown as a control for RNA loading. Results for nitrite production are the average Ϯ S.E. of three independent experiments, and iNOS protein expression and mRNA accumulation are representative of three independent experiments. of poly(IC) or poly(IC) and U0126 as indicated in the figure legends. Twenty-four hours after stimulation, the cells were harvested and lysed in Reporter Lysis Buffer, and luciferase activity was determined using the luciferase assay system according to the manufacturer's specifications (Promega). ␤-galactosidase assays were performed as previously described (41). Luciferase activities are reported relative to ␤-galactosidase activity to control for transfection efficiencies.
Immunocomplex Kinase Assay-The K71R mutation in human ERK1 was made using the Stratagene (La Jolla, CA) QuickChange PCRplasmid mutagenesis kit using 5Ј-TGGCCATCAGAAAGATCAGC-3Ј and 5Ј-GCTGATCTTTCTGATGGCCA-3Ј (mutated bases in bold) prim-ers according to the manufacturer's specifications. This mutation, which was confirmed by sequence analysis, has previously been shown to inhibit ERK kinase activity (42). RAW 264.7 cells were transiently transfected with 2 g of wild-type ERK1 or ERK1(K71R) using Superfect (Qiagen) according to the manufacturer's instructions. After a 48-h incubation at 37°C, poly(IC) was added, and the cells were cultured for 30 additional min. The cells were then harvested, ERK was immunoprecipitated, and the ability of immunoprecipitated ERK to phosphorylate myelin basic protein was examined as previously described (42). ERK kinase activity was quantitated by densitometry.
Densitometry and Image Analysis-Autoradiograms were scanned into NIH Image version 1.59 using a COHU high performance CCD camera (Brookfield, WI) and densities determined using NIH Image version 1.59 software.

Effects of MEK Inhibition on dsRNA-induced iNOS Expression, Nitrite Production, and IL-1 Expression and Release by RAW 264.7
Macrophages-Macrophage activation in response to dsRNA includes the expression of iNOS and production of nitric oxide (9,14). Also, dsRNA and viral infection have been shown to stimulate ERK activation in macrophages (15,24,27), and LPS-induced iNOS expression is sensitive to ERK inhibition (43). Therefore, the potential role of ERK in dsRNA-induced macrophage activation was examined using selective inhibitors (U0126 and PD98059) of MEK, the upstream kinase FIG. 2. MEK inhibition prevents dsRNA-induced IL-1 expression and release by RAW 264.7 macrophages. RAW 264.7 cells (4 ϫ 10 5 /400 l of DME) were pretreated with the indicated concentrations of U0126 for 30 min, poly(IC) (50 g/ml) was added, and the cells were cultured for 24 additional hours. IL-1 released into the culture supernatant was determined using the RINm5F cell bioassay (A), and pro-IL-1␤ protein expression was determined by Western blot analysis of the isolated cells (B). RAW 264.7 cells (5 ϫ 10 6 /2 ml DME) were pretreated with U0126 or Me 2 SO (vehicle control, DMSO) for 30 min, poly(IC) was added, and the cells were incubated for 6 additional hours at 37°C. Total RNA was isolated and used to determine IL-1␣ and IL-1␤ mRNA accumulation by Northern blot analysis (C). Cyclophilin mRNA accumulation is shown as a control for RNA loading. RAW 264.7 cells (4 ϫ 10 6 /2 ml DME), transiently transfected with the human IL-1␤ reporter construct XL-Luc, were pretreated for 30 min with U0126, poly(IC) was added, and the cells were cultured for 24 additional hours. Total cellular protein was isolated and luciferase activity determined as described under "Experimental Procedures." The results for IL-1 release and IL-1␤ promoter activity are the average Ϯ S.E. of three independent experiments, pro-IL-1␤ protein expression and IL-1 mRNA accumulation are representative of three independent experiments.
FIG. 3. Effects of PD98059 on dsRNA-induced IL-1 expression by RAW 264.7 macrophages. RAW 264.7 cells (5 ϫ 10 6 /2 ml DME) were pretreated with PD98059 for 30 min, poly(IC) was added, and the cells were incubated for 6 additional hours at 37°C. Total RNA was isolated and used to determine IL-1␣ and IL-1␤ mRNA accumulation by Northern blot analysis (A). Cyclophilin mRNA accumulation is shown as a control for RNA loading. RAW 264.7 cells (4 ϫ 10 5 /400 l of DME) were pretreated with the indicated concentrations of PD98059 for 30 min, poly(IC) (50 g/ml) was added, and the cells were cultured for 24 additional hours. The cells were then isolated and pro-IL-1␤ protein expression was determined by Western blot analysis (B). The concentration-dependent effects of PD98059 on dsRNA-induced p38 and JNK phosphorylation following a 30-min incubation were examined by Western blot analysis using phosphospecific (p38-P, JNK-P) antisera (C). Total JNK is shown as a loading control. The results IL-1 mRNA accumulation, pro-IL-1␤ protein expression, and p38 and JNK phosphorylation are representative of three independent experiments. that is responsible for the activation of ERK. Treatment of RAW 264.7 macrophages with 50 g/ml poly(IC) results in the accumulation of iNOS mRNA following a 6-h incubation, the expression of iNOS at the protein level and a 5-fold increase in nitrite production following a 24-h incubation (Fig. 1, A-C). The ERK selective inhibitor U0126, at concentrations of 0.01-5 M, does not attenuate iNOS mRNA accumulation, iNOS protein expression, or nitrite production by RAW 264.7 cells. Similar results for nitrite production and iNOS expression were obtained with the second MEK inhibitor, PD98059 at concentrations of 0.1-10 M (data not shown). MEK inhibition slightly increases the levels of nitrite produced by RAW 264.7 cells treated with poly(IC); however, this minor increase did not achieve statistical significance.
A second antiviral response activated by dsRNA in macrophages is the expression and release of the proinflammatory cytokine, IL-1 (4,9). A 24-h incubation of RAW 264.7 cells with poly(IC) (50 g/ml) results in an ϳ50-fold increase in IL-1 release and the accumulation of pro-IL-1␤ protein (Fig. 2, A  and B, respectively). Consistent with IL-1 release, treatment of RAW 264.7 cells with poly(IC) stimulates the accumulation of both IL-1␣ and IL-1␤ mRNA as determined by Northern blot analysis following a 6-h incubation, and IL-1␤ luciferase reporter activity following a 24-h incubation (Fig. 2, C and D,  respectively). The MEK selective inhibitor U0126 attenuates poly(IC)-induced IL-1␣ and IL-1␤ mRNA accumulation, IL-1␤ promoter activity, pro-IL-1␤ protein expression, and IL-1 release by RAW 264.7 cells, with maximal inhibition at 5 M.
To confirm these findings the effects of a second MEK selective inhibitor, PD98059 on IL-1␣ and IL-1␤ mRNA, and protein expression in response to poly(IC) treatment was examined. PD98059 at 10 M inhibits poly(IC)-induced IL-1␣ and IL-1␤ mRNA accumulation and pro-IL-1␤ protein expression (Fig. 3,  A and B). To confirm the selectivity of the ERK inhibitors, the effects of PD98509 and U0126 on dsRNA-induced p38 and JNK phosphorylation were examined. At concentrations that prevent dsRNA-induced IL-1␣ and IL-1␤ expression, PD98059 (10

FIG. 5. dsRNA stimulates ERK activation and dnERK inhibits dsRNA-induced IL-1 expression by macrophages.
A, RAW 264.7 cells (4 ϫ 10 5 /400 l DME) were pretreated with 5 M U0126 for 30 min, poly(IC) (50 g/ml) was added, and the cells were cultured for an additional 30 min. The cells were isolated and phosphorylated ERK (p44-P and p42-P) and p44 ERK (loading control) were determined by Western blot analysis. RAW 264.7 cells (4 ϫ 10 5 /2 ml of DME), transiently transfected with ERK or dnERK, were treated with 50 g/ml poly(IC) for 30 min, and ERK activity was determined by immunocomplex kinase assay (B). The expression levels of ERK and dnERK were determined by Western blot analysis and are shown in the inset (B). RAW 264.7 cells (4 ϫ 10 5 /2 ml DME), transiently transfected with ERK or dnERK, were treated with 50 g/ml poly(IC) (C) or EMCV (0.25 MOI, D) for 6 h, total RNA was isolated and used to examine IL-1␣ and IL-1␤ mRNA accumulation by RT-PCR. GAPDH mRNA accumulation was used as an internal control for the RT and PCR reactions. The results for ERK phosphorylation and IL-1 mRNA accumulation are representative of three independent experiments. The results for ERK kinase activity are the average of two independent experiments.  Fig. 3C) and U0126 (5 M, data not shown) do not attenuate poly(IC)-stimulated p38 or JNK phosphorylation. These findings suggest that dsRNA selectively regulates iNOS and IL-1 expression by macrophages, in that ERK appears to participate in the transcriptional activation of IL-1 but does not regulate iNOS expression in response to dsRNA.

Effects of MEK Inhibition on EMCV-induced pro-IL-1 Expression by RAW 264.7 Macrophages-To confirm that poly(IC)
recapitulates host responses to a viral infection; the effects of encephalomyocarditis virus (EMCV) infection on the expression of IL-1␤ by RAW 264.7 cells were evaluated. As shown in Fig. 4A, infection of RAW 264.7 cells with EMCV stimulates the accumulation of IL-1␤ following a 24-h incubation. Importantly, EMCV-induced IL-1␤ expression is attenuated by U0126 (5 M, Fig. 4B) at a concentration that prevents poly(IC)-induced IL-1 expression by RAW 264.7 cells (Fig. 2). These findings indicate that EMCV and poly(IC) elicit similar antiviral responses in RAW 264.7 cells and provide additional evidence to support a role for ERK in the regulation of IL-1␤ expression by macrophages in response to a viral infection.
dsRNA Stimulates ERK Activation and dnERK Inhibits dsRNA-induced IL-1 Expression by Macrophages-To confirm that U0126 prevents ERK activation, the effects of this MEK kinase inhibitor on ERK phosphorylation were examined by Western blot analysis. Treatment of RAW 264.7 cells for 30 min with poly(IC) (50 g/ml) results in the phosphorylation of ERK, an effect that is prevented by U0126 at concentrations (5 M) that inhibit dsRNA-induced IL-1 expression and release (Fig.  5A). In a similar manner PD98059 (10 M) prevents poly(IC)stimulated ERK phosphorylation (data not shown). Consistent with the stimulatory actions of poly(IC) on ERK phosphorylation, dsRNA activates ERK as determined by immunocomplex kinase assays. For these experiments, RAW 264.7 cells, tran-siently transfected with either wild type or dnERK (ERK1(K71R)), were incubated with or without poly(IC) for 30 min, and then the phosphorylation of myelin basic protein by immunoprecipitated ERK was examined. A 30-min incubation with poly(IC) stimulates an ϳ4.5-fold increase in ERK activity (myelin basic protein phosphorylation) in RAW 264.7 cells expressing wild-type ERK; however, poly(IC) fails to stimulate myelin basic protein phosphorylation in RAW 264.7 cells expressing dnERK (Fig. 5B). In this immunocomplex kinase assay wild type and dnERK are expressed at levels 2-3-fold higher than the levels of endogenous ERK (Fig. 5B, inset). RT-PCR was used to examine the role of ERK in the regulation of IL-1␣ and IL-1␤ mRNA accumulation by macrophages in response to dsRNA and EMCV infection. A 6-h incubation with poly(IC) results in the accumulation of both IL-1␣ and IL-1␤ mRNA in RAW 264.7 cells transfected with wild-type ERK (Fig. 5C). In contrast, IL-1␣ and IL-1␤ mRNA accumulation is attenuated in RAW 264.7 cells transfected with dnERK. In a similar fashion, EMCV-induced IL-1␣ and IL-1␤ mRNA accumulation is attenuated in RAW 264.7 cells transfected with dnERK (Fig. 5D). The transfection efficiency in these experiments was greater then 70% (as determined by ␤-galactosidase staining, and the levels of wild type and dnERK expression were comparable to those shown in Fig. 5B, inset). These findings indicated that dsRNA stimulates ERK activation and provide further evidence to support a role for ERK in the regulation of dsRNA-and EMCV-induced IL-1 expression by macrophages. IFN-␥-induced nitrite production (Fig. 6A) or iNOS expression (Fig. 6B) by PEC isolated from PKR Ϫ/Ϫ and PKR ϩ/ϩ mice. In contrast to RAW 264.7 cells, naive mouse peritoneal macrophages require two proinflammatory signals for iNOS expression and nitrite production (44), and we have previously shown that neither poly(IC) nor IFN-␥ alone stimulate iNOS expression or nitric oxide production by PEC (4,9). Also, we have shown that PKR is not required for poly(IC) ϩ IFN-␥-induced iNOS expression and nitric oxide production by mouse PEC (4).

MEK Inhibition Does Not Affect dsRNA-induced iNOS Expression and Nitrite Production but Prevents IL-1 Expression and Release by Peritoneal
While ERK does not appear to participate in the regulation of iNOS expression, it is required for poly(IC) and poly(IC) ϩ IFN-␥-induced IL-1 release by mouse macrophages. Treatment of PEC isolated from PKR Ϫ/Ϫ and PKR ϩ/ϩ mice for 24 h with either poly(IC) or poly(IC) ϩ IFN-␥ results in the release of IL-1 to similar levels, and this IL-1 release is prevented by U0126 (Fig. 7A). U0126 also prevents poly(IC) (data not shown) and poly(IC) ϩ IFN-␥-induced IL-1␤ protein accumulation in PEC isolated from PKR Ϫ/Ϫ and PKR ϩ/ϩ mice (Fig. 7B). These findings, which are consistent with the inhibitory actions of U0126 on IL-1 expression and release by RAW 264.7 cells, support a PKR-independent role for ERK in dsRNA-induced IL-1 expression and release by primary mouse macrophages.
To confirm that dsRNA stimulates ERK activation, and to determine if PKR is required for this activation, the effects of poly(IC) on ERK phosphorylation by PEC isolated from PKR Ϫ/Ϫ and PKR ϩ/ϩ mice were examined. As shown in Fig. 7C, poly(IC) stimulates the time-dependent phosphorylation of ERK that is first apparent following a 15-min incubation and that persists for up to 60 min. PKR does not appear to participate in ERK activation, as poly(IC) stimulates ERK phosphorylation to similar levels in PEC isolated form PKR Ϫ/Ϫ and PKR ϩ/ϩ mice. In addition, the stimulatory actions of a 30-min incubation with poly(IC) on ERK phosphorylation are prevented by U0126 (5 M, Fig. 7D). These findings indicate that the presence of PKR is not required for the activation of ERK by dsRNA in PEC. DISCUSSION Recently, Magun and co-workers (25) have proposed that dsRNA triggers two separate antiviral programs, cell suicide (apoptosis), and a survival pathway associated with proinflammatory cytokine production. At the core of this proposal is the ability of PKR to regulate each of these antiviral responses. The first antiviral pathway is that of apoptosis, a process of selfelimination that removes virally infected cells. This pathway of programmed cell death is activated by dsRNA, is dependent on PKR (45,46), and appears to be a widely used antiviral response as evidenced by the multiple strategies used by viruses to evade apoptosis (Ref. 8 and references therein). The second antiviral pathway activated by dsRNA or viral infection is the expression and release of proinflammatory cytokines such as interferons and interleukins. The role of PKR in the regulation of cytokine expression has not been clearly defined, in part because of results that support or refute a role for PKR in the regulation of NF-B and MAPK activation in response to dsRNA. PKR was originally believed to be required for dsRNAinduced NF-B activation and p38 phosphorylation based on studies in MEF isolated from PKR Ϫ/Ϫ mice (24,35). More recently, dsRNA and EMCV infection have been shown to stimulate NF-B and p38 activation in MEF isolated from PKR Ϫ/Ϫ mice to levels similar to MEF isolated from wild-type mice (25), suggesting that PKR-independent pathways also participate in the antiviral response.
In this study, we present evidence to further support the presence of a PKR-independent antiviral response pathway that is activated by viral infection and dsRNA and that regulates the expression and release of the proinflammatory cytokine IL-1. We show that dsRNA treatment or EMCV infection stimulates IL-1 expression and release in an ERK-dependent fashion. Selective inhibition of MEK, the upstream ERK kinase, prevents dsRNA and EMCV-induced IL-1 expression and release. The role of ERK in regulating the antiviral response appears to be selective for IL-1 expression as MEK inhibition does not modulate dsRNA-induced iNOS expression or nitric oxide production by macrophages. PKR does not appear to be required for this response, as dsRNA stimulates ERK activation and ERK-dependent IL-1 expression by macrophages isolated from PKR Ϫ/Ϫ mice. These findings suggest that PKR is not required for the activation of ERK or the expression and release of IL-1 by macrophages in response to dsRNA or viral infection. However, this conclusion must be tempered by the possibility that other "PKR-like" molecules may compensate for the absence of PKR (in PKR-deficient macrophages), and the recent evidence that the PKR Ϫ/Ϫ mice used in this study may be incomplete knockouts (47).
In summary, these studies have identified a novel mechanism by which viral infection and dsRNA stimulate the expression and release of IL-1 by macrophages. dsRNA or EMCV infection stimulates ERK activation and ERK participates in the transcriptional regulation of IL-1 expression by macrophages. PKR does not appear to be required for dsRNA-induced ERK activation or the transcriptional activation of IL-1 expression. The downstream targets of ERK that regulate IL-1 expression in response to dsRNA and virus infection have yet to be defined. One likely candidate is PU.1, a member of the ets family of transcription factors, and a transcription factor known to participate in the activation of IL-1 expression in response to LPS. In support of a role for PU.1 in the regulation of virus-induced IL-1 expression, we have recently shown by gel shift analysis that poly(IC) stimulates the DNA binding activity of PU.1 in macrophages in an U0126-sensitive manner. 2 Additional transcription factors also participate in the regulation of IL-1 expression in response to dsRNA. Inhibitors of NF-B have been shown to prevent dsRNA-induced IL-1 expression and release by mouse macrophages (9). In addition, interferon regulatory factor-4 in combination with PU.1 has been shown to synergistically activate the human IL-1 reporter by interacting at the distal PU.1 consensus binding element (32,48). Importantly, this transcriptional regulation of antiviral response genes following viral infection or treatment with dsRNA appears to be selective. We show that dsRNA-stimulated iNOS expression is not sensitive to inhibitors of ERK, while ERK inhibition prevents virus and dsRNA-stimulated IL-1 expression and release. Recently, we have identified a role for the calcium-independent phospholipase A 2 (iPLA 2 ) in the regulation of dsRNA and EMCV-induced iNOS expression, and the role of iPLA 2 in this antiviral response appears to be selective for iNOS as inhibitors of this phospholipase fail to modulate IL-1 expression in response to dsRNA (49). In this study we have not examined the upstream molecules that activate ERK; however, it is likely that the recently described receptor for dsRNA, TLR3 mediates ERK activation (50). These findings indicate that activation of the antiviral responses is cell type selective and that the pathways that regulate specific antiviral responses may be selective for individual target genes or cellular responses.