Genetic and Pharmacologic Evidence That Calcium-independent Phospholipase A2β Regulates Virus-induced Inducible Nitric-oxide Synthase Expression by Macrophages*

Recent evidence supports a regulatory role for the calcium-independent phospholipase A2 (iPLA2) in the antiviral response of inducible nitric-oxide synthase (iNOS) expression by macrophages. Because two mammalian isoforms of iPLA2 (iPLA2β and iPLA2 γ) have been cloned and characterized, the aim of this study was to identify the specific isoform(s) in macrophages that regulates the expression of iNOS in response to virus infection. Bromoenol lactone (BEL), a suicide substrate inhibitor of iPLA2, inhibits the activity of both isoforms at low micromolar concentrations. However, the R- and S-enantiomers of BEL display ∼10-fold greater potency for inhibition of the enzymatic activity of iPLA2γ and iPLA2β, respectively. In this study, we show that the iPLA2β-selective (S)-BEL inhibits encephalomyocarditis virus (EMCV)-induced iNOS expression, nitric oxide production, and iPLA2 enzymatic activity in macrophages in a concentration-related manner that closely resembles the inhibitory properties of racemic BEL. cAMP response element-binding protein (CREB) is one downstream target of iPLA2 that is required for the transcriptional activation of iNOS in response to virus infection, and consistent with the effects of BEL enantiomers on iNOS expression, (S)-BEL more effectively inhibits EMCV-induced CREB phosphorylation than (R)-BEL in macrophages. Using macrophages isolated from iPLA2β-null mice, virus infection fails to stimulate iNOS mRNA accumulation and protein expression, thus providing genetic evidence that iPLA2β is required for EMCV-induced iNOS expression. These findings provide evidence for a signaling role for iPLA2β in virus-induced iNOS expression by macrophages.

Recent evidence supports a regulatory role for the calcium-independent phospholipase A 2 (iPLA 2 ) in the antiviral response of inducible nitric-oxide synthase (iNOS) expression by macrophages. Because two mammalian isoforms of iPLA 2 (iPLA 2 ␤ and iPLA 2 ␥) have been cloned and characterized, the aim of this study was to identify the specific isoform(s) in macrophages that regulates the expression of iNOS in response to virus infection. Bromoenol lactone (BEL), a suicide substrate inhibitor of iPLA 2 , inhibits the activity of both isoforms at low micromolar concentrations. However, the R-and S-enantiomers of BEL display ϳ10-fold greater potency for inhibition of the enzymatic activity of iPLA 2 ␥ and iPLA 2 ␤, respectively. In this study, we show that the iPLA 2 ␤-selective (S)-BEL inhibits encephalomyocarditis virus (EMCV)-induced iNOS expression, nitric oxide production, and iPLA 2 enzymatic activity in macrophages in a concentration-related manner that closely resembles the inhibitory properties of racemic BEL. cAMP response elementbinding protein (CREB) is one downstream target of iPLA 2 that is required for the transcriptional activation of iNOS in response to virus infection, and consistent with the effects of BEL enantiomers on iNOS expression, (S)-BEL more effectively inhibits EMCVinduced CREB phosphorylation than (R)-BEL in macrophages. Using macrophages isolated from iPLA 2 ␤-null mice, virus infection fails to stimulate iNOS mRNA accumulation and protein expression, thus providing genetic evidence that iPLA 2 ␤ is required for EMCV-induced iNOS expression. These findings provide evidence for a signaling role for iPLA 2

␤ in virus-induced iNOS expression by macrophages.
The dsRNA 1 -dependent protein kinase (PKR) is a primary mediator of antiviral activities within infected cells. PKR, a 65-68-kDa serine-threonine protein kinase, is activated by binding to dsRNA that accumulates during viral infection and coordinates the antiviral response by inhibiting protein translation, up-regulating antiviral and proinflammatory gene expression, and inducing apoptosis (1,2). PKR suppresses protein translation in virus-infected cells by phosphorylation and subsequent inhibition of the eukaryotic translation initiation factor eIF2␣, whereas PKR-mediated gene transcription appears to be due to activation of transcription factors such as nuclear factor (NF)-B (3).
Whereas the regulation of the antiviral response has classically been thought to be mediated by PKR, recent reports support the existence of additional PKR-independent pathways that participate in antiviral activities within infected cells. In particular, treatment of PKR Ϫ/Ϫ mouse embryonic fibroblasts with dsRNA or infection with encephalomyocarditis virus (EMCV) stimulates inhibitory protein B (IB␣) degradation as well as NF-B nuclear localization, DNA binding activity, and promoter activation, suggesting that PKR is not absolutely required for virus-induced activation of NF-B-mediated gene expression (4). Furthermore, experimental evidence suggests PKR-independent pathways regulate dsRNA-and virus-induced expression of the NF-Bresponsive proinflammatory genes interleukin-1 (IL-1) (5), cyclooxygenase-2 (6), and the inducible isoform of nitric-oxide synthase (iNOS) (7) by macrophages.
The expression of iNOS and subsequent production of nitric oxide appear to play a primary role in innate immunity because viral infection of rodent (8) and human (9) macrophages stimulates iNOS expression and iNOS-deficient mice have increased viral titers and higher mortality than wild-type mice following virus infection (10,11). In an attempt to elucidate the mechanisms that regulate proinflammatory gene expression in macrophages during viral infection, we have recently identified a role for calcium-independent phospholipase A 2 (iPLA 2 ) in a novel PKR-independent signaling pathway that participates in the transcriptional regulation of iNOS (7). Treatment of macrophages with dsRNA or infection with EMCV results in an increase in membrane-associated iPLA 2 enzymatic activity and iNOS mRNA accumulation and protein expression. The suicide substrate inhibitor of iPLA 2 , bromoenol lactone (BEL), inhibits EMCV-induced iPLA 2 activation and iNOS expression in a concentration-dependent manner, supporting a role for iPLA 2 in the regulation of iNOS expression by macrophages. The mechanism by which iPLA 2 mediates the transcriptional activation of iNOS appears to be associated with membrane phospholipid hydrolysis and liberation of a lipid-derived mediator that is capable of activating downstream signaling targets that participate in the expression of iNOS. Importantly, we have shown that the inhibitory effects of BEL on dsRNA-induced iNOS expression by RAW264.7 cells can be overcome by the exogenous addition of a lipid product resembling those generated by iPLA 2 activity, lysoplasmenylcholine (7). iPLA 2 appears to selectively regulate iNOS expression in response to viral infection because inhibition of iPLA 2 activity does not attenuate virus-induced expression of other proinflammatory genes such as IL-1 or cyclooxygenase-2 (6, 7). iPLA 2 catalyzes the hydrolysis of the sn-2 fatty acid substituent of membrane phospholipids to generate a free fatty acid and a lysophospholipid (12,13). Two mammalian isoforms of iPLA 2 (iPLA 2 ␤ and iPLA 2 ␥) have been identified and characterized (14,15), and their enzymatic activity has been implicated in diverse cellular functions including homeostatic membrane phospholipid metabolism (16), secretagogue-induced insulin secretion from pancreatic ␤-cells (17), cell proliferation (18), and apoptosis (19). Historically, the functions of iPLA 2 have been delineated and characterized primarily by sensitivity to the suicide substrate inhibitor BEL. Although BEL is 1000-fold more selective for iPLA 2 than other phospholipases (20,21), distinguishing the biological roles of iPLA 2 ␤ and iPLA 2 ␥ by pharmacologic means had not been possible until the recent recognition that these two enzymes are differentially susceptible to inhibition by the enantiomers (R)-BEL and (S)-BEL. Studies by Gross and co-workers (22) have shown that (R)-BEL is 10-fold more selective for iPLA 2 ␥, whereas (S)-BEL is 10-fold more selective for iPLA 2 ␤. However, it is important to note that BEL enantiomers do not display absolute specificity for iPLA 2 isoforms because a significantly high enough concentration of either (R)-or (S)-BEL will inhibit both iPLA 2 ␤ and iPLA 2 ␥. Accordingly, when used at submaximally effective concentrations, the differential sensitivity of iPLA 2 isoforms to inhibition by the enantiomers of BEL allows for the identification of the iPLA 2 isoform that mediates a given biological response.
Using this pharmacological approach, the identity of the iPLA 2 isoform(s) that mediates the antiviral response of iNOS expression by macrophages was examined. We show that EMCV-induced iNOS expression and nitrite production by RAW264.7 cells and primary macrophages are more sensitive to inhibition by (S)-BEL than (R)-BEL. Furthermore, EMCV-induced iPLA 2 activity and the downstream phosphorylation and activation of the cAMP response element-binding protein (CREB), a transcription factor essential for virus-induced iNOS expression, are attenuated by (S)-BEL but largely unaffected by the presence of (R)-BEL at similar concentrations. Using a genetic approach to confirm these findings, we show that EMCV-induced iNOS expression and nitrite production are abolished in iPLA 2 ␤ Ϫ/Ϫ macrophages as compared with wild-type macrophages. These studies support a primary role for iPLA 2 ␤ in the regulation of virus-induced iNOS expression by macrophages.

EXPERIMENTAL PROCEDURES
Materials and Animals-EMCV was the generous gift of Dr. Ji-Won Yoon (University of Calgary, Calgary, Alberta, Canada) and grown and maintained as previously described (8). Poly(IC) was purchased from Sigma and prepared as previously described (23). The racemic mixture of BEL was obtained from Cayman Chemicals (Ann Arbor, MI), and (R)and (S)-BEL were the generous gift of Dr. Richard Gross (Washington University, St. Louis, MO) and prepared as previously described (22). PCR primers were purchased from Integrated DNA Technologies, Inc. (Coralville, IA). Rabbit anti-iNOS antiserum was a gift from Dr. Pam Manning (Pfizer, St. Louis, MO), rabbit anti-phospho-CREB antiserum was obtained from Upstate USA, Inc. (Charlottesville, VA), rabbit anti-ERK2 was a gift from Dr. John C. Lawrence, Jr. (University of Virginia, Charlottesville, VA), rabbit anti-STAT was purchased from Santa Cruz Biotechnology (Santa Cruz, CA), 3ZD monoclonal mouse anti-IL-1␤ was obtained from Biological Resources Branch, Division of Cancer Treatment, Diagnosis at the National Cancer Institute, and mouse anti-GAPDH antiserum was purchased from Ambion (Austin, TX). C57BL/ 6(J)ϫSV129 mice were obtained from The Jackson Laboratories (Bar Harbor, ME). The generation and maintenance of iPLA 2 ␤ ϩ/ϩ and iPLA 2 ␤ Ϫ/Ϫ mice have been previously described and were generously provided by Dr. John Turk (Washington University, St. Louis, MO) (24). The St. Louis University Institutional Animal Care and Use Committee has approved all animal studies contained in this report.
Cell Isolation, Culture, and Viral Infection-RAW264.7 cells were removed from growth flasks by treatment with 0.05% trypsin/0.02% EDTA for 5 min at 37°C. Primary mouse macrophages were isolated by peritoneal lavage as previously described (25). Following isolation, the cells were washed, plated at 4 ϫ 10 5 cells/400 l of Complete CMRL-1066 as previously described (23), and allowed to adhere for 2 h prior to the initiation of experiments. RAW264.7 cells or primary macrophages (4 ϫ 10 5 cells/400 l DMEM) were preincubated for 30 min at 37°C with the indicated concentrations of racemic BEL or R-or S-enantiomers, followed by the addition of EMCV (1 pfu/cell) or poly(IC) (50 g/ml), and continued in culture for 0.5, 6, or 24 additional h as indicated in the figure legends.
RT-PCR and Quantitative PCR Analysis-Total RNA was isolated from RAW264.7 cells using the RNeasy kit (Qiagen, Santa Clarita, CA) according to the manufacturer's instructions and used to prepare a first-strand cDNA library using the Superscript Preamplification System from Invitrogen. Standard PCR was performed using iPLA 2 ␤, iPLA 2 ␥, iNOS, and GAPDH oligonucleotide primers as previously described (26). Quantitative real-time PCR was performed using the QuantiTect SYBR Green PCR kit (Qiagen) according to the manufacturer's instructions, and iNOS mRNA levels were quantified and normalized to GAPDH. iPLA 2 ␤-specific primers were 5Ј-GTCATTATCCA-GCTTCTCATC-3Ј (forward) and 5Ј-GAGAGTTTCTTCACCTTGTTG-3Ј (reverse), and iPLA 2 ␥-specific primers were 5Ј-CAGTTTCAAAGGCTA-TTTTTGG-3Ј (forward) and 5Ј-TTTTTCCCTTGCTTTTGAAATAG-3Ј (reverse). iNOS and GAPDH primers have been described previously (26), and GAPDH was used as an internal control for all PCRs.
Nitrite and iPLA 2 Assays-Nitrite production was determined by the addition of 50 l of Greiss reagent to 50 l of cell culture supernatant and absorbance at 540 nm compared with a sodium nitrite standard curve (27). The activity of iPLA 2 was examined as previously described (7). In brief, macrophages (6 ϫ 10 6 ) were washed twice with phosphatebuffered saline, resuspended in 400 l of ice-cold PLA 2 homogenization buffer (250 mM sucrose, 10 mM KCl, 10 mM imidazole, 3.8 mM EDTA, and 2 mM dithiothreitol), and lysed by sonication at 4°C. The cell lysate was centrifuged at 14,000 ϫ g for 20 min at 4°C, and the supernatant fraction was further centrifuged at 100,000 ϫ g for 60 min. The membrane pellet was suspended in PLA 2 homogenization buffer. In previous studies we have shown that 75-80% of macrophage poly(IC)-stimulated iPLA 2 activity is membrane-associated (7). The activity of membraneassociated iPLA 2 was determined in 200-l reactions containing 100 l of iPLA 2 assay buffer (100 mM Tris, 4 mM EGTA, and 10% glycerol, pH 7.0), 8 g of membrane protein in 100 l PLA 2 homogenization buffer, and 100 M (16:0, [ 3 H]18:1) plasmenylcholine. Following a 5-min incubation at 37°C, the reactions were terminated by the addition of 100 l of butanol, followed by centrifugation at 2000 ϫ g for 5 min. Released radiolabeled fatty acid was isolated by application of 25 l of the butanol phase to channeled Silica Gel G plates, development in petroleum ether-diethyl ether-acetic acid (70:30:1, v/v/v), and subsequent quantification by liquid scintillation spectrometry. Protein content was determined by the Lowry assay with the use of freeze-dried bovine serum albumin as the protein standard (7,28).
Densitometry and Statistical Analysis-Autoradiograms were scanned into National Institutes of Health Image Version 1.59 using a COHU high performance charge-coupled device camera (Brookfield, WI), and band intensities were determined using National Institutes of Health Image Version 1.59 software. Statistical significance was determined using one-way analysis of variance with Bonferroni post-hoc analysis and reported when p Յ 0.05.

RESULTS
Macrophages Express iPLA 2 ␤ and iPLA 2 ␥-Virus infection has been reported to increase membrane phospholipid metabolism by accelerated phospholipid catabolism, although the specific enzymes participating in the cellular response to viral infection were not known with certainty (30,31). Consistent with these early reports, we have shown that treatment of macrophages with the viral dsRNA mimetic poly(IC) results in a 4-fold increase in membrane-associated iPLA 2 activity that is sensitive to inhibition by BEL. In a similar concentrationrelated manner, BEL also attenuates poly(IC)-and EMCVinduced iNOS expression by macrophages (7). Whereas these studies provide evidence to support a role for iPLA 2 in the antiviral response of iNOS expression by macrophages, the specific isoform of iPLA 2 activated by virus infection that contributes to the transcriptional regulation of iNOS has yet to be established. Two mammalian forms of iPLA 2 have been characterized, and both are expressed in macrophages under basal conditions and following a 6-h infection with EMCV as determined by RT-PCR (Fig. 1). Similar to RAW264.7 cells, peritoneal macrophages express both isoforms of iPLA 2 , and EMCV infection fails to modulate the expression of either isoform (data not shown). To confirm that the iPLA 2 ␥ PCR product was specific and not due to amplification of a homologous sequence of iPLA 2 ␤ cDNA, an iPLA 2 ␤-encoding plasmid (iPLA 2 ␤-pDONR) was used as the DNA template, and primers for iPLA 2 ␤ but not iPLA 2 ␥ produced a PCR product of the predicted size. These findings indicate that the iPLA 2 ␥ PCR product is specific and not due to mispriming or erroneous amplification of iPLA 2 ␤ cDNA. In addition, sequence analysis was used to confirm the identity of each PCR product (data not shown). These results indicate that macrophages express both iPLA 2 ␤ and iPLA 2 ␥ and suggest that either isoform could potentially participate in the regulation of virus-induced iNOS expression.
(S)-BEL Inhibits EMCV-induced iNOS Expression and Nitrite Production by Macrophages-We have recently shown that treatment of macrophages with the viral dsRNA mimetic poly(IC) or infection with EMCV results in iNOS expression and nitrite production, events that are inhibited in a concentration-related manner by the iPLA 2 inhibitors BEL and arachidonyl trifluoromethyl ketone (7). Until recently, it was impossible to distinguish biological effects mediated by an individual isoform of iPLA 2 pharmacologically using BEL be-cause the racemic mixture inhibits both iPLA 2 ␤ and iPLA 2 ␥ enzymatic activity at low micromolar concentrations (15,20). The (R)-and (S)-enantiomers of BEL have, however, recently been shown to selectively inhibit individual iPLA 2 isoforms in a narrow concentration range (22). Because the selectivity of these enantiomers for iPLA 2 ␤ and iPLA 2 ␥ is not absolute and is decidedly dependent upon concentration, we used submaximally effective concentrations of (R)-BEL and (S)-BEL to determine their differential effects on virus-induced iNOS expression by macrophages. Because EMCV infection induces membrane-associated iPLA 2 enzymatic activity, and iPLA 2 ␥ resides primarily in peroxisomes and mitochondria (15), we postulated that iPLA 2 ␥ may participate in the antiviral response and that virus-induced iNOS expression would be inhibited by (R)-BEL. To examine this possibility, the inhibitory actions of racemic BEL, as well as the (R)-and (S)-enantiomers of BEL, on EMCV-induced nitric oxide production ( Fig. 2A) and iNOS expression (Fig. 2B) were examined. Racemic BEL and (S)-BEL inhibit EMCV-induced nitrite production and iNOS expression by RAW264.7 cells in a similar concentration-related manner with Ͼ70% inhibition at 5 M. In contrast, (R)-BEL is much less effective at inhibiting EMCV-induced nitrite production and iNOS expression, with Ͻ40% inhibition at 5 M. Similar to RAW264.7 cells, racemic BEL and (S)-BEL inhibit EMCV ϩ IFN-␥-induced nitrite production and iNOS expression by primary macrophages to similar levels, whereas (R)-BEL fails to significantly inhibit nitric oxide production and iNOS expression (Fig. 3). Alone, (R)-BEL, (S)-BEL, or racemic BEL does not modulate basal nitrite production by RAW264.7 cells or primary mouse macrophages (data not shown). Because (S)-BEL is 10-fold more selective for iPLA 2 ␤ than iPLA 2 ␥, and (S)-BEL is significantly more effective at inhibiting EMCV-and EMCV ϩ IFN-␥-induced nitrite production and iNOS expression by macrophages than (R)-BEL (Figs. 2 and 3), these findings suggest that iPLA 2 ␤ is likely the predominant isoform Total RNA was isolated from RAW264.7 macrophages (1 ϫ 10 6 cells/2 ml DMEM) and used to examine the mRNA accumulation of iPLA 2 isoforms by RT-PCR. iPLA 2 ␤ and iPLA 2 ␥ mRNA accumulates to similar levels in mock-infected or RAW264.7 cells infected with EMCV for 6 h. iPLA 2 ␤, but not iPLA 2 ␥, oligonucleotide primers generate a PCR product from an iPLA 2 ␤-encoding plasmid (iPLA 2 ␤-pDONR), confirming the specificity of the amplification reaction. GAPDH mRNA accumulation was used as an internal control for the RT and PCR reactions. Results are representative of five independent experiments. that participates in the regulation of iNOS expression by macrophages.
(S)-BEL Inhibits dsRNA-induced iPLA 2 Enzymatic Activity-iPLA 2 has historically been ascribed a homeostatic role in membrane phospholipid maintenance, and thus there is a basal level of iPLA 2 enzymatic activity present in unstimulated macrophages. However, in previous studies, we have shown that dsRNA stimulates a membrane-associated phospholipase activity in macrophages that is inhibited by racemic BEL at concentrations (1-5 M) that also inhibit iNOS expression (7). Consistent with these previous findings, treatment of macrophages for 30 min with the viral mimetic poly(IC) results in a 3-fold increase in membrane-associated iPLA 2 activity, and this activation is sensitive to inhibition by racemic BEL and (S)-BEL at a concentration of 5 M (Fig. 4). At this concentration, racemic BEL and (S)-BEL attenuate dsRNA-and EMCVinduced iNOS expression and nitric oxide production (Fig. 1). However, the iPLA 2 ␥-selective (R)-BEL fails to attenuate dsRNA-stimulated iPLA 2 activity at a concentration of 5 M. In a similar manner, racemic BEL and (S)-BEL, but not (R)-BEL, inhibit membrane-associated iPLA 2 activity in EMCV-infected macrophages (data not shown). These data suggest that iPLA 2 ␤ is the isoform activated in macrophages during the antiviral response.
(S)-BEL Inhibits iPLA 2 -mediated Signaling and CREB Activation-We have previously identified the downstream targets of iPLA 2 that participate in the transcriptional regulation of iNOS expression in response to virus infection, particularly protein kinase A (PKA) and the subsequent phosphorylation and activation of the transcription factor CREB (7). Furthermore, we have shown that EMCV infection of RAW264.7 cells and primary mouse macrophages induces the rapid and transient phosphorylation of CREB at 15 and 30 min post-infection (32). As shown in Fig. 5, infection of RAW264.7 cells with EMCV results in CREB phosphorylation within 30 min. Racemic BEL and (S)-BEL, but not (R)-BEL, attenuate EMCVinduced CREB phosphorylation (Fig. 5B). The inhibitory actions of (S)-BEL on EMCV-induced CREB phosphorylation are consistent with the effects of this iPLA 2 ␤-selective inhibitor on poly(IC)-and EMCV-induced iNOS expression and provide further evidence to support a role for iPLA 2 ␤ in the regulation of iNOS expression by macrophages in response to virus infection.
EMCV-induced iNOS Expression Is Attenuated in iPLA 2 ␤ Ϫ/Ϫ Macrophages-As we have used the R-and S-enantiomers of BEL to provide pharmacological evidence consistent with a role for iPLA 2 ␤ in the regulation of virus-induced iNOS expression by macrophages, we used a genetic approach to further support our findings. To confirm that iPLA 2 ␤ participates in this antiviral response, the effects of EMCV infection on iNOS expression by macrophages isolated from iPLA 2 ␤ Ϫ/Ϫ mice were evaluated. In the presence of IFN-␥, EMCV infection of wild-type macrophages stimulates iNOS protein expression by 24 h (Fig.  6A), which is preceded by mRNA accumulation observed by 6 h post-infection (Fig. 6, B and C). Neither EMCV nor IFN-␥ alone is sufficient to induce iNOS expression, which is consistent with a requirement of two proinflammatory signals for the transcriptional activation of iNOS in primary macrophages (23). In contrast, the combination of EMCV ϩ IFN-␥ fails to stimulate iNOS protein expression (Fig. 6A) and mRNA accumulation (Fig. 6, B and C) by iPLA 2 ␤ Ϫ/Ϫ macrophages. The defect in virus-induced iNOS expression by iPLA 2 ␤ Ϫ/Ϫ macrophages is selective for iNOS and does not appear to be associated with a complete inhibition of the antiviral response because EMCV ϩ IFN-␥ stimulate IL-1␤ expression to similar levels in macrophages isolated from wild-type and iPLA 2 ␤ Ϫ/Ϫ mice (Fig. 6D). These findings are consistent with previous work in which we showed that BEL has no inhibitory effect on dsRNA-induced IL-1 expression by macrophages (7). These data provide genetic evidence to further support our findings using BEL enantiomers and suggest a role for iPLA 2 ␤ in the regulation of virus-induced iNOS expression by macrophages. DISCUSSION PKR is a central regulator of numerous cellular responses during viral infection, including activation of inflammatory gene expression, inhibition of protein translation, and induction of apoptosis. Recently, a number of PKR-independent pathways have been identified that participate in the regulation of transcription factors and inflammatory gene expression. Studies have provided evidence that dsRNA and EMCV acti-  vate NF-B to similar levels in both PKR ϩ/ϩ and PKR Ϫ/Ϫ mouse embryonic fibroblasts and macrophages (4,23). Additional transcription factors that participate in the antiviral response may also be independent of PKR because the dsRNAinduced activation of interferon regulatory factor-1 in macrophages does not require the presence of functional PKR (33). Furthermore, PKR is not required for virus-induced transcriptional activation of the inflammatory genes IL-1, cyclooxygenase-2, and iNOS because EMCV and dsRNA stimulate the expression of these genes to similar levels in mouse macrophages isolated from either PKR ϩ/ϩ or PKR Ϫ/Ϫ mice (6,23). Although NF-B activation is required, alone it is insufficient to stimulate the transcriptional activation of these inflammatory mediators. EMCV-or dsRNA-induced activation of these inflammatory genes requires a second signal transduction cascade, which appears selective for individual target genes. We have shown that dsRNA and EMCV stimulate IL-1 expression in an ERK-dependent manner (5), cyclooxygenase-2 expression requires the activation of c-Jun NH 2 -terminal kinase and p38 (34), 2 and iPLA 2 activation is essential for iNOS expression (7).
The aim of the current study was to further elucidate the role of iPLA 2 in the regulation of iNOS expression by macrophages in response to virus infection. In previous studies, the iPLA 2selective inhibitor BEL was shown to inhibit EMCV-and dsRNA-induced iNOS expression in a concentration-related manner that correlated with the inhibition of iPLA 2 enzymatic activity. We have identified PKA as one downstream target of iPLA 2 that participates in the transcriptional activation of iNOS by phosphorylating and activating the transcription factor CREB. Evidence to support this novel lipid-mediated signal transduction pathway in macrophages includes 1) the inhibition of dsRNA-induced iNOS expression by inhibitors of PKA, 2) the ability of direct PKA activation (using non-hydrolyzable analogues of cAMP) to overcome the inhibitory actions of BEL on dsRNA-induced iNOS expression, 3) dsRNA stimulation of CREB phosphorylation and CREB-dependent reporter activity in a BEL-sensitive manner, and 4) the observation that mutations in the proximal and distal cAMP response elements of the iNOS promoter result in near complete inhibition of EMCVinduced iNOS reporter activity. Furthermore, exogenous addition of an analogue of lysoplasmenylcholine, a product of iPLA 2 activity, reconstitutes dsRNA-induced iNOS expression in the presence of BEL. These findings are consistent with the iPLA 2dependent phosphorylation of CREB following 5 min of global ischemia in adult rat hearts (35) as well as the ability of lysoplasmenylcholine generated by iPLA 2 to stimulate CREB phosphorylation in endothelial cells (36).
BEL is an inhibitor of the group VI phospholipases (20). The two mammalian isoforms in this class of phospholipases are iPLA 2 ␤ and iPLA 2 ␥. Because BEL inhibits both isoforms at low micromolar concentrations, it is impossible to attribute a biological response to a specific iPLA 2 isoform based on inhibition by racemic BEL. Recently, Jenkins et al. (22) used chiral high pressure liquid chromatography to separate and purify the Rand S-enantiomers of BEL and subsequently demonstrated that (S)-BEL is 1 order of magnitude more selective for iPLA 2 ␤ than iPLA 2 ␥. We have examined the specific roles of iPLA 2 ␤ and iPLA 2 ␥ in the regulation of iNOS expression by macrophages in response to EMCV infection using these BEL enantiomers. Whereas mouse macrophages express both mammalian isoforms of iPLA 2 , the iPLA 2 ␤-selective (S)-BEL is 10-fold more effective at inhibiting EMCV-induced iNOS expression, nitrite production, and CREB phosphorylation as compared with (R)-BEL. Consistent with the inhibition of iNOS expression, (S)-BEL and racemic BEL inhibit dsRNA-induced iPLA 2 activity in RAW264.7 cells, whereas at the same concentration, (R)-BEL fails to inhibit iPLA 2 activity. The inability of (R)-BEL to inhibit iPLA 2 enzymatic activity suggests that iPLA 2 ␤, but not iPLA 2 ␥, is activated in response to dsRNA and that iPLA 2 ␤ selectively regulates the expression of iNOS. Consistent with this interpretation, EMCV-induced iNOS mRNA accumulation and protein expression are completely abrogated in mice specifically deficient in iPLA 2 ␤, despite the presence of iPLA 2 ␥. Taken together, these findings indicate that the participation of individual iPLA 2 isoforms in the antiviral response of iNOS expression is neither redundant nor compensatory such that iPLA 2 ␤ appears to be selectively activated in response to treatment with dsRNA or following virus infection of macrophages.
Whereas our data consistently display a differential sensitivity of virus-induced iNOS expression, nitrite production, CREB phosphorylation, and iPLA 2 enzymatic activity to BEL enantiomers, we failed to completely inhibit these events with BEL at concentrations of 5 M. Given that BEL is not a saturating inhibitor but rather a mechanism-based inhibitor that is only effective when iPLA 2 is in an active form, a portion of the BEL-insensitive component of this antiviral response is likely due to competition with exogenous phospholipids or inaccessibility due to enzyme compartmentalization. The apparent BEL insensitivity also derives, in part, from the fact that submaximally effective concentrations of BEL were used, and these concentrations were carefully chosen because it is only in this narrow concentration range that one sees differential effects of the inhibitors. Only at these concentrations is it possible to make meaningful observations about the participation of individual iPLA 2 isoforms to a given biological response. At a sufficiently high concentration, it is not possible to use inhibition by the individual R-and S-enantiomers of BEL to determine effects mediated by individual iPLA 2 isoforms because inhibition of iPLA 2 ␤ and iPLA 2 ␥ by (R)-and (S)-BEL is not absolute. In addition, the ability of racemic BEL to inhibit virus-induced iNOS expression and nitrite production to levels 2 S. A. Steer and J. A. Corbett, unpublished data. similar to (S)-BEL reflects that, at most, there is a theoretical 1.6-fold difference between potencies of racemic BEL and (S)-BEL because low concentrations of (R)-BEL inhibit the activity of iPLA 2 ␤ by ϳ30% (22).
Although BEL is a selective inhibitor of iPLA 2 , it has been shown to inhibit another enzyme involved in phospholipid metabolism, phosphatidate phosphohydrolase-1, thus caution must be exercised when assigning biological activities to iPLA 2 when studies are conducted solely using BEL. To substantiate our findings using pharmacological inhibitors and to confirm that iPLA 2 is a primary regulator of virus-stimulated iNOS expression, we examined the effects of EMCV infection on iNOS expression in macrophages isolated from the recently generated iPLA 2 ␤-deficient mice (24). Consistent with our pharmacological studies using the R-and S-enantiomers of BEL, EMCV-induced iNOS expression is abolished in iPLA 2 ␤deficient macrophages, whereas EMCV stimulates robust iNOS expression in wild-type controls. Furthermore, the role of iPLA 2 ␤ in the regulation of inflammatory gene expression in response to virus infection appears to be selective for iNOS because EMCV stimulates IL-1 expression to similar levels in macrophages isolated from iPLA 2 ␤ Ϫ/Ϫ or iPLA 2 ␤ ϩ/ϩ mice. These findings provide genetic evidence to further support our pharmacological studies and suggest a requirement for iPLA 2 ␤ in the transcriptional regulation of iNOS expression by macrophages during viral infection.
Whereas these results are consistent with a role for iPLA 2 ␤ in cell signaling, they also address the role of iPLA 2 in the regulation of normal macrophage function. Previous studies have suggested that the primary role for iPLA 2 in macrophages is the regulation of lipid homeostasis, to provide lysophospholipid acceptors for incorporation of arachidonic acid into membrane phospholipids (16). Our findings demonstrate that iPLA 2 also functions in macrophage intracellular signaling by providing products that activate kinases and transcription factors such as PKA and CREB (7).
Although we have implicated the involvement of iPLA 2 ␤ in the antiviral response and characterized its participation in a novel lipid signaling pathway that regulates virus-induced iNOS expression by macrophages, we have yet to identify the mechanism by which virus infection stimulates iPLA 2 ␤ enzymatic activity. Interestingly, the iPLA 2 -dependent regulation of iNOS expression in response to viral infection does not require the classical mediator of antiviral activities in infected cells, PKR, thus implicating iPLA 2 ␤ as an additional PKRindependent signaling pathway that participates in the antiviral response. Furthermore, iPLA 2 and NF-B activation, both of which are required for iNOS expression, appear to be regulated by independent mechanisms because dsRNA and EMCV are equally effective at stimulating NF-B activity in the presence of concentrations of BEL that prevent iNOS expression (7). Recently, we have obtained evidence that activation of signaling cascades that participate in the regulation of inflammatory gene expression by macrophages is receptor-mediated because EMCV infection activates iPLA 2 , NF-B, CREB, and mitogen-activated protein kinases within 15 min. Under these conditions, viral RNA accumulation and protein expression, early indicators of virus infection, are undetectable within infected macrophages. Additionally, treatment of macrophages with empty virions (capsid structures void of viral RNA) results in the time-dependent activation of iPLA 2 , NF-B, CREB, and mitogen-activated protein kinase in a manner similar to wildtype virus (32). The canonical receptors targeted by EMCV to facilitate cell attachment and entry are members of the cell adhesion molecule superfamily, and vascular cell adhesion molecule has been identified as a receptor for EMCV on vascular endothelial cells (37). We are currently investigating the involvement of this family of receptors in the activation of iPLA 2 ␤ during virus infection of macrophages.
FIG. 6. EMCV-induced iNOS expression is attenuated in iPLA 2 ␤ ؊/؊ macrophages. Peritoneal macrophages were isolated from iPLA 2 ␤ Ϫ/Ϫ and iPLA 2 ␤ ϩ/ϩ mice, plated (4 ϫ 10 5 cells/400 l of Complete CMRL-1066), and infected with EMCV (1 pfu/cell) in the presence (AϪC) or absence (D) of mouse IFN-␥ (150 units/ml). Following a 24-h infection, the cells were isolated, and iNOS expression was analyzed by Western blot analysis (A). Total RNA was isolated following a 6-h infection, and iNOS mRNA accumulation was determined by RT-PCR (B) or quantitative RT-PCR (C). GAPDH is shown as an internal control for the PCR reactions. The selectivity of iPLA 2 ␤-mediated gene expression during the antiviral response was analyzed by determining virus-induced IL-1␤ expression by Western blot analysis following a 24-h infection (D). Results are representative of two independent experiments.