Inhibition of agonist-induced activation of phospholipase C following poxvirus infection.

Recent studies indicate that viruses may influence polyphosphoinositide levels. This study has examined the effects of vaccinia virus infection on phospholipase C activity. Infection of BS-C-1 cells, an African Green Monkey kidney cell line, or A431 cells, a human carcinoma cell line, with vaccinia virus inhibits receptor-mediated phospholipase C activation. As a consequence, agonist-mediated Ca2+ mobilization in BS-C-1 cells also was inhibited by vaccinia virus infection. Alleviation of the inhibition of phospholipase C activation was observed in vaccinia virus-infected cells treated with cycloheximide without influencing uninfected cells. Treatment of infected cells with alpha-amanitin, an inhibitor of host mRNA synthesis but not virus mRNA synthesis, failed to alleviate the inhibition of phospholipase C activation. Together these results suggest that a virus-encoded gene product mediates the inhibition of phospholipase C activation without the need of a virus-induced host factor. Analysis of the processes involved in the formation of inositol (1,4,5)-trisphosphate and mobilization of intracellular Ca2+ indicate that the vaccinia virus gene product exerts its inhibitory effects at the level of phospholipase C activity. This may occur by either directly reducing the amount of phospholipase C, reducing the specific activity of phospholipase C, or by inhibiting the association of phospholipase C with its substrate, phosphatidylinositol 4,5-bisphosphate.

Recent studies indicate that viruses may influence polyphosphoinositide levels. This study has examined the effects of vaccinia virus infection on phospholipase C activity. Infection of BS-C-1 cells, an African Green Monkey kidney cell line, or A431 cells, a human carcinoma cell line, with vaccinia virus inhibits receptormediated phospholipase C activation.
As a consequence, agonist-mediated Ca2+ mobilization in BS-C-1 cells also was inhibited by vaccinia virus infection. Alleviation of the inhibition of phospholipase C activation was observed in vaccinia virus-infected cells treated with cycloheximide without influencing uninfected cells. Treatment of infected cells with a-amanitin, an inhibitor of host mRNA synthesis but not virus mRNA synthesis, failed to alleviate the inhibition of phospholipase C activation. Together these results suggest that a virus-encoded gene product mediates the inhibition of phospholipase C activation without the need of a virus-induced host factor.
Analysis of the processes involved in the formation of inositol (1,4,5)-trisphosphate and mobilization of intracellular Ca2+ indicate that the vaccinia virus gene product exerts its inhibitory effects at the level of phospholipase C activity. This may occur by either directly reducing the amount of phospholipase C, reducing the specific activity of phospholipase C, or by inhibiting the association of phospholipase C with its substrate, phosphatidylinositol 4,5-bisphosphate.
Poxviruses are large complex double-stranded DNA viruses that replicate in the cytoplasm of the infected cell in a highly regulated manner (1). Following virus entry into a cell and uncoating of the virion, virus enzymes packaged in the virion initiate early transcription which results in expression of virus-encoded proteins and enzymes and initiates a progression through the virus replication cycle (2). It has been estimated that approximately 75% of the poxvirus genome is required for the production of virus progeny, with the remaining 25% of the genome nonessential for virus replication in tissue culture cells (3). Although the functions of many of the * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. nonessential genes are unknown, they may determine pathogenicity and be vital for virus replication in uiuo.
Within several hours post-infection, poxvirus-encoded gene products initiate extreme changes in the physiology of the host cell which include inhibition of host protein (4), RNA (5), and DNA synthesis (6-8) as well as radical alterations in the internal architecture of the cell as virus gene products accumulate and form a viroplasm or virus factory (2). The effect these drastic changes have on the ability of infected cells to respond to external stimuli such as hormones or growth factors is unknown. The association of some growth factors or hormones with their specific receptors can induce protein, RNA, and DNA synthesis, and thus the ability of poxvirus gene products to alter these signaling events may be beneficial to poxvirus replication. One such signaling network that can induce protein, RNA, and DNA synthesis occurs through the activation of phospholipase C (PLC) ' (9). Phospholipase C cleaves phosphatidylinositol 4,5-bisphosphate (PIP2) into two second messengers, inositol (1,4,5)-trisphosphate ((1,4,5)IP3) and diacylglycerol, which initiate a cascade of events including increases in intracellular free Ca2+ and activation of protein kinase C, respectively (10). Alternatively, PLC activation has been associated with the secretion of inflammatory mediators (11,12), and thus the ability of poxviruses to inhibit PLC activity could be beneficial to poxvirus replication and spread.
25098 vaccinia virus infection can alter host cell responses to extracellular signals.
Therefore, this study has examined the effects of poxvirus infection on PLC activity. To distinguish the effects of VGF from other potential regulators of PLC, a mutant virus, vSC20 (24), that lacks a functional VGF was used to infect the African Green Monkey kidney cell line BS-C-1. A comparison of PLC activity in infected and uninfected cells should indicate whether poxvirus encoded gene products can regulate basal or agonist-stimulated PLC activity.
Cell Culture-BS-C-1 cells were grown in DMEM containing high glucose (4.5 g/liter), 10% FBS, L-glutamine (2 mM), and penicillinstreptomycin (500 units/ml and 500 pg/ml). BS-C-1 cells were labeled in six-well tissue culture plates in 2 ml/well of medium containing rny~-[~H]inositol (50 pCi/ml) for 3 days prior to the direct addition of poxvirus (50 plaque-forming units/cell) into the medium. The effective virus input was 5-15 plaque-forming units/cell during the 18-h infection period. Cell viability was measured by MTT (25). A431 human epidermoid carcinoma cells were generously provided by Dr. Graham Carpenter, Vanderbilt University and were cultured as described previously (26). A431 cells were labeled as described above for BS-C-1 cells; however, for the final 18 h the medium was aspirated, and DMEM in the absence of FBS with my~-[~H]inositol (50 pCi/ ml) was added.
/3H]Inositol Phosphate Measurements-Following culture of BS-C-1 cells with [3H]myoinositol for 3 days, the cells were infected with virus for 18 h, unless otherwise specified, in the continued presence of my~-[~H]inositol (50 pCi/ml). After infection the medium was aspirated and fresh medium (containing 1% FBS in the case of BS-C-1 cells or no FBS for A431 cells) either with or without agonist was added to the cells for various times. Cells were quenched by the rapid aspiration of the medium followed by the addition of 6% perchloric acid + 250 pg/sample of phytic acid. The plates were incubated at 4 "C for 20 min after which time the supernatants were removed and neutralized by the freon/trioctylamine method of Downes et al. (27) as modified by Shears et al. (28). The [3H]inositol phosphates were eluted from an anion exchange HPLC column (Adsorbosphere SAX, 5-pm particle size) using a linear ammonium phosphate gradient (0-1.0 M, pH 3.35) and radioactivity was measured on-line as described previously (29,30).
/3HlInositol Lipid Measurements-Following removal of the acidic supernatants from the tissue culture plates, the remaining precipitates were scraped from the plates and the total [3H]inositol lipids were extracted according to the method of Schacht (31). Briefly, the lipids were extracted with chloroform/methanol/2.4 N HCl (2.5:5:2.3), and the chloroform fraction was counted by liquid scintillation spectroscopy. For analysis of individual [3H]inositol lipids, the lipids within the chloroform fraction were dried down under a stream of Nz. Dried lipid samples were deacylated by the addition of monomethylamine reagent according to the method of Clarke and Dawson (32). Deacylated samples were analyzed by HPLC for the presence of glycerophosphoinositol (GPI), glycerophosphoinositol 4-phosphate (GPIP), and glycerophosphoinositol4,5-bisphosphate (GPIP,) which are the deacylated products of phosphatidylinositol (PI), phosphati-dylinositol4-phosphate (PIP), and PIP,, respectively (33).
Metabolism of Inositol Phosphates by BS-C-1 Cell Homogenates- ['HI (1,4,5)IP~ was added to BS-C-1 cell homogenates (250 pg cellular protein/ml) and incubated at 37 "C in an intracellular-like medium  (1,4,5)IP3 was monitored by elution from an Adsorbosphere SAX column and on-line monitoring of the inositol phosphates (see above).
Determination of Intracellular Ca2+ in BS-C-1 Cells-BS-C-1 cells were grown on glass coverslips for 1-2 days. For Ca2+ measurements the coverslips were placed in a Teflon chamber, and the cells were incubated with 0.5 p~ Fura-2/AM in DMEM + 10% FBS at 37 "C for 1 h. The cells were washed twice with phosphate-free nominally Ca2+-free HEPES-buffered Ringer's solution. Fluorescence emission of two to four cells was monitored at 515 nm following excitation at 350 and 380 nm with a photomultiplier-based spectrofluorimeter (Photon Technology International, Inc., Princeton, NJ) connected to a Nikon Diaphot microscope as described previously (26). Calcium ratios (350/380) were corrected for cell autofluorescence, determined by incubation with 10 p~ ionomycin plus 4 mM MnC1,.

RESULTS
Inhibition of ATP-induced Activation of Phospholipase C by Vaccinia Virus Infection-Treatment of BS-C-1 cells with maximally effective concentrations of ATP (500 KM) induced a rapid increase in (1,4,5)IP3 and its immediate metabolites (Fig. l ) , indicating the presence of functional P2 purinergic receptors. Infection of BS-C-1 cells with the VGF minus virus, vSC20, had no detectable effect on basal levels of (1,4,5)IP3 from 3-24 h post-infection (compare Fig. 1, A with B ) . This indicates that in the absence of VGF no vaccinia virus-derived gene products apparently activate PLC. In contrast, vSC20 infection dramatically inhibited the ATP-induced increase in (1,4,5)IP3 (compare Fig. 1, C with D). This inhibition was detectable as early as 9 h post-infection and was not a result of virus induced decreases in cell viability through 24 h postinfection.2 In addition, the generation of two metabolic products of (1,4,5)IP~, (1,3,4)IP3,and (1,3,4,5)IP4 also were inhibited, suggesting that poxvirus infection decreased the ability of BS-C-1 cells to produce (1,4,5)IP3 rather than increased the metabolism of (1,4,5)IP3. Similar results were obtained following infection of BS-C-1 cells with other orthopoxviruses, including wild-type WR strain of vaccinia virus that possesses a functional VGF (16-18) and cowpox virus (strain * G. Palumbo, unpublished results.

PLC Activity in Poxvirus-infected Cells
Brighton red (34)) (data not shown). Effect of uSC20 Infection on the P 2 Purinergic Receptor-One potential mechanism for the inhibition of phospholipase C activation by poxvirus infection would be through an alteration in the affinity or expression of the agonist receptor.
Generation of an ATP concentration response curve indicated that viral infection reduced the maximal response to ATP but did not substantially alter the apparent affinity of the receptor for ATP (Fig. 2 A ; average from three experiments -10 p~) .
In addition, circumvention of agonist-receptor interaction by direct activation of the G protein associated with PLC with aluminum fluoride (AlF;) (35,36) resulted in responses which also were inhibited in vSC20-infected cells (Fig. 2B).
These results indicate that the major inhibiting action of poxvirus infection occurs at a site subsequent to receptor activation.
Analysis of Phospholipase C Activation in A431 Cells-Since direct G protein activation of phospholipase C by AlF; treatment was inhibited by vaccinia virus, it was conceivable that the mechanism of inhibition could be through either a reduction in the level of the G protein expressed or an inhibition of the association of the G protein with PLC. To address these possibilities, the effects of virus infection on PLC-p activation, a G protein-dependent isozyme of phospholipase C (37,38), was compared with the effects of virus infection on PLC-7 activation, a G protein-independent tyrosine phosphorylation-dependent isozyme of phospholipase C (39,40), but see (41). By utilizing A431 cells, which possess both a G protein-dependent receptor (bradykinin, 100 nM, (42)) and a G protein-independent receptor (EGF, 30 nM, (41)), activation of both PLC isozymes were examined in the same cell.
Since EGF stimulates only minor changes in (1,4,5)IP3 (26), the percent maximal formation of (1,3,4,5)IP4 in response to EGF in infected or uninfected A431 cells was determined. Vaccinia virus infection inhibited both the G protein-dependent activation of PLC by bradykinin as well as the G proteinindependent activation of PLC by EGF (Fig. 3). These data demonstrate that vaccinia virus infection can inhibit PLC activation in different cell lines as well as inhibit the activity of different isozymes of PLC whose activation occurs either through a G protein-dependent or G protein-independent mechanism. Therefore, if the mechanism of inhibition of the different isozymes of PLC is the same in the two cell lines, then these data suggest that vaccinia virus infection inhibits

FIG. 3. Effect of virus infection on EGF stimulation of A43 1 cells.
Uninfected (closed bars) or vsc20-infected (hatched bars) A431 cells were stimulated with either the G protein-dependent PLC activator, bradykinin ( B K ) , or the G protein-indepedent PLC activator, EGF, for 30 s or 1 min, respectively. Shown is the percent maximal response to each agonist in uninfected cells. The data represents the mean & S.E. from three experiments of duplicate determinations.
[3H]Inositol phosphates were separated by HPLC. The percent catabolism of [3H] (1,4,5)IP3 is shown as the mean +-S.D. of duplicate determinations from one experiment of two.
PLC activation subsequent to G protein association with PLC.
Analysis of Inositol Lipids in BS-C-1 Cells-An additional mechanism for virus-mediated inhibition of (1,4,5)IP3 formation could be through a reduction in availability of the PIP, substrate for PLC due to either an overall decrease in total inositol lipids, a change in the ratio of the individual inositol lipids (PI to PIP to PIP,), or a decrease in the accessibility of PLC to PIP2. Following infection, there was  (1,3,4,5)IP4 (0); (1,3,4)IP3 (e); (1,3,4,6)Ip4 (W). no change in the total amount of inositol lipids (lipids from infected cells were 100.4 f 4.2% of lipids from uninfected cells, n = 9), and analysis of the proportion of the individual inositol lipids, PI, PIP, and PIP2, indicated that vaccinia virus infection did not lower, but if anything, appeared to increase the proportion of PIP2 (Table I). Therefore, vaccinia virus infection apparently does not inhibit agonist-induced increases in (1,4,5)IP3 by altering the amounts of inositol lipids within infected cells. Furthermore, since the level of PIPz is a sensitive reflection of cellular ATP (43)(44)(45), this finding indicates that it is unlikely that the effect of poxvirus infection on ( 1,4,5)IP3 levels results from a decreased concentration of ATP.
Effect of Vaccinia Virus Infection on ATP-stimulated Increase in Intracellular Ca2+ Levels-To determine if a physiological response resulting from phospholipase C activation was affected by vaccinia virus infection, Ca2+ mobilization was compared in uninfected and infected BS-C-1 cells. Ca2+ mobilization from internal stores occurs as a result of the interaction of (1,4,5)IP3 with specific intracellular receptors, and the depletion of these intracellular Ca2+ stores signals a sustained influx of Ca2+ from the extracellular space (46). Direct examination of Ca2+ mobilization can be achieved by measuring the intracellular calcium concentration ([Ca"],) in the absence of extracellular Ca2+. Thus, in the absence of extracellular ca2+, effects of vaccinia virus on (1,4,5)IP3 production should be reflected by the extent to which agonist mobilizes Ca2+ from internal stores.
The addition of ATP (20 p~) to uninfected cells resulted i n a rapid and transient increase in [Ca2+],, attributable to release of Ca2+ from internal stores (Fig. 6). Treatment of which empties the microsomal Ca2+ stores by inhibiting the microsomal Ca2+-ATPase, resulted in the further release of only a small amount of Ca2+. Stimulation of vaccinia virusinfected cells with ATP resulted in a diminished Ca2+ response, and the subsequent addition of thapsigargin resulted in a much larger release of Ca2+ as compared with uninfected cells (Fig. 6). This indicates that agonist stimulation of vaccinia virus-infected cells only partially mobilizes the ( 1,4,5)IP3-sensitive Ca2+ pool. These results demonstrate that vaccinia virus infection results in a reduced physiological Ca2+ response to agonist, consistent with the data demonstrating that vaccinia virus infection decreased agonist-induced production of ( 1,4,5)IP3.
Effect of Metabolic Inhibitors on the Generation of (1,4,5)IP3-To rule out the possibility that vaccinia virusinduced inhibition of the generation of second messengers was simply the result of the inhibition of host protein synthesis by virus (4), infected and noninfected cultures of BS-C-1 cells were treated with the protein synthesis inhibitor cycloheximide (100 pg/ml) 1 h prior to, and throughout, infection with vSC20. The inclusion of cycloheximide in noninfected cultures did not inhibit the ATP-induced generation of (1,4,5)IP3 (Fig. 7) or its metabolites (data not shown). This suggests that inhibition of protein synthesis cannot account for the inhibition of (1,4,5)IP3 generation observed in vSC20infected cells. In contrast, cycloheximide treatment ablated the ability of vSC20 to inhibit agonist-induced increases in (1,4,5)IP3 (Fig. 7). These data are consistent with a vaccinia virus-encoded gene product which either directly mediated or induced a host factor which mediated the inhibition of ( 1,4,5)IP3 generation.
To differentiate between these two possibilities BS-C-1 cells were treated with a-amanitin, a selective inhibitor of host RNA polymerase I1 but not the poxvirus encoded polymerase (47). A reversal of the virus-mediated inhibition of (1,4,5)IP3 generation would be consistent with a host-derived PLC Activity in Po gene product as the mediator of inhibition. Fig. 7 demonstrates that the addition of a-amanitin (4 pg/ml; -1 to 18 h) had no effect on the ability of vSC20 infection to inhibit agonist-induced increases in (1,4,5)IP3. This result is consistent with the expression of a virus-encoded protein mediating the inhibition of PLC.

DISCUSSION
This is the first report that infection by a poxvirus can alter inositol phosphate metabolism in cell culture. Evidence presented in this study is consistent with a virus-encoded mediator being responsible for the observed inhibition in agonist-induced increases in (1,4,5)IP3 formation which subsequently reduces mobilization of intracellular Ca2+ (Figs. 1 and 6). Treatment of poxvirus-infected cells with cycloheximide, an inhibitor of protein synthesis, reversed the virusinduced inhibition of inositol phosphate formation (Fig. 7). In contrast, treatment of uninfected BS-C-1 cells with cycloheximide failed to affect agonist-induced increases in (1,4,5)IP3, which suggests that a general virus-induced inhibition of host protein synthesis was not the mechanism of inhibition ( Fig. 7). This latter result suggests that expression of a poxvirus gene is required to obtain inhibition of ligandinduced inositol phosphate formation. In addition, the failure of a-amanitin to reverse the inhibition in poxvirus-infected cells suggests that an induction of the synthesis of a hostderived gene product is not required.
Stepwise analysis of the components involved in the inositol phosphate signaling pathway indicated that the inhibition in formation of ( 1,4,5)IP3 in vaccinia virus-infected BS-C-1 cells occurred at a step subsequent to receptor activation (Fig. 2) or G protein association with phospholipase C (Fig. 3). Following poxvirus infection the metabolism of (1,4,5)IP3 to either (1,4)IP2 or (1,3,4,5)IP4 (Figs. 4 and 5 ) also was unaffected. Analysis of the inositol lipids indicated that poxvirus infection had no inhibitory effect on either the total inositol lipid levels or the relative amounts of the individual inositol lipids PI, PIP, or PIP2 (Table I). In addition, vaccinia virus infection inhibited the physiological mobilization of Ca2+ mediated by (1,4,5)IP3 (Fig. 6). Together, these results indicate that the mechanism of inhibition of (1,4,5)IP3 formation in vaccinia virus-infected cells resides at the level of phospholipase C activity by either directly reducing the amount of Nxvirus-infected Cells phospholipase C, reducing the specific activity of phospholipase C, or by inhibiting the association of phospholipase C with its substrate phosphatidylinositol 4,5-bisphosphate.
One proposed mechanism by which the metabolism of PIP2 by PLC could be inhibited is through competitive binding of PIP2 by one of the actin binding proteins such as profilin (22,23). Thus, an increase in the amount of profilin within a cell would inhibit the ability of PLC to cleave PIP2. The Copenhagen and WR strains of vaccinia virus both encode a gene with strong homology to mammalian profilin (20, 21). However, infection with a profilin homologue deletion mutant (21) (kindly provided by Drs. B. Moss and R. Blasco, Laboratory of Viral Diseases) resulted in comparable inhibition of PLC activity as was observed with vSC20 or WR vaccinia v i r u~e s .~ This result does not rule out the possibility that the virus profilin homologue may play an important regulatory role in the activation of another form of PLC, but it does indicate that an additional vaccinia virus-encoded gene product inhibits PLC activity in the experimental protocol of this study.
In contrast to poxvirus-induced inhibition of inositol phosphate formation, several viruses have been found to enhance inositol phosphate levels either in the presence or absence of an exogenous agonist (48)(49)(50)(51). Only one virus gene product, the middle T antigen of polyoma virus, has been identified and linked to changes in levels of polyphosphoinositides. In an analogous manner to growth factor-induced signaling, the middle T antigen of polyoma virus associates with and activates a PI 3-kinase which produces a minor population of inositol lipids phosphorylated at the D-3 position of the inositol ring (51). In combination, these studies suggest that some viruses likely trigger activation or modification of host cell signaling pathways and thus possibly affect subsequent processes controlled by these pathways.
There are probably several genes encoded by poxviruses which are capable of influencing host cell pathways involved in the generation of second messengers. Evidence presented in this study suggests poxviruses encode a t least one molecule which is capable of inhibiting PLC activity, although the identity of this virus-encoded gene product is presently unknown. Several systems have demonstrated that PLC stimulation can induce the secretion of inflammatory mediators (11,12), and therefore, poxvirus inhibition of PLC activity may limit inflammation at the site of infection. Since an inflammatory response can limit virus infection (52), its down-regulation during infection may enhance virus survival in the host. In addition, it is likely that VGF can increase PLC activity, although it is not known if this can occur in infected cells and/or neighboring cells which could serve as targets for virus replication. Interestingly, the deletion of VGF from the parental WR strain of vaccinia virus decreases its virulence by approximately 5 orders of magnitude in the mouse intracranial LDS0 assay (24), which suggests that virus replication in vivo is enhanced by production of VGF. Studies are presently underway to identify poxvirus genes responsible for regulation of PLC so that the importance and function of these molecules in virus replication can be determined.