Proc. Natl. Acad. Sci. USAVol. 90, pp. 4475-4479, May 1993Biochemistry

Cytoplasmic phospholipase A2 activity and gene expression arestimulated by tumor necrosis factor: Dexamethasone blocksthe induced synthesis

(arachidonic acid/glucocorticoids)

WOLFGANG G. HOECK*, CHAKKODABYLU S. RAMESHAt, DAVID J. CHANG*, NANCY FAN*,AND RENU A. HELLER*tInstitutes of *Biochemistry and Cell Biology and tlmmunology and Biological Sciences, Syntex Research, 3401 Hillview Avenue, Palo Alto, CA 94304

Communicated by Ronald W. Davis, December 21, 1992

ABSTRACT The interaction of tumor necrosis factor a(TNF) with its two membrane-bound receptors initiates intra-cellular events in which arachidonic acid and its derivatives areinvolved. In HeLa cells, TNF treatment induces an arachidonicacid-selective, Ca2+-dependent cellular phospholipase A2(cPLA2). By itself, TNF causes a modest increase in cPLA2activity, but with the Ca2+ ionophore A23187 it provides astrong synergistic action. Within minutes in response to TNF,cPLA2 becomes phosphorylated and in the presence of Ca2+produces a 3- to 4-fold increase in activity. TNF also increasescPLA2 mRNA and protein expression, an estimated 5-foldincrease in an 8-hr period. This increase in cPLA2 activityoccurs, therefore, in a biphasic time-dependent manner. Dex-amethasone, known to antagonize the action of TNF, is hereshown to inhibit TNF-induced gene expression and to preventthe second phase of increase in cPLA2 activation. Our resultssuggest that the cPLA2 activation may provide a regulatoryfunction and may explain the proinflammatory action of TNF.

Tumor necrosis factor alpha (TNF) is a multifunctionalcytokine produced mainly by monocytes and macrophages(1). TNF is now considered to be one of the major "inflam-matory" cytokines, playing an essential role as an immuno-stimulant to increase host defense mechanisms against in-fections (2). In this role, TNF induces the production ofprostaglandins, leukotrienes, and platelet-activating factor,which serve as inflammatory mediators. Therefore, a varietyof events are initiated through the interaction ofTNF with itsmembrane-bound receptors. The presence of two divergentreceptors (3-6) and the multiple effects of TNF suggest thatvarious functions of TNF may be mediated by coupled orindependent receptor-activated signal transduction pathways(7).

Early events of TNF-mediated signal transduction appearto involve lipids as mediators (8-16). For example, TNFstimulates the release ofarachidonic acid (AA) from a varietyof cell types (11-13), and AA-depleted cells are less suscep-tible to the cytotoxic effects ofTNF (14). The induction oftheprotooncogene c-fos by TNF is mediated through lipoxyge-nase products of AA (15), which are also required forTNF-cycloheximide-induced cytotoxicity (16). Since AAmetabolites such as prostaglandins and leukotrienes play animportant role in inflammation (17, 18), the proinflammatoryproperties of TNF could be ascribed to its ability to releaseAA. Of the several enzymatic pathways responsible for therelease of AA, which is esterified to the sn-2 position ofphospholipids, phospholipase A2 is an important enzyme forthis hydrolytic cleavage reaction (18, 19). Thus, one possi-

bility is thatAA release by TNF is mediated through its effecton PLA2.

Phospholipase A2 enzymes can be grouped into two classesbased on their molecular weight and cellular distribution. The"'14-kDa small molecular weight granule-associated secre-tory PLA2 (sPLA2) enzymes are rich in disulfide bonds,require millimolar levels of Ca2+ for activity, but display noapparent selectivity for AA-containing phospholipids. Theypresumably play a role in inflammation, digestion, and scav-enging and are induced by TNF (20-23). On the other hand,the recently described and cloned high molecular weightcytosolic PLA2 (cPLA2) selectively liberates AA from thesn-2 position of membrane phospholipids (24-28) and iscoupled to ligand-regulated AA release (29). Therefore, it isan important enzyme for the biosynthesis of prostaglandinsand leukotrienes. Here, we demonstrate that receptor-mediated AA release occurs by TNF. TNF treatment phos-phorylates cPLA2 and with the Ca2+-ionophore A23187causes a biphasic release of AA. The initial rapid increase inactivity correlates with the phosphorylation of the protein,and the second rapid increase correlates with the synthesis ofnew cPLA2. The antiinflammatory glucocorticoid dexameth-asone does not block TNF-induced cPLA2 phosphorylation,but it inhibits TNF-induced cPLA2 synthesis.

MATERIALS AND METHODSCell Culture and PLA2 Assays. HeLa cells were grown as

described (7). Assays to determine the activity of sPLA2 inthe culture medium and cPLA2 in cell extracts were per-formed as described (30).

Release of [3H]AA from HeLa Cells. Cells (-2 x 105 perwell) were grown in 12-well Costar plates to 70-80% conflu-ency. They were incubated for 18 h at 37°C in 2 ml ofDulbecco's modified Eagle's medium (DMEM) containing10% (vol/vol) fetal calf serum and 0.5 ,uCi (18.5 kBq) of[3H]AA per ml of the medium. The cells were washed andincubated in 1 ml ofDMEM containing 0.5% fetal calf serumwith or without the addition of TNF, and the release of[3H]AA and its metabolites from the cells was measured bydetermining the amount of 3H in the medium by scintillationcounting.

[32P]Orthophosphate Labeling of cPLA2 and Its Immuno-precipitation. HeLa cells at 70-80% confluency were prein-cubated for 1 h in phosphate-free medium containing 1%dialyzed fetal calf serum and then were provided with freshphosphate-free medium containing 100 ,uCi of [32P]ortho-phosphate per ml and incubated for 3-4 h before stimulationwith test compounds for 20 min. Cells were lysed in 500 ,ul of

Abbreviations: TNF, tumor necrosis factor a; cPLA2, cellular phos-pholipase A2; AA, arachidonic acid; sPLA2, secretory PLA2.1To whom reprint requests should be addressed.

4475

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Proc. Natl. Acad. Sci. USA 90 (1993)

lysis buffer (50 mM Tris HCl, pH 7.8/1% Triton X-100/0.1%SDS/250 mM NaCl/5 mM EDTA) and centrifuged. Thesupernatant was immunoprecipitated with 2 ,ul of eithernormal rabbit serum or anti-cPLA2 antiserum (29), and theantigen-antibody complexes adsorbed to protein A-Sepha-rose were resolved by SDS/polyacrylamide gel electropho-resis.

Immunoblotting. Total cell extracts were prepared as de-scribed above. Protein concentration was determined byusing the Pierce BCA (Bradford Coomassie assay) proteinassay; 200 ,g of protein per lane was resolved in 9%SDS/polyacrylamide gels and blotted to Immobilon mem-branes (Millipore). Incubation with cPLA2 antibodies wasperformed and antigen-antibody complexes were visualizedby using the nonradioactive electrochemiluminescence(ECL) detection kit (Amersham).RNA Blot Analysis. Northern blot analysis on HeLa cell

total RNA was performed (3). Hybridization buffer contained4 x 106 cpm of cPLA2 riboprobe per ml prepared from a SalI fragment containing the majority of the cPLA2 codingsequence obtained from the plasmid pMT2-hcPLA2 (GeneticsInstitute, Boston) and cloned into the plasmid pGEM-3Z(Promega).

RESULTSTNF and A23187 Increase AA Release. Phospholipase A2

enzyme activity was measured by the release of [3H]AA fromprelabeled HeLa cells. The prelabeled cells at >90% conflu-ency were incubated in DMEM/0.5% FCS growth medium inthe presence or absence of 50 ng of TNF per ml, and therelease of 3H into the medium was determined after theindicated periods of time. A significant increase in the releaseof [3H]AA was evident in 2 h after TNF addition, and therelease continued to increase gradually over a 6-h period (Fig.1 Top). Considering the nature of the experiment, whichmeasures a total increase in [3H]AA in the medium andincludes the spontaneous release of radiolabel by the cells,the sensitivity of the assay is limited for measuring smallincreases in activity. Our results indicate that it takes acouple of hours for the increased activity to become evident,and this consistent increase becomes more obvious withtime. We also have assayed the growth medium of unlabeledHeLa cells under similar conditions for secreted PLA2 ac-tivity, but no increase in the medium was observed over a 4-hperiod (data not presented).

Since the requirement ofPLA2 enzymes for Ca2+ has beenwell documented (24, 28-30) and TNF alone does not appearto alter cellular Ca2+ levels (R.A.H., unpublished results),the effect of increased intracellular Ca2+ induced by themobilizing agent A23187 was examined. TNF alone in cellstreated for 15 min did not produce a noticeable effect but after4 h of treatment caused a modest increase in [3H]AA release(Fig. 1 Middle), results that are similar to those presented inFig. 1 Top. Stimulation of HeLa cells with only the Ca2+ionophore A23187 for 15 min or 4 h caused an -2.5-foldincrease in [3H]AA release over nontreated control cells. Butwhen HeLa cells were treated with TNF for either 15 min or4 h and then exposed to A23187 for 30 min, a 3- to 5-foldincrease in [3H]AA release was obtained. These data suggesta synergistic action ofTNF with A23187 in stimulating PLA2activity.

Biphasic Release of AA Occurs in Response to TNF withA23187. In time-course studies, TNF with A23187 showed abiphasic pattern of AA release (Fig. 1 Bottom). A firstrapid-but-transient 3- to 4-fold increase in AA release oc-curred in 15-30 min of treatment as compared with theionophore alone, followed by a second greater and moreprolonged action to produce a 16-fold increase in 8 h. Theantiinflammatory action of glucocorticoids is well known and

0

E0.U

Treatment Time (h)

5.0C;,,° 4.0x

a3 3.0u)

co)a> 2.0

Ea~ 1.0-0

cn6x

-oci)cna)

E0.

U BasalEl TNF* A2318* TNF+A23187

F15min 4h

Treatment Time

2 4 6TNF pretreatment in h

FIG. 1. (Top) Time-dependent release of [3H]AA from prelabeledHeLa cells. [3H]AA and its metabolites released into the medium areexpressed as total cpm released. Data points represent mean cpm oftriplicate samples for TNF treatment and duplicate samples ofcontrol cultures. The insert presents the mean difference in released3H at the indicated time between TNF-treated and -untreated cul-tures. (Middle) Release of [3H]AA in response to TNF and A23187alone or in combination. Two sets of HeLa cells were either nottreated or treated with TNF for 15 min or 4 h. One set was thentreated with 1 uM A23187 alone, and the other set treated with TNFwas followed by treatment with A23187 30 min before measuring thereleased radioactivity. (Bottom) Time course of [3H]AA release inresponse to TNF and A23187 and its inhibition by dexamethasone.HeLa cells were pretreated with TNF for the indicated time periodsin the presence (dashed line) or absence (solid line) of 1 j,Mdexamethasone and then were treated with 1 ,uM A23187 for 30 minbefore measuring released radioactivity.

has long been believed to occur via inhibition ofPLA2 activity(31). Therefore, we tested the action of the synthetic gluco-corticoid dexamethasone, and Fig. 1 Bottom (dashed line)shows that dexamethasone did not reduce the amount ofAAreleased within 15 min of treatment but did block the secondphase of prolonged increase in AA released from the cells.Also, dexamethasone was without effect on the AA releasedby A23187 alone.

4476 Biochemistry: Hoeck et al.

Proc. Natl. Acad. Sci. USA 90 (1993) 4477

cPLA2 Phosphorylation Occurs Within Minutes of TNFTreatment. It has been demonstrated (29) that cPLA2 isphosphorylated in response to a wide variety of agonists andthat this phosphorylation correlates with increased cPLA2activity. We have confirmed these observations with TNF.Since phosphorylated proteins show altered migration pat-terns on SDS/PAGE gels relative to their nonphosphorylatedcounterparts (29), we also analyzed the mobility of cPLA2 inresponse to TNF. In cytoplasmic extracts of resting HeLacells, analyzed by immunoblot analysis, cPLA2 protein wasobserved as a doublet of two equal intensity bands (Fig. 2A,forms I and II). TNF treatment caused a complete shift to theupper more slowly migrating form (form II) within 20 min atconcentrations ranging between 1 and 10 ng/ml (data notshown). Addition of the ionophore A23187 alone had no effect(Fig. 2A, last lane). Whether this shift was due to the phos-phorylation of the protein was tested by in vivo [32P]ortho-phosphate labeling of HeLa cells. The cPLA2 protein wasisolated by immunoprecipitation and analyzed by gel electro-phoresis and autoradiography. The immunoprecipitatedcPLA2 from untreated HeLa cells (Fig. 2B Upper panel, lanes2 and 4) was phosphorylated, and TNF treatment for 20 mincaused an 1.5- to 2-fold increase in the 32p incorporated intocPLA2, which was observed as an increase in band intensity(Fig. 2B, lane 3). Normal rabbit serum did not immunopre-

A A23187: - +

+ + + + + + - +

(min): 0 2 5 10 20 40 0 20 20 20

TNF: - - + -

B -6CD

0.

Dex: - - + +

TNF: +

I1-Lane: 1 2 3 4

C

C II

0-J

FIG. 2. Immunoblot and in vivo phosphorylation of cPLA2. (A)Immunoblot analysis. HeLa cells were treated for the indicated timewith TNF at 50 ng/ml, 1 AM A23187, or a combination of both.Cytoplasmic extracts were prepared, and 200 ,ug of total protein perlane was analyzed. The cPLA2 doublet is referred to as form I andform II. (B) In vivo phosphorylation of cPLA2. (B Upper) HeLa cellswere labeled with [32P]orthophosphate for 4 h and received notreatment (lanes 1, 2, and 4) or TNF at 50 ng/ml for 20 min (lane 3).Immunocomplexes were isolated by using either normal rabbit serum(lane 1) or cPLA2-specific antiserum (lanes 2-4) and separated on anSDS/9%o PAGE gel. The autoradiograph shows an overnight expo-sure. (B Lower) Immunoblot analysis of nonradioactive proteinsshown in B Upper and of the differentially migrating cPLA2 forms Iand II protein bands. Superimposition of the bands in B Upper andB Lower revealed form I to be nonphosphorylated cPLA2 and formII to be phosphorylated cPLA2. (C) Effect of dexamethasone oncPLA2 migration in SDS/PAGE. Immunoblot analysis of cPLA2after treatment of HeLa cells with TNF at 50 ng/ml, 1 AM dexa-methasone, or a combination of both. Forms I and II are as describedabove.

cipitate any phosphorylated proteins (Fig. 2B, lane 1). In anidentical experiment, cells were not radiolabeled, and animmunoblot of their extract was superimposed on the autora-diograph of the phosphorylated material (Fig. 2B Lowerpanel). Only the upper band of the cPLA2 doublet wasphosphorylated (form II), and TNF caused an increase in 32pincorporation as well as a complete shift of the cPLA2 proteininto this upper, slower migrating species. Since previousreports had suggested that glucocorticoids may induce serine/threonine protein phosphatases and dephosphorylate and in-activate PLA2 (31, 32), we analyzed cells in the presence orabsence of dexamethasone for the TNF-induced shift ofcPLA2 protein (Fig. 2C). Alone or in combination with TNF,dexamethasone failed to alter the migration ofcPLA2 protein.Therefore, at least in HeLa cells, dexamethasone does notalter the level of cPLA2 phosphorylation.TNF Increases Expression of cPLA2 mRNA in a Time- and

Concentration-Dependent Manner. To examine the molecularbasis for the second more prolonged increase in AA release,we analyzed the expression of cPLA2 mRNA. TNF wasadded to HeLa cells for different periods of time, and totalRNA was analyzed by Northern blots. A time-dependentincrease in cPLA2 mRNA expression was seen with theearliest visible increase occurring within 3 h of treatment(Fig. 3 Left). In 8 h, an -4-fold increase in steady-statemRNA levels was observed, which was maintained for up to18 h. We have also analyzed other cell lines, such as themouse adipogenic TAl cells and NIH 3T3 cells, and these tooshowed TNF induction ofcPLA2 mRNA (data not shown). Ina TNF dose-response, concentrations between 5 and 10ng/ml were adequate for the 4-fold increase in cPLA2 mRNA(Fig. 3 Right), and these concentrations of TNF are wellwithin a physiological range required to elicit the variousfunctions associated with this cytokine.TNF Also Increases the Total Amount of cPLA2 Protein in

HeLa Cells. The TNF-induced increase in cPLA2 mRNAexpression was reflected in a parallel increase in its proteinlevel. In immunoblot analysis of cytoplasmic extracts ofTNF-treated cells, a gradual increase in cPLA2 protein (Fig.4) was evident between 3 and 6 h oftreatment and by 18 h hadincreased 3- to 4-fold. Again, this increase in cPLA2 proteinlevels was TNF dose dependent (data not shown); thereforethe induction of both cPLA2 mRNA and protein occursconcomitantly.Dexamethasone Antagonizes the TNF-induced Increase in

cPLA2 mRNA Levels. In experiments described above (Fig. 1Bottom), dexamethasone had prevented the second phase ofincrease in cPLA2 activity by TNF but was without effect oncPLA2 phosphorylation (Fig. 2C). To explore the glucocor-ticoid-dependent inhibition of cPLA2 activity, the effect ofdexamethasone on cPLA2 mRNA expression was examined.In total RNA isolated from HeLa cells treated for 18 h with50 ng of TNF per ml plus various concentrations of dexa-methasone, the TNF-induced cPLA2 expression was inhib-ited (Fig. 5). The inhibition occurred at concentrations of 10and 100 nM and was also observed with the natural gluco-corticoid hydrocortisone. Other steroid hormones, like es-trogens or progestins, had no effect. In addition, dexameth-asone also appeared to slightly suppress the basal level ofcPLA2 mRNA.

DISCUSSIONA number of laboratories have provided evidence for theparticipation of AA and/or its metabolites in receptor-coupled intracellular signaling pathways (33, 34). We havepreviously demonstrated that the release of AA is a criticalstep in some of the TNF signaling pathways (14) and that theinduction of the protooncogene c-fos by TNF is mediated bylipoxygenase products of AA (15). More recently, we haveshown that lipoxygenase products are involved in TNF/

"

I I Asolilt.. Alma"Mm:4 -I.;:=--'Zmm ..:..:

Biochemistry: Hoeck et al.

..........

.............

Proc. Natl. Acad. Sci. USA 90 (1993)

Hours: 0 0.5 1 3 8 18

0

TNF(ng/ml): c 6 i 2_2:°

28S w ¾ K A ' -28S:# i' iwww cPLA2 -o W * * * *

18S- -185

*SS@ 16*1-f3-actin -@S@4PS

FIG. 3. Induction ofcPLA2 mRNA. (Left) Time course ofTNF treatment. Total RNA (20 ;.g) from HeLa cells treated with TNF at 50 ng/mlwas analyzed for the expression of the 3.4-kilobase cPLA2 mRNA and 2.0-kilobase 8-actin mRNA. An overnight exposure is shown. (Right)TNF concentration curve. HeLa cells were treated with the indicated concentrations ofTNF for 18 h, and total RNA was isolated and analyzedas described in Left.

cycloheximide-induced cytotoxicity as well as the inductionof manganous superoxide dismutase (16). Since AA is ex-clusively esterified to membrane phospholipids and is re-leased mainly by the action ofPLA2, it was likely that the AAreleased by TNF involved PLA2 action. Since the cytosolicPLA2 had been shown to be highly specific for the release ofAA (29) and was hormonally regulated (24, 27, 29), we choseto investigate the activation of this cPLA2 by TNF.The data presented here demonstrate that TNF in a dose-

dependent manner increases cPLA2 activity by two separatemechanisms. A more immediate action of TNF on cPLA2appears to be a phosphorylation of the enzyme, while along-term effect constitutes a sustained production of newcPLA2.- Phosphorylation by itself causes only a modestincrease in cPLA2 activity (29), and the gradual increase inPLA2 activity by TNF alone possibly reflects this action onnewly synthesized cPLA2 protein. In the presence of Ca2+,however, a 3- to 5-fold increase in AA release is observedwithin 15 min ofTNF treatment, and this increase correlateswell with the phosphorylation of cPLA2 protein. Thus, itappears that only a modest increase in activity occurs viaphosphorylation, but a far more enhanced PLA2 activity iscaused when Ca2+ is provided. Similar observations for aCa2+ requirement were reported for phorbol 12-myristate13-acetate (PMA)-induced cPLA2 activity in Chinese hamsterovary cells (29) and TNF-stimulated human neutrophils (35).The term "primed stimulation" was used to describe theinitial treatment with TNF that responded synergistically toa second agent such as the chemotactic peptide phenylalanyl-methionylarginyl-phenylaninamide (FMLP; ref. 35) or cal-cium ionophores (36). We now believe this priming action byTNF to be the phosphorylation of cPLA2. Therefore at sitesofinflammation it is conceivable that the other growth factorsand cytokines, by stimulating Ca2+ release, provide thesynergistic action with TNF to cause an exaggerated PLA2response and the ensuing inflammatory reactions.

TNF: - + + + + + +

Time (hr): 0 1 3 6 8 18 0

TNF is known to alter the phosphorylation state ofa numberof proteins (37). Dressler et al. (9) have recently described aceramide-activated serine/threonine protein kinase involvedin the TNF signaling pathway. This kinase belongs to a familyofprotein kinases with specificity similar to mitogen-activatedprotein 2 (MAP2) kinases (38), and purified cPLA2 has beenshown to be phosphorylated and activated by a MAP2 kinasein vitro (44). Whether the phosphorylation of cPLA2 by TNFis induced by this pathway to provide the priming functionremains to be determined.The long-term stimulation of AA release from HeLa cells

correlates well with the TNF induction ofcPLA2 message aswell as cPLA2 protein. The exact nature of the effect ofTNFon cPLA2 gene induction is not known. Earlier TNF has beenshown to activate a variety of transcription factors (15, 39),and only an examination of the cPLA2 promotor/enhancersequences will provide information of this transcriptionalcontrol and factors involved. Glucocorticoids here certainlyappear to act via inhibition of cPLA2 gene expression.However, their well-known antiinflammatory properties areascribed to their ability to inhibit PLA2 activity via inductionof phospholipid-binding proteins, called lipocortins (40), andof protein phosphatases (31, 32). Recent reports have ques-tioned the involvement of lipocortins (41), and at least inHeLa cells we find no evidence for the dephosphorylation ofcPLA2 by dexamethasone, a result that also agrees with thefailure of dexamethasone to inhibit the early stimulation ofAA release by TNF. Glucocorticoids exert their action onspecific genes via target regulatory elements for binding theactivated glucocorticoid receptor (42, 43), and here gluco-corticoids may operate via a repressor action by interferencewith the TNF-induced cPLA2 gene expression.

Dex (gM): - - N Y QyTNF (50 ng/mI): - + + + + X

-* w* Fw-28ScPLA2 3W

- 18S

cPLA2 -* :

fi-actin o 0 it. * * * lo:FIG. 4. Immunoblot analysis of cPLA2 protein. HeLa cells

treated with TNF at 50 ng/ml for the indicated lengths of time wereused to prepare cytoplasmic extracts, and cPLA2 protein was ana-lyzed as described in text. Two hundred micrograms of protein wasloaded into each lane.

FIG. 5. Effect of dexamethasone on cPLA2 mRNA levels. HeLacells untreated, treated with TNF at 50 ng/ml alone for 18 h, ortreated with TNF in the presence of a range of dexamethasoneconcentrations were used to isolate RNA for this analysis.

4478 Biochemistry: Hoeck et al.

Proc. Natl. Acad. Sci. USA 90 (1993) 4479

PLA2 enzymes reported to play an important role in theproinflammatory action ofTNF have all belonged to the classof small molecular weight secreted enzymes (20-22). Therecent discovery of an intracellular phospholipase A2 enzymehas added a new dimension to the study of phospholipases(27-30). The cPLA2 reported in this paper belongs to this classof enzymes, and our evidence supports its role in receptor-mediated release ofAA. Although we have emphasized a rolefor cPLA2, a contribution by sPLA2 ofAA release cannot becompletely ruled out. We believe, however, that the effect ofTNF observed here is primarily due to cPLA2 for the followingreasons. sPLA2 is a granule-associated, disulfide-containingenzyme released to the outside of the cell. No increase insPLA2 activity in the medium of TNF-treated HeLa cells wasobserved here during a 4-h period, while a substantial increasein [3H]AA is measured within this period and within minutesof TNF treatment in the presence of the Ca2+-mobilizingagent. Also, purified sPLA2 added extracellularly to thegrowth medium of cultured cells is unable to release AA (21).sPLA2 requires millimolar concentrations of Ca2+ for activity,and our experiments with TNF alone were conducted atphysiological Ca2+ levels normally present in the growthmedium. The time course of cPLA2 and sPLA2 induction(21-23) by various ligands is also different. For example, in thisstudy TNF with A23187 induce PLA2 activity within 15 minand the increase in cPLA2 expression within 3 h. On the otherhand, appearance of sPLA2 activity in cultured medium andsPLA2 in cells treated with various agonists including TNF isa relatively delayed response and requires at least 8-24 h.Specific inhibitors of the enzymes are not yet available, andalthough indomethacin and quinacrine have been used toinhibit sPLA2 they are not selective inhibitors (C.S.R., un-published results) and were therefore not used here.From the above discussion it is clear that TNF in the

presence of Ca2+ causes a dramatic increase in cPLA2 activity,while alone it brings about only a limited increase in cPLA2action. This dual action ofTNF, (i) the limited cPLA2 activityobserved in the absence of Ca2+ to conceivably provide alimited source ofAA and its metabolites that can function asintracellular mediators of TNF-induced responses and (ii) asuperinduction of cPLA2 activity in the presence of Ca2+ tooverproduce inflammatory mediators such as eicosanoids andplatelet-activating factor at sites of inflammation, can causetissue destruction and perhaps cytotoxicity.We thank Drs. Jim Clark and Lih-Lin Ling (Genetics Institute,

Boston) for generously providing us with cDNA clones and antiseraand helpful suggestions. Furthermore, we thank Dr. Robert Lewisfor supporting this project, Drs. Paul Findell and Paul Cannon forcritical reading of the manuscript, and Linda Miencier for prepara-tion of this manuscript.1. Vilcek, J. & Lee, T. H. (1991) J. Biol. Chem. 266, 7313-7316.2. Camussi, G., Albano, E., Tetta, C. & Bussolino, F. (1991) Eur.

J. Biochem. 202, 3-14.3. Heller, R. A., Song, K., Onasch, M. A., Fischer, W. H.,

Chang, D. & Ringold, G. M. (1990) Proc. Natl. Acad. Sci. USA87, 6151-6155.

4. Goodwin, R. G., Anderson, D., Jerzy, R., Davis, T., Brannan,C. I., Copeland, N. G., Jenkins, N. A. & Smith, C. A. (1991)Mol. Cell. Biol. 11, 3020-3026.

5. Lewis, M., Tartaglia, L. A., Lee, A., Bennett, G. L., Rice,G. C., Wong, G. H., Chen, E. Y. & Goeddel, D. V. (1991)Proc. Natl. Acad. Sci. USA 88, 2830-2834.

6. Loetscher, H., Pan, Y.-C. E., Lahm, H.-W., Gentz, R., Brock-haus, M., Tabuchi, H. & Lesslauer, W. (1990) Cell61, 352-359.

7. Heller, R. A., Song, K., Fan, N. & Chang, D. (1992) Cell 70,47-56.

8. Kronke, M., Schluter, C. & Pfizenmaier, K. (1987) Proc. Natl.Acad. Sci. USA 84, 469-473.

9. Dressler, K. A., Mathias, S. & Kolesnick, R. N. (1992) Science255,1715-1718.

10. Schutze, S., Pothoff, K., Machleidt, T., Berkovic, D.,Weigmann, K. & Kronke, M. (1992) Cell 11, 1-20.

11. Spriggs, D. R., Sherman, M. L., Imamura, K., Mohori, M.,Rodriguez, C., Robbins, G. & Kupe, D. W. (1990) Cancer Res.50, 7101-7107.

12. Atkinson, Y. H., Murray, A. W., Krilis, S., Vadas, M. A. &Lopez, A. F. (1990) Immunology 70, 82-87.

13. Clark, M. A., Chen, M.-J., Crooke, S. T. & Bomalaski, J. S.(1988) Biochem. J. 250, 125-132.

14. Reid, T., Ramesha, C. S. & Ringold, G. M. (1991) J. Biol.Chem. 266, 16580-16586.

15. Haliday, E. M., Ramesha, C. S. & Ringold, G. (1991)EMBOJ.10, 109-115.

16. Chang, D. J., Ringold, G. M. & Heller, R. A. (1992) Biochem.Biophys. Res. Commun. 188, 538-546.

17. Irvine, R. F. (1982) Biochem. J. 204, 3-16.18. Samuelsson, B., Dahlen, S. E., Lindgren, J. A., Rouzer, C. A.

& Serhan, C. N. (1987) Science 237, 1171-1176.19. Van den Bosch, H. (1980) Biochim. Biophys. Acta 604,191-246.20. Crowl, R. M., Stoller, T. J., Conroy, R. R. & Stoner, C. R.

(1991) J. Biol. Chem. 266, 2647-2651.21. Oka, S. & Arita, H. (1991) J. Biol. Chem. 266, 9956-9960.22. Schalkwijk, C., Pfeilschifter, J., Marki, F. & Bosch, H. V. D.

(1991) Biochem. Biophys. Res. Commun. 174, 268-275.23. Schalkwijk, C., Vervoordeldonk, M., Pfeilschifter, J., Marki,

F. & Bosch, H. V. D. (1991)Biochem. Biophys. Res. Commun.180, 46-52.

24. Gronich, J. H., Bonventre, J. V. & Nemenoff, R. A. (1988) J.Biol. Chem. 263, 16645-16651.

25. Channon, J. Y. & Leslie, C. C. (1990) J. Biol. Chem. 265,5409-5413.

26. Kramer, R. M., Roberts, E. F., Manetta, J. & Putnam, J. E.(1991) J. Biol. Chem. 266, 5268-5272.

27. Clark, J. D., Lin, L. L., Kriz, R. W., Ramesha, C. S., Sultz-man, L. A., Lin, A. Y., Milona, N. & Knopf, J. L. (1991) Cell65, 1043-1051.

28. Sharp, J. D., White, D. L., Chiou, X. G., Goodson, T., Gam-boa, G. C., McClure, D., Burgett, S., Hoskins, J., Skatrud,P. L., Sportsman, J. R., Becker, G. W., Kang, L. H., Roberts,E. F. & Kramer, R. M. (1991)J. Biol. Chem. 266,14850-14853.

29. Lin, L.-L., Lin, A. Y. & Knopf, J. L. (1992) Proc. Natl. Acad.Sci. USA 89, 6147-6151.

30. Clark, J. D., Milona, N. & Knopf, J. L. (1990) Proc. Natl.Acad. Sci. USA 87, 7708-7712.

31. Huang, K.-S., WaUlner, B. P., Mattaliano, R. J., Tizard, R.,Burne, C., Frey, A., Hession, C., McGray, P., Sinclair, L. K.,Chow, E. P., Browning, J. L., Ramachandran, K. L., Tang, J.,Smart, J. E. & Pepinsky, R. B. (1986) Cell 46, 191-199.

32. Zor, U., Her, E., Braquet, P., Ferber, E. & Reiss, N. (1991)Adv.Prostaglandin Thromboxane Leukotriene Res. 21, 265-271.

33. Hannigan, G. E. & Williams, B. R. G. (1991) Science 251,204-207.

34. Peppelenbosch, M. P., Tertoolen, L. G. J., den Hertog, J. & deLaat, S. W. (1992) Cell 69, 295-303.

35. Bauldry, S. A., McCall, C. E., Cousart, S. L. & Bass, D. A.(1991) J. Immunol. 146, 1277-1285.

36. DiPersio, J. F., Billing, P., Williams, R. & Gasson, J. C. (1988)J. Immunol. 140, 4315-4322.

37. Guy, G. R., Cao, X., Chua, S. P. & Tan, Y. H. (1992) J. Biol.Chem. 267, 1846-1852.

38. Seger, R., Ahn, N. G., Boulton, T. G., Yancopoulos, G. D.,Panayotatos, N., Radziejewska, E., Ericsson, L., Bratlien,R. L., Cobb, M. H. & Krebs, E. G. (1991) Proc. Natl. Acad.Sci. USA 88, 6142-6146.

39. Hohmann, H.-P., Brockhaus, M., Baeuerle, P. A., Remy, R.,Kolbeck, R. & VanLoon, A. P. G. M. (1990) J. Biol. Chem.265, 22409-22417.

40. Saris, C. J. M., Tack, B. F., Kristensen, T., Glenney, J. R.,Jr., & Hunter, T. (1986) Cell 46, 201-212.

41. Nakano, T., Ohara, O., Teraoka, H. & Arita, H. (1990)J. Biol.Chem. 265, 12745-12748.

42. Evans, R. M. (1988) Science 240, 889-895.43. Schule, R., Rangarajan, P., Kliener, S., Ransone, L. J., Bo-

lado, J., Yang, N., Verma, I. M. & Evans, R. M. (1990) Cell62,1217-1226.

44. Lin, L.-L., Wartman, M., Lin, A. Y., Knopf, J. L., Seth, A. &Davis, R. J. (1993) Cell 72, 269-278.

Biochemistry: Hoeck et al.