Stimulation of Phosphatidylcholine Hydrolysis, Diacylglycerol Release ...

3 downloads 0 Views 718KB Size Report
Oct 5, 2016 - Brendan D. Price, Jonathan D. H. Morris, Christopher J. Marshall, and Alan ...... Kamata, T., Sullivan, N. F. & Wooten, M. W. (1987) Oncogene 1,.
THEJOURNAL OF BIOLOGICAL CHEMISTRY 0 1989 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 264, No. 28,Issue of October 5, PP. 16638-16643,1989 Printed in U.S.A.

Stimulation of Phosphatidylcholine Hydrolysis, Diacylglycerol by Oncogenic RasIs a Release, and Arachidonic Acid Production Consequence of Protein KinaseC Activation* (Received for publication, February 3, 1989)

Brendan D. Price, JonathanD. H. Morris, ChristopherJ. Marshall, andAlan Hall$ From the Institute of Cancer Research, Chester Beatty Laboratories, Fulham Road, London S W3 6JB, United Kingdom

Swiss-3T3 cells were scrape-loaded with oncogeni- internal mitogenic signals independently of external growth cally activated p21'" protein. 10-20 min after intro- factor stimulation. ducing Va112p21" into the cell, diacylglycerol levels Exactly which signalling system(s) ras is involved in is not were increased, but levelsof inositol phosphates were clear. Stimulation of DNA synthesis by ras requires protein unaltered. However, cellular choline and phosphocho- kinase C(Lacal et al., 1987b; Morris et al., 1989), and a line levels were increased with a similar time course number of ras transformed cell lines do contain an activated to that observed for diacylglycerol production, sugprotein kinase C (Kamata et al., 1987; Wolfman and Macara, gesting that ras increases phosphatidylcholine turn- 1987). Since protein kinase C is activated by diacylglycerol over but not phosphatidylinositol turnover. Down-reg- (Nishizuka, 1984), which can be generated by the breakdown ulation of protein kinase C (by prolonged exposure to of phospholipids, e.g. PIP2' (Berridge, 1984), several groups phorbol esters prior to scrape loading) blocked the have looked for changes in phosphatidylinositol turnover and ability of ras protein to elevate the levels of diacyl- diacylglycerol levels in ras transformed cells, but the results glycerol, choline, and phosphocholine. Oncogenic ras have been inconsistent. Some groups have reported that ras can, therefore, cause a substantial increase in diacyl- causes an elevation of both basal and growth factor-stimulated glycerol (which correlates with increased phosphati- PIPz breakdown (Wakelam et al., 1986; Fleischman et al., dylcholine breakdown) in a protein kinase C-depend- 1986; Hancock et al., 1988; Maly et al., 1988), whereas others ent fashion. Va112p21'" also increased arachidonic show a decrease in growth factor stimulation (Benjamin et acid release, which was also dependent on protein ki- al., 1987; Parries et al., 1987). However, activation of protein nase C activation. Induction of DNA synthesis by on- kinase C can itself alter the inositol phosphate levels of cells cogenic ras was unaffected by inhibitors of prostaglan- (Brown et al., 1987), and we have previously shown that din synthesis, indicating that conversion of the re- oncogenic ras, by activating protein kinase C, may alter both leased arachidonic acid to various prostaglandins is the basal and growth factor-stimulated levels of inositol phosnot required for stimulationof DNA synthesis by ras. phates (Priceet al., 1989). The effect of p21" on phospholipid We suggest that ras rapidly activates protein kinase C, metabolism is therefore unclear. which inturn activates a numberof cellular signalling Most studies to datehave utilized cell lines transfected with systems, leading to a sustained increase in diacylglyc- normal or oncogenically activated ras genes. The problem erol levels. This elevation of diacylglycerol could sus- with this approach is that itis difficult to distinguish between tain protein kinase C activationoverthe12-15h the primary (short term) effects of ras and the secondary required for initiationof DNA synthesis. (long term) effects of clonal selection and cell transformation. We have recently utilized a novel method termed "scrape loading" (McNeil et al., 1984) to introduce oncogenically activated ras proteins into Swiss-3T3 cells (Morris et al., The three ras proteins (N-, Harvey-, and Kirsten-ras) can bind and hydrolyze GTP (McGrathet al., 1984) and may act 1989). This approach allows analysis of the immediate effects of adding atransformingprotein to the cell. In previous as guanine nucleotide-binding regulatory proteins inthe papers, we have shown that (i) protein kinase C activation is plasma membrane (Willingham et al., 1983; Willumsen et al., 1984). Oncogenically activated ras proteinscontain single essential for induction of DNA synthesis by ras and occurs within 10 min of scrape loading Va112p21" into thecells; (ii) amino acid substitutions and often have a decreased rate of protein kinase C activation occurs by a mechanism that does GTP hydrolysis. These mutant proteins arealso unresponsive not involve PtdIns breakdown; (iii) ras down-regulates PDGF tothe GTPase-activatingprotein,a cellular protein that and bombesin-stimulated PtdIns breakdown (Morris et al., stimulates GTP hydrolysis by normal p21" (Traheyand 1989; Price et al., 1989); (iv) morphological transformation by McCormick, 1987). These two effects allow oncogenically oncogenic ras occurs through a proteinkinase C-independent activated ras mutants toremain in an "active" (GTP-bound) route (Lloyd et al., 1989). conformation. Such mutants cause cell transformation (BarIn this paper, we have studied the effect of scrape-loaded bacid, 1987) and appear to act by constitutively activating Val12p21'"" on a variety of second messenger systems over the first 60 min after scrape loading. The results show that * This work was supported by the Cancer Research Campaign and Va112p21'"" increases PtdCho breakdown and causes release the Medical Research Council. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. $ To whom correspondence should be addressed. Tel.: 01-352-8133 (ext. 5177).

The abbreviations used are: PIP*, phosphatidylinositol 4,5-bisphosphate; PGExh,prostaglandin EI/E2; PMA, phorbol 12-myristate 13-acetate; PDGF, platelet-derived growth factor; PtdCho, phosphatidylcholine; PtdIns, phosphatidylinositol.

16638

16639

Ras Stimulates PtdCho Turnover ofdiacylglycerol and arachidonic acid metabolites. Downregulation of protein kinase C (with phorbol esters) blocked these changes in lipid metabolism caused by ras. Production of prostaglandins from ;arachidonic acid was not important for long-term stimulation of DNA synthesis or morphological transformation by the ras protein. In addition, cAMP levels were unaffected byras.

For estimation of DNA synthesis, confluent monolayers of cells were scraped as described and allowed to reattach to gelatin/fibronectincoated multiwells in 1 ml of Dulbecco's modified Eagle's medium containing 0.5 pCi/ml [6-3H]thymidine. [3H]Thymidine incorporation was measured as in Morris et al. (1989). RESULTS

Diacylglycerol Production-Cells were labeled with [ l(3)3H]glycerol, scraped i n the presence of ras protein or buffer, and then resuspended in incubation medium. Reactions were terminated at the indicated times and mono-, di-, and triacMaterials [methyl-3H]Choline,[1(3)-3H]glycerol,[5,6,8,9,11,12,14,15-3H]ara- ylglycerols separated and counted. The results are shown in chidonic acid, cAMP radioimmunoassay kit, and PDGF (v-sis) were Fig. lA.Scraping cells in buffer (Fig. lA, 0)elevates the basal from Amersham International. PMA, PGEI, indomethacin, choline, rate of diacylglycerol production. Previously (Morris et al., and lipid standards were fr'om Sigma. TLC plates were from Merck. 1989), we demonstrated that scraping cells does not activate BW755c was a gift from Wellcome. Harvey Va112p21'" and protein kinase C, so the increase in basal levels of diacylglycSer186.Va112p21rm wereprepared from an Escherichia coli expression erol seen in control cells (Fig. 1) is probably due to diacylvector as described previously (Hall and Self, 1986). glycerols that do not activate protein kinase C (e.g. 1,3- or 2,3-diacylglycerols, ether-linked &glycerides, or certain 1,2Methods diacylglycerols; Cabot and Jaken, 1984; Rando and Young, Cell Culture and Labeling Protocols-Swiss-3T3 cells were grown as described previously (Morris et al., 1989). For choline and arachi- 1984; Hannun e t al., 1986; Ganong et al., 1986; Sekiguchi et donic acid labeling, 30-mm dishes of confluent cells were labeled al., 1988). In cells scraped in the presence of Va112p2lrn", the overnight in serum-free Dul.becco's modified Eagle'smedium contain- levels of diacylglycerol are increased compared with controls ing [rnethyl-3H]choline(12 pCi/ml) or [5,6,8,9,11,12,14,15-3H]arachiThis increase occurs between 10 and 20 min (Fig. lA, 0,O). donic acid (4 pCi/ml). For glycerol labeling, cells were seeded at 5 X after scrape loading and continues for at least 60 min. The lo' cells per 30-mm dish in 10%fetal calf serum plus [1(3)-3H]glycerol time course of this increase in diacylglycerol appears to lag (20 pCi/ml) and allowed to reach confluence. Assays-Cells were preincubated for 2 X 20 min in 1ml of modified behind the activation of protein kinase C seen in scrapeHanks' (Ca*+-free)containing Tris (10 mM, pH 7.0), glucose (10 mM), loaded cells (Morris et al., 1989). Monoacylglycerols are unand bovine serum albumin (fatty acid-free, 1mg/ml). For choline- or changed until 40-60 min after scrape loading, after which arachidonic acid-labeled cells the Hanks' solution was supplemented time a small b u t significant increase is observed (Fig. lA, 0, with unlabeled choline (1 InM) or arachidonic acid (50 pM) to wash H).Triacylglycerol levels were unchanged Va112p21" by (conout unincorporated label. Cells were then washed once in phosphatebuffered saline and then scraped in 80 pl of buffer (10 mM Tris, pH trol = 7.68 k 0.57 cpm %; Va112p21'"" = 7.33 k 0.55 cpm %). As a control for the effects of oncogenic ras protein, a 7.0; 50 mM NaCI; 5 mM MgCl,) or buffer plus Va112p2lrn8(3 mg/ml) as described previously (Morris et al., 1989). Cells were resuspended biologically inactive protein, Ser186.Va112 p21'"" was scrapein 5 ml of modified Hanks', centrifuged at 800 X g, resuspended, and loaded. This mutant lacks the palmitoylation site (Cys-186) aliquoted into microcentrifuge tubes in a total volume of 180 pl of essential for membrane anchoring (Willumsen et al., 1984) Hanks'. For choline, diacylglycerol, and PGE, experiments, 2 x 30and does not transform cells.Table I shows that this defective mm dishes were combined per experiment. After incubation at 37 "C under the appropriate conditions, cells were centrifuged at 800 X g, protein does not stimulate either diacylglycerol or monoacylthe mediumremoved, and the reaction quenched with 675 plof 1.6 I I CHC13:MeOH (1:2). 180 pl of water was added and the phases split as in Morris et al.(1989). For measurement of cAMP levels, the reaction was stopped with 20 pl of perchloric acid (3 M), incubated for 10 min on ice, neutralized with 30 pl of triethanolamine (2 M ) , and left for 1 h on ice. Tubes were centrifuged for 3 min in a s t / Ai microcentrifuge, and an aliquot of the supernatant was assayed for cAMP using the cAMP ass.ay kit supplied by Amersham. To separate choline metabolites from [3H]choline-labeledcells, the aqueous phase of the CHC13:MeOH extract was dried under vacuum, resuspended in 50% MeOH, applied to Silica Gel 60 TLC plates, and separated in 0.6% NaClMeOH.35% NH3 (50:50:1). To separate the I i mono-, di-, and triacylglycerols from [3H]glycerol-labeledcells, the lipid phase was dried under nitrogen, resuspended in CHC13:MeOH (1:2), applied to Silica Gel60 TLC plates, and separated in hexane:diethyl ether:methanol:acetic acid (18040:64) (Habenicht et al., 1981).Arachidonic acid metabolite release was measured by counting an aliquot of the medium after centrifuging the cells. Cellular arachidonate and PGE? from the medium were measured as in Shier and Durkin (1982) and separated on Silica Gel 60 plates in ethyl ace) BW755c tate:acetic acidacetone (9O:lO:l). Indomethacin (2 p ~ and (10p ~ both ) decreased the stimulation of PGE2 synthesis by PDGF from 112% to around 13% (basal level = 0.16 cpm %) over a 10-min stimulation. Phosphocholine, choline, neutral lipids, or prostaglanTIME Onin.) dins were separated in adjacent tracks and identified with iodine vapor. In some experiments lipid or choline standards were added to FIG. 1. Effect of Va112p21'" on mono- and diacylglycerol the radiolabeled cell extracts. In all cases, the radiolabel comigrated levels. Cells were labeled with [3H]glycerol,scraped in buffer (conwith the appropriate stand.ard. After identification, equal areas of the trol; open symbols) or Va112p21"" (closed symbols), and incubated for gel, corresponding to the appropriate metabolite, were scraped di- the appropriate times. A, untreated cells; B , cells pretreated with rectly into scintillation vials and counted. Blank areas of the gel were PMA (40 h, 400 nM) to down-regulate protein kinase C; 0, 0, also counted to give backgrounds and were subtracted. Results are diacylglycerol; 0, monoacylglycerol. Results & S.E. (n = 6). cpm expressed as the number of counts in a particular metabolite as a in lipids per incubation, 49,000cpm.Val12p21""-scraped cells (0) percentage of the total counts in the lipid phase at time 0, termed were significantly different from controls ( p c 0.05) in A at 20, 40, cpm %. Inositol phosphates were measured as in Price et al. (1989). and 60 min. EXPERIMENTALPROCEDURES

"I

.,

Ras Stimulates PtdCho Turnover

16640 TABLEI

0 15

Stimulation of monoacylglycerol and dincylglycerotproduction Cells were labeled with [3H]glycerol,scraped in buffer (control) or buffer containing Va112p21re8or Ser186.Va112p2lr" (3 mg/ml), and incubated for 60 min. Some control cells were treated with PMA (40 nM) or PDGF (40 ng/ml) 10 min after scraping. Results f S.E. ( n = 6). cpm in lipids per incubation, 46,000cpm. Monoacylglycerol

Diacylglycerol

s

cpm %

Control

Ser186.Va112p21r" Va112p21'"'

0.311.02 f 0.03 0.24 f 0.01 1.08 0.51 f 0.07" 0.27 f 0.04 1.43 0.74 -t 0.03" 1.26

ChO,+PKC Ch0,-PKC

&

0.06

f 0.04 1.31 k 0.08" f 0.05"

PMA PDGF & 0.01" " p < 0.05 in two-sided t test compared with controls. glycerol production in scraped cells. Both PMA and PDGF can increase diacylglycerol (Table I) tolevels similar to ras in P-Cho,+PKC P-Cho, - PKC scraped cells and, in addition, PDGF can stimulatemonoac0 20 40 60 20 40 60 ylglycerol production. PtdCho Hydrolysis-Potential sources of diacylglycerol include both PIP, and PtdCho. We have previously demonstrated (Morris et al., 1989; Price et al., 1989) that scrape FIG.2. Effect of Va112p21" on the levels of choline and phosphocholine. Cells werelabeledwith [3H]choline,scrapedin loading Va112p21raEdoes notstimulateinositolphosphate production, either over a short (1 h) or a long (18 h) time buffer (control, open circles) or Va112p21'" (closed circles), and incubated for the indicated times. Levels of cellular choline (Cho) and course. Under the conditions of the experiment in Fig. lA, phosphocholine (PCho) were measured. Some cells were pretreated basal [3H]inositolphosphatelevels 60 min after scrape loading with PMA (40 h, 400 nM) to down-regulate protein kinase C (-PKC). Va112p21'"" were unchanged a t 151 f 17 cpm (control) and Control cells were treated with dimethyl sulfoxide (the solvent for 131 -+ 13 cpm (p2lra'; results f S.E., n = 3), indicating that PMA, 40 h, +PKC). PKC, proteinkinase C. cpm in lipids per the increasein diacylglycerol is not derived from elevated incubation, 486,000cpm. Va112p21ra8cells (0)weresignificantly PtdIns hydrolysis. Consequently, we have looked at alterna- different from controls (0)( p < 0.05) at t = 20, 40, and 60 min in tive sources for diacylglycerol. PtdCho turnover in Swiss-3T3 Cho, +PKC, and P-Cho,+ PKC. cells can lead to release of diacylglycerol, phosphocholine, and TABLEI1 choline (Takuwa et al., 1987). T o measure PtdCho turnover, Stimulation of choline and phosphocholine release cells were labeled with [methyl-3H]choline, scrape-loadedwith Cells were labeledwith [3H]choline, scraped in buffer, Va112p21r", ras, and the levels of choline andphosphocholine determined. or Ser186Va112p2lr" (3 mg/ml), and incubatedfor 60 min.Some Fig. 2 shows that Va112p21raaelevates cellular choline (Fig. 2; control cells were treated with PDGF(40 ng/ml) 10 min after scraping Cho, +PKC) and phosphocholine (Fig. 2; P-Cho, +PKC). in buffer.Results f S.E. ( n = 6 ) .cpm in lipids perincubation,463,000 This increase occurs 10 to 20 min after scrape loading and cpm. Phosphocholine Choline continues for at least 60 min, which is similar to the time __ cprn % course of diacylglycerolproduction(Fig.1A).Ser186. 0.133 f 0.029 0.076 & 0.026 Va112p21'"" does not elevate either choline or phosphocholine Control 0.138 f 0.021 Ser186.Va112p2lra" 0.089 & 0.020 levels (Table 11), whereas PDGF was capable of increasing 0.238 f 0.030" 0.147 0.019" Va112p21"" choline and phosphocholine to about the same level as ras 0.205 f 0.032" 0.128 r+ 0.015" PDGF (Table 11). PtdCho metabolism can also generate two other " p< 0.05 in two-sided t test compared with controls. choline compounds, CDP-choline (involved in the synthesis of PtdCho) and glycerophosphocholine (formed by deacylais abolished in tion of PtdCho). CDP-cholinelevels were very low in scraped the ras-dependent increase in phosphocholine cells and could not be measured accurately. Glycerophospho- down-regulated cells (Fig. 2, P-Cho, -PKC),while the increase choline levels were also low and unchanged by Va112p21" in choline is greatly reduced (Fig. 2, Cho, -PIE). In cells in which proteinkinase C had beendown-regulated, PDGF (data not shown). stimulation of both choline (control = 0.043 -+ 0.014 cpm %; C by Vall2p21""Since Activation of ProteinKinase Va112p21" activates protein kinaseC within 10 minof scrape PDGF = 0.064 f 0.023 cpm %) and phosphocholine (control loading (Morris etal., 1989), which can then activate PtdCho = 0.281 f 0.110 cpm %; PDGF = 0.278 f 0.063 cpm %) blocked. Theseresultsindicatethatthe turnover (Takuwa et al., 1987), it is important to determine releasewasalso whether PtdCho hydrolysis and diacylglycerol release (Figs. stimulation of PtdCho hydrolysis and diacylglycerol release by Va112p21"" is a secondary effect resulting from protein 1A and 2) precede or follow proteinkinase Cactivation. Protein kinase C can be down-regulated by prolonged treat- kinase C activation. Effect of Val12p21roson Arachidonic Acid Metabolism-We ment with PMA (40 h, 400 nM; Rodriguez-Pena and Rozenhave investigated arachidonic acid metabolism to see if this gurt, 1984) leaving no detectable protein on Western blots (Adams and Gullick, 1989; Morris et al., 1989). Fig. 1B shows might provide an early signal after scrape loading oncogenic that in cells lacking protein kinase C, Va112p21'"" no longer rats. Arachidonic acid may be releasedfrom phospholipids by converted, via the cyclooxstimulates either diacylglycerol (Fig. lB, 0, 0 ) or monoacyl- phospholipase A, and can then be glycerol (Fig. lB, 0, H) production. The stimulation of di- ygenase reaction, into various prostaglandins (Pace-Asciak acylglycerol release by PMA in normalcells (see Table I) was and Smith, 1983). We measured the release of arachidonic medium from cells labeledwith [3H] abolished in cells down-regulatedfor protein kinaseC (control acid metabolites into the arachidonic acid. The results areshown in Fig. 3. Stimulation = 1.03 f 0.08 cpm %; PMA = 1.10 f 0.09 cpm %). Similarly,

TurnoverPtdCho Ras Stimulates

16641 TABLE IV Effect of indomethacin on ValpBI"" DNA synthesis

Cells were scraped as describedunder "Methods," then allowed to reattach to gelatin/fibronectin-coated dishes. Insulin (1 pg/ml)or PDGF (40 ng/ml) was added as appropriate. Indomethacin (2 p M ) was added 30 min before scraping and was present throughout the experiment. Results f S.E. ( n 3j. ND = not done. CPM in [3H]-thymidine

TIME(mind

FIG. 3. Effect of Vall:2p21" on the levels of arachidonic acid metabolites. Cellswerelabeledwith [3H]arachidonicacid, scraped in buffer (control, open circles) or buffer containing Va112p21"" (closed circles), and incubated for the appropriate time. Cells were centrifuged and an aliquot of the medium counted. cpm perincubation in lipids,112,000 cpm. Va112p21" cells (0) were significantly different fromcontrols (0)( p < 0.05) at t = 30,40, and

Control Insulin PDGF + insulin Va112p21"" + insulin

-Indomethacin

+Indomethacin

493 f 37 4,633 f 1,095 33,444 f 2,212 24,254 f 1,070

628 f 47

ND 31,678 f 4,327 24,588 k 3,313

vates protein kinase C, mimicked the effects of ras (Table III), and, after down-regulation of protein kinase C, Va112p21'"" was no longer capable of stimulating release of 60 min. arachidonic acidmetabolites. Theseresultsareconsistent with release of arachidonic acid metabolites being due t o TABLE I11 protein kinase C activation. Release of arachidonic acid metabolites from cells Va112p21" Effects on CAMP Accumulation-PGFZ, and Cells were labeled with [313]arachidonic acid, then scraped in buffer PGE, are the major prostaglandins produced in Swiss-3T3 (control) or buffer containing Va112p21" orSer186.Va112p21r" (3 cells from the arachidonicacid released in response to ligand mg/ml), and incubated for 60 min. Cells treated with PMA to down- stimulation(ShierandDurkin, 1982; Burchand Axelrod, regulate protein kinaseC (400 nM, 40 h) are designated -PKC.PMA 1987). These two prostaglandins are capableof elevating PIP2 (40 nM) to stimulate cells was added 10 min after scraping. Results f S.E. ( n = 5). ND = not done. cpm in lipids per incubation,119,000. hydrolysis (PGF2,; Hesketh et al., 1988) and cAMP accumulation (PGE,; Rozengurt et al., 1983). We have already shown Arachidonic acid metabolites that levels of inositol phosphates are unaltered cells in scrape+PKC -PKC loaded with ras (Morris etal., 1989), so any increase in PGFz, cpm % is clearly insufficient to cause inositol phosphate formation. 11.0 f 4.0 9.8 f 1.9 Control We have also looked for a n increase in cAMP levels, caused 10.7 f 0.1 ND Ser186.Va112p21r" by PGEz synthesis from arachidonic acid or direct activation 10.7 f 4.0 18.3 f 0.6" Va112p21ra8 of adenylate cyclase by Va112p21"". Fig. 4 shows that oncoND 15.6 f 0.7" PMA genic ras has no effect on either the basallevel of cAMP (0, " p < 0.05 in two-sided t test compared with controls. 0 )or on stimulationof adenylate cyclase by PGE, (0,m). So Va112p21'"" does not directly activate or inhibit the adenylate cyclase system. Effect of Arachidonic Acid Metabolites on Va112p21m-induced DNA Synthesis-Cells scrape-loaded with Va112p21ra8 can (in thepresence of insulin) initiateDNA synthesis 12-15 h after scrape loading (Morris et al., 1989), and this stimulation is similar to that seen with PDGF plus insulin (Table IV). Metabolismof the arachidonicacid producedafter scrape loading oncogenic ras (Fig. 4) could allow prostaglandins to build up in the medium and provide autocrine stimulation of the cells. This may sustain signalling during the later stages of mitogenesis. To see if arachidonic acid metabolites are I I I involved in thisprocess, cells were pretreated with indometh0 10 20 30 40 acin, an inhibitor of the cyclooxygenase (Lapetina etal., 1978), TIME(minS) to block the first step in prostaglandin synthesis. Under these induced by eitherPDGFplus FIG. 4. Effect of Vall:2p21r" on the levels of CAMP. Cells conditions,DNAsynthesis were pretreated (15 min) in isobutylmethylxanthine(0.5 mM), scraped insulin or ras plus insulin was unaffected by indomethacin in buffer (control open symbols) or Va112p21"" (closed symbols), and (Table IV). This shows that prostaglandin synthesis is unincubated (with 0.5 mM isobutylmethylxanthine)for the appropriate important for ras induced DNA synthesis or early signalling time. cAMP levels were thten assayed. 0, 0 = control; 0, H = plus events. Similar results were obtained with the lipoxygenase PGE, (100 ng/ml). inhibitor BW755C (Walker et al., 1980) indicating that leukotriene synthesis is also unimportant. The stimulation of of arachidonic acid metabolite release by Va112p2lraSoccurs arachidonic acid metabolite release seen with Va112p21'"" in about 20 min after scraping and peaks after 40 to 60 min, Fig. 3 does not, therefore, seem to play an important role in which is slower than diacylglycerol release (Fig. 1)or PtdCho ras dependent transformationof cells. turnover (Fig. 2). Internal levels of free arachidonic acid were slightlyincreased by ras(control = 0.24 k 0.06 cpm %; DISCUSSION Va112p21'"" = 0.34 k 0.08 cpm %), indicating that arachidonic acid metabolite release comes from lipid sources and is not Wehave demonstratedthat10to 20 minafter scrape released from an internal arachidonic acid store. The data are loading oncogenic ras, diacylglycerol levels and PtdCho turnsummarized in Table 1111. Ser186.Va112p21ra" did not cause over are increased, but PtdIns turnover isunchanged. Downany arachidonic acid metabolite release. PMA, which acti- regulation of protein kinase C abolished the stimulation of

16642

TurnoverPtdCho Ras Stimulates

both PtdCho turnoverand diacylglycerol production. Scrapeloaded Va112p2lrRn also stimulated the release of arachidonic acid metabolites in a protein kinase C-dependent fashion, but these arachidonic acid metabolites were not required for induction of DNA synthesis by ras. The levels of CAMPin cells was unaffected by ras. Previously (Morris et al., 1989)we demonstrated that scrape loading Va112p2lra8causes rapid (5-10 min) phosphorylation of the 80-kDa protein substrate of protein kinase C. The kinase is thought to be activated by diacylglycerol(Nishizuka, 1984), but no detectable change in either diacylglycerol levels or PtdCho breakdown can be seen until at least 10-20 rnin after scrape loading ras, and thisincrease is blocked by downregulation of proteinkinase C. This implies that protein kinase C is activated prior to the observed increases in diacylglycerol. Since arachidonic acid (Shearman et al., 1989) and otherphospholipids (Hannun andBell, 1989; Nishizuka, 1989)have also been shown to modulate the activity of protein kinase C, it is possible that ras activates protein kinase C by a mechanism independent of diacylglycerol. Alternatively, ras may stimulate a small(localized) increase in diacylglycerolor perhaps a transient increase in a particular subclass of diacylglycerols in the first 5-10 min after scrape loading, but this diacylglycerol is rapidly metabolized. Preiss et al. (1986) and Lacal et al. (1987a) have also shown that levels of diacylglycerol are elevated in ras transformed cells, but this does not seem to be derived from increased PtdIns turnover (Wolfman and Macara, 1987; Morris et al., 1989).Lacal et al. (1987a) reported increased phosphocholine levels in ras transformed cells, which could indicate increased PtdCho hydrolysis, but the functional significance of this was not explored. The diacylglycerol produced in our system by ras is paralleled by an increase in PtdCho turnover and suggests that PtdCho hydrolysis may be the main source of diacylglycerol, although stimulation of either de novo synthesis (Fareseet al., 1987) or hydrolysis of other phospholipids (e.g. phosphatidylethanolamine; Kiss and Anderson, 1989) may contribute to this. Arachidonic acid release was also stimulated by Va112p21'"" and was blocked in cells in which protein kinase C had been down-regulated. Release of this fatty acid could be due to activation of phospholipase Az (Burch et al., 1988) or by a phospholipase C/diglyceride lipase pathway (Bell et al., 1979), but we are unable to distinguish between these. However, BarSagi and Feramisco (1986) have suggested that phospholipase Az may be activated when cells are micro-injected with ras. Prostaglandin synthesisfrom arachidonic acid release does not seem to be important in ras transformed cells, since inhibition of the cyclooxygenase and lipoxygenase pathways (by indomethacin and BW755C, respectively) had no effect on ras-induced DNA synthesis. Similarly, we could detect no change in either basal levels of CAMPor in PGE, stimulation of adenylate cyclase in cells loaded with Va112p21". In contrast, Tarpley et al. (1986) and Colletta et al. (1988) have reported that hormonal stimulation of adenylate cyclase is decreased in ras transformed cells, but this may be due to down-regulation of receptors, as has been noted in some ras transformed cells (e.g. Benjamin et al., 1987), and may be a result of chronic expression of ras. Induction of DNA synthesis by ras in scrape-loaded cells is dependent on protein kinase C, but morphological transformation is independent of protein kinase C (Morriset al., 1989; Lloyd et al., 1989). This implies that activation of protein kinase C is not the only signal generated by ras. Indeed, growth factor stimulation of DNA synthesis does not require protein kinase C, but does require ras function (Mulcahey et al., 1985; Lacal et al., 198713; Yu et al., 1987), suggesting that

activation of protein kinase C is not the primary (or only) effect of ras. However, once this kinase is activated by ras, it can induce secondary effects, including increased PtdCho turnover, elevated levels of diacylglycerol, and release of arachidonic acid which persist for at least 80 min. These elevated levels of diacylglycerol could then sustain protein kinase C activation over the 15-h period required for the progression of 3T3 cells through to initiation of DNA synthesis. This continued activationof protein kinase Cmay be an important aspect of transformation by oncogenic ras, especially in the absence of growth factors. It still remains, however, to determine the very early biochemical events regulated by ras that initiate and maintain this process. Acknowhdgments-We thank Annette Self for technical assistance and David Linstead for help and advice. REFERENCES Adams, J. C. & Gullick, W. J. (1989) Biochem. J. 257,905-911 Barbacid, M. (1987) Annu. Rev. Biochem. 5 6 , 779-827 Bar-Sagi, D. & Feramisco, S. R. (1986) Science 2 3 3 , 1061-1068 Bell, R. M., Kennerly, D. A., Stanford, N. & Majerus, P. W. (1979) Proc. Natl. Acad. Sci.U.S. A. 7 6 , 3238-3241 Benjamin, C. W., Tarpley, W. G. & Gorman, R. R. (1987) Proc. Natl. Acad. Sci. U. S. A . 84,546-550 Berridge, M. J. (1984) Biochem. J. 2 2 0 , 345-360 Brown, K. D., Blakeley, D. M.,Hamon, M. H., Laurie, M.S. & Corps, A. N. (1987) Biochem. J. 245,631-639 Burch, R. M. & Axelrod, J. (1987) Proc. Natl. Acad. Sci. U. S. A. 8 4 , 6374-6378 Cabot, M. C. & Jaken, S. (1984) Bwchem. Biophys. Res. Commun, 125, 163-169 Colletta, G., Corda, D., Schettini, G., Cirafici, A. M., Kohn, L. D. & Consiglio, E. (1988) FEBS Lett. 2 2 8 , 37-41 Farese, R. V., Konda,T. S., Davis, J. S., Standaert, M. L., Pollet, R. J. & Cooper, D. R. (1987) Science 236,586-589 Fleischman, L. F., Chahwala,S. B. & Cantley, L. (1986) Science 231, 407-410 Ganong, B. R., Loomis, C. R., Hannun, Y. A. & Bell, R. M. (1986) Proc. Natl. Acad. Sci. U. S. A. 8 3 , 1184-1188 Habenicht, A. J. R., Glomset, J. A., King, W. C., Nist, C., Mitchell, C. D. & Ross, R. (1981) J. Biol. Chem. 2 5 6 , 12329-12335 Hall, A. & Self, A. 3 . (1986) J . Biol. Chem. 261,10963-10965 Hancock, J. F., Marshall, C. J., McKay, I., Gardener, S., Houslay, M. D., Hall, A. & Wakelam, M. J. 0.(1988) Oncogene 3,187-193 Hannun, Y. A. & Bell, R. M. (1989) Science 243,500-507 Hannun, Y. A., Loomis, C. R. & Bell, R. M. (1986) J. Biol. Chem. 261,7184-7190 Hesketh, T. R., Morris, J. D. M., Moore, J. P. & Metcalfe, J. M. (1988) J. Biol. Chem. 2 6 3 , 11879-11886 Kamata, T., Sullivan, N. F. & Wooten, M. W. (1987) Oncogene 1 , 37-46 Kiss, Z. & Anderson, W. B. (1989) J . Bwl. Chem. 2 6 4 , 1483-1487 Lacal, J. C., Moscat, J. & Aaronson, S. A. (1987a) Nature 330,269212

Lacal, J. C., Fleming, T. P., Warren, B. S., Blumberg, P. M. & Aaronson, S. A. (1987b) Mol. Cell. Biol. 7,4146-4149 Lapetina, E. G., Chandrabose, K. & Cuatrecasas, P. (1978) Proc. Natl. Acad. Sci. U. S. A. 75,818-822 Lloyd, A., Patterson, H., Morris, J. D. M., Hall, A. & Marshall, C. J. (1989) EMBO J. 8,1099-1104 Maly, K., Doppler, W., Oberhuber, H., Meusberger, H., Hofmann, J., Jaggi, R. & Grunicke, H. H. (1988) Mol. Cell. Bwl. 8,4212-4216 McGrath, J. P., Capon, D. J.,Goeddel, D. V. & Levinson, A. D. (1984) Nature 310,644-649 McNeil, P. L., Murphy, R. F., Lanni, F. & Taylor, D. L. (1984) J. Cell Biol. 98, 1556-1564 Morris, J. D. M., Price, B. D., Lloyd, A., Marshall, C. J. & Hall, A. (1989) Oncogene 4, 27-31 Mulcahey, L. S., Smith, M. R. & Stacey, D. W. (1985) Nature 3 1 3 , 241-243 Nishizuka, Y. (1984) Nature 308,693-698 Nishizuka, Y.(1989) Nature 334,661-665 Pace-Asciak, C. R. & Smith, W. L. (1983) in The Enzymes (Boyer, P. D., ed) Vol. XVI, pp. 543-603, Academic Press, New York

TurnoverPtdCho Ras Stimulates

16643

Parries, G., Hoebel, R. & Racker, E. (1987) Proc. Natl. Acad. Sci. U. S. A . 84,2648-2652 Preiss, J., Loomis, C. R., Bishop, W. R., Stein, R., Niedel, J. E. & Bell, R. M. (1986) J. Biol. Chem. 261,8597-8600 Price, B. D., Morris, J. D. M., Marshall, C. J. & Hall, A. (1989) Biochem. J. 2 6 0 , 157-161 Rando, R. R. & Young, N. (1984) Biochem. Biophys. Res. Commun. 122,818-823 Rodriguez-Pena, A. & Rozengurt, E. (1984) Biochem. Biophys. Res. Commun. 1 2 0 , 1053-1059 Rozengurt, E., Stroobant, P.,Waterfield, M. D., Deuel, T. F. & Keehan, M. (1983) Cell 3 4 , 265-272 Sekiguchi, K., Tsukada, M., Ase, K., Kikkawa, U. & Nishizuka, Y. (1988) J. Biochem. (Tokyo) 103,759-765 Shearman, M. S., Naor, Z., Sekiguchi, K., Kishimoto, A. & Nishizuka, Y.(1989) FEBS Lett. 2 4 3 , 177-182

Shier, W. T. & Durkin, J. P. (1982) J . Cell. Physiol. 1 1 2 , 171-181 Takuwa, N., Takuwa, Y. & Rasmussen, H. (1987) Biochem. J. 2 4 3 , 647-653 Tarpley, W. G., Hopkins, N. K. & Gorman, R. R. (1986) Proc. Natl. Acud. Sci. U. S. A. 8 3 , 3703-3707 Trahey, M. & McCormick, F. (1987) Science 238,542-545 Wakelam, M. J. O., Davies, S. A., Houslay, M. D., McKay, I., Marshall, C. J. & Hall, A. (1986) Nature 3 2 3 , 173-176 Walker, J. R., Boot, J. R., Beverley, C. & Dawson, W. (1980) J. Pharm. Pharmucol. 32,866-871 Willingham, M. C., Banks-Schlegel, S. P. & Pastan, I. H. (1983) Exp. Cell. Res. 149, 141-149 Willumsen, B. M., Christensen, A., Hubbert, N. L., Papageorge, A. G . & Lowy, D. R. (1984) Nature 310,583-586 Wolfman, A. & Macara, I. G. (1987) Nature 325, 359-361 Yu, C.-L., Tsai, M-H. & Stacey, D. W. (1987) Cell 62,63-71