ADP-ribosylation of rho p21 inhibits lysophosphatidic acid-induced protein tyrosine phosphorylation and phosphatidylinositol 3-kinase activation in cultured Swiss 3T3 cells.

Botulinum C3 exoenzyme was used to specifically ADP-ribosylate and inactivate rho p21, and the effects of rho p21 inactivation on lysophosphatidic acid (LPA)-induced tyrosine phosphorylation were examined in cultured Swiss 3T3 cells. LPA induced a rapid increase in the tyrosine phosphorylation of a number of proteins. Pretreatment of the cells with the C3 exoenzyme caused ADP-ribosylation of rho p21 in the cells and selectively attenuated the phosphorylation of several proteins, including p43 mitogen-activated protein kinase, p125 focal adhesion kinase, and two proteins of 72 and 88 kDa. C3 exoenzyme pretreatment did not block the initial phosphorylation and activation of mitogen-activated protein kinase but suppressed its subsequent rise. In contrast, the enzyme treatment inhibited the induction of phosphorylation of the 72- and 88-kDa proteins and suppressed the basal and LPA-induced tyrosine phosphorylation of p125 focal adhesion kinase. In addition, immunoprecipitation of cell lysates with an antibody directed against the 85-kDa subunit of phosphatidylinositol 3-kinase (PI 3-kinase) co-precipitated a tyrosine-phosphorylated band of 180 kDa. C3 exoenzyme pretreatment suppressed both the phosphorylation of this band and PI 3-kinase activation associated with LPA stimulation. These findings suggest that rho p21 works as a link between the LPA receptor signal and the subsequent tyrosine phosphorylation and PI 3-kinase activation in these cells.

* This work was supported in part by Grants-in-aid for Scientific Research 04253213, 04255103, 05271103, 05404020, and 05670088 from the Ministry of Education, Science and Culture of Japan and by grants from the Mitsubishi Foundation, the Senri Life Science Foundation, and the Japanese Foundation on Metabolism and Diseases. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisemenl" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. chida, Tokyo 194, Japan. The rho gene products (rho ~2 1 s ) ' are members of the ras superfamily of small GTPases (1)(2)(3). Following receptor stimulation, these GTPases are converted from the inactive GDPbound form to the active GTP-bound form and link external stimuli to cellular responses such as growth, differentiation and secretion. rho p21s are unique among the ras-related GTPases, since they are substrates for botulinum C3 ADPribosyltransferase (3)(4)(5). This enzyme specifically modifies the Asn41 residue of rho p21s, and inhibits their biological activity, presumably by interfering with their interaction with downstream targets (6). Using this exoenzyme andor mutants of rho p21, the cellular functions of the rho p21s have been investigated. These studies have revealed that rho p21 mediates such cellular processes as stimulus-evoked cell adhesion (7)(8)(9) and motility (lo), regulation of smooth muscle contraction ( l l ) , G I to S phase progression in cell cycle (121, and cytokinesis of the fertilized egg (13,14). LPA acts on a cell surface receptor coupling to G-proteins, Gi and G,, and evokes a variety of biological responses (15, 16). It is a normal constituent of serum and responsible for the majority of its growth promoting activity (7,17). Ridley and Hall (7) reported that LPA characteristically induces focal adhesion and stress fibers in quiescent Swiss 3T3 cells, and that this response was mediated by rho p21. These findings suggest that rho p21 receives a signal from LPA in the cell and transduces it to a system regulating cell adhesion. However, little is known about this transduction mechanism.
We recently found that LPA rapidly induced the tyrosine phosphorylation of a number of proteins in Swiss 3T3 cells, including the MAP kinases and p125 FAK (18). I n the present study, we used botulinum C3 exoenzyme to examine the involvement of rho p21 in this process.
Preparation of Recombinant C3 Exoenzyme"C3 exoenzyme gene (5) was modified by a PCR-mediated site directed mutagenesis to produce a recombinant C3 exoenzyme that lacks the signal peptide and has dipeptide Met-Ala attached to Ser' of the mature exoenzyme. PCR was tides (5'-ACTGTTCATATGGCTAGCTATGCAGATACT"C-ACA-3' and performed with the cloned gene as template and synthetic oligonucleo-5'-TTATTGGATCCTATTA"AATATCATTGCTGTA3') as primers. The amplified fragment was cleaved with NdeI and BamHI and ligated with a PET-3a vector (21). M e r confirming DNA sequence, the recombinant plasmid, pET-3dC3, was introduced into Escherichia coli BL2UDE3)pLysE and expressed (5,21,22).

24535
( N l ('nltrrrv nnd I J ' A Stinrrrlntion-Swiss :lT3 mouse fihrohlasts w w r maintainrd and culturrd a s drscrihrd prrviously (12. 181. T h r crlls wvrr s w d r d :It a drnsity of 1.0 x 1078/wrll In six-wrll culture plates IFaIcon) and culturrd in cornplrtc~ rnrdiurn for :3 days. Thc. crlls were thtm culturcd with or without 5 pg/rnl rrcomhinant C 3 exornzyrnr. first for 4 X h in t.hr complc.tc. mcdium and thrn fnr 12 h in srrurn-free mcdium (:I 1 : l rnixturt-of Ihlhrcco's modifird E;~glr's mrdiurn and Ham's F-12 mrdium containing 5 mg/ml hovinc. s w u m alhumin and 0.05'; insulin-trans~.rrin-s(,l(,nitr s u p p l r r n r n t ) . 1,PA a t 0.2 p~ was t h e n nddrd t o t h r culturr. T h r cctlls wrrv incuh;ltrd ;It 37 ' C in n ( X ) : ! incuhator for thr indicatrd tinws and cxtractrd (]HI. I.l'A-inducrd inhihition of cAM1' zlccumulntion was studird by incuhnting the crlls wlth 3 p~ fnrskolin for 15 mln i n t h r prcwncca of 1 m\t isohutylmethylx;lnthine z~ddcd I O rnin prior to t h r forskolin addition. LPA a t 2 p~ w a s ;~ddcd at 5 . IO. a n d 14 min aftrr thv addition of forskolin. The rwction was stopprd t y ;Idding 10"; iccvold trichloroaccbtic acid. and CAMP content was drtrrrninrd using lL"I-cAMP assay svstrm from Amrrshnm Corp. and, consrquentlv, thr inactivation of rho p21 did not affrct the initial signal transduction of LPA, b r c a u s r 1,PA-mrdi;ttrd inhibition of CAMP grntvxtion occurrrd cqually in control and C3 rxornzymc-trc.;ttrd crlls (data not shown). Using these crlls, we exnmincd thcrffrcts ofthis treatmrnt on protein tyrosine phosphorylation. A s shown in Fig. lR, LPA induced a rapid tyrosine phosphorylntion of multiplc protrins. and thr C.7 rxornzyme t r e a t m r n t a t t r n u a t c d s o m r of thr phosphorylation. For rxamplr, phosphorylation of a protrin drsignated p72 occurred rapidly at 1 min aftrr 1,PA addition and rrmained high during thr rntirc 60-min incuhntion. This rrsponse was supprcssrd by t h r C3 rxocwzymr prrtrratmrnt. A similar supprrssion was ohsrrvrd for p88. On thr other hand, C3 cxoenzymc prctrratmrnt did not nffrct thr phosphorylation of a protein dcsignated p64. Wr previously idrntifird thr tyrosinc.-phosphorylat~,d protrin p43 as t h r MAP kinase, ERK-2 ( ] X ) . n r o s i n r phosphorylation of this protrin was drtrctcd at 1 min after LPA addition. incrcasrd to ;I maximum a t 5 min, and declined thereafter. Thc, C 3 prrtreatmrnt did not affect thc initial phosphorylation ohsrrvrd at 1 min (/nnc~.s 2 ard 2' ), hut significantly attrnuated phosphorylation by 5 min (lanvs .7 a n d 3 ' ) . To confirm this rffvrt, wr mc:tsurcd MAP kinase activity using myrlin basic protoin as a suhstr:rt(.. As shown in Fig. 2, MAP kinase was rapidly and tr;rnsic,ntly activated hy LPA addition; thc activity increased about 3-fold at 1 a n d 5 min hut declined quickly. In t h e C3 exoenzyme-treated cells, activation was also observed at 1 min. hut the activity had alrrady dccrcased by 5 min. These

200-
results a r r consistrnt with thc findings on tyrosinr phnsphorylation in Fig. 1R. In ordrr to identify pXX. immunoprc~ripitntion of t h r cell extracts from both control and ('3 rxcwnzvmr-treated cells was camed out using the CA3 monoclonal antibody specific for the p85 subunit of PI 3-kinase (19). Although this procedure precipitated the immunoreactive p85 protein, no precipitation of p88 was observed (Fig. 3A ). Instead, a tyrosinephosphorylated protein of 180 kDa, p180, was precipitated. This precipitation appeared specific, because no precipitation was found without anti-p85 antibody or with other antibodies such as anti-pl25 FAK (data not shown and Fig. 4). Since its phosphorylation was suppressed by C3 exoenzyme pretreatment (Fig. 3A ), we examined the PI 3-kinase activity associated with the immunoprecipitates. As shown in Fig. 3B, the activity of PI 3-kinase as determined by the 32P-phosphorylation of PI 4-monophosphate and PI 4,5-diphosphate was elevated at 5 min after LPA addition, and this activation was suppressed by C3 exoenzyme pretreatment.
The immunoblot shown in Fig. 3A showed that electrophoresis on an 8% polyacrylamide gel separated p72 into at least three proteins, tyrosine phosphorylation of all of which was significantly suppressed by C3 exoenzyme treatment. It also revealed that the enzyme treatment suppressed the phosphorylation of a group of proteins between 110 and 130 kDa, pll0-  control (lanes 1-3) and C3 exoenzyme-treated cells (lanes 4 4 ) were stimulated with 0.2 p~ LPA for 0 (lanes Z and 4). 1 (lanes 2 and 5 ) , and 5 min (lanes 3 and 6 ) . The cell extracts prepared were subjected to immunoprecipitation with a CA3 anti-p85 mouse monoclonal antibody. The immunoprecipitates were subjected to SDS-PAGE and immunoblotting with an anti-phosphotyrosine antibody. Lanes 1 ' 4 ' show the immunoblots for total cell lysates corresponding to lanes Z-6. B , PI 3-kinase activity in immunoprecipitate from control (open columns) and C3 exoenzyme-treated cells (hatched columns) stimulated with 0.2 p~ LPA for the indicated times was determined as described under "Experimental Procedures." The inset shows autoradiogram of thin layer chromatography of PI 3-kinase assay on the samples prepared from the control (lanes I , 3, and 5 ) and C3 exoenzyme- treated cells (lanes 2 , 4 , and 6 ) incubated with LPA for 0 (lanes I and 2) , 5 (lanes 3 and 41, and 10 min (lanes 5 and 6 ) . Experiments were repeated twice with essentially identical results.  Control (lanes I , 3. and 5 ) and C3 ex* enzyme-treated (lanes 2 . 4 , and 6 ) cells were incubated with 2 PSI LPA at 37 "C for 0 (lanes 1 and 2 ), 5 (lanes 3 and 4 ), and 10 (lanes $5 and 6 ) min. The extracts from these cells were then subjected to immunoprecipitation with an anti-p125 antibody and to immunoblotting with the anti-phosphotyrosine antibody as described under 'Experimental Procedures." A typical result of three independent experiments is shown.
p130. We previously identified one of these proteins as p125 FAK (18). To examine if the tyrosine phosphorylation and activation of this kinase is affected by the C3 exoenzyme treatment, immunoprecipitation of p125 FAK from the control and C3 exoenzyme-treated cells was performed and the extent of tyrosine phosphorylation was compared. As shown in Fig. 4, tyrosine phosphorylation of p125 was observed already before the LPA addition but increased at 5 and 10 min after the addition. C3 exoenzyme treatment suppressed both basal and LPA-induced phosphorylation.
The present study demonstrates that C3 exoenzyme pretreatment and consequently inactivation of rho p21 by ADPribosylation inhibited some of the LPA-induced tyrosine phosphorylation of the proteins. Because the initial transduction of LPA signaling as determined as inhibition of adenylate cyclase was not inhibited by C3 exoenzyme treatment, rho p21 most probably works downstream of second messenger generation. Recent studies have revealed that many agonists acting on G-protein-coupled receptors can induce tyrosine phosphorylation in Swiss 3T3 cells (26,27) and activate some enzymes activated by tyrosine phosphorylation such as PI 3-kinase (28).
The present study indicates that the activation of rho p21 may be a link between the stimulation of such receptors and some of the intracellular protein tyrosine phosphorylation events. It also suggests that the activation of PI 3-kinase by LPA is regulated a t least in part by a rho p21-dependent pathway. One contentious point is the relationship of these tyrosine phosphorylations to the site of rho p21 action. m o s i n e phosphorylation of p125 FAK is induced similarly by several agonists (29) and by cell adhesion to the substrate matrix (30)(31)(32). Since the activation of rho p21 leads to integrin activation and cell adhesion (7-91, it is conceivable that the observed phosphorylations are secondary to the LPA-induced cell adhesion. However, cytochalasin B treatment did not affect some of the phosphorylations (e.g. the phosphorylation of MAP kinase), although it caused cell rounding and suppressed the phosphorylation of p125 FAK.2 In addition, C3 exoenzyme treatment suppressed the tyrosine phosphorylation of other proteins such as p72 and p88, which responded much faster to LPA than p125. It is therefore likely that rho p21 activates a tyrosine kinase cascade in these cells which eventually leads to cell adhesion. Recently, Volberg et ai. (33) reported that the treatment of cultured Madin-Darby canine kidney cells with specific tyrosine phosphatase inhibitors induces focal adhesion and stress fiber formation, suggesting involvement of tyrosine phosphorylation in the induction of cell adhesion. ' b o modes of action for the ras-related small GTPases are now coming to focus. One is kinase regulation exemplified by m s p21, which complexes with c-raf and/or MAP kinase kinase kinase and regulates their activity (34). Recently Manser et ai. (35) reported that CDC42 p21 can complex with a nonreceptor tym sine kinase in a GTP-dependent fashion. The other is the regulation of translocation (36) shown by the rub family of GTPases, N. Kumagai, N. Morii, and S. Narumiya, unpublished observation.

_ _ -.
which promote unidirectional vesicular transport between the two compartments. Thus far rho p21 action has been analyzed by following changes in the cell phenotype, and there is no definitive evidence as to the mode of action of rho p21. This is the first study showing that rho p21 can change functional parameters in the cell, which suggests that rho p21 can directly regulate an intracellular signaling molecules such as a tyrosine kinase.