Specific inhibition of phorbol ester-stimulated phospholipase D by Clostridium sordellii lethal toxin and Clostridium difficile toxin B-1470 in HEK-293 cells. Restoration by Ral GTPases.

Activation of m3 muscarinic acetylcholine receptor (mAChR), stably expressed in human embryonic kidney (HEK)-293 cells, leads to phospholipase D (PLD) stimulation, a process apparently involving Rho GTPases, as shown by studies with Clostridium botulinum C3 exoenzyme and Clostridium difficile toxin B (TcdB). Direct activation of protein kinase C (PKC) by phorbol esters, such as phorbol 12-myristate 13-acetate (PMA), also induces PLD stimulation, which is additive to the mAChR action and which is only poorly sensitive to inactivation of Rho proteins by TcdB. To study whether Ras-like GTPases are involved in PLD regulation, we studied the effects of the TcdB variant TcdB-1470 and Clostridium sordellii lethal toxin (TcsL), known to inactivate Rac and some members of the Ras protein family, on PLD activities. TcdB-1470 and TcsL did not affect basal PLD activity and PLD stimulation by mAChR or direct G protein activation. In contrast, PMA-induced PLD stimulation was inhibited by TcdB-1470 and TcsL in a time- and concentration-dependent manner, without alteration in immunologically detectable PKC isozyme levels. In membranes of HEK-293 cells pretreated with TcdB-1470 or TcsL, basal and stable GTP analog-stimulated PLD activities measured with exogenous phosphatidylcholine, in the presence or absence of phosphatidylinositol 4,5-bisphosphate, were not altered. In contrast, pretreatment with TcdB-1470 and TcsL, but not TcdB, strongly reduced PMA-stimulated PLD activity. The addition of recombinant Rac1, serving as glucosylation substrate for TcdB, TcsL, and TcdB-1470, did not restore PLD stimulation by PMA. Furthermore, PMA-stimulated PLD activity, suppressed by prior treatment with TcdB-1470 or TcsL, was not rescued by the addition of recombinant Ras (RasG12V) or Rap proteins, acting as glucosylation substrates for TcsL only (Ras) or TcdB-1470 and TcsL (Rap). In contrast, the addition of recombinant Ral proteins (RalA and RalB), glucosylation substrates for TscL and TcdB-1470, but not for TcdB, to membranes of TcdB-1470- or TcsL-treated cells fully restored PLD stimulation by PMA without altering the strict MgATP dependence of PMA-induced PLD stimulation. RalA-mediated restoration of PMA-stimulated PLD activity in membranes of TcsL-treated cells was not enhanced by coaddition of RasG12V. In conclusion, the data presented indicate that TcdB-1470 and TcsL selectively interfere with phorbol ester stimulation of PLD and suggest an essential role of Ral proteins in PKC signaling to PLD in HEK-293 cells.

Ras and RalA, the latter acting apparently downstream of Ras (30). Finally, Ras itself has been reported to activate PLD in different cell types (31)(32)(33). Thus, GTPase cascades are apparently not only involved in cell proliferation and cytoskeleton organization, but they may also participate in PLD activation (34).
Another pathway for PLD stimulation involves isozymes of the protein kinase C (PKC) family, specifically members of the classical forms of PKC, as shown by studies in human neutrophils, in membranes of porcine brain, and with purified PLD (7,35,36). In various cellular systems, ARF and Rho proteins seem to be involved in PLD stimulation by PKC. Furthermore, activation of PLD by Ras proteins can apparently involve PKCdependent and PKC-independent pathways, depending on cell type and/or PLD isozyme (32). Finally, synergism between PLD stimulation by GTPases and phorbol ester-activated PKC has been demonstrated in several systems (35)(36)(37)(38)(39) and very recently also with purified PLD isozymes (7,9).
In human embryonic kidney (HEK)-293 cells, stably expressing the m3 muscarinic acetylcholine receptor (mAChR), various signal transduction components, including GTPases, PKC, tyrosine kinases, and phosphatidylinositol 4,5-bisphosphate (PtdIns(4,5)P 2 ), are apparently involved in PLD activation. We have recently shown that mAChR-mediated PLD stimulation in these cells involves small molecular weight GTPases of both the ARF and Rho protein families (17,27). As shown in studies with Clostridium difficile toxin B (TcdB), which inactivates Rho, Rac, and Cdc42, and with the selective Rho-inactivating Clostridium botulinum C3 exoenzyme, Rho proteins apparently regulate the supply of the essential PLD cofactor, PtdIns(4,5)P 2 (40,41). Direct activation of PKC by phorbol esters, such as phorbol 12-myristate 13-acetate (PMA), also induces marked PLD stimulation in HEK-293 cells, which is additive to the mAChR action and apparently does not involve ARF (17,42,43). Furthermore, PMA stimulation of PLD in HEK-293 cells was by far less sensitive to inactivation of Rho proteins by TcdB than mAChR stimulation (27). To examine whether Ras-like G proteins are involved in PLD stimulation in HEK-293 cells, we studied the effects of Clostridium sordellii lethal toxin (TcsL) and the variant TcdB, TcdB-1470, which induce similar cytopathic effects (44,45), on regulation of PLD activities. TcsL has recently been shown to inactivate Rac and some members of the Ras protein family (Ras, Rap, and Ral) by monoglucosylation (46 -48). We demonstrate here that TcsL and TcdB-1470, which glucosylates Rac, Rap, and Ral GTPases, potently and selectively inhibit PMA-induced PLD stimulation in HEK-293 cells and that this inhibition is selectively reversed by Ral proteins, suggesting an essential role for Ral GTPases in PKC signaling to PLD in HEK-293 cells.
Cell Culture-Culture conditions of HEK-293 cells stably expressing the m3 mAChR were as reported in detail before (27,40). For experiments, cells subcultured in Dulbecco's modified Eagle's medium/F-12 medium were grown to near confluence (175-and 75-cm 2 culture flasks or 145-mm culture dishes).
Toxin Treatment and Assay of PLD Activity in Intact Cells-For measurement of intact cell PLD activity, cellular phospholipids were labeled by incubating nearly confluent monolayers of cells for 20 -24 h with [ 3 H]oleic acid (2 Ci/ml) in growth medium. To exclude the possibility that the toxins interfere with the incorporation of radioactivity into phospholipids, the labeling medium was replaced before the cells were treated for 24 h or the indicated periods of time without and with TcsL or TcdB-1470 at the indicated concentrations. Thereafter, cells were detached from the culture flasks, resuspended twice in Hank's balanced salt solution, containing 118 mM NaCl, 5 mM KCl, 1 mM CaCl 2 , 1 mM MgCl 2 , and 5 mM D-glucose, buffered at pH 7.4 with 15 mM HEPES, at a cell concentration of 1 ϫ 10 7 cells/ml. The PLD activity assay was carried out for 60 min at 37°C in a total volume of 200 l containing 100 l of the cell suspension (1 ϫ 10 6 cells), 400 mM ethanol, and the indicated stimulatory agents. Stop of the reaction and isolation of labeled phospholipids and the specific PLD product, [ 3 H]phosphatidylethanol ([ 3 H]PtdEtOH), were as described before (27,40). The formation of [ 3 H]PtdEtOH is expressed as percentage of total labeled phospholipids.
Assay of PLD Activity in Permeabilized Cells-Prior to permeabilization, [ 3 H]oleic acid-prelabeled cells were treated for 24 h without and with TcdB-1470 at the indicated concentrations. Thereafter, the cells were detached from the culture flasks, washed twice by centrifugation and resuspension in PBS (137 mM NaCl, 2.7 mM KCl, 0.9 mM CaCl 2 , 0.5 mM MgCl 2 , 6.5 mM Na 2 HPO 4 , 1.5 mM KH 2 PO 4 , pH 7.2), and resuspended in assay buffer, containing 135 mM KCl, 5 mM NaHCO 3 , 5 mM EGTA, 4 mM MgCl 2 , 2 mM ATP, 1.5 mM CaCl 2 , 5.6 mM D-glucose, and 20 mM HEPES, pH 7.2, at a cell concentration of 1 ϫ 10 7 cells/ml. The assay of PLD activity was carried out for 60 min at 37°C in a total volume of 200 l, containing 100 l of the cell suspension (1 ϫ 10 6 cells), 10 M digitonin, and 400 mM ethanol as well as test agents (27).
Glucosylation of G Proteins-14 C glucosylation of recombinant GTPases or GTPases in HEK-293 cell lysates by TcdB, TcsL, or TcdB-1470 and analysis of 14 C-glucosylated proteins by SDS-PAGE and autoradiography were carried out as described before (46,48), with the following modifications. HEK-293 cells, grown to near confluence on 145-mm culture dishes, were treated without and with TcsL (100 ng/ml) or TcdB-1470 (300 pg/ml) for 24 h. Thereafter, the cells were detached from the culture dishes, washed once with ice-cold PBS, and resuspended in lysis buffer, containing 2 mM MgCl 2 , 0.1 mM phenylmethylsulfonyl fluoride, 10 g/ml leupeptin, 1 mM dithiothreitol, 5 mM GDP, and 50 mM HEPES, pH 7.5, followed by sonication on ice. HEK-293 cell lysates were prepared from the resulting supernatant by centrifugation for 10 min at 1000 ϫ g. The protein concentration was adjusted to 10 mg/ml. 14 C glucosylation was performed for 45 min at 37°C in a total volume of 30 l containing 30 M UDP-[ 14 C]glucose (100 nCi), 0.1 mM dithiothreitol, 5 mM GDP, 2 mM MgCl 2 , and 50 mM HEPES, pH 7.5, as well as TcdB (1 g/ml), TcdB-1470 (2 g/ml), or TcsL (2 g/ml). The reactions were started by the addition of 10 l of HEK-293 cell lysate (100 g of protein) or recombinant GTPases (2 g of protein). The reactions were terminated by the addition of SDS-PAGE sample buffer and 5-min incubation at 95°C. Thereafter, 14 C-glucosylated proteins were separated by SDS-PAGE on 12.5% (mass/volume) acrylamide gels and visualized by autoradiography, using a Kodak Biomax MS film in combination with Transcreen-LE (46,48).
Purification of Recombinant GTPases-The cDNAs of hRap1A, hRap2A, and hRap2B were inserted into the pGEX-2T vector (52). Fusion proteins containing an NH 2 -terminal glutathione S-transferase (GST) domain were induced in Escherichia coli by adding 1 mM isopropyl-1-thio-␤-D-galactopyranoside to the cultures. The bacteria were sonicated on ice in buffer A, containing 50 mM NaCl, 5 mM MgCl 2 , 1 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride, and 50 mM Tris-HCl, pH 7.5, and the crude membrane fractions were removed by centrifugation. The supernatant, containing GST-Rap1A, GST-Rap2A, and GST-Rap2B proteins, was incubated with glutathione-Sepharose beads for 30 min at 4°C. Thereafter, the beads were washed three times with buffer A to remove unbound proteins. Rap proteins were released from the parent GST fusion proteins bound to the beads by incubation with thrombin (10 units) overnight at 4°C in a buffer containing 150 mM NaCl, 5 mM MgCl 2 , 2.5 mM CaCl 2 , 1 mM dithiothreitol, and 50 mM Tris-HCl, pH 8. Thereafter, the beads were removed by centrifugation, and the excess of thrombin was removed by the addition of p-aminobenzamidine beads. The homogeneity of the recombinant Rap proteins was analyzed by Coomassie Blue staining of SDS-PAGE gels. The C-terminally truncated Rap1A used for the experiment shown in Fig.  8A was a generous gift of Dr. N. van den Berghe.
Similarly, hRac1 and the C-terminally truncated sRalA (amino acids 1-177) and hRalB (amino acids 1-178) were produced as GST fusion proteins in the protease-poor E. coli strain AD202 (53) and purified over a glutathione-Sepharose column. After cleavage by thrombin, the proteins were eluted, concentrated, and further purified by gel filtration. After concentration to approximately 10 mg/ml protein and dialysis against 50 mM Tris-HCl, 5 mM MgCl 2 , and 5 mM dithiothreitol, pH 7.6, the proteins were frozen in aliquots and stored at Ϫ80°C. Recombinant Ha-Ras (RasG12V) was kindly donated by Dorothee Vogt.
Immunoblot Analysis-For immunoblot analysis of PKC isozymes, cells were cultured in 75-cm 2 culture flasks to near confluence, before the addition of TcdB-1470 (300 pg/ml) for 24 h. Thereafter, the cells were detached from the culture flasks, pelleted by centrifugation, resuspended in PBS, and homogenized in a glass Teflon homogenizer. An aliquot of the homogenates and/or membranes (corresponding to about 50 g of protein) was subjected to SDS-PAGE on 10% (mass/volume) acrylamide gels for the separation of PKC isozymes. Proteins were transferred to nitrocellulose membranes and blotted with anti-c-PKC and anti-a-PKC antibodies (1:400 dilution, 1-h incubation). For detection of Ral and Ras proteins in HEK-293 cell membranes, 100 g of membrane proteins were subjected to SDS-PAGE on 15% (mass/volume) acrylamide gels, transferred to nitrocellulose membranes, and blotted with anti-RalA (1:5000 dilution, 1-h incubation), anti-RalB (1: 250 dilution, 1-h incubation), and anti-Ras (1:500 dilution, 1-h incubation) antibodies. For detection of hPLD1a and hPLD1b, we used an anti-hPLD1-specific peptide antibody (1:2000 dilution, 1-h incubation) (54).
Data Presentation-Data shown are mean Ϯ S.D. from one experiment performed in triplicate, repeated as indicated in the figure legends.
TcdB-1470 and C. sordellii TcsL have been reported to induce similar changes of cell morphology (tube-like cell shape with membrane blebbing), distinct from those induced by TcdB (44,45). Similar distinct cell morphology changes induced by TcdB at one hand and TcdB-1470 and TcsL on the other hand were observed in HEK-293 cells (data not shown). Therefore, we studied whether TcsL, known to inactivate Rac, Ras, Rap, and Ral proteins (46 -48), may induce similar differential inhibitory effects on mAChR-and phorbol esterstimulated PLD activities in HEK-293 cells as observed with TcdB-1470. Indeed, pretreatment of HEK-293 cells with TcsL had no effect on PLD stimulation by carbachol (data not shown) but caused a time-and concentration-dependent in-  (Fig. 3, left panel). When pretreated for 24 h, PMA-stimulated PLD activity was half-maximally reduced with 50 ng/ml TcsL, and 100 ng/ml TcsL inhibited PMA stimulation by 70% (Fig. 3, right panel). TcsL treatment had no effect on basal PLD activity.
To exclude the possibility that inhibition of PMA-induced PLD stimulation by TcdB-1470 and TcsL, causing drastic changes in cell morphology (44,45), was due to their effects on the cytoskeleton, cytochalasin B and C. botulinum C2 toxin were used as controls. Both agents, by distinct mechanisms, cause depolymerization of actin but notably without involving G proteins (55,56). Treatment of HEK-293 cells with cytochalasin B (5 g/ml, 15 min) and C2 toxin (20 ng/ml component I plus 40 ng/ml component II, 24 h) did not reduce PMA-stimulated [ 3 H]PtdEtOH accumulation, which amounted to 2.55 Ϯ 0.03% and 2.89 Ϯ 0.12% of total phospholipids, respectively, compared with 2.71 Ϯ 0.23% in control cells.
To further dissect m3 mAChR and PKC signaling to PLD, we studied the effects of TcdB-1470 and TcsL on PLD activation by direct G protein stimulation. In control cells, direct activation of heterotrimeric G proteins by AlF 4 Ϫ increased [ 3 H]PtdEtOH accumulation to a similar extent as mAChR activation by carbachol (1 mM). Pretreatment of the cells with TcdB-1470 (300 pg/ml, 24 h), which reduced PMA (0.1 M)-induced PLD activation by about 65%, did not alter PLD stimulation by either carbachol or AlF 4 Ϫ (Fig. 4A, left). Likewise, pretreatment of the cells with TcsL (100 ng/ml, 24 h) reduced PLD stimulation by PMA (0.1 M) by about 70%, without changing PLD activation by carbachol or AlF 4 Ϫ (Fig. 4A, right). In permeabilized HEK-293 cells, PLD activity can be stimulated by the stable GTP analog GTP␥S and PMA in an additive manner (40). As illustrated in Fig. 4B, in control digitonin-permeabilized cells pre- To study whether TcdB-1470 treatment may alter PLD per se, we measured PLD activities in membranes of control and TcdB-1470-treated HEK-293 cells with exogenous PtdCho as enzyme substrate, using two distinct assay conditions. TcdB-1470 treatment (300 pg/ml, 24 h) had no effect on basal PLD activity, measured with [ 3 H]PtdCho/sodium oleate/sodium cholate vesicles, either in the absence or in the presence of the stimulating PLD cofactor, PtdIns(4,5)P 2 (Fig. 5, left). Furthermore, basal and GTP␥S-stimulated PLD activities, measured with [ 3 H]PtdCho/PtdIns(4,5)P 2 vesicles, were not affected by prior treatment of HEK-293 cells with TcdB-1470 (Fig. 5, right  panel). In contrast, stimulation of PLD activity by PMA (0.1 M) was markedly reduced, by about 60%, in membranes of HEK-293 cells pretreated with TcdB-1470.
To examine whether the inhibition of PMA-stimulated PLD activity was caused by potential degradation of PKC isozymes, the levels of c-PKC (␣, ␤, ␥) and a-PKC (, ) isoforms were measured in membranes of control and TcdB-1470-treated HEK-293 cells with anti-PKC antibodies. A typical blot of three similar experiments performed in HEK-293 cell membranes is illustrated in Fig. 7, which indicates that the levels of the PKC isozymes were not different between control and TcdB-1470treated cells. Similar data were obtained in cell homogenates (data not shown).
Since TcsL and TcdB-1470 caused similar alterations in PLD stimulation, quite distinct from those induced by TcdB, it was tempting to speculate that TcsL and TcdB-1470 modify and inactivate a similar set of GTPases. To study this, we determined 14 C glucosylation of recombinant GTPases and GTPases in lysates of HEK-293 cells by TcdB, TcsL, and TcdB-1470. As shown in Fig. 8A, recombinant Rac1 served as glucosylation substrate for TcdB, TcsL, and TcdB-1470. In contrast, recombinant Rap and Ral proteins acted as substrates for TcsL and TcdB-1470 only, and Ras was exclusively glucosylated by TcsL. RhoA, which is a glucosylation substrate for TcdB (57) range. Pretreatment of intact HEK-293 cells with TcsL (100 ng/ml, 24 h) or TcdB-1470 (300 pg/ml, 24 h) reduced the subsequent 14 C glucosylation catalyzed by either toxin by at least 50%. Pretreatment of HEK-293 cells with TcsL and TcdB-1470 did not reduce the levels of immunologically detectable Ral and Ras proteins (data not shown).
Considering the GTPase specificities of TcdB, TcsL, and TcdB-1470 (see Refs. 46 -48 and Fig. 8A) on one hand and the distinct inhibitory patterns of these toxins on mAChR-and PMA-induced PLD stimulation on the other hand, it was tempting to speculate that a specific subset of GTPases participates in the PMA-induced PLD activation. Therefore, we studied the effects of various purified recombinant GTPases on PLD activity in membranes of toxin-pretreated HEK-293 cells. Rac proteins are common substrates of TcdB, TcsL, and TcdB-1470, suggesting that Rac proteins may not be essential for PKC signaling to PLD in HEK-293 cells. Indeed, as shown in Fig. 9, the addition of purified recombinant Rac1 had no effect on basal and PMA-stimulated PLD activities in control membranes and did not restore the TcdB-1470-inhibited PLD stimulation by PMA.
Since Ras proteins are substrates for TcsL, but not TcdB-1470, it was likely that the addition of Ras proteins might restore PLD stimulation by PMA in membranes of cells pretreated with TcsL, while being ineffective in TcdB-1470-treated membranes. Indeed, the addition of recombinant RasG12V, a constitutively active Ras mutant, to membranes of HEK-293 cells pretreated with TcdB-1470 (300 pg/ml, 24 h) did not restore PLD stimulation by PMA (Fig. 10, left panel). However, the addition of recombinant RasG12V also had no effect on PMA-induced PLD stimulation in membranes of cells pretreated with TcsL (100 ng/ml, 24 h) (Fig. 10, right panel). Similar to Rac1, the addition of RasG12V did not alter basal and PMA-stimulated PLD activities in membranes of control untreated cells. Thus, Rac and Ras proteins seem not to be crucial for stimulation of PLD by PKC in HEK-293 cells.
Since Rap and Ral proteins are common targets of TcsL and TcdB-1470, but not TcdB, their participation in the PMA-induced PLD activation was rather likely. However, as shown in Fig. 11, the addition of different Rap proteins (Rap1A, Rap2A, and Rap2B) did not restore the TcdB-1470-inhibited PLD stimulation by PMA. The addition of Ral proteins, which were detected by immunoblotting in the membrane preparations (data not shown), had no effects on basal and GTP␥S-or PMAstimulated PLD activities in control membranes (data not shown). Furthermore, as illustrated in Fig. 12A fects, whereas heat-inactivated Ral proteins were inactive (data not shown).
As recently reported (43), PLD stimulation by PMA in membranes of HEK-293 cells is absolutely dependent on the presence of MgATP (Fig. 12B). Half-maximal and maximal stimulation of PLD activity by PMA (0.1 M) was observed at about 50 M and 1 mM MgATP, respectively. In membranes of TcdB-1470-treated HEK-293 cells, the remaining PMA-stimulated PLD activity (about 30%) also exhibited a strict MgATP dependence, with half-maximal and maximal stimulation by PMA (0.1 M) occurring at about 10 M and 100 M MgATP, respectively. The addition of recombinant RalA fully restored the TcdB-1470-inhibited PLD stimulation by PMA, exhibiting an identical MgATP dependence as in control untreated membranes (Fig. 12B).
Finally, since TcsL inactivates Ras and Ral proteins and evidence has been provided that Ras/Ral signaling cascades can be involved in PLD stimulation (30,58,59), we examined whether Ral proteins alone can restore PLD stimulation by PMA in membranes of TcsL-treated cells or whether a combination of Ras and Ral proteins is required. As shown above (see Fig. 10), the addition of RasG12V alone to membranes of TcsLtreated HEK-293 cells had no effect on the TcsL-inhibited PMA-stimulated PLD activity (Fig. 13). In contrast, the addition of RalA fully restored PLD stimulation by PMA. The coaddition of RasG12V did not affect basal PLD activity; nor was the PMA-induced PLD stimulation restored by RalA further enhanced. An additive or synergistic effect of RasG12V addition also was not observed when RalA was added at a halfmaximally effective concentration (data not shown). DISCUSSION In the present study, we have investigated the involvement of Ras-like GTPases in signal transduction pathways leading to PLD stimulation in HEK-293 cells stably expressing the m3 mAChR. Previous studies indicated that at least two distinct signaling pathways evoke PLD activation in HEK-293 cells (42). PLD stimulation by the agonist-activated m3 mAChR is dependent on G proteins of the ARF and Rho protein family (17,27). While ARF proteins may directly stimulate PLD activity as recently reported for purified PLD1, by interacting with specific PLD domains (8,9), Rho proteins appear to participate indirectly in receptor-mediated PLD activation in HEK-293 cells, i.e. by controlling the supply of the stimulatory PLD cofactor, PtdIns(4,5)P 2 (17,27,40,41). In line with this notion, specific binding of partially purified rat brain PLD to PtdIns(4,5)P 2 has recently been described (60) as well as direct activation of purified recombinant PLD isozymes (hPLD1, mPLD2, and rPLD2) by PtdIns(4,5)P 2 (8,10,11). In contrast to mAChR activation, PLD stimulation in HEK-293 cells by phorbol ester-activated PKC is apparently independent of ARF and Rho proteins (17,27,43). This may reflect the existence of multiple signaling pathways, activating the lipase through dif- ferent signal transduction components, and/or the existence of different PLD isozymes, being under control of distinct cellular components. Indeed, the existence of different PLD isozymes in one cell type has recently been reported (61,62).
The role of different GTPases, specifically of members of the Rho and Ras protein family, in receptor and PKC signaling to PLD in HEK-293 cells was studied with the use of different large clostridial cytotoxins, belonging to a family of glucosyltransferases that monoglucosylate a specific subset of small GTP-binding proteins (46 -48, 57). Treatment of HEK-293 cells with C. difficile TcdB, which glucosylates Rho proteins (Rho, Rac, and Cdc42) (57), led to a specific inhibition of receptorstimulated PLD activity, with only minimal effects on PMAinduced PLD stimulation (27,40). Recently, TcsL, produced by C. sordellii strains 82 and 1522, has been shown to inactivate, by monoglucosylation, a different set of small G proteins, namely Rac, Ras, Rap, and Ral proteins, but not Rho and Cdc42, which are TcdB targets (46 -48). Since the variant of C. difficile TcdB, TcdB-1470, induces a TcsL-like cell morphology distinct from that induced by TcdB (44,45), it was tempting to speculate that TcdB-1470 may evoke its cellular effects due to the inactivation of those GTPases that are targets of TcsL. Indeed, as reported herein, TcdB-1470 glucosylates a similar set of small G proteins as TcsL, namely Rac, Rap, and Ral proteins, with the exception of Ras, which is not a target of TcdB-1470.
In the present study, we provide evidence that Ras-like GTPases, most likely Ral proteins, are involved in PKC signaling to PLD in HEK-293 cells. First, treatment of HEK-293 cells with TcsL and TcdB-1470 potently and specifically inhibited PLD stimulation by the phorbol ester PMA. This inhibition was observed both in intact and permeabilized cells where endogenous PtdCho was used as PLD substrate as well as in membrane preparations where an exogenous PLD substrate was added. It should be noted that TcdB-1470 and TcsL did not cause a complete (100%) loss of PMAinduced PLD stimulation (see e.g. Figs. 1-3), suggesting that a small part of PMA-stimulated PLD activity is independent of the G proteins being substrates for TcdB-1470 and TcsL. As recently reported with membrane-bound, partially purified, and purified PLD enzymes, PLD can be directly stimulated by purified phorbol ester-activated PKC-␣, which is apparently independent of the enzymatic kinase activity (8,9,36,39). In membranes of HEK-293 cells, PMA-induced PLD stimulation required the presence of MgATP (see Ref. 43 and Fig. 12), suggesting an enzymatic mechanism of action as also reported for PLD stimulation by PKC isozymes (PKC-␣, PKC-␤ 1 , and PKC-␥) in membranes of human neutrophils (35). Thus, the small part of PMA-stimulated PLD activity that was resistant to treatment of HEK-293 cells with TcdB-1470 and TcsL may be due to a direct interaction of a PKC isozyme with PLD, although it must be noted that the remaining PMA-stimulated PLD activity observed in membranes of TcdB-1470-treated cells was also MgATP-dependent. In contrast to PLD stimulation by PMA, stimulation of PLD activity by the m3 mAChR and direct activation of G proteins in intact or permeabilized HEK-293 cells as well as membrane preparations was not or was only marginally reduced by either TcdB-1470 or TcsL, and basal activity was not affected at all. On the other hand, mAChR-mediated PLD stimulation was potently inhibited by TcdB (27), which did not affect PMA-stimulated PLD activity. Second, as studied with TcdB-1470 and/or TcsL, the reduction in PMA-induced PLD stimulation cannot be attributed to a loss or altered activity of individual components potentially involved in PLD stimulation. Specifically, TcdB-1470 treatment did not reduce levels of c-and a-PKC isozymes. Furthermore, the MgATP dependence of PMA-induced PLD stimulation in membranes of TcdB-1470-treated HEK-293 cells reconstituted with RalA was identical to that of control membranes, indicating that the enzymatic mechanism of action of PKC isozyme(s) involved in PMA-stimulated PLD activity in HEK-293 cells was not altered by TcdB-1470. Third, in contrast to TcdB, which reduced cellular PtdIns(4,5)P 2 levels and basal PLD activity measured in the absence of PtdIns(4,5)P 2 (40,41), treatment of HEK-293 cells with TcdB-1470 and TcsL had only a slight effect on cellular PtdIns(4,5)P 2 level and did not reduce basal PLD activity measured in the absence of PtdIns(4,5)P 2 . Fourth, direct measurement of TcdB-1470and TcsL-catalyzed glucosylation of recombinant GTPases and GTPases endogenously expressed in HEK-293 cells suggested that Ral proteins are glucosylation substrates of either toxin. Finally, and most importantly, the addition of purified recombinant Ral proteins (RalA and RalB) fully restored PMA-induced PLD stimulation in membranes of HEK-293 cells pretreated with either TcsL or TcdB-1470. On the other hand, purified recombinant Rap proteins, which are also substrates of TcsL and TcdB-1470, were inactive in this respect. Furthermore, recombinant Rac1, serving as glucosylation substrate for TcdB, TcsL, and TcdB-1470, did not restore TcdB-1470-inhibited PMA-induced PLD stimulation. Interestingly, recombinant Ras (RasG12V), serving as glucosylation substrate for TcsL, but not TcdB-1470, did not rescue PMA-induced PLD stimulation, either in membranes of HEK-293 cells treated with TcdB-1470 or in those treated with TcsL. The small GTPases used in the reconstitution experiments were expressed in and purified from E. coli; thus, they were not post-translationally modified. In addition, to increase the purification yield, the Ral proteins were C-terminally truncated (28 and 27 amino acids). Since posttranslational modifications of Ras-like GTPases increase the signal transduction efficiency of these GTPases (63,64), this may explain why rather high concentrations of Ral proteins, in the 10 -30 M range, were required for restoration of PMA-stimulated PLD activity. Altogether, strong evidence is provided that phorbol ester-induced PLD stimulation in HEK-293 cells, which apparently involves the enzymatic activity of a PKC isozyme(s), is not direct, but involves Ras-like GTPases, most likely Ral proteins. As mentioned above, m3 mAChR-mediated PLD stimulation in HEK-293 cells is independent of phorbol ester-induced and Ral-mediated PLD stimulation. Since m3 mAChR activation causes marked phospholipase C activation in HEK-293 cells (41,42), it is likely that the PKC isozyme(s) responsible for PMA-stimulated PLD activity is different from the PKC isozyme(s) activated by the m3 mAChR, subsequent to the phospholipase C stimulation, or is present in a distinct cellular compartment.
Recently, a participation of Ral proteins in PLD stimulation in Balb/c and NIH 3T3 fibroblasts has been reported (30). However, in contrast to the data presented in this report in HEK-293 cells, the Ral-mediated PLD stimulation in those cells was caused by v-Src and was apparently not PKC-mediated. The v-Src-induced activation of PLD in Balb/c and NIH 3T3 fibroblasts required, in addition to Ral proteins, activated Ras and the Ral guanine nucleotide dissociation-stimulating factor, Ral-GDS. Furthermore, the Ral-mediated PLD activation was apparently independent of the GTP/GDP-binding state and the effector domain of RalA but was attributed to the unique NH 2 terminus of this G protein (30). Similar to that observed in Balb/c and NIH 3T3 fibroblasts (30), PMA-stimu-lated PLD activity in HEK-293 cells seems to be largely independent of the GTP/GDP binding state of Ral proteins but may not involve active Ras. First, full restoration of PMA-stimulated PLD activity in membranes of HEK-293 cells treated with TcsL or TcdB-1470 by Ral proteins was observed without adding GTP or GTP␥S. Second, PLD stimulation by PMA and GTP␥S in HEK-293 cells was simply additive (42), and the latter one was not affected by inactivation of Ras-like GTPases. Third, the addition of recombinant RasG12V, a constitutively active Ras mutant, to membranes of HEK-293 cells treated or not treated with TcdB-1470 or TcsL did not alter basal and GTP␥S-or PMA-stimulated PLD activities. Finally, the simultaneous addition of RasG12V and RalA had no additive or synergistic effect on TcsL-inhibited PMA-induced PLD stimulation. However, although the reconstitution experiments suggest that Ras is not essential for PKC-induced PLD stimulation mediated by Ral proteins, the possibility of a signaling cascade involving Ras, Ral-GDS, and Ral proteins in PKC-induced PLD stimulation in intact HEK-293 cells cannot be excluded and is presently under investigation. Recently, direct interaction of RalA with the ARF-responsive PtdIns(4,5)P 2 -dependent recombinant PLD1 isozyme has been demonstrated (65). HEK-293 cells apparently express hPLD1, as shown by immunoblot studies with anti-hPLD1-specific peptide antibodies (54) in membrane preparations (data not shown). Thus, it remains to be studied whether Ral proteins interact directly with hPLD1 in HEK-293 cells as well and whether this process is under the control of phorbol ester-activated PKC.
In conclusion, we demonstrate that C. sordellii TcsL and C. difficile TcdB-1470, which inactivate Ras-like GTPases, potently and specifically inhibit PLD stimulation by phorbol ester-activated PKC in HEK-293 cells. By comparing the effects of different large clostridial cytotoxins, namely TcdB, TcsL, and TcdB-1470, which glucosylate and thereby inactivate distinct members of the GTPase superfamily and, most importantly, by reconstitution with various purified GT-Pases, we provide evidence that Ral proteins are essentially required for PKC signaling to PLD in HEK-293 cells. In contrast, receptor stimulation of PLD activity, which was highly sensitive to inactivation of Rho proteins by TcdB, was resistant to inactivation of Ras-like GTPases by TcdB-1470 and TcsL, further strengthening the possibility that distinct signal transduction pathways leading to stimulation of PLD activity exist in HEK-293 cells.