Regulation of phospholipase D in L6 skeletal muscle myoblasts. Role of protein kinase c and relationship to protein synthesis.

The addition of vasopressin or 12-O-tetradecanoylphorbol-13-acetate (TPA) to prelabeled L6 myoblasts elicited increases in [14C]ethanolamine release, suggesting the activation of phospholipase D activity or activities. While the effects of both agonists on intracellular release were rapid and transient, when extracellular release of [14C]ethanolamine was measured, the effect of vasopressin was again rapid and transient, whereas that of TPA was delayed but sustained. Effects of both agonists on intra- and extracellular release were inhibited by the protein kinase C (PKC) inhibitor, Ro-31-8220, and PKC down-regulation by preincubation with TPA. The formation of phosphatidylbutanol elicited by vasopressin and TPA mirrored their effects on extracellular [14C]ethanolamine release in that the former was transient, whereas the latter was sustained. Responses to both agonists were abolished by PKC down-regulation. When protein synthesis was examined, the stimulation of translation by TPA and transcription by vasopressin were inhibited by Ro-31-8220. In contrast, down-regulation of PKC inhibited the synthesis response to TPA but not vasopressin. Furthermore, following down-regulation, the effect of vasopressin was still blocked by the PKC inhibitors, Ro-31-8220 and bisindolylmaleimide. Analysis of PKC isoforms in L6 cells showed the presence of alpha, epsilon, delta, mu, iota, and zeta. Down-regulation removed both cytosolic (alpha) and membrane-bound (epsilon and delta) isoforms. Thus, the elevation of phospholipase D activity or activities induced by both TPA and vasopressin and the stimulation of translation by TPA involves PKC-alpha, -epsilon, and/or -delta. In contrast, the increase in transcription elicited by vasopressin involves mu, iota, and/or zeta. Hence, although phospholipase D may be linked to increases in translation elicited by TPA, it is not involved in the stimulation of transcription by vasopressin.

Loss of skeletal muscle is an acute metabolic response to infection and neoplastic disease and results from a decrease in the rate of protein synthesis and an increase in the rate of protein degradation (e.g. see Refs. [1][2][3]. To reverse this process, an understanding of the signaling pathways regulating protein turnover is essential. We have used 12-O-tetradecanoylphorbol-13-acetate (TPA) 1 and vasopressin to investigate the coordinate regulation of protein turnover. Previous studies in this (4 -6) and other (7) laboratories have shown that these two agents both stimulate protein synthesis and reduce the release of N -methylhistidine, a marker of myofibrillar protein degradation from intact skeletal muscle in vitro and skeletal muscle cells in culture. However, the mechanism(s) mediating these effects are poorly understood.
Many extracellular signals such as hormones, neurotransmitters, and growth factors elicit their response by activating an intracellular signaling cascade that is initiated by the hydrolysis of membrane phospholipids (8 -11). In L6 skeletal muscle cells, we have demonstrated that TPA and vasopressin stimulate both protein synthesis and a phospholipase D (PLD) that degrades phosphatidylcholine. Furthermore, incubation of L6 cells with exogenous PLD mimicked the effects of TPA and vasopressin on protein synthesis, implying a link between the two events (4). Interestingly, however, TPA increased protein synthesis only during short term (90-min) incubations, which were neither blocked by the transcription inhibitor, actinomycin D, nor accompanied by increases in RNA, implying an effect only on translation. In contrast, the synthesis response to vasopressin was detected solely over longer term (6-h) incubations, where increases in RNA content and sensitivity to actinomycin D were observed, demonstrating effects on transcription (4).
One explanation of why TPA and vasopressin produced such temporally different effects on protein synthesis involves the hydrolysis by PLD of phospholipids other than phosphatidylcholine with the production of different phosphatidic acid (PA) species that either directly or indirectly manipulate translation and transcription. Although most studies of PLD to date have focused on phosphatidylcholine, it is now clear that phosphatidylethanolamine (PE) is also a substrate for PLD. TPA has been shown to activate a PLD that degrades PE in several cell types (e.g. [12][13][14][15], while as far as we are aware, there is no published evidence demonstrating that vasopressin elicits PE hydrolysis. Thus, to gain a greater insight into the mechanism(s) through which these two agonists may elicit temporally different increases in protein synthesis, we have initially focused in this study on PE metabolism and considered the possibility that TPA and vasopressin stimulate a PLD that degrades PE in L6 skeletal muscle myoblasts.
Alternative explanations of the temporally different effects of TPA, vasopressin, and exogenous PLD include the possibility that either the PLD preparation may have elicited its responses through a different mechanism, e.g. through the production of lysophosphatidic acid (16), or that the stimulation of PLD by TPA and/or vasopressin is not related to their effects on protein synthesis. In many cell types, studies using protein kinase C (PKC) inhibitors and down-regulation protocols have implicated PKC in the stimulation of PLD by TPA and vasopressin (e.g. see Refs. 10 and 17). In the second aspect of this study, we have used similar approaches to investigate both the regulation of PE hydrolysis by PLD and the possibility that activation of PLD can be dissociated from the stimulation of protein synthesis.
Determination of Intra-and Extracellular [ 14 C]Ethanolamine Watersoluble Products of Phospholipid Hydrolysis-Prior to each experiment, cells were transferred to 2 ml of serum-free DMEM containing [2-14 C]ethanolamine (0.1 Ci/ml) for 24 h. At the end of this labeling period, the cells were washed and incubated in fresh DMEM containing 5 mM unlabeled ethanolamine for an additional 2 h. After a further wash with DMEM, the cells were then incubated with or without agonists in the presence of 1 mM unlabeled ethanolamine and phosphoethanolamine. At the end of the incubation, the total medium was removed from the dishes, and to 0.8 ml of this, 3 ml of ice-cold chloroform/methanol (1:2, v/v) were added. After 30 min on ice, phase separation was achieved with chloroform (1 ml) and water (1 ml). The upper (water/methanol) phase was then removed. Extracellular metabolites were separated (23) and subjected to scintillation counting. In addition, increases in [ 14 C]ethanolamine were confirmed by thin layer chromatography (TLC) as described below for intracellular metabolites.
To analyze intracellular metabolites, following removal of the medium, each dish was washed three times with 1 ml of phosphatebuffered saline, and 1 ml of ice-cold methanol was added. The cells were scraped and washed into glass tubes with a further 1 ml of methanol. Chloroform (1 ml) and 12 M HCl (20 l) were added to each tube. The contents were mixed vigorously and extracted on ice for 30 min. Phase separation was achieved with chloroform (1.25 ml) and water (1.25 ml). The upper phase was removed, and the intracellular metabolites were separated (23) (4).
Down-regulation of PKC-Cells were incubated with 1 M TPA for 24 h prior to use. Determination of the PKC isoforms removed by this protocol was performed as described below.
Western Blot Analysis of PKC Isoforms-To obtain cytosolic and particulate fractions, L6 cells were washed three times with phosphatebuffered saline and scraped from the dish with 2 ϫ 100 l of Buffer A (20 mM Tris, pH 7.5, 0.25 M sucrose, 10 mM EGTA, 2 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, and 20 g/ml leupeptin) (24). The cells were then lysed by six passages through a 25-gauge needle, and cytosolic fractions were obtained by centrifugation at 100,000 ϫ g for 60 min. Membrane pellets were resuspended in the same volume of Buffer A supplemented with 1% Triton X-100 and incubated on ice for 60 min with gentle agitation. The supernatant following centrifugation at 100,000 ϫ g for 60 min was taken as the particulate fraction. To obtain total cell PKC content, 1% Triton X-100 was included in Buffer A when the cells were lysed, and the extract was incubated on ice for 60 min prior to centrifugation as above. The supernatant was taken as total cell PKC. Protein content was determined by the method of Bradford (25). Samples were stored, without boiling, at Ϫ70°C in SDS-polyacrylamide gel electrophoresis sample buffer (5% (w/v) SDS, 13% (w/v) glycerol, 60 mM Tris-HCl, 0.2% bromphenol blue, and 5% mercaptoethanol, pH 6.8). Equal amounts (15 g) of cytosolic or solubilized membrane protein were subjected to SDS-polyacrylamide gel electrophoresis (10% gels), and Western blot analysis was performed with a Mini Deca-Probe (26).
Measurement of Protein Synthesis-Following 24 h in serum-free DMEM, protein synthesis was measured by the addition of L-[2,6-3 H] phenylalanine for the final 60 min of incubations lasting 90 min or 6 h. Subsequent treatment to determine the specific radioactivity of proteinbound phenylalanine was as described previously (27).
Data Presentation-All lipid experiments were done with n ϭ 3 and those for protein synthesis with n ϭ 5. Each assay was performed at least twice under identical conditions. In studies designed to assess the effects of inhibitors, experiments were combined and first subjected to analysis of variance (ANOVA) (using set as a blocking factor) with the Genstat 5 version 3.1 statistics package to determine differences between control and agonist-treated samples. Then, by fitting a contrast using each experimental group, ((agonist with inhibitor) Ϫ inhibitor) Ϫ (agonist Ϫ control), it was possible to determine the effect of an inhibitor on any stimulated event. Where ANOVA showed no difference between experiments, the combined data are shown. Results are expressed as means Ϯ S.E., and significance was assessed by Student's t test.  (Fig. 2D). It was partially attenuated (ϳ70%) by the PKC inhibitor (Fig. 3D) and completely inhibited by down-regulation (Fig. 4D). To investigate this discrepancy, we considered the possibility that with the current approach, the inability to demonstrate sustained PLD activity in response to TPA might be due to insufficient butanol at the later time points. To examine this potential explanation, we added butanol for 20 min either at time 0 with and without agonists or at increasing times after agonist treatment. Using this approach, both TPA (14,124 Ϯ 1342, n ϭ 3; p Ͻ 0.001) and vasopressin (10382 Ϯ 753, n ϭ 3; p,0.001) significantly increased [ 3 H]PtdBuOH formation compared with control (2231 Ϯ 345) when butanol was added at time 0. However, when butanol was added 40 min after the agonists, a different picture emerged. TPA (10,314 Ϯ 1119, n ϭ 3; p Ͻ 0.001), but not vasopressin (2974 Ϯ 603, n ϭ 3) significantly increased [ 3 H]PtdBuOH production when compared with control (2182 Ϯ 489, n ϭ 3). Similar results were also obtained when butanol was added 100 min after agonists (data not shown  (24-h) incubation of L6 myoblasts with 1 M TPA to downregulate PKC abolished the ability of TPA to increase protein synthesis (Fig. 5A), but in complete contrast, the response to vasopressin was unaffected (Fig. 5B). TPA also failed to stimulate protein synthesis in the presence of the PKC inhibitor (Fig. 5C), and, unlike down-regulation of PKC, this inhibitor also partially attenuated (ϳ60%) the response to vasopressin at 1 M (Fig. 5D). To confirm the involvement of PKC in the stimulation of transcription by vasopressin, an additional inhibitor (bisindolylmaleimide; bIM) was used. Like Ro-31-8220, bIM partially blocked the effect (ϳ50%). Vasopressin increased protein synthesis from 12.1 Ϯ 0.3 to 15.8 Ϯ 0.3 dpm/g protein (n ϭ 10 from two combined experiments, p Ͻ 0.001) in the absence of 1 M bIM and from 10.5 Ϯ 0.3 to 12.2 Ϯ 0.2 dpm/g protein (n ϭ 10, by ANOVA; p Ͻ 0.001 versus bIM control and vasopressin alone) in its presence. Furthermore, the partial inhibition with both Ro-31-8220 and bIM was still present following the removal of specific isoforms of PKC by downregulation (Fig. 6).

Stimulation of [ 3 H]Choline Release and [ 3 H]PtdBuOH Formation by TPA and Vasopressin-In
Stimulation of Protein Synthesis by Exogenous PLD-We have previously shown that exogenous PLD elicits an increase in protein synthesis at both 90 min and 6 h in these cells (4). Preincubation with 1 M TPA had no significant effect on the ability of exogenous PLD to stimulate protein synthesis at 90 min (data not shown). In contrast, the response at 6 h was significantly attenuated by TPA preincubation. 5 units/ml PLD increased synthesis from 12.1 Ϯ 0.2 to 14.3 Ϯ 0.1 dpm/g protein (n ϭ 10, p Ͻ 0.001) in non-down-regulated cells but only from 12.0 Ϯ 0.3 to 13.5 Ϯ 0.1 dpm/g of protein (n ϭ 10, by ANOVA; significantly different from the increase observed in the absence of down-regulation, p Ͻ 0.05) in TPA-pretreated cells.
PKC Isoforms in L6 Myoblasts-Immunoblotting of total cell extracts demonstrated the presence of the following PKC isoforms (with approximate molecular masses): ␣ (81 kDa), ␦ (77 kDa), (115 kDa), ⑀ (88 kDa), (74 kDa), and (73 kDa). PKC-␤, -␥, and -were not detected (data not shown). When extracts were fractionated, ␣ was detected in the cytosol, whereas ⑀ and ␦ were found in the membrane. In contrast, , , and were present in both cytosolic and membrane fractions (Fig. 7A). Following a 24-h incubation with 1 M TPA, immunoblotting of total cell extracts demonstrated that PKC-␣ and -␦ could no longer be detected, and densitometric analysis showed that the PKC-⑀ content had decreased by ϳ90%. In contrast, PKC-, -, and -appeared resistant to down-regulation under these conditions (Fig. 7B). DISCUSSION We demonstrate in this study that TPA and vasopressin stimulate [ 14 C]ethanolamine release from prelabeled L6 myoblasts. These effects were elicited in the presence of a large excess of unlabeled ethanolamine and phosphoethanolamine, When taken together with effects on [ 3 H]PtdBuOH formation, the data imply that both TPA and vasopressin activate a PLDthatdegradesPE.Furthermore,thetime-andconcentrationdependent release of extracellular [ 14 C]ethanolamine in response to both agonists mirrored their effects on extracellular [ 3 H]choline release in these cells (4). Comparison of the timedependent effects on [ 3 H]PtdBuOH formation by TPA using the two different protocols suggests that some of the data in the literature and the conclusions drawn from it should be treated with caution. When sufficient butanol is present for transphosphatidylation to take place, the TPA response clearly correlates with the effect on extracellular [ 14 C]ethanolamine release; i.e. both events are sustained for at least 2 h. Interestingly, using the modified protocol to assess [ 3 H]PtdBuOH formation, we have found in C 2 C 12 myoblasts that the activation of PLD by TPA continues for at least 6 h. 2 TPA also stimulates ethanolamine release within 2 min in HeLa cells (12) and after a lag period (10 min) in HL-60 cells, NIH 3T3 fibroblasts, and baby hamster kidney cells (13). In contrast, as far as we are aware, this is the first demonstration that vasopressin stimulates a PLD that degrades PE in any cell type or tissue. Many of these previous studies have measured either total ethanolamine metabolites and failed to distinguish between intra-and extracellular release (e.g. see Refs. 13,14,28,and 29) or have measured only extracellular release (12). When intra-and extracellular release of choline has been measured, both vasopressin and TPA rapidly stimulated intracellular release. The vasopressin response had returned to control values within 10 min, and that for TPA was almost restored within 20 min (e.g. see Ref. 30). The different time course of intra-and extracellular release observed with TPA may reflect hydrolysis of PE pools in the inner and outer leaflet of the plasma membrane or in intracellular membranes. It might also involve the activity of both cytosolic and membrane-bound PLD activities (e.g. see Ref. 31 and see below). In addition to TPA and vasopressin, PE hydrolysis by PLD has also been reported to be stimulated by adenine nucleotides (28), platelet-derived growth factor and bombesin (29), lipid A (32), and endothelin (33).
Vasopressin, which generates diacylglycerol and thus stimulates PKC through inositol lipid hydrolysis (4), and TPA, activates PLD in many cell types, suggesting a role for PKC in the activation of PLD. In this and several other studies, inhibitors of PKC and down-regulation of PKC have been shown to attenuate TPA and vasopressin-stimulated PLD activity (e.g. see Refs. 34 -36). Our observation that down-regulation com- This suggests that the inability of Ro-31-8220 completely to inhibit TPA-induced extracellular release and vasopressin-induced intracellular release of [ 14 C]ethanolamine may be due to its action as a competitive inhibitor (18). Contrastingly, in other cell types, the activation of PLD by TPA may not involve PKC. For example, PKC inhibitors failed to block the activation of PLD by TPA in lymphocytes (37) and mast cells (38).
To gain further insight into the mechanism through which PKC regulates PLD in L6 cells, we examined which isoforms are present and subject to down-regulation. As far as we are aware, this is the first attempt to study PKC isoforms in L6 cells. A previous study in skeletal muscle and L8 skeletal muscle cells demonstrated the presence of PKC-␣, -␦, and -, but not PKC-␤ ␣nd -␥. Furthermore, ␣ and ␦, but not were down-regulated by TPA in L8 cells (39). These are very similar to our findings reported in this study and in C 2 C 12 skeletal muscle cells. 3 Other work has also reported the presence of PKC-␣ (40,41), -␦ (41,42), and -⑀ (41) in skeletal muscle.
The data showing a correlation between the disappearance of PKC-␣, -␦, and -⑀ and the loss of PLD activation by TPA and vasopressin clearly implicate one or more of these isoforms in this response. Their down-regulation upon TPA treatment has been observed in other cell lines (e.g. see Refs. 39,43,and 44). Furthermore, PKC-␣ has been shown to activate PLD in Madin Darby canine kidney cells (45) and CCL39 fibroblasts (46), while PKC-⑀ has been suggested to regulate PLD activity in rat mesangial cells (47). There is no evidence to date of a role for PKC-␦ in the regulation of PLD activity. It is not yet clear from our data if TPA and vasopressin activate the same PKC and/or PLD. Interestingly, although PKC-␣ is largely cytosolic in L6 and other cell lines (e.g. see Ref. 43), we have found PKC-⑀ to be membrane-associated, and this is also the case to some degree in U937 cells (48), rat6 fibroblasts (49), and renal mesangial cells (50). It is intriguing to suggest that isoform and/or location-specific PKC/PLD activities may be responsible for both the rapid but transient responses and the delayed but sustained stimulation observed with TPA. Such a possibility requires further investigation.
We have previously shown in L6 myoblasts that the EC 50 for vasopressin stimulation of transcription is 10-fold higher than that for [ 3 H]PtdBuOH formation and [ 3 H]choline release (4). Data from the current study show that this is also true for [ 14 C]ethanolamine release. Furthermore, the maximal stimu-lation of protein synthesis observed with exogenous PLD at 6 h was 12%, whereas an increase of 30% or more was elicited by vasopressin (4). Thus, while data in this and our previous study (4) support a link between the activation of PLD and the stimulation of translation by TPA, it also implies that, at best, activation of PLD is only part of the mechanism by which vasopressin increases protein synthesis in these cells. It is now clear that while PKC is involved in the stimulation of both PLD and transcription by vasopressin, the down-regulation protocol makes it possible to dissociate the two events completely. Thus, different isoforms of PKC mediate vasopressin effects on PLD and protein synthesis. Consequently, it appears that the increase in transcription we have previously observed with exogenous PLD in these cells (4) must involve an alternative mechanism to that employed by vasopressin (see below). This conclusion is also supported by the finding that the stimulation of protein synthesis at 6 h by exogenous PLD, but not vasopressin, is partially attenuated by PKC down-regulation. One possible mechanism involves the generation of lysophosphatidic acid and its action through an extracellular receptor (16).
Immunoblotting of the down-regulated cells implicates PKC-, -, and/or -in the stimulation of protein synthesis by vasopressin. However, all three isoforms have a high degree of sequence homology, raising the possibility of cross-reactivity between the antibodies. Studies have shown that the antibody cross-reacts with , 4 and, since we have been unable to obtain an alternative source of antibody, its presence in L6 cells remains unproven at present. The antibody to PKCclearly detects a protein of 115 kDa, but it fails to recognize any bands in the 70 -90-kDa range. Furthermore, the antibodies to and do not detect the 115-kDa protein. Thus, from the data available on the three atypical isoforms, it appears likely that only and are present. The inability of TPA to down-regulate PKC is also of interest, since when human PKC-was propagated in the baculovirus expression system and purified to homogeneity, it displayed high affinity TPA binding (51). However, this finding differs markedly from an earlier study in which only a very weak increase in TPA binding was observed in cellular extracts from PKCtransfectants (52). The difference between the two observations may be due to unknown factors present in the cellular extracts that prevent TPA binding.
Although the increase in translation elicited by insulin in L6 cells was not prevented by Ro-31-8220 (53), this inhibitor (54), but not PKC down-regulation, 3 also attenuated the stimulation of transcription by insulin, suggesting that vasopressin and insulin utilize the same subset of PKC isoforms. In many of the cell lines investigated so far, PKC-seems to be present as a cytosolic enzyme (e.g. see Ref. 55). As predicted from its structure, most studies, including the work reported here, indicate that PKCis resistant to TPA-induced translocation or down-regulation (e.g. see Refs. 50 and 56). In addition, both of the PKC inhibitors that partially blocked the effect of vasopressin on transcription have been shown to inhibit PKC-(e.g. see Ref. 57). The observation that higher concentrations of both agents (1 M) were required to partially inhibit PKC than to completely inhibit the ␣, ␦, and ⑀ isoforms (57) and our finding that Ro-31-8220 and bIM still elicited a partial inhibition of the vasopressin response following down-regulation further support a role for the down-regulation-resistant isoforms such as in mediating the vasopressin response. In addition, vasopressin has been shown to stimulate the translocation of PKC from the cytosol to a membrane fraction in human platelets (58), and in smooth muscle, phenylephrine elicits PKCtranslocation to the nucleus (59), suggesting a role in events such as gene expression.
PKChas been shown to be stimulated in vitro by phosphatidylinositol 3,4,5-trisphosphate (60), the phospholipid that is thought to be the physiologically important product of phosphatidylinositol-3-kinase (61). However, the phosphatidylinositol-3-kinase inhibitor, wortmannin (62), had no effect on the ability of vasopressin to stimulate transcription in L6 cells, suggesting that phosphatidylinositol-3-kinase is not involved in mediating this response. 3 PKC-is also activated by arachidonic acid (63), and preliminary data show that vasopressin increases arachidonic acid release from L6 cells. 5 Furthermore, we have previously shown that vasopressin stimulates mitogen-activated protein kinase in these cells (64) and mitogenactivated protein kinase is known to phosphorylate and activate cytosolic PLA 2 (65), releasing arachidonic acid. Thus, it is possible that vasopressin may, at least in part, stimulate protein synthesis in L6 cells through PKCvia mitogen-activated protein kinase, cytosolic PLA 2 , and arachidonic acid. This potential sequence of events is under further investigation.
Although PLD is clearly not involved in the stimulation of protein synthesis by vasopressin, both this agonist and TPA also reduce myofibrillar protein degradation in skeletal muscle (6,7). It remains to be determined whether PLD plays a role in the regulation of this component of protein turnover.