Ceramide-induced Inhibition of Akt Is Mediated through Protein Kinase C (cid:1)

We recently demonstrated that ceramide-coated bal-loon catheters limit vascular smooth muscle cell (VSMC) growth after stretch injury in vivo . In that study, inhibition of VSMC growth was correlated with a decrease in phosphorylation of the cell survival kinase Akt (pro-tein kinase B). Utilizing cultured A7r5 VSMCs, we have now examined the mechanism by which ceramide inhibits Akt phosphorylation/activation. Our initial studies showed that ceramide-induced inhibition of Akt phosphorylation was not mediated through diminution in phosphoinositide 3-kinase activity. As we have previously demonstrated that protein kinase C (cid:1) (PKC (cid:1) ) is a target of ceramide, we proposed an alternative signaling mechanism by which ceramide induces inhibition of Akt through activation of PKC (cid:1) . We demonstrate that C 6 - ceramide (but not the inactive analog dihydro-C 6 -cer-amide) induced PKC (cid:1) activity and also caused a selective increase in the association between Akt and PKC (cid:1) , without affecting PKC (cid:2) , in A7r5 cells. In addition, the ability of ceramide to significantly decrease platelet-derived growth factor-induced Akt phosphorylation or cell proliferation was abrogated in A7r5 cells overexpressing a dominant-negative mutant of PKC (cid:1) . Taken together, these data suggest that ceramide-mediated activation of PKC (cid:1) lecular Dynamics, Inc., Sunnyvale, CA) (3, 15, 25). The protein concentration was determined by the Bradford method (BioRad, Hercules, CA).To confirm the specificity of anti-PKC (cid:1) primary antibody, the mem- branes were stripped (15) and reprobed with additional PKC isoform-specific antibodies as described previously (4, 5). A protein band corre- sponding to the molecular mass of PKC (cid:1) was observed when the membranes were probed with anti-PKC (cid:1) antibody. In contrast, the protein band was not observed when the membranes were probed with antibody specific for PKC (cid:4) , PKC (cid:7) , or PKC (cid:3) . The anti-Akt-1 antibody used in this study is reactive with the Akt-1 isoform, with some cross-reactivity for the Akt-2 and Akt-3 isoforms. The anti-phospho-Akt(Ser 473 ) antibody recognizes phosphorylated Ser 473 (Akt-1), Ser 474 (Akt-2), and also Ser 472 (Akt-3) in the carboxyl- terminal regions of these Akt isoforms. The phosphorylation of Ser 473 / Ser 474 /Ser 472 is critical for the activation of Akt. In view of the possible recognition of all three Akt isoforms by the anti-Akt antibodies used in this study, the terms Akt and phosphorylated Akt have been used to denote Akt protein and Akt activation, respectively.

We recently demonstrated that ceramide-coated balloon catheters limit vascular smooth muscle cell (VSMC) growth after stretch injury in vivo. In that study, inhibition of VSMC growth was correlated with a decrease in phosphorylation of the cell survival kinase Akt (protein kinase B). Utilizing cultured A7r5 VSMCs, we have now examined the mechanism by which ceramide inhibits Akt phosphorylation/activation. Our initial studies showed that ceramide-induced inhibition of Akt phosphorylation was not mediated through diminution in phosphoinositide 3-kinase activity. As we have previously demonstrated that protein kinase C (PKC) is a target of ceramide, we proposed an alternative signaling mechanism by which ceramide induces inhibition of Akt through activation of PKC. We demonstrate that C 6ceramide (but not the inactive analog dihydro-C 6 -ceramide) induced PKC activity and also caused a selective increase in the association between Akt and PKC, without affecting PKC⑀, in A7r5 cells. In addition, the ability of ceramide to significantly decrease plateletderived growth factor-induced Akt phosphorylation or cell proliferation was abrogated in A7r5 cells overexpressing a dominant-negative mutant of PKC. Taken together, these data suggest that ceramide-mediated activation of PKC leads to diminished Akt activation and consequent growth arrest in VSMCs. The therapeutic potential for ceramide to limit dysregulated VSMC growth has direct applicability to vascular diseases such as restenosis and atherosclerosis.
Ceramide is a sphingolipid-derived second messenger that has been implicated in growth arrest, differentiation, and/or apoptosis (1)(2)(3). Using human embryonic kidney cells, we have recently demonstrated that ceramide-induced growth arrest is associated with changes in protein kinase signaling cascades, including protein kinase C (PKC) 1 and mitogen-activated pro-tein kinase (4,5). In particular, ceramide-mediated PKC activation leads to inhibition of cell growth (4). In addition, we (3) and others (6 -10) have shown that ceramide inhibits cell growth through inhibition of the cell survival kinase Akt/protein kinase B in vivo and in vitro. However, the precise mechanism by which ceramide inhibits Akt activation has not been elucidated.
Multiple lines of evidence suggest that Akt is a downstream target of phosphoinositide 3-kinase (PI3K) in different cell types, including vascular smooth muscle cells (VSMCs) (11)(12)(13)(14)(15)(16). Whether ceramide inhibits Akt activation through a PI3Kdependent mechanism is still a subject of controversy in the literature. Recent studies have shown that ceramide-induced inhibition of Akt occurs in a PI3K-dependent (6) and PI3Kindependent (17,18) manner. Of particular importance is the fact that PKC has been identified as a putative PI3K-independent regulator of Akt (19). Direct binding of PKC to Akt (20,21) serves to negatively regulate Akt activation (19). As we (4) and others (22)(23)(24) have shown that PKC is a target of ceramide, it is now hypothesized that PKC may mediate ceramide-induced inhibition of Akt.
Cell Culture and Treatments-Rat embryonic thoracic aorta smooth muscle-derived A7r5 cells were obtained from American Type Culture Collection (Manassas, VA). The cells were incubated in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum at 37°C in a humidified atmosphere of 95% air and 5% CO 2 . After attainment of confluency (70 -80%), the cells were incubated in serum-free DMEM containing 0.5% bovine serum albumin for 20 -24 h. For experiments involving treatment with C 6 -, C 2 -, dihydro-C 2 -, or dihydro-C 6 -ceramide, the following protocol was used. A stock solution (200 mM) of ceramide derivative(s) was initially prepared in Me 2 SO. This was diluted 200-fold with serum-free DMEM containing 5% bovine serum albumin to obtain a working ceramide concentration of 1 mM (with constant stirring for 45 min at room temperature). To expose the cells to the ceramide derivative, the working ceramide solution was added to serum-free DMEM containing 0.5% bovine serum albumin to obtain the final concentration (1 or 10 M) as indicated in the figure legends.
[ 3 H]Thymidine Incorporation Studies-Serum-starved A7r5 cells were pretreated without (control) or with ceramide derivative(s) for 1-2 h and then exposed to PDGF (10 ng/ml) for an additional 20 h in the presence of the above agents. During the last 4-h incubation period, the cells were labeled with [ 3 H]thymidine (0.5 Ci/ml). After labeling, the cells were washed with ice-cold phosphate-buffered saline and then exposed to 10% trichloroacetic acid (3 ϫ 10 min). After complete removal of the trichloroacetic acid, the acid-insoluble material was extracted with 0.1 N NaOH. The incorporation of [ 3 H]thymidine into acid-insoluble DNA was determined with a liquid scintillation counter (15).
Western Blot Analyses-Both control and treated A7r5 cells were washed with ice-cold phosphate-buffered saline and exposed to lysis buffer (50 mM HEPES (pH 7.4), 150 mM NaCl, 1 mM NaF, 1 mM EDTA, 1 mM EGTA, 1 mM Na 3 VO 4 , 0.2% Nonidet P-40, 10 g/ml leupeptin, 10 g/ml pepstatin, and 10 g/ml aprotinin). The cell lysates were centrifuged at 14,000 ϫ g for 10 min at 4°C. The supernatants (40 g of protein) from the cell lysates were subjected to 10% SDS-PAGE, and the resolved proteins were transferred to nitrocellulose membranes (Hybond C, Amersham Biosciences, Inc., Uppsala, Sweden). The membranes were blocked in 5% nonfat milk in Tris-buffered saline for 1 h and then incubated with the appropriate primary antibody (1:1000 dilution) for 2 h at room temperature. After incubation, the membranes were washed with Tris-buffered saline (3 ϫ 10 min). The membranes were then incubated with the appropriate secondary antibody (1:5000 dilution) for 2 h at room temperature. After three more washes with Tris-buffered saline, the protein bands were detected by the enhanced chemiluminescence method and quantitated by laser densitometry (Molecular Dynamics, Inc., Sunnyvale, CA) (3,15,25). The protein concentration was determined by the Bradford method (BioRad, Hercules, CA).
To confirm the specificity of anti-PKC primary antibody, the membranes were stripped (15) and reprobed with additional PKC isoformspecific antibodies as described previously (4,5). A protein band corresponding to the molecular mass of PKC was observed when the membranes were probed with anti-PKC antibody. In contrast, the protein band was not observed when the membranes were probed with antibody specific for PKC␣, PKC␦, or PKC⑀.
The anti-Akt-1 antibody used in this study is reactive with the Akt-1 isoform, with some cross-reactivity for the Akt-2 and Akt-3 isoforms. The anti-phospho-Akt(Ser 473 ) antibody recognizes phosphorylated Ser 473 (Akt-1), Ser 474 (Akt-2), and also Ser 472 (Akt-3) in the carboxylterminal regions of these Akt isoforms. The phosphorylation of Ser 473 / Ser 474 /Ser 472 is critical for the activation of Akt. In view of the possible recognition of all three Akt isoforms by the anti-Akt antibodies used in this study, the terms Akt and phosphorylated Akt have been used to denote Akt protein and Akt activation, respectively.
PI3K Assays-For the intact cell experiments, serum-starved A7r5 cells were pretreated with C 2 -, C 6 -, or dihydro-C 6 -ceramide or vehicle for 1 h and then incubated without or with PDGF (10 ng/ml) for 5 min. The supernatants (400 g of protein) from the cell lysates were immunoprecipitated with anti-p85 PI3K antibody by overnight incubation at 4°C, followed by a 2-h incubation at 4°C with GammaBind G-Sepharose (Amersham Biosciences, Inc.). The immunocomplexes were washed once with lysis buffer and twice with kinase buffer containing 20 mM Tris-HCl (pH 7.5), 100 mM NaCl, 0.5 mM EGTA, and 10 mM MgCl 2 . PI3K assays were performed with these immunocomplexes for 15 min at 35°C in the presence of 20 g/ml phosphatidylserine and 20 M [ 32 P]ATP (10 Ci of ATP/assay), using 200 g/ml phosphoinositide as the exogenous substrate. The kinase reactions were terminated by the addition of 200 l of 1 M HCl/methanol. Lipids were extracted twice with 200 l of chloroform and then dried under nitrogen. After solubilizing the lipids in 40 l of chloroform, the samples were subjected to TLC (Whatman Silica Gel 60, aluminum). The lipids were separated using a solvent system consisting of chloroform/methanol/H 2 O/NH 4 OH (60:47:11.3:2). Phosphoinositide 3-phosphate (PI3P), formed by the catalytic action of PI3K, was visualized by autoradiography and quantitated by laser densitometry. As a negative control, mock immunoprecipitations were performed using lysis buffer, which revealed negligible formation of 32 P-labeled PI3P.
For the cell-free experiments, serum-starved A7r5 cells were treated without (control) or with PDGF (10 ng/ml) for 5 min. After immunoprecipitation, the immunocomplexes were treated with C 2 -, C 6 -, or dihydro-C 6 -ceramide (1 M each) for 10 min at 35°C. PI3K assays with these ceramide-treated immunocomplexes were performed as described above.
In Vitro Kinase Assays for Immunoprecipitated PKC-Serumstarved A7r5 cells were exposed to vehicle (control), dihydro-C 6 -ceramide (1 M, 1 h), C 6 -ceramide (1 M, 1 h), IL-1␤ (20 ng/ml, 5 min), or PDGF (10 ng/ml, 5 min). The supernatants (200 g of protein) from the cell lysates were subjected to immunoprecipitation with anti-PKC antibody for 3 h at 4°C. The immunocomplexes were captured by incubation with GammaBind G-Sepharose for 2 h at 4°C. The immunocomplexes were then washed twice with phosphate-buffered saline and once with kinase buffer containing 50 mM HEPES (pH 7.5), 100 mM NaCl, 10 mM MgCl 2 , 50 mM NaF, 1 mM NaVO 4 , 1 mM dithiothreitol, and 0.1% Tween 20. Subsequently, the immunocomplexes were subjected to in vitro kinase assays for 15 min at 35°C in the presence of 40 g/ml phosphatidylserine and 1 Ci of [␥-32 P]ATP (10 mCi/mmol) using myelin basic protein (MBP; 10 g) as the substrate. The kinase reactions were terminated by the addition of SDS sample buffer. After heating for 5 min at 95°C, the samples were separated by 12% SDS-PAGE and transferred to nitrocellulose membranes. The phosphorylation of MBP was assessed by liquid scintillation counting of the excised MBP bands. Control experiments showed that MBP phosphorylation was dependent upon exogenous phosphatidylserine (4).
Co-immunoprecipitation of PKC with Akt-Serum-starved A7r5 cells were pretreated without (control) or with C 6 -ceramide (1 M) or its inactive analog dihydro-C 6 -ceramide (1 M) for 1 h and then exposed to IGF-1 (50 ng/ml) for 5 min. The supernatants (200 g of protein) from the cell lysates were subjected to immunoprecipitation with anti-Akt-1 antibody and then assessed for co-immunoprecipitation with PKC or PKC⑀ by probing the blots with the respective anti-PKC primary antibody. The protein bands corresponding to PKC or PKC⑀ were visualized by the enhanced chemiluminescence method and quantitated by laser densitometry. Equal loading of samples was verified by reprobing the blots with anti-Akt-1 antibody.
Transient Transfection Studies with PKC Constructs-A7r5 cells were transiently transfected with wild-type or dominant-negative mutant PKC constructs (a generous gift from Dr. J. Moscat) using Superfect ® (QIAGEN Inc.) according to the manufacturer's instructions. Superfect reagent consists of non-lipid-activated dendrimer molecules, which assemble DNA into compact structures and optimize DNA entry into cells. The wild-type construct is full-length PKC in a pcDNA3 expression vector. The dominant-negative mutant construct is a kinase-defective mutant that contains a point mutation in the catalytic domain. Briefly, confluent (ϳ70%) A7r5 cells were incubated with 1 ml of DMEM (supplemented with 10% fetal bovine serum) along with a mixture of plasmid DNA (2 g) and Superfect (10 l) for 2 h at 37°C. Subsequently, the culture media were removed, and the cells were incubated with fresh DMEM containing 10% fetal bovine serum for 24 h to allow for protein expression. The transfected cells were then subjected to serum starvation by incubating in DMEM devoid of fetal bovine serum for 24 h. Transfection efficiency was determined to be ϳ25-30% as assessed by green fluorescent protein cotransfection. Serum-starved cells were pretreated with C 2 -or C 6 -ceramide (1 M) for 1 h and then incubated with PDGF (10 ng/ml) for 5 min (Fig. 6) or 20 h (Fig. 7). For the data shown in Fig. 6, the cell lysates (40 g of protein) were immunoprecipitated with anti-PKC antibody, subjected to SDS-PAGE, and then probed with anti-phospho-Akt(Ser 473 ) antibody. For Fig. 7, [ 3 H]thymidine incorporation studies were performed as described above.
Statistical Analyses-The results are expressed as the means Ϯ S.E. of three or more experiments. The data were analyzed by one-way analysis of variance (ANOVA), followed by unpaired t tests, corrected by Bonferroni's method. In select experiments, non-parametric data were analyzed by Mann-Whitney one-way ANOVA tests, followed by Dunn's correction method. In those experiments in which the control optical density values were set to 100%, the S.E. for each of these control values was reported using the nontransformed data.

RESULTS
Ceramide Induces Growth Arrest in A7r5 Cells-We have previously shown that ceramide inhibits growth factor receptor-stimulated mitogenesis in rat glomerular mesangial cells (25,26) and human embryonic kidney cells (4,5). In the present study, we examined whether ceramide induces growth arrest in rat aortic vascular smooth muscle-derived A7r5 cells. We initially chose a ceramide concentration (10 M) that was previously shown to maximally induce growth arrest without inducing necrosis or apoptosis. As assessed by [ 3 H]thymidine incorporation studies, cell-permeable ceramide had negligible effects on basal DNA synthesis (Fig. 1). However, pretreatment of A7r5 cells with ceramide led to a significant decrease in PDGF-stimulated DNA synthesis. These results suggest that ceramide inhibits proliferation of A7r5 cells and support our in vivo studies demonstrating that cell-permeable ceramides limit neointimal hyperplasia after stretch injury in rabbit carotid arteries (3).
Ceramide Inhibits PDGF-stimulated Akt Phosphorylation-To determine whether ceramide-induced inhibition of cell growth occurs as a consequence of its regulatory effect on the pro-survival kinase Akt, we examined the effects of ceramide pretreatment on PDGF-stimulated Akt phosphorylation in A7r5 cells. We have previously shown that PDGF (10 ng/ml) or IGF-1 (50 ng/ml) stimulation of A7r5 cells leads to maximal phosphorylation of Akt within 5-10 min, which correlates with increases in Akt activity (15). Hence, in the present study, we used a time point of 5 min to demonstrate Akt phosphorylation in response to PDGF. We now demonstrate that the cell-permeable bioactive C 2 -ceramide (but not the inactive dihydro-C 2ceramide) caused significant inhibition of PDGF-induced Akt phosphorylation in A7r5 cells (Fig. 2). These studies suggest that ceramide pretreatment leads to inhibition of growth factor-induced Akt activation in VSMCs.
Ceramide Does Not Inhibit PI3K Activity in A7r5 Cells-We next investigated whether ceramide-induced inhibition of Akt is mediated through PI3K in A7r5 cells. Our initial experiments using PI3K assays with intact A7r5 cells revealed that there was ϳ3.5-fold increase in the formation of PI3P in response to PDGF stimulation (Fig. 3A). Pretreatment of intact A7r5 cells with either C 2 -or C 6 -ceramide, followed by stimulation with PDGF, did not lead to alterations in basal as well as PDGF-stimulated increases in PI3K activity. In addition, dihydro-C 6 -ceramide did not have any effect on PI3K activity. Similarly, in cell-free experiments, stimulation of A7r5 cells without (control) or with PDGF, followed by treatment of PI3K immunocomplexes with C 2 -, C 6 -, or dihydro-C 6 -ceramide, did not reveal significant changes in basal or PDGF-stimulated PI3K activity by ceramide derivatives (Fig. 3B). The data from both intact cell and cell-free experiments indicate that the mechanism by which ceramide mediates its inhibitory effects on Akt activation is not directly dependent upon PI3K. Moreover, these data indicate that both C 2 -and C 6 -ceramide do not regulate PI3K at a concentration that we have previously shown to activate PKC (4) and to inhibit PKC⑀ (5).
Ceramide and IL-1␤ Activate PKC in A7r5 Cells-We next investigated an alternative mechanism by which ceramide could inhibit Akt phosphorylation. Based on the previous work of Doornbos et al. (19), we examined whether ceramide may function as the endogenous signal regulating PKC-dependent inhibition of Akt. Therefore, we assessed the ability of ceramide to activate PKC. Confirming previous studies using other cell types (3,24,27), we now demonstrate enhanced PKC activity in A7r5 cells in response to ceramide. As shown in Fig. 4, exposure of A7r5 cells to C 6 -ceramide (1 M, 1 h), but not dihydro-C 6 -ceramide (1 M, 1 h), resulted in a significant increase in PKC activity. We also demonstrate that IL-1␤ (but not PDGF) could mimic the effects of exogenous ceramide in stimulating PKC activity in A7r5 cells. Previously, we (4,28) and other investigators (29,30) reported that inflammatory cytokines such as IL-1␤ induce ceramide formation in multiple cell types. Attesting to the direct action of ceramide upon PKC, we have previously shown that ceramide is a cofactor for immunoprecipitated as well as recombinant PKC activity (4). Together, these studies show that ceramide activates PKC in A7r5 cells.
Ceramide Induces an Association between Akt and PKC in A7r5 Cells-As we have shown that ceramide activates PKC while inhibiting PDGF-stimulated Akt, we next determined whether ceramide orchestrates a direct interaction between Akt and PKC in A7r5 cells. The rationale for utilizing IGF-1 (instead of PDGF) in these experiments was to demonstrate that the effects of growth factors on these interactions are not specific for PDGF. As shown in Fig. 5, we observed a strong selective association of PKC (but not PKC⑀) with Akt after treatment of A7r5 cells with C 6 -ceramide. Dihydro-C 6 -ceramide did not induce an association of PKC or PKC⑀ with Akt. Growth factors such as IGF-1 alone did not exert any effect on the basal association between Akt and PKC in A7r5 cells. However, IGF-1 produced a significant inhibition of C 6 -ceramide-induced increases in Akt and PKC association. These studies suggest that ceramide may function as an endogenous signal to induce a selective association between PKC and Akt in VSMCs.  Fig. 5 demonstrate that ceramide treatment induced the association of PKC with Akt. However, a simple interaction between two proteins does not imply direct inhibition of Akt. Therefore, we assessed the critical role of PKC in ceramideinduced inhibition of Akt phosphorylation. A7r5 cells were subjected to transient transfection studies utilizing wild-type and dominant-negative mutant PKC constructs. In control experiments, there was an increased expression of PKC protein in cells transfected with wild-type PKC constructs compared with empty vector controls (2.24 Ϯ 0.23 versus 0.66 Ϯ 0.04 (arbitrary units); p Ͻ 0.05). In addition, there was no change in the expression of PKC␣ in cells transfected with wild-type PKC constructs compared with empty vector controls (0.10 Ϯ 0.02 versus 0.11 Ϯ 0.04; not significantly different). As shown in Fig. 6, PDGF produced significant increases in Akt phosphorylation in cells overexpressing wild-type PKC constructs, consistent with growth factor stimulation of the PI3K pathway. Pretreatment of A7r5 cells (overexpress-ing wild-type PKC) with either C 2 -or C 6 -ceramide caused significant attenuations in PDGF-stimulated Akt phosphorylation. In contrast, in cells transfected with the dominantnegative mutant PKC constructs, pretreatment with either Intact cell (A) or cell-free (B) PI3K experiments were performed to assess the ability of ceramide to regulate PI3K activity. For A, serumstarved A7r5 cells were pretreated with or without C 2 -ceramide (C2-Cer; 1 M), C 6 -ceramide (C6-Cer; 1 M), or dihydro-C 6 -ceramide (DH-C6-Cer; 1 M) for 1 h and then incubated without or with PDGF (10 ng/ml) for 5 min. The cell lysates were subjected to immunoprecipitation with anti-p85 PI3K antibody, followed by PI3K assays as described under "Experimental Procedures." For B, serum-starved A7r5 cells were incubated without or with PDGF (10 ng/ml) for 5 min, and the cell lysates were subjected to immunoprecipitation with anti-p85 PI3K antibody. These immunocomplexes were subsequently treated with ceramide analogs (1 M each) for 10 min at 35°C. PI3K assays with these ceramide-treated immunocomplexes were then performed as described under "Experimental Procedures." The data shown are the means Ϯ S.E. of three to six separate experiments. *, p Ͻ 0.05 compared with vehicle control (unpaired Mann-Whitney test).

FIG. 4. Ceramide and IL-1␤ stimulate PKC activity in A7r5 cells.
Serum-starved A7r5 cells were exposed to vehicle (Control), dihydro-C 6 -ceramide (DH-C6-Cer; 1 M, 1 h), C 6 -ceramide (C6-Cer; 1 M, 1 h), IL-1␤ (20 ng/ml, 5 min), or PDGF (10 ng/ml, 5 min). The cell lysates were subjected to immunoprecipitation with anti-PKC antibody, followed by in vitro kinase assays for PKC using MBP (10 g) as the substrate. The phosphorylated MBP was resolved by SDS-PAGE and quantitated by liquid scintillation counting of the excised MBP bands as described under "Experimental Procedures." The data shown are the means Ϯ S.E. of three to four separate experiments. *, p Ͻ 0.05 compared with vehicle (repeated measures one-way ANOVA, followed by Bonferroni's t test).

FIG. 5. Ceramide enhances interaction of PKC (but not PKC⑀) with Akt in A7r5 cells.
Serum-starved A7r5 cells were pretreated with vehicle (Control), C 6 -ceramide (C 6 -Cer; 1 M, 1 h), or dihydro-C 6ceramide (DH-C 6 -Cer, DH; 1 M, 1 h) and then incubated without or with IGF-1 (50 ng/ml) for 5 min. The cell lysates were subjected to immunoprecipitation (IP) with anti-Akt-1 antibody, and the blots were probed with anti-PKC or anti-PKC⑀ antibody as described under "Experimental Procedures." To verify equal loading of samples, the blots were reprobed with anti-Akt-1 antibody. The blots shown in A are representative of a single experiment. The data shown in B are the means Ϯ S.E. of five separate experiments. *, p Ͻ 0.001 compared with vehicle control; #, p Ͻ 0.005 compared with C 6 -ceramide treatment alone (unpaired Mann-Whitney test). IB, immunoblot; CO-IP, co-immunoprecipitation. C 2 -or C 6 -ceramide had no inhibitory effects on PDGFstimulated Akt phosphorylation. In all cases, dihydro-C 6ceramide had no effect on Akt phosphorylation. These experiments suggest that PKC is a necessary component for ceramide-induced inhibition of Akt activation.
PKC Mediates Ceramide-induced Inhibition of Cell Growth-Because we have demonstrated that ceramide-activated PKC mediated inhibition of Akt activation, we next examined whether PKC is also required for ceramide-induced growth arrest. Therefore, we assessed the effects of ceramide analogs on growth factor-induced DNA synthesis in A7r5 cells overexpressing wild-type or dominant-negative mutant constructs. Transient transfections with wild-type PKC led to a significant diminution in PDGF-stimulated DNA synthesis after prior treatment with ceramide (Fig. 7A). These data are in conformity with those shown in Fig. 1. In cells overexpressing dominant-negative mutant PKC constructs, pretreatment with C 6 -ceramide had no significant inhibitory effects on PDGF-stimulated DNA synthesis (Fig.  7B). Dihydro-C 6 -ceramide was without any effect on PDGFstimulated DNA synthesis after overexpression of either construct. Thus, these observations demonstrate that ceramideactivated PKC leads to marked inhibition of growth factor-induced Akt activation as well as cell growth, independent of PI3K.

DISCUSSION
In addition to its well established role as a cell survival kinase, Akt has been implicated in cell proliferation in different cell types, including VSMCs (13-16, 31-34). Activation of Akt at the membrane is a highly regulated event involving several lipid-regulated pleckstrin homology domain-containing kinases as well as PI3K-generated phosphoinositide-3-phosphate derivatives (13, [35][36][37][38][39]. These events lead to conformational changes in Akt-1, with subsequent phosphorylation at Thr 308 and Ser 473 . Fully activated, phosphorylated Akt is able to translocate to the nucleus, where it regulates the protein synthesis of numerous pro-mitogenic and pro-survival transcription factors (40 -44).
Recent studies have suggested that the endogenous sphingolipid-derived second messenger ceramide induces growth arrest, in part, via dephosphorylation of Akt (9,10,45). However, the mechanism leading to inhibition of Akt has not been conclusively demonstrated. As ceramide does not directly inhibit Akt (7), multiple indirect mechanisms for ceramide regulation of Akt have been proposed. An obvious potential mediator for ceramide-induced inhibition of Akt is PI3K. Conflicting results have suggested PI3K-dependent (6) as well as PI3K-independent (17,18) regulation of Akt by ceramide. There are also several reports demonstrating activation of Akt, independent of PI3K (46,47). Our studies argue for a PI3K-independent mechanism for ceramide-induced Akt inhibition in VSMCs. PKC has been implicated as a putative mediator of PI3Kindependent Akt inactivation (19). It has been suggested that PKC is a negative regulator of PDGF-induced increases in Akt activity (19). The ability of PKC to negatively regulate Akt may imply that an interaction between PKC and Akt abrogates the phosphorylation of Akt, yet the precise inhibitory/ regulatory sites for PKC-dependent phosphorylation of Akt currently remain undefined. In addition, the endogenous regulator for PKC-dependent inhibition of Akt has not been determined. These studies suggest that ceramide may function as one such endogenous mediator, regulating PKC-dependent inhibition of Akt phosphorylation in VSMCs. Our studies further support the biological significance of PKC as one of the potential targets for ceramide.
The role of ceramide in binding to and activating PKC is still somewhat controversial. It has been hypothesized that ceramide directly interacts with the single cysteine-rich domain of PKC (26,48). Our previous studies using both immunoprecipitated PKC and human recombinant PKC clearly demonstrate that ceramide (but not dihydroceramide) directly induces PKC bioactivity (4). Supporting our findings, ceramide has been shown to bind to PKC as determined by kinetic analyses and in vitro phosphorylation studies (24,27). In contrast, a radioiodinated photoaffinity-labeled ceramide analog was unable to directly interact with immunoprecipitated PKC (49). These apparent contradictions in the literature may be due to structural differences in the ceramide analogs utilized. Even if ceramide does not directly interact with PKC, our studies still document that ceramide treatment leads to a selective interaction between Akt and PKC and that activated PKC is necessary for ceramide-induced inhibition of Akt.
In addition to the controversy surrounding ceramidestimulated PKC, there is disagreement regarding the signaling function of PKC. There are several reports in the literature that suggest that PKC mediates mitogenic signaling pathways (50 -53). These studies discuss PKC as a downstream signaling target for PI3K and phosphoinositide 3,4,5-trisphosphate-dependent kinase-1. Other reports suggest that PKC may exert growth-arresting effects by directly inhibiting Akt (19,20). Our present studies, using dominantnegative mutant PKC constructs, firmly suggest that ceramide activation of PKC in VSMCs leads to growth arrest. It is also possible that PKC may serve as a bifunctional modulator, as suggested by Muller et al. (24). Thus, it is possible that ceramide and phosphoinositide-3-phosphate derivatives may differentially activate and couple PKC to both pro-and anti-mitogenic effectors. This is a similar scenario to previously published data demonstrating that ceramide preferentially couples activated PKC to upstream elements in the anti-mitogenic stress-activated protein kinase cascade, and not the pro-mitogenic extracellular signal-regulated kinase cascade (4).
In addition to activation of PKC, ceramide has been suggested to activate protein phosphatases that may dephosphorylate Akt, thereby preventing translocation into the nucleus (10,17,18). Even though we did not directly assess the putative role of ceramide-activated protein phosphatases in mediating PKC-dependent inhibition of Akt, studies demonstrate that ceramide activates PKC under conditions in which ceramide activates protein phosphatases (10,56,57). Yet, the literature regarding the role of ceramide-activated protein phosphatases in regulating Akt inhibition is somewhat controversial. Ceramide has been shown to dephosphorylate Akt-1 at Ser 473 and Thr 308 , suggesting a role for a ceramide-activated protein phosphatase (9,10). However, studies by Zhou et al. (7) and Summers et al. (8) suggest that ceramide-induced inactivation of Akt is independent of protein phosphatases. These discrepancies in the literature may be a consequence of nonspecific or toxic actions of the pharmacological inhibitors of protein phosphatase-1 and -2A utilized in these studies. Using a molecular approach, it has been suggested that ceramide inhibition of Akt occurs via a protein phosphatase-independent mechanism (19). Specifically, PKC was able to inhibit the activity of a constitutively active Akt-1 mutant that is not regulated by phosphorylation/dephosphorylation at Ser 473 and Thr 308 .
The mechanism(s) by which growth factors inhibit ceramideinduced increases in the interactions of PKC with Akt remains to be verified. Establishing ceramide as a mediator of PKC/ Akt interactions offers several possible explanations for this observation. Growth factor-generated diglycerides may compete with ceramide at the putative ceramide-binding site on PKC (58) or activate other PKC isoforms linked to mitogenesis. Alternatively, growth factor treatment induces the activation of ceramidase, which forms pro-mitogenic sphingosine species (28,59). The ability of growth factors to inhibit ceramide-induced PKC/Akt interactions probably does not involve activation of PKC because we (this study) and others (19,60) have shown that growth factors induce a marginal increase in PKC activity. Regardless of the mechanism, growth factor treatment reduces the actions of inflammatory cytokines or ceramides in inducing PKC/Akt interactions.
We have shown that ceramide-induced inhibitions of Akt activation and cell growth are mediated by PKC activation, independent of PI3K. The role of ceramide in selectively inducing PKC⅐Akt complex formation may illustrate one mechanism by which cytokine receptor-induced ceramide formation may limit cell proliferation in a pro-inflammatory environment. Inhibition of Akt by cell-permeable ceramide analogs may have direct applicability to the control of dysregulated VSMC proliferation in vivo.