Redox-regulated Signaling by Lactosylceramide in the Proliferation of Human Aortic Smooth Muscle Cells*

Previously, our laboratory reported that lactosylceramide (LacCer) stimulated human aortic smooth muscle cell proliferation via specific activation of p44 mitogen-activated protein kinase (MAPK) in the p21 ras /Raf-1/MEK2 pathway and induced expression of the transcription factor c-fos downstream to the p44 MAPK signaling cascade (Bhunia A. K., Han, H., Snowden, A., and Chatterjee S. (1996) J. Biol. Chem. 271, 10660–10666). In the present study, we explored the role of free oxygen radicals in LacCer-mediated induction of cell proliferation. Superoxide levels were measured by the lucigenin chemiluminescence method, MAPK activity was measured by immunocomplex kinase assays, and Western blot analysis and c-fos expression were measured by Northern blot assay. We found that LacCer (10 μm) stimulates endogenous superoxide production (7-fold compared with control) in human aortic smooth muscle cells specifically by activating membrane-associated NADPH oxidase, but not NADH or xanthine oxidase. This process was inhibited by an inhibitor of NADPH oxidase, diphenylene iodonium (DPI), and by antioxidants,N-acetyl-l-cysteine (NAC) or pyrrolidine dithiocarbamate. NAC and DPI both abrogated individual steps in the signaling pathway leading to cell proliferation. For example, the p21 ras ·GTP loading, p44 MAPK activity, and induction of transcription factor c-fos all were inhibited by NAC and DPI as well as an antioxidant pyrrolidine dithiocarbamate or reduced glutathione (GSH). In contrast, depletion of GSH byl-buthionine (S,R)-sulfoximine up-regulated the above described signaling cascade. In sum, LacCer, by virtue of activating NADPH oxidase, produces superoxide (a redox stress signaling molecule), which mediates cell proliferation via activation of the kinase cascade. Our findings may explain the potential role of LacCer in the pathogenesis of atherosclerosis involving the proliferation of aortic smooth muscle cells.

Glycosphingolipid (GSL) 1 and its metabolic products have been shown to play critical roles as bioregulators of a variety of processes such as cell proliferation (1,2), cell mobility (3), and programmed cell death (apoptosis) (4). Lactosylceramide (Lac-Cer), a ubiquitous GSL, has been implicated in diverse biological functions (5). For example, we have found that LacCer exhibits a time-and concentration-dependent proliferation of aortic smooth muscle cells (ASMC) (1). Because proliferation of smooth muscle cells is considered a hallmark in the pathogenesis of atherosclerosis, we previously measured the levels of LacCer and other GSLs in human subjects who had this disease. We found that the levels of LacCer and glucosylceramide were markedly higher in the plaque and calcified plaque than in unaffected aorta from patients who died from atherosclerosis at The Johns Hopkins Hospital (6). Recently, we observed that LacCer stimulated the activation of p44 mitogen-activated protein kinase (p44 MAPK) and the expression of the transcription factor c-fos (7), which perhaps regulates the genes essential for cell proliferation. Moreover, upstream activators of p44 MAPK; p21 ras , Raf, and MEK2, but not MEK1, are involved in the activation of p44 MAPK by LacCer. Although our findings suggested that LacCer stimulated this kinase cascade, it is unclear whether LacCer itself or second messengers generated by LacCer are responsible for the activation of this signaling cascade.
A class of highly diffusible and ubiquitous molecules, termed reactive oxygen species, has recently been recognized to act as signaling intermediates for cytokines including interleukin-1 and tumor necrosis factor-␣ (8,9). The reactive oxygen species (ROS) encompass species such as superoxide (O 2 . ), hydrogen peroxide, nitric oxide, and hydroxyl radicals (10). Oxidative stress, which is an excess production of ROS, plays a role in different pathological conditions such as atherosclerosis and cancer (11,12). In addition, O 2 . has numerous effects on cell function including induction of growth, regulation of kinase activity, and inactivation of endothelial derived relaxation factor, nitric oxide (13,14). Thus superoxide and its metabolites can function as intracellular and intercellular second messengers, transducing receptor stimulation into biochemical response. Because they are very small, rapidly diffusible, and highly reactive, free radical and redox stress are now thought to participate in cellular signaling (13,15,16). Current evidence indicate that the different stimuli use reactive oxygen species as signaling messengers to activate transcription factors and induce gene expression (17,18). The functional role of GSLs, particularly LacCer, in generating free oxygen radicals has not been reported to the best of our knowledge. In the present study, we found that LacCer stimulated the generation of O 2 . via activation of NADPH oxidase in H-ASMC. Alteration in the redox status by the LacCer-dependent production of O 2 .
stimulated the loading of GTP to Ras, p44 MAPK activation, c-fos expression, and cell proliferation.

MATERIALS AND METHODS
Isotopes and Chemicals-[␥-32 P]ATP (6000 mCi/mmol), [␣-32 P]dCTP (3000 Ci/mmol), and [ 32 P]orthophosphoric acid (H 3 PO 4 ) (carrier-free) and [ 3 H]thymidine were purchased from Amersham Life Science Inc. Diphenylene iodonium (DPI) obtained from LC Laboratories. Glycosphingolipids and all other chemicals were purchased from Sigma. Bovine erythrocyte membrane-derived LacCer was a gift from Prof. T. Taki (Department of Biochemistry, Tokyo Medical and Dental University, Japan). Human plaque intima-derived LacCer was prepared in our laboratories. The purity of glycosphingolipids (Ͼ99%) was assessed by high pressure liquid chromatography and/or high pressure thin layer chromatography. Anti-p21 ras antibody and anti-MAPK antibody (Specific for p44 MAPK and p42 MAPK) were obtained from Upstate Biotechnology, Inc. (Lake placid, NY). cDNAs for c-fos and glyceraldehyde-3-phosphate dehydrogenase were a generous gift from Prof. Daniel Nathans (The Johns Hopkins University) and Dr. D. Dewitt (Department of Biochemistry, Michigan State University), respectively. The polyethyleneimine TLC plates were purchased from E. M. Separations (Gibbstown, NJ).
Isolation of LacCer from Human Atherosclerotic Plaque-Isolation, purification, characterization, and quantitation of LacCer from human atherosclerotic plaque intima was pursued according to the standard published protocol in our laboratory (6).
Vehicle for Glycosphingolipids-Stock solution of LacCer were prepared in chloroform-methanol (1:2, v/v), dried under a stream of nitrogen, dissolved in dimethyl sulfoxide (Me 2 SO) and added to culture medium to give the desired concentrations of LacCer. Cells incubated with 0.01% Me 2 SO served as a control. PMA, staurosporine, or DPI stock solutions were prepared in Me 2 SO and stored at Ϫ20°C until use. Aqueous solutions of NAC and allopurinol were prepared either in medium or in buffer. produced at each time point was calculated by comparison with a standard curve generated using xanthine/xanthine oxidase. GSH Assay-H-ASMC were incubated with 10 M LacCer, and control cells were incubated with Me 2 SO (final concentration 0.01%). In some experiments, cells were preincubated with 100 g/ml BSO, an inhibitor of de novo GSH synthesis, followed by incubation with 10 M LacCer. At various time intervals cells were harvested, and cell pellets were suspended in a solution of 25% metaphosphoric acid and 100 mM potassium phosphate buffer (pH 8.0), sonicated for 10 min, and centrifuged at 30,000 ϫ g for 30 min at 4°C. The supernatant was used for the GSH assay fluorometrically at 420 nm (excitation wave length 350 nm) using o-phthalialdehyde (Sigma) as a fluorescence reagent (21).

Measurement of Superoxide Production in
Cell Fractionation and NADH/NADPH Oxidase Assay-Confluent H-ASMC were incubated with 10 M LacCer. At different time intervals, cells were washed with ice-cold phosphate-buffered saline, harvested, and homogenized in lysis buffer containing 20 mM potassium phosphate buffer, pH 7.0, 1 mM EGTA, 10 g/ml aprotinin, 0.5 g/ml leupeptin, 0.7 g/ml pepstatin, and 0.5 mM phenylmethylsulfonyl fluoride. The membrane (both plasma membrane and mitochondrial membrane) and cytosol was prepared by centrifugation of cell homogenate at 29,100 ϫ g for 20 min at 4°C (22). Membrane was resuspended in the original volume of lysis buffer. NADH and NADPH oxidase activity was measured in both cytosolic and membrane fraction as described previously by the lucigenin chemiluminescence method (22). Briefly, the reaction mixture contained 50 mM phosphate buffer (pH 7.0), 1 mM EGTA, 150 mM sucrose, and 500 M lucigenin as the electron acceptor and either 100 M NADPH or 100 M NADH as an electron donor. The reaction was initiated by the addition of membrane homogenate (150 -200 g of protein). Luminescence was monitored as described above. In some experiments, NADPH oxidase activity was measured in the membrane preparations in the presence of 1 mM KCN (mitochondrial poison). Protein content was measured by the method of Lowry et al. (23) with bovine serum albumin serving as a standard.
Western Blot Analysis-Immunocomplexes, prepared as described above, were subjected to electrophoresis in 12.5% SDS-polyacrylamide gel electrophoresis. The protein was then transferred onto a polyvinylidine membrane and blotted with anti-MAP kinase antibody as described previously (7).
Ras⅐GTP Loading Assay-H-ASMC were metabolically labeled with [ 32 P]orthophosphate in phosphate-free media for 16 h as described previously (7) and incubated with 10 M LacCer with or without antagonists. At different time points, cells were lysed in radioimmune precipitation lysis buffer and cell lysates were immunoprecipitated with human p21 ras antibody (Upstate Biotechnology). Immunoprecipitates were washed, and bound nucleotides (GTP and GDP) associated with p21 ras were eluted with 2 mM EDTA, 2 mM dithiothreitol, 0.2% SDS, 0.5 mM GTP, and 0.5 mM GDP at 68°C for 20 min as described (25). The eluted nucleotides were separated on polyethyleneimine TLC plates using 0.75 M KH 2 PO 4 (pH 3.4) as a solvent and then exposed to an x-ray film.
Northern Blot Analysis of c-fos Expression-Total cellular RNA was isolated from H-ASMC that had been preincubated with 15 mM NAC for 30 min, 5 M diphenylene iodonium for 30 min, 100 g/ml L-buthionine (S,R)-sulfoximine (BSO) for 4 h, followed by stimulation with 10 M LacCer for 1 h prior to RNA extraction (26). Twenty micrograms of total RNA were separated by electrophoresis on 1% agarose gel (25 mM MOPS, pH 7.8, 1 mM EDTA, 1% (w/v) formaldehyde), transferred to a -probe blotting membrane (Bio-Rad), and hybridized with 32 P-labeled c-fos cDNA probes as described. As a control, the blot was stripped off and reprobed with 32 P-labeled cDNA for glyceraldehyde-3-phosphate dehydrogenase analysis and photographed.
Cell Proliferation Assay-H-ASMC were incubated with different antagonists with or without LacCer (10 M) for 24 h. Next, [ 3 H]thymidine (5 Ci/ml media) was added. After incubation for 2 h, cells were washed with phosphate-buffered saline. The incorporation of [ 3 H]thymidine in H-ASMC was measured as described previously (1).
Lactate Dehydrogenase Assay-Lactate dehydrogenase activity in the culture medium was measured as described (1). To test whether the membrane integrity of cells was lost upon incubation of cells with LacCer, we measured LDH activity in the culture medium. No increase of LDH activity was observed in culture medium up to 50 M LacCer in the incubation mixture (Fig. 2E). These data indicate that membrane integrity was intact during LacCer-(10 M) induced O 2 . generation.
Effect of LacCer and L-Buthionine (S,R)-Sulfoximine on the Intracellular Level of GSH in H-ASMC-Intracellular level of GSH was decreased in a time-dependent manner by LacCer (Fig. 4A). A low level of GSH was observed upon incubation of cells with BSO (100 g/ml) and also with BSO plus LacCer (Fig.  4A). The level of GSH was also decreased in a concentrationdependent manner in cells incubated with BSO only (Fig. 4B) . production) were further increased (about 8-fold compared with control) upon incubation of BSO-preincubated cell with LacCer (Fig. 4C).
LacCer Activates NADPH Oxidase-At various time points, following stimulation of H-ASMC with LacCer, NADPH oxidase activity was measured in a plasma membrane preparation with NADPH as a co-factor. NADPH oxidase activity in control cells (vehicle only) was 2.9 Ϯ 0.03 nmol/min/mg of protein but was increased 4-fold within 2.5 min, as compared with control (Fig. 5A). NADPH oxidase activity was 3-fold higher in cells incubated with LacCer at 10 min compared with control. The mitochondrial poison KCN did not inhibit LacCer-induced NADPH oxidase activity (data not shown). Preincubation of LacCer-stimulated plasma membrane preparations with 5 M DPI, an inhibitor of NADPH oxidase (30, 31), completely abrogated LacCer stimulated NADPH oxidase activity (Fig. 5A). No NADPH oxidase activity was observed in cytosolic fraction (  5C). An addition of exogenous LacCer after the isolation of plasma membrane preparations of nonstimulated cells did not alter NADPH oxidase activity (data not shown . generation (Fig. 6A). A moderate inhibition of  (Fig. 7B).

Effect of LacCer on Ras⅐GTP Loading in Cells Preincubated with NAC, BSO, and DPI-LacCer exerted a time-dependent
stimulation of Ras⅐GTP loading compared with control. Maximum stimulation of Ras⅐GTP loading occurred 1 min after incubation of cells with LacCer (Fig. 8A). NAC and DPI both abrogated LacCer-induced Ras⅐GTP loading (Fig. 8, B and C). In contrast, cells preincubated with BSO alone showed significant (above base-line value) (Fig. 8D) Ras⅐GTP loading. Coincubation of cells with BSO and LacCer exerted further effects on Ras⅐GTP loading within 1 min, and loading was sustained at this high level for 2.5 min (Fig. 8E). This observation indicates that free radicals are involved in LacCer-mediated Ras activation.
Effect of NAC, BSO, and DPI on LacCer-induced c-fos Expression-Northern blot assay revealed that LacCer increased the c-fos mRNA levels approximately 4-fold (densitometric scan not shown) compared with control (Fig. 9). This phenomenon was abrogated by NAC and DPI. In contrast, BSO had a moderately additive effect on the LacCer-mediated induction of c-fos mRNA levels in H-ASMC. Interestingly, LacCer did not alter the mRNA levels of c-jun and c-myc in H-ASMC (7).
Inhibition of LacCer-mediated Cell Proliferation by NAC and DPI-We found that LacCer stimulated the proliferation of H-ASMC approximately 5-fold as compared with control (Fig.  10). NAC and DPI abrogated the LacCer-mediated induction in cell proliferation. In contrast, BSO and BSO plus LacCer stimulated cell proliferation approximately 5.5-fold. DISCUSSION LacCer, a ubiquitous GSL, plays a pivotal role in the biosynthesis of complex GSL (5). However, its biological function is not well understood. Our laboratory has reported a close relationship between increased levels of LacCer and hyperproliferation in diverse human diseases. For example, in human atherosclerotic plaque (6), familial hypercholesterolemia (36,37), and human polycystic kidney disease, an increased cellular/tissue level of LacCer was accompanied by cell hyperproliferation (38). Among several GSLs investigated, we found that LacCer exerts the highest stimulation in H-ASMC proliferation (1), and LacCer from human atherosclerotic plaque tissue was significantly more effective in stimulating H-ASMC proliferation than LacCer from the unaffected aorta (6). Such studies indicated that one of the biological functions of LacCer may involve cell proliferation. Next, we showed that LacCer specifically stimulated the activation of Ras⅐GTP loading, Raf-1, and MEK2 upstream to p44 MAPK, and the expression of growth response early gene c-fos, downstream to the p44 MAPK signaling pathway (7). However, these studies did not elucidate whether LacCer itself or second messengers generated by Lac-Cer regulated this pathway. In this report we provide evidence that NADPH oxidase dependent O 2 . generation was increased in H-ASMC upon LacCer addition in a time-and concentrationdependent manner. This, in turn, activated the p21 ras ⅐GTP loading, activation of the kinase cascade, and induction of c-fos mRNA that finally led to cell proliferation. Our hypothetical model depicting LacCer-mediated redox signaling leading to the proliferation of aortic smooth muscle cells is summarized in Fig. 11 generation in intact cells is endogenous.
To determine which reactive oxygen species-generating enzymes are involved in LacCer-induced O 2 . generation, we took advantage of inhibitors known specifically to inhibit NADPH oxidase/NADH oxidase and xanthine oxidase. We found that LacCer specifically stimulated NADPH oxidase activity in a time-dependent manner, but not NADH oxidase. Moreover, O 2 . production in LacCer-stimulated intact cells was completely blocked by DPI, a flavoprotein containing NADPH oxidase inhibitor. Previously, DPI has been used to demonstrate its specific effect on the inhibition of NADPH oxidase (30,31). The nonspecific effects of DPI on other flavoproteins by direct binding have also appeared. Since mitochondrial poison KCN did not inhibit LacCer-induced NADPH oxidase activation (data not shown) it appears that the LacCer-induced O 2 . production was due to plasma membrane-bound NADPH oxidase. Allopurinol has been shown to specifically inhibit xanthine oxidase (34,35). We found that allopurinol did not inhibit LacCerinduced O 2 . production. Staurosporine is a well known potent inhibitor of PKC (32,33). Similarly, preincubation of cells with PMA (100 nM) for 24 h depletes PKC activity (32,33 9. Northern blot analysis of LacCer-induced c-fos mRNA expression in H-ASMC preincubated with NAC, DPI, and BSO. Cells were preincubated with 15 mM NAC for 30 min, 5 M DPI for 30 min, and 100 g/ml BSO for 24 h and then stimulated with 10 M LacCer for 1 h. Total RNA was extracted, and 10 g of total RNA was separated on 1% agarose gel, transferred onto a -probe blotting membrane, and hybridized with 32 P-labeled c-fos cDNA. impair the p44 MAPK activation and cell proliferation by Lac-Cer (7). Taken together, our data suggest that LacCer-induced O 2 . production in H-ASMC occurs predominantly due to PKC- To prevent oxidative damage and allow survival in an oxygen environment, mammalian cells have developed an elaborate antioxidant defense system that includes nonenzymatic antioxidants. The major nonprotein thiol and potent antioxidant in the cells is GSH, which principally buffers the intracellular redox state. We found that LacCer decreased the intracellular GSH level. BSO, an inhibitor of de novo GSH-synthesizing enzyme ␥-glutamyl cysteine synthetase, decreased basal GSH level. GSH level was further decreased upon the addition of LacCer to cells preincubated with BSO. The decrease in intracellular GSH may lead to alterations in the activity of redoxsensitive enzymes, including protein-tyrosine kinases, p21 ras , and MAP kinases (15,41). Moreover, it was reported earlier that the generation of O 2 . produced an increase in intracellular pH (40). Since p21 ras is one of the redox-sensing proteins (41), alteration of redox status by LacCer via production of endogenous O 2 . caused the activation of p21 ras by loading GTP. We have previously shown that activation of p21 ras transmits its activation signal to p44 MAPK via the Raf and MEK2 pathway (7). Moreover, it is evident from our study that depletion of GSH increased the susceptibility of p21 ras to oxidative stress generated by LacCer, whereas inhibition of O 2 . production with DPI and reduction of O 2 . level by antioxidant PDTC (data not shown) or NAC abrogated LacCer-induced p21 ras ⅐GTP loading. This observation is consistent with a previous report that O 2 .
caused activation of p21 ras in Jurkat cell lines (41). Previously, we reported that p44 MAPK activation was the target of LacCer-mediated signal transduction (7). In the present study, phosphorylation/activation of p44 MAPK was abrogated by PDTC, NAC, and DPI. Therefore, it is possible that LacCer-mediated O 2 . production via the activation of NADPH oxidase specifically phosphorylates p44 MAPK but not p42 MAPK. We were interested in whether GSH supplementation caused inhibition of p44 MAPK phosphorylation and activation. Since GSH is not transported across the membrane, we incubated cells with GSE, a readily transported derivative of GSH (42). Preincubation of GSE blocked the p44 MAPK phosphorylation/activation by LacCer. In contrast, in vivo depletion of GSH by BSO increased phosphorylation/activation of p44 MAPK by LacCer as compared with control. Thus, decreased levels of GSH potentiate the sensitivity of cells to LacCer and enhance mitogenic signaling via activation of p44 MAPK. In contrast, phosphorylation/activation of p44 MAPK but not p42 MAPK was abrogated by inhibitors of superoxide-generating enzyme DPI and antioxidant PDTC or NAC. Although the precise MAPK-dependent cellular alterations engendering a modified response to oxidants remain to be defined, the present study provides strong support for a crucial role for the p44 MAPK signaling pathway in regulating cell proliferation in response to oxidative stress induced by LacCer. The protooncogene c-fos functions as an inducible transcription factor in signal transduction processes (43). Elevated ex-pression of the c-fos gene via alteration of redox state has been previously shown to accompany cell proliferation. Also, in our previous study, LacCer specifically induced the expression of the c-fos transcriptional factor (7, 43) downstream to p44 MAPK. As expected, DPI and NAC both abrogated LacCerinduced c-fos mRNA expression. In contrast, BSO stimulated c-fos mRNA expression. These results support the notion that oxidative stress generated by LacCer induced expression of c-fos mRNA as well as cell proliferation.
Exogenously added GSLs, e.g. LacCer, to the culture medium may be incorporated into the plasma membrane, the lipophillic ceramide moiety being inserted into the lipid bilayer and thus increasing the proportion of this lipid in cell membrane (44). An increased level of LacCer also occurs in patients with hypercholesterolemia (36), perhaps contributing to an increase in the production of O 2 . in endothelial cells (45). In fact, recently, we have observed that LacCer can stimulate O 2 . production in human arterial endothelial cells via the activation of NADPH oxidase. 2 How LacCer activates NADPH oxidase is not clear from our study and requires further investigation. In summary, LacCer-mediated generation of low levels of superoxide may constitute a novel biochemical signaling pathway in H-ASMC proliferation. Our findings may explain the potential role of LacCer in the pathogenesis of atherosclerosis involving the proliferation of aortic smooth muscle cells.