UV-C Light Induces Raft-associated Acid Sphingomyelinase and JNK Activation and Translocation Independently on a Nuclear Signal*

The initiation of UV light-induced signaling in mammalian cells is largely considered to be subsequent to DNA damage. Several studies have also described ceramide (CER), a lipid second messenger, as a major contributor in mediating UV light-induced c-Jun N-terminal kinase (JNK) activation and cell death. It is demonstrated here that UV-C light irradiation of U937 cells results in the activation and translocation of a Zn2+-independent acid sphingomyelinase, leading to CER accumulation in raft microdomains. These CER-enriched rafts aggregate and play a functional role in JNK activation. The observation that UV-C light also induced CER generation and the externalization of acid sphingomyelinase and JNK in human platelets conclusively rules out the involvement of a nuclear signal generated by DNA damage in the initiation of a UV light response, which is generated at the plasma membrane.

The initiation of UV light-induced signaling in mammalian cells is largely considered to be subsequent to DNA damage. Several studies have also described ceramide (CER), a lipid second messenger, as a major contributor in mediating UV light-induced c-Jun N-terminal kinase (JNK) activation and cell death. It is demonstrated here that UV-C light irradiation of U937 cells results in the activation and translocation of a Zn 2؉ -independent acid sphingomyelinase, leading to CER accumulation in raft microdomains. These CERenriched rafts aggregate and play a functional role in JNK activation. The observation that UV-C light also induced CER generation and the externalization of acid sphingomyelinase and JNK in human platelets conclusively rules out the involvement of a nuclear signal generated by DNA damage in the initiation of a UV light response, which is generated at the plasma membrane.
Mammalian cells respond to UV light irradiation by activating a complex signaling network that implies radical oxygen species (ROS) 1 production, activation of transcription factors, and stimulation of kinases (1). Among the latter, the c-Jun N-terminal kinase (JNK), a main regulator of the AP-1 transcription factor, is considered as one of the most critical components of the UV light response. Indeed, the JNK/AP-1 pathway has been implicated in various UV light effects depending on the cellular model, including tumor promotion (2), apoptosis (3), and cell cycle arrest (4). Therefore, it is not surprising that the characterization of the signal transduction pathway leading to JNK activation has attracted a great deal of attention. From these studies, ceramide (CER), a lipid second messenger, has emerged as a major contributor in mediating UV light-induced JNK activation (5). Further studies have confirmed the general function of CER in stress-activated JNK activation and in mediating cell death (6).
The mechanism by which CER is produced upon UV light activation has been investigated. Hitherto, two main metabolic pathways have been identified for CER accumulation, namely hydrolysis from sphingomyelin (SM) through sphingomyelinase (SMase) stimulation and de novo synthesis by CER synthase activation. The latter appears not to be involved in UV light-induced CER production. Indeed, UV-A, -B, and -C light induce SM hydrolysis in most cellular models because of SMase stimulation (7), although in human keratinocytes a third, nonenzymatic mechanism of CER formation has been described (8). Despite some controversies, it appears that both neutral SMase (N-SMase) and acid SMase (A-SMase) have been implied in UV light-induced CER production and apoptosis, depending on the cellular origin and experimental conditions (9,10). However, the functional role of A-SMase, but not N-SMase, in UV light-induced JNK activation has been established (11).
Although A-SMase stimulation appears to be a critical upstream event for JNK activation, the enzyme and how it operates following UV light irradiation have not been determined. The A-SMase gene encodes for at least two forms of A-SMase produced by post-translational processing, namely a lysosomal form (L-A-SMase), which is lacking in Niemann-Pick disease, and a so-called secretory form (S-A-SMase) identified by Schissel et al. (12). S-A-SMase targets the plasma membrane, requires exogenous Zn 2ϩ for activity, and has been involved in the cellular response to inflammatory cytokines (13). Finally, a third form of Zn 2ϩ -independent A-SMase has been identified recently that is present on the cell membrane surface of CD95or CD40-stimulated cells (14). Finally, it remains to be determined what role, if any, does UV-C light-mediated DNA damage and, hence, a potential intranuclear signaling cascade have in initiating A-SMase activation.
Based on these considerations, we hypothesized that UV light irradiation may induce the activation and translocation of a Zn 2ϩ -independent A-SMase independently of a nuclear signal, resulting in CER formation in raft microdomains. We also hypothesized that CER-enriched membrane platform formation plays a role in JNK activation.

MATERIALS AND METHODS
Drugs and Reagents-Silica gel 60 thin layer chromatography plates were from Merck (Darmstadt, Germany). Aquasafe 300 scintillation mixture was purchased from EG & G Wallac (Evry, France). SR33557 was kindly provided by Dr. J. M. Herbert (Sanofi-Synthélabo, Toulouse, France), and maltose-binding protein⅐lysenin was provided by Dr. T. Kobayashi (Lipid Biology Laboratory, RIKEN, Saitama, Japan.) All other drugs and reagents, unless specified, were purchased from Sigma (St. Quentin Fallavier, France) or Alexis Biochemicals (Paris, France).
Cell Culture-The human myeloblastic cell line U937, obtained from the American Type Culture Collection (Manassas, VA) was cultured in RPMI 1640 medium at 37°C in 5% CO 2 . The culture medium was supplemented with 10% heat-inactivated fetal calf serum, 2 mM glutamine, 100 units/ml penicillin, and 100 g/ml streptomycin (all from Eurobio, Les Ulis, France). Human blood platelet concentrates were obtained from the local blood bank (Centre Régional de Transfusion Sanguine, Toulouse, France). Normal human lymphoblast cells or Niemann-Pick disease lymphoblast MS1418 was a generous gift from Prof. T. Levade (INSERM U466, CHU Rangueil, Toulouse, France).
Cell Irradiation-U937 cells were irradiated with UV-C light (254 nm) in PBS for 30 s, corresponding to 30 joules/m 2 at a concentration of 1 million cells/ml. Sphingomyelin Hydrolysis-SM quantitation was performed by labeling cells to isotopic equilibrium with 0.5 Ci/ml [methyl-3 H]choline (81 Ci/mM; Amersham Biosciences, Sarclay, France) for 48 h in complete medium. Cells were then washed and resuspended in complete medium for kinetic experiments. Radiolabeled SM was extracted and quantified by scintillation counting as described previously (15,16).
Metabolic Cell Labeling and Quantitation of Ceramide-Total cellular CER was processed by labeling cells to isotopic equilibrium with 1 Ci/ml [9,10-3 H]palmitic acid (53.0 Ci/mmol; Amersham Biosciences) for 48 h in complete medium as described previously (16). Cells were then washed and resuspended in serum-free medium for kinetic experiments. Lipids were extracted and resolved by thin layer chromatography developed in chloroform/methanol/acidic acid/formic acid/water (65: 30:10:4:2, by volume) in up to two-thirds of the plate and then in chloroform/methanol/acetic acid (94:5:5, by volume). CER was scraped and quantitated by liquid scintillation spectrometry. Lipid standards were used to identify the various metabolic products. Alternatively, total cellular CER quantitation was performed using Escherichia coli diacylglycerol kinase (Amersham Biosciences) according to previously published procedures (17).
Sphingomyelinase Activities-Each fraction (150 l aliquot) was assayed for the presence of different SMase activities (18). Volumes were adjusted to 250 l, and reactions were started by adding 250 l of substrate solution. For the measurement of the A-SMase activity, this solution consisted of [choline-methyl-14 C]SM (54.5 mCi/mol; 1 ϫ 10 5 dpm/assay (ϳ1 nmol/assay); PerkinElmer Life Sciences), 0.1% (w/v) Triton X-100, and 10 mM EDTA in 200 mM sodium acetate buffer (pH 5). For the N-SMase assay, the substrate solution consisted of [cholinemethyl-14 C]SM (1 ϫ 10 5 dpm/assay) and 0.1% (w/v) Triton X-100 in 200 mM Tris-HCl buffer (pH 7.4) containing 10 mM dithiothreitol and 10 mM MgCl 2 . After a 2-h incubation at 37°C, reactions were terminated by adding 300 l of H 2 O and 2.5 ml of chloroform/methanol (2:1; v/v). Phases were separated by centrifugation (1000 ϫ g for 5 min), and the amount of released radioactive phosphocholine was determined by subjecting 700 l of the upper phase to scintillation counting.
The amount of radiolabeled substrate that was hydrolyzed during an assay never exceeded 10% of the total amount of substrate added. For calculation of the specific activities in total cell homogenates, values were corrected for protein content, reaction time, and specific activity of the substrate.
Isolation of Membrane Raft Microdomains-Raft microdomains were isolated from cells as described (19). For each isolation, 100 ϫ 10 6 cells were washed twice with PBS. Cells were pelleted by centrifugation, resuspended in 1 ml of ice-cold MES-buffered saline (150 mM NaCl and 25 mM MES, pH 6.5), containing 1% (w/v) Triton X-100. After 30 min on ice, cells were further homogenized by 10 strokes of a Dounce homogenizer on ice. 1.5 ml of ice-cold MES-buffered saline was added, and 2 ml of this suspension was mixed with 2 ml of 80% (w/v) sucrose in MESbuffered saline. This mixture was subsequently loaded under a linear gradient consisting of 8 ml of 5-40% (w/v) sucrose in MES-buffered saline. All solutions contained the protease inhibitors 100 M phenylmethylsulfonylfluoride, 1 mM EDTA, and 1 M each aprotinin, leupeptin, and pepstatin A. Gradients were centrifuged in a Beckman SW 41 swinging rotor at 39,000 rpm for 20 h at 4°C. 12 fractions of 1 ml each were collected (from top to bottom), vortexed, and stored at Ϫ80°C. The protein content of both fractions and the total initial cell suspension were measured using bovine serum albumin as standard (20). GM1 was used as a marker of rafts (21,22). Cholesterol depletion was performed by incubating U937 cells for 30 min at 37°C in buffer A (140 mM NaCl, 5 mM KCl, 5 mM KH 2 PO 4 , 1 mM MgSO 4 , 10 mM HEPES, pH 6.5, 5 mM glucose, 0.2% bovine serum albumin (BSA) (w/v), and 10 mM methyl ␤-cyclodextrin (M␤CD)) (18).
Slot Blot-20 l of light, heavy, and raft fractions was blotted onto a nitrocellulose filter (Hybond-C, Amersham Biosciences) using a slot blot apparatus from Bio-Rad Laboratories. After blocking with 10% nonfat milk in Tris-buffered saline-Tween 20 (0.1%) for 2 h, the filter was incubated overnight at 4°C with the ␤-subunit of cholera toxin, which has an affinity for GM1. The filter was then washed, and bound proteins were detected by enhanced chemiluminescence (Amersham Biosciences).
Fluorescence-activated Cell Sorter Analysis-Cells were irradiated with 30 J/m 2 of UV-C light, fixed for 10 min in 4% paraformaldehyde (w/v) in PBS. Cells were then washed and further incubated for 45 min with a rabbit polyclonal anti-A-SMase or rabbit polyclonal anti-JNK (Santa Cruz Biotechnology) at 2 g/ml. Cells were then washed in PBS containing 1% fetal calf serum and stained for 45 min with 7.5 g/ml FITC-labeled goat anti-rabbit (Jackson ImmunoResearch Laboratories). After a final PBS-BSA wash, cells were analyzed on a fluorescence-activated cell sorter (FACScalibur, BD Biosciences).
Determination of ROS-Production of ROS was detected using a C2938 fluorescent probe (Molecular Probes). Briefly, exponentially growing cells were labeled with 0.5 M C2938 for 1 h and then irradiated with 30 J/m 2 UV-C light. The cells were washed in PBS, and cell fluorescence was determined using flow cytometry on a FACScan cytometer (BD Biosciences).
Confocal Microscopy-Cells were irradiated with 30 J/m 2 of UV-C light and incubated with 15 g/ml cholera toxin subunit B conjugated to Cy5 (Molecular Probes) for 20 min. Cells were fixed for 10 min in 4% paraformaldehyde (w/v) in PBS and washed with PBS containing 3% BSA (w/v) and 1 mM HEPES (PBS-BSA). Cells were then incubated for 45 min with either a rabbit polyclonal anti-A-SMase, mouse monoclonal anti-CER 15B4 (Alexis (Coger), Paris, France), or maltose-binding protein-conjugated lysenin followed by mouse anti-maltose-binding protein antiserum (New England Biolabs) (43,44). Cells were then washed in PBS-BSA and stained for 45 min with 200 ng/ml FITC-labeled goat anti-rabbit, anti-mouse (green emission), or Cy3-labeled goat antimouse (red emission) (Jackson ImmunoResearch Laboratories). After a final PBS-BSA wash, cells were mounted on glass coverslips with Dako mounting medium (Dako, Trappes, France). Control staining were performed with secondary antibodies alone. Slides were examined with a Carl Zeiss LSM 510 confocal microscope (Carl Zeiss, Oberkochen, Germany) using a Plan-Apochromat 63ϫ objective (1.4 oil). An argon laser at 488 nm was used to excite FITC (emission 515-540 nm), and a helium-neon laser was filtered at 633 or 550 nm to excite Cy5 and Cy3, respectively (emission 680 and 570 nm), as regulated by LSM software (Zeiss). For co-localization, images were recorded in multitracking mode, and images were obtained using IMARIS co-localization software (Bitplane, Zurich, Switzerland).
Protein Kinase Assays-Cells extracts were prepared by lysing cells in buffer containing 20 mM HEPES, pH 7.4, 12 mM EDTA, 250 mM NaCl, 1% Nonidet P-40, 2 g/ml leupeptin, 2 g/ml aprotinin, 1 mM phenylmethylsulfonyl fluoride, 0.5 g/ml benzamidine, and 1 mM dithiothreitol. Cells extracts (150 -250 g/sample) were immunoprecipitated with 0.3 g of anti-JNK (Santa Cruz Biotechnology) for 60 min at 4°C. Immune complexes were collected by incubation with protein A/G-Sepharose beads (Pierce) for 60 min at 4°C. The beads were washed extensively with lysis buffer and kinase buffer (20 mM HEPES, pH 7.4, 1 mM dithiothreitol, and 25 mM NaCl). Kinase assays were performed for 15 min at 30°C using myelin basic protein (Sigma) as a substrate for JNK activity in 20 mM HEPES, pH 7.4, 10 mM MgCl 2 , 1 mM dithiothreitol, and 10 Ci of [␥-32 P]ATP (ICN, Orsay, France). Reactions were stopped with the addition of 15 l of 2ϫ SDS sample buffer, boiled for 5 min, and subjected to SDS-PAGE (9%). Phosphorylated MBP was visualized by staining with Coomassie Blue, the dried gel was analyzed by autoradiography, and the corresponding bands were scraped and quantitated by scintillation counting.
Statistics-The Student's t test was performed to evaluate the statistical significance.

UV-C Light-induced CER Generation Through A-SMase but
Not N-SMase Activation-U937 cells were prelabeled with [9,10-3 H]palmitic acid or [methyl-3 H]choline to equilibrium for 48 h and then irradiated in PBS with UV-C light at 30 J/m 2 . As shown in Fig. 1A, UV-C light induced an increase in CER levels (which peaked at 5-15 min after irradiation; data not shown). Concomitant with CER generation we observed significant SM hydrolysis (Fig. 1B), which suggested the involvement of a SMase.
To determine which enzyme was responsible for SM hydrolysis and the subsequent CER generation, we evaluated the effect of UV-C light on both N-SMase and A-SMase activities. We observed that UV-C light did not affect N-SMase but led to a ϳ30% increase of a zinc-independent A-SMase activity, which peaked within 5-15 min after irradiation (Fig. 1, C and D). To confirm this result, U937 cells were preincubated for 1 h with SR33557, a potent A-SMase inhibitor (23), and then irradiated with 30 J/m 2 UV-C light. As expected, SR33557 significantly abrogated UV-C light induced A-SMase activity in a dose-dependent manner (Table I and Fig. 1D) and also abrogated CER generation (data not shown).
UV-C Light-induced Activation of A-SMase in Rafts-To evaluate the potential role of plasma membrane rafts in UV-C light-induced A-SMase activation, cells were irradiated with UV-C light at 30 J/m 2 . 12 min post-UV-C light irradiation (corresponding to the peak of UV light-C light-triggered A-SMase stimulation), cells were lysed in cold Triton X-100 and fractionated on a sucrose density gradient. We observed basal A-SMase activity throughout the gradient (Fig. 2, A and B).
Under UV-C light treatment, A-SMase was observed to in-crease exclusively in raft fractions. We have previously characterized U937 rafts whereby the Triton-insoluble material (fractions 4 -6) expressed high levels of SM that co-migrated, as shown in Fig. 2C, on the density gradient with the raft marker ganglioside GM1 (24). Again, no significant increase in raft-associated N-SMase activity was observed (data not shown).
To confirm the role of rafts microdomains, we pretreated U937 cells with the cholesterol-sequestering agent M␤CD under conditions where it is still possible to isolate rafts but with their content in cholesterol significantly reduced (ϳ50% decrease; data not shown) (18,24). As shown in Fig. 2D, the CER generation induced by UV-C light was significantly decreased in M␤CD-treated cells. These results strongly suggest that raft microdomains were essential for UV-C light-triggered CER generation.
UV-C Light-induced A-SMase Translocation to the Plasma Membrane Outer Leaflet-We next investigated the subcellular distribution of A-SMase in UV-C light-treated cells, UV-C light induced a time-dependent recruitment of A-SMase to the outer leaflet of the cell membrane as revealed by flow cytometry analysis performed with a FITC-coupled anti-A-SMase antibody on non-permeabilized cells (Fig. 3A). Confocal analysis of A-SMase accumulation on the cell surface was detected as early as 10 min and peaked at 12 min, after which the fluorescence signal decreased and returned to basal level after 15 min (data not shown). These results suggested that, upon UV light activation, a fraction of internal A-SMase was rapidly and transiently externalized to the cell surface. Moreover, the temporal association between A-SMase stimulation and externalization suggested that activation was a critical event for enzyme relocalization. To clarify this important issue, we evaluated the effect of SR33557 on A-SMase externalization. As shown in Fig.  3B, inhibition of the enzyme resulted in significant inhibition of its relocalization. Because SR33557 at 30 M inhibits basal SMase activity by Ͼ75% but does not lead to A-SMase degradation (23), one can rule out an artifactual event. This result does suggest, however, that for the enzyme to be externalized and reach the SM-enriched outer membrane (40), it must be in an active form. We next evaluated the effect of brefeldin A on A-SMase externalization. As shown in Fig. 3C, inhibition of the vesicular trafficking by brefeldin also abrogated the enzyme externalization.
UV-C Light-induced A-SMase Translocation to Raft Microdomains-Based on our biochemical studies, we hypothesized that upon UV-C light activation, A-SMase relocalization was not a random process but that A-SMase was redirected toward raft microdomains. We first investigated whether UV-C light activation could interfere with raft distribution by using a Cy3-coupled antibody directed against flotillin (data not  shown) or Cy5-coupled cholera toxin, which specifically binds the raft component ganglioside GM1. As shown in Fig. 4A, UV light irradiation induced a rapid and marked reorganization of rafts into larger platforms, producing a capping effect on the cells at 12 min post-irradiation (Ͻ10% GM1 clustering in con-trols (depending on the experiment) compared with Ͼ80% in irradiated cells). Moreover, in irradiated cells A-SMase colocalized with cholera toxin, suggesting that the enzyme translocated into raft microdomains (Fig. 4B). Based on these results, we hypothesized that CER production preferentially occurred at the raft level. Indeed, upon UV-C light activation CER co-localized with GM1 as shown in Fig. 4C. These results strongly suggested that plasma membrane microdomains are essential constituents in UV-C light-mediated A-SMase activation and externalization. Because SM is an essential constituent of plasma membrane rafts, we elected to investigate SM distribution in irradiated U937 cells. As shown in Fig. 4D, the SM-specific toxin lysenin (here coupled to maltose-binding protein) bound uniformly to the plasma membrane of control U937 cells, as described previously for normal fibroblasts (41). However, there appeared to be a modest redistribution after irradiation that colocalized with GM1 aggregation.
To further confirm the role of rafts in our study, we pretreated U937 cells with M␤CD. As shown in Table II, UV-C light-induced A-SMase externalization was completely inhibited in M␤CD-treated cells. Moreover, we observed, using con-

FIG. 2. UV-C light-induced activation of A-SMase in raft microdomains. A, U937 cells were either untreated (E) or treated (•)
with UV-C light at 30 J/m 2 . 12 min post-irradiation, cells were lysed in cold Triton X-100 and fractionated on a sucrose density gradient. Aliquots were collected and analyzed for A-SMase activity. Enzyme activities present in each fraction are expressed against the aliquot-associated protein levels. Bar denotes raft fractions (4 -6). Results are representative of three independent experiments performed in triplicate. *, p Ͻ 0.01. B, A-SMase activities are expressed compared against the amount of proteins in each fraction, light (pooled fractions 1-3), rafts (4 -6) and heavy (7)(8)(9)(10). Open and closed bars denote cells untreated or treated with UV-C light, respectively. Results are mean of triplicate determinations of a representative experiment (one of three independent experiments). *, Ͻ 0.01. C, rafts (pooled fractions 4 -6) were identified by the ganglioside GM1, which is enriched in this microdomain and not in the light fractions (1-3) and heavy fractions (7)(8)(9)(10). Results are representative of three independent experiments. D, U937 cells were prelabeled with [9,10-3 H]palmitic acid to equilibrium for 48 h and then cholesterol-depleted using M␤CD (f) as described under "Materials and Methods" or left as is (Ⅺ). CER levels were quantitated after UV-C light irradiation; results are representative of three independent experiments. *, p Ͻ 0.05. Non-irradiated control A-SMase values were 4966 Ϯ 86 dpm and 3970 Ϯ 175 dpm in the absence and the presence of M␤CD, respectively. focal microscopy, that both CER and GM1 redistribution in UV-C light-treated cells were inhibited (as expected) in M␤CDtreated cells (data not shown). These results suggested that UV-C light activated an ordered signaling cascade consisting of A-SMase activation and relocalization into raft microdomains, SM consumption, and CER release, resulting in the formation of large CER-enriched platforms. To further confirm this hypothesis, we compared A-SMase externalization under UV-C light treatment in normal and Niemann-Pick disease lymphoblasts. Again, UV-C light treatment induced significant A-SMase externalization in normal lymphoblasts, which was comparable with that observed in U937 cells but not in Niemann-Pick disease cells (Table II).
ROS Regulates A-SMase Activation and Translocation to Raft Microdomains-In an attempt to characterize the signaling pathways leading to A-SMase activation induced by UV-C light, we first studied the role of ROS in UV-C light signaling. Indeed, treatment of U937 cells with UV-C light resulted within 5 min in a burst of H 2 O 2 production (Fig. 5A), stimulation of A-SMase activity (Fig. 5B), and translocation of A- SMase activity to the outer leaflet of the plasma membrane (Fig. 5C). Moreover, inhibiting UV-C light-induced ROS production by PDTC abolished both A-SMase activation (Fig. 6A) and membrane externalization (Fig. 6B).

JNK Is Activated and Relocalized into Raft Fractions Following UV-C Light
Irradiation-CER has been described as triggering distinct intracellular signaling pathways, including the stimulation of JNK (5). Therefore, we evaluated whether UV-C light could activate this pathway through CER production. These experiments showed that, in U937 cells, treatment with UV-C light resulted in rapid (as early as 10 min) and prolonged (up to 24 h) JNK phosphorylation (Fig. 7A). Furthermore, the inhibition of A-SMase, by pretreatment with SR33557 (data not shown) or desipramine, another potent A-SMase inhibitor (23), blocked UV-C light-induced JNK kinase activity using the myelin basic protein as substrate (Fig. 7B). Moreover, we observed that under UV-C light treatment, JNK and a fraction of P-JNK was redistributed toward raft microdomains (Fig. 7C) and that this translocation was also inhibited by desipramine and M␤CD (data not shown). By confocal microscopy, we clearly observed P-JNK externalization in UV-C light-treated cells, a portion of which colocalized with GM1 (Fig. 7D). These results suggested that SM-derived CER, resulting in CERenriched rafts, mediated the stimulatory effect of UV-C light on these signaling kinases.

UV-C Light-induced CER Production and JNK Activation Is Not Dependent on a Nuclear
Signal-It has been suggested that cell signaling induced by UV-damaged DNA in the nucleus is rapidly transferred to the cytosol, leading to downstream events (25,26). To determine whether a nucleus is necessary for UV-C light-induced CER generation, we attempted to prepare U937 cytoplasts. However, karyophilic staining revealed that enucleation efficiency was Ͻ70%, making it impossible to interpret a clear result. Therefore, we elected to irradiate platelets with UV-C light. As shown in Fig. 8, A and B, exposure to UV-C light induced CER generation and A-SMase translocation to the external leaflet of the plasma membrane similar to that in observed in intact U937 cells. Furthermore, because platelets also express JNK (42), we were able through flow cytometry analysis to demonstrate JNK externalization in UV-C light-treated platelets (Fig. 8C). Hence, as described previously by Devary et al. (27), UV-C light-induced JNK activation was observed to be independent of a nucleus. These experiments conclusively rule out the involvement of a nuclear signal generated by DNA damage in the production of CER. DISCUSSION In this study we show that UV-C light irradiation induced a rapid and transient increase in cellular CER concentration. CER production correlated with the stimulation of an A-SMase, whereas N-SMase activity was unaffected. Moreover, SR33557 and desipramine, both inhibitors of A-SMase (23), prevented not only A-SMase stimulation but also CER generation. These results suggest that an A-SMase plays a critical role in the mediation of SM cycle activation by UV-C light. Characterized as a Zn 2ϩ -independent A-SMase, which has been shown previously to be present on the cell membrane surface of CD95-or CD40-stimulated cells (14). Gulbins and co-workers have described this enzyme as mainly residing inside secretory vesicles and reported that, upon stimulation, at least a fraction of the enzyme is rapidly translocated to the cell surface in cluster-like structures concomitant with extracellularly oriented CER (identified using the anti-ceramide 15B4 monoclonal antibody) (28). The specificities of the anti-A-SMase and anti-CER antibodies has been described previously (45), and we confirmed this by Western blot analysis (normal versus Niemann-Pick disease lymphoblasts) and thin layer chromatography immunostaining on U937 cells (data not shown). Our study revealed striking similarities between the A-SMase involved in UV-C light response and that described by Gulbins and colleagues.  a Cells were irradiated with UV-C light at 30 J/m 2 . 12 min after UV-C light irradiation, A-SMase externalization was determined by fluorescence-activated cell sorter cytometry on non-permeabilized cells using FITC-rabbit anti-A-SMase. b U937 cells were pretreated or not pretreated with M␤CD as described under "Materials and Methods," followed by UV-C light irradiation.
c Normal and Niemann-Pick disease (NPD) type A (cell line MS1418) Epstein-Barr virus-transformed lymphoid cells.
d Results are expressed as the percentage of total cell surface fluorescence compared to untreated controls and the mean of three independent experiments Ϯ S.D. e p Ͻ 0.02.
Indeed, not only was UV light-induced A-SMase stimulation found to be Zn 2ϩ -independent, but UV-C light irradiation also resulted in an increase in A-SMase localized on the outer leaflet of the cell surface. This finding suggested that, upon UV-C light stimulation, A-SMase probably translocated from the cytoplasm to the external leaflet of the plasma membrane by intracellular vesicles, as this event was found to be brefeldinsensitive (29). Because we were unable to detect any modification of the total intracellular distribution of A-SMase by confocal microscopy following UV-C light irradiation, only a very small sub-population fraction of the enzyme was implicated.
Little is known about the regulation of A-SMase in the context of stress response. ROS has been described as one of the major signaling components of the UV light response (30,31). For this reason, we investigated whether ROS could be involved in the stimulation and/or the translocation of the enzyme. Our study shows that the antioxidant PDTC, a ROS scavenger, not only inhibited A-SMase stimulation but also prevented its translocation to the cell surface. Indeed, UV light-induced ROS generation could be localized in raft microdomains (32). Of course, one cannot rule out the possibility that UV-C light induced cholesterol peroxidation; however, UV-C light did not induce ROS generation in Niemann-Pick disease cells (data not shown). Hence, our study not only shows that ROS operated upstream of A-SMase but also suggests that stimulation of the enzyme is required for its translocation to the cell surface. This hypothesis is supported by the fact that SR33557, an inhibitor of A-SMase activity, also abrogated its relocalization to the cell surface.
This study also shows that UV-C light induced significant redistribution of GM1, suggesting a major reorganization of raft microdomains. Indeed, we found that non-irradiated cells displayed dispersed GM1 distribution at the surface of the cell, whereas upon UV activation GM1 concentrated to one pole of the cell in a majority of cells. One could speculate that UV-C light induces the aggregation of raft components, leading to the formation of a major polarized signaling microdomain. Indeed, a model has been suggested in which the T cell receptor-associated signaling machinery initiates raft aggregation by promoting F-actin reorganization, permitting full activation of the tyrosine phosphorylation cascade, reorganization of the actin cytoskeleton, and sustained cell signaling and activation (33,34). It has been proposed that it is the consumption of SM in rafts by A-SMase that results in the formation of large CERenriched membrane domains which, in turn, could facilitate recruitment and activation of signaling molecules (35,36). Using the SM-specific toxin lysenin, however, we did not observe a decrease in cell membrane labeling but rather a slight redistribution of SM toward GM1-enriched regions. Further studies aimed at quantifying raft-associated SM, including a comparison of SM levels in the inner versus outer leaflet, are needed to confirm raft-associated SM-hydrolysis.
Finally, our study questions the involvement of a nuclear signal generated by UV light-induced DNA damage in the induction of A-SMase activation and CER generation. Indeed, it is generally postulated that UV light response occurs by induction of a nuclear signaling cascade by damaged DNA (25). However, Karin and colleagues proposed that UV light response (such as JNK activation) was likely to be initiated at or near the plasma membrane through alterations at the cell surface leading to receptor clustering (27,37). In our study we also observed UV-C light-induced JNK activation in U937 cells, as well as its translocation to the outer surface of raft microdomains. Furthermore, our results are consistent with the hypothesis of Karin and colleagues, as we observed that the initiation of UV cell response (i.e. CER generation, A-SMase and JNK activation, and externalization to the external plasma membrane) was similarly observed in platelets. A recent study also proposed that mitochondrial CER generation was implicated in UV signaling (38), perhaps as a consequence of mitochondrial DNA damage (39). In the field of cell response to UV light, our observations should provide new investigative paths into UV light-induced photoaging, immunosuppression, and photocarcinogenesis.