Overcoming Resistance to C-Rays in Squamous Carcinoma Cells by Poly-Drug Elevation of Ceramide Levels

Overcoming resistance to c-rays in squamous carcinoma cells by poly-drug elevation of ceramide levels

Gersende Alphonse1, Clara Bionda1, Marie-The´ re` se Aloy1, Dominique Ardail1, Robert Rousson1 and Claire Rodriguez-Lafrasse*,1

1Department of Biochemistry, INSERM U189, Lyon-Sud Medical School, BP12, 69921 Oullins Cedex, France.

Recent strategies to sensitize radioresistant tumours of pro- or antiapoptotic proteins (Solary et al., 2001), are based on combining c-irradiation with inducers whereas, previously, cancer therapies have usually of apoptosis. We report that the combination of involved nonselective agents that induce DNA damage three inhibitors of sphingolipid metabolism, DL-threo-1- or interfere with DNA repair. Shifting the balance to the phenyl-2-decanoylamino-3-morpholino-1-propanol.HCl(DL- proapoptotic pathway may increase apoptotic cell kill- PDMP) þ imipramine7D-erythro-2-(N-myristoylamino)- ing and represents a potential improvement in cancer 1-phenyl-1-propanol (D-MAPP), with 10-Gy irradiation therapy. triggers both mitotic and apoptotic killing in radio- Several studies have attempted to overexpress the resistant SQ20B squamous carcinoma cells. In these cells, proapoptotic proteins involved in the formation of apoptosis is defective due to a lack of ceramide generation apoptosomes, such as Apaf-1 and caspase-9, which led upstream, which cannot be explained by sphingomyelinase to a clear increase in the radiosensitivity of U-373MG (neutral and acidic) deficiency or rapid derivation to the glioma cells (Shinoura et al., 2001). Other approaches, sphingolipid pathway. We present evidence of a functional like the overexpression of Bax (Lee et al., 1999; Lin et al., transduction death pathway when ceramide generation is 2001) or p21 WAF1/CIP1 antisense oligonucleotides restored, which involves the mitochondrial-mediated path- (Kokunai et al., 2001), also enhanced radiation-induced way coupled to alterations in redox status and to executive apoptosis. caspases activation. The poly-drug treatment restored It has been proposed that ceramide is a proapoptotic apoptosis to levels similar to those observed in radio- lipidic messenger, involved in the response to various sensitive SCC61squamous carcinoma cells. Simultaneous stimuli in a variety of normal and neoplastic cells (Perry, exposure to c-irradiation and poly-drug treatment acted 1999; Kolesnick and Fuks, 2003). In Jurkat leukaemia synergistically in SQ20B cells to produce a marked cells and SCC61 squamous carcinoma cells sensitive to increase in both mitochondrial dysfunction and caspase anti-Fas and g-irradiation, we demonstrated that cleavage, which led to a 7.8-fold increase in apoptosis ceramide is a determining factor in the commitment within 48 h, relative to irradiated cells. Moreover, the phase of apoptosis (Rodriguez-Lafrasse et al., 2001; results suggest that the ceramide released by irradiation or Alphonse et al., 2002). Ceramide acts upstream from poly-drug treatment converges upon common cellular and conveys the apoptotic signal to the mitochondria, targets. Modulation of endogenous ceramide levels by leading to the subsequent activation of executive inhibitors of sphingolipid metabolism may represent a new caspases. In resistant SQ20B adenocarcinoma cells, the cellular target for the sensitization of radioresistant lack of ceramide generation in response to both anti-Fas tumours to c-ray therapy. and g-irradiation makes the whole pathway ineffective Oncogene (2004) 23, 2703–2715. doi:10.1038/sj.onc.1207357 and leads to radioresistance (Alphonse et al., 2002). Published online 29 March 2004 Several works have also correlated defective ceramide generation or signalling with resistance to radiation- Keywords: SQ20B cells; SCC61 cells; g-irradiation; induced apoptosis (Chmura et al., 1997a, b; Bruno et al., apoptosis; inhibitors of sphingolipid metabolism 1998; Sautin et al., 2000). Moreover, Selzner et al. (2001) demonstrated that ceramide levels are reduced by more than 50% in human colon tumours relative to those of healthy colon mucosa collected from the same patients. Introduction In a recent study, Riboni et al. (2002) demonstrated an inverse correlation between the malignant progression New therapeutic strategies attempt to amplify or restore of human astrocytomas (low- and high-grade tumours) the apoptotic death pathway by targeting the expression and the ceramide content measured in surgical speci- mens. They also found that low ceramide levels, which confer rapid growth and apoptotic-resistant features to *Correspondence: C Rodriguez-Lafrasse; tumours, were associated with low patient survival. That E-mail: [email protected] Received 30 April 2003; revised 7 November 2003; accepted 12 study thus supports the potential benefits of ceramide- November 2003 based chemotherapy. Overcoming radioresistance by ceramide generation G Alphonse et al 2704 Therefore, manipulation of ceramide levels may be Results particularly relevant in the treatment of cancer (Radin, 2003), and more specifically in amplifying the efficiency Similar enzymatic activities of neutral and acid of radiotherapy. This was previously achieved either by sphingomyelinases, and glucosylceramide synthase in the addition of short-chain or natural ceramide to SQ20B and SCC61 cells in response to 10-Gy irradiation tumour cells (Sautin et al., 2000; Lozano et al., 2001; We first verified that the lack of ceramide generation in Selzner et al., 2001, Kimura et al., 2003) or by the addition of chemical agents that interfere with its radioresistant SQ20B cells does not result from enzymatic dysregulation. The activities of three enzymes metabolism in cell cultures or in xenografts in nude implicated in the metabolism of ceramide were measured mice. For instance, Chmura et al. (1997a) reversed in SQ20B cells between 1 and 24 h after irradiation and the radioresistance of the SQ20B tumour model compared with those of radiosensitive SCC61 cells. The in vitro and in vivo by increasing sphingomyelinase total cellular content of sphingosine, glucosylceramide, activity by the addition of chelerythrine. Raisova and sialic acid were also determined in SQ20B cells. et al. (2002) increased the ceramide content and The basal levels of neutral sphingomyelinase activity induced apoptosis in HaCaT keratinocytes and human were very similar in the radioresistant and radiosensitive melanoma cells by the addition of a ceramidase cell lines (Figure 1). A dose of 10-Gy irradiation induced inhibitor. We previously used a combination of three different a gradual increase in neutral sphingomyelinase activity from 8 to 24 h in both cell lines (P 0.05). No significant inhibitors of sphingolipid metabolism, DL-threo-1- o activation of acid sphingomyelinase activity was ob- phenyl-2-decanoylamino-3-morpholino-1-propanol. HCl served between 1 and 24 h after irradiation in either cell (DL-PDMP) (a specific inhibitor of glucosylceramide line. Basal levels of this enzyme were 5.6470.44 mkat/mg synthase), D-erythro-2-(N-myristoylamino)-1-phenyl-1- protein in the radioresistant cell line, and threefold propanol (D-MAPP) (an inhibitor of ceramidase), and lower in the radiosensitive cell line. No significant imipramine (an amphiphilic amine that disturbs lipid variation in glucosylceramide synthase activity was turnover in biological membranes), to amplify the observed after irradiation in either cell line. However, endogenous ceramide signal in lymphoid cell lines the basal level of this enzyme was higher in the derived from leukaemic, Niemann-Pick A, or normal subjects (Rodriguez-Lafrasse et al., 2002). We demon- radiosensitive SCC61 cells. Furthermore, the total cellular content of glucosylceramide, sphingosine (which strated that endogenously formed ceramide triggers reflects ceramidase activity), and sialic acid (which time- and concentration-dependent apoptotic cell death through the induction of mitochondrial injury reflects cellular ganglioside content) did not differ between control and irradiated SQ20B cells (Figure 2). and the activation of the caspase pathway. In this study, Taken together, these results demonstrate that the we take advantage of this pharmacological model to lack of ceramide release in the radioresistant SQ20B cell overcome the radioresistant phenotype of SQ20B line in response to 10-Gy irradiation did not result from squamous carcinoma cells. We studied the death a deficit in sphingomyelinase activity as measured under response of this SQ20B cell line (SF2: 0.82) and also in vitro conditions, from highly stimulated glucosylcer- of SCC61 cells (SF2: 0.36), a radiosensitive control cell amide activity, or from rapid derivation to the line derived from the same histological type, to the sphingolipid pathway. combined effects of 10-Gy irradiation and the combined poly-drug, DL-PDMP þ imipramine7D-MAPP. We present evidence for the usefulness of generating Radioresistance of SQ20B cells does not result from intracellular ceramide by a poly-association of inhibitors neutral sphingomyelinase inhibition by elevated of sphingolipid metabolism to potentiate the cytotoxi- glutathione levels city of radiotherapy in resistant tumours. This proceeds through the activation of both clonogenic and apoptotic As basal glutathione levels were higher in resistant death, which were not detectable in the response of SQ20B cells than in SCC61 cells (Alphonse et al., 2002), SQ20B cells to irradiation only. Moreover, we demon- we postulated that in SQ20B cells in vivo, neutral strate for the first time that, in SQ20B cells, the sphingomyelinase inhibition may be secondary to its apoptotic death pathway is present but not functional inhibition by elevated glutathione levels (Liu and in the absence of a certain threshold of ceramide Hannun, 1997). We therefore attempted to depress the generation. The enhancement of apoptotic death implies cellular glutathione content with buthionine sulphox- a marked increase in both mitochondrial dysfunction imine (BSO) treatment to overcome sphingomyelinase and caspase cleavage coupled to changes in redox status inhibition and to increase ceramide levels. (reactive oxygen species (ROS) production and total SQ20B cells were pretreated with 500 mM BSO 12 h glutathione depletion). Apoptotic cell death reached before irradiation. As shown in Table 1, treatment of levels similar to those observed in radiosensitive SCC61 control cells with BSO induced a significant decrease in cells. Modulation of endogenous ceramide levels by the total glutathione within 24 h (Po0.05) of the pretreat- combined poly-drug (DL-PDMP þ imipramine7D- ment. This decrease was associated with a burst of ROS MAPP) may thus represent a new cellular target for production (Po0.05), a decrease in mitochondrial the sensitization of radioresistant tumours to g-ray transmembrane potential (Dcm), and a subsequent therapy. increase in apoptosis. The combination of 10-Gy

Oncogene Overcoming radioresistance by ceramide generation G Alphonse et al 2705

Figure 1 Kinetics of neutral sphingomyelinase, acid sphingomyelinase, and glucosylceramide synthase activities in SQ20B (left panel) and SCC61 (right panel) cell lines exposed to 10-Gy irradiation. A time-matched control was assessed at each point. Each value represents the mean7s.d. of one experiment performed in triplicate for sphingomyelinase activities and the mean7s.d. of two separate experiments performed in duplicate for glucosylceramide synthase

Table 1 Effect of 500 mM BSO, with or without 10-Gy irradiation, on ceramide and glutathione levels, ROS production, mitochondrial transmembrane potential, and apoptotic induction in the SQ20B cell line 10 Gy BSO BSO+10Gy

% of controls 24 h 48 h 72 h 24 h 48 h 72 h 24 h 48 h 72 h

Glutathione 807286738273 1 0.5 0.5 0.7 0.5 0.3 ROS 12478 11376 13174 17773 32874 57877 46579 639714 776722 High Dcm 9572.9 94.4759575917789778375727572766774 Apoptosis 10777 11079 12176 275712 271710 313711 34477 427711 531713 Ceramide 10271 10672 11974 87.772 10372 103728876 11472 11672

BSO was added to the medium 12 h before irradiation, and the cells were collected 24, 48, or 72 h after irradiation. Each value (expressed as percent of controls) represents the mean 7 s.d. of two separate experiments performed in duplicate

Oncogene Overcoming radioresistance by ceramide generation G Alphonse et al 2706

Figure 2 Kinetics of glucosylceramide, sphingosine, and sialic acid concentrations in SQ20B cells exposed to 10-Gy irradiation. A time-matched control was assessed at each point. Each value represents the mean7s.d. of two separate experiments performed singly or in duplicate

irradiation and BSO treatment ampliﬁed these changes. Interestingly, no variation in intracellular ceramide concentration was observed in response to BSO or BSO þ 10-Gy irradiation. These experiments show that depleting the intracellular glutathione pool may induce apoptosis via the oxidative pathway, but does not inﬂuence ceramide release, suggesting a lack of neutral sphingomyelinase inhibition by elevated glutathione levels.

Overcoming resistance to clonogenic and apoptotic death by increasing endogenous ceramide levels in radioresistant SQ20B cells As a 10-Gy irradiation (Alphonse et al., 2002) did not trigger mitotic or apoptotic cell death in SQ20B cells, we tested whether the combination of DL- PDMP þ imipramine7D-MAPP with or without 10- Gy irradiation would increase endogenous ceramide levels and overcome resistance to cell death. Because kinetic studies have been previously validated in different cellular models (Alphonse et al., 2002; Rodri- guez-Lafrasse et al., 2002), ceramide was quantified in SQ20B and SCC61 cells, 24 and 48 h after treatment. In order to determine the nature of cell death involved in response to ceramide generation, the mitotic cell death was quantified using clonogenic assay, and the apoptotic death was determined by flow cytometry using TUNEL labelling and cell cycle analysis. As expected (Alphonse et al., 2002), 10-Gy irradiation did not trigger any significant ceramide release in resistant SQ20B cells at either time (Figure 3). The presence of DL-PDMP þ imipramine induced a 210712% increase in ceramide relative to control cells at 48 h (Po0.01). The combination of DL- PDMP þ imipramine with D-MAPP induced an even more pronounced elevation of ceramide, to a concentra- Figure 3 Effect of 20 mMDL-PDMP (P) þ 20 mM imipramine tion of 4.570.7 nmol/pmol P (340723.7% of control (I)75 mMD-MAPP (M) with or without 10-Gy irradiation, on cells; Po0.01) after 48 h. The combination of 10-Gy ceramide levels in SQ20B (top panel) or SCC61 (bottom panel) cell irradiation and chemical treatments induced an addi- lines. Poly-drug treatment was performed 30 min before irradiation. After 24 or 48 h, cells were collected by low-speed centrifuga- tional ceramide release producing a nearly cumulative tion and ceramide was quantified by HPLC with fluorimetric pattern, with up to 6.171.1 nmol/pmol P of ceramide detection. Each value represents the mean7s.d. of three separate (405723% of controls; Po0.01) at 48 h. In comparison, experiments performed in duplicate

Oncogene Overcoming radioresistance by ceramide generation G Alphonse et al 2707 irradiation of SCC61 radiosensitive cells induced a cells (SF2: 0.36). The pretreatment of SQ20B cells with significant rise in ceramide levels to 150 and 175% of DL-PDMP þ imipramine before irradiation resulted in control values at 24 and 48 h (Po0.05), respectively. increased cytotoxicity at all radiation doses. The profile Incubation of SCC61 cells with DL-PDMP þ imipramine of the survival curve was similar to that for the or DL-PDMP þ imipramine þ D-MAPP induced a very radiosensitive SCC61 cells (SF2 of 0.33 for SQ20B cells significant increase in ceramide levels at 24 h (Po0.01) treated with DL-PDMP þ imipramine versus 0.36 for of incubation, with no further increase at 48 h (Po0.01). SCC61). Combining the poly-drug (DL-PDMP The combination of irradiation and sphingolipid in- þ imipramine þ D-MAPP) with irradiation led to a hibitors induced a marked increase but no additive further decrease in the percentage of surviving cells effect. Interestingly, ceramide levels seemed to reach a (SF2: 0.18). Interestingly, the combination of DL- maximum of about 6.0 nmol/pmol P in both cell lines, in PDMP þ imipramine with irradiation did not cause response to irradiation with chemical treatments. more significant cytotoxic effects in SCC61 cells (SF2: The dose–response curves for SQ20B and SCC61 cells 0.19). in response to g-irradiation with or without DL- The relative contribution of apoptosis to ceramide- PDMP þ imipramine7D-MAPP are presented in induced cell death was then examined (Figure 5). Figure 4. They clearly show that SQ20B cells are much Exposure of SQ20B cells to 10-Gy irradiation did not more resistant to g-irradiation (SF2: 0.82) than SCC61 induce significant apoptosis at 24 or 48 h, whereas g- irradiation of SCC61 cells led to a 280% increase in apoptotic cells at 48 h, relative to control cells (Po0.01). The addition of DL-PDMP þ imipramine or DL- PDMP þ imipramine þ D-MAPP to SQ20B cells for 48 h increased the percentage of TUNEL-positive cells by 37.676.8% (Po0.01) and 66.071.1% (Po0.01) of the total population, respectively. The triple poly-drug provoked an apoptotic cell death rate of 72.275.4% (Po0.01) in the SCC61 cell line at 48 h. Poly-drug treatment combined with irradiation increased the number of apoptotic cells to 87.375.7% at 48 h for SQ20B cells and 86.672.1% for SCC61 cells. Interest- ingly, changes in ceramide and apoptosis levels were directly correlated insofar as they reached approxi- mately the same rate in both cell lines. Because the apoptotic cell rate should be estimated by another independent technique, quantification of the sub-G1 peak after propidium iodide labelling was used to confirm the results. The percentage of cells in sub-G1 was greatly increased in both cell lines treated with DL- PDMP þ imipramine7D-MAPP with or without irradiation at both times. Despite differences in the sensitivity of the two methods, both revealed the potentiation of radiation-induced apoptosis by combined DL-PDMP þ imipramine7D-MAPP in SCC61 and SQ20B cells. These results demonstrate firstly that the apoptotic machinery is present in SQ20B multiresistant cells and that increasing endogenous ceramide levels with the poly-drug restored the defective pathway. Secondly, they clearly prove that ceramide elevation can overcome resistance to g-radiation and apoptosis.

Mitochondrial propagation of the death signal is initiated by endogenous ceramide Figure 4 Clonogenic cell death assays performed on SQ20B and To date, two conflicting hypotheses have been proposed SCC61 cell lines treated or not with 20 mMDL-PDMP (P) þ 20 mM imipramine (I)75 mMD-MAPP (M). Cells were preincubated with concerning the apoptotic death transduction pathway. DL-PDMP þ imipramine7D-MAPP for 30 min and irradiated at The first implicates both mitochondrial alteration and 0.5, 1, 2, 3, or 5 Gy. Drugs were removed after 24 h by changing the caspase cleavage, whereas the second argues that the culture medium. The surviving fraction after each treatment was death signal is independent of the mitochondria normalized to the surviving fraction in the corresponding control (plating efficiency) and survival curves were fitted using the linear and only involves caspase activation (Scaffidi et al., quadratic model. Each value represents the mean of one experi- 1998). To determine whether the ceramide pool liberated ment performed in triplicate by poly-drug treatment induced apoptosis via a

Oncogene Overcoming radioresistance by ceramide generation G Alphonse et al 2708

Figure 5 Apoptotic induction in SQ20B (top panel) and SCC61 (bottom panel) cell lines in response to 20 mMDL-PDMP (P) þ 20 mM imipramine (I)75 mMD-MAPP (M) with or without 10-Gy irradiation. At 24 or 48 h after treatment, apoptotic cells were quantiﬁed by ﬂow cytometry either by the TUNEL method or by analysis of hypodiploid cells (percentage of sub-G1 cells). Each value represents the mean7s.d. of three or two separate experiments performed in duplicate

mitochondrial-dependent or -independent pathway, we PDMP þ imipramine þ D-MAPP induced a decrease in explored the involvement of this organelle. intracellular glutathione, which reached 50% at 48 h Because a decrease in Dcm only reflects mitochondrial (Po0.01) for SQ20B cells and 63% (Po0.01) for membrane dysfunction, ROS production and total SCC61 cells. The combination of poly-drug treatment glutathione depletion were studied to clarify the and irradiation caused no further decrease in glu- intracellular redox status of the cell. tathione levels. Interestingly, the basal level of total As previously published, 10-Gy irradiation does not glutathione was twofold higher in SQ20B cells than in induce significant variation in Dcm in SQ20B cells SCC61 cells, suggesting that radioresistant cells are (Figure 6a), in contrast to SCC61 cells, of which up to more protected from oxidative stress than radiosensitive 71.4711.4% of cells had a high Dcm 48 h after cells. In parallel with the decrease in glutathione, after irradiation. Moreover, the poly-drug provoked a sig- 48 h of DL-PDMP þ imipramine þ D-MAPP treatment, nificant decrease in Dcm in SQ20B cells, in that only a massive production of intracellular ROS (Figure 6c) 65.872.8% of cells exhibited high Dcm at 48 h occurred in 2971% (Po0.01) of SQ20B cells and compared with 97.371.4% of control cells (Po0.01). 42.570.4% (Po0.01) of SCC61 cells. Combining poly- By comparison, under the same experimental condi- drug treatment with 10-Gy radiation induced an even tions, only 50.174.4% of SCC61 cells exhibited high more pronounced production of intracellular ROS in Dcm. Moreover, as previously observed for apoptotic the two cell lines, as expected. status, the addition of 10-Gy radiation to treatment with We here present evidence that the mitochondrial DL-PDMP þ imipramine7D-MAPP resulted in a cumu- apoptotic machinery can be activated in a radioresistant lative effect that increased the number of cells with high cell line when a significant endogenous ceramide signal Dcm to 1.676.7% in the radioresistant cell line and to is generated. 50.174.4% in the radiosensitive cells. As shown in Figure 6b, ceramide accumulation was Caspase activation by the elevation of endogenous associated with a significant drop in total glutathione. In ceramide the SQ20B cell line, 10-Gy irradiation induced no change in intracellular glutathione content, whereas We first demonstrated that the transduction of the death intracellular glutathione decreased to 44% at 48 h in signal, activated by endogenous ceramide in radio- SCC61 cells. Incubation of both cell lines with DL- resistant cells, involved the mitochondria. We then

Oncogene Overcoming radioresistance by ceramide generation G Alphonse et al 2709

Figure 6 Mitochondrial response in SQ20B (top panel) and SCC61 (bottom panel) cell lines to the combined effect of 10-Gy irradiation and 20 mMDL-PDMP (P) þ 20 mM imipramine (I)75 mMD-MAPP (M). Irradiation was performed 30 min after the addition of the double or triple drug combination. Variations in mitochondrial transmembrane potential (left panel) were measured by flow cytometry using the JC-1 probe. Total glutathione (middle panel) was quantified from cellular extracts by HPLC with fluorimetric detection. Intracellular ROS content (right panel) was analysed by flow cytometry using a dihydroethidium probe (HE). The relative fluorescence intensity was taken as a measure of ROS production. Each value represents the mean7s.d. of three separate experiments performed in duplicate investigated the role of caspase activation in the effective 10-Gy irradiation caused the activation of this caspase. phase of apoptosis. Global quantification of caspase The proforms of both caspases did not vary at 24 h in cleavage was first studied using flow cytometry. Im- adherent SCC61 cells, whereas at 48 h a decrease in the munoblotting was then used to determine more precisely native forms was observed in treated and irradiated the involvement of caspase-8 and caspase-3. cells. The active p43/41 and p17 forms of caspase-8 and In SQ20B cells (Figure 7), significant caspase activa- caspase-3, respectively, appeared weakly in the floating tion occurred 24 h after poly-drug treatment (Po0.01) cells after 24 h of treatment, whereas a vast increase in and increased to 694% of control cell levels at 48 h these cleavage products was observed at 48 h. (Po0.01). The percentage of caspase activation in- The overall results of the study suggest that the creased to about the level observed in SCC61 cells. ceramide released from irradiated SCC61 cells or from Consistent with this increase in ceramide, caspase SQ20B cells treated with the poly-drug derives from a activation was potentiated in both cell lines when common metabolic pool. irradiation was combined with drug treatment. The proteolytic cleavage of pro-caspase-8 (Figure 8, upper panel) and pro-caspase-3 (Figure 8, middle panel) Discussion was followed by immunoblotting of the proteins from adherent and floating cells, separately. GAPDH In the present work, we examined the combined effect of (Figure 8, lower panel) was used as a loading control 10-Gy irradiation and a poly-drug treatment (20 mMDL- to confirm changes in caspase-3 and caspase-8 levels in PDMP þ 20 mM imipramine75 mMD-MAPP) that en- adherent SCC61 and SQ20B cells. hances endogenous ceramide levels, in two head and Expression of the p55 and p54 proforms of caspase-8 neck squamous carcinoma cell lines (SQ20B and SCC61) was detectable in adherent SQ20B cells at 24 and 48 h. A with opposite radiosensitivities. Simultaneous exposure slight decrease in these proforms, without the appear- of SQ20B radioresistant cells to g-irradiation and poly- ance of cleavage products, was observed 48 h after drug treatment produced a synergistic response in terms DL-PDMP þ imipramine or DL-PDMP þ imipramine þ of mitotic and apoptotic cell death. We demonstrated D-MAPP treatments combined with 10-Gy irradiation. that increasing endogenous ceramide levels in these cells The pro- and cleaved forms of caspase-8 were detectable overcame apoptosis resistance by restoring the death in the floating population of treated cells, and associated signal transduction pathway, which had been ineffective

Oncogene Overcoming radioresistance by ceramide generation G Alphonse et al 2710 2001) apoptotic pathway. The lack of ceramide signal in SQ20B cells could result from at least two enzymatic mechanisms, a deficiency in ceramide production or an ultrarapid detoxification caused by the metabolism of ceramide into intermediate compounds such as glucosylceramide or sphingosine derivatives. We previously established (Alphonse et al., 2002; Rodriguez-Lafrasse et al., 2002), in agreement with others (Chmura et al., 1997b; Pena et al., 2000; Jaffrezou et al., 2001), that ceramide generation triggered in response to irradiation arises mainly from the sphingomyelinase-dependent catabolism of sphingomyelin. The loss of ceramide production, by neutral or acid sphingomyelinase deficiency, has previously been shown to confer resistance to g-irradiation (Santana et al., 1996; Chmura et al., 1997b; Michael et al., 1997; Bruno et al., 1998; Sautin et al., 2000). Therefore, we compared the kinetic responses of neutral and acid sphingomyelinases in SQ20B and SCC61 cell lines. From this study, we can exclude a deficiency in these two enzymatic activities in resistant SQ20B cells. Furthermore, we have demonstrated, by depleting the intracellular glutathione pool with BSO, that the lack of ceramide release was not related to any in vivo inhibition of neutral sphingomyelinase by high glutathione concentrations. This result is in accordance with the report of El-Assaad et al. (2003), who demonstrated that glutathione supplementation or depletion did not alter radiation-induced ceramide accumulation in Molt-4 leukaemia cells. A difference in glucosylceramide synthase activity in SQ20B resistant and SCC61 sensitive cells has never been demonstrated. However, increased activity might be expected because glucosylceramide synthesis is an important mechanism by which ceramide-induced apoptosis is averted. High cellular levels of glucosylcer- Figure 7 Total caspase activation measured in response to the amide have been extensively correlated with multidrug combined effect of 10-Gy irradiation and 20 mMDL-PDMP resistance (Lavie et al., 1996; Lucci et al., 1998; Liu et al., (P) þ 20 mM imipramine (I)75 mMD-MAPP (M). The two or three drugs were added to the medium 30 min before irradiation, and the 1999). Modulation of glucosylceramide synthase activity cells were collected 24 or 48 h after treatment. Caspase activation by antisense transfection reversed resistance to adria- was determined by flow cytometric analysis using the CaspACETM mycin (Liu et al., 2000), and overexpression of this FITC-VAD-FMK In Situ Marker. Each value represents the enzyme significantly inhibited doxorubicin-induced 7 mean s.d. of two separate experiments performed in duplicate apoptosis (Itoh et al., 2003). However, some authors have challenged the involvement of glucosylceramide and glucosylceramide synthase in the phenomenon of following irradiation owing to a lack of ceramide multidrug resistance. Thus, Tepper et al. (2000) estab- generation (Alphonse et al., 2002). The enhancement lished that the overexpression of glucosylceramide of apoptosis implied a marked increase in both synthase does not increase the resistance of cells to mitochondrial dysfunction and caspase cleavage. Inter- cytotoxic agents such as anti-Fas or etoposide. Further- estingly, cell death reached levels similar to those more, Veldman et al. (2003) demonstrated that both observed in radiosensitive SCC61 cells. Moreover, these genetic and pharmacological alterations in cellular results suggest that the ceramide released by irradiation ceramide glycosylation capacity failed to sensitize or poly-drug treatment converges upon common cellular melanoma cells to anticancer drugs such as doxorubicin. targets at the level of the mitochondria and the caspase The synthesis of sphingosine-1-phosphate from cera- effector. mide catabolism has also been implicated in the radio- The p53 mutation is not a major determinant of resistance phenomenon, because it can act upon caspase SQ20B radioresistance because, as observed for most activation by inhibiting the release of cytochrome c and cancers, this protein was also mutated in the radio- Smac/DIABLO from the mitochondria into the cytosol sensitive SCC61 cell line, which was derived from a (Cuvillier and Levade, 2001). We could not measure any tumour of the same histological type. On the contrary, significant difference in the kinetic study of free the absence of the p53 apoptotic pathway strengthens sphingosine, which is formed upstream of sphingosine- the role of the p53-independent ceramide (Shi et al., 1-phosphate, or in glucosylceramide levels. These two

Oncogene Overcoming radioresistance by ceramide generation G Alphonse et al 2711

Figure 8 Western blot of caspase-8 (upper panel), caspase-3 (middle panel), and GAPDH (lower panel) in SQ20B (left panel) and SCC61 (right panel) cell lines after 24 or 48 h incubation with 20 mMDL-PDMP (P) þ 20 mM imipramine (I)75 mMD-MAPP (M) with or without g-irradiation (g). Pro-caspase-8 and -3 were monitored by immunoblotting with specific antibodies. Total cellular protein (20 mg) was loaded onto an SDS–polyacrylamide gel. The native forms of caspase-8 and -3 are 55/54 and 32 kDa, respectively, and their cleavage products are 43/41 and 17 kDa, respectively. C: control cells compounds were also detected during chromatographic apoptotic death in either cellular model, as previously ceramide quantification. Because all enzymatic activities described in lymphoid cell lines (Rodriguez-Lafrasse were measured in vitro, we cannot exclude the possibility et al., 2002). Contrary to our findings, Bielawska et al. that the substrate was inaccessible to the enzyme in vivo. (1996) reported a threefold elevation in ceramide levels A worthy hypothesis for the absence of ceramide release after the addition of 20 mMD-MAPP to HL60 cells. in SQ20B cells may be a defect in the organization, Moreover, Raisova et al. (2002) reported a twofold number, or functionality of the membranous micro- elevation in the death rate after 25 mMD-MAPP domains (Jacobson and Dietrich, 1999; Kurzchalia and treatment of HaCaT keratinocytes. Recently, Choi Parton, 1999; Ramstedt and Slotte, 2002, Gulbins and et al. (2003) showed that treatment with D-MAPP Kolesnick, 2003) because the neutral (Grazide et al., induced an in vitro dose-dependent increase in apoptosis 2002) and acid sphingomyelinase activities of plasma in mouse glomerular mesangial cells and an in vivo membranes are preferentially localized in rafts (Mendez increase in the number of apoptotic cells on the mucosal et al., 2001; Gulbins and Kolesnick, 2003). surface of the duodenum in mice. DL-PDMP has also Amplification of endogenous ceramide by the inhibi- been shown to chemosensitize neuroblastoma cells to tion of its metabolism was achieved in SQ20B and taxol and vincristine (Sietsma et al., 2000) and HepG2 SCC61 cells treated with a combination of 20 mMDL- hepatoma cells to doxorubicin (Di Bartolomeo and PDMP þ 20 mM imipramine75 mMD-MAPP. The ad- Spinedi, 2001). This agent is recognized as a glucosyl- vantage of combining several drugs that interfere with ceramide synthase inhibitor (Lavie et al., 1996), but a the endogenous ceramide pathway, rather than blocking direct effect on lysosomes cannot be excluded because, one specific pathway (Raisova et al., 2002) or adding like imipramine, DL-PDMP may have lysosomotropic short-chain exogenous ceramide (Sugiki et al., 2000), properties (Rosenwald and Pagano, 1994). Recent should be in reducing the concentration of potentially studies report that damage of the lysosomal membrane toxic chemical agents and blocking different metabolic and the consequent release of lysosomal contents into routes only partially. The separate addition of these the cytosol may trigger apoptosis (Li et al., 2000). compounds did not generate ceramide release or Cathepsin D, a lysosomal protease, can either induce the

Oncogene Overcoming radioresistance by ceramide generation G Alphonse et al 2712 cleavage of Bid, which leads to cytochrome c release and et al., 2002). The present work confirms that endogen- subsequent caspase activation (Stoka et al., 2001), or ous ceramide can induce apoptosis via a caspase- trigger Bax activation, which induces the selective dependent transduction pathway, in agreement with release of mitochondrial AIF (Bidere et al., 2003). previous reports (Ito et al., 1999; Tepper et al., 1999; Imipramine could then act at the lysosomal level, but Thomas et al., 1999; Alphonse et al., 2002). It also also through modification of the lipid raft structure and demonstrates that the cascade of caspases, inactivated in functions (Yokogawa et al., 2002). SQ20B cells in response to irradiation or anti-Fas, can Of principal interest in this paper is the demonstration be switched on following the generation of an intracel- that increasing endogenous ceramide levels by inhibition lular ceramide signal. Finally, it confirms our previous of its metabolism clearly overcomes radioresistance. observation (Alphonse et al., 2002), which places Because the mode of cell death (apoptotic versus mitotic) caspase-8 activation downstream from ceramide accu- that determines cellular radiosensitivity is still unclear, mulation in response to 10-Gy irradiation. both modes have been explored in this study. We have The experiments equally suggest that the endogenous produced evidence that ceramide can induce cell killing ceramide generated by poly-drug treatment of SQ20B or by both mitotic and apoptotic mechanisms. It was SCC61 cells derives from the same intracellular pool as previously reported that increasing intracellular cera- the ceramide generated in the SCC61 cell line in mide levels, either by the inhibition of alkaline response to 10-Gy irradiation, because no difference ceramidase or by the addition of exogenous C2- could be demonstrated in the kinetics or mechanisms of ceramide, can induce growth arrest in HL60 cells the responses. (Bielawska et al., 1996) and breast tumour cells Overall, we conclude from our experimental data that (Gewirtz, 2000), respectively. In most cases, this G1 the modulation of endogenous ceramide levels may arrest further led to mitotic death (McIlwrath et al., represent a new cellular target for the sensitization of 1994). We have demonstrated that, despite the absence radioresistant tumours to g-irradiation therapy. of ceramide release, the apoptotic machinery is present in SQ20B cells and that it can be activated when the ceramide signal is restored. For the first time, we Materials and methods achieved a significant increase in ceramide levels in this Chemical and biological reagents cellular model, which was directly associated with the triggering of apoptosis. These results suggest that DL-PDMP and D-MAPP were obtained from Biomol Research ceramide must reach a definite threshold level to trigger Laboratories Inc. (Plymouth Meeting, USA); BSO and intracellular events but, on the other hand, cannot imipramine from Sigma (St Louis, MO, USA) Ceramide (type III) from bovine brain sphingomyelin and free sphingoid bases exceed an upper limit. used for calibration curves were purchased from Sigma. To date, the majority of studies based on radio- 5,50,6,60-tetrachloro-1,10,3,30-tetraethylbenzimidazolylcarbo- sensitive cell lines have reported that radiation-induced cyanine iodide (JC-1) was purchased from Molecular Probes apoptosis proceeds through changes in mitochondrial (Leiden, The Netherlands) and CaspACEt FITC-VAD-FMK function and structure (Petit et al., 1997; Daniel, 2000). from Promega (Charbonnie` res, France). Anti-caspase-8 mono- We found evidence that the absence of mitochondrial clonal antibody was supplied by Euromedex (Mundolsheim, dysfunction displayed in SQ20B cells after irradiation France), anti-caspase-3 polyclonal antibody by R&D Systems could be overcome by the generation of an endogenous (Abingdon, UK), and anti-GAPDH monoclonal antibody by ceramide signal. As a consequence of poly-drug treat- Biodesign International (Saco, Maine, USA). The secondary ment, ROS were generated in both cell lines, in antibodies used for caspase-8, GAPDH and caspase-3 immunoblots were, respectively, horseradish peroxidase agreement with Garcia-Ruiz et al. (1997), who reported (HRP)-conjugated goat anti-mouse IgG (Santa CruzBiotech- ROS production by isolated rat liver mitochondria in nology, Santa Cruz, USA), and HRP-conjugated rabbit anti- response to C2-ceramide. A significant decrease in goat IgG (Zymed, San Francisco, CA, USA). All other glutathione has also been measured. Lavrentiadou et al. chemical reagents were of analytical grade. (2001) recently reported that the ROS-mediated depletion of glutathione plays a key role in ceramide Cell culture generation by removing its inhibitory effect on neutral SCC61 and SQ20B cell lines derived from human head and et al sphingomyelinase (Liu and Hannun, 1997; Liu ., neck squamous carcinomas were kindly provided by Dr J 1998). Therefore, we propose an amplification loop in Bourhis (Institut Gustave Roussy, Villejuif, France) and Dr V which glutathione depletion is triggered by the induction Favaudon (Institut Curie, Orsay, France), respectively. of ROS and this drop in glutathione activates ceramide Cells were grown in Dulbecco’s modified Eagle’s medium formation by the activation of neutral sphingomyeli- (DMEM) and Glutamax-I (0.86 mg/ml) supplemented with nase. A sustained decrease in Dcm was also observed in 10% (v/v) heat-inactivated foetal calf serum, 0.4 mg/ml response to the poly-drug treatment of SQ20B and hydrocortisone, 100U/ml penicillin, 100 mg/ml streptomycin, SCC61 cells. Experiments conducted on different radio- and 0.1% fungizone, and were maintained at 371Cina sensitive cell lines also confirm that exogenous humidified atmosphere containing 5% CO2. (Decaudin et al., 1998; Ghafourifar et al., 1999) or endogenous ceramide (Tepper et al., 1999; Thomas et al., Irradiation and experimental procedures 1999; Rodriguez-Lafrasse et al., 2001; Alphonse At 16 h before treatment, cells were seeded in 25-cm2 culture 4 2 et al., 2002) can directly trigger the loss of Dcm (Birbes flasks at a density of 6.8 Â 10 cells/cm for SQ20B and

Oncogene Overcoming radioresistance by ceramide generation G Alphonse et al 2713 6.8 Â 104 cells/cm2 for SCC61. Cells were irradiated with a 100 000 g. The second pellet containing the plasma membrane Saturne 42 irradiator, at a dose of 10 Gy delivered at a rate of was then incubated for 2 h in 100 mM Tris (pH 7.4), 5 mM 14 2.4 Gy/min. After incubation at 371C for 24 or 48 h, the MgCl2, 0.05% Triton X-100 containing [choline-methyl- C]- cellular monolayer was scraped for the ceramide assay, lysed sphingomyelin. for Western blot analysis, or trypsinized for all other Glucosylceramide synthase activity was measured as de- experiments. The cell pellet obtained by centrifugation scribed by Matsuo et al. (1992). Briefly, about 100 mgof was either frozen (ceramide, glutathione, and protein proteins was incubated for 30 min in a reaction mixture analyses), fixed (cell cycle analysis), or analysed immediately containing 50 mM 2-[N-morpholino]ethane-sulphonic acid (ROS analysis and mitochondrial transmembrane potential (MES) buffer (pH 6.5), 5 mM MnCl2,5mM MgCl2, 0.5% measurement). 3-([3-cholamidopropyl]dimethylammonio)-1-propanesulphonate D-MAPP and DL-PDMP were stored in ethanol at À201C (CHAPS), and 20 mM uridine diphosphate-[14C]glucose. The and used at final concentrations of 5 and 20 mM, respectively. reaction was terminated by the addition of 1 ml chloroform : Imipramine was stored frozen in water and used at a working methanol (2 : 1). Extracted lipids were resolved by thin-layer concentration of 20 mM. DL-PDMP þ imipramine7D-MAPP chromatography (TLC) using migration in two mobile phases: was added directly to the culture medium and incubated for 24 the first was chloroform : methanol : water (60 : 35 : 8) for half or 48 h. When the inhibitors of sphingolipid metabolism were the length of the plate, and the second was chloroform : combined with irradiation, they were added to the culture methanol : water (65 : 25 : 4) to the end of the plate. After medium 30 min before irradiation and maintained at those autoradiography, the amount of glucosylceramide synthesized levels until the end of the experiment. was estimated by scraping the spots that co-migrated with the BSO was added to the SQ20B culture medium 12 h before unlabelled standard and then counting the radioactivity. irradiation at a final concentration of 500 mM. HPLC analysis Clonogenic cell death assay Ceramide, glucosylceramide, and free sphingoid bases were Cell survival was assessed by a standard colony-formation quantified by high-performance liquid chromatography 2 assay. SQ20B and SCC61 cells were seeded in 25-cm culture (HPLC) with fluorimetric detection, as previously described 2 flasks at densities of 60–1500 and 180–1800 cells/cm , respec- (Rodriguez-Lafrasse et al., 2001). Briefly, total cellular lipid tively, to yield 10–200 colonies after irradiation with doses of extracts were divided into three parts that were either 7 0.5, 1, 2, 3, or 5 Gy. DL-PDMP þ imipramine D-MAPP was hydrolysed for 1 h at 1001C with 1 M KOH in methanol (total added to the culture medium 30 min before irradiation. The sphingoid base content), or for 1.5 h at 371C in 0.1 M KOH in medium was changed 24 h after irradiation to remove the chloroform : methanol (2 : 1, v/v) (free sphingoid bases con- drugs. After 6 days at 371C, corresponding to six cellular tent). The long-chain bases present in those two aliquots were divisions, the cells were washed with phosphate-buffered saline derivatized with o-phthalaldehyde, then separated and quanti- (PBS), fixed with methanol, and stained with Giemsa (Sigma). fied by HPLC as described by Rodriguez-Lafrasse et al. (1994). Colonies containing more than 50 cells were scored and the The third aliquot was used to determine the total lipidic percentages of surviving cells were calculated. The surviving phosphorus content by the method of Fiske and Subbarow fraction after each treatment was normalized to the surviving (1925). fraction for the corresponding control (plating efficiency) and Total and oxidized cellular glutathione was quantified from survival curves were fitted using the linear quadratic model. cellular extracts prepared as described by Neuschwander-Tetri and Roll (1989). After derivatization with o-phthalaldehyde, Quantification of sialic acid glutathione was quantified after separation by HPLC with Cellular concentrations of sialic acid were determined by the fluorimetric detection at 340-nm excitation and 420-nm resorcinol assay. Briefly, cells were lysed in water and total emission (Rodriguez-Lafrasse et al., 2001). lipids were extracted in chloroform : methanol : water (5 : 5 : 1) for 24 h at 41C. After evaporation, lipid extracts were Flow cytometric analysis incubated for 20 min at 1001C in 400 ml of water and 400 ml The terminal transferase dUTP nick end-labelling (TUNEL) of resorcinol mixture (resorcinol 0.3%, CuSO4 0.25 mM, HCl 30%). The reaction was terminated by the addition of 0.8 ml of reaction was carried out by labelling cells with the in situ Cell butanol : butyl acetate (15 : 85). After centrifugation at 2000 g Death Detection Kit (Roche, Meylan, France). Briefly, after for 3 min, the upper organic phases were analysed by spectro- collection of both adherent and floating cells by centrifugation, 6 photometry at 580 nm and quantified by comparison with a 1–2 Â 10 cells were washed twice with PBS containing 1% standard curve. bovine serum albumin, and fixed for 1 h with 4% paraformal- dehyde in PBS. After one wash in PBS, the cells were permeabilized with a mixture of 0.1% Triton and 0.1% Enzymatic activities sodium citrate for 2 min at 41C. After another wash, cells were Acid sphingomyelinase activity was measured in vitro as labelled for 1 h at 371C in the dark with TUNEL reaction described by Vanier et al. (1980). Briefly, cells were lysed in mixture containing dUTP–fluorescein isothiocyanate (FITC) water and enzymatic activity was determined using 0.2 M and terminal deoxynucleotidyl transferase (TdT). After two sodium acetate buffer (pH 5) containing [choline-methyl-14C]- final washes, the cells were resuspended and 10 000 events were sphingomyelin (2.4 kBq/nmol; Amersham Pharmacia Biotech, analysed by flow cytometry. Orsay, France) as substrate and 0.1% Triton X-100 as The percentage of cells in sub-G1 phase in the course of the detergent. cell cycle was taken as an index of apoptosis. SCC61 Neutral sphingomyelinase activity was determined after cell (1.7 Â 106 cells) and SQ20B (1.5 Â 106 cells) cells were pelleted fractionation of plasma membranes. The cells were lysed in and fixed in ice-cold 70% ethanol for at least 24 h. The cells 20 mM Tris (pH 7.2) buffer containing 0.1% Triton X-100 and (adherent and floating together) were then washed with PBS antiprotease. After a first centrifugation at 1000 g, the pellet and incubated for 15 min in the dark at room temperature in was discarded and the supernatant centrifuged for 1 h at 0.25 ml PBS, 0.25 ml RNase (1 mg/ml), and 0.5 ml propidium

Oncogene Overcoming radioresistance by ceramide generation G Alphonse et al 2714 iodide (0.5 mg/ml). Analysis was performed on a Coulter Epixs 5% non-fat milk (TBSTM) for caspase 3 analysis. After flow cytometer. overnight incubation at 41C with the primary antibodies (anti- Mitochondrial transmembrane potential (Dcm) was mea- caspase-8 antibody was diluted to 1 : 1000 in PBSTM; anti- sured with the lipophilic cation JC-1 probe, which selectively caspase 3 to 1 : 2000 in TBST, 2% non-fat milk), the enters into mitochondria. After drug treatment and/or membranes were washed and incubated for 1 h with perox- irradiation, cells were collected by trypsination and incubated idase-conjugated secondary antibodies at a dilution of 1 : 7500 with 5 mg/ml JC-1 for 20 min at 371C. Green fluorescence for caspase 8 and 1 : 2000 for caspase 3. After the final washes, corresponding to monomer formation in the dye was proteins were visualized with chemiluminescence reagents and immediately analysed by flow cytometry. exposed to Hyperfilm. Analysis of GAPDH was performed in Intracellular ROS content was also measured in living cells, parallel using an anti-GAPDH antibody at a dilution of with the fluorescent probe dihydroethidium (HE), which 1 : 15 000 and a secondary antibody at a dilution of 1 : 7500. permeates the cell membrane and can be oxidized into red fluorescent ethidium (Eth) in the presence of the superoxide Statistical analysis anion. Cells were incubated with 4 mM HE for 20 min at 371C in the dark and analysed by flow cytometry with an FL-2 filter. Numerical data are expressed as means7s.d. The statistical Caspase activation was determined using the CaspACEt significance of differences between the values calculated for FITC-VAD-FMK In Situ Marker. After chemical and/or different experimental treatments was determined with Stu- physical treatments, cells were collected by trypsination and dent’s t-test at Po0.05. incubated at 371C in the dark with 10 mM FITC-VAD-FMK. After 20 min, a minimum of 10 000 events were analysed by flow cytometry to determine caspase activation. Acknowledgements This work was supported by the Institut National de la Sante´ et de la Recherche Me´ dicale and by the French ETOILE Western blot analysis of caspase-8, caspase-3, and GAPDH project (Espace de Traitement Oncologique par les Ions For the analysis of caspase-8, caspase-3, and GAPDH, cell Le´ gers). We thank Colette Berthet and all the technicians lysates were prepared as previously described (Alphonse et al., from the Biochemical Laboratory of the Cardiovascular and 2002). Briefly, the extracted proteins were separated by sodium Pneumological Hospital (Lyon, France) for ceramide and dodecyl sulphate–polyacrylamide gel electrophoresis (SDS– glutathione quantification, and also the technicians from the PAGE) on 12% (caspase 8) or 14% (caspase 3) polyacrylamide haematology laboratory (Lyon-Sud Hospital Center) for their gels, and then transferred onto nitrocellulose membrane. help with flow cytometry. We are grateful to the personnel of Nonspecific binding was blocked for 1 h at room temperature the Radiotherapy Department (under Dr P. Romestaing) for with PBS, 0.05% Tween, 3% non-fat milk (PBSTM) for allowing us access to the irradiators of the Lyon-Sud Hospital caspase 8 or with Tris-buffered saline (TBS), 0.05% Tween, Center.

References

Alphonse G, Aloy MT, Broquet P, Gerard JP, Louisot P, Fiske CH and Subbarow Y. (1925). J. Biol. Chem., 66, Rousson R and Rodriguez-Lafrasse C. (2002). Int. J. Radiat. 375–400. Biol., 78, 821–835. Garcia-RuizC, Colell A, Mari M, Morales A and Bidere N, Lorenzo HK, Carmona S, Laforge M, Harper F, Fernandez-Checa JC. (1997). J. Biol. Chem., 272, Dumont C and Senik A. (2003). J. Biol. Chem., 278, 11369–11377. 31401–31411. GewirtzDA. (2000). Breast Cancer Res. Treat., 62, 223–235. Bielawska A, Greenberg MS, Perry D, Jayadev S, Shayman Ghafourifar P, Klein SD, Schucht O, Schenk U, Pruschy M, JA, McKay C and Hannun YA. (1996). J. Biol. Chem., 271, Rocha S and Richter C. (1999). J. Biol. Chem., 274, 12646–12654. 6080–6084. Birbes H, Bawab SE, Obeid LM and Hannun YA. (2002). Adv. Grazide S, Maestre N, Veldman RJ, Bezombes C, Maddens S, Enzyme Regul., 42, 113–129. Levade T, Laurent G and Jaffrezou JP. (2002). FASEB J., Bruno AP, Laurent G, Averbeck D, Demur C, Bonnet J, 16, 1685–1687. Bettaieb A, Levade T and Jaffrezou JP. (1998). Cell Death Gulbins E and Kolesnick R. (2003). Oncogene, 22, Differ., 5, 172–182. 7070–7077. Chmura SJ, Mauceri HJ, Advani S, Heimann R, Beckett MA, Ito A, Uehara T, Tokumitsu A, Okuma Y and Nomura Y. Nodzenski E, Quintans J, Kufe DW and Weichselbaum RR. (1999). Biochim. Biophys. Acta, 1452, 263–274. (1997a). Cancer Res., 57, 4340–4347. Itoh M, Kitano T, Watanabe M, Kondo T, Yabu T, Taguchi Chmura SJ, Nodzenski E, Beckett MA, Kufe DW, Quintans J Y, Iwai K, Tashima M, Uchiyama T and Okazaki T. (2003). and Weichselbaum RR. (1997b). Cancer Res., 57, Clin. Cancer Res., 9, 415–423. 1270–1275. Jacobson K and Dietrich C. (1999). Trends Cell Biol., 9, 87–91. Choi MS, Anderson MA, Zhang Z, Zimonjic DB, Popescu N Jaffrezou JP, Bruno AP, Moisand A, Levade T and Laurent and Mukherjee AB. (2003). Gene, 315, 113–122. G. (2001). FASEB J., 15, 123–133. Cuvillier O and Levade T. (2001). Blood, 98, 2828–2836. Kimura K, Markowski M, Edsall LC, Spiegel S and Gelmann Daniel PT. (2000). Leukemia, 14, 2035–2044. EP. (2003). Cell Death Differ., 10, 240–248. Decaudin D, Marzo I, Brenner C and Kroemer G. (1998). Int. Kokunai T, Urui S, Tomita H and Tamaki N. (2001). J. Oncol., 12, 141–152. J. Neurooncol., 51, 111–119. Di Bartolomeo S and Spinedi A. (2001). Biochem. Biophys. Kolesnick R and Fuks Z. (2003). Oncogene, 22, 5897–5906. Res. Commun., 288, 269–274. Kurzchalia TV and Parton RG. (1999). Curr. Opin. Cell Biol., El-Assaad W, Kozhaya L, Araysi S, Panjarian S, Bitar FF, 11, 424–431. BazE, El-Sabban ME and Dbaibo GS. (2003). Biochem. J., Lavie Y, Cao H, Bursten SL, Giuliano AE and Cabot MC. 376, 725–732. (1996). J. Biol. Chem., 271, 19530–19536.

Oncogene Overcoming radioresistance by ceramide generation G Alphonse et al 2715 Lavrentiadou SN, Chan C, Kawcak T, Ravid T, Tsaba A, Rodriguez-Lafrasse C, Alphonse G, Broquet P, Aloy MT, Rasooly R and Goldkorn T. (2001). Am. J. Respir. Cell Mol. Louisot P and Rousson R. (2001). Biochem. J., 357, Biol., 25, 676–684. 407–416. Lee JU, Hosotani R, Wada M, Doi R, Kosiba T, Fujimoto K, Rodriguez-Lafrasse C, Alphonse G, Aloy MT, Ardail D, Miyamoto Y, Tsuji S, Nakajima S, Nishimura Y and Gerard JP, Louisot P and Rousson R. (2002). Int. J. Cancer, Imamura M. (1999). Eur. J. Cancer, 35, 1374–1380. 101, 589–598. Li W, Yuan X, Nordgren G, Dalen H, Dubowchik GM, Rosenwald AG and Pagano RE. (1994). J. Lipid Res., 35, Firestone RA and Brunk UT. (2000). FEBS Lett., 470, 35–39. 1232–1240. Lin J, Page C, Jin X, Sethi AO, Patel R and NunezG. (2001). Santana P, Pena LA, Haimovitz-Friedman A, Martin S, Green Anticancer Res., 21, 831–839. D, McLoughlin M, Cordon-Cardo C, Schuchman EH, Fuks Liu B and Hannun YA. (1997). J. Biol. Chem., 272, Z and Kolesnick R. (1996). Cell, 86, 189–199. 16281–16287. Sautin Y, Takamura N, Shklyaev S, Nagayama Y, Ohtsuru A, Liu B, Andrieu-Abadie N, Levade T, Zhang P, Obeid LM and Namba H and Yamashita S. (2000). Thyroid, 10, 733–740. Hannun YA. (1998). J. Biol. Chem., 273, 11313–11320. Scafﬁdi C, Fulda S, Srinivasan A, Friesen C, Li F, Tomaselli Liu YY, Han TY, Giuliano AE, Ichikawa S, Hirabayashi Y KJ, Debatin KM, Krammer PH and Peter ME. (1998). and Cabot MC. (1999). Exp. Cell Res., 252, 464–470. EMBO J., 17, 1675–1687. Liu YY, Han TY, Giuliano AE, Hansen N and Cabot MC. Selzner M, Bielawska A, Morse MA, Rudiger HA, Sindram D, (2000). J. Biol. Chem., 275, 7138–7143. Hannun YA and Clavien PA. (2001). Cancer Res., 61, Lozano J, Menendez S, Morales A, Ehleiter D, Liao WC, 1233–1240. Wagman R, Haimovitz-Friedman A, Fuks Z and Kolesnick Shi YQ, Wuergler FE, Blattmann H and Crompton NE. R. (2001). J. Biol. Chem., 276, 442–448. (2001). Radiat. Environ. Biophys., 40, 301–308. Lucci A, Cho WI, Han TY, Giuliano AE, Morton DL and Shinoura N, Sakurai S, Asai A, Kirino T and Hamada H. Cabot MC. (1998). Anticancer Res., 18, 475–480. (2001). Int. J. Cancer, 93, 252–261. Matsuo N, Nomura T and Imokawa G. (1992). Biochim. Sietsma H, Veldman RJ, Kolk D, Ausema B, Nijhof W, Biophys. Acta, 1116, 97–103. Kamps W, Vellenga E and Kok JW. (2000). Clin. Cancer McIlwrath AJ, Vasey PA, Ross GM and Brown R. (1994). Res., 6, 942–948. Cancer Res., 54, 3718–3722. Solary E, Plenchette S, Sordet O, Rebe C, Ducoroy P, MendezAJ, Lin G, Wade DP, Lawn RM and Oram JF. Filomenko R, Bruey JM, Droin N and Corcos L. (2001). (2001). J. Biol. Chem., 276, 3158–3166. Therapie, 56, 511–518. Michael JM, Lavin MF and Watters DJ. (1997). Cancer Res., Stoka V, Turk B, Schendel SL, Kim TH, Cirman T, 57, 3600–3605. Snipas SJ, Ellerby LM, Bredesen D, Freeze H, Neuschwander-Tetri BA and Roll FJ. (1989). Anal. Biochem., Abrahamson M, Bromme D, Krajewski S, Reed JC, Yin 179, 236–241. XM, Turk V and Salvesen GS. (2001). J. Biol. Chem., 276, Pena LA, Fuks Z and Kolesnick RN. (2000). Cancer Res., 60, 3149–3157. 321–327. Sugiki H, Hozumi Y, Maeshima H, Katagata Y, Mitsuhashi Y Perry DK. (1999). Biochem. Soc. Trans., 27, 399–404. and Kondo S. (2000). Br. J. Dermatol., 143, 1154–1163. Petit PX, Zamzami N, Vayssiere JL, Mignotte B, Kroemer G Tepper AD, de Vries E, van Blitterswijk WJ and Borst J. and Castedo M. (1997). Mol. Cell. Biochem., 174, 185–188. (1999). J. Clin. Invest., 103, 971–978. Radin NS. (2003). Biochem. J., 371, 243–256. Tepper AD, Diks SH, van Blitterswijk WJ and Borst J. (2000). Raisova M, GoltzG, Bektas M, Bielawska A, Riebeling C, J. Biol. Chem., 275, 34810–33487. Hossini AM, Eberle J, Hannun YA, Orfanos CE and Geilen Thomas RL, Matsko CM, Lotze MT and Amoscato AA. CC. (2002). FEBS Lett., 516, 47–52. (1999). J. Biol. Chem., 274, 30580–30588. Ramstedt B and Slotte JP. (2002). FEBS Lett., 531, 33–37. Vanier MT, Revol A and Fichet M. (1980). Clin. Chim. Acta, Riboni L, Campanella R, Bassi R, Villani R, Gaini SM, 106, 257–267. Martinelli-Boneschi F, Viani P and Tettamanti G. (2002). Veldman RJ, Mita A, Cuvillier O, Garcia V, Klappe K, Medin Glia, 39, 105–113. JA, Campbell JD, Carpentier S, Kok JW and Levade T. Rodriguez-Lafrasse C, Rousson R, Pentchev PG, Louisot P (2003). FASEB J., 17, 1144–1146. and Vanier MT. (1994). Biochim. Biophys. Acta, 1226, Yokogawa K, Ishizaki J, Ohkuma S and Miyamoto K. (2002). 138–144. Methods Find. Exp. Clin. Pharmacol., 24, 81–93.

Oncogene