In Vitro and in Vivo Induction of Apoptosis by Sphingosine and N,N-Dimethylsphingosine in Human Epidermoid Carcinoma KB-3-1 and Its Multidrug-resistant Cells1

Sphingolipid breakdown products, including ceramide and sphingosine, regulate cell growth, cell differentiation, and apoptosis. We examined the effect of various agents, including sphingolipids, on apoptosis induction in human epidermoid carcinoma KB-3-1 and its multidrug-resistant (MDR) subclone KB-C2 cells which express P-glycoprotein. Adriamycin (ADM) induced apoptosis in KB-3-1 cells but not in KB-C2 MDR cells at the concentration of 50 g/ml. On the other hand, 15 p t sphingosine or its methylated derivative N,N-dimethylsphingosine (DM5) induced apoptosis in both cell types in vitro. These results suggested that KB-C2 MDR cells were resistant to apoptosis induction by ADM but sensitive to that by sphingosine and DM5. Ceramide and sphingosine-1-phosphate, the initial metabolites of sphingosine, failed to induce apoptosis under the same experimental condition as sphingosinefDMS. The protein kinase C (PKC) inhibitors H7 and staurosporine did not induce apoptosis in either cell line, suggesting that PKCindependent signaling is involved in apoptosis induced by sphingosine and DM5, although both sphingosine and DM5 have been shown to down-regulate PKC. Furthermore, DM5 significantly inhibited the growth of KB-3-1 as well as Received 2/2/96: revised 9/28/96; accepted 10/2 /96. The costs of publication of this article were defrayed in part by the payment of page charges. This article must thereflre be hereby marked adrertise,nent in accordance with 1 8 U.S.C. Section 1734 solely to indicate this fact. I Supported by funds from the Biomembrane Institute. it1 part under research contracts with Otsuka Pharmaceutical Co. and Seikagaku Corp.. and by National Cancer Institute Outstanding Investigator Grant CA42505 (to S-i. H.). 2 Present address: Fred Hutchinson Cancer Research Center. I I 24 CoIumbia Street, M62l. Seattle, WA 98104 and Departmetit of Pathohiology. University of Washington. Seattle. WA 98195. To whom requests (or reprints should be addressed. at Fred Hutchinson Cancer Research Center, I I 24 Columbia Street, M62 I . Seattle, WA 98104. Phone: (206) 667-2844: Fax: (206) 667-6519. KB-C2 MDR tumors in vivo, with evidence of increased apoptosis. The intracellular level of exogenously added 13H]sphingosine or I’4C]DMS did not differ between the KB-3-1 parent cell line and its MDR subclone KB-C2, whereas that of [‘4CIADM was reduced in KB-C2 MDR cells compared to KB-3-1 cells. These results suggest that P-glycoprotein acts as a transporter for ADM but not for sphingosine or DM5. Furthermore, DM5 at the concentrations which induce apoptosis in KB-C2 cells did not affect the level of [‘4C]ADM. Because sphingosine and DM5 induce apoptosis regardless of P-glycoprotein expression, they may provide a new strategy and a promising approach to the treatment of anticancer drug-resistant cancer. INTRODUCTION Sph4 has been shown to modulate transmembrane signaling through PKC-dependent or -independent pathways ( 1-4). We have demonstrated that treatment of human promyelocytic leukemia HL-60 cells with a phorbol ester increased the endogenous levels of ceramide and Sph and resulted in induction of differentiation and apoptosis (5. 6). Furthermore, Sph miniicked the phorbol ester in inducing apoptosis. Similarly. treatment of human neutrophils with tumor necrosis factor rt caused apoptosis and increased endogenous Sph levels (7). These results suggest that endogenous Sph production may play an important role in regulating apoptosis in those cells. A metabolically stable, methylated derivative of Sph, DMS. was also shown to induce apoptosis in HL-60 cells and other solid carcinoma cells (8) and to inhibit tumor growth in vitro and in viva (9). It has been increasingly recognized that many anticancer agents act by inducing apoptosis (10-12), and that insensitivity to apoptosis induction by certain agents may be due to multidrug resistance ( I 0, 1 1 ). Overexpression of a membrane glycoprotein termed the P-glycoprotein is known to be one cause of multidrug resistance ( 13, 14) by its action as a pump molecule which transports hydrophobic anticancer agents out of the cells. Pglycoprotein is normally expressed in the epithelia of excretory organs and the endothelia of the blood-brain barrier, suggesting that it provides a protective mechanism against cndogenous toxins produced metabolically and exogenous toxins present in the diet and environment ( 15). Sph has cationic amphophilic properties which suggest it may serve as a substrate for Pglycoprotein (2). In this study. we examined whether Sph or 4 The abbreviations used are: Sph. sphingosine: PKC. protein kinase C: DMS. N.N-dimethylsphingosine: MDR. multidrug resistant: TUNEL. terminal deoxynucleotidyltransferase-mediated dUTP-hiotin nick end labeling: MU’. 3-(4.5-dimethylthiazol-2-yl )-2.5-diphenyltetrazolium bromide: ADM. Adriamyciti. Research. on September 21, 2017. © 1997 American Association for Cancer clincancerres.aacrjournals.org Downloaded from 258 Sphingosine/DMS-induced Apoptosis in MDR Cells DM5 would induce apoptosis in MDR cells with high P-glyeoprotein expression by using the epidermoid carcinoma cell line KB-3-l and its MDR subelone KB-C2. We further examined the capability of the P-glyeoprotein-expressing cells to pump these sphingolipids out of the cell using ADM for comparison. MATERIALS AND METHODS Materials. DM5 ( I 6) and Sph1-phosphate ( I 7) were synthesized as described previously. [‘4C]DMS was prepared from Sph with ‘4CH3I and K2C01 reacted in a methanol solution. The following materials were purchased from the indicated suppliers: [3HISph (New England Nuclear, Boston, MA), [‘4C]doxorubiein hydrochloride (Amersham Corp., Arlington Heights, IL), biotin-21-dUTP 3’ end labeling kit (Clontech, Palo Alto, CA), avidin-horseradish peroxidase (ICN, Costa Mesa. CA), and Resolution TLC, autoradiography enhancer spray (L. M. Corp., Chestnut Hill, MA). All other reagents were purchased from Sigma Chemical Co. Cell Culture. Human epidermoid carcinoma KB cells were obtained from Dr. M. M. Gottesman (NIH, Bethesda, MD). KB-3-1 cells were grown in DMEM containing 10% heat-inactivated fetal bovine serum supplemented with 100 units/ml penicillin G, 100 p.g/ml streptomycin, 2 mtvi L-glutamine, and I % sodium pyruvate. KB-C2 cells which express P-glycoprotein were originally selected from KB-3-l cells with increasing concentrations of colehicine and maintained in 2 p.g/ml colehicine (18, 19). KB-C2 cells were cultured without colchicine 2 or 3 days before the experiments. Analysis of DNA Fragmentation. Cultured cells (4 X 10”) were treated with 15 p.M C2-eeramide, DM5, Sph, or Sph1-P, or with 50 p.M H7 or 50 p.g/ml ADM for 15 h at 37#{176}C. All sphingolipids were stored in stock solution in 50% ethanol and diluted directly into culture media (final ethanol coneentration < 0. 1 %). Control experiments were performed with ethanol (0. 1 %) as the vehicle. Cells were harvested, washed, and incubated in 10 mM Iris-HC1, 100 m i NaCl, 25 m i EDIA, and 0.5% SDS (pH 8.0) with 500 p.g/ml proteinase K at 50#{176}Cuntil the mixture became clear. DNA was extracted as described previously (20) by phenol/chloroform and incubated with 0.3 mg/ml RNase A for 1 h at 37#{176}C. After precipitation, DNA samples (-2 p.g) were eleetrophoretically separated on a I .5% agarose gel containing ethidium bromide. DNA was visualized by UV light and the gel was photographed with a Polaroid camera. Measurement of Apoptosis by Flow Cytometry. Cultured cells (4 X l0 ) were treated with different agents as described above. Apoptotie cells were quantitatively evaluated by flow cytometry as described previously (8). After 400 X g centrifugation, the resultant cell pellet was suspended in 0.3 ml hypotonie fluorochrome solution containing 50 p.g/ml propidium iodide, 0.1% sodium citrate, and 0.1% Iriton X-I00. Samples were placed overnight in the dark at 4#{176}C, and then the fluorescence of individual nuclei was analyzed using an EPICS flow cytometer (Coulter Electronics, Hialeah, FL). Metabolism of [3H]Sph and [‘4C]DMS in KB Cells. KB-3-l or KB-C2 cells were plated at 2 X l0 cells in 0.5 ml media in 24-well plates and incubated at 37#{176}C overnight. Cells were treated with S p.M Sph including 9 nM [3H]Sph (0. 1 p.Ci) or t’4CIDMS (2 X i0 epm). At each indicated time, the media were removed and the reaction was terminated by the addition of I .875 ml of ice-cold ehloroformlmethanol/concentrated HCI (100:200:1). Lipids were extracted according to the method of Bligh and Dyer (21). Samples from the lower chloroform phase were dried, resuspended in small volumes of chloroformlmethanol (2: 1) solution, and applied to silica gel high-performance thin-layer chromatography plates ( Merck, Darmstadt, Germany). The plates were developed in butanol/acetic acid/water (3: 1 : I ), and bands were visualized by primulin and identified under UV light. After treatment of TLC plates with enhancer, autoradiography was performed with Kodak X-OMAT film at -80#{176}Cfor 2 to 3 days. Radioactive areas were scraped and counted by a liquid scintillation counter. Accumulation of [‘4C]ADM in KB Cells. Cells (3 X 105/well in 24-well plates) were washed with Dulbeeco’s PBS and incubated in 0.5 ml serum-free DMEM containing 50 msi HEPES with or without DMS for 30 mm at 37#{176}C.Cells were then incubated with 1.5 p.M [‘4C]ADM (0.075 p.Ci) for I h at 37#{176}C. After incubation, the cells were trypsinized and washed with PBS and the radioactivity was counted. Cell Viability by MIT Assay. Cells (8 X 103/well in 96-well plates) were cultured with various concentrations of DM5, ADM, or DM5 plus ADM for 24 h at 37#{176}C. ell viability was assessed by a tetrazolium dye uptake assay (MIT) as described previously (22). Liposome Preparation. Liposomes were prepared as deseribed previously (23). Briefly, egg phosphatidyleholine/cholesterol/DMS (4.5:4.5: 1 , molar ratio) was dried under N, stream, hydrated in PBS, and sonicated by a bath-type sonicator (Laboratory Supply, Hicksville, NY) twice for 10 mm with a 30-mm interval. Tumor Growth Inhibition in Vivo by Different Agents. The effect of DMS on tumor growth in vita was examined and compared with that of other agents such as SphI -phosphate and ADM. Animals used for the present study were 8to 9-week-old female BALB/e nude mice. KB-3-1 or KB-C2 cells (2-5 X 10”) were injected into the s.c. tissue in the backs of nude mice. Tumor volume was calculated with the formula (tumor length X width X height) X 0.5. When tumors reached the size of 100-200 mm3. treatment was initiated by i.p. injection of different agents. DM5 and Sph-l-phosphate, each at 0.5 mgI mouse, were dissolved in 1.25% DMSO and ADM was in 0.9% NaC1. DM5 or Sph1 -phosphate was administered three times at 8-15-h intervals on days 0 and 1. ADM at 0.17 mglmouse (8 mg/kg) was administered on day 0. Identification of Apoptotic Cells in Vivo Using TUNEL Method. lumor samples were removed 15 h after i.p. injeetion of 0.5 mg free DMS and fixed in 10% buffered formaldehyde. Apoptotie cells were stained for apoptosis using the TUNEL method as described previously (24). Tissue sections were incubated with 20 p.g/ml proteinase K for I 5 mm at room temperature followed by blocking of endogenous peroxidase with H,O,. Sections were incubated with 0.6 units/mI terminal deoxynueleotidyltransferase and 0.75 p.M biotin-dUTP in terminal deoxynueleotidyltransferase buffer [30 mvi Tris-HC1 (pH 7.2), 140 mM sodium cacodylate, and 1 mtvi CoC1,) for 90 mm at 37#{176}C. The reaction was terminated with TB buffer (30 msi sodium citrate and 300 msi NaCI) and incubated with 4% BSA Research. on September 21, 2017. © 1997 American Association for Cancer clincancerres.aacrjournals.org Downloaded from