Ceramide glycanase activities in human cancer cells

Bioscience Reports, Vol. 19, No. 5, 1999

Ceramide Glycanase Activities in Human Cancer Cells Manju Basu,1,2,3 Patrick Kelly,3 Peter O’Donnell,3 Maria Miguel,3 Mathew Bradley,3 Sandro Sonnino,4 Sipra Banerjee,5 and Subhash Basu2,3 Ceramide glycanase (CGase) activities have been detected in different human tumor cells (colon, carcinoma Colo-205; neuroblastoma, IMR-32; breast cancer lines, SKBr3 and MCF7). However, the level of enzymatic activity is lower in these cells compared to that present in other mammalian tissues reported before (Basu, M., Kelly, P., Girzadas, M. A., Li, Z., and Basu, S. Methods Enzymol. (in press)). The majority of CGase activity was found in the 100,000g soluble supernatant fraction isolated from all these cell lines and tissues. Using the soluble enzyme, the requirement for optimum CGase activity was found to be consistent with previous observations found for rat and rabbit tissues (Basu, M., Dastgheib, S., Girzadas, M. A., O’Donnell, P. H., Westervelt, C. W., Li, Z., Inokuchi, J. I., and Basu, S. (1998) Acta Pol. Biochim. 42:327). The CGase activities from both Colo205 and IMR-32 cells are optimum at a protein to detergent ratio of one. All the mammalian CGases, including human cancer cells, show an optimum pH between 5.5 and 5.8 in sodium acetate buffer. The CGase activities from cancer cells are found to be cationindependent; however, mercury, zinc, and copper ions seem to inhibit the enzyme activity substantially in both tumor cells lines. The mercury ion inhibition of CGase activities from all different sources indicates a possible structural homology in the CGase proteins. Radiolabeled substrates, labeled at the sphingosine double bond or at the 3-position of sphingosine without modifying double bond of sphingosine were used in this investigation. Both were active substrates with all enzyme preparations isolated from different cancer cells (apparent Km, 500 UM for nLcOse5[3H-DT]Cer and 350 UM for GgOse4[sph3-3H]Cer with Colo-205 enzyme). Structural analogues of ceramide and sphingosine (LPPMP, L-PDMP, alkylamines, and Tamoxifen) inhibited cancer cell CGase activities in vitro. KEY WORDS: ceramide glycanase; cancer cells; glycosphingolipid; sphingosine; ceramide; apoptosis; PPMP; PDMP. ABBREVIATIONS: GSL, Glycosphingolipid; GLT, Glycosyltransferases; CGase, Ceramide Glycanase; TDC, Sodium Taurodeoxycholate; DDQ, Dichloro-dicyanobenzoquinone; PPMP, l-phenyl-2-hexadecanoylamino-3-morpholino-l-propanol, HC1; PDMP, l-phenyl-decanoylamino-3-morpholino-l-propanol, HC1; GgOse4Cer, GalB13GalNAcB1-4GalB1-4Glc-Cer, Gangliotetraosylceramide; nLcOse5Cer, Gala1-3GalB14GlcNAcj31-3Galj3l-4Glc-Cer or neolactopentaosylceramide; DMS, N'N'Dimethylsphingosine; SPP, Sphingosine-1-phosphate; MDR, Multiple Drug resistance. 1

Present address: Bayer Corporation, Diagnostic Division, Elkhart, Indiana 46515. To whom correspondence should be addressed. 3 Department of Chemistry and Biochemistry, University of Notre Dame, Notre Dame, IN 46556. E-mail: [email protected]/[email protected] 4 Universita’ degli Studi di Milano, Dipartimento di Chimica e Biochimica Medica, LITA-Segrate; Via Fratelli Cervi 93, 20090 Segrate (Milano) Italy. 5 Cleveland Clinic Foundation Research Institute, 9500 Euclid Avenue, NC-2-118, Cleveland, OH 44195. 2

449 0144-8463/99/1000-0449$16.00/0 © 1999 Plenum Publishing Corporation

450

Basu et al.

INTRODUCTION Ceramide, the indispensable moiety of a glycosphingolipid or sphingomyelin has become in recent years a key molecule being implemented in various cellular processes among which apoptosis is the most central event [1–4]. Coupling of ceramide to specific signaling cascades is both stimulus and cell-type specific. Administration of short chain ceramides can mimic apoptotic cell death [1–5]. Until now the sphingomyelin cycle has been considered the major pathway for production of ceramide under physiological condition [5–7] which then leads the way to different cellular processes including apoptosis, cell differentiation, cell proliferation, stress response, and signaling responses [8–10]. In addition to the sphingomyelinase cycle, ceramide can also be formed either by acylation of sphingosine or by another catabolic pathway which involves glycosphingolipid breakdown. A little over a decade ago an enzyme, ceramide glycanase (CGase, ceramide glycanase, endoglycosceramidase), which catalyzes the one-step cleavage of a glycosphingolipid (GSL) with the liberation of ceramide and corresponding oligosaccharide, was reported from bacteria and also from lower eukaryotes [11–15]. Structural studies of GSLs have greatly benefited from the discovery of this enzyme. The first report of a mammalian CGase, however, came from our laboratory [16], followed by the characterization of CGase activity in various other mammalian tissues [17– 20]. Presence of this enzyme in mammalian systems indicates a possible involvement of GSLs in signaling cascade via the action of this enzyme. A hormonal regulation of the enzyme has been suggested since the level of CGase activity was found to vary in rat mammary tissue during gestation and lactation [17]. The notion that the mammalian CGase will emerge as an important component of the signal cascade pathway or of the apoptotic processes, like sphingomyelinase, [5–7] is a tempting speculation. MATERIALS AND METHODS Sodium taurodeoxycholate, sodium taurocholate, Triton X-100, tamoxifen, sphingosine, ceramide and other reagent grade chemicals were purchased from Sigma Chemical Company (St. Louis, MO). Tritiated Sodium borohydride (NaBT 4 ) was purchased from Du Pont Company (Boston, MA). The SG81 chromatographic papers were purchased from Fisher Scientific Products. PPMP and PDMP are gift samples kindly provided by Dr. J.-I. Inokuchi (Japan). Breast cancer cell lines (SKBr3 and MCF7) have been maintained in the laboratory of Dr. Sipra Banerjee, Cleveland Clinic Foundation, Cleveland, OH. Both Colo-205 (human colon carcinoma cells) and IMR-32 (human neuroblastoma) cell lines were purchased from ATCC and maintained in our laboratory. The tissue culture mediums, penicillin, streptomycin, and trypsin were purchased from GIBCO; and fetal bovine serum from Rehatuin, New York. The glycolipids used in this study were prepared in our laboratory according to published methods [21]. Pentaglycosylceramide (Gala13GalB1-4GlcNAcB1-3GalB1-4Glc-Cer; nLcOse5Cer) was isolated and purified from bovine blood [22]. Lacto-neo-tetraosylceramide (GalB1-4GlcNAcB1-3GalB1-4GlcCer; nLc4) was obtained from partial hydrolysis of nLcOse5Cer using fig a-galactosidase [23]. Gangliotetraosyl-ceramide (GalB l-3GalNAcB1-4GalB1-4Glc-Cer;

Mammalian Ceramide Glycanase

451

GgOse4Cer) was obtained by formic acid-hydrolysis of ganglioside GM1, isolated and purified from bovine brain. Radiolabeling of the Substrate The glycolipid substrates used for the CGase assay are being labeled using one of the following two methods. (a) The glycosphingolipid substrates were tritiated at the double bond of ceramide (3H-DT-GSL) according to the method of Schwarzman [24]. Briefly, 1 mg of nLcOse5Cer (Gala1-3GalB1-4GlcNAcB1-3GalB1-4Glc-Cer), GgOse4Cer (GalB13GalNAc/31-4Gal/31-4Glc-Cer) or any other glycolipid was suspended in tetrahydrofuran and radiolabeled with NaBT4 in the presence of palladium chloride as catalyst at room temperature for 6 to 8 hr with shaking followed by treatment with excess NaBH4 for another 4 to 6 hr. The unreacted NaBT4 was removed by passing the reaction mixture through a SepPak C-18 column. The labeled GSL was further purified by Biosil column, and specific activity was determined with a densitometric scanner before using as substrate. (b) Recently, the glycolipids have also been labeled using a second method reported by Chigorno et al. [25] where the 3-position of sphingosine is being tritiated (3-sph-3H-GSL). In this method the 3-hydroxyl of sphingosine is oxidized first in the presence of DDQ (dichloro-dicyano-benzoquinone) and anhydrous toluene followed by reduction with NaBT4. The unreacted sodium borotritide is removed using C18 Sep-Pak column followed by silica gel 200-400 mesh column chromatography. The purity of the labeled substrates is checked by thin layer chromatography (TLC). The advantage of this labeling is that the double bond of sphingosine remain unchanged while the 3-position of the sphingosine is labeled. Unlabeled corresponding glycolipids were added to each of the labeled substrates to make desired specific activity (2–25 × 106cpm/Umol) before use as a substrate. Maintenance of Tumor Cells in Culture Colo-205 cells are maintained in RPMI-1640 medium containing penicillin (100 Ug/ml) and streptomycin (100 units/ml) and 10% fetal bovine serum under 95% humidity and 5% CO2 at 37°C. The cells are subcultured using PBS/EDTA once they reach confluency. The medium is changed once or twice during each passage. Conditions for maintaining IMR-32 cells are the same except minimum essential medium (MEM) is used. Both the cell lines are harvested using PBS/EDTA once they reach confluency and stored in 0.05 M HEPES buffer, pH 7.0 containing 0.32 M sucrose, 0.001 M EDTA, and 0.1% B-mercaptoethanol. Preparation of Enzyme The harvested cells (Colo-205, IMR-32 and breast cancer cells) were homogenized in 0.32 M sucrose containing 0.05 M HEPES, pH 7.0, 0.001 M EDTA, and 0.1%

452

Basu et al.

mercaptoethanol using Polytron 10ST homogenizer. The homogenate was centrifuged at 100,000g for 1 hr and the resulting supernatant was used as the enzyme source for ceramide glycanase assay. Enzyme Assay

(a) Assay Conditions With Radiolabeled Substrates. The incubation mixture contained the following components in micromoles, unless otherwise stated, in a final volume of 50 Ul: [3H]-GSL substrate, 1 nmol (20 × 106cpm/Umol); detergent, taurodeoxycholate, 20 to 30Ug; sodium acetate buffer, pH 5.5, 10 Umol; enzyme protein up to 50 Ug. After incubation for 2 to 4 hrs at 37°C, 50 Ul each 2-propanol and hexane was added to the reaction mixture, vortexed, and spun at 5000 rpm for 5 min. The upper layer was then quantitatively spotted on SG81 paper, and descending chromatography was performed using chloroform : methanol (9:1) solvent. The cleaved labeled ceramide moved one inch behind the solvent front which was quantitatively determined using toluene scintillation-counting technique by Beckman scintillation counter [15–20]. (b) Assay Condition for Unlabeled Substrates. The method was exactly the same when unlabeled substrates were used except 30 nmole substrates were used for each experimental tube. After incubation and separating the layers by addition of heptane and propanol, each layer was chromatographed on thin layer plates. The upper layer containing cleaved ceramide was chromatographed using chloroform : methanol (9:1). Butanol:acetic acid:water (2:1:1) was used as the chromatographic solvent for the lower layer which contained cleaved oligosaccharide chains. The ceramide was identified using coomassie stain [26] while diphenylamine spray was used for the identification of oligosaccharides. RESULTS AND DISCUSSION

The substrate used for majority of the experiments reported in this article was 3-Sphingosine-labeled gangliotetraosylceramide (sph-3-3H-Gg4). The supernatant fraction obtained after centrifugation of the cell homogenates had been used as the enzyme source. The enzymatic activity had been found to be distributed in the soluble and membrane-bound fraction in the ration of 60% and 40%, respectively. Almost all of the CGase activity could be extracted from the membrane-bound fraction using sodium taurodeoxycholate detergent. The pattern of distribution of the enzyme suggests both cytosolic and membrane-bound CGase could act on the GSL substrates present under physiological conditions. Kinetic Parameters of Carcinoma CGases

(i) Requirement for CGase in Cancer Cells. The optimum conditions (as described in the text) for the CGase activities in breast cancer cells are given in Table 1. Both metastatic (SKBr3) and non-metastatic (MCF7) breast cancer cell lines showed reasonable CGase activity when tested with 3-sph-3H-Gg4 substrate under

Mammalian Ceramide Glycanase

453

the optimum conditions as observed for other sources previously [15–20]. The CGase activities from these breast cancer cells were dependent on the presence of detergent in the incubation mixture as seen in Table 1. Like CGases from other sources, the breast cancer cell CGases were also cation-independent and remain unaffected by EDTA. However, these activities were inhibited by the mercury ion, found to be a common feature for all the CGases reported so far. Further investigation is in progress to evaluate the significance of CGase activities in these metastatic and nonmetastatic breast cancer cells. The requirements for optimum CGase activities in human colon carcinoma (Colo-205) and human neuroblastoma (IMR-32) cell lines are summarized in Table 2. Like the breast cancer cells these two tumor cells were also dependent on the presence of detergent for exhibiting optimum CGase activities. Sodium taurodeoxycholate had been found to be the best detergent for either cell lines at the protein to detergent ratio between 1 to 2. As seen in Table 2 and Fig. 1, these activities were partially cation-independent in nature. However, zinc and copper ions in addition to mercury were also found to be completely inhibitory for these cancer cell lines while partial inhibition was seen with Fe3+ (Fig. 1). From the results given in Table 2, CGase activity from IMR-32 cells appeared to be more sensitive to cations than Table 1. Requirement for Human Breast Cancer Cell CGase Activities [3H]Ceramide cleaved

Condition Complete - Detergent + MnCl2* + HgCl2* + EDTA* - Enzyme

SKBr3 MCF7 (metastatic) (nonmetastatic) (nmol/mg prot/2 hr)

1.2 0.5 1.2 0.3 1.2 0.3

2.2 1.0 1.95 0.36

1.2 0.3

*(5 mM added). The incubation mixture contained the following components in micromoles, unless otherwise stated, in a final volume of 50 Ul: [3H]-GSL substrate, 1 nmol (20 × 106cpm/Umol); detergent, taurodeoxycholate, 20 to 30Ug; sodium acetate buffer, pH5.5, 10 Umol; enzyme protein (breast cancer cell soluble supernatants) up to 50 Ug. After incubation for 2 to 4 hours at 37°C, 50 Ul each 2-propanol and hexane was added to the reaction mixture, vortexed, and spun at 5000 rpm for 5 min. The upper layer was then quantitatively spotted on SG81 paper, and descending chromatography was performed using chloroform : methanol (9:1) solvent. The cleaved 3Hceramide moved one inch behind the solvent front which was quantitatively determined using toluene scintillation-counting technique.

454

et al.

Fig. 1. Effect of cations on cancer cell CGase activities. Effect of various cations on both Colo-205 and IMR-32 CGase activities are shown in this figure. Various cations as well as EDTA were added to the incubation mixtures in the concentration of 0.5 mM. The cleaved [3H] ceramide have been estimated by chromatographic method as described previously after 2 hour incubation. For Colo-205 and IMR-32 CGase proteins the 100% activity represents 18.5 nmol/mg prot/hr and 15 nmol/mg prot/hr, respectively.

Table 2. Requirement of CGase Activities in Human Colon Carcinoma and Neuroblastoma Cells CGase activities Conditions Complete - Enzyme –TDC – TDC + Sod.Cholate – TDC + Taurocholate + MnCl2* + HgCl2* + EDTA*

IMR-32 Colo-205 (nmol/mg/hr)

9.3 1.9 2.8 2.6 3.9 7.0 1.33 8.74

10.2

2.7 3.4 5.3 5.0

9.3 1.6 7.0

*(0.5mM). The incubation conditions remained the same as described for Table 1 except that the enzyme preparation of Colo-205 and IMR-32 soluble supernatants were used. The cleaved ceramide were quantitated as above.

Mammalian Ceramide Glycanase

455

the activity from Colo-205. The CGase activities from both these cell lines remained unaffected by EDTA. (Table 2, Fig. 1). (ii) Substrate Specificity. Differentially labeled substrates 3H-DH-nLc5 and 3Sph- H-Gg4 had been used to study the effect of substrate structure on enzymatic activities. The apparent Km for either substrate seemed comparable when Colo-205 supernatant was used as enzyme source although the Vmax were different. (Fig. 2a and 2b). Similar results were also observed with IMR-32 CGase activity (data not shown). Other glycolipids with differential labeling are being prepared in the laboratory and will be tested. Unlabeled neutral and acidic glycolipid of lacto-, globo-, and ganglio-series glycolipids were tested as substrates using TLC assay as described in the method section. The TLC results indicated that both Colo-205 and IMR-32 CGase proteins preferred globo-series glycolipids although both lacto- and ganglioseries neutral GSLs are being hydrolyzed well in both cases. 3

Inhibitors of Carcinoma CGases

(i) Inhibition by Substrate Analogues. Previous results from our laboratory indicated that the mammalian ceramide glycanases were hydrophobic in nature, and were purified using hydrophobic columns [17–20]. Other CGase proteins from lower organisms were also found to be hydrophobic in nature [14]. The ceramide glycanase cleaves the glycosidic bond between the ceramide moiety and the oligosaccharide chain of a glycosphingolipid where some hydrophobic interactions may be involved. The glucosyltransferase (GlcT-1) catalyzes the formation of the same glycosidic linkage between the ceramide and a monosaccharide resulting in the synthesis of glucosylceramide, GlcCer from ceramide, the very first step in the biosynthesis of the majority of GSLs [27–29]. It is well known that PDMP (l-phenyl-2-decanoylamino-3-morpholino-l-propanol, HC1) is a potent inhibitor for GlcT-1 in low concentration [30]. We previously reported a similar kind of inhibition of mammalian as well as mollusc CGase activities by PPMP (l-phenyl-2-hexadecanoylamino-3-morpholino-l-propanol, HC1), a higher analog of PDMP [17–20]. Similarly the CGase activities from both Colo-205 and IMR-32 were inhibited by L-PPMP and PDMP as seen in Table 3. The concentration of these analogues needed for inhibition of CGase activities were much higher than that needed for the inhibition of GlcT-1, probably due to the more rigid structural feature of a GSL compared to that of free ceramide. Recent report from our laboratory indicated that the inhibition by PPMP was of competitive nature when purified clam CGase was used [15]. However, in the case of mammalian CGases this inhibition seemed to be of a mixed nature, and a detailed study is in progress. The other substrate analogues like ceramide (non-hydroxy fatty acid containing as well as natural), sphingosine, and sphingomyelin all appeared to inhibit the CGase activities substantially (Table 3). Detailed studies with rat CGase showed that alkyl amines containing C2–C18 chain lengths inhibited the reaction [17–19]. Similarly inhibition of CGase activities in both rat and rabbit tissues with a higher chain length of synthetic sphingosine (above C12) had been reported also [20]. This indicates that the interaction of the alkyl side chains of the ceramide is probably involved in the

456

Basu et al.

Fig. 2. Effect of substrate concentrations on Colo-205 CGase activity. The substrate concentration curves with two differentially labeled substrates are shown with Colo-205 CGase activity. The conditions remained the same as before except the concentrations of the radiolabeled substrates were varied. Panels a and b represent the data for [DH3 H]nLcOse5Cer and [3-Sph-3H]GgOse4Cer, respectively.

Mammalian Ceramide Glycanase

457

Table 3. Effect of Substrate Analogues on Carcinoma Cell CGase Activities CGase activity remaining Addition (1.0 mM) None L-PPMP D-PDMP Ceramide (hFA/nhFA) D-Sphingosine Sphingomyelin

Colo-205 (%)

IMR-32

100 30 27

100 10 30

20 5 17

36 26 12

The incubation conditions remained the same as in Table 2 with the exception of Colo-205 soluble supernatant was used as the only enzyme source. The substrate analogues, as indicated in the table were added in the concentration between 0.5 to 1.0 mM during incubation. The cleaved ceramide was estimated as described in the Table 1 legend. The results are expressed as percent. The 100% values are 20 nmol/mg prot/hr for Colo-205 and 16.5 nmol/mg prot/hr for IMR-32, respectively and the percent calculation as expressed in the table are based on these values.

CGase catalyzed reaction. Recently, a model has been proposed for the mode of action of ceramide in signaling the process where the alkyl chain of ceramide interacts with the hydrophobic cavity of a signaling protein. This implies that through protrusion of the alkyl chain, ceramide provides a lipid anchor to recruit protein to the membrane [31]. N'N'-dimethylsphingosine (DMS) is a potent inhibitor of sphingosine kinase, the enzyme responsible for formation of sphingosine-1-phosphate (SPP), in certain cell types [32]. DMS is also known to increase the ceramide level in a variety of cell lines thus affecting the ceramide/sphingosine rheostat which might account for the pro-apoptotic effect of DMS. SPP, on the other hand, counteracts ceramidemediated stress-activated protein kinase [34]. Perhaps the balance between intracellular levels of ceramide and SPP and their regulatory effects on different family members of mitogen-activated protein kinase (MAPK) determines the fate of the cell [32–34]. The role of DMS and of SPP on CGase activity has not been tested yet. (ii) Inhibition by Other Chemical Agents. One of the major reasons of unsuccessful chemotherapy has been correlated to the multiple drug resistance (MDR) of the cancer cells. Presence of the elevated levels of glucosylceramide (GlcCer), the major intermediate for almost all the glycosphingolipid during biosynthesis, in varieties of MDR cancer cells and a low level of GlcCer in drug-sensitive cancer cells implements glucosylceramide as a tumor marker in the MDR cells that could be used for early detection [35, 36]. It was shown recently that overexpression of GlcT1 (glucosylceramide synthetase) in wild type MCF-7 breast cancer cells induced

458

Basu et al.

resistance to adriamycin and ceramide in the transfected MCF-7 breast cancer cells [37]. Tamoxifen, a synthetic anti-estrogen, among other chemicals had been successfully used in MDR breast cancer cells to lower the GlcCer levels [38]. Tamoxifen inhibited the formation of lactosylceramide and gangliosides in human melanoma cells [39]. From these and other related studies it is apparent that GSLs may be playing a significant role in the tumorigenic process, and a current hypothesis suggests the use of inhibitors to slow GSL synthesis which might make the cancer cells non-metastatic [40]. However the role of ceramide in cancer progression or remission is yet to be determined. We have tested the effect of tamoxifen, 4-OH tamoxifen, and cis-platin, a platinum conjugate well-known chemotherapeutic agent used for almost all cancer chemotherapy, on cancer cell CGase activities. Previously, it was reported from our laboratory that mammalian CGase activities from rat and rabbit tissues were inhibited by tamoxifen but not by cis-platin [18]. As seen in Table 4, both tamoxifen and 4-OH tamoxifen inhibited the CGase activity while cis-platin did not show much inhibition. A previous report from our laboratory indicated that cis-platin inhibits DNA replication by inhibiting DNA polymerase-a [41]. The inhibition of this DNA replication enzyme is accompanied by the release of zinc from the DNA-binding zinc-finger domain of the polymerase-a [42]. This has been confirmed by complete relaxation matrix analysis of transverse NOE data obtained from the Pt-bound nonapeptide (ERFKCPCT) NMR studies [43]. However, the pathway of apoptosis from the inhibition of DNA biosynthesis or CLIP-174 (another zinc-protein) is not known yet [44]. The balance between ceramide and its metabolites as well as certain glycosphingolipid levels play a crucial role in cellular proliferation, differentiation, and apoptosis [31–34]. Certain glycosphingolipids were also identified recently as markers for MDR cancer cells which can be reduced by inhibiting their synthesis [35–38]. PDMP, the well known inhibitor for GSL biosynthesis, has been used widely to inhibit GSL biosynthesis at various steps [30, 45, 46]. PPMP, a higher analog of PDMP, had been shown to inhibit mammalian as well as mollusc CGase activities Table 4. Effect of Chemotherapeutic Drugs on CGase Activity CGas activity remaining Addition (1 mM) None Tamoxifen 4-Hydroxy Tamoxifen cis-Platin

IMR-32 Colo-205 (nmol/mg pr/2 hr)

22.8

18.2

3.8 6.3

3.0 5.0

12.4

21.0

Colo-205 soluble supernatant was used as enzyme source. The incubation conditions remained the same except that the chemotherapeutic drugs were added in the incubation mixture in the concentrations as indicated. The cleaved ceramide was estimated after incubation as described before.

Mammalian Ceramide Glycanase

459

at a concentration higher than that needed for inhibition of synthetic enzymes [15– 20]. Recently, MAPP (myristoly-amino-phenyl-propanol) inhibition of ceramidase has been reported also [47]. The functional relationship during germinal center Bcell apoptosis between globotriaosylceramide, the germinal center (GC) B-cell differentiation antigen CD77, and B-cell restricted cell differentiation antigen CD19, was demonstrated recently in Daudi cells by treating the cells with PPMP [48]. Glycosphingolipids which comprise an integral part of all eukaryotic cells contribute to various cellular functions [49, 50]. Functional aspect of GSLs have been of great interest for the past few years [51, 52]. In addition, ceramide, an integral structural part of all GSLs, and its metabolites have emerged as one of the regulatory components in the signal transduction and apoptotic pathways [1-10, 31-34, 51, 52]. Previously, a correlation of mammalian ceramide glycanase has been reported with development and with hormonal balance [17, 20]. Functional significance of the ceramide glycanase had been suggested recently in regulation of homeostasis of the cell surface GSL contents of B16 melanoma cells [53]. Bacterial endoglyco-ceramidase (EGCase) had been used for the cleavage of B16 melanoma cell surface GSLs [11, 13] and the recycling of the ceramide was monitored. A novel form of homeostasis in the mammalian cells has been suggested which is coupled to the GSLsynthesizing system and can provide a defense mechanism against microbial EGCase [53]. The breakdown of sphingomyelin has been considered until now the most probable pathway for the generation of ceramide under physiological conditions [17–20, 54–56]. The discovery of the mammalian ceramide glycanase however suggests another interesting functional possibility for glycosphingolipids by producing ceramide via a one-step breakdown of GSL which may in the future become an important tool for studying the mechanism of action of ceramide in cellular processes. ACKNOWLEDGEMENTS Part of this work was supported by NIH Grants NS-18005 (Jacob Javitz Award) and CA-14764 to S.B. and a grant-in-aid from Bayer Corporation Elkhart, IN to M.B. Our special thanks to Dr. Jin-ichi Inokuchi for gift samples of HPLC purified L-PPMP and -PDMP and to Mrs. Dorisanne Nielsen for her help in the preparation of the final draft of the manuscript. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.

Hofmann, K., and Dixit, V. M. (1998) Trends Biochem. Sci. 10:374–377. Perry, D. K., and Hannun, Y. (1998) Biochim. Biophys. Acta 1436:233–243. Haimovitz-Friedman, A., Kolesnick, R. N., and Fuks, Z. (1997) Br. Med. Bull. 53:539–553. Veldman, R. J., Klappe, K,, Hoekstra, D., and Kok, J. W. (1998) Biochem. J. 331:563–569. Kolesnick, R. N., and Kronke, M. (1998) Ann. Rev. Physiol. 60:343–377. Hannun, Y. A. (1994) J. Biol. Chem. 269:3125–3128. Liu, B., Obeid, L. M., and Hannun, Y. (1997) Semin. Cell Dev. Biol. 8:311–322. Hannun, Y. (1996) Science 274:1855–1859. Futerman, A. H. (1998) Biochemistry (Mosc.) 63:74–83. Geilen, C. G., Wieder, T., and Orfanos, C. E. (1997) Arch. Dermatol. Res. 289:559–566. Ito, M., and Yamagata, T. (1986) J. Biol. Chem. 261:14278–14282. Li, Y. T., and Li, S. C. (1989) Methods Enzymol. 179:479–487.

460

Basu et al.

13. Ito, M., and Yamagata, T. (1989) Methods Enzymol. 179:488–496. 14. Basu, S. S., Dastghieb-Hosseini, S., Hoover, G., Li, Z., and Basu, S. (1994) Anal. Biochem. 222:270– 274. 15. Dastgheib, S., Basu, S., and Basu, M. Methods Enzymol. (in press). 16. Westervelt, C. W. et al. (1989) Glycoconjugate J. 6:405. 17. Basu, M. et al. (1997) Ind. J. Biochem. Biophys. 34:142–149. 18. Basu, M. et al. (1998) Acta Pol. Biochim. 42:327–335. 19. Basu, M., Kelly, P., Girzadas, M. A., Li, Z., and Basu, S. Properties of animal ceramide glycanases. Methods Enzymol. (in press). 20. Basu, M., Kelly, P., O’Donnell, P., Miguel, M., and Basu, S. (1998) Glycobiology 8:Abstract 94. 21. Das, K. K., Basu, S., and Basu, M. (1984) Anal. Biochem. 143:125–134. 22. Chein, J. L., Li, S. C., Laine, R. A., and Li, Y. T. (1978) J. Biol. Chem. 253:4031–4035. 23. Li, S. C., and Li, Y. T. (1972) Methods Enzymol. 28:714–720. 24. Schwarzman, G., and Sandhoff, K. (1987) Methods Enzymol. 138:138–160. 25. Chigorno. V., Valesecchi, M., Nicolini, M., and Sonnino, S. (1997) Ind. J. Biochem. Biophys. 34: 150–156. 26. Nakamura, K., and Handa, S. (1984) Anal. Biochem. 142:406–410. 27. Basu, S., Kaufman, B., and Roseman, S. (1973) J. Biol. Chem. 248:1385–1394. 28. Basu, S. (1991) Glycobiology 1:469–475. 29. Basu, S., Basu, M., Dastgheib, S., and Hawes, J. W. (1999) in: Comprehensive Natural Products Chemistry, D. Barton and K. Nakanishi (eds.), Vol 3, B. M. Pinto (ed.), Elsevier, New York, pp. 107–128. 30. Radin, N., and Inokuchi, J. I. (1991) Trends Glycosci. GlycoTech. 3:200–213. 31. Kronke, M. (1997) Cytokine Growth Factor Rev. 8:103–107. 32. Edsall, L. C., Van Brocklyn, J. R., Culvillier, O., Kleuser, B., and Spiegel, S. (1998) Biochemistry 37:12892–12898. 33. Cuvillier, O. et al. (1996) Nature 381:800–803. 34. Spiegel, S. et al. (1998) Biochemistry (Mosc.) 63:69–73. 35. Lavie, Y., Cao, H., Bursten, S. L., Giuliano, A. E., and Cabot, M. C. (1996) J. Biol. Chem. 271:19530–19536. 36. Lucci, A., Cho, W. I., Han, Y. T., Giuliano, A. E., Morton, D. L., and Cabot, M. C. (1998) Anticancer Res. 18:475–480. 37. Liu, Y. Y., Han, T. Y., Giuliano, A. E., and Cabot, M. C. (1999) J. Biol. Chem. 274:1140–1146. 38. Lavie, Y. et al. (1997) J. Biol. Chem. 272:1682–1687. 39. Cabot, M. C., Giuliano, A. E., Volner, A., and Han, T. Y. (1996) FEBS Lett. 394:129–131. 40. Radin, N. S. (1999) Biochem. Pharmacol. 57:589–595. 41. Kelley, T. J., Moghaddas, S., Bose, R., and Basu, S. (1993) Cancer Biochem. Biophys. 13:135–146. 42. Bose, R. N., Li, D., Kennedy, M., and Basu, S. (1995) J. Chem. Soc. Commun. 1731–1732. 43. Bose, R. N., Li, D., Yang, W.-W., and Basu, S. (1999) J. Biomol. Struc. Dynamics 16:1075–1085. 44. Pierre, P., Pepperkok, R., and Kries, T. E. (1994) J. Cell. Sci. 107:1909–1920. 45. Radin, N. S., Shyaman, J. A., and Inokuchi, J. I. (1993) Adv. Lipid Res. 26:183–213. 46. Chatterjee, S., Cleveland, T., Shi, W.-Y., Inokuchi, J., and Radin, N. S. (1996) Glycoconjugate J. 13:481–486. 47. Bielawska, A. et al. (1996) J. Biol. Chem. 271:12646–12645. 48. Khine, A. A., Firtel, M., and Lingwood, C. A. (1998) J. Cell Physiol. 176:281–291. 49. Hakomori, S., Handa, K., Iwabuchi, K., Yamamura, S., and Prinetti, A. (1998) Glycobiology 8-xi-xix. 50. Hakomori, S. (1989) Adv. Cancer Res. 52:257–331. 51. Hakomori, S. (1990) J. Biol. Chem. 265:18713–18716. 52. Hannun, Y. A., and Bell, R. M. (1989) Science 243:500–507. 53. Ito, M., and Komori, H. (1996) J. Biol. Chem. 271:12665–12660. 54. Chatterjee, S. (1993) Adv. Lipid Res. 26:27–48. 55. Hannun, Y. A., and Obeid, L. M. (1994) Trends Biochem. Sci. 20:73–77. 56. Obeid, L. M., Linardic, C. M., Karolak, L. A., and Hannun, Y. A. (1993) Science 259:1769–1771.