ALTERED SPHINGOLIPID METABOLISM IN N -(4-HYDROXYPHENYL)RETINAMIDE-RESISTANT A2780 HUMAN OVARIAN CARCINOMA CELLS

In the present work, we studied the effects of fenretinide (HPR), a hydroxyphenyl derivative of all- trans -retinoic acid, on sphingolipid metabolism and expression in human ovarian carcinoma A2780 cells. A2780 cells, which are sensitive to a pharmacologically achievable HPR concentration, become 10-fold more resistant after exposure to increasing HPR concentrations. Our results showed that HPR was able to induce a dose- and time dependent increase in cellular ceramide levels in sensitive, but not in resistant cells. This form of resistance in A2780 cells was not accompanied by the overexpression of multidrug resistance specific proteins MDR1 and MRP, whose mRNA levels did not differ in sensitive and resistant A2780 cells. HPR-resistant cells were characterized by an overall altered sphingolipid metabolism. The overall content in glycosphingolipids was similar in both cell types, but the expression of specific glycosphingolipids was different. More in detail, our findings indicated that glucosylceramide levels were similar in sensitive and resistant cells, but resistant cells were characterized by a 6-fold lower expression of lactosylceramide levels and by a 6-fold higher expression of ganglioside levels than sensitive cells. The main gangliosides from resistant A2780 cells were identified as GM3 and GM2. The possible metabolic mechanisms leading to this difference were investigated. Interestingly, the mRNA levels of glucosylceramide and lactosylceramide synthases were similar in sensitive and resistant cells, while GM3 synthase mRNA level and GM3 synthase activity were remarkably higher in resistant cells.


INTRODUCTION
Sphingolipid metabolism plays a pivotal role in the mechanism of apoptosis induced in tumor cells. Ceramide, produced under physiological (tumor necrosis factor α, γinterferon, interleukines) and pharmacological (anticancer drugs, including daunorubicin, vincristine, 1-α-D-arabinofuranosylcytosine, retinoids) stimuli by sphingomyelin hydrolysis or by de novo biosynthesis, is a mediator of apoptosis and an inhibitor of cell proliferation in a variety of tumor cell lines (reviewed in 1-2).
Cell cultures. An HPR-resistant cell line, A2780/HPR, was developed from parental A2780 cells by in vitro incubation of A2780 cells with increasing concentrations of HPR as previously described (34). Briefly, cells surviving 60 rounds of selections in HPR containing medium (3 rounds at 1 µM, 11 rounds at 2 µM, 8 rounds at 3 µM and 38 rounds at 5 µM) were cloned by limiting dilution. One clone (A2780/HPR) was expanded and when tested for HPR sensitivity demonstrated a 10 fold resistance to HPR, which was slightly reversible upon drug removal for 5 rounds (34). Therefore A2780/HPR cells were continuously maintained in 5 µM HPR and seeded without the drug in all the experiments. A2780 and A2780/HPR cells were grown in monolayer in RPMI 1640 medium (Biowhittaker, Belgium) containing 10% fetal bovine serum (Gibco, Gaithersburg, MD, USA) in 5% CO 2 at 37 o C.
Effects of C2Cer on A2780 and A2780/HPR cells. Cells were seeded at density of 10,000 cells per well in 96-well tissue culture plates, treated on the next day with different concentrations of C2Cer in absence of FBS and incubated for one additional day. Control cultures received the same amount of ethanol as the treated cultures (0.1%). Cell number was estimated by using the sulforhodamine B (SRB) assay (49). Two analyses were performed and four replicate wells were used for each analysis. Apoptotic cells were identified by the Terminal dUTP nick-end labeling (TUNEL) method using an in situ cell death detection kit (Roche Diagnostics, Indianapolis, IN) according to the manufacturer's instructions. Briefly, A2780 and A2780/HPR cells were treated 24 hr after seeding with 2.5 µM C2Cer in absence of FBS, incubated for one additional day, harvested and washed twice in PBS. After centrifugation, cells were fixed in 4% paraformaldehyde for 1 hour at 26°C, and permeabilized with 0.1% Triton X-100 and 0.1% sodium citrate in PBS. After washing, the cells were resuspended in TUNEL reaction mixture containing FITC-dUTP and terminal-deoxy-nucleotidyl-transferase (TdT). Control cells were suspended in the TUNEL reaction mixture containing FITC-dUTP without TdT and incubations were performed for 1 hour at 37°C before washing the cells twice. The number of TUNELpositive cells, as detected by fluorescent microscopy, was assessed on at least 100 cells in two different smears and referred to the whole cell population.
Northern analysis. Total cellular RNA was isolated using the TriReagent method (Molecular Research Center, Cincinnati). Twenty µg of total RNA were fractionated in formaldehyde/Mops agarose gel and blotted on Hybond N+ nylon membranes (Amersham Corp., Arlington Heights, IL). Filters were hybridized with cDNA probes for multidrug resistance (MDR1), multidrug resistance-associated protein (MRP), GlcCer synthase, LacCer synthase, and GM3 synthase, obtained by RT-PCR using A2780/HPR cDNA as template. Briefly, single-stranded cDNA was synthesized from 2 µg RNA by reverse transcription, then amplified by PCR. The primer pairs used for PCR amplification are the following: MDR1: sense, 5'-AAAAAGATCAACTCGTAGGAGTG-3', and antisense, 5'-GCACAAAATACACCAACAA-3' (161 pb amplified product). MRP: sense, 5'-CCACCTCCTCATTCGCATCCACCTTG-3', and antisense, 5'-GGAAACCATCCACGACCCTAATCCCT-3' (296 pb amplified product). Lipid extraction and determination. In the experiments described above, at the end of the treatment periods, cells adherent to the dishes were harvested in ice-cold water (2 mL) by scraping with a rubber policeman. Cells floating in the culture medium were collected by centrifugation. Adherent and floating cells were analyzed to determine the content of radioactivity associated with lipids. Samples were lyophilized and lipids were extracted twice with chloroform/methanol 2:1 by vol (1 st extraction 1.5 ml, 2 nd extraction 0.25 ml) (51). The total lipid extracts were subjected to a two-phase partitioning as previously described (52), resulting in the separation of an aqueous phase containing gangliosides and in an organic phase containing all other lipids. Aliquots of total lipid extracts, aqueous and organic phases were analyzed by HPTLC as described below, followed by radioactivity imaging for quantification of radioactivity.
Identity of radioactive lipids separated by HPTLC (using HPTLC Silica Gel 60 plates from Merck, Germany) was assessed by comigration with standard lipids and confirmed by susceptibility of compounds to the following enzymatic and chemical treatments (51). The total lipid extract from about 50 mg of cell proteins was analysed for mass lipid content. The phospholipid content was determined in the organic phase as phosphate after perchloric acid digestion by the method of Bartlett (53). Gangliosides from the aqueous phase and phospholipids, and cholesterol from the organic phase were separated by HPTLC as described below. Identification of lipids after separation was assessed by comigration with standard lipids and confirmed by their susceptibility to enzymatic and chemical treatments as previously described (51). After chromatographic separation, compounds were chemically detected and their amount was determined by densitometry as described below. Gangliosides from the aqueous phase were characterized by HPLC-ESI-MS (46) (see below).
GM3 synthase assay. Cells cultured in 100-mm dishes as described above were harvested using a plastic scraper and washed two times with phosphate-buffered saline.
GM3 synthase activity was assayed as previously described (54-55) with some modifications. Cells were resuspended in 150 mM sodium cacodylate-HCl buffer, pH 6.6  Ganglioside, neutral GSL and phospholipid species were quantified after separation on HPTLC followed by specific detection with a p-dimethylaminobenzaldehyde reagent   Figure 4 shows the patterns of radioactive lipids extracted from A2780 and A2780/HPR cells after HPTLC separation. In both cell lines, radioactive bands co-migrating with standard ceramide, glucosylceramide, PE, lactosylceramide and SM were detectable. In the case of A2780/HPR cells, at least two intense radioactive bands migrating below SM were also present. To allow a better resolution of these lipid profiles, the total lipid extracts were subjected to a two-phase partitioning, resulting in the separation of an aqueous phase containing gangliosides and an organic phase containing all less polar lipids (including radioactive ceramide, PE, neutral glycosphingolipids and SM), that were further analyzed.
As reported in Table 1, most radioactive lipids were associated with the organic phase in both A2780 and A2780/HPR cells. Remarkably, independently from the chase time, the radioactivity associated with the aqueous phase in the case of resistant A2780/HPR cells was 3 times higher than in sensitive A2780 cells. The quantitative data, summarized in Table 2

Expression levels of GlcCer synthase, LacCer synthase and GM3 synthase in A2780
and A2780/HPR cells. To investigate potential mechanisms underlying the observed differences in sphingolipid levels in A2780 and A2780/HPR cells, the mRNA expression of the enzymes GlcCer synthase, LacCer synthase, and GM3 synthase was evaluated in parental and resistant cells. A2780/HPR cells showed GlcCer synthase and LacCer synthase mRNA levels similar to those of A2780 cells, whereas GM3 synthase mRNA was markedly increased (Figure 8). In A2780 cells, the mRNA levels of GM3 synthase were below the detection limit of our Northern blot analysis. Using RT-PCR, we were able to detect in A2780 cells a faint band relative to GM3 synthase.
To asses if observed changes in the mRNA levels of GM3 synthase were paralleled by differences in the enzyme levels, we measured the activity of GM3 synthase in A2780 and A2780/HPR cells by the mean of a cell-free assay using radioactive LacCer as substrate. As indicated in Figure 9, the in vitro activity of GM3 synthase was 20-fold higher in resistant A2780/HPR cells than in sensitive A2780 cells. Interestingly, a parallel between ceramide generation in response to HPR treatment and sensitivity to HPR was observed in cells of lymphoid origin (64). Different acute lymphoblastic leukemia cell lines respond to cytotoxic HPR treatment with an increase in ceramide. On the other hand, HPR is not cytotoxic and has no effect on ceramide levels in non-malignant lymphoid cell types (64).
In the well studied case of multidrug resistance  (12). In these cells, a reduced synthesis of LacCer was observed, that resulted in lower levels of LacCer, GM3 and GM2 and higher levels of galactosylceramide and SM.
According to this knowledge, the resistance to a retinoid in A2780 cells herein reported, does not seems to be related to MDR. In fact, our data indicate that levels of MDR1 mRNA and MRP mRNA, as well as those of GlcCer and GlcCer synthase mRNA were substantially unchanged in A2780/HPR cells respect to parental A2780 cells.
Instead striking differences were observed between A2780 and A2780/HPR cells in the metabolism of more complex glycosphingolipids. In A2780/HPR cells, the sphingolipid metabolism is markedly oriented toward the synthesis of gangliosides. We characterized the main gangliosides of A2780/HPR cells as GM3 and GM2, and their levels were 6-fold higher than in parental cells, while LacCer, the direct neutral glycolipid precursor of gangliosides, was proportionally lower. Finally, differences were observed in the mRNA levels of the enzyme GM3 synthase, that is more expressed in A2780/HPR than in sensitive A2780 cells. This difference was reflected by the in vitro activity of GM3 synthase, that was 20-fold higher in resistant A2780/HPR cells than in sensitive A2780 cells.
Thus, the data presented in this paper indicate that all the biosynthetic pathway downstream of ceramide is more active in A2780/HPR cells than in the parental sensitive cell line, leading to the expression of higher levels of GM3 and GM2 gangliosides. This particular alteration in sphingolipid metabolism associated with the development of resistance to an apoptosis inducing drug is reported here for the first time.
It could be argued that an overall increased flow through the ganglioside biosynthetic pathway is also a mechanism, even if quite elaborate, to scavenge ceramide thus           Table 3. Activity of acid, alkaline and neutral ceramidase in A2780 and A2780/HPR cells.
Ceramidase activity was measured by the mean of a cell-free assay at pH 4. 5, 9.