We reported that safingol (L-threo-dihydrosphingosine) synergized the cytotoxicity of the ceramide-generating retinoid, N-(4-hydroxyphenyl)retinamide (fenretinide, 4-HPR) in human cancer cell lines in vitro. The mechanism by which safingol synergized 4-HPR cytotoxicity is not known. Safingol inhibited protein kinase C activity, and D,L-threo-dihydrosphingosine inhibited sphingosine kinase, in platelets, in cell extracts, or in short-term (minutes) drug exposures. However, the significance of these observations to chronic exposure in whole cells is not clear, and in rodent cells, safingol was metabolized to L-threo-ceramide and L-threo-glucosylceramides. Determining safingol metabolism in human cancer cells is crucial to understanding the mechanism of safingol + 4-HPR cytotoxic synergy, and to optimize clinical strategies for safingol administration. Using 3H-safingol, we determined the uptake and metabolism of safingol in human neuroblastoma cell line, CHLA-90, and ALL leukemia cell line, MOLT-4. Safingol and its metabolites were analyzed by thin-layer chromatography and scintillography, cell viability was measured using DIMSCAN analysis. To partially mimic infusional delivery, cells were incubated with 1 uM and 3 uM safingol for 24 hrs. Lipid analysis showed that safingol was acylated to L-threo-dihydroceramide (L-t-(DH)cer) and L-threo-dihydrosphingomyelin (L-t-(DH)SM). No safingol-derived ceramides or glucosylceramide were detected. At +4 hrs in 1 and 3 uM safingol in CHLA-90 cells, free safingol was 47% and 53%, L-t-(DH)cer was 47% and 44%, and L-t-(DH)SM was 6% and 3%, of remaining label, respectively. At +4 hrs in 1 and 3 uM safingol in MOLT-4 cells, free safingol was 20% and 60%, L-t-(DH)cer was 70% and 30%, and L-t-(DH)SM was 10% and 10%, of remaining label, respectively. To mimic a bolus delivery, cells were incubated with 6 uM safingol for 2 hrs and then medium was changed. In CHLA-90, at +4 and +24 hrs after washout, free safingol was 35% and 10%, L-t-(DH)cer was 50% and 30%, and L-t-(DH)SM was 15% and 60%, of remaining label, respectively. In contrast, in MOLT-4, at +4 and +24 hrs after washout, free safingol was 55% and 15%, L-t-(DH)cer was 42% and 75%, and L-t-(DH)SM was 3% and 10% of remaining label, respectively. Short chain L-t-(DH)ceramides (C2, C6 and C8) did not increase 4-HPR cytotoxicity in CHLA-90 and MOLT-4 cells. We conclude that 1) safingol was not metabolized to L-threo-ceramide or L-threo-glucosylceramide in two human cancer cell lines; 2) safingol metabolism differed somewhat in two human cancer cell lines; 3) L-t-(DH)cer most likely does not synergize 4-HPR cytotoxicity; 4) free safingol likely synergizes 4-HPR cytotoxicity; 5) delivery schedule may affect the kinetics of cellular safingol levels. This study has partially clarified the mechanism of safingol + 4-HPR cytotoxic synergy, and provided crucial information to optimize safingol delivery for in vivo efficacy studies.
[Proc Amer Assoc Cancer Res, Volume 45, 2004]