Sphingosine 1-phosphate generated in the endoplasmic reticulum membrane activates release of stored calcium

Sphingosine and sphingosine derivatives induce Ca2+ release from inositol 1,4,5-trisphosphate (InsP3)-sensitive Ca2+ pools in permeabilized cells (Ghosh, T. K., Bian, J., and Gill, D. L. (1990) Science 248, 1653-1656). To further assess the mechanism of sphingoid base-mediated Ca2+ release, the effects of sphingosine and sphingosine derivatives on Ca2+ fluxes were characterized using a microsomal membrane vesicle fraction (B3) enriched in rough endoplasmic reticulum (ER) prepared from cells of the DDT1MF-2 cell smooth muscle line (Ghosh, T. K., Mullaney, J. M., Tarazi, F. I., and Gill, D. L. (1989) Nature 340, 236-239). Addition of 15 microM sphingosine to Ca2+ pump-loaded B3 vesicles induced a delayed but thereafter rapid Ca2+ release from vesicles which was dependent on the presence of ATP and was blocked by ADP. Sphingosylphosphorylcholine (SPC) released Ca2+ to the same extent (more than 80% of pumped Ca2+), but in contrast to sphingosine, there was no lag and the effect was independent of ATP or ADP. The EC50 for sphingosine and SPC in activating Ca2+ release was 1 and 3 microM, respectively. Such observations are consistent with the view that sphingosine, unlike SPC, must be modified by an ATP-requiring kinase activity located within the ER membrane. Sphingoid bases do not appear to release Ca2+ through InsP3 receptors since heparin had no effect on sphingoid base-mediated Ca2+ release. Sphingosine 1-phosphate (sph-1-P), the likely active Ca(2+)-releasing derivative of sphingosine, was synthesized by phospholipase D-catalyzed cleavage of SPC, purified, and tested for Ca(2+)-releasing activity. sph-1-P at 10 microM induced Ca2+ release from both B3 vesicles and permeabilized DDT1MF-2 cells to exactly the same extent as sphingosine. Unlike sphingosine, the effect of sph-1-P was immediate and not blocked by ADP. Using B3 membrane vesicles incubated with [gamma-32P]ATP and sphingosine under the same conditions as Ca2+ flux studies, a labeled band was detected on TLC which ran identically with authentic sph-1-P. Formation of this labeled product was prevented by removal of exogenous sphingosine and blocked by ADP. Sphingosine- but not SPC-mediated Ca2+ release was blocked by 10 mM oxalate. 10 mM oxalate also blocked the formation of 32P-labeled sph-1-P indicating that it is an inhibitor of sph-1-P formation. The studies establish that the ER membrane contains the necessary kinase to convert sphingosine to sph-1-P which functions as a powerful mediator of Ca2+ release through a non-InsP3 receptor-mediated mechanism in the same ER membrane, perhaps reflecting a novel Ca2+ signaling pathway.