The enigma of sphingosine 1-phosphate synthesis: a novel role for endothelial sphingosine kinases.

Sphingosine 1-phosphate (S1P) is well known as a key sphingolipid messenger molecule in the cardiovascular system, yet many fundamental features of its synthesis, transport, and action remain enigmatic. Indeed, the very name “sphingolipid” was coined to reflect the mysterious characteristics of this class of lipids. It is now well established (through the work of Timothy Hla and of many others subsequently) that S1P binds to and activates a family of G protein–coupled S1P receptors located in vascular endothelial cells, cardiac myocytes, blood platelets, and vascular smooth muscle cells (among other cell types) and elicits a broad range of physiological responses (reviewed elsewhere1). In vascular endothelial cells, S1P elicits such diverse responses as cell survival, proliferation, angiogenesis, cell migration, permeability, and endothelial NO synthase activation. The intracellular signaling pathways stimulated by S1P have been extensively characterized, yet the cellular origins of S1P and the pathways for its transport and action in cardiovascular tissues remain incompletely understood. The concentration of S1P in normal human plasma is within the range of several hundred nanomolar, and this amphipathic lipid is highly protein bound, mostly to HDL and albumin.2 Importantly, S1P receptors bind their ligand with an affinity in the nanomolar range; therefore, many important questions remain. What controls S1P binding to plasma proteins? Where does all this plasma S1P come from? Important clues to the latter question are provided in a recent report from the laboratory of Timothy Hla, published in this issue of Circulation Research3 and reporting the novel and important discovery: the vascular endothelium is an important source of plasma S1P through the actions of sphingosine kinases (SphKs) in endothelial cells. For many years, conventional wisdom held that blood platelets represent the principal source for S1P in the plasma. Elegant studies by Igarashi4 (no relation to J.I.) showed that blood platelets contain abundant quantities of sphingosine kinase, the key enzyme involved in synthesizing S1P from sphingosine, and many laboratories have shown that blood platelets can secrete S1P following platelet activation by thrombin and other ligands. Moreover, platelets are characterized by having a low activity of S1P lyase, a key enzyme involved in S1P degradation. Thus, blood platelets contain abundant S1P, and can secrete the lipid in response to agonist stimulation. So convincing were these observations that S1P was dubbed by many investigators, including ourselves, as a “platelet-derived lipid mediator.” There is no doubt that platelets do contain and secrete S1P, but are they the sole, or even the principal, source of S1P in plasma? Recent research suggests that the answer is no. Indeed, a recent article5 reported that erythrocytes are a major source of plasma S1P. Venkataraman et al3 used multiple complementary experimental approaches to test their hypothesis that endothelial cells are a key source of plasma S1P. They first explored the half-life of plasma S1P using a new H-labeled synthetic reagent C17-S1P. This lipid differs slightly from endogenous S1P (C18-S1P), and these 2 lipid species are readily distinguishable in high-performance liquid chromatographic analysis, permitting the authors to determine the abundance of the exogenous H-labeled C17-S1P following injection of the compound into the mouse tail vein. The authors estimated the t1/2 of plasma C17-S1P to be on the order of 15 minutes, suggesting the presence of highly active S1P-producing and -degrading pathways within the body. The authors next analyzed plasma S1P levels in mice that had targeted deletions of the critical S1P synthetic enzyme sphingosine kinase, of which there are 2 mammalian isoforms (SphK1 and SphK2). To test the contribution of hematopoietic cells to plasma S1P, the authors performed bone marrow transplant experiments in which marrow from wild-type mice was transfused into genetargeted SphK1/SphK2-deficient mice (these mice were engineered to have a homozygous deletion of SphK1 but were heterozygous for SphK2 deletion because complete deletion of all 4 SphK alleles is embryonic lethal6). The authors subjected either wild-type or SphK-deficient mice to whole body irradiation to eliminate hematopoietic cells, then transfused bone marrow derived from mice having the other genotype into the irradiated animals. At baseline, the mice that were genetically deficient in SphK had lower plasma S1P levels, and transfusion of wild-type cells to these mice was able to rescue plasma S1P levels, as expected. However, to the surprise of the authors, the plasma S1P concentration of wild-type mice did not decrease after total body irradiation and remained at normal levels following transfusion of bone marrow from the SphK-deficient animals. The authors also found that pharmacological treatments that cause anemia or thrombocytopenia similarly failed to attenuate plasma S1P levels. So, where is the S1P coming from? These observations raised the possibility that some other nonhematopoietic cell type must produce sufficient S1P such The opinions expressed in this editorial are not necessarily those of the editors or of the American Heart Association. From the Department of Cardiovascular Physiology (J.I.), Kagawa University, Kagawa, Japan; and Cardiovascular Division (T.M.), Department of Medicine, Brigham and Women’s Hospital and Harvard Medical School, Boston, Mass. Correspondence to Thomas Michel, MD, PhD, Cardiovascular Division, Brigham and Women’s Hospital, Thorn Building, Room 1210A, 75 Francis St, Boston, MA 02115; E-mail [email protected] (Circ Res. 2008;102:630-632.) © 2008 American Heart Association, Inc.