The sphingomyelin pathway in tumor necrosis factor and interleukin-1 signaling

Abstract

Richard Kolesnick and David W. Golde The Laboratories of Signal Transduction and Molecular and Cellular Hematology Memorial Sloan-Kettering Cancer Center New York, New York 10021 E:vidence suggests that tumor necrosis factor a (TNFa) and interleukin-18 (IL-18) employ the sphingomyelin path- way to effect signal transduction by their receptors. This pathway is initiated by hydrolysis of plasma membrane sphingomyelin to ceramide by the action of a sphingomye- linase. Ceramide serves as asecond messenger, stimulat- ing a serine/threonine ceramide-activated protein kinase to transduce the cytokine signal, in part through mitogen- activated protein (MAP) kinase and transcription factors such as NF-KB (Figure 1). The extent to which this signal- ing system is used in inflammation, immune responses, and apoptosis is not known, but accumulating evidence suggests that it is a commonly employed pathway that could be exploited therapeutically. The Sphingomyelin Metabolic Pathway Sphingomyelin is preferentially concentrated in the outer leaflet of the plasma membrane of most mammalian cells; it is comprised of a long chain sphingoid base backbone (predominantly sphingosine), a fatty acid, and a phospho- choline head group (Figure 2). The fatty acid in amide linkage at the second position of the sphingoid base consti- tutes ceramide. Hydrolysis of the phosphodiester bond by a sphingomyelinase to yield ceramide and phosphocho- line is the only clearly defined mechanism for sphingomye- lin degradation in mammalian cells (Kolesnick, 1991). Sphingomyelin was considered only a structural ele- ment of the plasma membrane. However, 1,2-diacylglycerol (DG), a physiologic activator of protein kinase C, stimu- lated rapid sphingomyelin degradation toceramide in GH3 rat pituitary cells (Kolesnick, 1991). Little of the generated ceramide was deacylated to sphingoid bases, potential inhibitors of protein kinase C (Hannun and Bell, 1989), prompting a search for additional derivatives of ceramide. Several investigations established the existence of a spe- cific metabolic pathway from sphingomyelin to ceramide l-phosphate (see Bajjalieh et al., 1989; Kolesnick, 1991). The sphingomyelin metabolic pathway is similar to the phosphoinositide signal transduction pathway. The cen- tral lipids in these pathways, ceramide and DG, both serve as substrates for the same bacterial DG kinase, implying they possess structural similarity. Their phosphorylated forms, ceramide l-phosphate and phosphatidic acid, were, therefore, also structurally similar. Further, neutral sphingomyelinase, the enzyme that initiates the sphingo- myelin pathway, is a phospholipase C concentrated in the plasma membrane (Kolesnick, 1991) like the enzyme ini- tiating the phosphoinositide pathway. Ceramide also ap- pears to be an ideal candidate second messenger since it readily redistributes across a membrane bilayer (Lipsky and Pagano, 1985). Since DG utilized a specific kinase, protein kinase C, for signaling, it was considered that cera- mide might stimulate a kinase. Ceramide does not activate protein kinase C. identification of Ceramidektivated Protein Kinase and Phosphatase Because natural ceramide contains long and very chain saturated or monounsaturated fatty acids and is poorly soluble in aqueous solutions, cell-permeable ana- logs were synthesized. A ceramide was constructed con- taining octanoic acid, N-octanoyl sphingosine (CS-cer), which is analogous to 1,2dioctanoylglyceroI, the hydro-