Sphingolipids: The Sphinx of Lipids
By Tom Brock, Ph.D.
The lipids are an extraordinarily diverse collection of molecules that serve almost every conceivable function. Some lipids are well-known by name and purpose to most biologists, while others remain as enigmatic as the Great Sphinx, looming before the pyramids of Egypt. Consider the sphingolipids, whose name reveals little about their structure or function. They can be as mysterious to the biologist as the Sphinx was to the first European explorers in Cairo. In fact, the Egyptian sphinx was designed with a human head on a lion's body. The figure of the lion, symbolizing the fierce hunter Sekhmet, was the foundation for many different sphinxes. Atop each were the different heads of Egyptian royalty. From this inspiration, perhaps, sprang the sphingolipids, whose foundation is the fierce molecule ceramide (Figure 1). Reknown for initiating apoptosis, ceramide is the messenger of death. And like the sphinx, the different sphingolipids have different head groups. While Egyptologists would prefer to delve more deeply into the secrets of the sphinx, our interests turn here to reveal more mysteries of the sphingolipids.
Perhaps the best known sphinx is the Great Sphinx of Giza, standing guard over the Second Pyramid of Khafra. Of similar fame amongst lipidologists is sphingomyelin, which contains phosphocholine as its head group (Figures 1 and 2). Less commonly, phosphoethanolamine can serve as the head group. Sphingomyelin is abundant on the outer leaflet of cellular plasma membranes. In the fluid surface of the cell, sphingomyelin and cholesterol form lipid rafts, which serve as platforms for proteins and protein assemblies involved in signal transduction.
A complicated story emerges when sugars are used as the head group, as is found in glycosphingolipids. These are present on the cell membranes of organisms from bacteria to man and, in vertebrates, are abundant in the brain. For example, galactosylceramide (GalCer), which has galactose attached to ceramide, is one of the most abundant molecules in the brain. There are well over a hundred different sugar units or assemblies that have been identified on the heads of different glycosphingolipids. The ‘neutral' versions commonly contain mono- or oligoglycosyl units, while the ‘acidic' varieties include sulfo-, phospho- and phosphono-linked sugars. Like sphingomyelin, glycosphingolipids coat the outer leaflet of the cell membrane. Here, they modulate cell-cell interactions by binding with complementary molecules on adjacent membranes. Of no less importance, they regulate the activities of neighboring proteins in their own membrane leaflet. These processes are essential for the development of multicellular organisms and for normal intercellular communication.
Ceramide is the simplest of the sphingolipids: with hydrogen as its ‘head' group, it is a diol. Like the other sphingolipids, it occurs in abundance in the outer leaflet of the cell membrane. It is synthesized de novo in the endoplasmic reticulum by the serine palmitoyl transferase pathway, starting with serine and palmitate. Perhaps more interesting, ceramide can be rapidly released from sphingomyelin by sphingomyelinases (SMase, or sphingomyelin phosphodiesterases). An acid SMase (aSMase) functions in lysosomes, while the neutral SMases (nSMases) are active at the plasma membrane. The aSMase can be activated by the TNF Receptor-Associated Death Domain (TRADD) and the FAS-Associated Death Domain (FADD) linked to the TNF receptor, indicating that ceramide generated through its activation is important in driving apoptosis. Deficiencies in aSMase activity, on the other hand, lead to sphingomeylin accumulation and Niemann-Pick disease.
A second sphingomyelinase, the nSMase2, is palmitoylated, which targets it to the inner leaflet of the plasma membrane. It is activated by oxidative and mechanical stress, UV light and TNF, in concert with FAN (factor associated with neutral nSMase), to produce ceramide acutely, leading to apoptosis. Slower nSMase2 activity, as induced in confluent epithelial cells, promotes differentiation and is necessary for normal monolayer formation: a defect in nSMase2 action results in impaired epithelial barrier function, as is found in atopic dermatitis. A third neutral form, nSMase3, is found on the Golgi and endoplasmic reticulum (ER), as well as the plasma membrane and is activated by DNA damage and nongenotoxic stresses. This nSMase is dysregulated in some forms of cancer.
Ceramide consists of sphingosine joined with a fatty acid by an amide linkage (Figure 3). The fatty acid chain length and degree of saturation varies, which affects the solubility and biological properties of the ceramide species. Sphingosine is an 18 carbon chain that terminates with two hydroxyl groups on either side of an amino group, with a single double bond. It is, in fact, the unusual structure of sphingosine which conjured up the mysteries of the sphinx to early chemists.
Ceramide is enzymatically cleaved to sphingosine and fatty acid by ceramidases. The acid ceramidase, also known as N-acylsphingosine amidohydrolase 1, is encoded by the gene ASAH1. This ceramidase is a heterodimer consisting of two transcript variants, the non-glycosylated α variant and the post-translationally truncated and glycosylated β variant. It is particularly important in lysosomal degradation of ceramide in that deficiency of this acid ceramidase causes a lysosomal storage dysfunction known as Farber disease. The neutral ceramidase, also known as N-acylsphingosine amidohydrolase 2, is an integral membrane protein that is important in the generation of sphingosine as a signaling molecule.
Sphingosine is phosphorylated to give S1P by two distinct sphingosine kinases, SPHK1 and SPHK2. SPHK1, the better studied form, is activated by many stimuli, including TGF-β, IL-1β, TNF-α, Platelet-derived Growth Factor (PDGF), HGF, LPS, Respiratory Syncytial Virus (RSV), insulin, anaphylatoxins, and BCR/ABL. Phosphorylation of Ser311 by Extracellular Signal-Regulated Kinase (ERK), reversed by PP2A, causes plasma membrane targeting and activation of SPHK1. SPHK1 is best known as a survival, or anti-apoptosis, enzyme, with additional positive effects on cell motility and proliferation, resulting from the production of S1P. These actions contribute to tumor growth and metastasis. In addition, SPHK1-derived S1P activates endothelium, regulating endothelial barrier homeostasis, primes neutrophils, activates macrophages and promotes phagosome maturation, and increases immune cell motility and function.
While some of the actions of SPHK2-derived S1P overlap those of SPHK1, SPHK2 may promote, rather than prevent, apoptosis. SPHK2 can also be phosphorylated by ERK, which leads to activation and increased cell motility.
S1P was first thought to have its effects intracellularly, acting as a second messenger, interacting with and modulating the activities of specific target proteins. While this remains a possibility, current research focuses on the G protein-coupled receptors that respond to S1P. These receptors were initially identified as EDG (endothelial differentiation gene) receptors and were orphan receptors. With the identification of S1P as a ligand for five of the EDG receptors, these have been renamed: S1P1 (EDG1), S1P2 (EDG5), S1P3 (EDG3), S1P4 (EDG6), and S1P5 (EDG8). S1P1 and S1P3 were first isolated from endothelial cells, while S1P2 was first found on rat brain and vascular smooth muscle cells, S1P4 was found on dendritic cells and S1P5 on rat PC12 (prostate cancer) cells. The five S1P receptors share high sequence identity with the CB and lysophosphatidic receptors, which are also GPCRs for lipid ligands.
Cells overexpressing S1P1 and S1P3 have been shown to signal primarily through Gi, activating ERK and PLC, while inhibiting adenylate cyclase (AC)-mediated increase of cAMP in a pertussis toxin-sensitive manner (Figure 4). In addition, these receptors activate PI3K, leading to Rho-mediated cell motility through Akt and Rac signaling. While S1P2 and S1P3 activate ERK, PLC and PI3K via Gi, both can also activate PLC through Gq. Commonly, the main role of S1P2 is to oppose or balance the signaling of S1P1 through its G12-mediated activation of Rho, which inhibits Rac. S1P4 and S1P5 have been less well studied. Both the activation of ERK and PLC by S1P4 and the inhibition of AC by S1P5 are pertussis toxin-sensitive, showing a role for Gi in this signaling. In lymphocytes, activation of Cdc42 from S1P4 contributes to S1P-mediated chemotaxis. Both S1P4 and S1P5 also suppress Rac via G12/13-activated Rho. Moreover, S1P5 activation of Jun N-terminal Kinase (JNK) and ERK is insensitive to pertussis toxin, through an unknown mechanism.
The cycling and turnover of S1P receptors is important in regulating the response of cells to S1P. The levels of S1P are relatively high in the blood but low in the lymph and tissues. As a result, circulating leukocytes and mature endothelial cells are constantly exposed to S1P, resulting in receptor turnover and refractory responsiveness to S1P. In contrast, lymphocytes circulating in the lymph, as well as resident tissue leukocytes (e.g., mast cells, macrophages, and dendritic cells) are highly responsive to S1P. The forced turnover of S1P receptors, using S1P analogs, has proven to be an effective way to suppress lymphocyte emigration from lymphoid organs, resulting in allograft protection.
The enigmatic sphingolipids and their derivates are, like the Great Sphinx, beginning to reveal their secrets. Some of the most exciting discoveries are presented in recent reviews.1-4
Figure 1. Sphingolipids, like the sphinx, are composed of diverse subunits
Figure 2. Changing the head group changes the sphingolipid
Figure 3. Sphingosine, derived from ceramide, is phosphorylated by sphingosine kinases
Figure 4. Extracellular S1P activates five different GPCRs. Key signaling pathways are indicated in boxes
1. Rivera, J., Proia, R.L., and Olivera, A. Nat. Rev. Immunol.8, 753-763 (2008).
2. Hannun, Y.A., and Obeid, L.M. Nat. Rev. Mol. Cell Biol.9, 139-150 (2008).
3. Marsolais, D., and Rosen, H. Nat. Rev. Drug Discov.8, 297-307 (2009).
4. Hla, T., Venkataraman, K., and Michaud, J. Biochim. Biohys. Acta1781, 477-482 (2008).