Sphingolipids are important signalling molecules and thus it is very important to understand how they are generated. Sphingolipid biosynthesis shows a conserved compartmentalization in eukaryotic cells. Their synthesis begins in the endoplasmic reticulum and is completed in the Golgi apparatus. We now know quite a bit about the topology of the reactions in the pathway, but certain critical steps, such as ceramide synthesis, are still poorly understood. In the present paper, we discuss the latest views on this subject.
Sphingolipids are important components of membranes, in particular the plasma membrane, where they are thought to play important roles in membrane structure. In addition to their roles in membrane structure, hydrolysis of sphingomyelin can generate ceramide, an important intracellular signalling molecule . Ceramide can subsequently be hydrolysed to fatty acid and sphingosine, the latter also implicated in signalling pathways. Sphingosine can be phosphorylated by sphingosine kinase to generate a sphingosine-1-phosphate, a potent ligand for the EDG (endothelial differentiation gene) receptors involved in important steps of development, immune cell circulation and other physiologically relevant pathways .
These signalling molecules can also be generated as intermediates of the de novo synthesis of sphingolipids, and certain physiological responses seem to depend more highly on de novo synthesis than degradation. We have studied the de novo synthesis and organization of sphingolipids in yeast. Their synthesis is strongly induced upon heat stress , and we have shown that the synthesis of sphinganine is required for translation of heat-shock mRNAs upon heat stress .
The sphingolipid biosynthetic pathway in yeast is similar to the pathway in animal cells (1) and shares a similar spatial organization with the steps up to and including generation of ceramides taking place in the ER (endoplasmic reticulum) and most subsequent steps occurring in the Golgi compartment . Palmitoyl-CoA and serine are condensed by SPT (serine palmitoyltransferase) to form 3-keto-sphinganine. This entry step into the sphingolipid biosynthetic pathway is a likely point for metabolic control, although little is known about how this enzyme is regulated. Some experiments from animal cells and mathematical modelling suggest that flux through the pathway could be regulated by availability of the substrates of SPT . The precise topology of SPT is not known with certainty, but it has been proposed that the enzyme has its active site exposed to the cytoplasm and no evidence exists for the presence of one of the substrates, serine, in the lumen of the ER. The next step, catalysed by 3-sphinganine reductase, generates sphinganine while oxidizing NADPH and has also been proposed to occur on the cytoplasmic face of the ER .
Sphingolipid biosynthesis in yeast
Sphinganine can be further hydroxylated to form 4-OH sphinganine and both of these long chain bases can be condensed with fatty acyl-CoA to obtain dihydroceramide or phytoceramide respectively. The yeast ceramide synthase is composed of three subunits, Lag1p, Lac1p and Lip1p . Lag1p and Lac1p are two highly homologous proteins that have a genetically redundant function and therefore deletion of both genes is required to negate ceramide synthase activity [8,9]. Lip1p is essential for ceramide synthase activity. Yeast ceramide synthase prefers fatty acyl-CoA with long chain fatty acids (C26) and this is seen in cells because the vast majority of ceramides have C26 fatty acids. On the other hand, mammalian cells attach a wide range of fatty acids to sphinganine to produce a variety of different ceramides with preferences to be incorporated into different complex sphingolipids. Nevertheless, mammalian homologues of Lag1p are capable of directing ceramide synthesis in yeast . Interestingly, when mammalian Lag1 homologues, now collectively called the LASS genes, are overproduced in mammalian cells, they lead to increases in ceramides with particular chain lengths [5,11]. This implies that Lag1 and its homologues are the subunits that confer fatty acyl chain length specificity. It is an enigma how a hydrophobic membrane protein such as the ceramide synthase can at once recognize the length of the fatty acyl chain and at the other end of the fatty acid perform a reaction involving hydrophilic groups.
In contrast with the previous steps, the topology of the ceramide synthase reaction is less clear. One of the substrates, sphinganine is probably generated on the cytoplasmic side of the ER . However, when exogenous substrate is added to yeast cells, it is first phosphorylated on the cytoplasmic side of the ER, then dephosphorylated, most likely in the ER lumen, before incorporation into ceramide [13,14]. Therefore this substrate may be able to access the enzyme from either side of the membrane. We have worked on the topology of the ceramide synthase enzyme. The Lip1 subunit of the ceramide synthase is a single-pass transmembrane protein with the bulk of the molecule localized in the lumen of the ER. The cytoplasmically disposed residues are not required for enzyme localization or for its activity . The Lag1 and Lac1 subunits of the ceramide synthase are polytopic membrane proteins whose predicted topology varies depending upon the algorithm used. The number of predicted transmembrane domains varies as well as the choice of their locations. Therefore only an experimental approach is likely to shed light on to the topology of this enzyme. We have investigated this issue using two independent methods. Our results suggest that the protein spans the membrane eight times with both the N- and C-termini on the cytoplasmic side of the ER. The Lag homology domain would span two transmembrane α-helices, with the critical residues for the enzymatic reaction most likely to be close to the middle of the lipid bilayer. From these results, we speculate that the ceramide synthase enzyme might contain a hydrophilic core within the ER membrane that would permit the enzymatic reaction to take place in an appropriate environment for both the hydrophobic fatty acyl chain, which could remain in the membrane, and for the hydrophilic CoA and reaction involving charged residues.
Once ceramide is made, it must be transported from the ER to the Golgi compartment. In mammalian cells, major progress has been made in understanding this mechanism because a protein, CERT (ceramide transporter), has been identified that participates directly in this transport step [15,16]. We have shown previously that ER–Golgi membrane contacts are most likely to be important for ceramide transport . This makes it tempting to speculate that CERT also works at membrane–membrane contact sites.
Once in the Golgi compartment, ceramide is converted into the complex sphingolipids. In yeast, the first product is IPC (inositolphosphorylceramide) and the reaction is carried out by the Aur1 protein . It would be very interesting to know precisely why all eukaryotic cells compartmentalize their sphingolipid biosynthetic pathways and separate ceramide synthesis from complex sphingolipid synthesis. Further additions to IPC are a mannose residue and an inositol phosphate to form mannosyl IPC and mannosyl di-inositol-phosphorylceramide and these reactions also take place in the Golgi compartment .
Non-Vesicular Intracellular Traffic: Biochemical Society Focused Meeting held at Goodenough College, London, U.K., 15–16 December 2005. Organized and edited by S. Cockcroft (University College London, U.K.) and T. Levine (Institute of Ophthalmology, London, U.K.).