Interorganelle Trafficking of Lipids: Preface for the Thematic Review Series

Lipids are the major constituents of all cell membranes and play dynamic roles in organelle structure and function. In eukaryotic cells, different organelles are characterized by different lipid compositions, but with only a few exceptions (e.g. cardiolipin) this tends not to be an absolute presence or absence of a particular lipid class in a specific organelle. Since most lipids are synthesized in the endoplasmic reticulum (ER), the cell must have a system to transport the newly synthesized lipids to their final destinations. For example, cholesterol is synthesized in the ER, but the level of cholesterol is quite low in this organelle and highest in the plasma membrane (PM) at steady state. Although we have known that the intracellular transport of lipids is required within cells because the loci of synthetic enzymes is usually restricted, elucidation of the mechanisms of transport has been quite slow, particularly in comparison with the progress in understanding molecular mechanisms of protein transport. Pioneering studies by Kaplan and Simoni in 1985 showed that the transport pathways for cholesterol and phosphatidylcholine from the ER to the PM are distinct from the typical transport pathway for proteins based on the aspects of energy dependence and drug sensitivity 1, 2. Yet, the molecular machinery responsible for ER-to-PM transport of cholesterol or PtdCho has not been identified. Biochemical studies on lipid-transfer proteins (LTPs) started in the late 1960s by using in vitro assay systems 3, and although many LTPs have been identified, purified and their cDNAs cloned 4, in many cases we still cannot demonstrate the transport of the cargo lipids via the LTPs in living cells. As an alternative to biochemical approaches with LTPs, Pagano and coworkers developed cytological approaches using fluorescent lipid analogs 5, 6. Findings with fluorescent lipids have provided important insights into the dynamics of lipid transport in intact cells and cell biology of biomembranes. However, there still remains some uncertainty about the extent to which the behavior of such analogs reflects that of their natural counterparts in cells 7-9. Multiple genetic studies and experiments with classical inhibitors of vesicle trafficking processes have demonstrated that robust interorganelle lipid movement can occur by non-vesicular routes. Significant research is now focused upon identifying the non-vesicular transport mechanisms that in some cases are mediated by specific LTPs and protein complexes assembled at membrane contact sites 10. In addition, specialized biological functions of membrane lipids have been increasingly appreciated and range from the transient lipid micro-domains within the plane of the membrane (lipid-rafts) to that of signaling modulators. In the coming months, Traffic will publish a series of reviews that highlight some of most exciting advances in the field. This overview is a preview of what is to come. The sites where different organelle membranes come into close apposition within nanoscale ranges (∼10–50 nm), are now commonly referred to as membrane contact sites. The close proximity of different organelle membranes was observed in the early era of cell biology by classical electron microscopy 11. The earliest recognition of organelle contact appeared as an association between mitochondria and the ER in gill-relevant cells of fish 12. A much finer description of the contact structures was later resolved by 3D-image reconstructing cryo-electron tomography 13, 14. Almost all types of organelles are now suggested to have their own contact sites and these contact sites have been the subject of increased attention 15-19. Molecular elucidation of interorganelle contact sites began from the description of a membrane fraction associated with mitochondria (named MAM) 20, which was demonstrated to be a subdomain of the ER. MAM exhibited an enriched activity of phosphatidylserine synthase, an ER-resident enzyme 20. Later, genetic approaches revealed various genes for the formation of ER–mitochondria membrane contact sites in the model single cell eukaryote Saccharomyces cerevisiae 21-23. As mitochondria have a double membrane system, consisting of an outer membrane and an inner membrane, the metabolism of lipids in mitochondria often requires ‘intra-mitochondrial’ trafficking processes in addition to ER-to-mitochondria trafficking processes 24. Tamura and his colleagues will summarize the latest findings on inter-ER/mitochondria and intra-mitochondrial transport of lipids. To many investigators unfamiliar with the cell and molecular biology of lipids, the multiplicity of lipid classes and metabolic routes and subcellular locales of their synthesis can appear daunting. Glycerophospholipids are the predominant class of lipids in biomembranes from bacteria 25, 26, archaea 27, 28, yeast 29, plants and mammals 30-32. Vance will provide an overview describing the landscape of phospholipid synthesis and transport in mammalian cells. Cholesterol is also a fundamental membrane lipid of all vertebrate cells, and it has gathered great attention on both physiological and pathological aspects. In addition, cholesterol metabolites such as steroid hormones and bile acids play crucial roles in mammals. While the biochemical pathways of sterol metabolism have been extensively elucidated, the mechanisms underlying intracellular transport of sterols remain surprisingly poorly understood. Nevertheless, the finding that proteins containing a variety of lipid-binding domains play a key role in the synthesis of steroid hormones has led to a model of cholesterol binding-domains participating in multiple physiological intracellular lipid-transport events 33, 34. Papadopoulos and colleagues summarize the historical perspective, most recent advances and future perspectives on intracellular trafficking of sterols. Sphingolipids are also ubiquitously present in eukaryotes 35, 36. The de novo synthetic pathway of ceramide, the precursor to complex sphingolipids synthesis, is conserved through evolution 37, 38. However, the chemical structures of the polar head groups of complex sphingolipids are highly diverse among different organisms as well as between tissues and cell types within the same organism, while those of complex glycerolipids are well conserved from bacteria to human 35, 36 . Ceramide is synthesized at the ER, while the attachment of polar head groups to ceramide mainly occurs in the Golgi apparatus 35, 36. In mammalian cells, the synthesis of glucosylceramide has been suggested to occur largely in the cis Golgi region, while the synthesis of sphingomyelin occurs in the medial or trans regions 39, 40. Thus, it is a mystery how cells can dispense ceramide to the trans Golgi site for sphingomyelin synthesis without its pre-consumption at the cis Golgi site for glucosylceramide synthesis. The discovery of the ceramide transport protein CERT in 2003 provided a good explanation: CERT delivers ceramide from the ER to the trans Golgi region in a non-vesicular manner, thereby bypassing the cis Golgi altogether 35, 41. This mechanism is in contrast to proteins, which transit through the cis Golgi before reaching the trans Golgi 42. Consistent with the function of CERT, it contains a Golgi-binding PH domain and an ER-binding FFAT motif, allowing it to form membrane contact sites between the ER and the Golgi 15, 41. Other LTPs involved in the transfer of different sphingolipids have also been identified 43, 44. Yamaji and Hanada summarize recent advances in understanding the function of LTPs in the intracellular trafficking of sphingolipids. The family of lipids is not restricted to the structural constituents of membranes but also encompasses the lipophilic vitamins A and E. Uptake and transport of lipid vitamins is vital for normal development and sustainability of multicellular organisms, and its genetic deficiency often appears as hereditary diseases. Moreover, obvious relationships between genetic mutations and resulting biochemical dysfunctions have sometimes led to prophylactic or therapeutic strategies: e.g. vitamin E supplementation can ameliorate ataxia in isolated vitamin E deficiency, which is caused by mutations in the gene encoding the α-tocopherol (vitamin E) transfer protein 45, 46. Kono and Arai summarize recent advances in lipid vitamin transfer proteins and familial vitamin deficiency. Plant cells have at least three membrane features that are distinct from animal cells: First, plant cells have a unique organelle, the chloroplast. Secondly, plant cells contain a large vacuole. Third, the PMs of plant cells are surrounded by a cell wall rich in cellulose, which prevents cells from bursting in hypo-osmotic environments. The membrane lipid composition in plant cells, in particular that of the chloroplast, is considerably different from that of animal cells: e.g. monogalactosyldiacylglycerol and its derivatives are the major lipid class in chloroplasts and other plastids, while galactosyl glycerolipids are absent or rare in mammalian cells. Accordingly, plant cells have transport systems for plant-specific lipid classes 47, 48. Plant cells also differ from animal cells in their subcellular localization of lipid synthetic enzymes, thereby requiring specialized routes of lipid trafficking not found in animal cells. Benning and his colleagues summarize recent advances on lipid trafficking in plant cells, especially into and out of chloroplasts. Membrane lipids frequently participate in the interaction between host cells and pathogens (i.e. toxins, viruses, bacteria, fungi and protists) 49-52. For example, specific types of glycosphingolipids in the PM of host cells serve as membrane receptors of several bacterial toxins such as cholera toxin and Shiga-like toxins. All lipid constituents of viral envelopes are derived from host cells. In addition, the obligate intracellular bacterium Chlamydia trachomatis behaves as a metabolic parasite, and requires host cell-derived sphingolipids for its growth in the parasitophorous vacuole ‘inclusions’ 53. Although non-obligate intracellular microbes appear to have all the metabolic pathways for the synthesis of their essential lipids, invading microbes often steal host lipids for their own aims. One simple but elegant strategy for this theft is to hijack host-encoded lipid transport systems to direct newly synthesized lipids to the microbes 54, 55. Invasive microbes also frequently modify host membrane lipids to enhance entry into host cells, facilitate formation of parasitic or proliferative vacuoles inside the cells, and promote exit from the cells 56, 57. In addition, while protein lipidation rarely occurs in bacteria, bacterial effectors injected into host cells sometimes undergo host-mediated lipidation 58. Helms and his colleagues summarize lipid trafficking between pathogens and host cells, an understanding of which may provide the molecular basis for new strategies to tackle infectious diseases. Structural analysis of proteins using X-ray crystallography and NMR has provided profound insight into how proteins, including LTPs, recognize their substrates. The 3D structures of many LTPs have been resolved at the atomic level, which has greatly enhanced our insight in the interactions between LTPs and their lipid ligands 59-64. Although it remains difficult or impossible to see real time behavior of membrane-bound proteins at the atomic resolution level (∼2 Å), such limitations should be overcome by rapidly developing new techniques for structural biology in the near future. Brown and his colleagues summarize the new insights into LTP function derived from recent structural studies. Many of the reviews in this series focus on ‘inter’-organelle trafficking of lipids. However, we cannot forget that lipids are transported between leaflets of the same lipid bilayer. This process occurs, when a lipid that is synthesized on the luminal side of an organelle membrane is subsequently translocated to the non-luminal side of the same membrane, and then later transported to another organelle by a non-vesicular mechanism. Thus, in some pathways intra-membrane transport can be an obligate step prior to inter-membrane transfer. Similarly, a lipid molecule that reaches the cytosolic face of a target organelle membrane may traverse the bilayer before it can undergo further metabolism or exert its activity in the lumen of the organelle. Therefore, inter- and intra-membrane transport of lipids may be functionally connected in cells. In addition, the directional translocation of particular lipid species, such as phosphatidylserine, to the cytosolic leaflet contributes to the asymmetric structure of the PM. Regulated disruption of PM phospholipid asymmetry is a signaling device important in apoptosis and blood clotting. Several types of proteins harboring multiple transmembrane spanning domains have been identified as lipid translocases (flippases, floppases and scramblases) for transbilayer movement of lipids 65-68. Graham and his colleagues summarize recent advances in our understanding of lipid translocases, and also introduce their roles for connecting intra- and inter-organelle transport of lipids. In many instances, our views of lipid transport have been dominated by a paradigm encompassing a donor membrane, an acceptor membrane, a cargo lipid and a single carrier that may be a soluble protein for interorganelle traffic, and an integral protein for intra-membrane traffic. Historically, this simplified view has been useful, but broader perspectives now need to be considered that include recruitment of soluble and membrane proteins to specialized membrane domains to assemble lipid transport machines. This recruitment process is likely to be influenced by local lipid environments either enriched in cargo lipids, or signaling lipids such as polyphosphoinositides, or both. The interactions of donor and acceptor membranes may also be regulated by processes similar to receptor-ligand recognition reactions, to assemble mature lipid transport complexes between membranes. Tethering complexes between donor and acceptor membranes may play a role in stabilizing the association of donor and acceptor membranes either with, or without participation in the transport reaction per se. In the article by Voelker and his colleagues, some elements of the formation of mature transport processes driven by both specific lipids and protein associations between donor and acceptor membranes are discussed as part of the mechanism for transporting newly synthesized phosphatidylserine in yeast, from the ER to the locus of phosphatidylserine decarboxylase 2. Although the first intracellular LTP was identified more than half a century ago, it took many years to demonstrate that LTPs play crucial roles in the transport of their lipid ligands. Indeed, various studies have demonstrated that some LTPs are not essential for transport events that are anticipated to occur during the metabolism of specific lipid types in cells: e.g. knock-out of the gene encoding the phosphatidylcholine transfer protein, a typical LTP, showed no dramatic impact on phosphatidylcholine metabolism in mice 69. This might be due to functional redundancy, but it also invoked the scientifically sound skepticism that the biochemical property of an LTP in vitro might be irrelevant to its physiological functions in vivo. In addition, as the genomes have been resolved in more organisms, additional types of proteins have been identified that contain lipid-transfer domains. However, it has been difficult to envision lipid transport properties for some of these proteins. For instance, proteins having the lipid-transfer START domain in plants often contain the DNA-binding homeobox domain, which suggests their role as transcription factors with yet to be identified lipid ligands as their regulators 70. Physiological roles of LTPs as ‘lipid-binding regulators’, not as ‘lipid carriers’, are best exemplified by the Sec14 protein. Sec14p was genetically identified as a product of a gene essential for transport of secretory proteins from the trans Golgi network in the yeast S. cerevisiae, and it was also biochemically identified to be a phosphatidylinositol transfer protein 71. Several other LTPs have amino acid sequences similar to that of the ‘lipid-transfer domain’ of Sec14p (CRAL/TRIO domain), and some members of the CRAL/TRIO LTP family are most likely to mediate intracellular transport of their specific lipid ligands 45. However, a series of elegant studies showed that the primary function of Sec14p is not inter-membrane transfer of lipids but rather regulation of effectors by changing its conformation depending on bound lipid ligands 72. Such regulatory or scaffolding functions with lipid ligands are also suggested in other lipid-transfer domain families 73, 74. Mousely and Bankaitis review the recent advances in this field. In conclusion, some LTPs have been shown to participate in inter-membrane transport of lipids in cells, while some other LTPs are better regarded as regulators by using their lipid substrates as molecular switching ligands. Several LTPs may play two or more roles, and it is likely that we will discover even more diverse roles for LTPs in the near future. Cytosolic LTPs are topologically inaccessible to lipids oriented to the luminal side of organelles. Hence, there is no doubt that intra-membrane transport systems and inter-organelle vesicular transport systems are also important for the network of lipid trafficking events. More comprehensive understanding of lipid traffic in cells will hopefully be available in the next decade. We thank all contributors to this thematic review series for their constructive comments to the preface. This work was supported by the Japan Society for the Promotion of Science KAKENHI Grant number 25670065 and also by Takeda Science Foundation (to K.H.)