Intracellular lipid transport

Although much of my laboratory now works on various aspects of lipid biology, my interest in lipids started accidentally. We were studying membrane protein traffic, and the transferrin receptor and the low-density lipoprotein receptor were the major objects of our interest. While recognizing the importance of membrane lipids, I mainly ignored them. I also assumed that much was known about lipid trafficking — just not by me. Eventually, I realized that much is known about lipid trafficking, but there are fundamental issues in this very important area of biology that are not well understood by anybody.
Basic mechanisms for maintaining distinct lipid compositions in different organelles are only partially understood, which means that this is an area where fundamental principles are still awaiting discovery. In addition to being a fascinating area of scientific inquiry, intracellular lipid transport plays a key role in dyslipidemias, which are a growing health problem throughout the world.

cholesterol trafficking pathway
Illustration of some cholesterol trafficking pathways. Courtesy of David B. Iaea.

My first foray into lipid transport was using fluorescent lipids as a control for a membrane protein trafficking experiment. We found that, after endocytosis in fibroblasts, the recycling of the lipid analog was kinetically and morphologically indistinguishable from the recycling of the transferrin receptor. This supported our hypothesis that specific protein-protein interactions were not required for rapid and efficient recycling of transferrin receptors.
This satisfying result was published, but it left some gnawing questions. The lipid analog we studied recycled to the plasma membrane with nearly 100 percent efficiency, but obviously some lipids were required to form the membranes that went to late endosomes. Were there lipids that would be targeted preferentially to these organelles? If so, how? While we did make some progress on this by showing that some fluorescent lipid analogs could be sorted efficiently to late endosomes, this type of sorting remains poorly understood for natural lipids in cells.
Our results with fluorescent lipids were an example of one mechanism for lipid sorting: segregation of a subset of lipids during the formation of vesicles and tubules in membrane vesicular trafficking. In a recent Lipid News column, Patricia Bassereau discussed the role of lipid curvature induced by proteins on the selection of lipids into highly curved membranes, such as those formed in vesicular membrane trafficking. While the preference of individual lipid molecules for curved regions does not impose a strong selection, curvature can contribute to lipid sorting in lipid mixtures in which the composition is close to a phase separation boundary (see Callan-Jones et al and Sorre et al).
Just as lipids are recycled at the plasma membrane, there must be similar mechanisms to sort lipid components in anterograde and retrograde transport at each step of the biosynthetic pathways. While general principles based on lipid phase separation and curvature preferences also are likely to play a role in these secretory pathways, much remains to be learned about how this works.
The second major mechanism for lipid sorting involves nonvesicular transport processes that exchange lipids among membranes. There are several examples of lipids that are delivered from a specific donor organelle to a specific acceptor, based in large part on binding specificity of the carrier proteins.
One example is the transport of ceramide by the ceramide-transfer protein, CERT, which has a ceramide-binding START domain. A pleckstrin homology domain can target CERT to Golgi membranes, and a FFAT motif binds the endoplasmic reticulum protein VAP. Thus, CERT can shuttle efficiently ceramide from its site of synthesis on the cytoplasmic side of the ER to the cytoplasmic side of the Golgi, where it can be converted to glucosylceramide. This type of selective nonvesicular transport process plays a role in determining the specific membrane composition of different organelles. However, in general, we do not know the relative contributions of vesicular and nonvesicular transport pathways to the flow of lipid between organelles.
Similarly, cholesterol, which is mainly synthesized in the ER, reaches organelles, including the plasma membrane, independently of vesicle transport pathways. Our work on LDL internalization naturally led to questions about how cholesterol gets from lysosomes to the ER, where the cell’s sterol regulatory machinery is located. Using fluorescent sterols, we have made some progress in identifying nonvesicular sterol transport mechanisms, but overall it has been challenging to identify the carriers for cholesterol (or ergosterol in yeast).
In yeast, members of the oxysterol binding protein family have been proposed as nonvesicular sterol carriers, but elimination of many (or transiently all seven) of the OSBPs does not fully block sterol transport between the ER and the plasma membrane. In mammalian cells, an additional family of lipid-binding proteins, the START domain proteins, has been proposed to play a role in nonvesicular transport of sterols and other lipids, but much more work is required to understand their role (see Mesmin et al and Clark et al).
The third mechanism for regulating lipid composition of organelles is enzymatic transformations in specific organelles. This can include modifications of head groups as well as exchange of acyl chains. These reactions are carried out by enzymes that are localized to specific organelles, which can lead to local changes in the lipid composition. Additionally, these transformations can lead to changes in curvature preference or in the susceptibility to extraction of a lipid and binding to a nonvesicular transport protein. For example, removal of an acyl chain from a glycerophospholipid creates a lysolipid that has different curvature preferences and is easier to extract from the bilayers compared with the parent lipid.
A major challenge in the field is to understand how all of these mechanisms are integrated to maintain the proper balance of lipid compositions in various organelles. An intriguing finding that provides a possible general mechanism for such integration is that several lipid transport proteins also are involved in regulating the vesicle transport machinery. It has been proposed that these proteins may serve as coincidence detectors to ensure that an appropriate set of lipids is available in a donor compartment before allowing the formation of a transport vesicle or tubule.
Studies in a variety of yeast mutants have been interpreted as indicating that a phosphatidylinositol transport protein, Sec14, coordinates lipid levels with membrane transport in the trans-Golgi network and endosomal compartments. Members of the OSBP family have been implicated in the transport of sterols and other lipids, but they also can regulate vesicle formation and the structure of the trans-Golgi network. It is unclear if these proteins play a significant role in sterol transport among organelles or if they are primarily lipid sensors that regulate metabolic pathways and membrane trafficking.
New tools, including lipidomics and high-resolution fluorescence microscopy along with genetics and molecular biology methods, finally are allowing us to make significant headway in understanding the details of intracellular lipid transport. It is likely that sophisticated computational modeling and systems biology approaches will be required to develop an integrated understanding of the many processes that play a role in determining lipid distribution.

Fred MaxfieldFrederick R. Maxfield (frmaxfie@ is a professor in the biochemistry department at Weill Cornell Medical College.