Four recent studies take a close look at how molecules cross the gap between membrane contacts within a cell, offering insights into crucial cellular activities. Three of the papers elucidate the transport of cholesterol; the fourth suggests that such contact sites may set up distinct regions of the plasma membrane with specialized roles. Though the basic mechanisms still require plenty more investigation, clarifying these mechanisms could provide researchers new ways to target cholesterol pathways, or perhaps offer ways to alter cell membrane signaling molecules that regulate processes such as cell growth and division.
Cholesterol is the most abundant sterol in mammals and a major component of the plasma membrane. Sterols move through the cell not only through membrane vesicles but also via membrane contact sites.
Some molecules that transport sterol across contact sites do so via a protein module called the StART domain. Previous studies showed that the StART domain resembles a cup with a lid, which can be filled by a single sterol molecule. But until now, the structure of the domain had not been examined with sterol bound to it.
In three of the new studies, published in EMBO Journal, PNAS, and the Journal of Biological Chemistry, researchers report accomplishing this by examining the sterol-bound domain of yeast proteins in the LAM family (lipid transfer proteins anchored at membrane contact sites—also called ltc for “lipid transfer at contact site”). The LAM proteins contain a domain closely resembling the StART domain (called the StART-like StARkin, or VASt domain). Some LAM proteins shuttle sterol between organelles, and others operate between the plasma membrane and the endoplasmic reticulum, a vast network of membranes within the cell.
All three studies showed that sterol entered the cup, but stayed in the top portion, and the lid closed only partially. This positioning may enable sterol to slip quickly in and out of the cavity, says Tim Levine, a cell biologist at University College London who was not involved in the study.
And speed is indeed key—in vitro experiments with artificial vesicles containing the LAM StARkin domains showed faster sterol transport than previous reports. That may help put to rest arguments that the LAM proteins act too slowly to be bona fide transporters, says Levine, who recently helped identify this group of proteins.
Levine says that sterol transfer proceeds when the domain lands on the membrane, orients itself in the right fashion in some way, and somehow gets a kick to open the lid. The sterol then eases into the cavity, and the domain diffuses to the other side of the gap to release the sterol. The flexibility of the other parts of the LAM protein, which spans the membrane, enables the StARkin domain to move around. “It’s lovely … bendable,” he says.
One of the studies hints at what drives the StARkin domain to ping pong between the two membranes. In vitro assays suggest that a mammalian LAM protein can transport not only sterol but another lipid, PI(4,5)P2, a cell signaling mediator. That finding hints that this protein may operate via a “counter-exchange mechanism” moving PI(4,5)P2 to one membrane and cholesterol to another, says Karin Reinisch, lead author of the study and a professor of cell biology at Yale University. Reinisch’s data also suggest that the other two mammalian LAM proteins bind to and shuttle sterol.
A fourth study, in eLIFE, takes a broader view. Marina Besprozvannaya, a postdoctoral researcher at University of California Davis, and her colleagues report that a human ortholog of a LAM protein, GRAMD1a, localizes to the contact between the endoplasmic reticulum and the plasma membrane, consistent with a role as a cholesterol transporter at that location. They also looked at GRAMD2a, which does not have a StARkin domain, showing that it spans the space between the endoplasmic reticulum and the plasma membrane, where it binds to PI(4,5)P2. GRAMD2a seems to help tether the membranes and also marks the site for recruitment of a molecule central to calcium transport.
GRAMD1a and GRAMD2a not only have distinct functions, they are also located at distinct contact sites, the researchers found. William Prinz, a cell biologist at the National Institutes of Health in Bethesda, MD, says this study fits with an emerging view that membrane contact sites help set up specialized regions of the plasma membrane. He explains the idea: “You have special domains in the plasma membrane that have a special lipid and protein composition, and that is driven by them touching the endoplasmic reticulum.” Perhaps GRAMD2a and other tethers act to organize specific types of contacts, he adds.
Jodi Nunnari, lead author of the eLIFE study and a cell biologist at UC Davis, says the findings lead to a host of new questions. Do the various kinds of membrane tethers—GRAMD2a being one—interact with each other to cooperatively build different types of membrane contacts? How does lipid composition affect tethering? “This study has really pushed forward an understanding of the complexity of these contact sites in human cells,” she says, “How dynamic are these structures, and how does that tune the functional output of cells?”
Nunnari is also interested in the more practical implications. She notes that GRAMD1a and GRAMD2a have both been linked to cancer though the biology behind this connection is not well understood. Prinz adds that the new sterol-bound structures may also be interesting for researchers who study diseases associated with toxic sterol accumulation and defects in cholesterol biosynthesis.