Most of us are at least a little concerned about cholesterol, but for some scientists, it's an obsession. Investigators who study cholesterol closely—deep inside the cell—are finding that it has some rather fascinating properties.

"It turns out cholesterol actually has a very important role in organizing the cell membrane," says Philip A. Beachy, an HHMI investigator at The Johns Hopkins University School of Medicine. "This is a huge area of cell biology right now, and cholesterol is an important player."

HHMI investigator Matthew P. Scott, a biologist at Stanford University School of Medicine, believes cholesterol can shed light on how cells communicate. In what he calls an "exquisite interplay," cells send signals to each other as they multiply—transmitting information that determines a cell's ultimate function and destiny. Thus, "cholesterol is arising as a very interesting molecule for signaling processes, in addition to its known role in affecting membrane properties," Scott says.

MYSTERIOUS PROTEIN
Scott, who had been studying the signaling proteins that direct body patterning during embryonic development in the fruit fly Drosophila and in mice, found himself drawn into the study of human disease when NPC1, one of two proteins behind a devastating disorder called Type C Niemann-Pick disease (NPC), turned out to resemble Patched, a protein that regulates development.

"When the NPC1-Patched connection was found, a new opportunity arose to learn about development from cholesterol metabolism and vice versa," says Scott. NPC1 normally helps transport excess cholesterol out of the cell. But if the protein is disrupted, as in NPC, transport processes inside the cell fail, high levels of cholesterol accumulate, and certain cells in the brain and other tissues are damaged or killed. The unfortunate few who are born with the disease rarely live past childhood, because basic bodily functions deteriorate progressively and there is no known treatment.

"The cell has developed elegant control systems to handle cholesterol because it has such a profound impact on the properties of the cell," says Scott. "But much remains to be learned about the roles of cholesterol in cell properties and signaling mechanisms. The NPC proteins are providing a way to try to understand the components of the cholesterol-processing pathway."

Two articles published in the March 4, 2003, issue of the Proceedings of the National Academy of Sciences (PNAS) describe the structure and cholesterol-binding properties of NPC2, the second protein that can cause NPC, which was discovered three years ago by Peter Lobel, professor of pharmacology at Robert Wood Johnson Medical School-University of Medicine and Dentistry of New Jersey (UMDNJ). NPC2 is responsible for about 5 percent of NPC cases, but apart from the fact that it can bind cholesterol, little else has been reported about it until now. Lobel recruited HHMI investigator Ann M. Stock, also at UMDNJ, to determine the crystal structure of the protein, and together the two labs reported a protein structure with unexpected features.

Other proteins known to bind cholesterol contain a large cavity. The interior of the cavity is hydrophobic, which means that it repels water, but it attracts highly insoluble molecules such as cholesterol. The hydrophobic interior of NPC2 lacks a large cavity, or pocket, that could accommodate a large molecule such as cholesterol, but it contains several small cavities. "What was a surprise to us when we determined the structure was that NPC2 lacked an obvious binding pocket," says Stock. "The structure of NPC2 is basically a sandwich of two beta sheets with an intriguing pore leading into the protein's interior. When we inspected the structure, it became obvious that this could be the binding site, but it is too small to accommodate the cholesterol molecule. The NPC2 protein must change its shape when it binds cholesterol."

Confirmation of their hypothesis came at the annual meeting of the Ara Parseghian Medical Research Foundation in 2002, when Scott and M.D.-Ph.D. student Dennis C. Ko reported results of their genetic studies of NPC2, which are now published in PNAS back-to-back with Stock and Lobel's work. (The Parseghian Foundation was established by the legendary Notre Dame football coach, whose three grandchildren were born with NPC. The foundation funds Lobel, Scott, and Stock's work, and its annual meeting brings together dozens of scientists whose work is focused on the disease.)

By studying evolutionarily conserved regions among NPC2 genes in several species, Ko identified and mutated several key amino acids in the protein. He then tested their ability to bind cholesterol and to reverse cholesterol buildup in cells that lack NPC2 function. "What we found were three mutations unable to bind cholesterol and also unable to rescue NPC2-deficient cells," Ko says.

The three amino acids that affect cholesterol binding when mutated are located in the same hydrophobic region of the NPC2 protein that, based on their structural work, Stock and Lobel suggested would bind cholesterol. However, although the convergence of the structural and genetic studies strengthens the case of NPC2 as a cholesterol-binding protein, "exactly what's happening isn't clear," says Stock. "Cholesterol is very insoluble, so it makes some sense that it does not exist in a free state to any significant extent. It may be passed from protein to protein inside the cell. It is logical to hypothesize that cholesterol binding by NPC2 somehow facilitates delivery of cholesterol for export out of the cell. But it is a relatively unexplored field, and there are presently far more questions than answers."

PATCHED AND SMOOTHENED
Answers may come from the study of NPC1's evolutionary relative Patched and its binding partners Hedgehog and Smoothened, a trio of proteins that constitute a pathway for regulating basic body patterning during embryonic development. When Hedgehog, a secreted signaling molecule, binds to Patched on the cell surface, it releases Smoothened to transmit signals to the nucleus for eventual activation of a plethora of genes that direct cell specialization. It turns out that cholesterol is a crucial piece of this pathway.

In 1996, a seminal Science paper from the laboratory of Philip Beachy at Johns Hopkins showed that the Hedgehog signal is formed through an unusual autocatalytic process that involves cleavage of a precursor protein into two parts and the attachment of a cholesterol molecule to the signaling portion of the molecule.

"It was a bolt from the blue," says Beachy. "The specific modification of a protein by cholesterol is truly unique."

Beachy speculates that cholesterol's insolubility may have been advantageous during evolution to restrict the movement of the nascent differentiation signal. "You can't give an accurate accounting of what evolution was doing, but it may have been advantageous at the dawn of multicellularity to restrict Hedgehog's effects to the cell right next door," says Beachy. "Then you could have three cell types: the one that makes it, the one right next door, and the one just a little farther away. Maybe evolution co-opted this autoprocessing domain to add cholesterol and restrict the range of Hedgehog signaling."

Since his discovery of cholesterol's role in development, Beachy too has been drawn into the study of human genetic disease. He became intrigued by a family of human syndromes, the most well-known being Smith-Lemli-Opitz syndrome (SLOS), caused by a defect in the final step of cholesterol synthesis. Lack of cholesterol in these diseases leads to birth defects—such as brain and facial malformations, reduced branching of the lungs, and defects in the development of the nervous system—that resemble those associated with loss of Hedgehog signaling.

"Naturally, we thought what's probably happening in these patients is that there is not enough cholesterol, Hedgehog isn't getting made properly, and so there's a defect in the signaling," says Beachy. However, after using a mouse model of SLOS in a series of experiments, Beachy's team ruled out Hedgehog processing as the culprit. Instead, the scientists discovered that reduced cholesterol levels affect the cell's ability to respond to Hedgehog signals.

In a series of studies using mutant forms of Patched and Smoothened proteins in cholesterol-depleted cells, the scientists narrowed the effect to Smoothened's inability to undergo conformational change from the inactive to active form. In a paper published in the April 2003 issue of Nature Genetics, they showed that the level of cholesterol in the cell membrane in SLOS patients is too low to support the critical transition. "The state of the membrane must be such that Smoothened can make this transition from inactive to active, and cholesterol seems to be a key part of that," says Beachy.

But a mystery remains. The genetic and biochemical studies, says Beachy, suggest a missing piece of the puzzle—a regulator in between Patched and Smoothened—even though no such player is known to act between the two proteins.

In an article in the August 22, 2002, issue of Nature, Beachy and colleagues reported that Patched does not directly suppress Smoothened, as had previously been suggested. Instead, Patched may transport a small molecule, whose identity is not yet known, that affects Smoothened activity. Beachy notes that Patched and NPC1 are both related to a family of bacterial transporters called the resistance-nodulation-division proteins. Another related protein, Dispatched, is responsible for releasing the Hedgehog signal from the cell—perhaps, speculates Beachy, by releasing the cholesterol component that binds Hedgehog to the cell membrane.

"We've been studying how Patched regulates Smoothened, and we think that Patched may transport a small molecule that then regulates Smoothened," says Beachy. "That may be what's in between Patched and Smoothened. But we don't know its identity yet."

What is known is that the Hedgehog signal itself and possibly other pathway components associate with cholesterol-rich domains of the cell membrane, known as rafts, which have been implicated in several cellular processes, including signal transduction and vesicle traffic—the very process that is defective in NPC disease.

"We were all taught 20 years ago that membranes were just this sort of sea of lipids with proteins floating in them," says Beachy. "Well, it's not so. Parts of membranes are much more ordered." In fact, lipid rafts have emerged as a crucial element of many signal-transduction pathways, most prominently in white-blood-cell formation, immune-cell activation, and neuron signaling.

Download this story in Acrobat PDF format.
(requires Acrobat Reader)

Photos: Bill Denison, Robert Cardin, Marc Bryan-Brown

Reprinted from the HHMI Bulletin,
September 2003, pages 10-19.
©2003 Howard Hughes Medical Institute