If you have an ACS member number, please enter it here so we can link this account to your membership. (optional)

ACS values your privacy. By submitting your information, you are gaining access to C&EN and subscribing to our weekly newsletter. We use the information you provide to make your reading experience better, and we will never sell your data to third party members.


Analytical Chemistry

Simulating Life's Envelopes

Models provide clues about lipid behavior in cell membranes, but they may have reached their limits

by Celia Henry Arnaud
February 9, 2009 | A version of this story appeared in Volume 87, Issue 6

Lipids In Flux
Credit: Courtesy of Aurelia Honerkamp-Smith & Sarah Keller
This three-component unilamellar vesicle starts with a uniform distribution of lipids (top left). Slightly above (top right) and below (bottom left) the critical temperature, the lipid composition and domain boundaries fluctuate. Far below the critical temperature (bottom right), phase-separated domains appear.
Credit: Courtesy of Aurelia Honerkamp-Smith & Sarah Keller
This three-component unilamellar vesicle starts with a uniform distribution of lipids (top left). Slightly above (top right) and below (bottom left) the critical temperature, the lipid composition and domain boundaries fluctuate. Far below the critical temperature (bottom right), phase-separated domains appear.

THE PLASMA MEMBRANE, which surrounds biological cells, consists of hundreds—possibly even thousands—of different lipids that are arranged in a bilayer. Membrane proteins embedded in the bilayer occupy a large fraction of the membrane surface, and the cytoskeleton, a latticework of intracellular protein filaments, attaches to the inner side of the membrane. This complexity makes the plasma membrane and its constituent lipids difficult to study directly.

That's where model systems come in. These stripped-down constructions allow scientists to probe the behavior of lipid membranes under carefully controlled circumstances. Scientists have garnered a lot of insight about the membrane from model systems, but these idealized versions of membranes may be reaching the limits of what they can reveal about biology.

Cell biologists and biophysicists use lipid model systems to gain a physical and chemical understanding of the plasma membrane. Such models typically contain three components—an unsaturated lipid, a saturated lipid, and cholesterol. These three components stand in for the multitude of lipids found in the natural cell membrane. Even such simplified mixtures can answer questions about the behavior of the lipid portion of the membrane.

With model systems "you can get at fundamental physical chemical questions," says Barbara A. Baird, a chemistry professor at Cornell University. Such questions include how the lipids organize themselves into multiple liquid phases, called domains, and under what conditions those phases form.

Of particular interest to researchers studying membrane model systems is whether membranes can spontaneously form coexisting liquid phases in the absence of proteins. "You can understand how lipids work and then extrapolate and use that as a model of how they might work in a biological system," Baird says.

Model systems suggest membrane lipids are in a "very peculiar state," says Jay T. Groves, a chemistry professor at the University of California, Berkeley. The largely linear, oily molecules, each with a polar, hydrophilic end, seem to be near a critical point in their phase diagram—a combination of composition, pressure, and temperature at which two coexisting phases become identical.

The remarkable thing is that this behavior occurs in lipid mixtures similar to those in cells. This suggests that cell membranes might hover around a critical point in the lipid phase diagram, where even small changes in conditions can trigger large changes in the membrane. "It's like the cell evolved itself a solvent that doesn't resist all the different things it would need to do," Groves says, referring to the way lipids serve as a versatile solvent conducive to signaling and other interactions on and among cells.

Model systems are valuable because "you can really affirm with no ambiguity whatsoever that lipids and cholesterol have these physical properties as a mixture," Groves says. "Those physical tendencies don't go away when you put this mixture into the membrane of a cell."

One of the key issues that can be addressed with model systems is lipid phase behavior. Using three-component systems, independent groups led by biophysicists Gerald Feigenson at Cornell and Sarah Keller at the University of Washington, Seattle, see lipid mixtures separate into coexisting phases. Seeing such phases in model systems is a first step toward answering the question of whether such domains exist in the intact cell membrane.

In a quest for biologically relevant model systems, Feigenson is moving toward more complicated four-component systems—three lipids plus a "judiciously chosen protein." He thinks such a mixture is the minimum for a model system to approximate a real system.

GIANT PLASMA membrane vesicles, or "blebs," offer an even closer approximation. Blebs, released by cells either naturally or by laboratory inducement, have compositions similar to cell membranes. They are more complicated than synthetic model systems but simpler than intact cell membranes, in part because they lack connection to an underyling cytoskeletal network. The group led by Baird and David Holowka at Cornell is studying the phase behavior of blebs. "I see this kind of work as a bridge between the well-defined model systems and the more complex biological systems," Baird says. Nevertheless, she notes, "it's pretty tricky business trying to relate it to a biological system and to a model system."

With model systems "you can get at fundamental physical chemical questions."

Mass spectrometric analysis shows that the lipid composition of blebs appears similar to the composition of the cell membrane, to the extent that the composition of the cell membrane is actually known, Baird says. Plasma membrane preparations from cells are often contaminated with lipids from the membranes surrounding internal organelles, which have different lipid compositions from the plasma membrane, she says. "The cleanest work on membranes was done on red blood cells because they don't have those internal membranes," Baird notes.

The work of Sarah L. Veatch, a postdoc in Baird's group, suggests that blebs exist in a state near a critical point on a phase diagram (ACS Chem. Biol. 2008, 3, 287). At such points, the system goes through wide fluctuations and easily switches between a single phase and multiple phases. "This kind of fluctuating system can be harnessed to cause a rather dramatic change with the appropriate signal," such as a changing temperature, Baird says.

Veatch studied membrane behavior by fluorescence microscopy. At 20 ºC, micrometer-sized domains form in the membranes. Extrapolating those findings to 37 ºC suggests that nanometer-sized domains should form in biological membranes at physiological temperatures.

"The idea that Sarah Veatch can take blebs and see exactly the same behavior near critical points that we routinely see in purely synthetic vesicles is hugely exciting," Keller says. "It says that even our ridiculously simplified system just might really be biologically relevant."

One of the unanswered questions about cell membrane phase behavior involves the existence of so-called lipid rafts, which are minuscule, patchlike domains that are believed to be involved in protein clustering and cell signaling. These rafts are thought to consist of a more highly ordered cholesterol and sphingolipid-rich liquid phase (the "liquid ordered" phase) interspersed with a less ordered liquid phase (the "liquid disordered" phase). Although liquid-ordered phases have been seen in model systems, the evidence for their existence in natural cell membranes is hard to find. Under physiological conditions, the domains may be too small or too transient to see with conventional tools.

Induced Structure
Credit: Biochemistry © 2008
In a supported lipid bilayer, domains in one layer (green) can induce the formation of domains in a lipid mixture that wouldn't usually form domains (red).
Credit: Biochemistry © 2008
In a supported lipid bilayer, domains in one layer (green) can induce the formation of domains in a lipid mixture that wouldn't usually form domains (red).

Model membrane systems also give researchers a way to study asymmetric bilayers, in which two layers have different compositions, much like the outer and inner "leaflets" of the cell membrane. Asymmetric membranes allow researchers to address whether lipid domains in one leaflet can induce the formation of domains in the other, even if the lipid composition of the second leaflet doesn't usually form domains.

One way to make these asymmetric membranes is with polymer-tethered supported membranes, in which a polymer layer lifts a model membrane from its underlying substrate. The polymer supports make the asymmetric bilayers easier to prepare by giving them added stability, says Lukas K. Tamm of the University of Virginia.

"Using polymer-tethered membranes of asymmetric lipid compositions, we and others have shown that raft-mimicking domains in the bottom layer can induce corresponding domains of equal size on top," says Christoph Naumann, a chemistry professor at Indiana University-Purdue University, Indianapolis.

Keller and Marcus D. Collins have observed similar coupling in unsupported asymmetric planar bilayers. By tuning the lipid composition of only one leaflet, they can induce phase separation in either both leaflets or neither leaflet (Proc. Natl. Acad. Sci. USA 2008, 105, 124).

Credit: Courtesy of Fred Heberle & Gerald Feigenson
Credit: Courtesy of Fred Heberle & Gerald Feigenson

A DANGER of model systems is in taking them too literally, says Michael Edidin, a biologist at Johns Hopkins University. Just one of the problems is that people study model systems under conditions in which domains are on the micrometer scale and thus easy to see. "You just don't see domains that large in cell membranes," he says.

Groves similarly believes that people would be wise to exercise caution in their extrapolations of behavior observed in model systems to cells. The structures seen in model systems provide clues about how lipids might behave in cell membranes, but the correlation will not be exact. Instead, Groves believes the models show that the lipids in membranes exist in a state that allows them to respond quickly to changes in their environment.

Edidin sounds another cautionary note about the energy of membranes. The energy needed for cell membrane processes or functions is much higher than the comparatively weak interactions between the lipid acyl chains, he says, but these higher energies are not available in model systems. "You're talking about ATP being split to do things like endocytosis and exocytosis," which are much more highly energetic processes than the much weaker interactions between lipids in model membranes, such as hydrophobic effects, hydrogen bonding, and van der Waals forces.

Yet another pitfall of models is the possibility of introducing artifacts with measurements. For example, the intense light in fluorescence microscopes can generate free radicals that induce changes in some model membranes, Feigenson says. "It's a problem with chemically simple mixtures, but it might not be such a problem with real cells, because they have reducing systems that will find and destroy light-induced free radicals."

Groves doesn't think that increasing the complexity of model systems will be useful. "Complex model systems can give you a false sense of security that you can take them more seriously" than simple models, he says. "We have no way of knowing that the complex model system resembles the actual cell membrane any more than simpler systems."

In fact, the time has come to move on from model systems, Groves says. "There will always be a role for model systems in studies of cell membranes. However, for my own work, I think we have learned enough from them. The questions we're asking are about the cell membrane and not about soft condensed matter physics. The model systems help us understand the interesting soft condensed matter physics of these systems, but not the cell biology" of the membrane.

That may be, but some researchers will continue working with model systems to further elucidate the role of lipids in cell membranes. "The models have to get more complicated, and the resolution in cells has to get better," Edidin says, conceding however that "there's going to be some upper limit reached where you might as well look at an isolated native membrane."


This article has been sent to the following recipient:

Chemistry matters. Join us to get the news you need.