The paradigm of biological membranes has recently gone through a major update. Instead of being simply fluid and homogeneous, recent studies have shown that membranes are characterized by transient domains with varying fluidity. Of particular interest have been the highly ordered domains known as lipid rafts, which have been suggested to take part in a variety of dynamic cellular processes such as membrane trafficking, signal transduction, and regulation of the activity of membrane proteins. Lipid rafts contain large amounts of cholesterol, which stresses the importance of this key molecule often found in a variety of different cellular processes.

The coupling between lipid rafts and membrane protein functions is a particularly appealing idea, as it highlights how cellular functions are related to and depend on membrane properties. However, despite the importance of rafts, their properties such as the exact lipid composition, characteristic sizes and lifetimes, and even the precise nature of the lipid phases, have remained open issues. This is largely due to the short time and length scales associated with lipid rafts, which renders experimental studies of these complex domains profoundly difficult. The available data proposes, though, that the sizes of lipid rafts range from a few nanometers upwards. This is good news for computational modeling, since it implies that atomistic computer simulations can help to clarify many of the open questions.

Lipid rafts as nano-sized islands

What do we actually know about lipid rafts? Not much. The mere existence of rafts has actually been debated, since there seems to be no unique definition for a raft. What researchers agree on is the fact that cellular membranes are about 5 nm thick interfaces that are highly heterogeneous, characterized by lateral domains whose sizes and compositions fluctuate in time. What is also generally accepted is the existence of domains enriched in cholesterol and some other molecular components such as sphingolipids. These domains are the core of the concept of lipid rafts. In living cells, the sizes of rafts seem to extend from a few to hundreds of nanometers, which implies that—roughly speaking—rafts can be considered as nano-sized cholesterol-containing islands floating in a sea of other lipids.

Over these extremely small scales, inertia is totally irrelevant, meaning that the motion of nano-sized particles is completely governed by viscous forces. In physics this is known as the low Reynolds number limit, but in more concrete terms, one may prefer an example from the macroscopic world: if the nano-sized raft in a cell membrane were instead a macroscopic raft (made out of planks of wood) floating on a sea, the sea around the raft would be maple syrup instead of water. Maple syrup, as we know, is both tasty and very viscous. Obviously, swimming in syrup is rather different from a pool of water (try it), but this is what biological systems such as lipid rafts feel when they move around. As the world seen by individual biomolecules is very different from the one seen by us, it is obvious that the biological relevance of rafts is also related to the functions of nano-scale motors. These motors are membrane proteins many of which may favor being in rafts.

It has been shown that certain proteins function properly only if they are embedded in rafts. If they are transferred to a different membrane environment, they cease to function properly. Obviously, lipid rafts and membranes have a role to play in the activation and functioning of membrane proteins, implying that they are actually dependent of each other.

Proteins in lipid rafts—proteins, lipids, and syrup

We have recently performed extensive atom-scale simulations using the supercomputers at CSC to compare the properties of raft-like membranes with non-raft membrane domains. The studies have shown a number of intriguing properties that are specific to rafts. High concentrations of sphingolipids and cholesterol enhance the lateral packing of lipids and increase the overall ordering of lipids in a membrane, thus giving rise to domains that are particularly tightly packed. Consequently, the rate of molecular motion in the membrane plane is reduced substantially compared with any other fluid membrane type studied before. It is likely that changes in lipid dynamics may in itself contribute to the dynamical partitioning of membrane proteins, as they spend more time in the raft domains due to slower diffusion. Thus, rafts allow more time, e.g., for cross-linking between the proteins to occur.

What is more, the elasticity of raft membranes was discovered to be reduced significantly when compared with non-raft membranes. For example, we observed 5–14 fold differences in the area compressibility and also substantial changes in the bending rigidity when non-raft membranes were replaced with rafts. This is of particular interest, since in experiments it has been observed that the activity of a variety of membrane proteins correlates with the elastic parameters of membranes: protein functionality depends on whether it is located in a flexible or in a more unbending cell membrane environment. The elastic parameters of membranes in turn originate from their lateral pressure profile, which describes the pressure distribution exerted by the membrane on the protein embedded in a membrane.

When we compared rafts with many other membrane types, a careful analysis revealed large differences in the distribution of lateral pressure within the membrane. Consequently, while the simulated systems did not include proteins, one can conclude that all membrane proteins which undergo anisotropic structural changes between functional states are likely to be affected by the lateral pressure profile. As an example, one may consider proteins, which tilt their helices when opening the channel, such as the mechanosensitive channel MscL.We found that the free energy difference between open and closed states of MscL changed from 1.0  to 4 – 11 when single-component bilayers were replaced with rafts.

While these are just numbers, the bottom line is that a change in lipid composition surrounding a protein may contribute as much as 25–50% of the free energy barrier needed to activate or deactivate a membrane protein. This is an appealing example of the dramatic role played by cellular membranes in the functions of proteins. In order to understand how membrane proteins work, one must also understand how membranes are involved in their feelgood play. ■

Perttu Niemelä, Helsinki University of Technlogy
Samuli Ollila, Tampere University of Technology
Marja Hyvönen, Wihuri Research Institute
Mikko Karttunen, University of Western Ontario (Canada)
Ilpo Vattulainen, Tampere University of Technology

Perttu Niemelä is a biophysicist working at Helsinki University of Technology. Samuli Ollila specializes on molecular modeling at Tampere University of Technology. Marja Hyvönen is a senior associate at the Wihuri Research Institute, and Mikko Karttunen is a professor at The University of Western Ontario, Canada. Ilpo Vattulainen is professor of biophysics at Tampere University of Technology and also is a group leader at Helsinki University of Technology and in Helsinki Institute of Physics.

More information

• K. Simons K and E. Ikonen. Functional rafts in cell membranes. Nature 387, 569–572 (1997).
• P. S. Niemelä, S. Ollila, M. T. Hyvönen, M. Karttunen, and I. Vattulainen. Assessing the Nature of Lipid Rafts.
PLoS Computat. Biol. 3, 314–320 (2007).
• J. Aittoniemi, T. Róg, P. Niemelä, M. Pasenkiewicz-Gierula, M. Karttunen, and I. Vattulainen. Tilt: Major Factor in Sterols’ Ordering Capability in Membranes. J. Phys. Chem. B 110, 25562–25564 (2006).
• T. Murtola, T. Rog, E. Falck, M. Karttunen, and I. Vattulainen. Transient Ordered Domains in Single-Component Phospholipid Bilayers. Phys. Rev. Lett. 97, 238102 (2006).
• S. Vainio, M. Jansen, M. Koivusalo, T. Rog, M. Karttunen, I. Vattulainen, and E. Ikonen. Signifi cance of sterol structural specifi city: Desmosterol cannot replace cholesterol in lipid rafts. J. Biol. Chem. 281, 348–355 (2006).