Interactions of Small Molecules and Peptides in Membranes

A. Pohorille, M. A. Wilson, C. Chipot, M. H. New and K. Schweighofer, in Computational Molecular Biology, (J. Leszczynski, Ed.), Elsevier, Amsterdam, 1999, pp 485-536.

Although the atomic-level simulation of the interactions of small molecules and proteins with membranes has only become feasible recently, it has already enriched our understanding of this important class of phenomena, largely through an improved understanding of the balance between hydrophobic and electrostatic forces in membranes, and the roles played by membrane fluctuations and ordering. As we have discussed, a coherent picture of the equilibrium distribution and dynamic behavior of small molecules in membranes is emerging. Progress has also been made in understanding unassisted transmembrane ion and proton transport, paving the way to the recently emerging studies on the assisted transport of charges through channels. While studies of membrane proteins are less mature, initial information on the structure of simple membrane-bound proteins and their aggregates has been produced and models of peptide folding at interfaces have been studied. Simulations are also playing an important role in understanding the complex, long-ranged interactions between membranes.

The advances in massively parallel computer hardware and software promise to expand the timescales accessible to atomic-level simulations into the microsecond range. In fact, the first such simulations have started to emerge []. This would allow markedly extend the range of problems in the structural equilibria of protein aggregates that can be reliably studied in simulations. Such studies will provide detailed information on the factors determining the stability of protein aggregates in membranes. Direct observation of the local relaxations, and possibly phase transitions, of lipids near a membrane-integral protein should also become accessible to atomic-level simulations. The added efficiencies of advanced simulation techniques will also facilitate progress in understanding the mechanism of ion transport through simple channels, folding of peptides at the membrane surface, their insertion into the bilayer and the dependence of these processes on the type of the membrane.

Even with these new capabilities many interesting phenomena appear to be beyond the reach of simulation. In particular, the complete folding of a large, membrane protein is not likely to be possible in the foreseeable future. Thus, theoretical progress in understanding the mechanism of action of action of membrane receptors and pumps relies heavily on the availability of high-resolution protein structures supplied from experiment. Obtaining such structures has proven to be extremely difficult but recent progress in this area, manifested by the publication of high resolution X-ray crystal structures of the K channel [] and bacteriorhodopsin [] and in rapid improvements in solid state NMR methodologies, is very encouraging. Starting with a detailed, atomic-level structure, properly designed computer simulations of even large integral proteins or their assemblies may be both possible and profitable. For problems not amenable to detailed mechanistic studies other, more approximate theoretical methods may be useful. Currently, however, they suffer from several, potentially very serious, problems. In particular, they do not faithfully capture heterogeneity of the membrane environment and membrane fluctuations. Here, results of atomic-level computer simulations may guide efforts to improve these methods.

Achieving the full potential of mechanistic studies will require the development of new simulation methodologies. For example, proton transport systems require truly quantum statistical treatments but efficient techniques for the incorporation of quantum effects into large-scale simulations are currently immature. In a similar vein, techniques for the inclusion of polarization effects need further development. Furthermore, improvements in sampling methods may greatly accelerate equilibration of slow degrees of freedom in the membrane and improve mixing in multicomponent systems.

In general, understanding the mechanism of membrane-related functions and the structure of the molecules participating in these processes is taking the central stage in cellular biophysics and structural biology. For theorists it offers both great opportunities and challenges. This review summarized early efforts to take advantage of these opportunities and meet these challenges but, undoubtedly, there is much more to come.