Investigating ion channels using computational methods

Ion channels are of enormous biological importance and are also susceptible to extraordinarily precise measures of function and mechanism, based on to patch clamp technology. The richness of functional data makes ion channels excellent prototype proteins for the development of computational methods for understanding the relationships between macromolecular sequence, structure, and function. In this chapter we outline the overall strategy that is emerging for analyzing and simulating channel structure and function. The essence of the strategy is to describe the channel in a hierarchy of levels of coarse graining, with the appropriate simulation method at each level, and to connect the levels by statistical mechanical theory. We illustrate this strategy with two examples from our laboratory in which we are able to successfully predict the channel permeability directly from the structure; water permeation in gramicidin and potassium permeability in the KcsA ion channel. Remarkably, the detailed simulations of the permeant motions in the channel reveal knock-on, knock-off mechanisms for permeation that resemble simplified models used by channel biophysicists for years in the absence of structural data on the channels. In addition to simulation and statistical mechanics, bioinformatics is also emerging as a useful computational approach for understanding ion channels. The first task in this effort is to develop a sound classification system for channels, including channel-specific rules for inferring similarities and differences in channel sequences. Once this is done, the classification scheme can be used as a context for inferring functional attributes and structural motifs from channel sequence data.