Principal Investigator Shin-Ho Chung Project r06

Department of Chemistry, Machine VP

Faculty of Science

Co-Investigators Matthew Hoyles, Serdar Kuyucak,

Department of Theoretical Physics, Research School of Physical Sciences and Engineering

Permeation of Ions through Membrane Channnels

Ion channels are the transistors of the body: they are small groups of proteins spanning the cell membrane that respond to electrical or chemical signals by opening and allowing ions to cross. A cell may have millions of channels, which together allow it to rapidly receive, transmit, and process electrochemical signals. Channels are needed to make nerves and muscles work, including the heart and the brain. A great deal is known about what channels do, but nobody knows much about how they work or what their detailed structure is. Current experimental methods cannot investigate these directly. Molecular biology can change the primary structure of the protein, but the effects on its shape are unknown. X-ray crystallography could reveal the structure of a channel if it could be crystallized, but no one has been able to do this. This project aims to investigate channel structure and mechanism by building computer simulations of channels. The computer simulations will allow us to make definite predictions from our theories about channels, and we can then compare these predictions with experiment and refine the theories.

What are the basic questions addressed?

We aim to investigate the mechanisms of conductance, selectivity, and gating, and how these are related to the structure of channels. The mechanism of conductance is now widely believed to be by diffusion rather than by hopping from binding site to binding site, but questions remain, such as whether the ion is compleatly or partially dehydrated as it passes through the channel. Selectivity is of two types, that between cations and anions, which may be due to long range electrostatic effects, and that between ions of the same charge, which must be due to short range forces. For example, some channels will pass potassium but block sodium, even though the potassium ion is larger. Gating mechanisms are devices used by channels to open and close in response to stimuli. Many gating mechanisms have been proposed but it is not known which mechanisms are used in channels. Some channels open in response to a stimulus and then close shortly afterwards (while the stimulus is still present). Such channels may have two independent gating mechanisms.

What are the results to date and future of the work?

We plan to construct three programs, the first to calculate potential profiles from macroscopic electrostatics with dielectrics, the second to calculate potential profiles and diffusion rates in the constricted region of the channel using molecular dynamics, and the third to calculate conductance for the whole channel using Brownian dynamics, based on data from the first two.

The first program is complete, and has been used to investigate the effect that different shapes and the presence of dipoles have on potential barriers in channels. These results show that large vestibules increase the barrier in a channel, and that dipoles or charges in the channel wall are necessary to reduce the barrier and allow the channel to conduct. The program uses an iterative method to calculate the potential from the dielectric boundary formed by the channel wall. This method is very flexible: it can be used with an arbitrary boundary, provided the boundary is smooth and it does not rely on cylindrical symmetry. The iterative method has been compared with an analytical solution for a prolate spheroid: the results are nearly identical.

What computational techniques are used and why is a supercomputer required?

Our Brownian dynamics program will require a large amount of memory, available only on a supercomputer, as we will be using large look-up tables to calculate the long range forces quickly. Molecular dynamics does not require a large amount of memory, but demands a very large amount of computer power. It could be done either on a supercomputer or a large number of workstations. Our iterative method for finding the electrostatic potential works by dividing the boundary between water and protein into many small sectors. It then calculates the electric field due to fixed charges at each sector, deduces the surface charge at each sector from the electric field and that sector's polerizability, and repeats the process, now including the electric field from other sectors as well as that from fixed charges. The method takes from 10 to 100 iterations to converge, and each iteration takes time proportional to the square of the number of sectors. Our implementation on the VP is highly vectorizable (over 90% vectorization) and uses the large memory of the VP to increase efficiency. An implementation on a workstation would be over 500 times slower, as a less efficient algorithm would be needed to conserve memory.


Energy Barrier Presented to Ions by the Vestibule of the Biological Membrane Channel M. Hoyles, S. Kuyucak, and S. Chung. Biophys. J. 70 accepted