Maps of the energy barriers faced by an ion in different parts of the channel. Areas with high energy barriers are colored red, while those with low energy barriers are colored blue. An ion following the lowest energy pathway (traced by the heavy dotted lines) would assume configurations a-b-d-f. Alternatively, ions may follow a secondary low-energy pathway (traced by the thin dashed lines) that follows the configurations a-c-d-f. Click to enlarge
 

Channel cartography

What the calculations revealed was astonishing. The model indicated a progression where potassium ions slide in and out of the binding sites in unison. Just as one ion leaves the channel, another slides into place outside the mouth, and the ions already in the channel roll one valley, or binding site, closer to the exit. Furthermore, Roux says, "Many biophysicists had been relatively uncomfortable having so many ions in a small pore; the energy repelling one another was thought to be enormous...The simulation showed us, yes, the ions can move in this pore, and the repulsion in the pore is actually much more modest than people had expected."

In addition to five interior valleys or sites where the simulation found that ions were most likely to be located, Roux's calculations predicted two as-yet-undetected sites at the mouth of the channel, just outside the cell membrane. The energetic slope at each of these valleys was very gentle and allowed ions to keep moving rather than linger.

Just as Berneche and Roux completed their analyses, MacKinnon and colleagues made available a second X-ray crystallographic image showing the potassium channel at much higher resolution. The snapshot revealed several potassium ions in the act of passing through the channel, data that closely agreed with the positions predicted by Roux and Berneche. The crystallography work and the prediction were published in the same issue Nature in 2001. In an accompanying review, biochemist Christopher Miller of Brandeis University writes of Roux and Berneche's achievement, "This successful prediction in advance of the facts--a rarity in computational biochemistry--enhances the confidence of skeptical experimentalists in methods and parameters used in this theoretical work."

Now that the basis of ion conduction is clearer, the next challenge will be to understand how the potassium channel opens and closes its doors to allow an ion to pass. The potassium channel itself looks like an upside-down teepee made up of four identical molecular subunits. Each subunit makes up part of the door that closes the channel entrance. Recent X-ray crystallographic images, which have caught the door in its open position, show that each subunit must move in order to open the channel. "How does this happen? Do they open one by one? Or at the same time? That will definitely be an interesting thing to look at," Roux says. Once he finds out, with help from NCSA's computing resources, rush hour will never feel as bewildering again.

This research is supported by the National Science Foundation, the National Institutes of Health, the Merck Research Laboratory, and the Keck Foundation.