Weill Cornell Scientists Help Elucidate How Membrane Channels Conduct Potassium Ions

Theoretical Computer Simulations Remarkably Predict Actual Findings of X-Ray Crystallography

Nov 28, 2001


One of the marvels of life is how the nervous systems of organisms can operate so rapidly and precisely, governing motion and consciousness. Nerve cells, like other cells, have membranes spanned by potassium-conducting channel proteins, but exactly how do these channel proteins work? Now, several papers in the November 1 issue of Nature, including one by BenoΒt Roux and Simon Bernèche of Weill Cornell Medical College, answer this question in dazzling detail, greatly deepening our understanding of some of the basic phenomena of the living world.

Dr. Roux is a Professor of Biochemistry and a Professor of Physiology and Biophysics, and Mr. Bernèche is his graduate student. They achieved their results by performing computer simulations (known as "molecular dynamics") of the potassium channel protein—showing not only in theory but, as it turns out, with uncanny accuracy, the positions and activity of the ions in and around the channel.

Two other related papers in Nature were contributed by a group at The Rockefeller University, led by Dr. Roderick MacKinnon, and Dr. Christopher Miller of Brandeis University provided a commentary to explain the significance of all the papers.

Dr. MacKinnon's group used X-ray crystallography to analyze and describe the potassium channel at the atomic level, showing, among other things, that potassium ions occupy positions just outside the extracellular end of the channel. One of these is coordinated in front by four protein carbonyl groups reaching out from the channel, and behind by solvent. As Dr. Miller puts it, "Remarkably . . . Bernèche and Roux anticipated that a K+ [potassium] ion would be localized in solvent, right at the two sites where it is actually seen in the X-ray work." In other words, the Weill Cornell scientists' theoretical work with molecular dynamics was borne out by the Rockefeller University scientists' direct observations with X-ray crystallography.

Dr. Miller comments: "This successful prediction in advance of the facts—a rarity in computational biochemistry—enhances the confidence of skeptical experimentalists in the methods and parameters used in this theoretical work."

Dr. Miller has remarked to a journalist that, "Roux had been talking about the existence of these sites more than a year ago, and nobody believed it. Now, MacKinnon has seen it."

He goes on: "This tends to fortify an additional conclusion emerging from the work—that not only are the K+-binding configurations energetically similar, but so are the transitions between them. This means that the entire conduction process is energetically barrierless. So, overall, K+ permeation looks rather like the concerted action of those steel pendulum toys known as Newton's balls. An ion entering the pore on the left expels an ion on the right into solution, while the four ions within and just outside the pore all shift one position to the right . . . ; if this occurs on energetically level ground, the process will be rapid."

Dr. BenoΒt Roux

As Dr. Roux observes of his research in general, "The objective of my research is to investigate the structure, dynamics, and function of complex macromolecular biological systems using computational methods. My efforts are aimed at understanding the fundamental processes involving the cell membrane, such as permeation, excitability, signaling, and fusion."

Among the specific questions that Dr. Roux's laboratory of molecular dynamics focuses on are: What are the microscopic mechanisms by which channels achieve a high conductivity and remain selective? What is the molecular basis for the specificity and gating of an ion channel? Which specific amino acids are responsible for the selectivity to potassium, sodium, chloride, and calcium ions? What is the molecular basis for the voltage gating of an ion channel? How can one use limited information from experiments to deduce the structure of membrane proteins? His latest paper in Nature provides answers to several of these questions.

Dr. Roux, who obtained his Ph.D. in biophysics from Harvard, joined the Weill Cornell faculty in 1999 as a structural biology recruit under the College's Strategic Plan for Research. Through this Plan, the institution is supporting young researchers who, it is hoped, will lay the foundations of fundamental breakthroughs in the biomedical sciences. Structural biology, Dr. Roux explains, is a field that hypothesizes that the activity and function of biological molecules can be understood from the fundamental laws of chemistry and physics once their three-dimensional structure is determined to atomic resolution. But, unlike many in the field, who are experimentalists—crystallographers, nuclear magnetic resonance spectroscopists, and so on—Dr. Roux is a computational biophysicist, who attempts to understand how proteins work by using theoretical considerations based on microscopic energies, atomic motions, fluctuations, and other such manifestations of the laws of molecular physics and chemistry.