They are the gatekeepers of the cell, responsible for the movement of ions across its outer membrane. And 2003 was the year in which the scientist who painstakingly deciphered many of their structures received the ultimate honour for his work — just months after his latest data challenged orthodox views in the field.

Beginning with a landmark paper1 published in 1998, Rod MacKinnon of Rockefeller University in New York has determined the detailed structures of several important ion channels in the cell membrane2,3,4,5,6. Many of his colleagues thought he was embarking on a fool's errand, as proteins embedded in lipid membranes are notoriously difficult to crystallize for X-ray structural analysis. The magnitude of his achievement helps to explain why he was awarded a share of this year's Nobel Prize in Chemistry — a remarkably rapid response from the usually conservative awarding committee.

Indeed, ion channels have not yet yielded all of their secrets, and it was in investigating a remaining mystery that MacKinnon this year ignited a heated debate. Among the most important ion channels, involved in regulating neural impulses, are potassium channels that can be switched on and off according to the voltage across a cell's membrane. In May, MacKinnon's group published the first crystal structure of one of these 'voltage-gated' potassium channels7,8. “It's a major breakthrough,” comments electrophysiologist Chris Miller of Brandeis University in Waltham, Massachusetts, one of MacKinnon's early mentors.

In the absence of a detailed structure, researchers interested in such channels had used electrophysiological and molecular data to develop a mechanism for voltage gating. Voltage-gated potassium channels were known to consist of four identical protein subunits around a central core, and most researchers had assumed that the voltage-sensitive domains would be situated in the heart of each subunit. They were thought to undergo delicate, screw-like movements in response to changes in voltage, allowing potassium ions to rush through.

In contrast, MacKinnon's structure places the voltage sensors on the exterior of each subunit, extending into the cell membrane. According to his suggested mechanism, the sensors open the channel by heaving through the membrane in large, paddle-like strokes — more like weightlifting than the molecular ballet envisaged earlier.

Critics claim that MacKinnon's mechanism is unrealistic. “Previous and continuing experimental studies suggest that the paddle model will not hold up,” says electrophysiologist Richard Horn of Jefferson Medical College in Philadelphia.

But MacKinnon is undaunted. “I think that these new ideas, which are based on very solid structural and functional data, are correct,” he asserts.