Stereo view of a bacterial ClC channel. Reproduced with permission from Dutzler et al. © 2002 Macmillan Magazines Ltd.

ClC chloride channels are a widespread family of anion-selective channels. Many examples underscore the relevance of ClC channels: mutations in ClC-1 produce myotonia, and loss of ClC-5 causes Dent's disease, a renal pathology in which abnormal amounts of protein appear in the urine. In the nervous system, the absence of ClC-3, which is normally expressed in synaptic vesicles, leads to retinal and hippocampal degeneration in mice.

Like every other channel, ClC channels have been extensively mutated and studied under the electron microscope in attempts to unravel their structure. But unlike several cation channels, neither ClC nor any other anion channel had succumbed to the power of crystallography. Now Dutzler and his colleagues report on their success in solving the crystal structure of two prokaryotic ClC channels. Their data provide us with unprecedented details on the structure of these proteins, and help us to explain some of our previous biophysical findings. The new structure takes our understanding of the chemical and physical basis of anion selectivity to new heights.

ClC channels are thought to be dimeric and to have two pores. Dutzler et al. confirmed this prediction and obtained unequivocal evidence that each pore is formed independently by one monomer. Moreover, the authors found a structural peculiarity within each monomer: the amino- and carboxy-terminal halves are structurally similar, giving rise to an internal repeated pattern that had not been identified in the primary sequence. Importantly, the two halves have opposite orientations in the membrane: they run antiparallel to create a pseudo axis of symmetry. This arrangement allows the amino-acid residues that constitute the chloride-binding site to be in close proximity, despite being in different halves of the molecule. Similar to what had been found in potassium channels, the chloride-binding site in ClC channels is made of partial (not full) charges, providing a favourable electrostatic environment for anions, but preventing them from binding too tightly to the pore.

In addition to locating the chloride-binding site, Dutzler et al. identified another negative charge in the conduction pathway — the side chain of a glutamate residue. The authors speculated that this glutamate must move out of the way for conduction to occur, and reasoned that chloride entry into the pore might elicit this structural rearrangement. This idea makes perfect sense in view of previous observations that chloride is necessary for ClC channel gating.

The X-ray structures of these two bacterial ClC channels represent a real breakthrough in the study of anion channels. They will pave the way for a deeper understanding of anion conduction and selectivity, placing it on a par with our modern insights into the workings of cation channels.