The open pore conformation of potassium channels Youxing Jiang, Alice Lee, Jiayun Chen, Martine Cadene, Brian T. Chait & Roderick MacKinnon Howard Hughes Medical Institute, Laboratory of Molecular Neurobiology and Biophysics and Laboratory of Mass Spectrometry and Gaseous Ion Chemistry, Rockefeller University, 1230 York Avenue,New York, New York 10021, USA ........................................................................................................................................................................................................................... Living cells regulate the activity of their ion channels through a process known as gating. To open the pore, protein conformational changes must occur within a channel’s membrane-spanning ion pathway. KcsA and MthK, closed and opened K 1 channels, respectively, reveal how such gating transitions occur. Pore-lining ‘inner’ helices contain a ‘gating hinge’ that bends by approximately 308. In a straight conformation four inner helices form a bundle, closing the pore near its intracellular surface. In a bent configuration the inner helices splay open creating a wide (12 A ˚ ) entryway. Amino-acid sequence conservation suggests a common structural basis for gating in a wide range of K 1 channels, both ligand- and voltage-gated. The open conformation favours high conduction by compressing the membrane field to the selectivity filter, and also permits large organic cations and inactivation peptides to enter the pore from the intracellular solution. Potassium and other ion channels are allosteric proteins that switch between closed and opened conformations in response to an external stimulus in a process known as gating. Depending on the channel type, the gating stimulus can be the binding of a ligand, the membrane electric field, or both. A central issue in ion channel biophysics concerns the nature of the pore conformational changes that accompany channel gating. What do the opened and closed structures of the pore look like? In general, little is known about protein conformational changes in membrane proteins, and yet for ion channels these changes are crucial to every aspect of their function: ion conduction, gating and pharmacology. Here, we address the following questions regarding these three areas of ion channel function. For the conduction mechanism of K þ channels, when the pore opens, how wide does it become, how does opening change the electric field across the pore, and how accessible is the K þ selectivity filter to the intracellular solution? For the gating mech- anism, what are the mechanics of pore opening, are the confor- mational changes within the membrane large or small? Finally, for K þ channel pharmacology, which protein surfaces become exposed when the pore opens, and how might pharmacological agents interact with the closed versus opened state of the channel? Mutational experiments from numerous laboratories have placed important structural constraints on K þ channel gating. In particu- lar, mutational studies of the Shaker voltage-dependent K þ chan- nel 1,2 —interpreted in the context of the KcsA K þ channel structure 3 —identified what is probably the pore’s gate. In the KcsA structure four a-helices (inner helices) line the pore’s intra- cellular half, forming a right-handed bundle (inner helix bundle) near the intracellular opening (Fig. 1) 3,4 . The inner helix bundle marks the point at which the Shaker pore becomes inaccessible to thiol-reactive compounds and metal ions applied to the cytoplasmic side of the channel when the channel is closed 1,5 . Thus, it would seem that the extracellular half of the pore, where the selectivity filter is located, is dedicated to the function of K þ selectivity, and that the intracellular half, where the pore is lined by inner helices, is dedicated to gating. In the KcsA structure the pore is very wide (about 12 A ˚ diameter) at the centre of the membrane in what is called the central cavity, but closer to the intracellular opening, at the level of the inner helix bundle, the pore diameter narrows to about 4.0 A ˚ (the separation between van der Waals surfaces of protein atoms) (Fig. 1). At this narrow segment, hydrophobic side chains from the inner helices line the pore, creating what would seem to be an inhospitable environ- ment for a K þ ion. Is the gate closed in the KcsA crystal structure? Functional measurements offer the best insight into the probable conformational state. In membranes, the KcsA K þ channel has a very low open probability even under the acidic pH conditions known to open the channel 6,7 . When the intracellular carboxy- terminal 35 amino acids are truncated from KcsA, the open probability is even lower, remaining effectively near zero (C. Miller, personal communication). The crystal structures (Fig. 1) are of the truncated KcsA K þ channel 3,4 ; if structure matches function, then the truncated KcsA K þ channel has a closed gate. Exactly how do the inner helices move to open the pore? This has been a difficult question to answer in the absence of direct measurements of closed and opened channel structures. One attempt, using electron paramagnetic resonance (EPR) with spin- labelled KcsA channels, suggested that the inner helices rotate and translate (relative to the KcsA crystal structure), causing a very subtle diameter increase at the inner helix bundle (Protein Data Figure 1 Structural elements of the K þ channel pore. Two subunits of the KcsA K þ channel are shown with the extracellular side on top. The selectivity filter is orange and the central cavity is marked by a red asterisk. Three helical segments include, from N to C terminus, the outer helix (M1), pore helix (P) and inner helix (M2). The gate is formed by the inner helix bundle (Bundle). The figure was prepared using RIBBONS 28 . articles NATURE | VOL 417 | 30 MAY 2002 | www.nature.com 523 © 2002 Nature Publishing Group