Chloride Ion Conduction Without Water Coordination in the Pore of ClC Protein YOUN JO KO, WON HO JO Department of Materials Science and Engineering, Seoul National University, Seoul 151-742, Korea Received 22 September 2008; Revised 29 January 2009; Accepted 30 April 2009 DOI 10.1002/jcc.21432 Published online 23 June 2009 in Wiley InterScience (www.interscience.wiley.com). Abstract: In the present work, we have found by an atomistic molecular dynamics simulation that hydrogen atoms originating from the residues of a prokaryotic ClC protein (EcClC) stabilize the chloride ion without water mole- cules in the pore of ClC protein. When the chloride ion conduction is simulated by pulling a chloride ion along the pore axis, the free energy barrier for chloride ion conduction is calculated to be low (4 kcal/mol), although the chlo- ride ion is stripped of its hydration shell as it passes through the dehydrated pore region. The calculation of the num- ber of hydrogen atoms surrounding the chloride ion reveals that water molecules hydrating the chloride ion are replaced by polar and non-polar hydrogen atoms protruding from the protein residues. From the analysis of the pair interaction energy between the chloride ion and these hydrogen atoms, it is realized that the hydrogen atoms from the protein residues stabilize the chloride ion at the dehydrated region instead of water molecules, by which the energetic penalty for detaching water molecules from the permeating ion is compensated. q 2009 Wiley Periodicals, Inc. J Comput Chem 31: 603–611, 2010 Key words: ClC channel; membrane protein; ion channel; chloride ion conduction; molecular dynamics simulation Introduction In living cells, ion pumps generate transmembrane ion gradients by consuming a large portion of their metabolic energy. The generated gradients are spent by many ion channels to perform the fundamental process such as transduction of electrical signal, regulation of cell volume, and control of electrolyte transport. For this purpose, ion channels are evolved to be selective to spe- cific ion species. ClC channels are well known for their anion selectivity. Although they conduct other monovalent anions with low efficiency, they are called as ClC channels because the ma- jority of anions in biological system is chloride ion. ClC chan- nels are found in many organisms from bacteria to human, and are involved in various physiological functions. 1,2 Miller 3,4 first discovered the existence of chloride ion channel from the electric organ of Torpedo ray (ClC-0) and proposed that the channel has a double-barrel structure with two independ- ent pores. Later, many workers identified the primary structure of ClC channel as a dimeric structure. 5–10 After this structure identification, a three-dimensional structure of prokaryotic ClC protein has been revealed by X-ray crystallography. 11 Later, a more precise experiment 12 has found that the negatively charged glutamate residue located at the center of the conducting pore is primarily responsible for gating. Motivated by these fascinating results, many theoretical approaches have been made to investi- gate the ion transit pathway and to elucidate the mechanism of the gating phenomenon by using various methods. 13–22 In general, an ion channel has its unique structure to reduce the energetic cost caused by ion penetration through the narrow pore of the channel. For example, a potassium channel, KscA, has a cage-like structure composed of carbonyl oxygen atoms protruding from the backbone of protein, which enables a potas- sium ion to reside stably at the selectivity filter. 23 According to the X-ray crystallographic results of ClC protein, 11,12 a water molecules has not been found at the putative translocation path- way for chloride ion. In this circumstance, it is assumed that a chloride ion is stripped of its hydration shell during its conduc- tion through the dehydrated pore, which inevitably causes a large energy barrier for chloride ion conduction. Recently, nota- ble studies have revealed that several members of ClC family are not channels but H 1 /Cl 2 exchangers in which the flux of chloride ion is coupled to the flux of proton in the opposite direction. 24–26 In addition to these reports, Accardi and his col- leagues proposed that the proton pathway would be separated from the chloride ion pathway. 27 If their proposal is correct, it is reasonable to assume that the pore of ClC protein is dehydrated, although the dehydrated pore is disadvantageous for the conduc- tion of chloride ion, to prevent the formation of a continuous water network through which protons are able to be conducted in an uncontrolled way by the so-called ‘‘Grotthuss’’ or ‘‘hop- and-turn’’ mechanism. 28 Nonetheless, none of the previous Correspondence to: W. H. Jo; e-mail: whjpoly@snu.ac.kr q 2009 Wiley Periodicals, Inc.