Accessibility to residues in transmembrane segment four of the glycine receptor

Abstract

Glycine receptors (GlyRs) are members of the ligand-gated ion channel superfamily. Each subunit has four transmembrane segments (TM1–TM4). Several studies suggest that amino acids in all four TMs face into a water-filled, alcohol and anesthetic binding cavity in the extracellular portion of the transmembrane domain. TM4 should contribute a “wall” to this cavity, but the residues involved are unknown. Here, we determined the ability of an alcohol analog, propyl methanethiosulfonate (propyl MTS), to covalently react with twelve GlyR TM4 positions (I401–I412) after mutating the original amino acids to cysteines. Reactivity of a cysteine with propyl MTS implies that the cysteine is exposed to water. W407C, I409C, Y410C, and K411C showed altered receptor function following reaction with propyl MTS in the presence or absence of glycine. The cysteine mutations alone eliminated the effects of ethanol for I409C, Y410C, and K411C, and reduced the effects of octanol for I409C and isoflurane for K411C. The ability of propyl MTS to reduce isoflurane and chloroform potentiation was examined in the reactive mutants. Potentiation by isoflurane was significantly reduced for I409C after reaction. These data demonstrate water-accessibility of specific TM4 positions in the GlyR and suggest involvement of these residues with alcohol and anesthetic action.

Introduction

Glycine receptors (GlyRs) are members of the cys-loop superfamily of ligand-gated ion channels. Each subunit of the pentameric receptor has four transmembrane segments (TM1–TM4). GlyRs are targets of alcohols and volatile anesthetics, and their function is potentiated by these drugs. Three amino acids were identified as critical for alcohol and/or volatile anesthetic action on GlyRs (and the homologous GABA_A receptor): I229 (in TM1), S267 (in TM2) and A288 (in TM3) (Greenblatt and Meng, 1999, Jenkins et al., 2001, Mihic et al., 1997, Ueno et al., 2000, Wick et al., 1998, Yamakura et al., 1999, Ye et al., 1998). All three sites are located in the extracellular portion of the transmembrane domain. These three amino acids are believed to face into a water-filled, drug-binding cavity at the center of the transmembrane domain of each receptor subunit. This model is supported by mutagenesis, substituted cysteine accessibility method (SCAM), and electrophysiological studies, as well as biochemical crosslinking between S267C and A288C, and molecular modeling data (Jenkins et al., 2001, Jung et al., 2005, Lobo et al., 2004a, Lobo et al., 2004b, Mascia et al., 2000, Mihic et al., 1997, Wick et al., 1998, Williams and Akabas, 1999, Yamakura et al., 2001). We hypothesize that the drug-binding cavity is bound by amino acids contributed by all four TMs, including TM4.

While amino acids on TM1, TM2, and TM3 have been the subject of study in the GlyR, the orientation of amino acids involved in alcohol and anesthetic action suggest that residues in the extracellular portion of TM4 also contribute to form a wall of the drug-binding cavity. In the recently published 4 Å nicotinic acetylcholine receptor structure, the mostly lipid-facing TM4 forms a left-handed, alpha helical bundle with the other three transmembrane segments, and has probable contacts with TM1 and TM3 (Miyazawa et al., 2003, Unwin, 2005). Experimental evidence from mutagenesis in the GABA_A receptor suggest that positions in TM4 are important for anesthetic action (Jenkins et al., 2002). When 12 positions in the extracellular portion of TM4 were mutated to the bulky amino acid tryptophan in the α1 subunit of the GABA_A receptor, a number of positions were found to either increase or decrease the effects of the volatile anesthetics: isoflurane, halothane, and chloroform (Jenkins et al., 2002). Presently, it is unknown which amino acids in TM4 of the glycine receptor are contributing to the drug-binding pocket. Homology modeling of the GlyR implicated that the amino acids in the extracellular portion of TM4, Y406 to Y410, are the most likely participants in drug binding (Bertaccini et al., 2005).

Here, we tested 12 single cysteine mutants (I401C-I412C) in TM4 of the GlyR α1 subunit for their accessibility to the sulfhydryl-specific reagent propyl methanethiosulfonate (MTS) using the substituted cysteine accessibility method (SCAM; Karlin and Akabas, 1998). These amino acids were selected for mutation because they are on the extracellular side of the transmembrane domain, and we hypothesize that amino acids in this region could be facing into the water-filled, drug-binding pocket. Reaction with MTS reagents requires that the sulfhydryl side chain of the substituted cysteine is ionized, and ionization occurs predominantly in the presence of water (Karlin and Akabas, 1998). Because alcohols and anesthetics are proposed to bind in a water-filled cavity at the center of the transmembrane domain, SCAM results would provide information about the structure of TM4 and indicate possible candidate positions for drug action. As in our previous studies, we used an alkyl MTS reagent (rather than more conventionally used sulfhydryl-specific reagents), because when bound to a cysteine an alkyl thiol is structurally similar to an alcohol or anesthetic molecule bound covalently to the protein (Lobo et al., 2004a, Mascia et al., 2000). Here, we tested for reactivity under conditions where the channel was either closed or open (in the absence or presence of glycine) using two-electrode voltage clamping in Xenopus oocytes.

Covalent reaction with propyl MTS results in altered channel function, and we found that four positions covalently reacted with propyl MTS (W407C, I409C, Y410C, and K411C), indicating that these positions are water-accessible. To our knowledge, this study is the first to determine water-accessibility to residues in TM4 for any ligand-gated ion channel. We tested whether these four mutants were responsive to alcohols (ethanol and octanol) and tested two anesthetics (isoflurane and chloroform) both before and after reaction with propyl MTS. These experiments tested the effect of the single cysteine substitutions on drug modulation and tested whether reaction with propyl MTS could occlude potentiation by anesthetic molecules, one criterion previously proposed for an alcohol/anesthetic binding site (Mascia et al., 2000). Additionally, following reaction of I409C with propyl MTS, potentiation of the glycine response by isoflurane was significantly reduced. These data suggest that there are water-accessible sites in TM4 of the glycine receptor and that these positions are important for alcohol and anesthetic action.