Cysteine Modification 35
ligand-gated ion channels (pLGIC). After the cloning of the nicotinic acetylcholine
receptor subunits (nACh) (Noda et al. 1982 , 1983a, b; Numa et al. 1983 ) and the
subsequent cloning of the genes encoding GABAA, glycine and 5-HT 3 receptor sub-
units (Grenningloh et al. 1987 ; Schofield et al. 1987 ; Maricq et al. 1991 ), it became
clear that they were all part of a single gene superfamily. The cloning of the genes
encoding the various subunits, the development of site-directed mutagenesis and
the ability to heterologously express both native and mutant channels set the stage
for structure-function studies. Evidence from several labs suggested that the sec-
ond membrane-spanning segment (M2) lined the nACh receptor channel (Giraudat
et al. 1986 ; Hucho et al. 1986 ; Imoto et al. 1986 , 1988 ; Leonard et al. 1988 ; Revah
et al. 1990 ), but no systematic method was available to identify all of the channel-
lining residues. The substituted cysteine accessibility method (SCAM) provided a
systematic approach to identify the channel-lining residues (Akabas et al. 1992 ; Xu
and Akabas 1993 ; Akabas et al. 1994a, b; Xu and Akabas 1996 ; Karlin and Akabas
1998 ). SCAM avoided an issue that had plagued site-directed mutagenesis struc-
ture-function studies; that was, how to determine whether the effects of a mutation
on channel function were local, at the site of the mutation, or due to more global,
albeit subtle, changes in protein structure. Although the effects of each cysteine
mutant on the apparent affinity for agonist were assessed, the major experimental
question was whether the charged, water soluble MTS reagents could react with the
engineered cysteine. Because MTS reagents react a billion-fold faster with the ion-
ized thiolate relative to the un-ionized thiol (Roberts et al. 1986 ; Karlin and Akabas
1998 ), we inferred that MTS reactive residues must be, at least transiently, on the
water-accessible surface, because only residues that were, at least transiently, on the
water-accessible surface would ionize.
Identification of M2 Segment Channel-Lining Residues The initial SCAM
study of the nACh α and GABAA α1 subunit M2 segments showed remarkable
concordance in terms of the reactive residues identified in the presence of ago-
nist (Fig. 3 ) (Akabas et al. 1994a; Xu and Akabas 1996 ). For ease of comparing
results in Cys-loop family subunits an index numbering system was introduced
for the M2 segment residues; 0’ is the conserved positively charged residue near
the N-terminal end of M2, the index numbers become positive towards the C-ter-
minus and negative towards the N-terminus (Miller 1989 ). A subsequent SCAM
study of the 5-HT 3 A receptor identified a similar set of residues (Fig. 3 ) (Reeves
et al. 2001 ). It is striking that in three separate channels, using three different
sulfhydryl reagents, MTSEA (nACh), pCMBS (GABA) and MTSET (5-HT 3 A), a
similar subset of M2 segment residues were identified in the presence of agonist.
It is important to recognize that in the presence of agonist on the time scale of
a typical SCAM experiment, tens of seconds to minutes, the channels undergo
transitions between open and desensitized states. In general, it is not possible to
determine whether reaction occurred in the open or desensitized state. Previous
studies of the nACh receptor had used photoaffinity labeling with channel blockers
and peptide sequencing to identify channel-lining residues (Giraudat et al. 1986 ,
1987 ; Hucho et al. 1986 ; Oberthur et al. 1986 ; Revah et al. 1990 ). The residues