Search

Previous studies by several laboratories have identified a narrow sequence region of the nicotinic acetylcholine receptor (AChR) alpha subunit, flanking the cysteinyl residues at positions 192 and 193, as containing major elements of, if not all, the binding site for cholinergic ligands. In the present study, we used a panel of synthetic peptides as representative structural elements of the AChR to investigate whether additional segments of the AChR sequences are able to bind alpha-bungarotoxin (alpha-BTX) and several alpha-BTX-competitive monoclonal antibodies (mAbs). The mAbs used (WF6, WF5, and W2) were raised against native Torpedo AChR, specifically recognize the alpha subunit, and bind to AChR is inhibited by all cholinergic ligands. WF6 competes with agonists, but not with low mol. wt. antagonists, for AChR binding. The synthetic peptides used in this study were approximately 20 residue long, overlapped each other by 4-6 residues, and corresponded to the complete sequence of Torpedo AChR alpha subunit. Also, overlapping peptides, corresponding to the sequence segments of each Torpedo AChR subunit homologous to alpha 166-203, were synthesized. alpha-BTX bound to a peptide containing the sequence alpha 181-200 and also, albeit to a lesser extent, to a peptide containing the sequence alpha 55-74. WF6 bound to alpha 181-200 and to a lesser extent to alpha 55-74 and alpha 134-153. The two other mAbs predominantly bound to alpha 55-74, and to a lesser extent to alpha 181-200. Peptides alpha 181-200 and alpha 55-74 both inhibited binding of 125I-alpha-BTX to native Torpedo AChR. None of the peptides corresponding to sequence segments from other subunits bound alpha-BTX or WF6, or interfered with their binding. Therefore, the cholinergic binding site is not a single narrow sequence region, but rather two or more discontinuous sequence segments within the N-terminal extracellular region of the AChR alpha subunit, folded together in the native structure of the receptor, contribute to form a cholinergic binding region. Such a structural arrangement is similar to the "discontinuous epitopes" observed by X-ray diffraction studies of antibody-antigen complexes [reviewed in Davies et al. (1988)].

Our present view of the membrane topology of nicotinic acetylcholine receptors (nAChR) heavily depends on hydropathy plots (Kyte kath Doolittle, 1982; Noda et al., 1982; Claudio et al., 1983) and epitope mapping (Lindstrom et al., 1984; Maelicke et al., 1984; Neumann et al., 1984; Ratnam kath Lindstrom, 1984; Plumer et al., 1984; Young et al., 1985; Fuchs kath Safran, 1986; Ratnam et al., 1986a,b). The latter is usually performed with antibodies raised against rather long synthetic peptides. We have criticized this approach (Maelicke et al., 1984; Plumer et al., 1984) as it may lead to ambiguous results: Recent structural data have established that the antigen recognition site of antibodies is quite large and that antibody-antigen interaction probably involves multipoint attachment (Amit et al., 1986). Thus, only under limiting conditions will the charge density pattern defining an antibody binding site be directly correlated to the primary structure of this particular region of the antigen. The following results of a topological study of the nAChR from Torpedo electric organ (Maelicke et al., 1989) may exemplify this problem.

Antibodies were raised against eight synthetic peptides matching preselected portions of the amino acid sequence of nicotinic acetylcholine receptor (nAChR) from Torpedo marmorata. To increase the probability of obtaining antibodies specific for the exact sequence of the immunizing peptide, peptides of only five to seven amino acids in length were employed. Even under these limiting conditions some of the polyclonal rabbit immune sera showed cross-reactivity with other peptides and/or other sequence regions of the receptor. Further studies with polyclonal and monoclonal sera suggested that conformation and charge pattern rather than linear sequence are the essential determinants of antibody epitopes. Application of antibodies for topological studies therefore requires that the antibody specificity for a particular region of the antigen has been firmly established. Epitope mapping with the eight anti-peptide immune sera provides information on the accessibility to antibody of matching sequences within the receptor molecule. We find the sequence portions alpha 81-85, alpha 127-132, and alpha 190-195 to be freely accessible both at membrane-bound and at purified receptor. Binding of anti-alpha 387-392 serum does not prove accessibility of this region as the serum cross-reacts strongly with peptide fragments corresponding to the regions alpha 165-200 and beta 190-200 of nAChR from Torpedo californica. To permit binding of anti-alpha 137-142 immune serum, treatment of the receptor with endoglycosidase is required, showing that Asn-141 indeed is glycosylated in native nAChR. The homologous sequence of the other subunits differing only in one sequence position from alpha 137-142 is not accessible in native nAChR to antibody, indicating clear differences in folding of the receptor polypeptides. Sequence portions alpha 395-401 and alpha 161-166 must first be exposed by appropriate treatment to permit binding of respective serum. These results and previous epitope mapping studies by other laboratories are discussed with respect to the limited sequence specificity of antibodies.

Neurotransmitter receptors are key components for neuronal function. A major breakthrough in the study of these molecules has been provided by cloning and sequencing of their genes. By this approach the amino acid sequence of the precursors of several receptors, among them the GABA receptor (Schofield et al., 1987), the glycine receptor (Grenningloh et al., 1987), s adrenergic receptors (Kubo et al., 1986a, Kubo et al., 1986b, Bonner, et al., 1987) and muscarinic receptors (Yarden et al., 1986, Dixon et al., 1986, Kobilka et al., 1987) from different tissues and different nicotinic receptors (McCarthy et al., 1986, Boulter et al., 1986, Hermans-Borgmeyer et al., 1986, Goldman et al., 1987, Baldwin et al, 1988, Deneris et al., 1988) have been obtained. In the case of receptors which function via a chemically gated ion channel, extensive cloning and sequencing of their genes has led to the intriguing discovery that their molecular structure can be very similar, irrespective of their ion specificity. For example, the sequences of the GABA receptor and the glycine receptor, which both contain an anion (Cl-) gating unit, are very similar to the sequence of the nicotinic receptors (AcChR), which contain a cation (Na+) channel (Schofield et al., 1987, Grenningloh et al., 1987). Elucidation of the molecular structure of these receptors is necessary to understand the structural basis for their function and for the ion discrimination operated by these very similar molecules. Regrettably, direct studies of receptor structure by X-ray crystallographic approaches are unsuitable for neurotransmitter receptors, with the possible exception of the nicotinic receptor from Torpedo electric organ, because of their scarcity. New experimental strategies need to be devised to obtain structural information on these molecules, and the available sequence information should be the key towards this endeavour.