Membrane fragments from electric tissue of Torpedo californica containing nicotinic acetylcholine receptor are composed of four different polypeptide chains with molecular weights of 40,000 (a), 48,000 (a), 62,000 ('y), and 66,000 (6). The a and a chains are still present in all and y ahd 6 in some of the receptor preparations after Triton X-100 extraction and purification by affinity chromatography. All components of the receptor react covalently with the photoaffinity label 4-azido-2-nitrobenzyltrimethylammonium fluoroborate, the 6 chain incorporating less of the reagent as compared to the a and # chains. Agonists and antagonists containing a quaternary ammonium group protect all chains against the label; the principal neurotoxin from Naja naja siamensis protects the a chain only. We conclude that the a chain binds the neurotoxin from Naja naja, the a and fi chains are involved in the binding of ligands with quaternary ammonium groups, and the function of the y and 6 chains remains to be determined. The quaternary structure of the nicotinic acetylcholine receptor appears to be complex. The protein extracted from the electric tissue of Electrophorus electricus has been shown to be an oligomer (1-5), possibly a pentamer (2) consisting of two different types of polypeptide chains. Two to four different polypeptide chains have been reported for receptor purified from Torpedo californica (6)(7)(8). The function of the different chains is unclear, because only one of them binds the neurotoxin from Naja naja siamensis (9). After reduction with dithiothreitol only the component having a molecular weight of 38,000 binds the affinity label 4-(N-maleimido)benzyltri[3H]methylammonium iodide (6, 10). In this paper experiments with the photoaffinity label 4-azido-2-nitrobenzyltrimethylammonium fluoroborate are reported. This arylazide inactivates membrane-bound erythrocyte acetylcholine esterase and frog satorius muscle acetylcholine receptor after irradiation (11). Since nonspecific reaction with other proteins of the membrane occurred and could not be selectively prevented by appropriate nitrene scavengers, it was concluded that the reaction might not represent a true photoaffinity labeling but an ordinary affinity labeling by a long-lived intermediate generated by photolysis (12). In our experiments reported here with detergent-solubilized and purified receptor the arylazide turned out to be a valuable tool for the investigation of the receptor structure.A central question remains: what is the structure of the acetylcholine receptor in the postsynaptic membrane? Does the molecule extracted by detergent and purified by affinity chromatography represent the true entity or is it a preparative artifact? Finally, what is the structure of the functional postsynaptic membrane? It has been shown by electron microscopy of negative stained and freeze-etched membranes that the postsynaptic membrane contains doughnut-like particles believed to represent receptor molecules (13,14). Receptor-rich membrane fragments can be puri...