The renal glomerulus of mice lacking s-laminin/laminin beta 2: nephrosis despite molecular compensation by laminin beta 1.

S-laminin/laminin beta 2, a homologue of the widely distributed laminin B1/beta 1 chain, is a major component of adult renal glomerular basement membrane (GBM). Immature GBM bears beta 1, which is replaced by beta 2 as development proceeds. In mutant mice that lack beta 2, the GBM remains rich in beta 1, suggesting that a feedback mechanism normally regulates GBM maturation. The beta 2-deficient GBM is structurally intact and contains normal complements of several collagenous and noncollagenous glycoproteins. However, mutant mice develop massive proteinuria due to failure of the glomerular filtration barrier. These results support the idea that laminin beta chains are functionally distinct although they assemble to form similar structures. Laminin beta 2-deficient mice may provide a model for human congenital or idiopathic nephrotic syndromes.

Myasthenia gravis-like syndrome induced by expression of interferon gamma in the neuromuscular junction.

Abnormal humoral responses toward motor end plate constituents in muscle induce myasthenia gravis (MG). To study the etiology of this disease, and whether it could be induced by host defense molecules, we examined the consequences of interferon (IFN) gamma production within the neuromuscular junction of transgenic mice. The transgenic mice exhibited gradually increasing muscular weakness, flaccid paralysis, and functional disruption of the neuromuscular junction that was reversed after administration of an inhibitor of acetylcholinesterase, features which are strikingly similar to human MG. Furthermore, histological examination revealed infiltration of mononuclear cells and autoantibody deposition at motor end plates. Immunoprecipitation analysis indicated that a previously unidentified 87-kD target antigen was recognized by sera from transgenic mice and also by sera from the majority of human MG patients studied. These results suggest that expression of IFN-gamma at motor end plates provokes an autoimmune humoral response, similar to human MG, thus linking the expression of this factor with development of this disease.

Mutational analysis of muscle nicotinic acetylcholine receptor subunit assembly.

The structural elements required for normal maturation and assembly of the nicotinic acetylcholine receptor alpha subunit were investigated by expression of mutated subunits in transfected fibroblasts. Normally, the wild-type alpha subunit acquires high affinity alpha bungarotoxin binding in a time-dependent manner; however, mutation of the 128 and/or 142 cysteines to either serine or alanine, as well as deletion of the entire 14 amino acids in this region abolished all detectable high affinity binding. Nonglycosylated subunits that had a serine to glycine mutation in the consensus sequence also did not efficiently attain high affinity binding to toxin. In contrast, mutation of the proline at position 136 to glycine or alanine, or a double mutation of the cysteines at position 192 and 193 to serines had no effect on the acquisition of high affinity toxin binding. These data suggest that a disulfide bridge between cysteines 128 and 142 and oligosaccharide addition at asparagine 141 are required for the normal maturation of alpha subunit as assayed by high affinity toxin binding. The unassembled wild-type alpha subunit expressed in fibroblasts is normally degraded with a t1/2 of 2 h; upon assembly with the delta subunit, the degradation rate slows significantly (t1/2 greater than 13 h). All mutated alpha subunits retained the capacity to assemble with a delta subunit coexpressed in fibroblasts; however, mutated alpha subunits that were not glycosylated or did not acquire high affinity toxin binding were rapidly degraded (t1/2 = 20 min to 2 h) regardless of whether or not they assembled with the delta subunit. Assembly and rapid degradation of nonglycosylated acetylcholine receptor (AChR) subunits and subunit complexes were also observed in tunicamycin-treated BC3H-1 cells, a mouse musclelike cell line that normally expresses functional AChR. Hence, rapid degradation may be one form of regulation assuring that only correctly processed and assembled subunits accumulate, and ultimately make functional receptors in AChR-expressing cells.

BIP associates with newly synthesized subunits of the mouse muscle nicotinic receptor.

A slow conformational change in newly synthesized acetylcholine receptor subunits is thought to be a requisite step in the biogenesis of this multi-subunit transmembrane glycoprotein. Previously, we demonstrated that this early conformational change within the alpha-subunit was inefficient and dependent upon disulfide bond formation (Blount, P. and J.P. Merlie. 1990. J. Cell Biol. 111:2613-2622). Here we show that newly synthesized acetylcholine receptor subunits and subunit complexes in the muscle-like cell line, BC3H-1, are associated with Bip, a ubiquitous binding protein of the endoplasmic reticulum. Characterization of the Bip/alpha-subunit complex in stably transfected fibroblasts revealed that Bip associates with newly synthesized unassembled alpha-subunit and some alpha gamma and alpha delta subunit complexes. Significantly, Bip does not associate well with the more mature form of the alpha-subunit containing an intramolecular disulfide bridge. Hence, Bip may play an important role in the conformational maturation and/or editing of unassembled AChR subunits and subunit complexes in vivo.

Assembly intermediates of the mouse muscle nicotinic acetylcholine receptor in stably transfected fibroblasts.

We have used fibroblast clones expressing muscle nicotinic acetylcholine receptor alpha and gamma, and alpha and delta subunits to measure the kinetics of subunit assembly, and to study the properties of the partially assembled products that are formed. We demonstrate by coimmunoprecipitation that assembly intermediates in fibroblasts coexpressing alpha and delta subunits are formed in a time-dependent manner. The alpha and gamma- and the alpha and delta-producing transfected cells form complexes that, when labeled with 125I-alpha-bungarotoxin, migrate in sucrose gradients at 6.3S, a value consistent with a hetero-dimer structure. An additional peak at 8.5S is formed from the alpha and gamma subunits expressed in fibroblasts suggesting that gamma may have more than one binding site for alpha subunit. The stability and specificity of formation of these partially assembled complexes suggests that they are normal intermediates in the assembly of acetylcholine receptor. Comparison of the binding of 125I-alpha-bungarotoxin to intact and detergent-extracted fibroblasts indicate that essentially all of the binding sites are retained in an intracellular pool. The fibroblast delta subunit has the electrophoretic mobility in SDS-PAGE of a precursor that does not contain complex carbohydrates. In addition, alpha gamma and alpha delta complexes had lectin binding properties expected of subunits lacking complex oligosaccharides. Therefore, fibroblasts coexpressing alpha and gamma or alpha and delta subunits produce discrete assembly intermediates that are retained in an intracellular compartment and are not processed by Golgi enzymes.

Neural regulation of muscle acetylcholine receptor epsilon- and alpha-subunit gene promoters in transgenic mice.

The effects of denervation were investigated in mice with transgenes containing promoter elements from the muscle acetylcholine receptor epsilon- and alpha-subunit genes. The promoter sequences were coupled to a nuclear localization signal-beta-galactosidase fusion gene (nlacZ) as a reporter. While many postsynaptic specializations form in the embryo, expression of the epsilon subunit is induced during the first two postnatal weeks. When muscles were denervated at birth, before the onset of epsilon expression, epsilon nlacZ still appeared at the former synaptic sites on schedule. This result suggests that the nerve leaves a localized "trace" in the muscle that can continue to regulate transcription. An additional finding was that epsilon nlacZ expression was much stronger in denervated than in intact muscles. This suggests that the epsilon promoter is similar to the other subunits in containing elements that are activated on cessation of neural activity. However, even after denervation, epsilon nlacZ expression was always confined to the synaptic region whereas alpha nlacZ expression increased in nuclei along the entire length of the fiber. This suggests that while the epsilon gene is similar in its activity dependence to other subunit genes, it is unique in that local nerve-derived signals are essential for its expression. Consequently, inactivity enhances epsilon expression only in synaptic nuclei where such signals are present, but enhances expression throughout the muscle fiber. Truncations and an internal deletion of the epsilon promoter indicate that cis-elements essential for the response to synaptic signals are contained within 280 bp of the transcription start site. In contrast to these results in young animals, denervation in older animals leads to an unexpected reduction in nlacZ activity. However, mRNA measurements indicated that transgene expression was increased in these animals. This discordance between nlacZ mRNA and enzyme activity, demonstrates a previously unknown limitation of nlacZ as a reporter gene in transgenic animals.

The mammalian 43-kD acetylcholine receptor-associated protein (RAPsyn) is expressed in some nonmuscle cells.

Torpedo electric organ and vertebrate neuromuscular junctions contain the receptor-associated protein of the synapse (RAPsyn) (previously referred to as the 43K protein), a nonactin, 43,000-Mr peripheral membrane protein associated with the cytoplasmic face of postsynaptic membranes at areas of high nicotinic acetylcholine receptor (AChR) density. Although not directly demonstrated, several lines of evidence suggest that RAPsyn is involved in the synthesis and/or maintenance of such AChR clusters. Microscopic and biochemical studies had previously indicated that RAPsyn expression is restricted to differentiated, AChR-synthesizing cells. Our recent finding that RAPsyn is also produced in undifferentiated myocytes (Frail, D.E., L.S. Musil, a. Bonanno, and J.P. Merlie, 1989. Neuron. 2:1077-1086) led to to examine whether RAPsyn is synthesized in cell types that never express AChR (i.e., cells of other than skeletal muscle origin). Various primary and established rodent cell lines were metabolically labeled with [35S]methionine, and extracts were immunoprecipitated with a monospecific anti-RAPsyn serum. Analysis of these immunoprecipitates by SDS-PAGE revealed detectable RAPsyn synthesis in some (notably fibroblast and Leydig tumor cell lines and primary cardiac cells) but not all (hepatocyte- and lymphocyte-derived) cell types. These results were further substantiated by peptide mapping studies of RAPsyn immunoprecipitated from different cells and quantitation of RAPsyn-encoding mRNA levels in mouse tissues. RAPsyn synthesized in both muscle and nonmuscle cells was shown to be tightly associated with membranes. These findings demonstrate that RAPsyn is not specific to skeletal muscle-derived cells and imply that it may function in a capacity either in addition to or instead of AChR clustering.

Subtle neuromuscular defects in utrophin-deficient mice.

Utrophin is a large cytoskeletal protein that is homologous to dystrophin, the protein mutated in Duchenne and Becker muscular dystrophy. In skeletal muscle, dystrophin is broadly distributed along the sarcolemma whereas utrophin is concentrated at the neuromuscular junction. This differential localization, along with studies on cultured cells, led to the suggestion that utrophin is required for synaptic differentiation. In addition, utrophin is present in numerous nonmuscle cells, suggesting that it may have a more generalized role in the maintenance of cellular integrity. To test these hypotheses we generated and characterized utrophin-deficient mutant mice. These mutant mice were normal in appearance and behavior and showed no obvious defects in muscle or nonmuscle tissue. Detailed analysis, however, revealed that the density of acetylcholine receptors and the number of junctional folds were reduced at the neuromuscular junctions in utrophin-deficient skeletal muscle. Despite these subtle derangements, the overall structure of the mutant synapse was qualitatively normal, and the specialized characteristics of the dystrophin-associated protein complex were preserved at the mutant neuromuscular junction. These results point to a predominant role for other molecules in the differentiation and maintenance of the postsynaptic membrane.

ARIA/HRG regulates AChR epsilon subunit gene expression at the neuromuscular synapse via activation of phosphatidylinositol 3-kinase and Ras/MAPK pathway.

AChR-inducing activity (ARIA)/heregulin, a ligand for erbB receptor tyrosine kinases (RTKs), is likely to be one nerve-supplied signal that induces expression of acetylcholine receptor (AChR) genes at the developing neuromuscular junction. Since some RTKs act through Ras and phosphatidylinositol 3-kinase (PI3K), we investigated the role of these pathways in ARIA signaling. Expression of activated Ras or Raf mimicked ARIA-induction of AChR epsilon subunit genes in muscle cells; whereas dominant negative Ras or Raf blocked the effect of ARIA. ARIA rapidly activated erk1 and erk2 and inhibition of both erks also abolished the effect of ARIA. ARIA stimulated association of PI3K with erbB3, expression of an activated PI3K led to ARIA-independent AChR epsilon subunit expression, and inhibition of PI3K abolished the action of ARIA. Thus, synaptic induction of AChR genes requires activation of both Ras/MAPK and PI3K signal transduction pathways.

Mutagenesis of the 43-kD postsynaptic protein defines domains involved in plasma membrane targeting and AChR clustering.

The postsynaptic membrane of the neuromuscular junction contains a myristoylated 43-kD protein (43k) that is closely associated with the cytoplasmic face of the nicotinic acetylcholine receptor (AChR)-rich plasma membrane. Previously, we described fibroblast cell lines expressing recombinant AChRs. Transfection of these cell lines with 43k was necessary and sufficient for reorganization of AChR into discrete 43k-rich plasma membrane domains (Phillips, W. D., C. Kopta, P. Blount, P. D. Gardner, J. H. Steinbach, and J. P. Merlie. 1991. Science (Wash. DC). 251:568-570). Here we demonstrate the utility of this expression system for the study of 43k function by site-directed mutagenesis. Substitution of a termination codon for Asp254 produced a truncated (28-kD) protein that associated poorly with the cell membrane. The conversion of Gly2 to Ala2, to preclude NH2-terminal myristoylation, reduced the frequency with which 43k formed plasma membrane domains by threefold, but did not eliminate the aggregation of AChRs at these domains. Since both NH2 and COOH-termini seemed important for association of 43k with the plasma membrane, a deletion mutant was constructed in which the codon Gln15 was fused in-frame to Ile255 to create a 19-kD protein. This mutated protein formed 43k-rich plasma membrane domains at wild-type frequency, but the domains failed to aggregate AChRs, suggesting that the central part of the 43k polypeptide may be involved in AChR aggregation. Our results suggest that membrane association and AChR interactions are separable functions of the 43k molecule.

Molecular forms of N-CAM and its RNA in developing and denervated skeletal muscle.

The neural cell adhesion molecule (N-CAM) is present in both embryonic and perinatal muscle, but its distribution changes as myoblasts form myotubes and axons establish synapses (Covault, J., and J. R. Sanes, 1986, J. Cell Biol., 102:716-730). Levels of N-CAM decline postnatally but increase when adult muscle is denervated or paralyzed (Covault, J., and J. R. Sanes, 1985, Proc. Natl. Acad. Sci. USA., 82:4544-4548). To determine the molecular forms of N-CAM and N-CAM-related RNA during these different periods we used immunoblotting and nucleic acid hybridization techniques to analyze N-CAM and its RNA in developing, cultured, adult, and denervated adult muscle. As muscles develop, the extent of sialylation of muscle N-CAM decreases, and a 140-kD desialo form of N-CAM (generated by neuraminidase treatment) is replaced by a 125-kD form. This change in the apparent molecular weight of desialo N-CAM is paralleled by a change in N-CAM RNA: early embryonic muscles express a 6.7-kb RNA species which hybridizes with N-CAM cDNA, whereas in neonatal muscle this form is largely replaced by 5.2- and 2.9-kb species. Similar transitions in the desialo form of N-CAM, but not in extent of sialylation, accompany differentiation in primary cultures of embryonic muscle and in cultures of the clonal muscle cell lines C2 and BC3H-1. Both in vivo and in vitro, a 140-kD desialo form of N-CAM and a 6.7-kb N-CAM RNA are apparently associated with myoblasts, whereas a 125-kD desialo form and 5.2- and 2.9-kb RNAs are associated with myotubes and myofibers. After denervation of adult muscle, a approximately 12-15-fold increase in the levels of N-CAM is accompanied by a approximately 30-50-fold increase in N-CAM RNA, suggesting that N-CAM expression is regulated at a pretranslational level. Forms of N-CAM and its RNA in denervated muscle are similar to those seen in perinatal myofibers.

Acetylcholine receptor-associated 43K protein contains covalently bound myristate.

Torpedo electroplaque and vertebrate neuromuscular junctions contain high levels of a nonactin, 43,000-Mr peripheral membrane protein referred to as the 43K protein. 43K protein is associated with the cytoplasmic face of postsynaptic membranes at areas of high acetylcholine receptor density and has been implicated in the establishment and/or maintenance of these receptor clusters. Cloning of cDNAs encoding Torpedo 43K protein revealed that its amino terminus contains a consensus sequence sufficient for the covalent attachment of the rare fatty acid myristate. To examine whether 43K protein is, in fact, myristoylated, mouse muscle BC3H1 cells were metabolically labeled with either [35S]cysteine or [3H]myristate and immunoprecipitated with a monospecific antiserum raised against isolated Torpedo 43K protein. In cells incubated with either precursor, a single labeled species was specifically recovered that comigrated on SDS-PAGE with 43K protein purified from Torpedo electric organ. Approximately 95% of the 3H labeled material released from [3H]myristate-43K protein by acid methanolysis was extractable in organic solvents and eluted from a C18 reverse-phase HPLC column exclusively at the position of the methyl myristate internal standard. Thus, 43K protein contains authentic myristic acid rather than an amino or fatty acid metabolite of [3H]myristate. Myristate appears to be added to 43K protein cotranslationally and cannot be released from it by prolonged incubation in SDS, 2-mercaptoethanol, or hydroxylamine (pH 7.0 or 10.0), characteristics consistent with amino terminal myristoylation. Covalently linked myristate may be responsible for the high affinity of purified 43K protein for lipid bilayers despite the absence of a notably hydrophobic amino acid sequence.

Denervation supersensitivity in skeletal muscle: analysis with a cloned cDNA probe.

Motor neurons regulate the acetylcholine sensitivity of the muscles they innervate: denervated muscle fiber become "supersensitive" to acetylcholine, due to insertion of newly synthesized acetylcholine receptors (AChRs) in the plasma membrane. We used hybridization analysis with a cloned cDNA specific for AChR alpha-subunit to compare the abundance of AChR mRNA in innervated and denervated adult mouse muscles. Within 3 d of denervation, levels of AChR mRNA increased 100-fold; levels of actin mRNA changed little. The increase in AChR mRNA level was sufficiently large and rapid to account for denervation supersensitivity.

Fiber type- and position-dependent expression of a myosin light chain-CAT transgene detected with a novel histochemical stain for CAT.

We recently generated and characterized transgenic mice in which regulatory sequences from a myosin light chain gene (MLC1f/3f) are linked to the chloramphenicol acetyltransferase (CAT) gene. Transgene expression in these mice is specific to skeletal muscle and graded along the rostrocaudal axis: adult muscles derived from successively more caudal somites express successively higher levels of CAT. To investigate the cellular basis of these patterns of expression, we developed and used a histochemical stain that allows detection of CAT in individual cells. Our main results are as follows: (a) Within muscles, CAT is detected only in muscle fibers and not in associated connective tissue, blood vessels, or nerves. Thus, the tissue specificity of transgene expression observed by biochemical assay reflects a cell-type specificity demonstrable histochemically. (b) Within individual muscles, CAT levels vary with fiber type. Like the endogenous MLC1f/3f gene, the transgene is expressed at higher levels in fast-twitch (type II) than in slow-twitch (type I) muscle fibers. In addition, CAT levels vary among type II fiber subtypes, in the order IIB greater than IIX greater than IIA. (c) Among muscles that are similar in fiber type composition, the average level of CAT per fiber varies with rostrocaudal position. This position-dependent variation in CAT level is apparent even when fibers of a single type are compared. From these results, we conclude that fiber type and position affect CAT expression independently. We therefore infer the existence of separate fiber type-specific and positionally graded transcriptional regulators that act together to determine levels of transgene expression.

Mapping of myogenin transcription during embryogenesis using transgenes linked to the myogenin control region.

During vertebrate embryogenesis, the muscle-specific helix-loop-helix protein myogenin is expressed in muscle cell precursors in the developing somite myotome and limb bud before muscle fiber formation and is further upregulated during myogenesis. We show that cis-acting DNA sequences within the 5' flanking region of the mouse myogenin gene are sufficient to direct appropriate temporal, spatial, and tissue-specific transcription of myogenin during mouse embryogenesis. Myogenin-lacZ transgenes trace the fate of embryonic cells that activate myogenin transcription and suggest that myogenic precursor cells that migrate from the somite myotome to the limb bud are committed to a myogenic fate in the absence of myogenin transcription. Activation of a myogenin-lacZ transgene can occur in limb bud explants in culture, indicating that signals required for activation of myogenin transcription are intrinsic to the limb bud and independent of other parts of the embryo. These results reveal multiple populations of myogenic precursor cells during development and suggest the existence of regulators other than myogenic helix-loop-helix proteins that maintain cells in the early limb bud in the myogenic lineage.

Clustering and immobilization of acetylcholine receptors by the 43-kD protein: a possible role for dystrophin-related protein.

Recombinant acetylcholine receptors (AChRs) expressed on the surface of cultured fibroblasts become organized into discrete membrane domains when the 43-kD postsynaptic protein (43k) is co-expressed in the same cells (Froehner, S.C., C. W. Luetje, P. B. Scotland, and J. Patrick, 1990. Neuron. 5:403-410; Phillips, W. D., M. C. Kopta, P. Blount, P. D. Gardner, J. H. Steinbach, and J. P. Merlie. 1991. Science (Wash. DC). 251:568-570). Here we show that AChRs present on the fibroblast cell surface prior to transfection of 43k are recruited into 43k-rich membrane domains. Aggregated AChRs show increased resistance to extraction with Triton X-100, suggesting a 43k-dependent linkage to the cytoskeleton. Myotubes of the mouse cell line C2 spontaneously display occasional AChR/43k-rich membrane domains that ranged in diameter up to 15 microns, but expressed many more when 43k was overexpressed following transfection of 43k cDNA. However, the membrane domains induced by recombinant 43k were predominantly small (< or = 2 microns). We were then interested in whether the cytoskeletal component, dystrophin related protein (DRP; Tinsley, J. M., D. J. Blake, A. Roche, U. Fairbrother, J. Riss, B. C. Byth, A. E. Knight, J. Kendrick-Jones, G. K. Suthers, D. R. Love, Y. H. Edwards, and K. E. Davis, 1992. Nature (Lond.). 360:591-593) contributed to the development of AChR clusters. Immunofluorescent anti-DRP staining was present at the earliest stages of AChR clustering at the neuromuscular synapse in mouse embryos and was also concentrated at the large AChR-rich domains on nontransfected C2 myotubes. Surprisingly, anti-DRP staining was concentrated mainly at the large, but not the small AChR clusters on C2 myotubes suggesting that DRP may be principally involved in permitting the growth of AChR clusters.

Separable regulatory elements governing myogenin transcription in mouse embryogenesis.

Expression of the myogenic helix-loop-helix (HLH) protein myogenin in muscle cell precursors within somites and limb buds is among the earliest events associated with myogenic lineage determination in vertebrates. Mutations in the myogenin promoter that abolish binding sites for myogenic HLH proteins or myocyte enhancer factor-2 (MEF-2) suppressed transcription of a linked lacZ transgene in subsets of myogenic precursors in mouse embryos. These results suggest that myogenic HLH proteins and MEF-2 participate in separable regulatory circuits leading to myogenin transcription and provide evidence for positional regulation of myogenic regulators in the embryo.

ACh receptor-rich membrane domains organized in fibroblasts by recombinant 43-kildalton protein.

Neurotransmitter receptors are generally clustered in the postsynaptic membrane. The mechanism of clustering was analyzed with fibroblast cell lines that were stably transfected with the four subunits for fetal (alpha, beta, gamma, delta) or adult (alpha, beta, epsilon, delta) type mouse muscle nicotinic acetylcholine receptors (AChRs). Immunofluorescent staining indicated that AChRs were dispersed on the surface of these cells. When transiently transfected with an expression construct encoding a 43-kilodalton protein that is normally concentrated under the postsynaptic membrane, AChRs expressed in these cells became aggregated in large cell-surface clusters, colocalized with the 43-kilodalton protein. This suggests that 43-kilodalton protein can induce AChR clustering and that cluster induction involves direct contact between AChR and 43-kilodalton protein.

Chromosomal localization of muscle nicotinic acetylcholine receptor genes in the mouse.

The chromosomal localization of the genes encoding the four subunits of muscle nicotinic receptor was determined by analyzing restriction fragment length polymorphisms between two mouse species Mus musculus domesticus (DBA/2) and Mus spretus (SPE). Analysis of the progeny of the interspecies mouse backcross (DBA/2 X SPE) X DBA/2 showed that the alpha-subunit gene cosegregates with the alpha-cardiac actin gene on chromosome 17, that the beta-subunit gene is located on chromosome 11, and that the gamma- and delta-subunit genes cosegregate and are located on chromosome 1.

Molecular basis of the two nonequivalent ligand binding sites of the muscle nicotinic acetylcholine receptor.

We have stably expressed in fibroblasts different pairs of alpha and non-alpha subunits of the mouse muscle nicotinic acetylcholine receptor (AChR). The gamma and delta, but not the beta, subunits associated efficiently with the alpha subunit, and they extensively modified its binding characteristics. The alpha gamma and alpha delta complexes formed distinctly different high affinity binding sites for the competitive antagonist d-tubocurarine that, together, completely accounted for the two nonequivalent antagonist binding sites in native AChR. The alpha delta complex and native AChR had similar affinities for the agonist carbamylcholine. In contrast, although the alpha gamma complex contains the higher affinity competitive antagonist binding site, it had an affinity for carbamylcholine that was an order of magnitude less than that of the alpha delta complex or the AChR. The comparatively low agonist affinity of the alpha gamma complex may represent an allosterically regulated binding site in the native AChR. These data support a model of two nonequivalent binding sites within the AChR and imply that the basis for this nonequivalence is the association of the alpha subunit with the gamma or delta subunit.