Rings of negatively charged amino acids determine the acetylcholine receptor channel conductance

The structure-function relationship of the nicotinic acetylcholine receptor (AChR) has been effectively studied by the combination of complementary DNA manipulation and single-channel current analysis. Previous work with chimaeras between the Torpedo californica and bovine AChR delta-subunits has shown that the region comprising the hydrophobic segment M2 and its vicinity contains an important determinant of the rate of ion transport through the AChR channel. It has also been suggested that this region is responsible for the reduction in channel conductance caused by divalent cations and that segment M2 contributes to the binding site of noncompetitive antagonists. To identify those amino acid residues that interact with permeating ions, we have introduced various point mutations into the Torpedo AChR subunit cDNAs to alter the net charge of the charged or glutamine residues around the proposed transmembrane segments. The single-channel conductance properties of these AChR mutants expressed in Xenopus laevis oocytes indicate that three clusters of negatively charged and glutamine residues neighbouring segment M2 of the alpha-, beta-, gamma- and delta-subunits, probably forming three anionic rings, are major determinants of the rate of ion transport.     

A cytoplasmic region determines single-channel conductance in 5-HT3 receptors

5-hydroxytryptamine type 3 (5-HT3) receptors are cation-selective transmitter-gated ion channels of the Cys-loop superfamily. The single-channel conductance of human recombinant 5-HT3 receptors assembled as homomers of 5-HT3A subunits, or heteromers of 5-HT3A and 5-HT3B subunits, are markedly different, being 0.4 pS (refs 6, 9) and 16 pS (ref. 7), respectively. Paradoxically, the channel-lining M2 domain of the 5-HT3A subunit would be predicted to promote cation conduction, whereas that of the 5-HT3B subunit would not. Here we describe a determinant of single-channel conductance that can explain these observations. By constructing chimaeric 5-HT3A and 5-HT3B subunits we identified a region (the 'HA-stretch') within the large cytoplasmic loop of the receptor that markedly influences channel conductance. Replacement of three arginine residues unique to the HA-stretch of the 5-HT3A subunit by their 5-HT3B subunit counterparts increased single-channel conductance 28-fold. Significantly, ultrastructural studies of the Torpedo nicotinic acetylcholine receptor indicate that the key residues might frame narrow openings that contribute to the permeation pathway. Our findings solve the conundrum of the anomalously low conductance of homomeric 5-HT3A receptors and indicate an important function for the HA-stretch in Cys-loop transmitter-gated ion channels.     

Structural homology of Torpedo californica acetylcholine receptor subunits

The nicotinic acetylcholine receptor (AChR) from the electroplax of the ray Torpedo californica is composed of five subunits present in a molar stoichiometry of alpha 2 beta gamma delta (refs 1-3) and contains both the binding site for the neurotransmitter and the cation gating unit (reviewed in refs 4-6). We have recently elucidated the complete primary structures of the alpha-, beta- and delta-subunit precursors of the T. californica AChR by cloning and sequencing cDNAs for these polypeptides. Here, we report the whole primary structure of the gamma-subunit precursor of the AChR deduced from the nucleotide sequence of the cloned cDNA. Comparison of the amino acid sequences of the four subunits reveals marked homology among them. The close resemblance among the hydrophilicity profiles and predicted secondary structures of all the subunits suggests that these polypeptides are oriented in a pseudosymmetric fashion across the membrane. Each subunit contains four putative transmembrane segments that may be involved in the ionic channel. The transmembrane topology of the subunit molecules has also been inferred.     

Cloning, sequencing and expression of cDNA for a novel subunit of acetylcholine receptor from calf muscle

The nicotinic acetylcholine receptor (AChR) from fish electric organ has a subunit structure of alpha 2 beta gamma delta, and this is thought to be also the case for the mammalian skeletal muscle AChR. By cloning and sequencing the complementary or genomic DNAs, we have previously elucidated the primary structures of all four subunits of the Torpedo californica electroplax and calf muscle AChR and of the alpha- and gamma-subunits of the human muscle AChR; the primary structures of the gamma-subunit of the T. californica AChR and the alpha-subunit of the Torpedo marmorata AChR have also been deduced elsewhere. We have now cloned DNA complementary to the calf muscle messenger RNA encoding a novel polypeptide (the epsilon-subunit) whose deduced amino-acid sequence has features characteristic of the AChR subunits and which shows higher sequence homology with the gamma-subunit than with the other subunits. cDNA expression studies indicate that the calf epsilon-subunit, as well as the calf gamma-subunit, can replace the Torpedo gamma-subunit to form the functional receptor in combination with the Torpedo alpha-, beta- and delta-subunits.     

Role of acetylcholine receptor subunits in gating of the channel

The Torpedo and calf acetylcholine receptors and hybrids composed of subunits from the two species have been produced in Xenopus oocytes by the use of the cloned complementary DNAs. Single-channel current measurements indicate that these receptors form channels of similar conductance but with different gating behaviour.     

Extracellular domains mediating epsilon subunit interactions of muscle acetylcholine receptor.

Ligand-gated ion channels, a major class of cell-surface proteins, have a pseudosymmetric structure with five highly homologous subunits arranged around a central ion pore. The correct assembly of each channel, whose subunit composition varies with cell type and stage of development, requires specific recognition between the subunits. Assembly of the pentameric form of the acetylcholine receptor from adult muscle (AChR; alpha 2 beta epsilon delta) proceeds by a stepwise pathway starting with the formation of the heterodimers, alpha epsilon and alpha delta. The heterodimers than associate with the beta subunit and with each other to form the complete receptor. We have now determined which parts of the subunits mediate the interactions during assembly of the adult form of the receptor from mouse muscle by using a chimaeric subunit in which the N-terminal and C-terminal extracellular domains are derived from the epsilon subunit with the remainder from the beta subunit. The epsilon and beta subunits were chosen because the epsilon subunit forms a heterodimer with the alpha subunit in the pathway for assembly of the receptor, whereas the beta subunit does not. The epsilon beta chimera can substitute for the epsilon but not the beta subunit in the oligomeric receptor, indicating that the alpha subunit specifically recognizes an extracellular domain of the epsilon subunit.     

Primary structure and functional expression of a developmentally regulated skeletal muscle chloride channel.

Skeletal muscle is unusual in that 70-85% of resting membrane conductance is carried by chloride ions. This conductance is essential for membrane-potential stability, as its block by 9-anthracene-carboxylic acid and other drugs causes myotonia. Fish electric organs are developmentally derived from skeletal muscle, suggesting that mammalian muscle may express a homologue of the Torpedo mamorata electroplax chloride channel. We have now cloned the complementary DNA encoding a rat skeletal muscle chloride channel by homology screening to the Cl- channel from Torpedo. It encodes a 994-amino-acid protein which is about 54% identical to the Torpedo channel and is predominantly expressed in skeletal muscle. Messenger RNA amounts in that tissue increase steeply in the first 3-4 weeks after birth, in parallel with the increase in muscle Cl- conductance. Expression from cRNA in Xenopus oocytes leads to 9-anthracene-carboxylic acid-sensitive currents with time and voltage dependence typical for macroscopic muscle Cl- conductance. This and the functional destruction of this channel in mouse myotonia suggests that we have cloned the major skeletal muscle chloride channel.     

Repeat I of the dihydropyridine receptor is critical in determining calcium channel activation kinetics.

Membrane depolarization causes many kinds of ion channels to open, a process termed activation. For both Na+ channels and Ca2+ channels, kinetic analysis of current has suggested that during activation the channel undergoes several conformational changes before reaching the open state. Structurally, these channels share a common motif: the central element is a large polypeptide with four repeating units of homology (repeats I-IV), each containing a voltage-sensing region, the S4 segment. This suggests that the distinct conformational transitions inferred from kinetic analysis may be equated with conformational changes of the individual structural repeats. To investigate the molecular basis of channel activation, we constructed complementary DNAs encoding chimaeric Ca2+ channels in which one or more of the four repeats of the skeletal muscle dihydropyridine receptor are replaced by the corresponding repeats derived from the cardiac dihydropyridine receptor. We report here that repeat I determines whether the chimaeric Ca2+ channel shows slow (skeletal muscle-like) or rapid (cardiac-like) activation.     

The heteromeric cyclic nucleotide-gated channel adopts a 3A:1B stoichiometry

Cyclic nucleotide-gated (CNG) channels are crucial for visual and olfactory transductions. These channels are tetramers and in their native forms are composed of A and B subunits, with a stoichiometry thought to be 2A:2B (refs 6, 7). Here we report the identification of a leucine-zipper-homology domain named CLZ (for carboxy-terminal leucine zipper). This domain is present in the distal C terminus of CNG channel A subunits but is absent from B subunits, and mediates an inter-subunit interaction. With cross-linking, non-denaturing gel electrophoresis and analytical centrifugation, this CLZ domain was found to mediate a trimeric interaction. In addition, a mutant cone CNG channel A subunit with its CLZ domain replaced by a generic trimeric leucine zipper produced channels that behaved much like the wild type, but less so if replaced by a dimeric or tetrameric leucine zipper. This A-subunit-only, trimeric interaction suggests that heteromeric CNG channels actually adopt a 3A:1B stoichiometry. Biochemical analysis of the purified bovine rod CNG channel confirmed this conclusion. This revised stoichiometry provides a new foundation for understanding the structure and function of the CNG channel family.     

T-type calcium channel regulation by specific G-protein betagamma subunits

Low-voltage-activated (LVA) T-type calcium channels have a wide tissue distribution and have well-documented roles in the control of action potential burst generation and hormone secretion. In neurons of the central nervous system and secretory cells of the adrenal and pituitary, LVA channels are inhibited by activation of G-protein-coupled receptors that generate membrane-delimited signals, yet these signals have not been identified. Here we show that the inhibition of alpha1H (Ca(v)3.2), but not alpha(1G) (Ca(v)3.1) LVA Ca2+ channels is mediated selectively by beta2gamma2 subunits that bind to the intracellular loop connecting channel transmembrane domains II and III. This region of the alpha1H channel is crucial for inhibition, because its replacement abrogates inhibition and its transfer to non-modulated alpha1G channels confers beta2gamma2-dependent inhibition. betagamma reduces channel activity independent of voltage, a mechanism distinct from the established betagamma-dependent inhibition of non-L-type high-voltage-activated channels of the Ca(v)2 family. These studies identify the alpha1H channel as a new effector for G-protein betagamma subunits, and highlight the selective signalling roles available for particular betagamma combinations.     

Channel properties of an insect neuronal acetylcholine receptor protein reconstituted in planar lipid bilayers

A pentameric membrane protein composed of four types of polypeptide has been identified as the minimal structural unit responsible for the electrogenic action of acetylcholine on electrocytes and muscle cells. Because many populations of central and peripheral neurons also have nicotinic acetylcholine receptors (AChRs), considerable effort has recently gone into identifying the neuronal receptor. The central nervous tissue of insects contains very high concentrations of nicotinic AChRs, and we have recently purified an alpha-toxin binding protein, a putative AChR, from neuronal membranes of locusts. It is a component of high relative molecular mass, clearly composed of identical subunits, a structure predicted for an ancestral AChR protein. To verify that the purified polypeptides not only represent ligand binding sites but that they are indeed functional receptors, we have now reconstituted the isolated protein in a planar lipid bilayer. We show that in this system cholinergic agonists activate functional ion channels, that have properties comparable to those exhibited by the peripheral AChRs in vertebrates; thus, for the first time a functional acetylcholine receptor channel has been identified in nerve cells.     

Quantification of Ca2+-activated K+ channels under hormonal control in pig pancreas acinar cells

Ca2+- and voltage-activated K+ channels are found in many electrically excitable cells and have an important role in regulating electrical activity. Recently, the large K+ channel has been found in the baso-lateral plasma membranes of salivary gland acinar cells, where it may be important in the regulation of salt transport. Using patch-clamp methods to record single-channel currents from excised fragments of baso-lateral acinar cell membranes in combination with current recordings from isolated single acinar cells and two- and three-cell clusters, we have now for the first time characterized the K+ channels quantitatively. In pig pancreatic acini there are 25-60 K+ channels per cell with a maximal single channel conductance of about 200 pS. We have quantified the relationship between internal ionized Ca2+ concentration [( Ca2+]i) membrane potential and open-state probability (p) of the K+ channel. By comparing curves obtained from excised patches relating membrane potential to p, at different levels of [Ca2+]i, with similar curves obtained from intact cells, [Ca2+]i in resting acinar cells was found to be between 10(-8) and 10(-7) M. In microelectrode experiments acetylcholine (ACh), gastrin-cholecystokinin (CCK) as well as bombesin peptides evoked Ca2+-dependent opening of the K+ conductance pathway, resulting in membrane hyperpolarization. The large K+ channel, which is under strict dual control by internal Ca2+ and voltage, may provide a crucial link between hormone-evoked increase in internal Ca2+ concentration and the resulting NaCl-rich fluid secretion.     

A large anion-selective channel has seven conductance levels

Ion channels have generally been found to have two predominant conductance levels thought to be associated with 'open' and 'closed' states, but intermediate (subconductance) states have also been reported. We have now found that a large conductance, anion-selective channel in pulmonary alveolar epithelial cells can adopt any of six open levels of conductance that are integer multiples of 60-70 pS. The channel is usually either fully open or fully closed. The frequencies of the different conductance levels are inconsistent with the notion that there are six independent channels. We suggest that the channel consists of six conducting pathways in parallel, 'co-channels', with a shared gating mechanism that can synchronously render all of them non-conducting. Other channels with lower maximum conductance may operate in a similar way but multiple conductance levels would not easily be detected because of a less favourable signal-to-noise ratio.     

A chloride channel widely expressed in epithelial and non-epithelial cells.

Chloride channels have several functions, including the regulation of cell volume, stabilizing membrane potential, signal transduction and transepithelial transport. The plasma membrane Cl- channels already cloned belong to different structural classes: ligand-gated channels, voltage-gated channels, and possibly transporters of the ATP-binding-cassette type (if the cystic fibrosis transmembrane regulator is a Cl- channel). The importance of chloride channels is illustrated by the phenotypes that can result from their malfunction: cystic fibrosis, in which transepithelial transport is impaired, and myotonia, in which ClC-1, the principal skeletal muscle Cl- channel, is defective. Here we report the properties of ClC-2, a new member of the voltage-gated Cl- channel family. Its sequence is approximately 50% identical to either the Torpedo electroplax Cl- channel, ClC-0 (ref. 8), or the rat muscle Cl- channel, ClC-1 (ref. 9). Isolated initially from rat heart and brain, it is also expressed in pancreas, lung and liver, for example, and in pure cell lines of fibroblastic, neuronal, and epithelial origin, including tissues and cells affected by cystic fibrosis. Expression in Xenopus oocytes induces Cl- currents that activate slowly upon hyperpolarization and display a linear instantaneous current-voltage relationship. The conductivity sequence is Cl- greater than or equal to Br- greater than I-. The presence of ClC-2 in such different cell types contrasts with the highly specialized expression of ClC-1 (ref. 9) and also with the cloned cation channels, and suggests that its function is important for most cells.     

Localized Ca2+ and calcium-activated potassium conductances in terminals of a barnacle photoreceptor

Calcium channels are found in the presynaptic terminals of neurones, where they have a key role in synaptic transmission. They are also found in the somata of many cells, in dendrites and along a few axons. In no cell is the actual distribution of these channels known in detail, because there are no known toxins or other agents suitable for labelling calcium channels, and the current through these channels is usually too small to be quantified with extracellular electrodes. However, several experiments have suggested that the density of the channels is less in the axon than in the cell body or terminal region. Here we have used the indicator dye Arsenazo III in conjunction with an array of photodetectors to examine the spatial influx of calcium in the presynaptic terminal region of the giant barnacle, Balanus nubilus. In these cells, calcium entry occurs in a restricted region less than 50 micron in length, which corresponds closely to the region of synaptic contact with second-order cells. Outside this area the magnitude of calcium entry is reduced at least 50-fold. With reasonable assumptions it follows that the calcium channel density is equally localized. In addition, we demonstrate that these cells have a calcium-activated potassium conductance. Since calcium entry is restricted to the synaptic zone, this conductance must be effective only in this region.     

Monoclonal antibodies to the acetylcholine receptor by a normally functioning auto-anti-idiotypic mechanism

Recently we described a procedure for preparing antibodies to the acetylcholine receptor (AChR) based on immunoglobulin idiotypes and on the hypothesis that, regardless of functional differences, macromolecules of the same specificity will show structural homologies in their binding sites. Antibodies were prepared in rabbits to a structurally constrained agonist of AChR, trans-3,3'-bis[alpha-(trimethylammonio)methyl]azobenzene bromide (BisQ). These antibodies mimicked the binding specificity of AChR in its activated state--agonists were bound with affinities that were in accord with their biological activities and antagonists were bound poorly. Rabbits were then immunized with a specifically purified preparation of anti-BisQ to elicit a population of antibodies specific for the binding sites of anti-BisQ. A portion of the anti-idiotypic antibodies produced in the second set of rabbits cross-reacted with determinants on AChR preparations from Torpedo californica, Electrophorus electricus and rat muscle. Moreover, several of the rabbits showed signs of experimental myasthenia gravis, in which circulating AChR antibodies are typically found. To devise a more direct route to monoclonal anti-receptor antibodies we based our strategy on acceptance of the concept of the anti-idiotypic network theory of Jerne. According to this theory, injection of an antigen elicits, in addition to antibodies to the antigen, other populations that include anti-idiotypic antibodies directed at the combining sites of the antigen-specific antibodies. If the antigen-specific antibodies recognize a ligand of a receptor, then the anti-idiotypic antibodies should bind receptor. Thus, when a mouse is immunized with a bovine serum albumin conjugate of BisQ (BisQ-BSA), it should be possible to expand populations of spleen cells that secrete antibodies which bind anti-BisQ and AChR, in addition to populations specific for BisQ. Fusion of the spleen cells with an appropriate myeloma line should yield monoclonal anti-AChR antibodies. Here we report the success of this approach and its implications.     

Conductances of single ion channels opened by nicotinic agonists are indistinguishable

Hypotheses concerning the mechanism by which acetylcholine-like agonists cause ion channels to open often suppose that the receptor-ionophore complex can exist in either of two discrete conformations, open and shut. On the basis of noise analysis it has been reported that certain agonists open ion channels of lower conductance than usual, though many potent agonists give similar conductances, and hence that differences in the conductance of ion channels opened by different agonists may contribute to differences in efficacy. Here we have reinvestigated this question by recording single ion channel currents evoked by acetylcholine-like agonists on embryonic rat muscle in tissue culture and on adult frog muscle endplate. Ten different agonists (Fig. 1) were tested, including several that noise analysis has suggested have a low conductance. The single-channel conductance was found to be the same, within a few per cent, for all 10 agonists. It seems that noise analysis has given erroneously low conductances in some cases. Therefore efficacy differences do not depend on differences in single-channel conductance evoked by various agonists but presumably on the position of the open-shunt equilibrium of the agonist-channel complexes.     

The protein kinase A-regulated cardiac Cl- channel resembles the cystic fibrosis transmembrane conductance regulator.

Stimulation of beta-adrenoceptors in cardiac ventricular myocytes activates a strong chloride ion conductance as a result of phosphorylation by cyclic AMP-dependent protein kinase (PKA). This Cl- conductance, which is time- and voltage-independent, counters the tendency of the simultaneously enhanced Ca2+ channel current to prolong the ventricular action potential. Using inside-out giant patches excised from guinea-pig myocytes, we show here that phosphorylation by the PKA catalytic subunit plus Mg-ATP elicits discrete Cl- channel currents. In almost symmetrical Cl- solutions (approximately 150 mM), unitary current amplitude scales with membrane potential, and reverses sign near 0 mV, to yield a single channel conductance of approximately 12 pS. Opening of the phosphorylated channels requires hydrolysable nucleoside triphosphate, indicating that phosphorylation by PKA is necessary, but not sufficient, for channel activation. The properties of these PKA-regulated cardiac Cl- channels are very similar, if not identical, to those of the cystic fibrosis transmembrane conductance regulator (CFTR), the epithelial cell Cl- channel whose regulation is defective in patients with cystic fibrosis. The full cardiological impact of these Cl- channels and of their possible malfunction in patients with cystic fibrosis remains to be determined.     

Crystal structure and mechanism of a calcium-gated potassium channel

Ion channels exhibit two essential biophysical properties; that is, selective ion conduction, and the ability to gate-open in response to an appropriate stimulus. Two general categories of ion channel gating are defined by the initiating stimulus: ligand binding (neurotransmitter- or second-messenger-gated channels) or membrane voltage (voltage-gated channels). Here we present the structural basis of ligand gating in a K(+) channel that opens in response to intracellular Ca(2+). We have cloned, expressed, analysed electrical properties, and determined the crystal structure of a K(+) channel (MthK) from Methanobacterium thermoautotrophicum in the Ca(2+)-bound, opened state. Eight RCK domains (regulators of K(+) conductance) form a gating ring at the intracellular membrane surface. The gating ring uses the free energy of Ca(2+) binding in a simple manner to perform mechanical work to open the pore.