MuSK induces in vivo acetylcholine receptor clusters in a ligand-independent manner.

Muscle-specific receptor tyrosine kinase (MuSK) is required for the formation of the neuromuscular junction. Using direct gene transfer into single fibers, MuSK was expressed extrasynaptically in innervated rat muscle in vivo to identify its contribution to synapse formation. Spontaneous MuSK kinase activity leads, in the absence of its putative ligand neural agrin, to the appearance of epsilon-subunit-specific transcripts, the formation of acetylcholine receptor clusters, and acetylcholinesterase aggregates. Expression of kinase-inactive MuSK did not result in the formation of acetylcholine receptor (AChR) clusters, whereas a mutant MuSK lacking the ectodomain did induce AChR clusters. The contribution of endogenous MuSK was excluded by using genetically altered mice, where the kinase domain of the MuSK gene was flanked by loxP sequences and could be deleted upon expression of Cre recombinase. This allowed the conditional inactivation of endogenous MuSK in single muscle fibers and prevented the induction of ectopic AChR clusters. Thus, the kinase activity of MuSK initiates signals that are sufficient to induce the formation of AChR clusters. This process does not require additional determinants located in the ectodomain.

Assembly and clustering of acetylcholine receptors containing GFP-tagged epsilon or gamma subunits: selective targeting to the neuromuscular junction in vivo.

Acetylcholine receptor (AChR) gamma and epsilon subunits were tagged by green fluorescent protein (GFP) to analyse assembly and targeting in live muscle fibers at the neuromuscular junction. N- or C-terminal fusion polypeptides showed no fluorescence upon transfection of HEK cells. When GFP was inserted into the cytoplasmic loop connecting putative transmembrane regions M3 and M4, the gamma/GFP and epsilon/GFP subunits were fluorescent and formed together with the alpha, beta, and delta subunits GFP-tagged AChR complexes that were integrated into the plasma membrane. As the AChR were also clustered by rapsyn, the results indicate that the cytoplasmatic domains of the gamma and epsilon subunits may not be required for assembly and rapsyn-dependent clustering. The gamma/GFP and epsilon/GFP subunit-containing receptors were expressed in X. laevis oocytes and have affinities for acetylcholine similar to that of the wild-type receptors. Direct gene transfer into single muscle fibers reveals that gamma/GFP or epsilon/GFP polypeptides are expressed at the site of injection and are transported within the endoplasmatic reticulum. When reaching subsynaptic regions, both gamma/GFP or epsilon/GFP subunits compete with endogenous epsilon subunits to assemble GFP-tagged receptors, which are selectively targeted to the postsynaptic membrane.

Electrical activity and postsynapse formation in adult muscle: gamma-AChRs are not required.

Skeletal muscle fibers will not accept hyperinnervation by foreign motor axons unless they are paralyzed, suggesting that paralysis makes them receptive to innervation, e.g., by upregulating extrasynaptic expression of gamma-AChRs and/or of the agrin receptor MuSK. To examine the involvement of these parameters in paralysis-mediated synapse induction, ectopic expression of agrin, a factor from motor neurons controlling neuromuscular synapse formation, was made dependent on the administration of doxycycline in innervated adult muscle fibers. In response to doxycycline-induced agrin secretion, adult fibers did form ectopic postsynaptic specializations, even when they were electrically active, lacked fetal AChRs, and expressed normal low levels of MuSK. These data demonstrate that paralysis and changes associated with it are not required for agrin-induced postsynapse formation. They suggest that paralyzed muscle induces synapse formation via the release of factors that make motor neurites contact muscle fibers and secrete agrin.

Gene transfer into individual muscle fibers and conditional gene expression in living animals.

Pressure injection of DNA directly into individual fibers of surgically exposed soleus muscle leads to efficient and reliable expression of the transgene. Conditionally regulated gene expression in a single muscle fiber was analyzed in vivo by co-injecting a tetracycline-regulated lacZ reporter construct and a transactivator (rtTA) expression vector. The tetracycline-responsive element revealed significant basal transcriptional activity that was further increased by rtTA even in the absence of the effector doxycycline (dox). The high basal activity of the simple two-component system precludes tight gene regulation in muscle. Concomitant expression of the silencer tTS(Kid), however, reduced the basal activity to low or undetectable levels. This allowed the specific activation of the tetracycline-responsive element by the application of dox. Direct gene transfer can thus be employed to express transgenic proteins in distinct muscle fibers at spatially defined regions and to regulate gene expression conditionally.

Different functions of fetal and adult AChR subtypes for the formation and maintenance of neuromuscular synapses revealed in epsilon-subunit-deficient mice.

Mice deficient in epsilon-subunits of the acetylcholine receptor (AChR) channel die prematurely due to severe AChR deficiency that leads to the progressive reduction in AChR density at the neuromuscular endplate [Witzemann, V., Schwarz, H., Koenen, M., Berberich, C., Villarroel, A., Wernig, A., Brenner, H.R. & Sakmann, B. (1996) Proc. Natl Acad. Sci. USA, 93, 13286-13291]. The mice may serve as a model for studying AChR-related myasthenic diseases. The postnatal development of the subsynaptic apparatus takes place in the absence of the adult type, epsilon-subunit-containing receptors which normally replace the fetal gamma-subunit-containing receptors. During later development the secondary folds of the postsynaptic membrane disappear concomitant with the decrease in AChR density, so that the flattened-out membrane with its remaining nicotinic receptors is in close proximity to the subsynaptic cytoplasmatic compartment and the subsynaptic myonuclei. The decrease in AChR concentration is correlated with a decrease of postsynaptic rapsyn, but has less effect on agrin, a neuronally released aggregating factor for AChRs. Thus, despite the presence of agrin at the synapse, AChR expression is not maintained at the level required to stabilize normal synaptic structure comprising secondary postsynaptic membrane folds. Collectively the results suggest that the postnatal switch from the global, activity-sensitive gamma-subunit gene transcription to the synapse-specific, activity-independent epsilon-subunit gene transcription is not required for the formation and differentiation of synapses but is essential for the maintenance of the highly organized structure of the neuromuscular endplate.

Waglerin-1 selectively blocks the epsilon form of the muscle nicotinic acetylcholine receptor.

Neonatal mice resist the lethal effect of Waglerin-1. Because Waglerin-1 blocks the nicotinic acetylcholine receptor of mature end-plates, the appearance of lethality may result from the epsilon- for gamma-subunit substitution. In support of this hypothesis, adult knockout (KO) mice lacking the gene coding for the epsilon-subunit resist the lethal effect of Waglerin-1. In contrast, heterozygous litter mates respond to Waglerin-1 like adult wild-type mice. In vitro application of 1 microM Waglerin-1 inhibited spontaneous miniature end-plate potentials and evoked end-plate potentials of adult wild-type and heterozygous KO mice. Both miniature end-plate potentials and end-plate potentials of neonatal wild-type and adult homozygous KO mice resisted Waglerin-1. Waglerin-1 decreased the end-plate response of adult wild-type mice to iontophoretically applied acetylcholine (ACh) with an IC50 value of 50 nM; 1 microM Waglerin-1 decreased the ACh response to 4 +/- 1% of control for adult heterozygous KO mice. In contrast, 1 microM Waglerin-1 decreased the ACh response to 73 +/- 2% of control for wild-type mice less than 11 days old and had no effect on the ACh response of adult homozygous KO mice. Between 11 and 12 days after birth, the suppressant effect of Waglerin-1 on wild-type end-plate responses to ACh dramatically increased. Waglerin-1 reduced binding of alpha-bungarotoxin to end-plates of adult but not neonatal wild-type mice. These data demonstrate that Waglerin-1 selectively blocks the mouse muscle nicotinic acetylcholine receptor containing the epsilon-subunit.

Identification and characterization of a novel splice variant of MuSK.

MuSK is a receptor tyrosine kinase that initiates the formation of neuromuscular junctions in response to agrin. Little is known about the ligand-induced activation and kinase-dependent signalling that leads to the clustering of acetylcholine receptors. The ectodomain of these molecule is composed of four Ig-like domains. We describe here the isolation of a novel MuSK splice variant that lacks the third Ig-like domain in its ectodomain. The corresponding RNA is the result of alternative splicing which eliminates two exons. There is 10 times less mRNA for this shorter form than for the long form of MuSK and both forms are regulated coordinately. They decrease strongly after birth and are elevated in denervated muscle. Gene transfer by muscle injection of MuSK DNA into individual muscle fibers demonstrates that kinase-induced acetylcholine receptor clustering caused by overexpression of the two kinases does not depend on the presence of the third Ig-like domain.

Formation of postsynaptic-like membranes during differentiation of embryonic stem cells in vitro.

To analyze the formation of neuromuscular junctions, mouse pluripotent embryonic stem (ES) cells were differentiated via embryoid bodies into skeletal muscle and neuronal cells. The developmentally controlled expression of skeletal muscle-specific genes coding for myf5, myogenin, myoD and myf6, alpha 1 subunit of the L-type calcium channel, cell adhesion molecule M-cadherin, and neuron-specific genes encoding the 68-, 160-, and 200-kDa neurofilament proteins, synaptic vesicle protein synaptophysin, brain-specific proteoglycan neurocan, and microtubule-associated protein tau was demonstrated by RT-PCR analysis. In addition, genes specifically expressed at neuromuscular junctions, the gamma- and epsilon-subunits of the nicotinic acetylcholine receptor (AChR) and the extracellular matrix protein S-laminin, were found. At the terminal differentiation stage characterized by the formation of multinucleated spontaneously contracting myotubes, the myogenic regulatory gene myf6 and the AChR epsilon-subunit gene, both specifically expressed in mature adult skeletal muscle, were found to be coexpressed. Only the terminally differentiated myotubes showed a clustering of nicotinic acetylcholine receptors (AChR) and a colocalization with agrin and synaptophysin. The formation of AChRs was also demonstrated on a functional level by using the patch clamp technique. Taken together, our results showed that during ES cell differentiation in vitro neuron- and muscle-specific genes are expressed in a developmentally controlled manner, resulting in the formation of postsynaptic-like membranes. Thus, the embryonic stem cell differentiation model will be helpful for studying cellular interactions at neuromuscular junctions by "loss of function" analysis in vitro.

Induction by agrin of ectopic and functional postsynaptic-like membrane in innervated muscle.

Two factors secreted from the nerve terminal, agrin and neuregulin, have been postulated to induce localization of the acetylcholine receptors (AChRs) to the subsynaptic membrane in skeletal muscle fibers. The principal function ascribed to neuregulin is induction of AChR subunit gene expression and to agrin is the aggregation of AChRs. Here we report that when myoblasts engineered to secrete an agrin fragment were placed into the nerve-free region of denervated rodent muscle, the host muscle fibers expressed AChR epsilon-subunit gene transcripts, characteristic of the neuromuscular synapse in adult muscle. Transcripts were colocalized with agrin deposits and AChR clusters that were resistant to electrical muscle activity. More directly, single innervated muscle fibers injected intracellularly with agrin expression plasmids in their extrasynaptic region developed a functional ectopic postsynaptic membrane with clusters of adult-type AChR channels and acetylcholinesterase and accumulation of myonuclei. The results demonstrate that agrin is the principal neural signal that induces the formation of the subsynaptic apparatus in the muscle fiber and controls locally, either indirectly or directly, the transcription of AChR subunit genes and the aggregation of AChRs.

An amino acid exchange in the second transmembrane segment of a neuronal nicotinic receptor causes partial epilepsy by altering its desensitization kinetics.

The alpha4 subunit of the neuronal nicotinic acetylcholine receptor is the first gene shown to be involved in a human idiopathic epileptic disease. A missense mutation, leading to the replacement of serine 248 by phenylalanine in the second transmembrane segment, had been detected in patients with autosomal dominant nocturnal frontal lobe epilepsy. The properties of the wild type receptor composed of alpha4 and beta2 subunits and the mutant receptor where alpha4 subunits carried the mutation at serine 248 were compared by means of cDNA manipulation and expression in Xenopus oocytes. The mutant receptor exhibited faster desensitization upon activation by acetylcholine and recovery from the desensitized state was much slower than in the wild type receptor. We conclude that the reported mutation causes seizures via a diminution of the activity of the alpha4beta2 neuronal nicotinic acetylcholine receptor.

Acetylcholine receptor epsilon-subunit deletion causes muscle weakness and atrophy in juvenile and adult mice.

In mammalian muscle a postnatal switch in functional properties of neuromuscular transmission occurs when miniature end plate currents become shorter and the conductance and Ca2+ permeability of end plate channels increases. These changes are due to replacement during early neonatal development of the gamma-subunit of the fetal acetylcholine receptor (AChR) by the epsilon-subunit. The long-term functional consequences of this switch for neuromuscular transmission and motor behavior of the animal remained elusive. We report that deletion of the epsilon-subunit gene caused in homozygous mutant mice the persistence of gamma-subunit gene expression in juvenile and adult animals. Neuromuscular transmission in these animals is based on fetal type AChRs present in the end plate at reduced density. Impaired neuromuscular transmission, progressive muscle weakness, and atrophy caused premature death 2 to 3 months after birth. The results demonstrate that postnatal incorporation into the end plate of epsilon-subunit containing AChRs is essential for normal development of skeletal muscle.

Polyamine spider toxins and mammalian N-methyl-D-aspartate receptors. Structural basis for channel blocking and binding of argiotoxin636.

Recombinant N-methyl-D-aspartate receptors composed of NR1/NR2A subunits were expressed in Xenopus oocytes to analyse the voltage-dependent and use-dependent channel blocking activity of argiotoxin636. Functional assays demonstrate that the toxin competes with other open channel blockers such as Mg2+ and MK-801. Direct binding or competition assays using radiolabeled ligands and isolated rat brain membranes, in contrast, reveal no specific binding or yield binding constants which differ by orders of magnitude from the IC50 values of the functional assays. One explanation is that argiotoxin636 does not bind with high affinity to the inhibitory site in the N-methyl-D-aspartate-receptor channel under in vitro conditions when membranes are depolarised. The structure of argiotoxin636 was investigated by NMR spectroscopy. In solution the positively charged argiotoxin636 acquires an extended conformation and its dimensions might allow permeation deep into the channel. In the absence of direct structural information on the channel protein, the detailed analysis of blockade in conjunction with structural information, as provided here, may be of aid in the deduction of structural features of glutamate-receptor channel ion pores.

Pharmacology of the nicotinic acetylcholine receptor from fetal rat muscle expressed in Xenopus oocytes.

The fetal rat muscle nicotinic acetylcholine receptor was expressed in Xenopus oocytes. Using the voltage-clamp technique, the response to a range of agonists was measured, listed in order of (decreasing) activity efficacy: anatoxin > or = epibatidine > acetylcholine > DMPP (1,1-dimethyl-4-phenylpiperazinium) > > cytisine > pyrantel > nicotine > coniine > tubocurare > lobeline. The agonist responses were compared with the steric and electrostatic properties of the molecules, using molecular modelling. Single-channel current were measured in outside-out patches for acetylcholine, nicotine, cytisine, anatoxin and epibatidine. The conductance of the single channels was independent of the type of agonist. The mean open times were characteristic of the agonist applied. Tubocurare, better known for its antagonist properties, was also a partial agonist. Single-channel currents were also observed for tubocurare, and for methyllycaconitine in patches with a very high density of the muscle nicotinic acetylcholine receptor, and these were blocked by alpha-bungarotoxin. The agonist properties of physostigmine, galanthamine and their methyl derivatives were also investigated. The conductance of the channels observed in outside-out patches was similar to that obtained for the classical agonists. The single-channel currents observed for physostigmine, galanthamine and their methyl derivatives were blocked by alpha-bungarotoxin, methyllycaconitine and mecamylamine, in contrast to previously reported studies on neuronal and adult muscle nicotinic acetylcholine receptors.

Structural determinants of channel conductance in fetal and adult rat muscle acetylcholine receptors.

1. The structural basis of the developmentally regulated increase in endplate channel conductance in rat, where the gamma-subunit of the fetal muscle acetylcholine receptor (gamma-AChR) is replaced by the epsilon-subunit in the adult muscle receptor (epsilon-AChR), was investigated by analysing the structure of gamma- and epsilon-subunit genes and by expressing recombinant AChR channels of different molecular composition in Xenopus oocytes and measuring their single-channel conductance. 2. The gamma- and epsilon-subunit genes each have twelve exons. In both subunits, the four homologous segments, designated M1, M2, M3 and M4, which are thought to contribute to the formation of the pore, are encoded by four separate exons, E7, E8, E9 and E12. 3. Chimaeric epsilon(gamma)- or gamma(epsilon)-subunits were constructed from the parental epsilon- and gamma-subunits, respectively. Exchanging the four hydrophobic segments (M1-M4) of the gamma-subunit for those of the epsilon-subunit and vice versa completely reversed the difference in conductance between gamma-AChR and epsilon-AChR channels. 4. Effects of single- and multiple-point mutations in M1-M4 segments of gamma- and epsilon-subunits indicate that the major determinants of the difference in conductance between fetal and adult endplate channels are located in the M2 segment. The key differences are the exchange of alanine/threonine (gamma-subunit) for serine/isoleucine (epsilon-subunit) in M2, and the lysine (gamma-subunit) for glutamine (epsilon-subunit) exchanges in the regions flanking the M2 segment.

Local neurotrophic repression of gene transcripts encoding fetal AChRs at rat neuromuscular synapses.

The spatio-temporal expression patterns of mRNA transcripts coding for acetylcholine receptor (AChR) subunits and myogenic factors were measured in denervated rat soleus muscle and in soleus muscle chronically paralyzed for up to 12 d by conduction block of the sciatic nerve by tetrodotoxin (TTX). In denervated muscle the AChR alpha-, beta-, gamma-, and delta-subunit mRNAs were elevated with highest expression levels in the former synaptic and the perisynaptic region and with lower levels in the extrasynaptic fiber segments. In muscle paralyzed by nerve conduction block the alpha-, beta-, gamma-, and delta-subunit mRNA levels increased only in extrasynaptic fiber segments. Surprisingly, in the synaptic region the gamma-subunit mRNA that specifies the fetal-type AChR, and alpha-, beta-, delta-subunit mRNAs were not elevated. The expression of the gene encoding the epsilon-subunit, which specifies the adult-type AChR, was always restricted to synaptic nuclei. The mRNA for the regulatory factor myogenin showed after denervation similar changes as the subunit transcripts of the fetal AChR. When the muscle was paralyzed by nerve conduction block the increase of myogenin transcripts was also less pronounced in synaptic regions compared to extrasynaptic fiber segments. The results suggest that in normal soleus muscle a neurotrophic signal from the nerve locally down-regulates the expression of fetal-type AChR channel in the synaptic and perisynaptic muscle membrane by inhibiting the expression of the gamma-subunit gene and that inhibition of the myogenin gene expression may contribute to this down-regulation.

Differential expression patterns of five acetylcholine receptor subunit genes in rat muscle during development.

The spatial and temporal expression patterns of five genes which encode the alpha-, beta-, gamma-, delta- and epsilon-subunits of the nicotinic acetylcholine receptor in skeletal muscle were followed during development in the rat by in situ hybridization analysis. Three major developmental phases, characterized by specific expression patterns, could be distinguished. (i) During myogenic differentiation alpha-, beta-, gamma- and delta-subunit genes are activated and transcripts are expressed in muscle precursor cells at embryonic day 12 (E12) and during subsequent cell fusion. (ii) Following innervation of myotubes at approximately E15-E17 the mRNA of the alpha-, beta-, gamma- and delta-subunit genes accumulate in synaptic and decrease in extrasynaptic fibre regions during early synaptogenesis. The mRNA of the epsilon-subunit gene becomes detectable first in subsynaptic nuclei 2-3 days after innervation has occurred. (iii) During postnatal development alpha-, beta- and delta- subunit transcript levels are reduced predominantly in extrasynaptic fibre segments and show significant differences in distribution depending on the muscle subtype whereas the gamma-subunit mRNA disappears completely within the first postnatal week in all muscles. In contrast, the epsilon-subunit gene is transcribed only in subsynaptic myonuclei throughout development and in the adult muscle.

Internodal cells of the giant green alga Chara as an expression system for ion channels.

Internodal cells of the giant green alga Chara corallina were utilized as an expression system for two nicotinic acetylcholine receptor subtypes (nAChR) derived from rat muscle. From Chara internodes that were pressure-injected with the respective cRNA, cytoplasmic droplets were formed, and functional expression of channel proteins in the membrane delineating the droplets was confirmed by patch-clamp techniques. The droplet membrane was recently identified as the original tonoplast, patch-clamp recordings on cytoplasmic droplets were easily performed and single channel activity could be measured with high resolution. The properties of the recombinant nAChR observed in this membrane were similar to those reported for the channel expressed in Xenopus oocytes, and for the native channel recorded in situ. The Chara expression system is, therefore, suitable for functional expression of animal messenger RNAs, without the risk of signal contamination by intrinsic ion channels. It offers a low-cost and practical alternative to the use of Xenopus oocytes for the investigation of heterologously expressed ion channels.

PVP-containing solutions for analysis of divalent cation-dependent NMDA responses in Xenopus oocytes.

The electrophysiological analysis of Ca(2+)-conducting ion channels in Xenopus oocytes is difficult due to secondary intracellular effects induced by Ca2+. In the presence of polyvinylpyrrolidone (PVP) membrane currents can be recorded in nominally divalent cation-free solutions. The Ca(2+)-permeable recombinant NMDA receptors of the NR1/NR2A subtype were used as assay system and the results show that PVP has no effect on NMDA receptor-induced currents. Ca2+ and Ba2+ depress NMDA-induced currents at submillimolar concentrations probably by interfering with the Na+/K+ flux. This block is fully reversible as also observed for Mg2+ but shows in contrast no pronounced voltage dependence. PVP-containing solutions may be useful for the analysis of divalent cation-dependent ion channels.

Nerve-dependent induction of AChR epsilon-subunit gene expression in muscle is independent of state of differentiation.

To determine if the expression of the epsilon-subunit of the acetylcholine receptor by the subsynaptic nuclei in skeletal muscle is dependent on the state of differentiation of the muscle, we have compared the spatiotemporal distribution of epsilon-subunit transcripts during synapse formation in fetal and adult muscle. Both during "ontogenic" synaptogenesis in the fetus and during "ectopic" synaptogenesis in the adult animal the motor nerve induced focally the expression of the epsilon-subunit mRNA in subsynaptic nuclei. The temporal expression patterns at both types of developing synapses were similar. The results support the view that in muscle developing in vivo epsilon-subunit gene transcription and its stabilization in subsynaptic nuclei is exclusively controlled by the motor neuron, independently of the developmental state of the muscle nuclei. Thus, both nerve and muscle remain plastic in their respective abilities to induce and express the synapse-specific combination of AChR subunit genes.

Characterization of the functional role of E-box elements for the transcriptional activity of rat acetylcholine receptor epsilon-subunit and gamma-subunit gene promoters in primary muscle cell cultures.

The expression of gamma and epsilon subunits of the acetylcholine receptor from mammalian skeletal muscle is regulated independently during myogenic differentiation and innervation. Genomic DNA fragments containing 5'-flanking sequences of the epsilon-subunit and gamma-subunit genes were characterised by a series of 5' deletions fused to the chloramphenicol-acetyltransferase gene and transiently expressed by transfection of primary cultures of rat muscle cells and non-muscle cells. A 6.3-kb epsilon-subunit fragment can be reduced to yield a 270-bp fragment that confers 5-10-times higher expression levels in muscle cells compared to in non-muscle cells. The region composed of nucleotides -185 to -128 increases the transcriptional activity moderately while the 14-bp palindrome containing a single E box at nucleotides -88 to -83 may interact with the promoter but has no enhancer properties in muscle cells. From a 1.1-kb genomic fragment of the gamma-subunit gene, 167 bp were sufficient for muscle-specific expression. Two promoter-proximal E-box elements enhance promoter activity in muscle and mediate transactivation by myogenic factors. Myogenin and myf5 were much more efficient than MRF4 or MyoD1 which exerted only little transactivation. Cotransfection experiments show that increased expression of Id in primary muscle cells inhibits chloramphenicol-acetyltransferase expression mediated by the gamma-subunit gene promoter and support the view that myogenic factors play an important role in the transcriptional regulation of the gamma-subunit gene.