Globular domains of agrin are functional units that collaborate to induce acetylcholine receptor clustering.

Agrin, an extracellular matrix protein involved in neuromuscular junction formation, directs clustering of postsynaptic molecules, including acetylcholine receptors (AChRs). This activity resides entirely in the C-terminal portion of the protein, which consists of three laminin-like globular domains (G-domains: G1, G2 and G3) and four EGF-like repeats. Additionally, alternate mRNA splicing yields G-domain variants G2(0,4) with 0- or 4-amino-acid inserts, and G3(0, 8,11,19) with 0-, 8-, 11- or 19-amino-acid inserts. In order to better understand the contributions of individual domains and alternate splicing to agrin activity, single G-domains and covalently linked pairs of G-domains were expressed as soluble proteins and their AChR clustering activity measured on cultured C2 myotubes. These analyses reveal the following: (1) While only G3(8) exhibits detectable activity by itself, all G-domains studied (G1, G2(0), G2(4), G3(0) and G3(8)) enhance G3(8) activity when physically linked to G3(8). This effect is most pronounced when G2(4) is linked to G3(8) and is independent of the order of the G-domains. (2) The deletion of EGF-like repeats enhances activity. (3) Increasing the physical separation between linked G1 and G3(8) domains produces a significant increase in activity; similar alterations to linked G2 and G3(8) domains are without effect. (4) Clusters induced by two concatenated G3(8) domains are significantly smaller than all other agrin forms studied. These data suggest that agrin G-domains are the functional units which interact independently of their specific organization to yield AChR clustering. G-domain synergism resulting in biological output could be due to physical interactions between G-domains or, alternatively, independent interactions of G-domains with cell surface receptors which require spatially localized coactivation for optimal signal transduction.

WW and EF hand domains of dystrophin-family proteins mediate dystroglycan binding.

The membrane-spanning dystrophin glycoprotein complex mediates an indirect linkage between the actin-based cytoskeleton and the extracellular matrix. Although expressed by diverse cell types, genetic lesions of members of this complex result in muscular dystrophy phenotypes emphasizing the importance of these interactions in muscle cells. We have characterized interactions between dystrophin family members and dystroglycan: cytoskeletal and transmembrane components of the complex, respectively. Our results demonstrate that both the WW and EF hand domains of dystrophin and utrophin, an autosomal homologue of dystrophin, directly bind the cytoplasmic domain of dystroglycan. Furthermore, alpha-dystrobrevin, a more distantly related dystrophin family member which lacks a WW domain but contains the EF hand domain, binds dystroglycan. This is the first demonstration of a direct interaction between a dystrobrevin or utrophin and dystroglycan, and has implications for the organization of the dystrophin glycoprotein complex and the use of dystrophin homologues in muscular dystrophy therapy.

A functional role for specific spliced variants of the alpha7beta1 integrin in acetylcholine receptor clustering.

The clustering of acetylcholine receptors (AChR) on skeletal muscle fibers is an early event in the formation of neuromuscular junctions. Recent studies show that laminin as well as agrin can induce AChR clustering. Since the alpha7beta1 integrin is a major laminin receptor in skeletal muscle, we determined if this integrin participates in laminin and/or agrin-induced AChR clustering. The alternative cytoplasmic domain variants, alpha7A and alpha7B, and the extracellular spliced forms, alpha7X1 and alpha7X2, were studied for their ability to engage in AChR clustering. Immunofluorescence microscopy of C2C12 myofibers shows that the alpha7beta1 integrin colocalizes with laminin-induced AChR clusters and to a much lesser extent with agrin-induced AChR clusters. However, together laminin and agrin promote a synergistic response and all AChR colocalize with the integrin. Laminin also induces the physical association of the integrin and AChR. High concentrations of anti-alpha7 antibodies inhibit colocalization of the integrin with AChR clusters as well as the enhanced response promoted by both laminin and agrin. Engaging the integrin with low concentrations of anti-alpha7 antibody initiates cluster formation in the absence of agrin or laminin. Whereas both the alpha7A and alpha7B cytoplasmic domain variants cluster with AChR, only those isoforms containing the alpha7X2 extracellular domain were active. These results demonstrate that the alpha7beta1 integrin has a physiologic role in laminin-induced AChR clustering, that alternative splicing is integral to this function of the alpha7 chain, and that laminin, agrin, and the alpha7beta1 integrin interact in a common or convergent pathway in the formation of neuromuscular junctions.

Agrin inhibits neurite outgrowth but promotes attachment of embryonic motor and sensory neurons.

Agrin is a secreted glycoprotein with the ability to cluster cell surface molecules, including the nicotinic acetylcholine receptor (AchR) on muscle cells. Alternate splicing of agrin mRNA results in a family of agrin proteins which differ in their clustering potency. Neuronal-specific isoforms with the highest clustering activity play a role in clustering postsynaptic proteins at the neuromuscular junction. However, the function of agrin isoforms expressed in many nonneuronal tissues, and only weakly active in clustering assays, remains obscure. Monolayer cultures of Chinese hamster ovary (CHO) cells expressing a neuronal (agrin-19) or a nonneuronal (agrin-0) form of agrin were used to assay the effect of agrin on neurite outgrowth and cell attachment. These results were compared to outgrowth on control CHO cells expressing only drug resistance and on regions of CHO-agrin monolayers not expressing detectable levels of agrin. Neurite extension on confluent monolayers of agrin-0- or -19-expressing CHO cells was reduced substantially below that of controls. In one experiment neurite lengths were compared at 2 and 3 days after plating and suggested that neurite outgrowth may be stopped and not simply retarded. Attachment of sensory or motoneurons was nearly twofold higher to agrin monolayers than to control cells, showing that the inhibition is not a result of a nonpermissive environment. An agrin construct missing the C-terminal half, removing the major site of variability and clustering activity, was also tested. This construct did not reduce outgrowth, suggesting that the C-terminal half of the protein may be important in stopping growth as well as inducing clustering. These results expand the role of agrin in synaptogenesis as it may provide a stop signal at the myofiber surface and may anchor the presynaptic fibers to the eventual motor endplate .

Alternative RNA splicing that determines agrin activity regulates binding to heparin and alpha-dystroglycan.

Agrin is a component of the extracellular matrix that regulates aspects of neuromuscular junction differentiation. Identification of agrin-binding proteins has lead to the suggestion that alpha-dystroglycan is a muscle cell surface proteoglycan that mediates agrin activity. To further test this hypothesis, we have compared the ability of differentially active agrin isoforms to interact with a model component of proteoglycans, heparin, as well as with the putative proteoglycan alpha-dystroglycan. We demonstrate that an alternately spliced exon (encoding the sequence lysine, serine, arginine, lysine: Y site) is necessary for agrin-heparin interactions. We also show that alternate splicing at another site (Z site) dramatically affects interaction of alpha-dystroglycan with agrin. We propose a model in which multiple distinct domains of agrin interact with both protein and sugar moieties of alpha-dystroglycan. The isoform-specific binding of agrin to alpha-dystroglycan is consistent with a functional role for this interaction during synaptogenesis.

A role for dystrophin-associated glycoproteins and utrophin in agrin-induced AChR clustering.

Synapse formation is characterized by the accumulation of molecules at the site of contact between pre- and postsynaptic cells. Agrin, a protein implicated in the regulation of this process, causes the clustering of acetylcholine receptors (AChRs). Here we characterize an agrin-binding site on the surface of muscle cells, show that this site corresponds to alpha-dystroglycan, and present evidence that alpha-dystroglycan is functionally related to agrin activity. Furthermore, we demonstrate that alpha-dystroglycan and adhalin, components of the dystrophin-associated glycoprotein complex, as well as utrophin, colocalize with agrin-induced AChR clusters. Thus, agrin may function by initiating or stabilizing a synapse-specific membrane cytoskeleton that in turn serves as a scaffold upon which synaptic molecules are concentrated.

Structural domains of agrin required for clustering of nicotinic acetylcholine receptors.

Agrin is an extracellular matrix component which promotes the clustering of nicotinic acetylcholine receptors (nAChRs) and other proteins at the neuromuscular junction. This aggregation process is one of the earliest steps in synapse formation. Expression of highly active isoforms of agrin, generated by alternative splicing, is restricted to neurons in the central nervous system (CNS) including motoneurons. In the experiments reported here we investigate the regions of agrin necessary for nAChR clustering activity using two different methods. First, we expressed truncated soluble forms of the agrin protein in mammalian cells and assessed their clustering activity. Second, we generated a panel of monoclonal antibodies (mAbs) against agrin and mapped their epitopes. Several mAbs block agrin-induced aggregation of nAChRs. One of the mAbs, Agr86, binds exclusively to the CNS-specific splicing variants and thus identifies an epitope common only to these more active isoforms. Mapping of the Agr86 epitope suggests that alternative splicing results in a distributed conformational change in the agrin protein. Taken together our data suggest that four domains in the C-terminal 55 kDa of agrin contribute to its nAChR clustering activity.

Developmental regulation of highly active alternatively spliced forms of agrin.

Agrin is an extracellular matrix protein involved in clustering acetylcholine receptors during development of the neuromuscular junction. We have previously shown that alternative splicing at three sites generates multiple forms of rat agrin and that a novel 8 amino acid insert is the most important in determining biological activity. In the present study we have examined the expression of agrin during development with particular emphasis on determining the tissue distribution of the splicing variants at each site. Our principal observation is that the variants containing the sequence most responsible for biological activity are expressed exclusively in neural tissue and that their expression is highly regulated during development. We also show that muscle expresses less active forms and that agrin immunoreactivity during synaptogenesis is initially not limited to synaptic sites, but becomes progressively restricted to the synapse as development proceeds.

The ability of agrin to cluster AChRs depends on alternative splicing and on cell surface proteoglycans.

Agrin, which induces acetylcholine receptor (AChR) clustering at the developing neuromuscular synapse, occurs in multiple forms generated by alternative splicing. Some of these isoforms are specific to the nervous system; others are expressed in both neural and nonneural tissues, including muscle. We have compared the AChR clustering activity of agrin forms varying at each of the three identified splicing sites, denoted x, y, and z. Agrin isoforms were assayed by applying either transfected COS cells, with agrin bound to their surfaces, or soluble agrin to myotubes of the C2 muscle line, or of two variant lines having defective proteoglycans. Dramatic differences in activity were seen between z site isoforms and lesser differences between y site isoforms. The most active agrin forms contained splicing inserts of 4 amino acids at the y site and 8 amino acids at the z site. These forms are found exclusively in neural tissue. All forms were active on C2 myotubes in cell-attached assays, but muscle forms were less active than neural forms. AChR clustering activity of all agrin forms was decreased when assayed on the proteoglycan-deficient lines, suggesting that proteoglycans may help mediate the action of agrin. As neural agrin forms are more active than muscle forms, they are likely to play a primary role in synaptogenesis.

RNA splicing regulates agrin-mediated acetylcholine receptor clustering activity on cultured myotubes.

Agrin is a component of the synaptic basal lamina that induces the clustering of acetylcholine receptors (AChRs) on muscle fibers. A region near the carboxyl terminus of the protein exists in four forms that are generated by alternative RNA splicing. All four alternatively spliced forms of agrin are active in inducing AChR clusters on rat primary and C2-derived muscle fibers. In contrast, only two forms of the protein, each containing an 8 amino acid insert, are capable of inducing clusters on myotubes of S27 cells, a C2 variant that has defective proteoglycans. These two forms are also most active in inducing clusters on chick myotubes. This pattern of differential activity suggests that RNA splicing of agrin transcripts and interactions with proteoglycans or other components of basal lamina have important roles in regulating the localization of neurotransmitter receptors at synaptic sites.

Agrin and the organization of the neuromuscular junction.

Agrin is a component of the synaptic extracellular matrix and may regulate the organization of acetylcholine receptors and other synaptic molecules in both synapse regeneration and development. Analyses of cDNAs encoding agrin define a number of structural domains, including regions of homology to laminin, Kazal protease inhibitors, and epidermal growth factor repeats.

Agrin mediates cell contact-induced acetylcholine receptor clustering.

One of the important events in synapse formation is the accumulation of neurotransmitter receptors beneath the presynaptic nerve terminal. Agrin is a component of the synaptic basal lamina that induces the clustering of acetylcholine receptors when bath-applied to muscle fibers in culture. When a cDNA encoding a putative agrin protein is transfected into cells, the molecule is secreted and concentrated on the extracellular surface. Coculture of transfected cells with muscle fibers induces the formation of receptor patches at contact sites. These results demonstrate that expression of a single gene encoding agrin confers receptor clustering that is restricted to specific sites of cell-muscle contact.

A role for hydrophobic residues in the voltage-dependent gating of Shaker K+ channels.

A leucine heptad repeat is well conserved in voltage-dependent ion channels. Herein we examine the role of the repeat region in Shaker K+ channels through substitution of the leucines in the repeat and through coexpression of normal and truncated products. In contrast to leucine-zipper DNA-binding proteins, we find that the subunit assembly of Shaker does not depend on the leucine heptad repeat. Instead, we report that substitutions of the leucines in the repeat produce large effects on the observed voltage dependence of conductance voltage and prepulse inactivation curves. Our results suggest that the leucines mediate interactions that play an important role in the transduction of charge movement into channel opening and closing.

Synuclein: a neuron-specific protein localized to the nucleus and presynaptic nerve terminal.

We used an antiserum against purified cholinergic synaptic vesicles from Torpedo and expression screening to isolate a cDNA clone encoding synuclein, a 143 amino acid neuron-specific protein. A cDNA clone was also isolated from a rat brain cDNA library that encodes a highly homologous 140 amino acid protein. The amino terminal 100 amino acids of both proteins are comprised of an 11 amino acid repeating unit that contains a conserved core of 6 residues. The synuclein gene is expressed only in nervous system tissue, not in electric organ, muscle, liver, spleen, heart, or kidney. In the electric organ synapse Torpedo synuclein-immunoreactive proteins are found in 3 major molecular-weight classes of 17.5, 18.5, and 20.0 kDa. In the neuronal cell soma the 17.5 kDa species is predominant and immunoreactivity is localized to a portion of the nuclear envelope.

Histidine-rich basic peptide: a cardioactive neuropeptide from Aplysia neurons R3-14.

We have previously demonstrated that neurons R3-14 of the Aplysia abdominal ganglia specifically express a gene encoding a 108-amino acid neuropeptide precursor. This precursor is postranslationally processed by cleavage of a signal sequence and two internal dibasic residues resulting in three peptides. The peptide products are colocalized in dense core granules throughout the R3-14 processes that innervate the efferent vein of the gill and the auricle. Gel filtration and reverse phase high-pressure liquid chromatography (rpHPLC) were used to purify a 4.9-kDa peptide produced by the R3-14 neurons. We call this peptide the histidine-rich basic peptide (HRBP), which reflects its primary structure. In vitro tension measurements of cannulated Aplysia hearts revealed dose-dependent cardioexcitatory actions of HRBP. HRBP increased both beat frequency and amplitude with a threshold of 10(-7) M. HRBP increased the amplitude of ventricular contractions in a dose-dependent manner, whereas the frequency of contraction is unaffected. In contrast both the amplitude and frequency of auricular contractions were enhanced. High concentrations of HRBP also had a positive tonotropic effect on the auricle. HRBP was also demonstrated to have actions on tissue of the gut. Circular muscles of the crop adjacent to the anterior gizzard showed infrequent spontaneous contractions. Both HRBP and acetylcholine (ACh) induced repetitive contractions of this muscle. Circular muscles of the posterior gizzard had a high degree of spontaneous activity when continually perfused. Contraction amplitude and frequency was increased by HRBP and ACh, whereas contractility was inhibited by Phe-Met-Arg-Phe-amide (FMRFamide).

Proteolytic processing of a peptide precursor in Aplysia neuron R14.

The large neurons of the mollusc Aplysia are useful for studying the biogenesis of neuropeptides in single cells. Neuron R14 in the abdominal ganglion synthesizes large quantities of a 10-kDa neuropeptide precursor. The amino acid sequence of this precursor has been defined by analysis of the nucleotide sequence of a cDNA clone. We labeled proteins in vivo by microinjection of radioactive amino acids into individual R14 neurons. The labeled peptides were fractionated by high performance liquid chromatography and subjected to Edman degradation, thus enabling us to determine post-translational processing sites. Cleavage of the signal sequence was observed and at two internal sites. Cleavage at these internal sites occurs at basic amino acids and results in three products, a 2.9-, a 4.9-, and a 1.4-kDa peptide. These studies of protein processing serve as a basis for further investigations of the biogenesis and physiological activities of the neuropeptides.

Low molecular weight proteins of Aplysia neurosecretory cells.

The proteins of identified cells from the Aplysia californica central nervous system were labeled with radioactive amino acids and fractionated on SDS acrylamide gels containing 6 M urea. Most of the large cells contain prominent, cell-specific protein products in the molecular weight range between 3 and 30 KD. The molecular weights of the largest specific prevalent protein products are in good agreement with the predicted molecular weights of precursors as determined from an analysis of cDNA clones homologous to mRNA's specifically expressed in several of these neurons. Biologically active peptides have been found in many of these cells. These data, and other indirect evidence suggests that the synthesis of a large amount of a particular protein in this molecular weight range is indicative of the synthesis of a neurosecretory product. We conclude that most, if not all, large neurons in the Aplysia central nervous system are peptidergic.