Association of muscle-specific kinase MuSK with the acetylcholine receptor in mammalian muscle
1997; Springer Nature; Volume: 16; Issue: 16 Linguagem: Inglês
10.1093/emboj/16.16.4951
ISSN1460-2075
Autores Tópico(s)Cellular Mechanics and Interactions
ResumoArticle15 August 1997free access Association of muscle-specific kinase MuSK with the acetylcholine receptor in mammalian muscle Christian Fuhrer Christian Fuhrer Section on Synaptic Mechanisms, Laboratory of Cellular and Molecular Regulation, National Institute of Mental Health, NIH, Bethesda, MD, 20892 USA Search for more papers by this author Janice E. Sugiyama Janice E. Sugiyama Section on Synaptic Mechanisms, Laboratory of Cellular and Molecular Regulation, National Institute of Mental Health, NIH, Bethesda, MD, 20892 USA Search for more papers by this author Robin G. Taylor Robin G. Taylor Section on Synaptic Mechanisms, Laboratory of Cellular and Molecular Regulation, National Institute of Mental Health, NIH, Bethesda, MD, 20892 USA Search for more papers by this author Zach W. Hall Corresponding Author Zach W. Hall Section on Synaptic Mechanisms, Laboratory of Cellular and Molecular Regulation, National Institute of Mental Health, NIH, Bethesda, MD, 20892 USA Search for more papers by this author Christian Fuhrer Christian Fuhrer Section on Synaptic Mechanisms, Laboratory of Cellular and Molecular Regulation, National Institute of Mental Health, NIH, Bethesda, MD, 20892 USA Search for more papers by this author Janice E. Sugiyama Janice E. Sugiyama Section on Synaptic Mechanisms, Laboratory of Cellular and Molecular Regulation, National Institute of Mental Health, NIH, Bethesda, MD, 20892 USA Search for more papers by this author Robin G. Taylor Robin G. Taylor Section on Synaptic Mechanisms, Laboratory of Cellular and Molecular Regulation, National Institute of Mental Health, NIH, Bethesda, MD, 20892 USA Search for more papers by this author Zach W. Hall Corresponding Author Zach W. Hall Section on Synaptic Mechanisms, Laboratory of Cellular and Molecular Regulation, National Institute of Mental Health, NIH, Bethesda, MD, 20892 USA Search for more papers by this author Author Information Christian Fuhrer1, Janice E. Sugiyama1, Robin G. Taylor1 and Zach W. Hall 1 1Section on Synaptic Mechanisms, Laboratory of Cellular and Molecular Regulation, National Institute of Mental Health, NIH, Bethesda, MD, 20892 USA *Corresponding author. E-mail: [email protected] The EMBO Journal (1997)16:4951-4960https://doi.org/10.1093/emboj/16.16.4951 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info During synaptogenesis at the neuromuscular junction, a neurally released factor, agrin, causes the clustering of acetylcholine receptors (AChRs) in the muscle membrane beneath the nerve terminal. Agrin acts through a specific receptor which is thought to have a receptor tyrosine kinase, MuSK, as one of its components. In agrin-treated muscle cells, both MuSK and the AChR become tyrosine phosphorylated. To determine how the activation of MuSK leads to AChR clustering, we have investigated their interaction in cultured C2 myotubes. Immunoprecipitation experiments showed that MuSK is associated with the AChR and that this association is increased by agrin treatment. Agrin also caused a transient activation of the AChR-associated MuSK, as demonstrated by MuSK phosphorylation. In agrin-treated myotubes, MuSK phosphorylation increased with the same time course as phosphorylation of the β subunit of the AChR, but declined more quickly. Although both herbimycin and staurosporine blocked agrin-induced AChR phosphorylation, only herbimycin inhibited the phosphorylation of MuSK. These results suggest that although agrin increases the amount of activated MuSK that is associated with the AChR, MuSK is not directly responsible for AChR phosphorylation but acts through other kinases. Introduction Differentiation of the synaptic apparatus at the neuromuscular junction occurs through the exchange of signals between nerve and muscle cells. One of the earliest signals is the release of agrin by nerve terminals (McMahan, 1990; Reist et al., 1992). Agrin acts on a specific receptor in the muscle cell membrane to cause the clustering of acetylcholine receptors (AChRs) beneath the nerve terminal (Nastuk et al., 1991). Due to alternative splicing, agrin exists in several isoforms that differ in AChR clustering activities and tissue distribution (Ferns et al., 1992; Ruegg et al., 1992). The most active forms are expressed exclusively in neurons, whereas the predominant agrin forms in muscle are much less effective in aggregating AChRs (Ferns et al., 1993; Hoch et al., 1993). In addition to its action on the AChR, neural agrin causes other synaptic components to be concentrated in association with the AChR clusters (Wallace, 1989), suggesting that it is a general organizing factor for postsynaptic differentiation. In genetically altered recombinant mice that do not express agrin, differentiated synapses fail to form and few nerve-associated AChR clusters are seen on muscle fibers (Gautam et al., 1996). The muscle receptor for neurally released agrin has not been fully defined, but a trk-related receptor tyrosine kinase, MuSK, appears to be one of its components (Jennings et al., 1993; Valenzuela et al., 1995). Like the AChR, MuSK is localized in the synapse in mature muscles and its expression is increased extrasynaptically following denervation (Valenzuela et al., 1995). Neural agrin induces tyrosine autophosphorylation of MuSK, and appears to bind a complex containing MuSK and one or more muscle-specific factors (Glass et al., 1996). Most importantly, mice in which the MuSK gene has been disrupted have a phenotype that is similar to that seen in agrin-deficient mice, in that the muscle fibers lack differentiated synapses and nerve-associated AChR clusters (DeChiara et al., 1996). In addition, cultured myotubes from the MuSK−/− mice fail to cluster AChRs in response to agrin (Glass et al., 1996). Another possible constituent of the receptor for agrin is α-dystroglycan, which is the predominant agrin-binding protein in muscle, and which is a component of the dystrophin–utrophin glycoprotein complex of skeletal muscle (Bowe et al., 1994; Campanelli et al., 1994; Gee et al., 1994; Sugiyama et al., 1994). Although α-dystroglycan binds to agrin, its functional role in agrin-mediated synaptic organization has remained unclear (Sealock and Froehner, 1994) because it binds to active neural as well as inactive muscle isoforms of agrin (Sugiyama et al., 1994; Bowen et al., 1996) and because short recombinant C-terminal fragments of neural agrin, which do not bind to α-dystroglycan, retain significant AChR clustering activity (Gesemann et al., 1995, 1996; Hopf and Hoch, 1996). In addition to the activation of MuSK, agrin induces tyrosine phosphorylation of the β subunit of the AChR (Wallace et al., 1991; Qu and Huganir, 1994; Meier et al., 1995; Ferns et al., 1996). In mammalian muscle, this phosphorylation is transient, with a time course that precedes the time course of AChR cluster formation. Although not demonstrated to be required for AChR cluster formation in muscle cells, phosphorylation of the AChR β subunit is well correlated with the formation of AChR clusters under several conditions, suggesting that it may play a critical role (Ferns et al., 1996; Meier et al., 1996). In support of this idea, inhibitors that block tyrosine phosphorylation of the AChR also block AChR cluster formation. Thus, herbimycin, a specific inhibitor of tyrosine kinases, and staurosporine, a more general inhibitor of protein kinases, block both agrin-induced AChR phosphorylation and AChR clustering (Wallace, 1994; Ferns et al., 1996). Agrin-induced tyrosine phosphorylation of the AChR could, in principle, occur by several mechanisms; agrin-activated MuSK could phosphorylate the AChR directly or it could act through other kinases. We have previously investigated AChR phosphorylation in mammalian muscle by src family kinases (Fuhrer and Hall, 1996). Our experiments showed that src binds to, and can phosphorylate, fusion proteins derived from the cytoplasmic loop of the AChR β subunit; the binding region on the β subunit contains a conserved tyrosine that is phosphorylated in the receptor isolated from the electric organ of Torpedo (Wagner et al., 1991). Immunoprecipitation experiments showed that muscle AChR is associated with both src and fyn (Fuhrer and Hall, 1996), raising the possibility that src phosphorylates the AChR, and that fyn becomes bound as a result of this phosphorylation (Swope and Huganir, 1993, 1994). Another route leading to AChR phosphorylation might involve direct phosphorylation of the β subunit by MuSK. In support of this idea, the AChR β subunit is tyrosine phosphorylated when the AChR, MuSK and the postsynaptic 43 kDa protein, rapsyn, are co-expressed in transfected QT-6 quail fibroblasts (Gillespie et al., 1996). As this phosphorylation requires MuSK kinase activity, the β subunit of the AChR could be a direct substrate for MuSK, as proposed by Gillespie et al. (1996); alternatively, MuSK activation could be an early step in the pathway leading to AChR phosphorylation. In this system, phosphorylation of the AChR β subunit depends on co-expression of both MuSK and rapsyn, raising the possibility that all three proteins are part of a complex involved in AChR clustering. To clarify the relationship between MuSK and AChR phosphorylation, we have investigated the association between the two proteins in unstimulated and agrin-stimulated mammalian C2 muscle myotubes. We find that MuSK is associated with the AChR and that receptor-bound MuSK rapidly becomes tyrosine phosphorylated after addition of neural agrin. Block of tyrosine phosphorylation of the AChR β subunit by the kinase inhibitor staurosporine can be achieved without affecting MuSK activation, however, suggesting that another tyrosine kinase is responsible for agrin-induced phosphorylation of the AChR. Results MuSK is associated with the AChR in C2 myotubes In our initial experiments, we sought to detect interaction between MuSK and the AChR in C2 myotubes using co-precipitation experiments. For this purpose, we used a polyclonal antiserum raised against the C-terminus of MuSK that specifically recognized mouse MuSK on immunoblots (Figure 1A and B) and in immunoprecipitations (see Figure 4C). As shown previously by Glass et al. (1996), MuSK migrated as a band, or sometimes a doublet, with an apparent molecular weight of ∼110 kDa. To test for its association with the AChR, we first isolated the AChR from C2 myotube extracts with α-bungarotoxin covalently coupled to Sepharose beads. We then looked for the presence of MuSK by stripping the beads with SDS sample buffer, separating the proteins by SDS–PAGE and immunoblotting with the MuSK-specific antibody. The immunoblots show that MuSK was present on the beads (Figure 1C); its presence required the AChR, since when AChR precipitation was prevented by the addition of excess α-bungarotoxin to the C2 extract prior to incubation with the beads, no MuSK was detected. MuSK was also not detected when unconjugated Sepharose beads were used as a control. In each case, the presence or absence of the AChR on the beads was confirmed by stripping the blots and reprobing with an antibody that recognizes the β subunit of the AChR. By comparing the content of MuSK in AChR isolates with the content in the initial lysate, we estimate that 0.29 ± 0.02% (mean ± SD, data from three experiments) of the total MuSK in the cell lysates is bound to the toxin–Sepharose beads. A similar comparison for the AChR (Figure 1F) shows that 14.9 ± 4.7% of the total AChR in C2 extracts can be recovered on the beads. Taken together, these values suggest that at least 2% of the total MuSK in C2 myotubes is associated with the AChR. Figure 1.Association of MuSK with the AChR in C2 myotubes. (A and B) Characterization of MuSK antibodies. Extracts made from C2 myotubes or COS cells transfected with a mouse MuSK expression vector were analyzed by immunoblotting with pre-immune and immune sera as indicated. (C) Co-isolation of MuSK with AChRs. C2 extracts were incubated with α-bungarotoxin coupled to Sepharose beads (TS) to isolate the AChR. As controls, 10 μM free toxin was added to compete for receptor binding to toxin–Sepharose (+T), or non-conjugated Sepharose beads were used (CS). AChR-associated MuSK was analyzed by SDS–PAGE and immunoblotting with MuSK-specific antibodies. A fraction (0.3%) of the initial lysate was loaded as a standard (L). For the lower panel, the blot was stripped and reprobed with mAb 124, which is reactive with the AChR β subunit. (D) Co-immunoprecipitation of the AChR with MuSK. C2 myotube extracts were incubated with the indicated amounts of MuSK antibodies, and the presence of AChRs in the immunoprecipitates was assessed by non-reducing SDS–PAGE and mAb 124 immunoblotting. Controls included a portion (0.5%) of the initial lysate (L) as well as an incubation of MuSK antibodies with blank lysis buffer (−lysate sample). (E) The AChR is not associated with the transferrin receptor. Toxin–Sepharose precipitations of the AChR were analyzed by immunoblotting using antibodies reactive with the transferrin receptor. L represents 0.3% of the initial lysate, F the efficiency of AChR recovery using toxin–Sepharose. After precipitation with toxin–beads, the indicated percentages of samples were analyzed by mAb 124 immunoblotting. For comparison, fractions of the initial lysate were loaded. Download figure Download PowerPoint Several experiments were carried out to determine whether the co-precipitation of MuSK with the AChR represents a specific association rather than non-specific trapping of membrane proteins. First, the extent of MuSK co-precipitation using toxin–beads was constant and reproducible over several experiments. Second, the association of MuSK with the AChR isolated on toxin–beads was not significantly changed by cell lysis and washing under stringent conditions of high salt (0.5 M NaCl) and detergent (1% Triton X-100 and 0.5% deoxycholate; data not shown). Such conditions resulted in the loss of other proteins such as β-dystroglycan (C.Fuhrer, J.E.Sugiyama and Z.W.Hall, unpublished observations), a membrane component of the dystrophin–utrophin glycoprotein complex that is found at synapses (Matsumura et al., 1992). Third, as a measure of non-specific binding of membrane proteins, we analyzed preparations of AChRs isolated by toxin–beads for the presence of the transferrin receptor, a prominent membrane protein found at the plasma membrane and in endosomal compartments (Ralston and Ploug, 1996). In an immunoblot analysis with specific monoclonal antibodies, no detectable transferrin receptor was seen in AChR precipitates (Figure 1E). Finally, we tested for association of MuSK with the AChR by immunoprecipitating MuSK and looking for the presence of the AChR by immunoblotting with the antiserum against the AChR β subunit. Although some non-specific binding to protein A–Sepharose occurred, the AChR was clearly detected in the MuSK immunoprecipitates (Figure 1D). Taken together, we believe that association of MuSK with the AChR is specific and that the two proteins form a stable complex which excludes other membrane proteins. Agrin increases the association between MuSK and the AChR We then used the co-precipitation assays to examine the effects of agrin on the association of MuSK with the AChR. C2 myotubes were incubated for increasing times with 5 nM of the soluble C-terminal fragment of either the most active, neural-specific (C-Ag12,4,8) or the predominant muscle (C-Ag12,0,0) isoform of agrin (Ferns et al., 1993). We examined the association of MuSK and the AChR at three times after the beginning of the treatment: 40 and 90 min, when agrin-induced tyrosine phosphorylation of the AChR β subunit is near its peak, but few clusters are seen, and 16 h, when AChR phosphorylation has declined, but cluster formation is maximal (Ferns et al., 1996). As shown by both α-bungarotoxin–Sepharose precipitation followed by MuSK immunoblotting and by MuSK immunoprecipitation followed by AChR Western blotting, agrin increases the interaction between MuSK and the AChR (Figure 2A and B). This effect is specific for neural, but not muscle, agrin and is seen after a 16 h treatment; quantitation shows that the increase is ∼2.5-fold and is highly significant (Figure 2C). Agrin's effects were most pronounced at a concentration of 5 nM; we observed smaller changes at lower amounts (0.2 nM; data not shown). In control experiments, the amounts of AChRs were the same in all lanes; hence agrin did not affect the level of the AChRs in C2 cells (Figure 2A, bottom; see also Jones et al., 1996). Figure 2.Neural agrin leads to increased association of MuSK with the AChR. C2 myotubes were stimulated for the indicated times with 5 nM neural (4,8) or muscle (0,0) agrin. (A) Cell extracts were incubated with α-bungarotoxin–Sepharose beads to isolate AChRs, and receptor-associated MuSK was visualized by SDS–PAGE and MuSK immunoblotting. In a control sample, 10 μM free toxin was added to block AChR binding to the beads (+T). To confirm that equal amounts of receptor are present in the appropriate lanes, the blot was stripped and reprobed with mAb 88B, which is specific for the γ and δ AChR subunits (lower panel). (B) Agrin-treated cells were lysed and subjected to immunoprecipitation with MuSK antiserum followed by immunoblotting with mAb 124. As controls, we omitted MuSK antibodies (−ab) or the lysate (−L). We also analyzed a fraction (0.3% in A, 0.1% in B) of the total C2 cellular extracts as a standard (L). (C) Quantitation of the effect of neural agrin on the amount of AChR-associated MuSK. Experiments as described in (A) were quantitated by densitometric scanning of films, and values of untreated cells were set to 100%. Data represent mean ± SD of four experiments. * Differs significantly from control, P <0.01 (by ANOVA followed by Bonferroni's t-test). Download figure Download PowerPoint Agrin-induced activation of AChR-associated MuSK correlates with phosphorylation of the AChR β subunit To determine whether AChR-associated MuSK is activated by agrin, we treated C2 myotubes for increasing times with 0.1 nM neural or muscle agrin. AChRs were then isolated with α-bungarotoxin–Sepharose beads and the associated proteins examined by immunoblotting with phosphotyrosine-specific antibodies (Figure 3). After 15 min of treatment with neural agrin, a doublet of immunoreactive bands was detected that migrated at ∼110 kDa. Immunoreactivity remained elevated at 40 and 90 min but, after 5 and 15 h, the tyrosine-phosphorylated bands were no longer detected. By stripping and reprobing the blot with MuSK-specific antibodies, we identified this doublet as AChR-associated MuSK. Phosphotyrosine-immunoreactive bands were not observed when free excess toxin was added to the C2 extracts prior to AChR isolation, when unconjugated Sepharose was used or when the cells were treated with muscle agrin. Figure 3.Neural agrin causes a rapid and transient activation of AChR-associated MuSK. Cells were treated for increasing times with 0.1 nM neural (4,8) or muscle (0,0) agrin. AChRs were isolated with α-bungarotoxin–Sepharose precipitation and activation of associated MuSK was visualized by phosphotyrosine immunoblotting. A doublet of bands with the same molecular weight as MuSK (∼110 kDa) becomes transiently tyrosine phosphorylated, with a peak at ∼40 min. Controls included addition of 10 μM free toxin (+T) and the use of non-conjugated control Sepharose (CS). The identity of the ∼110 kDa doublet (MuSK) was confirmed by stripping and reprobing the blot with MuSK antibodies (lower panel). Download figure Download PowerPoint Figure 4.Dose dependence and time course of agrin-induced MuSK activation and AChR β subunit phosphorylation. (A and B) C2 myotubes were treated for increasing times with the indicated concentrations of neural or muscle agrin. After lysis, cell extracts were split into two parts; the first was incubated with α-bungarotoxin–Sepharose beads to isolate AChRs (top and middle panels), whereas the second part was subjected to immunoprecipitation with MuSK-specific antibodies (bottom panels). Both precipitates were analyzed by SDS–PAGE and phosphotyrosine immunoblotting. Protein bands were identified by their molecular weight and by stripping and reprobing blots with the appropriate antisera (not shown). Agrin-induced phosphorylation of AChR-associated MuSK, AChR β subunits and of the total cellular pool of MuSK correlate over a wide range of agrin concentrations. (C) Controls for the specificity of MuSK immunoprecipitation. Cells were stimulated for 15 min with 15 nM of neural agrin. Lysates were subjected to immunoprecipitation with pre-immune or immune anti-MuSK sera with or without the addition of an excess (2.5 mg/ml) of free immunizing MuSK peptide. Samples were analyzed by phosphotyrosine immunoblotting. Download figure Download PowerPoint We then used this system to compare three phosphorylation events induced by agrin: activation of the total cellular pool of MuSK, activation of AChR-associated MuSK and phosphorylation of the AChR β subunit. For these experiments, C2 myotubes were treated for increasing times with a range of neural and muscle agrin concentrations. Cell extracts were then split into two parts; one part was incubated with α-bungarotoxin–Sepharose beads to isolate AChRs, whereas the other part was subjected to immunoprecipitation with MuSK-specific antibodies. Both precipitations were analyzed by phosphotyrosine immunoblotting. We first examined the agrin concentration dependence of the phosphorylations. As shown in Figure 4A, 5 nM of neural agrin caused phosphorylation of AChR-associated MuSK, of the AChR β subunit and of total MuSK within as little as 5 min, whereas in each case muscle agrin had no effect. At lower concentrations of agrin, all three phosphorylation events were visible after 40 min, but not after 5 min, of treatment. A slight reduction at that time in the extent of phosphorylation of AChR-bound MuSK was accompanied by decreased phosphorylation of both the AChR β subunit and total MuSK. Finally, at the lowest concentration of agrin tested, none of the three phosphorylation events was detected. We then performed a time course of agrin treatment, using two concentrations of neural agrin. Activation of both AChR-associated MuSK and of total MuSK was maximal after 40 min of treatment with 5 or 0.1 nM neural agrin, and decreased to close to background levels after a 15 h incubation (Figure 4B). In contrast, phosphorylation of the AChR β subunit followed a similar initial time course, but showed a slower decline and was still visible after 15 h. We noted a difference in the time course of MuSK activation at 5 nM as compared with 0.1 nM agrin: the high concentration induced a rapid activation of MuSK within 5 min, whereas at the low concentration, activation occurred with slower kinetics, as it was only detected after 15 min and reached a peak after 40 min. Control experiments confirmed the specificity of immunoprecipitation with MuSK antibodies; thus no agrin-induced tyrosine phosphorylation of MuSK was seen when using the pre-immune serum or when adding an excess of free MuSK C-terminal immunizing peptide to MuSK immunoprecipitations (Figure 4C). In summary, the phosphorylation events induced by agrin are correlated: phosphorylation of AChR-associated MuSK shows a similar dependence on agrin concentration and initial time course as does phosphorylation of total MuSK and of the AChR; dephosphorylation of both total and AChR-bound MuSK, however, occurs more rapidly than that of the AChR. Staurosporine blocks agrin-induced AChR β subunit phosphorylation, but not activation of AChR-associated MuSK To investigate further the relationship between MuSK activation and AChR phosphorylation, we used the tyrosine kinase inhibitors, herbimycin and staurosporine, that were shown previously to block agrin-induced AChR clustering and tyrosine phosphorylation of the AChR β subunit (Wallace, 1994; Ferns et al., 1996). C2 myotubes were incubated with either 1 μM herbimycin or 10 nM staurosporine for 6 h before adding agrin, conditions that have been shown to be optimal for inhibition of AChR clustering and phosphorylation. Cell extracts were then prepared and subjected to α-bungarotoxin–Sepharose or MuSK immunoprecipitation followed by phosphotyrosine immunoblotting. Under all conditions tested, herbimycin blocked activation of both AChR-bound and total MuSK as well as phosphorylation of AChR β subunits. In contrast, staurosporine had no effect on agrin-induced activation of MuSK, either for the total pool or for that associated with the AChR, but strongly decreased phosphorylation of the β subunit of the AChR, as previously reported (Figure 5). When cells were treated with increasing concentrations of staurosporine, ranging from 1 to 20 nM, agrin-stimulated phosphorylation of the AChR β subunit was progressively inhibited, but no effect was seen on activation of MuSK (Figure 6). Most dramatically, at 20 nM of staurosporine, agrin-induced AChR phosphorylation was completely abolished, whereas activation of neither total nor AChR-bound MuSK was significantly altered. These experiments show that staurosporine does not affect agrin-induced MuSK activation, but blocks phosphorylation of the β subunit of the AChR caused by agrin. We have thus found conditions that uncouple activation of MuSK from AChR phosphorylation, suggesting that MuSK does not directly phosphorylate the AChR β subunit. As we have reported previously that src can phosphorylate the AChR, and that AChRs isolated from C2 myotubes are associated with src-related kinases, we examined the effects of agrin on association of fyn and src with the AChR. As shown in Figure 7, treatment of cells with 2.5 nM of neural agrin did not change the relative amounts of either kinase bound to the AChR. In addition, neural agrin (0.1 or 2.5 nM) did not affect the overall level of tyrosine phosphorylation of src and fyn, as measured by kinase immunoprecipitation using specific antibodies, followed by phosphotyrosine immunoblotting. Figure 5.Effects of herbimycin and staurosporine on agrin-induced phosphorylation of MuSK and the AChR β subunit. C2 myotubes were pre-treated for 6 h with 1 μM herbimycin (H), 10 nM staurosporine (S) or a DMSO carrier control (c). Cells were then incubated with the indicated amounts of agrin for 5 or 40 min, and cells extracts were analyzed by precipitation with α-bungarotoxin–Sepharose beads followed by phosphotyrosine immunoblotting (top and middle). In some cases, parallel dishes of C2 myotubes were treated with agrin (5 and 40 min with 1 and 0.1 nM of agrin, respectively), and analyzed by precipitation with MuSK antibodies followed by phosphotyrosine immunoblotting (bottom). Herbimycin blocks all phosphorylation events, whereas staurosporine leaves MuSK activation intact, but strongly reduces AChR β subunit phosphorylation. Download figure Download PowerPoint Figure 6.Staurosporine does not interfere with agrin-induced MuSK activation, but blocks agrin-induced phosphorylation of the AChR β subunit. (A) Myotubes were pre-incubated for 5 h with the indicated concentrations of staurosporine or a carrier control (c) and then stimulated for 40 min with 1 nM neural agrin. Lysates were processed by α-bungarotoxin–Sepharose or MuSK antibody precipitation and analyzed by phosphotyrosine immunoblotting. Increasing the concentration of staurosporine up to 20 nM progressively reduces agrin-induced AChR β subunit phosphorylation without substantially altering MuSK phosphorylation. (B) Quantitation of staurosporine's effects by densitometric scanning. Data from four experiments as described in (A) were evaluated and are shown as mean ± SD. Phosphorylation signals from cells not treated with agrin were set to 0%. Agrin-induced control signals of cells incubated with carrier, lacking staurosporine, were set to 100%. All phosphorylation signals were normalized for the AChR contents of cells as visualized by reprobing blots containing toxin precipitates with mAb 88B. Download figure Download PowerPoint Figure 7.Agrin does not change the amount of fyn and src associated with the AChR, nor does it affect the overall phosphotyrosine level of total src and fyn. (A) C2 myotubes were incubated for up to 15 h with 2.5 nM of neural agrin. Cell lysates were analyzed by precipitation with α-bungarotoxin–Sepharose beads and immunoblotting with antibodies specific for fyn or src; 0.1% of the total lysate was included as a control (L). (B) C2 cells treated with neural or muscle agrin as indicated were subjected to immunoprecipitation with fyn- or src-specific antibodies followed by phosphotyrosine immunoblotting. Over the course of several experiments, the signals shown in (A) and (B) displayed some variation; however, no significant agrin effect was seen. The figure shows one representative experiment for each treatment (A and B). Download figure Download PowerPoint Discussion The principal result of this work is the demonstration that MuSK, the receptor tyrosine kinase that is thought to mediate the effects of agrin in clustering AChRs in muscle cells, is closely associated with the AChR in C2 myotubes, even in the absence of neural agrin. Treatment of the myotubes with agrin results in tyrosine phosphorylation of the AChR-associated MuSK in a manner that is similar to agrin-induced tyrosine phosphorylation of the AChR β subunit. M
Referência(s)