The Muscle Protein Dok-7 Is Essential for Neuromuscular Synaptogenesis

See allHide authors and affiliations

Science  23 Jun 2006:
Vol. 312, Issue 5781, pp. 1802-1805
DOI: 10.1126/science.1127142


The formation of the neuromuscular synapse requires muscle-specific receptor kinase (MuSK) to orchestrate postsynaptic differentiation, including the clustering of receptors for the neurotransmitter acetylcholine. Upon innervation, neural agrin activates MuSK to establish the postsynaptic apparatus, although agrin-independent formation of neuromuscular synapses can also occur experimentally in the absence of neurotransmission. Dok-7, a MuSK-interacting cytoplasmic protein, is essential for MuSK activation in cultured myotubes; in particular, the Dok-7 phosphotyrosine-binding domain and its target in MuSK are indispensable. Mice lacking Dok-7 formed neither acetylcholine receptor clusters nor neuromuscular synapses. Thus, Dok-7 is essential for neuromuscular synaptogenesis through its interaction with MuSK.

Skeletal muscle is controlled by motor neurons, which contact the muscle at the neuromuscular junction, a synapse that uses the neurotransmitter acetylcholine (1, 2). To achieve sufficient sensitivity to the neurotransmitter, acetylcholine receptors (AChRs) on the muscle must be densely clustered on the postsynaptic side of the neuromuscular junction (1, 2). Failure of AChR clustering is associated with disorders in neuromuscular transmission, including congenital myasthenic syndrome and myasthenia gravis (3, 4). The presynaptic motor-nerve terminal secretes the glycoprotein agrin to activate postsynaptic MuSK (5). This agrin-dependent activation of MuSK is essential to establish the postsynaptic apparatus, including the clustering of AChRs, via the AChR-associated protein Rapsyn (68). Nevertheless, before innervation, MuSK-dependent AChR clusters can form at the endplate area of myotubes, suggesting a mechanism of postsynaptic specialization that is independent of agrin and innervation (911). Furthermore, neuromuscular synapses can form independently of agrin in mice that lack acetylcholine, which appears to antagonize postsynaptic differentiation (12, 13). Thus, in addition to agrin, there may be another element that can achieve MuSK activation and trigger postsynaptic specializations at the neuromuscular junction. MuSK contains a phosphotyrosine-binding domain (PTB domain) target motif Asn-Pro-X-Tyr encompassing Tyr553 in the juxtamembrane region, which is essential for proper functioning in vivo (14). The binding partner for this motif has remained elusive.

By searching databases, including GenBank, the European Molecular Biology Laboratory, and the DNA Data Bank of Japan, for a previously unidentified member of the Dok-family of proteins, each of which has a PTB domain, we identified Dok-7 and cloned human cDNA encoding 504 amino acids. Like other members, Dok-7 has pleckstrin-homology (PH) and PTB domains in the N-terminal portion and Src homology 2 (SH2) domain target motifs in the C-terminal region (fig. S1) (1517). Cloning of mouse (Mus musculus) and puffer fish (Takifugu rubripes) Dok-7 cDNA revealed a highly conserved structure (fig. S2). Like agrin and MuSK, no ortholog was found in invertebrates such as the fruit fly (Drosophila melanogaster) and nematode (Caenorhabditis elegans). Northern blot analysis of human tissues showed that Dok-7 mRNA is preferentially expressed in skeletal muscle and in the heart (fig. S3A), and immunoblot analysis identified a 55-kD Dok-7 protein in the thigh muscle, diaphragm, and heart but not in the liver or spleen (fig. S3B). Furthermore, immunostaining of mouse skeletal muscles, including the sternocleidomastoid, extensor digitorum longus, and gastrocnemius, with antiserum to Dok-7 highlighted the accumulation of Dok-7 at neuromuscular junctions (Fig. 1, A to C), which are composed of the postsynaptic membrane with its densely clustered AChRs in close juxtaposition with the presynaptic nerve terminal. Therefore, we denervated a mouse gastrocnemius muscle by sciatic nerve resection to confirm the muscular, and thus postsynaptic, localization of Dok-7. One week after the operation, synaptophysin, a component of the presynaptic vesicle, was completely abolished in denervated muscles (fig. S4). However, the muscular localization of Dok-7 and AChRs remained intact, indicating a postsynaptic localization of Dok-7 at neuromuscular junctions (Fig. 1, D to F). Because postsynaptic differentiation and neuromuscular synapse formation are initiated at the endplate zone of skeletal muscle during embryogenesis (911), we performed a whole-mount in situ hybridization and found that Dok-7 transcripts are expressed in the central region encompassing the endplate area of the diaphragm muscles at day 14.5 of embryonic development (E14.5), when AChRs cluster in a nerve- and agrin-independent manner (fig. S5). Together, these results suggest that Dok-7 has the appropriate distribution to be involved in the neuromuscular junction.

Fig. 1.

Forced expression of the muscle protein Dok-7 activates MuSK and induces AChR clustering. (A to F) Postsynaptic localization of Dok-7 at neuromuscular junction. Dok-7 and AChR were visualized with antibodies (αDok-7) and α-bungarotoxin, respectively, at an innervated (IN) or denervated (DN) neuromuscular junction. Scale bars, 20 μm. (G) Dok-7 induces autophosphorylation of MuSK. Whole-cell lysates (WCL) or anti-Myc immunoprecipitates (IP: αmyc) prepared from 293T cells transfected with plasmids expressing Dok-7 and either Myc-tagged MuSK (WT) or MuSK-KA (KA) were subjected to immunoblotting (IB). PY, phosphotyrosine. (H) Forced expression of Dok-7 activates the MuSK pathway. Anti-MuSK IP, α-bungarotoxin precipitates (BP), or WCL from C2 myotubes transfected with plasmids for Dok-7 were subjected to IB. (I and J) Forced expression of Dok-7 induces aneural AChR clustering in C2 myotubes. Abundant clusters of AChRs formed in C2 myotubes transfected with Dok-7 expression plasmids (J), but only a few small clusters formed in the control (Mock) (I). Scale bars, 200 μm.

Given the requirement for MuSK's PTB target motif and presumably its binding partner in postsynaptic specialization (14, 18, 19), we next examined the interaction of MuSK with Dok-7, which has a PTB domain, in 293T cells. These heterologous cells do not express either protein detectably, and forced expression of MuSK in these cells induced weak autophosphorylation (20, 21). Forced expression of Dok-7 induced an intense tyrosine phosphorylation of MuSK but not the kinase-inactive mutant with a Lys/Ala substitution (MuSK-KA), indicating that Dok-7 induced the autophosphorylation of MuSK (Fig. 1G). This activity was unique to Dok-7; no other mammalian Dok-family proteins induced phosphorylation of MuSK (fig. S6). It was also conserved; Dok-7 from puffer fish was able to activate even mammalian MuSK. Also, in C2 myotubes, the forced expression of Dok-7 induced tyrosine phosphorylation of MuSK and the β subunit (AChRβ1) of the AChR complex, which is known to be tyrosine-phosphorylated upon activation of MuSK (22) (Fig. 1H). Furthermore, this forced expression induced numerous clusters of AChRs, and the number of AChR clusters correlated with the amount of Dok-7 expression plasmid (Fig. 1, I and J; fig. S7A and supporting online material). The exogenous Dok-7–induced AChR clusters were elaborately branched, and their complicated architecture resembled the differentiated “pretzel-like” AChR clusters formed in vivo (fig. S7, B and C). In addition, forced expression in myotubes of Dok-7 that had been fused with enhanced green fluorescent protein (EGFP) induced Dok-7 and AChR coclustering (fig. S7, D to I), as observed at postsynaptic areas in vivo (Fig. 1, C and F).

Because the regulatory interaction of Dok-7 with MuSK as described above implies their physical interaction, we examined whether Dok-7 binds to MuSK by way of the PTB domain in 293T cells. MuSK was coimmunoprecipitated with Dok-7 but not with Dok-7 carrying three Arg/Ala substitutions (Dok-7–RA) in the PTB domain (Fig. 2A). Consistently, the mutant MuSK carrying either a Tyr/Phe substitution at Tyr553 (MuSK-YF) or an Asn/Ala substitution at Asn550 (MuSK-NA) in the PTB target motif was not coimmunoprecipitated with Dok-7 (fig. S9 and Fig. 2B). The failure of the MuSK-KA kinase-inactive mutant to be coimmunoprecipitated with Dok-7 confirms the requirement of tyrosine phosphorylation for the binding of MuSK with Dok-7 via the PTB domain. These results indicate that Dok-7 binds to MuSK through the PTB domain in a manner dependent on the tyrosine phosphorylation of its target motif in MuSK.

Fig. 2.

Dok-7 interacts with MuSK by way of the PTB domain. (A and B) The PTB domain, its target, and kinase activity are essential for Dok-7 binding to MuSK. Anti-Myc IP or WCL from 293T cells transfected with plasmids for Dok-7 and Myc-tagged MuSK or their mutants, including MuSK-KA, were subjected to IB. (C and D) The PTB domain and C-terminal region are indispensable for the Dok-7–induced activation of MuSK and AChR clustering in fully differentiated C2 myotubes. Anti-MuSK IP or WCL from C2 cells transfected with expression plasmids for Dok-7 (WT), Dok-7–ΔC (ΔC), or Dok-7–RA (RA) were prepared at day 3 or 6 upon differentiation into myotubes and subjected to IB (C). The number of AChR clusters (mean ± SD) counted at day 7 is shown (D). Differentiation was achieved by day 6, whereas only a few myotubes had formed by day 3.

Nevertheless, mutations in the PTB domain (Dok-7–RA) or PTB target motif (MuSK-NA or -YF) did not block activation of MuSK, at least in heterologous cells (Fig. 2, A and B). In addition, the N- and C-terminal deletion mutants of Dok-7 (Dok-7-ΔN and -ΔC) revealed that the C-terminal moiety, but not the PH domain, of Dok-7 is dispensable for MuSK activation in heterologous cells (fig. S10). Also, the forced expression of Dok-7–RA or Dok-7–ΔC induced MuSK activation even in C2 cells at day 3 of differentiation into myotubes (Fig. 2C), when very few myotubes have formed. Unexpectedly, however, the PTB domain and C-terminal portion were indispensable for Dok-7–induced MuSK activation and AChR clustering in fully differentiated C2 myotubes at days 6 and 7 of differentiation (Fig. 2, C and D). In addition, the PH domain, responsible for membrane localization in general, was indispensable for the activation of MuSK in fully differentiated myotubes (fig. S11), as was seen in heterologous cells (fig. S10). Together these findings suggest that a negative regulatory mechanism preventing MuSK activation is established upon differentiation into myotubes, which is accompanied by increased expression of MuSK and Dok-7 (fig. S12). Trace phosphorylation of MuSK in myotubes might allow physical interaction with Dok-7, in turn facilitating dimerization and/or conformational changes in MuSK that are necessary for its sustained activation.

MuSK-deficient myotubes do not form agrin-dependent or -independent clusters of AChRs unless MuSK is reintroduced (18, 19, 23). To confirm whether Dok-7–mediated AChR clustering is dependent on MuSK, we introduced Dok-7 into MuSK-deficient myotubes. Unlike its effect in C2 myotubes, forced expression of Dok-7 induced no AChR clustering in the MuSK-deficient myotubes; however, additional expression of wild-type MuSK resulted in robust clustering of AChRs in these cells (Fig. 3A). Furthermore, the MuSK-KA and MuSK-YF mutant each failed to complement the MuSK deficiency, regardless of exogenous Dok-7. These findings demonstrate that Dok-7–induced AChR clustering in myotubes depends on Dok-7 interaction with MuSK and subsequent activation of MuSK catalytic activity. Thus, we examined the regulatory interaction of Dok-7 with a MuSK mutant [MuSK-Val/Met (MuSK-VM)] that carries a Val790 to Met substitution. This mutation is causally associated with the congenital myasthenic syndrome by way of an as yet unclear mechanism (24). As observed with MuSK-YF (Fig. 2B), forced expression of Dok-7 in 293T cells induced the autophosphorylation of MuSK-VM, but its coimmunoprecipitation with Dok-7 was barely detectable in these heterologous cells (fig. S13). Forced expression of Dok-7 with MuSK-VM induced only very weak AChR clustering in MuSK-deficient myotubes (Fig. 3A). Therefore, the congenital myasthenic syndrome–associated Val790 to Met mutation impaired interaction of MuSK with Dok-7, suggesting a possible cause of neuromuscular junction dysfunction in these patients.

Fig. 3.

Dok-7 is essential for activation of the MuSK-pathway to AChR clustering in myotubes. (A) MuSK is required for Dok-7–induced AChR clustering. MuSK-deficient myotubes were transfected with the indicated plasmids. The number of AChR clusters (mean ± SD) per EGFP-positive myotube is shown. MuSK-VM is a congenital myasthenic syndrome–associated mutant. (B) Activation of the MuSK pathway requires Dok-7. C2 myotubes transfected with Dok-7 siRNA (siD7) or the control (Ctrl) without (–) or with (+) agrin treatment for 15 min were studied as in Fig. 1H. Both short [10 s (10″)] and long [1 min (1′)] exposures are shown for the anti-PY IB of the anti-MuSK IP. (C) Dok-7 is essential for AChR clustering. C2 myotubes were transfected with Dok-7 siRNA (siD7) or the control (Ctrl) with or without agrin treatment for 12 hours. The number of AChR clusters (mean ± SD) is shown.

To examine the effects of Dok-7 down-regulation in myotubes, we used a small interfering RNA (siRNA) designed specifically to block its expression. Inhibition of Dok-7 suppressed the tyrosine phosphorylation of MuSK and AChRβ1 in C2 myotubes, demonstrating its essential role in the aneural, basal catalytic activity of MuSK (Fig. 3B). Indeed, MuSK-dependent spontaneous AChR clustering was suppressed by this siRNA-mediated inhibition (Fig. 3C). Moreover, the inhibition of Dok-7 impaired the agrin-dependent activation of MuSK, the phosphorylation of AChRβ1, and the subsequent formation of AChR clusters (Fig. 3, B and C). Thus, we conclude that Dok-7 is essential for aneural activation of MuSK and AChR clustering in myotubes and is also crucial for agrin-dependent activation of MuSK and AChR clustering. Nonetheless, our results do not exclude the possibility that Dok-7 might also play a role downstream of MuSK. Indeed, Dok-7 and MuSK were synchronously tyrosine phosphorylated upon treatment of myotubes with agrin (fig. S14).

We generated mice lacking Dok-7 to explore its role in vivo (fig. S15). Like mice lacking MuSK or agrin (6, 7), all Dok-7–deficient (Dok-7–/–) mice were immobile at birth and died shortly thereafter (26 homozygotes were observed among the first 137 pups), although their wild-type and heterozygous littermates appeared normal. Also, the alveoli of the mutant mice were not expanded at birth (fig. S15D), indicating a failure to breathe and suggesting a severe defect in neuromuscular transmission in the skeletal muscles. Consistently, there were no detectable AChR clusters in the endplate area of the diaphragm muscle in Dok-7–/– embryos at either E14.5 or E18.5 (Fig. 4, E and K). Because nascent AChR clusters are formed in a nerve- and agrin-independent manner at E13.5 to E16.5, whereas most neuromuscular junctions are formed in a nerve- and agrin-dependent manner at E18.5, our findings indicate a requirement for Dok-7 in both types of MuSK-dependent postsynaptic specialization, although we cannot exclude the possibility that nascent AChR clustering is a prerequisite for nerve- and agrin-dependent AChR clustering (911). Consistent with this finding, Dok-7 transcripts were expressed in the endplate area of the diaphragm muscle (fig. S5). In addition, axonal branches extending from the motor nerve trunk were aberrantly long in the endplate area of Dok-7–/– diaphragms at E18.5 and, unlike the controls, did not terminate near the nerve trunk (Fig. 4, G and J). Overall, these pre- and postsynaptic abnormalities are indistinguishable from those found in mice lacking MuSK (7), demonstrating an essential role in vivo for Dok-7 in neuromuscular synaptogenesis, a MuSK-dependent vital process.

Fig. 4.

Dok-7 is essential for neuromuscular synaptogenesis in vivo. Diaphragm muscles were prepared from the wild-type control (WT) or Dok-7–/– embryos at E14.5 (A to F) or E18.5 (G to L) and subjected to whole-mount anti-neurofilament and α-bungarotoxin staining, to visualize nerve and AChR, respectively. Scale bars, 100 μm.

MuSK-dependent postsynaptic specialization during neuromuscular synaptogenesis appears to be controlled by multiple regulatory mechanisms (2, 25). We have shown that Dok-7 may be a muscle-intrinsic activator of MuSK by demonstrating its essential role in the aneural activation of MuSK and subsequent AChR clustering in cultured myotubes. This conclusion is further supported by our findings that mice lacking Dok-7 showed marked disruption of neuromuscular synaptogenesis that was indistinguishable from the disruption found in MuSK-deficient mice. Thus, neuromuscular synaptogenesis requires Dok-7 within the skeletal muscle. Dok-7 dysfunction may be involved in the pathogenesis of neuromuscular junction disorders.

Supporting Online Material

Materials and Methods

SOM Text

Figs. S1 to S15


References and Notes

Stay Connected to Science

Navigate This Article