Dok-7 Mutations Underlie a Neuromuscular Junction Synaptopathy

See allHide authors and affiliations

Science  29 Sep 2006:
Vol. 313, Issue 5795, pp. 1975-1978
DOI: 10.1126/science.1130837


Congenital myasthenic syndromes (CMSs) are a group of inherited disorders of neuromuscular transmission characterized by fatigable muscle weakness. One major subgroup of patients shows a characteristic “limb girdle” pattern of muscle weakness, in which the muscles have small, simplified neuromuscular junctions but normal acetylcholine receptor and acetylcholinesterase function. We showed that recessive inheritance of mutations in Dok-7, which result in a defective structure of the neuromuscular junction, is a cause of CMS with proximal muscle weakness.

In congenital myasthenic syndromes (CMSs), the safety margin for neuromuscular transmission is compromised, resulting in characteristic myasthenic fatigable muscle weakness. A search of candidate genes has resulted in the identification of mutations in nine genes that encode proteins at the neuromuscular junction (NMJ) (1), but there are additional CMSs for which the underlying molecular defect remains to be defined (2). One of these syndromes causes a characteristic “limb girdle” pattern of muscle weakness. Because muscle biopsies have not found clear-cut defects in the functional components of signal transmission but revealed abnormally small NMJs (3), it was concluded that this disorder may arise from the defective formation or maintenance of the synaptic structure of the NMJ.

Studies, mainly with knock-out mice, have elucidated key processes in the initiation and maintenance of the specialized postsynaptic structures at the NMJ (4). Agrin, which is released from the nerve terminal, activates muscle-specific tyrosine kinase (MuSK) located in the postsynaptic membrane, which leads to the precise aggregation and localization of the acetylcholine receptors (AChRs) through their association with the cytoplasmic anchoring protein rapsyn. However, additional key components contribute to this pathway (57). Dok-7, a member of the Dok family of cytoplasmic molecules, can induce the aneural activation of MuSK and the subsequent clustering of AChR in cultured myotubes (8).

We screened genomic DNA from patients with suspected CMS for mutations within the seven coding exons of the human dok-7 gene by amplification using polymerase chain reaction (PCR) and bidirectional DNA sequencing [see supporting online material (SOM)], and we identified frameshift mutations in 16 unrelated patients. In three additional patients, a frameshift mutation was identified in combination with a nonsense mutation, a splice site mutation, and a missense change of a conserved residue, Gly180→Ala180 (G180A), within the phosphotyrosine binding domain (PTB) (Fig. 1A and Table 1). In two other patients, we identified a frameshift mutation but have yet to identify a second mutation. Sixteen of the 21 patients that had frameshift mutations harbored 1124_1127dupTGCC, indicating that this was a common mutation in this group of patients and that allele-specific PCR can facilitate genetic screening. When DNA from family members was available, we observed that the disease cosegregated with recessive inheritance of the Dok-7 frameshift mutations (Fig. 1B). Sequencing of the entire coding region of Dok-7 from 25 control individuals revealed no frameshift mutations, and allele-specific PCR did not detect 1124_1127dupTGCC in 204 control chromosomes. Similarly, BsaH I restriction endonuclease digests of amplicons of exon 5 showed the segregation of G180A with recessive disease inheritance and did not detect G180A in 200 control chromosomes, suggesting that this mutation is also likely to be pathogenic. When DNA from family members was not available, subcloning and sequencing of amplicons from genomic DNA, or cDNA derived from muscle biopsies, demonstrated that the mutations were on separate alleles. Previous screening of DNA from index cases 2 to 5, 8, 11, 20, and 21 did not detect mutations in the genes encoding the AChR subunits (CHRNA, CHRNB, CHRND, and CHRNE), RAPSN, COLQ, or CHAT, where defects are known to cause CMS. Thus, we identified at least 19 unrelated CMS patients in which disease was associated with recessive mutations in Dok-7.

Fig. 1.

Identification of mutations in Dok-7. (A) Position of mutations (arrows) in the functional domains of Dok-7. PH, pleckstrin homology domain. The C-terminal domain contains the target motifs of the Src homology 2 domain. (B) Segregation of disease with recessive inheritance of mutant alleles. Colored traces show DNA sequencing electrophoretograms of respective frameshift mutations. Mutation 1124_1127dupTGCC is labeled as 1127dupTGCC and mutation 1339_1342dupCTGG as 1342dupCTGG. AS PCR, allele-specific PCR for 1127dupTGCC. The presence (plus sign) or absence (minus sign) of a frameshift mutation was detected by bidirectional DNA sequencing as shown in the electrophoretograms; BsaH I restriction digests were used to detect 539G>C in exon 5. Affected family members (black symbols) carry two mutant alleles. Unaffected members harbor one (half-black symbols) or no (white symbols) mutant alleles. F1, F3, F7, and F11 refer to the respective families of patients in Table 1. Arrows indicate the index cases.

Table 1.

Mutations in Dok-7 identified in CMS patients. c., cDNA nucleotide numbering from the adenine of the initiating ATG codon; p., protein sequence; fs, frameshift; IVS, intron; n.d., not determined; dash, mutation not yet identified. The number that comes after the “X” indicates the number of frameshift missense amino acids that are present before a nonsense codon.

Index caseEthnic originIdentified mutationsAlteration in coding sequenceDok-7 exon
1 UK, white c.1124_1127dupTGCC p.Pro376ProfsX30 7
c.1339_1342dupCTGG p.Gly447AlafsX70 7
2 UK, white c.1124_1127dupTGCC p.Pro376ProfsX30 7
c.1124_1127dupTGCC p.Pro376ProfsX30 7
3 UK, white c.1124_1127dupTGCC p.Pro376ProfsX30 7
c.1263insC p.Ser422LeufsX94 7
4 UK, white c.1124_1127dupTGCC p.Pro376ProfsX30 7
c.1263insC p.Ser422LeufsX94 7
5 UK, white c.1124_1127dupTGCC p.Pro376ProfsX30 7
c.1263insC p.Ser422LeufsX94 7
6 UK, white c.1124_1127dupTGCC p.Pro376ProfsX30 7
c.1124_1127dupTGCC p.Pro376ProfsX30 7
7 Indian subcontinent c.548_551delTCCT p.Phe183CysfsX61 5
c.1124_1127dupTGCC p.Pro376ProfsX30 7
8 UK, white c.1143insC p.Glu382ArgfsX24 7
c.1143insC p.Glu382ArgfsX24 7
9 UK, white c.1143insC p.Glu382ArgfsX24 7
c.1339_1342dupCTGG p.Gly447AlafsX70 7
10 UK, white c.1124_1127dupTGCC p.Pro376ProfsX30 7
c.1339_1342dupCTGG p.Gly447AlafsX70 7
11 UK, white c.539G>C G180A 5
c.1124_1127dupTGCC p.Pro376ProfsX30 7
12 UK, white c.1124_1127dupTGCC p.Pro376ProfsX30 7
c.1124_1127dupTGCC p.Pro376ProfsX30 7
13 UK, white c.1124_1127dupTGCC p.Pro376ProfsX30 7
- - -
14 UK, white 1143insC p.Glu382ArgfsX24 7
- - -
15 Finnish c.1378insC p.Glu460ProfsX58 7
c.1508insC p.Pro503ProfsX15 7
16 French-Canadian c.1263insC p.Ser422LeufsX94 7
c.1124_1127dupTGCC p.Pro376ProfsX30 7
17 Portuguese c.1357_1370del14 p.Arg452ArgfsX61 7
c.1124_1127dupTGCC p.Pro376ProfsX30 7
18 Spanish IVS2-1G>T n.d. intron 2
c.1124_1127dupTGCC p.Pro376ProfsX30 7
19 German c.1124_1127dupTGCC p.Pro376ProfsX30 7
c.1124_1127dupTGCC p.Pro376ProfsX30 7
20 German c.601C>T p.R201X 5
c.1124_1127dupTGCC p.Pro376ProfsX30 7
21 Spanish c.1124_1127dupTGCC p.Pro376ProfsX30 7
c.1124_1127dupTGCC p.Pro376ProfsX30 7

Because mice lacking Dok-7 die shortly after birth (8), we determined how the allelic C-terminal truncations that were seen in the CMS patients affect Dok-7 function. Dok-7 harboring the common mutation 1124_1127dupTGCC was cloned into pcDNA3.1 (Fig. 2A) and coexpressed with Myc-tagged MuSK in human embryonic kidney (HEK) 293 cells. The mutant Dok-7 showed robust expression and induced phosphorylation of MuSK (fig. S1A). Pull-down assays with an antibody to Myc (anti-Myc) showed coimmunoprecipitation of MuSK and mutant Dok-7 (fig. S1B), indicating that the C-terminal truncation does not disrupt Dok-7–MuSK interaction in the HEK 293 cells. However, because the regulatory interaction of Dok-7 with MuSK is more sensitive to deletion mutations in Dok-7 in differentiated myotubes (8), we tested whether mutant Dok-7 affects MuSK function when overexpressed in C2C12 myotubes. Although the mutant Dok-7 again showed robust expression, immunoblots conducted after pull-down assays with anti-MuSK or α-bungarotoxin–Sepharose (which binds AChRs) showed reduced phosphorylation of MuSK (one-tailed paired t test, P = 0.037, three independent experiments) and of the muscle AChR β subunit (P = 0.016, three independent experiments), which is a known down-stream target of MuSK, indicating partially impaired MuSK function (Fig. 2, B and C).

Fig. 2.

The 1124_1127dupTGCC mutation (1127dupTGCC) impairs Dok-7–induced post-synaptic specialization in myotubes. (A) Schematic representation of wild-type Dok-7 (WT) and its mutant harboring a 1124_1127dupTGCC mutation, showing the position of the frame-shift (indicated by the asterisk) and the resulting region of missense amino acids (light gray area). (B to E) The 1127dupTGCC mutation impairs Dok-7–induced MuSK activation and the subsequent AChR clustering in myotubes. (B) Anti-MuSK immunoprecipitates (IP), α-bungarotoxin (BuTx) eluates, or whole-cell lysates (from C2C12 myotubes transfected with plasmids expressing wild-type Dok-7 or the 1127dupTGCC mutant) were subjected to immunoblotting (IB). α-PY, antibody to phosphotyrosine. Quantification of phosphorylation is shown in (C); error bars indicate mean ± SD (n = 3). The number, size, and morphology of the AChR clusters induced by each transfection were visualized and analyzed according to the length of the longest axis (D) and the morphology of individual clusters (E). 240 to 248 clusters were analyzed for each transfection in (D) and (E); results are shown as the mean ± SD of three experiments. P+C is the sum of the numbers of perforate-type (P) and c-shape type (C) AChR clusters. The numbers of P+C and branched-type (B) AChR clusters are shown [top panel in (E)] as a percentage of the total number of clusters, including the plaque-type (PL) cluster. Representative clusters are shown in the bottom panels of (E). Scale bars, 20 μm.

We next analyzed the AChR clusters that were induced in C2C12 myotubes after the expression of wild-type or mutant Dok-7–1124_1127dupTGCC. Compared to the wild type, Dok-7–1124_1127dupTGCC induced the formation of fewer clusters that were 5 μmorgreaterin size [90 ± 9.41 versus 37 ± 5.29 clusters per 10 fields; mean ± SD; one-way analysis of variance (ANOVA); P < 0.0001; wild type versus Dok-7–1124_1127dupTGCC; three independent experiments], with a shift toward smaller clusters (Fig. 2D, two-way repeated measures ANOVA, P < 0.001, three independent experiments). AChR clusters formed in cultured myotubes show a variety of morphologies, including plaque-shaped, perforated, c-shaped, and branched types (9), that may reflect the maturation of the NMJ seen in vivo. The branched shape of the clusters corresponds most closely to the morphology of the mature NMJ (9). Analysis of clusters induced by the Dok-7–1124_1127dupTGCC mutation, as compared to the wild type, showed a significant reduction in the number of branched-type plaques (Fig. 2E, one-tailed paired t test, P = 0.026, three independent experiments), suggesting that the mutation attenuates the maturation of the synaptic structure. Thus, in cultured myotubes, the truncation of Dok-7 by 1124_1127dupTGCC impairs MuSK activity and its ability to shape the specialization of postsynaptic structures.

Clinical examination of 19 out of 21 index cases and the affected brother of case 11 was undertaken in Oxford or Munich (table S1). All of the patients displayed electromyographic evidence of a defect in neuromuscular transmission (determined by abnormal decrement and/or abnormal single-fiber studies), and all but three patients developed weakness within the first five years of life. The clinical onset of disease was generally characterized by difficulty in walking after initially achieving normal walking milestones. Features typically seen in patients with rapsyn mutations, such as congenital joint deformity and squint, were not present. In adulthood, a proximal weakness of the affected patients' upper and lower extremities was evident, and most had weakness in the trunk and neck regions. All of the patients had weak facial muscles, and all but two patients had ptosis. Eye movements were generally unaffected. Anticholinesterase medication either had no effect or made the weakness worse, although a short-lived initial response was occasionally seen.

An analysis of motorpoint muscle biopsies from a cohort of CMS patients with limb girdle proximal weakness showed that many features of the NMJ were normal, including the quantal release per unit area of synaptic content and the size and kinetics of the miniature end-plate currents (3). However, two major abnormalities were identified: a reduced size of the NMJs (Tables 2 and 3) and reduced postsynaptic folding (3). We identified C-terminal domain frameshift mutations in Dok-7 in DNA available from six out of seven patients included in the study by Slater et al. (Table 2) and from three patients with reduced NMJ size that were studied in Oxford (Table 3). The postsynaptic fold index in biopsies from patients harboring Dok-7 mutations that were studied at Newcastle was 5.51 ± 1.2 (mean ± SD, n = 7) versus 8.05 ± 1.34 (mean ± SD, n = 8) for control patients.

Table 2.

Vastus lateralis biopsies from patients harboring mutations in Dok-7 studied in Newcastle. For the index cases, the numbers refer to those in Table 1, in which the index case mutations are identified. The number associated with LGM (limb-girdle myasthenia) refers to patients whose vastus lateralis biopsies were reported in (3); patient 1 was not included in that study. The total area of acetylcholinesterase (AChE)–stained NMJs was measured. ns, not significant; MEPP, miniature end-plate potential; MEPC, miniature end-plate current.

Index caseSex, age at biopsyAChRs/NMJ (× 107)Total AChE area (μm2)AChRs/AChE areaMEPP amplitude (mV)MEPC amplitude (nA)Quantal content (m)
1 M, 22 1.06 69.6 0.015 0.64 nd 13.8
8 (LGM1) F, 12 nd 109.0 nd 0.45 2.04 14.7
9 (LGM3) M, 14 nd 82.1 nd 0.43 3.84 12.4
11 (LGM8) F, 38 0.65 47.1 0.014 0.62 6.13 13.2
12 (LGM5) M, 22 1.26 107.9 0.012 0.56 5.40 4.8
13 (LGM4) F, 26 0.90 78.6 0.012 0.51 4.41 11.5
14 (LGM7) M, 20 0.46 82.4 0.006 0.33 4.26 15.5
Mean ± SD (n) 0.87 ± 0.32 (5) 82.3 ± 21.5 (7) 0.012 ± 0.003 (5) 0.51 ± 0.11 (7) 4.35 ± 1.41 (6) 12.3 ± 3.6 (7)
Controls, mean ± SD (n) 2.34 ± 0.4 (2) 146.8 ± 26.9 (8) 0.016 ± 0.006 (2) 0.71 ± 0.19 (8) 3.59 ± 0.47 (6) 21.2 ± 4.8 (8)
P, unpaired t test 0.003 0.009 0.23, ns 0.023 0.24, ns 0.0013
Table 3.

Intercostal muscle biopsies in patients harboring mutations in Dok-7 studied in Oxford. Index case numbers refer to those in Table 1, in which the index case mutations are identified. The length of AChE-stained NMJs was measured as length in sarcomeres, where one sarcomere equals 3.2 μm.

Index caseSex, age at biopsyAChRs/NMJ (× 107)Length of AChE stained area (μm)AChRs/AChE lengthMEPP amplitude (mV)MEPC amplitude (nA)Quantal content (m)
2View inline M, 41 0.84 17.6 0.048 0.32View inline nd nd
3 F, 53 0.90 15.3 0.059 0.50 nd 14
4View inline F, 44 0.60 22.7 0.027 0.56 nd nd
Controls, mean ± SD (n) 1.2 ± 0.3 (7) 24 ± 2.9 (3) 0.042 (3) 0.8 ± 0.1 (4) 20 ± 3.2 (4)
  • View inline* Data reported in (18).

  • View inline Determined in the presence of eserine; nd, not done.

  • Our results demonstrate that mutations in Dok-7 affect the size and structure of the NMJ and underlie a CMS with a characteristic pattern of weakness, particularly affecting the proximal muscle groups. Impaired activation of MuSK is implicated as the pathogenic mechanism in these patients with Dok-7 mutations. The reduction in NMJ size, and hence efficacy, that results from mutations affecting the C-terminal region of Dok-7 could not have been predicted from earlier experiments, in which the complete inactivation of the gene encoding Dok-7 in mice virtually abolished NMJ formation (8). Many NMJ-associated proteins, including rapsyn (10), ColQ (11), and various different kinases such as Abl and Src/Fyn (12, 13), have been proposed as interacting partners with MuSK. The truncation of the C-terminal end of Dok-7 may reduce MuSK catalytic activity directly, affect the docking of other signal-transducing molecules, and/or affect the binding of MuSK to other NMJ-associated proteins. Subsequently, the pre- and postsynaptic structures are affected, leading to the reduced synaptic size.

    Because the onset of symptoms resulting from Dok-7 mutations is usually observed in early childhood and not at birth, the mutations probably have little pathogenic effect on initial synapse formation but instead exert their effects through aberrant synaptic maturation or maintenance. This suggestion is consistent with our functional studies showing that 1124_1127dupTGCC impairs the maturation of the postsynaptic structure in cultured myotubes. Dok-7 that harbored a truncated C-terminal region induced MuSK activation during the early stages of the differentiation of C2C12 cells into myotubes but not when the myotubes were fully differentiated (8).

    Mutations in Dok-7 appear to be a common cause of CMS in patients of different European or Indian ethnic origins (Table 1). Genetic screening of exon 7 in dok-7 should facilitate the diagnosis of this disorder, although, because a limb girdle pattern of muscle weakness has been reported for other myasthenic disorders (1416), additional genes are probably involved (supporting online text). Our findings emphasize the importance of synaptic structure as a determinant of functional efficacy (17) and indicate that Dok-7–associated CMS should be classified as a synaptopathy rather than a channelopathy.

    Supporting Online Material

    Materials and Methods

    SOM Text

    Fig. S1

    Table S1


    References and Notes

    View Abstract

    Stay Connected to Science

    Navigate This Article