In 38 patients and in their affected siblings, the disease manifested at birth or in the first 2 years of life. Only one patient presented at 5 years of age. This contrasts with previous reports of onset as late as the fifth decade of life.9,16
Most patients had ptosis of varying severity; in 5 it was asymmetric, a finding thought to be uncommon in congenital myasthenias.17
As previously observed,9,18
strabismus was relatively common, but 9 patients in our series have had constant or episodic ophthalmoparesis. Therefore, ophthalmoparesis is not a reliable negative criterion for distinguishing rapsyn-CMS from CMS caused by low expressor mutations in the AChR
subunit as previously suggested.9
Still, except for the CMS caused by mutations in DOK7
, ophthalmoparesis is less common than in other forms of CMS. Facial and bulbar weakness were common, often associated with neck muscle weakness. Proximal muscle weakness was as or more severe than distal weakness. One patient had only mild distal but no proximal weakness, but his proximal muscles fatigued abnormally on exertion and he also had ptosis, ophthalmoparesis, and facial weakness. Out-of-proportion weakness of the foot dorsiflexors was reported a feature of the late-onset phenotype,9
but it was not detected in our series of early onset patients. Intermittent exacerbations occurred in more than half of the patients. Arthrogryposis manifesting at birth occurred in less than a third of the patients.
Patient 38 with a homozygous N88K mutation had limb-girdle weakness and abnormal fatigability present at least since the age of 5 years. He had no ocular, facial, or bulbar symptoms, intermittent exacerbations, and his clinical course was stable. His muscle biopsy showed no tubular aggregates. This patient’s case extends the genetic causes of the congenital limb-girdle myasthenias beyond mutations in Dok-7, or with tubular aggregates in muscle with no mutations in Dok-7.19–21
Not all patients had a decremental EMG response at rest. Six with mild weakness of muscles accessible for repetitive nerve stimulations had none, but in 5 of these single-fiber EMG demonstrated abnormal jitter and blocking. Three patients with mild weakness had a transient decremental EMG response at 2 Hz immediately after a conditioning train at 10 Hz for 5 minutes. This is a nonspecific finding in different types of CMS; it is specific only if it persists for 5-10 minutes after stimulation, in which case it would point to a defect in choline acetyltransferase.22
The in vitro electrophysiologic and electron microscopy studies of intercostal or anconeus muscle EPs in 7 patients showed multiple small synaptic contacts dispersed over an extended length of the muscle fiber. Electron microscopy revealed EPs with few or no junctional folds, and a patchy distribution of AChR on the highly simplified postsynaptic membrane. The safety margin of neuromuscular transmission in rapsyn-CMS is compromised by the decreased amplitude of the MEPP which is due to the decreased number of AChRs per EP, and by the paucity of junctional folds which decreases the input resistance of the postsynaptic membrane and hence the amplitude of the MEPP.23
Interestingly, the decrease in the number of AChRs per EP (mean decrease 32% of normal) is less marked than in patients with low-expressor mutations in AChR subunits in whom it is typically less than 10% of normal. This likely explains why the amplitude of the MEPP and MEPC is not as markedly reduced, and why in some patients the EMG decrement is more difficult to detect, than in patients with low-expressor mutations of the AChR.
Since the discovery that mutations in rapsyn cause a CMS,8
a total of 45 rapsyn mutations have been identified, 15 by us14,15,24
and 30 by other investigators.9,10,16,18,25–31
The mutations are dispersed through the entire RAPSN
gene. N88K occurred with high frequency, as observed in other series.9,18,26,32
Six mutations reported here are novel and family analysis indicates they are recessive.
The mutated arginine at codon 164 (p.R164H) and 242 (p.R242W), and the mutated leucine at codon 326 (p.L326P), are highly conserved among species ().
p.R164H is located in TPR5. Another mutation at the same location (p.R164C) was shown to reduce co-clustering of AChR with rapsyn by about 50%.29
Although arginine and histidine are both positively charged amino acids, histidine has an imidazole group which likely affects the secondary structure of rapsyn.
p.R242W is located in the linker region between TPR6 and TPR7. Replacing the positively charged arginine with the larger nonpolar tryptophan may distort the structure of the close TPR7, interfering with its role in co-clustering of AChR with rapsyn. A new mutation was previously reported in rapsyn-CMS 4 amino acids downstream (p.A246V).18
p.L326P is the first mutation observed in the coiled-coil domain (298-331) of rapsyn. This domain interacts with the long cytoplasmic loop of each AChR subunit and is essential for clustering AChRs.4
Leucine at position 326 contributes to the continuous hydrophobic surface of the coiled-coil domain. The integrity of this hydrophobic surface has been shown to be essential for AChR clustering by site-directed mutagenesis studies.4
Regarding the novel p.V50_S55del, 3 of the 5 deleted amino acids are conserved across species (). The deletion is located in the TPR2, one of the most important domains for rapsyn self-clustering. The deletion may also affect the folding of the protein. Rapsyn mutant lacking TPR2 does not form clusters efficiently and exerts a dominant negative effect when expressed with wild type rapsyn by disrupting rapsyn self-association.33
The nonsense mutation p.Q325X results in truncated rapsyn lacking part of the coiled coil domain and the C-terminal Ring-H2 that mediates association of rapsyn with AChR subunits and dystroglycan.4,7
The mutant rapsyn also lacks the extreme C-terminus essential for MuSK-induced tyrosine phosphorylation of the AChR β subunit.34
The splice-site mutation c.966 + 1GT>AG may result in exon skipping, intron retention, creation of an intronic pseudo-exon, or activation of cryptic splice site.35
Point mutations at position IVS + 1G>A more likely lead to cryptic 5′ splice sites.36
Therefore, the splice site likely affects the coiled-coil and the C-terminal RING-H2 domains of rapsyn.
We found no genotype-phenotype correlation except the association of the homozygous c.-38A>G rapsyn E-box mutation in Near Eastern patients with mandibular prognathism and a mild clinical course.14,37
In particular, heterozygous truncation mutations are not associated with a more severe phenotype and the phenotypic spectrum of p.N88K homozygosity in our series encompasses respiratory failure at birth (patient 4), mild ptosis with intermittent exacerbations (patient 12), and pure limb-girdle weakness (patient 38). Arthrogryposis occurred in patients with missense, frameshift, in-frame, and promoter region mutations. The type of mutation did not correlate with the clinical course, which included stability, improvement, or progressive worsening. The type of mutation also did not correlate with the extent of the EMG decrement or with the decreased amplitude of the MEPP and MEPC.