Congenital Myasthenic Syndromes: Recent Advances

Congenital myasthenic syndromes (CMS) can arise from presynaptic, synaptic, or postsynaptic defects. Mutations of the acetylcholine receptor (AChR) that increase or decrease the synaptic response to acetylcholine (ACh) are a common cause of the postsynaptic CMS. An increased response occurs in the slow-channel syndromes. Here, dominant mutations in different AChR subunits and in different domains of the subunits prolong the activation episodes of AChR by either delaying channel closure or increasing the affinity of AChR for ACh. A decreased synaptic response to ACh occurs with recessive, loss-of-function mutations. Missense mutations in the low-affinity, fast-channel syndrome and in a disorder associated with mode-switching kinetics of AChR result in brief activation episodes and reduce the probability of channel opening. Mutations causing premature termination of the translational chain or missense mutations preventing the assembly or glycosylation of AChR curtail the expression of AChR. These mutations are concentrated in the ϵ subunit, probably because substitution of the fetal γ for the adult ϵ subunit can rescue humans from fatal null mutations in ϵ. Recent molecular genetic studies have also elucidated the pathogenesis of the CMS caused by absence of the asymmetric form of acetylcholinesterase from the synaptic basal lamina. Endplate acetylcholinesterase deficiency is now known to be caused by mutations in the collagenic tail subunit of the asymmetric enzyme that prevents the association of the collagenic tail subunit with the catalytic subunit or its insertion into the basal lamina.

Congenital myasthenic syndromes (CMS) are heterogeneous disorders arising from presynaptic, synaptic, or postsynaptic defects. In each CMS, the specific defect compromises the safety margin of neuromuscular transmission by one or more mechanisms. The clinical phenotypes of CMS are often similar; therefore, precise diagnosis requires correlation of clinical, in vitro electrophysiological, morphological, and, whenever possible, molecular genetic studies.¹

Prior to 1990, the investigations involving patients with CMS were based on clinical, morphologic, and conventional microelectrode studies. Since then, four developments have paved the way for molecular analysis of CMS. First, by 1993, the complementary DNA sequences of the α, β, δ, and ϵ subunits of adult and of the γ subunit of fetal human acetylcholine receptor (AChR) were known, allowing molecular genetic analysis. Second, in the early 1990s, Milone et al²succeeded in patch-clamping endplates in human intercostal muscles, permitting analysis of the activity of single AChR channels. Third, the use of mammalian expression systems facilitated detailed analysis of how human AChR mutants alter the kinetics of the AChR channel. Coincident with this, we¹hypothesized that a kinetic abnormality of AChR at the single-channel level predicts, and that severe endplate AChR deficiency may predict, one or more mutations in the subunits of AChR. This hypothesis was subsequently confirmed by the discovery of mutations in different subunits of AChR that either increase^3-7or decrease^7-14the synaptic response to acetylcholine (ACh). Fourth, the recent cloning of the catalytic^15,16and collagenic tail (ColQ)¹⁷subunits of asymmetric acetylcholinesterase (AChE) opened the door to identifying the genetic causes of endplate AChE deficiency.¹

Since 1994, we and other investigators have identified 56 AChR subunit gene mutations in 69 CMS kinships. Table 1indicates the identified mutations according to their functional consequences and subunit locations. It includes 34 published and 17 unpublished CMS mutations observed in our laboratories, three slow-channel mutations in the α subunit¹⁸and a frame-shifting rearrangement in the ϵ subunit described by Croxen et al,¹⁹and a slow-channel mutation in the β subunit detected by Gomez et al.²⁰Interestingly, 38 of the 56 mutations and all 27 null mutations occur in the ϵ subunit of AChR, highlighting the susceptibility of the ϵ subunit gene to mutation.

The clues for the diagnosis of a slow-channel CMS consist of selectively severe weakness of the forearm extensor muscles, a repetitive compound muscle action potential response to single-nerve stimuli that is accentuated by edrophonium, and a prolonged and biexponentially decaying miniature endplate current. Eleven slow-channel CMS mutations have been reported to date.^3-6,18,20-23The different mutations occur in different AChR subunits and in different functional domains of the subunits ( Figure 1, A). Each is dominant, causing a pathologic gain of function.

The phenotypic consequences of the slow-channel CMS mutations stem from prolonged opening episodes of the AChR channel. These cause (1) cationic overloading of the junctional sarcoplasm and an endplate myopathy with loss of AChR from degenerating junctional folds and (2) a depolarization block due to staircase summation of prolonged endplate potentials.^3-5Patch-clamp studies at the endplate, mutation analysis, and expression studies in human embryonic kidney fibroblast (HEK) cells dissected three types of slow-channel CMS. Those residing in the M2 domain, which lines the channel pore, act predominantly by slowing channel closure.^3,5,6A mutation near the ACh binding site on the α subunit increases affinity of AChR for ACh, causing repeated reopenings of the channel during the prolonged ACh occupancy.⁴Another type of CMS has features of the two preceding types and the mutations reside in the M1 or M2 domain.^5,6,22

Recent studies indicate that quinidine, a long-lived open-channel blocker of AChR, is beneficial in the slow-channel CMS. Fukudome et al²⁴demonstrated that drug levels attainable in clinical practice shorten and even normalize the prolonged opening episodes of mutant slow-channel CMS acetylcholine receptors expressed in HEK cells, and Harper and Engel²⁵found that doses of the drug producing serum levels of 0.7 to 2.5 µg/mL (2.1-7.7 µmol/L) benefit patients with slow-channel CMS by clinical and electromyographic criteria.

Mutation analysis in two patients revealed two mutations in different alleles of the ϵ subunit gene: a common ϵP121L mutation plus a null mutation in the second ϵ allele, so that ϵP121L defined the clinical phenotype. In these patients, the postsynaptic response to ACh is markedly diminished, although the number of AChR per endplate is normal.^8,26Patch-clamp studies show infrequent AChR channel events, abnormally brief activation episodes due to diminished channel reopenings during ACh occupancy, and increased resistance to desensitization by ACh.⁸Genetically engineered ϵP121L-AChR expressed in HEK cells has a markedly decreased rate of channel opening and shows greatly reduced affinity for ACh in the open-channel and desensitized states.⁸

It is interesting to note that the ϵP121L mutation and the slow-channel CMS mutations have opposite effects: slow-channel CMS mutations increase the duration of activation episodes, enhance ACh binding affinity, increase desensitization by ACh, and cause an endplate myopathy; ϵP121L shortens the duration of activation episodes, reduces ACh binding affinity, decreases desensitization by ACh, and leaves no anatomical footprint.

Recently we encountered a second low-affinity, fast-channel mutation, αV285I, combined with a low-expressor mutation, αF233V, in the other α allele.²⁷Detailed kinetic studies of αV285I are in progress.

In this disorder, an in-frame duplication in the long cytoplasmic loop of ϵ, ϵ1254ins18, appears in combination with a cysteine-loop null mutation, ϵC128S.¹³The ϵ1254ins18 mutation, which determines the phenotype, causes mode switching in the kinetics of receptor activation in which the normal high efficiency of gating is accompanied by two new modes that gate inefficiently. In the two abnormal modes the channel opens more slowly and closes more rapidly than normal. The ϵ1245ins18 AChR at the endplate shows abnormally brief activation episodes during steady-state agonist application, and appears electrically silent during the synaptic response to ACh. The phenotypic consequences are endplate AChR deficiency, simplification of the postsynaptic region, and compensatory expression of fetal AChR that restores electrical activity at the endplate and rescues the phenotype.

Severe endplate AChR deficiency can result from different types of recessive mutations in AChR subunit genes. The mutations are either homozygous or, more frequently, heterozygous. Morphologic studies show an increased number of endplate regions distributed over an increased span of the muscle fiber. The integrity of the junctional folds is preserved, but some endplate regions are simplified and smaller than normal. The distribution of AChR on the junctional folds is patchy and the density of the reaction for AChR is attenuated. Conventional microelectrode studies reveal a decreased amplitude of the miniature endplate potentials and currents, and frequently high or higher than normal quantal release by nerve impulse. Single-channel recordings at the endplate^7,10,13or immunocytochemical studies¹²often reveal the presence of fetal AChR that harbors the γ (γ-AChR) instead of the adult ϵ subunit at the endplate.

Different types of recessive mutations causing severe endplate AChR deficiency have now been identified ( Figure 1, B): (1) Mutations causing premature termination of the translational chain—these mutations are frame shifting,^{7,11,12,14,28}occur at a splice site,^9,14,28or produce a stop codon directly.⁷(2) Missense mutation in a signal peptide region (ϵG-8R).⁸(3) Missense mutations in residues essential for assembly of the pentameric receptor—mutations of this type were observed in the ϵ subunit at an N-glycosylation site (ϵS143L),⁸in cysteine 128 (ϵC128S), a residue that is an essential part of the C128-C142 disulfide loop in the extracellular domain,¹³and in arginine 147 (ϵR147L) in the extracellular domain, which lies between isoleucine 145 and threonine 150, residues that contribute to subunit assembly.⁷(4) Missense mutations affecting both AChR expression and kinetics. For example, ϵR311W⁷and ϵ1254ins18¹³in the long cytoplasmic loop between M3 and M4 decrease, whereas ϵP245L in the M1 domain⁷increases the open duration of channel events. In the case of ϵR311W and ϵP245L, the kinetic consequences are modest and are likely overshadowed by the reduced expression of the mutant gene.

There are two possible reasons that recessive mutations causing AChR deficiency are concentrated in the ϵ subunit. First, expression of the fetal type γ subunit, although at a low level, may compensate for absence of the ϵ subunit,^7,12,13whereas patients harboring null mutations in subunits other than ϵ might not survive for lack of a substituting subunit. Second, the gene encoding the ϵ subunit, and especially the exons coding for the long cytoplasmic loop, have a high GC content, which likely predisposes to DNA rearrangements.

In skeletal muscle, AChE exists in homomeric globular forms of type T catalytic subunits (ACHE_T) and heteromeric asymmetric forms composed of 1, 2, or 3 tetrameric ACHE_Tattached to a collagenic tail (ColQ). Asymmetric AChE is concentrated at the endplate where its ColQ anchors it into the basal lamina. The ACHETgene has been cloned in humans^15,16; COLQcomplementary DNA has been cloned in Torpedo²⁹and rodents,³⁰but not in humans. In endplate AChE deficiency (endplate AD), the normal asymmetric species of AChE are absent from muscle.³¹Endplate AD could stem from a defect that prevents binding of ColQ to ACHE_Tor the insertion of ColQ into the basal lamina. In recent studies of six patients with endplate AD, Ohno and coworkers¹⁷found no mutations in ACHET. They therefore cloned human COLQcomplementary DNA, determined the genomic structure and chromosomal localization of COLQ, and then searched for mutations in this gene. Their search revealed six recessive truncation mutations of COLQin the six patients. Coexpression of each COLQmutant with wild-type ACHE_Tin fibroblasts showed that a mutation proximal to the ColQ attachment domain for ACHETprevents association of ColQ with ACHE_T; mutations distal to the attachment domain generate a mutant species of AChE composed of one ACHE_Ttetramer and a truncated ColQ strand. The mutant species lack part of the collagen domain and the entire C-terminal domain of ColQ, or only the C terminal domain of ColQ that is required for formation of the triple collagen helix, and this likely prevents their insertion into the basal lamina. Additional observations³²indicate that endplate AD can also arise from missense mutations in COLQ. Finally, rare cases of endplate AD could stem from defects in the basal lamina that prevent the binding of ColQ.

Since 1994, molecular analysis of the CMS has provided clear insights into disease mechanisms and highlighted functionally significant domains of AChR and AChE. In the coming years molecular studies will undoubtedly be applied to presynaptic CMS, like those that alter the release of ACh quanta by nerve impulse or those that prevent the filling of synaptic vesicles with ACh. It is also likely that the molecular studies will provide clues for conventional and gene therapy and lead to the identification of novel CMS.

Accepted for publication April 24, 1998.

Work in the authors' laboratories was supported by grants NS6277 (Dr Engel) and NS31744 (Dr Sine) from the National Institutes of Health, Bethesda, Md, and a research grant from the Muscular Dystrophy Association, Tucson, Ariz (Dr Engel).

Reprints: A. G. Engel, MD, Mayo Clinic, 200 First St SW, Rochester, MN 55905.