|Home | About | Journals | Submit | Contact Us | Français|
Centronuclear myopathy is a pathologically diagnosed congenital myopathy. The disease genes encode proteins with membrane modulating properties (MTM1, DNM2, and BIN1) or alter excitation-contraction coupling (RYR1). Some patients also have myasthenic symptoms but electrodiagnostic and endplate studies in these are limited. A sporadic patient had fatigable weakness and a decremental EMG response. Analysis of centronuclear myopathy disease- and candidate- genes identified no mutations. Quantitative endplate structure and in vitro microelectrode studies revealed simplified postsynaptic regions, endplate remodeling with normal nerve terminal size, normal synaptic vesicle density, and mild acetylcholine receptor deficiency. The amplitude of the miniature endplate potential was decreased to 60% of normal. Quantal release by nerve impulse was reduced to 40% of normal due to a decreased number of releasable quanta. The safety margin of neuromuscular transmission is compromised by decreased quantal release by nerve impulse and by a reduced postsynaptic response to the released quanta.
Centronuclear myopathies (CNM) represent a group of clinically and genetically heterogenous congenital myopathies sharing a common pathology of centronucleated muscle fibers without a significant number of necrotic or regenerating fibers. To date, mutations in genes encoding myotubularin (MTM1), dynamin 2 (DNM2), amphiphysin 2 (BIN1) and skeletal muscle ryanodine receptor (RYR1) have been identified in nearly 70% of CNM patients .
Myotubularin is a phosphoinositide phosphatase important for endosomal trafficking . Non-quantitative EM studies find disorganization of the T-tubules in zebrafish model of myotubularin-CNM . Dynamin 2 is a large GTPase involved in the release of nascent vesicles during endocytosis and intracellular membrane trafficking. Dynamin 2 induces membrane curvature in both clathrin-dependent and clathrin-independent endocytosis and associates with amphiphysin on phagosomes . Amphiphysin 2 is involved in membrane remodeling, is important for organization of the T-tubules, and is regulated by phosphoinositides . T-tubules have been described as sometimes slightly dilated in amphiphysin 2-CNM . At the neuromuscular junction, amphiphysin promotes assembly of neuronal dynamin 1 at the neck of clathrin coated synaptic vesicles undergoing endocytosis . Dynamin 1 then binds to clathrin and adaptor protein 2 (AP-2) and acts as a molecular scissor to effect fission of the coated vesicle from the presynaptic membrane . RYR1 is the calcium release channel of the sarcoplasmic reticulum (SR) essential for excitation-contraction coupling [9,10].
The age of onset and severity of weakness in CNM patients vary widely depending on the causative gene. X-linked CNM due to mutations of MTM1 causes severe neonatal-onset and often lethal weakness. DNM2 mutations typically give rise to a childhood-onset, autosomal dominant myopathy with mild weakness [11,12]. RYR1 and less commonly BIN1 mutations underlie an autosomal recessive CNM with an intermediate phenotype between MTM1- and DNM2-CNMs [10,13]. Four intermediate phenotype CNM patients with different dominant RYR1 mutations were also reported [9,10]. Recently, a heterozygous mutation in MTMR14, the gene encoding myotubularin-related protein 14, was observed in two Brazilian patients with an intermediate phenotype. One of these patients also carried a known DNM2 mutation and the other has an asymptomatic parent carrying the same MTMR14 mutation , suggesting that MTMR14 is a modifying rather than a causative gene, and this was recently confirmed by a study in a zebrafish model of MTMR14-CNM .
Apart from limb weakness, ptosis and ophthalmoparesis are common manifestations in different CNM subtypes [6,16–24]. Some CNM patients also experience abnormal fatigability [6,16,25–27]. The constellation of these symptoms suggests an associated defect of neuromuscular transmission but a decremental EMG response, abnormal SFEMG, or changes in EP morphology were documented in only few CNM patients [6,16,18,19,21,22] (Table 1). Moreover, some previously reported ultrastructural changes represent preparatory artifacts. Most of these studies were done in the pre-genetic era but a recently reported CNM patient with fatigable weakness and a decremental EMG response is homozygous for a missense mutation in BIN1 . Detailed analysis of endplate ultrastructure and neuromuscular transmission have not been available to date.
We here report a CNM patient with fatigable weakness and a decremental EMG response. We analyze the structure of intercostal muscle endplates by quantitative electron microscopy and the parameters of neuromuscular transmission by in vitro microelectrode studies. Mutation analysis of currently identified and candidate CNM genes revealed no mutations.
All human studies were in accord the regulations of the Mayo Institutional Review Board.
Acetylcholinesterase (AChE) and the acetylcholine receptor (AChR) were localized in frozen sections with a monoclonal anti-AChE antibody and rhodamine-labeled α-bungarotoxin (α-bgt) . Endplates were localized for electron microscopy and quantitatively analyzed by established methods . Peroxidase-labeled α-bgt was used for the ultrastructural localization of AChR . The number of AChRs per endplate (EP) was measured with [125I]α-bgt, as described previously . The amplitude of miniature EP potential (MEPP), and EP potential (EPP), estimates of the number of quanta released by nerve impulse (m), the number of releasable quanta (n), and the probability of quantal release (p) were determined as previously described [32–34].
We directly sequenced all exons and flanking untranslated regions of all known causative genes of CNM (MTM1, DNM2, BIN1, RYR1 and MTMR14) and candidate genes including genes encoding amphiphysin 1 (AMPH), dynamin 1 (DNM1), and Sorting Nexin 9 (SNX9). Genomic sequences of all genes were obtained from GenBank. DNM1 encodes a neuron-specific dynamin and AMPH codes for a neuron-specific amphiphysin isoform and both proteins play an important role in synaptic vesicle endocytosis [7,35]. SNX9 is a ubiquitously expressed protein that recruits dynamin-2 to sites of endocytosis and stimulates its GTPase activity. SNX9 also contains a BAR domain that promotes membrane curvature like amphiphysin .
A 39-year-old man was the product of a normal pregnancy. There were no perinatal complications. His early motor development was normal but he never ran well and fell easily while running. At age 13 years, he noted difficulty raising his arms overhead, combing his hair, and climbing stairs. He had had mild ptosis when tired but no other symptoms referable to cranial muscles. His symptoms progressed slowly until the age of 25 years but then became stable or even slightly improved, but he also experienced episodes of increased fatigability lasting 3–4 weeks and low grade myalgias sometimes related to overexertion, alcohol consumption, or aphthous stomatitis. The serum CK level was intermittently elevated to 3 to 10 times above the upper limit of normal. He had no episodes of pigmenturia. The parents are fourth degree cousins. There are no other similarly affected family members.
Physical examination revealed high-arched feet and hammertoes. There was hypertrophy of gastrocnemius and atrophy of thigh, upper arm, and pectoral muscles. He rose from the floor with a Gowers maneuver (Fig. 1). The arm elevation time was close to a minute, but he struggled to keep his arms in the horizontal position. The leg elevation time was only 25 seconds. He rose from a low-sitting position ten times but with increasing difficulty. Repeated shoulder abductions against resistance caused increasing weakness of the deltoid muscles. There was no weakness of the cranial muscles, and the pupillary light reflex was normal. There was no ptosis or fatigability of eyelid elevators. Detailed manual muscle testing showed mild to moderately severe weakness of pelvic and pectoral girdle muscles with selectively severe involvement of the gluteus maximus and toe extensor muscles and mild weakness of the distal limb muscles. The tendon reflexes were hypoactive to absent in the arms but normally active at the knees and ankles. The rest of the neurologic examination was unremarkable.
Tests for acetylcholine receptor and MuSK antibodies were negative. The serum CK level was 878 U/L (normal, 52–336 U/L).
Standard nerve conduction studies gave normal results. Repetitive nerve stimulation at 2-Hz of the musculocutaneous and spinal accessory nerves revealed a 19% decrement in the biceps and a 35% decrement in the trapezius muscle of the fourth compared to the first evoked compound muscle action potential. Edrophonium decreased the decremental response in the trapezius to 23%, and 3,4-diaminopyridine (3,4-DAP) diminished it to 13%, of the initial value. Needle EMG examination of limb muscles showed unstable, short-duration, low-amplitude motor unit potentials.
Based on results of the EMG studies, we treated the patient with 60 mg pyridostigmine twice daily and 10 mg 3,4-DAP three times daily. Two days later he could climb 60 steps without having to rest whereas previously he could climb only 20 steps due to fatigue and myalgias. There was now only subtle weakness of the gluteus maximus and tibialis anterior muscles and slight fatigable weakness of the deltoid muscles. A few weeks later, however, he experienced myalgias after taking both medications and now uses only 60 mg pyridostigmine when he is more active than usual.
Biopsy specimens of intercostal, serratus anterior and latissimus dorsi muscles showed similar findings. The muscle fibers varied from 30 to 120 μm in diameter except serratus anterior contained few atrophic fibers consisting of clumps of nuclei surrounded by a thin rim of cytoplasm (Fig. 2). Nearly all intercostal and serratus anterior and 80% of the latissimus dorsi fibers displayed one or more internal nuclei. Rare fibers were necrotic and replaced by macrophages or were regenerating. There was a moderate increase in perimysial fibrous and fatty connective tissue. In NADH dehydrogenase-reacted section, some fibers displayed small irregular attenuations of enzyme activity or aberrant or coiled myofibrils. Rare fibers showed a radial arrangement of the myofibrils. Tubular aggregates were absent. As adjudged by ATPase-reacted sections, most fibers in serratus anterior and latissimus dorsi and all fibers in the intercostal muscle were histochemically type 1. In latissimus dorsi, type 1 fibers had a smaller mean diameter than type 2 fibers.
Patient EPs observed in transversely sectioned frozen sections expressed AChE normally but the expression of AChR was mildly attenuated compared to control EPs (Fig. 3). In AChE reacted teased muscle fibers, 28 of 48 fibers displayed multiple synaptic contacts consisting of 2 to 4 ovoid EPs each about 10μm in diameter (Fig. 3E) or multiple small pleomorphic EP regions, with the smallest regions reacting for AChE with reduced intensity, arrayed over an extended length of the fiber (Fig. 3F). The latter finding also occurs in other myasthenic disorders associated with ongoing destruction and remodeling of the postsynaptic region [37,38].
Fifty-five patient and 162 control EP regions were inspected for changes of EP conformation. Forty-one of 55 patient EP regions displayed one or more of the following abnormalities: simplified (Fig. 4A and C) or absent (Figs. 4B and and5)5) junctional folds, partial occupancy of the postsynaptic region by nerve terminal, abundant junctional sarcoplasm harboring numerous dilated vesicles ranging from 200 to 900 nm in diameter some of which resembled dilated components of the sarcoplasmic reticulum (Fig. 5), and occasional Schwann cell encasement of the nerve terminals (Table 2).
Extrajunctional muscle fiber regions showed streaming of few Z-disks or focal dilations of the sarcoplasmic reticulum, but no clearly identifiable T-tubule abnormalities.
Morphometric analysis of 40 to 48 EP regions showed the nerve terminal area decreased to 75% of the control value but this did not reach statistical significance. The synaptic vesicle density (number/μm2 of nerve terminal area) was not significantly different from that at control EP regions. In contrast, the postsynaptic area of folds and clefts was reduced to 40%, and the postsynaptic membrane density was decreased to 76%, of the control value (Table 3).
Indirect stimulation of muscle elicited no contractions. Therefore all measurements were obtained in the absence of curare (Table 4). The MEPP amplitude was reduced to 59% of normal and its decay time was prolonged 1.5-fold. The quantal content of the EPP was reduced to 42% of normal. The probability of quantal release (p) was normal but the number of releasable quanta (n) was only half-normal. The number of α-bungarotoxin binding sites per EP was mildly reduced.
Depolarization of the nerve terminal by raising the [K+] from 5 to 20 mM in the bath increased the MEPP frequency 65-fold in patient and 41-fold in control fibers, indicating the decreased quantal content of the patient EPP was not likely due to reduced Ca2+ ingress into the depolarized nerve terminal.
Direct sequencing of the known CNM genes MTM1, DNM2, BIN1, RYR1, of the modifier gene MTMR14, and of candidate genes AMPH, DNM1 and SNX9 revealed no mutations.
We here present the first detailed structural and electrophysiologic analysis of the neuromuscular junction in a CNM associated with a myasthenic syndrome. The myasthenic symptoms were mitigated by cholinergic agonists but their use was limited by myalgias. Although it is now known that a myasthenic component can be a facet of CNMs, it is not known whether it is present in all CNMs. Therefore it will be important to carefully search for a defect of neuromuscular transmission in all CNM patients.
Intercostal muscle studies show formation of new EP regions on individual fibers and simplified or absent junctional folds at some EPs (Table 2). At individual EP regions the postsynaptic area of folds and cleft is reduced to 40% and the postsynaptic membrane density is decreased to 65% of the corresponding control value. These alterations can at least in part be attributed to immaturity of newly formed EP regions. In contrast, the nerve terminal area and the density of the synaptic vesicles in the nerve terminal were not significantly different from normal.
The numerous dilated vesicles in the junctional sarcoplasm of many EPs were of special interest. Some of these vesicles were empty but others had a fine granular content similar to that present in lateral vesicles of the sarcoplasmic reticulum, and some of these vesicles appeared in a triadic configuration. The functional significance of this finding remains unknown.
The in vitro electrophysiologic studies reveal reduction of the quantal content of the EPP (m) to ~40% of normal and of the MEPP amplitude to ~60% of normal. In addition, the decay time of the MEPP is 1.5 times longer than normal (Table 4). Because m is a function of the probability of quantal release (p) and the number of releasable quanta (n) according to the relationship m = n × p , and because p is normal, the decrease in m is attributed to the decrease in n. The parameters affecting n include the nerve terminal volume , synaptic vesicle density , and integrity of the synaptic vesicle cycle. The total nerve terminal volume per EP was not determined but synaptic vesicle density in individual nerve terminals was not significantly altered. A defect in the synaptic vesicle cycle could be due to impaired exocytosis or endocytotic retrieval of the vesicles. Impaired vesicle exocytosis is unlikely to reduce the synaptic vesicle density but decreased endocytotic retrieval could reduce it in the wake of physiologic activity; this, however, would not be reflected by our analysis because the EPs for EM studies were fixed in the resting state. Future quantitative EM studies comparing the synaptic vesicle density before and after tetanic stimulation or fluorescent dye studies to monitor vesicle traffic in patient and control muscles will be required to distinguish between these possibilities.
The decrease of the MEPP amplitude can be attributed to a combination of factors. (1) Simplification of the postsynaptic regions reduces the input resistance of the EP and hence the amplitude of the MEPP . (2) The number of AChRs per EP and the AChR index were decreased to 78% and 76% of normal, respectively. (3) The immature EP regions may express immature AChRs containing the γ instead of the ε subunit, and opening events of the γ-AChR are of lower amplitude and of longer duration than those of the adult ε-AChR. Presence of γ-AChR at some EPs would also account for the 1.5-fold prolonged decay time of the MEPP.
Our genetic studies exclude with reasonable certainty mutations in currently recognized disease and candidate genes for CNM. This implies that other CNM disease genes likely exist and await discovery. That the physiologic abnormality in our patient involves the synaptic vesicle cycle predicts that the disease gene, like previously identified CNM disease genes, has membrane modulating properties.
This work was supported by National Institutes of Health R01 Grant NS6277 and by a Research Grant form the Muscular Dystrophy Association.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.