Congenital neuromuscular disease with uniform type 1 fiber and RYR1 mutation

Congenital neuromuscular disease with uniform type 1 fiber (CNMDU1) was first described in 1983.¹ It is a rare disorder pathologically characterized by the exclusive presence of type 1 muscle fiber (>99%) without any specific structural abnormality such as cores, nemaline bodies, or centrally placed nuclei. Clinically, it shares common features with congenital myopathy; including early onset, mild proximal muscle weakness, hypo- or areflexia, normal creatine kinase levels, and myopathic electromyography findings. So far, at least 12 cases have been reported.^1–10 However, its genetic cause and molecular pathomechanism are still unknown.

We are aware of a rare existence of CNMDU1 case with a family history of central core disease (CCD) in our own series⁶ and in the previous report.⁷ In addition, our recent study on CCD revealed that patients with a heterozygous C-terminal mutation in the gene encoding ryanodine receptor 1 (RYR1) have nearly exclusively type 1 fibers, in addition to well-demarcated, mostly singular and centrally located, “typical” cores,¹¹ suggesting a tight relationship between uniform type 1 fiber and RYR1 mutations, especially those in the C-terminal domain. We therefore hypothesized that CNMDU1 may be caused by RYR1 mutation.

METHODS

RESULTS

Mutation screening.

Four of 10 patients (40%) had heterozygous sequence variations predicted to change amino acids in the C-terminal domain of RYR1 (figure 1, A through D). In Patient 1, we identified a 20-bp deletion (c.13013_13032delCAGCAGTGACGCGCGCTGGG, p.A4338fs) in exon 91, which resulted in a premature stop codon at the 4,575th amino acid. The presence of the deletion mutation was confirmed by sequencing of the cloned fragments. No family members including the parents and siblings carried the mutation, suggesting that the mutation in Patient 1 is a de novo mutation. Patient 2 had a missense mutation (c.14582G→A, p.R4861H) in exon 101, which had been previously reported in patients with CCD.¹⁴ In Patient 3, a novel missense mutation of c.14680G→C (p.A4894P) was identified in exon 102. Patient 4 had a substitution of two consecutive nucleotides (c.14761_14762TT→AC, p.F4921T) in exon 102, which was previously reported in his father with CCD.^6,11 We confirmed the two-nucleotide change in one allele and the absence in the other by sequencing of the cloned fragments in Patient 4. All three amino acids replaced by missense mutations, R4861, A4894, and F4921, were highly conserved across the RYR1 species including human, pig, rabbit, mouse, frog, and C. elegans (data not shown). These substitutions were not found in either 300 Japanese control chromosomes or in the Japanese Single Nucleotide Polymorphisms database.¹⁵

• Download figure
• Open in new tab
• Download powerpoint

Figure 1 Mutation analysis and Western blot analysis of RYR1

(A) A novel deletion mutation of c.13013_13032del20 (p.A4338fs) is identified in exon 91 in Patient 1. Deleted nucleotides (enclosed in box) are shown in wild-type sequence. (B) In Patient 2, a known missense mutation c.14582G→A (p.R4861H) is detected in exon 101. (C) A novel missense mutation of c.14680G→C (p.A4894P) is identified in Patient 3 in exon 102. (D) In Patient 4, substitution of two consecutive nucleotides c.14761_14762TT→AC (p.F4921T) is identified in exon 102. Arrows mark the site of mutation. (E) Western blot analysis of muscles from Patient 1, Patient 3, and control. All show a 565-kd band corresponding to the predicted size of full-length RYR1 (upper arrow). Only Patient 1 has a smaller sized band (about 513 kd) corresponding to the truncated RYR1 mutant (lower arrow).

The substitution c.11266C→G (p.Q3756E) in exon 79 previously reported as nonpathogenic¹⁷ was found in 4 of 10 patients: 1 with RYR1 mutation and 3 without RYR1 mutation. This substitution was also reported in the Japanese population diversity to be 11.4% in the Japanese Single Nucleotide Polymorphisms database. Fifteen silent single-nucleotide polymorphisms were also identified (data not shown). For the six patients without RYR1 mutation, we were able to amplify and sequence the C-terminal domain of RYR1 in muscle cDNA and confirmed the absence of any mutation including aberrant splicing.

No mutations were found in either FKBP1A or CACNA1S. Two patients with congenital myopathy with marked type 1 fiber predominance did not have any mutation in RYR1.

We also tried to sequence the C-terminal domain of RYR1 using DNA from the original three patients first reported to have CNMDU1.¹ However, as only paraffin-embedded muscles were available, the quality of DNA did not allow us to successfully amplify the regions except for exons 96 and 100 wherein no mutation was found.

Pathologic findings.

The mean age of biopsy between patients with and without mutations (3.2 ± 2.3, n = 4; 3.2 ± 4.9 years, n = 6) did not differ; all patients showed fiber size variation, regardless of the RYR1 mutations (table; figure 2). Endomysial fibrosis was mild except in two patients without RYR1 mutations (Patients 7 and 8). There were no necrotic or regenerating fibers, although a small number of fibers with internally placed nuclei were seen. No group atrophy was noted. No nemaline bodies, ragged-red fibers, or rimmed vacuoles were seen. Intermyofibrillar network was well organized in all fibers without any core or core-like structure. Type 1 fibers comprise more than 99% of fibers. A small number of type 2 fibers were seen except in two patients (Patients 3 and 4), even though the percentage was less than 1%. These type 2 fibers were either type 2B or 2C, and no type 2A fibers were observed.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2 Muscle biopsy

(A) Histochemistry. Patient 1 is a boy age 3 years 5 months (upper panel) with a mutation in RYR1. There is mild endomysial fibrosis. Patient 7 is an 8-month-old boy (lower panel) without RYR1 mutation. Marked fibrosis is observed. In both patients, pathologic findings show marked variation in fiber size, well-organized intermyofibrillar network, and with all fibers composed of type 1 on myosin ATPase staining at pH 4.2. Bar = 20 μm. (B) Electron micrograph. Neither Z streaming nor loss of mitochondria is seen. Bar = 5 μm.

On electron microscopy, none of 50 fibers observed showed either loss of mitochondria or disorganized myofibrillar structure such as Z-line streaming.

DISCUSSION

This is the first genetic study for CNMDU1. In 4 of 10 patients (40%), we identified a heterozygous mutation in the C-terminal domain of the gene encoding RYR1, which is virtually exclusively expressed in the skeletal muscle, forming the homotetrameric structures in the sarcoplasmic reticulum membrane, and functions as a Ca²⁺ release channel.¹⁸ As RYR1 mutations have been associated with three different diseases (CCD, multiminicore disease, congenital myopathy with cores and rods) and MH,^19–22 therefore, CNMDU1 may be the fifth disease linked to RYR1 mutations.

Among four mutations that we identified, c.14582G→A (p.R4861H) was previously associated with CCD in Europeans,¹⁴ and the c.14761_14762TT→AC (p.F4921T) was previously reported in the father of Patient 4, who had CCD.¹¹ Two other mutations, c.13013_13032del20 (p.A4338fs) and c.14680G→C (p.A4894P), were novel ones. The 20-bp deletion mutation is predicted to cause a frame-shift, leading to a premature stop codon and removal of the C-terminal 464 amino acid residues from the protein.

The predicted transmembrane helices have been described as M1 to M10.^23,24 However, the recent in vitro study suggested that M1 to M4 regions are actually located in the cytosol and that only M5 to M10 are the transmembrane domains.²⁵ According to this model, the deletion mutation identified in Patient 1, which was predictably located between M3 and M4, should truncate the protein after the M3 region, losing all transmembrane domains. Furthermore, the previous study showed that the mutant RYR1 truncated after M3 region can still exist in the cytosol even without being anchored to the membrane.²⁵ Indeed, the truncated RYR1 protein was present in the patient's muscle as confirmed in Western blot analyses. Our results raise a possibility that the truncated RYR1 mutant may somehow be associated with the wild-type RYR1 and disrupt its function. However, the limited amount of the sample did not allow further investigation to clarify the interaction between wild-type and mutant RYR1.

Interestingly, c.14680G→C (p.A4894P) affects the same nucleotide and amino acid site with c.14680G→A (p.A4894T), which was found in the MH patient in our previous study.¹³ Pathologically, the MH patient with p.A4894T had a normal mosaic pattern of fiber type distribution and not uniform type 1 fiber. Core-like structure was observed in only a few fibers. Proline differs from other amino acids in its structure of imino acid; that is, the side chain of proline forms a cyclic structure.²⁶ Therefore, a single amino acid change from alanine to proline may lead to a different structural and thereby functional change in RYR1 from that in p.A4894T, resulting in uniform type 1. It is an interesting issue as to whether p.A4894P mutation is also associated with MH. However, no sample was available for in vitro contraction²⁷ or calcium-induced calcium release test²⁸ in Patient 3.

We did not find any RYR1 mutation in six patients in our cohort, suggesting the presence of another causative gene for CNMDU1 and the genetic heterogeneity of the disease, even though there still remains a possibility that mutations may exist in unexamined regions such as the majority of introns. We did not find any RYR1 mutation in two patients having congenital myopathy with marked type 1 fiber predominance in which type 1 fibers account for less than 99%, suggesting that the RYR1 mutation in the C-terminal domain may be tightly associated with uniform type 1 fiber, namely, >99% type 1 fibers, albeit a greater number of patients are needed to make a definite conclusion.

We could amplify only two exons in the C-terminal domain in DNA of the patients first reported to have CNMDU1.¹ Although the original patients were clinically similar to our patients with RYR1 mutation, in terms of early onset, mild muscle weakness, delayed motor milestones, and pathologic features, their age at the time of biopsy (ages 9 and 12) was higher in comparison with our patients, raising the possibility that the original patients may have had a genetically distinct disorder.

Excitation–contraction (EC) uncoupling caused by RYR1 mutation is thought to be closely associated with CCD.²⁹ In vitro studies have shown that two RYR1-binding proteins, FKBP1A and CACNA1S, directly participate or modulate EC coupling in skeletal muscle.^30,31 In addition, 1% of MH patients have mutations in the RYR1-binding region in CACNA1S.³² Therefore, we sequenced FKBP1A and CACNA1S, but we did not find a mutation in any patient, suggesting that these genes may not or only rarely be associated with CNMDU1.

In our study, CNMDU1 patients with RYR1 mutations have mild clinical features compared with those without mutations, in terms of poor fetal movement, asphyxia, infantile hypotonia, respiratory distress, mental retardation, myopathic facies, and high arched palate. This supports the idea that CNMDU1 may be genetically heterogeneous. Most remarkably, none of the patients with RYR1 mutations had mental retardation, whereas five of six patients without RYR1 mutations had it. Three of five patients had ventricular dilatation or brain atrophy on brain imaging, suggesting that the mental retardation might occur with a perinatal history of asphyxia or another primary abnormality of unknown origin.

Regarding pathologic findings, CNMDU1 patients either with or without mutations in RYR1 had similar myopathic changes: mild to marked variation in fiber size. The majority of type 2 fibers, albeit few in number, found in our patients were type 2C, indicating that mature type 2 fibers are even fewer. Patients without RYR1 mutations had more pathologic variation than those with mutations, suggesting that those without mutations might have genetically different causes.

Solely from the clinical features, it is difficult to differentiate between CNMDU1 patients with RYR1 mutations and CCD patients with C-terminal mutation in RYR1. Both groups of patients show muscle weakness and delayed motor milestones. The frequency of asphyxia, mental retardation, myopathic facies, high arched palate, and skeletal deformities is similar. Furthermore, uniform type 1 fiber is a characteristic pathologic finding in both groups.¹¹ Two mutations of c.14582G→A (p.R4861H) and c.14761_14762TT→AC (p.F4921T) were identified in both CNMDU1 and CCD patients, and all the patients showed type 1 fiber uniformity despite the absence or presence of cores. This result suggests that type 1 fiber uniformity is closely associated with C-terminal RYR1 mutation. Although additional study is required, there still remains a possibility that CNMDU1 and CCD are closely related diseases, regardless of the presence or absence of cores.

In support of this notion, the father of Patient 4 had CCD, while no cores were observed in the patient's sample.⁶ A similar family case was also reported: a 4-month-old girl had CNMDU1 in a family with CCD due to p.Y4864C mutation in exon 101 of RYR1.⁷ In both families, CNMDU1 was identified in younger children, whereas CCD was found in older family members. These findings suggest that the core may be formed later in the course of disease at least in some patients. Alternatively, cores may not be formed in CNMDU1 patients for factors that are yet to be known.

The fact that we were unable to find distinct pathologic changes other than type 1 fiber uniformity can be due to many possibilities. One is that CCD and CNMDU1 may be a part of a spectrum, as mentioned above. Interestingly, in all familial cases including the one previously reported by others, adults had CCD, whereas children showed CNMDU1, suggesting that cores might not be present in their early lives. In fact, age at biopsy in CNMDU1 patients with C-terminal mutations (3.2 ± 2.3 years, n = 4) was more than 1 year lower than that in CCD patients with C-terminal mutations (4.4 ± 3.0 years, n = 14) in our series,¹¹ although there is a significant overlap between the two age groups. Nevertheless, we have never found a case with muscle pathology falling between CNMDU1 and CCD, that is, uniform type 1 fiber with cores only in a few fibers. In addition, electron microscopic study of our patient (albeit only one was available) did not show any sign of core formation. These observations may cast some doubt on the notion that CNMDU1 and CCD are part of a spectrum.

Another possibility is that CNMDU1 is actually CCD, and the absence of cores in CNMDU1 may be attributed to the site of sampling. This, however, is less likely because we have sampled a wide range of sites. MRI studies in CCD have confirmed a distinct pattern of muscle involvement, mostly involving muscle of the lower extremities.³³ In our series, even biceps brachii, which is relatively spared based on clinical examination, actually shows core formation (data not published). The third possibility is that CNMDU1 is a distinct entity, and thus the absence of other pathologic findings may not be influenced by the choice of area sampled.

Previously, we have identified 14 CCD patients with heterozygous C-terminal mutation in RYR1.¹¹ Combining this number with that from this study, 4 of 18 patients (22%) with a heterozygous C-terminal mutation are associated with CNMDU1, suggesting that CNMDU1 may not be a rare condition at least among those with the RYR1 mutation in the C-terminal domain.

The pathomechanism for the development of uniform type 1 fiber is still unknown. Skeletal muscle fiber type formation is thought to be regulated by nerve control and succeeding intracellular signal transduction including Ca²⁺ release that lead to transcriptional activation of fiber type-specific genes. In slow muscle fibers, the lower amplitude and longer duration of Ca²⁺ transition facilitate activation of calcineurin, the Ca²⁺/calmodulin-dependent phosphatase, which dephosphorylates and activates two kinds of transcription factors; nuclear factor of activated T cell (NFAT) and myocyte enhancer factor 2 (MEF2) resulted in significant activation of slow myosin heavy chain 2 gene (slow MyHC2) promoter, expressed in avian skeletal muscle.^34–36 In contrast, in fast fibers, high-amplitude calcium sparks induced by infrequent phasic firing of the motor nerves are insufficient to keep activation of calcineurin. When calcineurin is inactivated, phosphorylated NFAT cannot enter the nucleus and the slow fiber-specific program is down-regulated, resulting in the predominant transcription of genes encoding fast fiber-specific proteins.³⁷

The inhibition of RYR1 by ryanodine treatment has been reported to induce the activation of slow MyHC2 promoter in fast muscle fibers via NFAT- and MEF2-dependent transcriptional pathway.³⁸ It strongly suggests that the loss of function of the RYR1 channel contributes to slow MyHC2 gene expression. This naturally raises a possibility that the C-terminal mutations found in this study may cause loss of function of RYR1, which leads to the activation of slow fiber-specific program in fast fibers.

In fact, all missense mutations found in CNMDU1 in this study are located in the pore-forming segment of RYR1 (figure 3). This pore-forming segment is located close to the luminal end of the RYR1 channel and is supposed to form the selectivity filter, which plays a critical role in the selection of permeating ions.^39,40 According to this hypothetical model, binding of ryanodine would be expected to change the conformation of this selectivity filter and the inner conduction pore.⁴¹ Therefore, mutations in the pore-forming segment may well alter the interaction of the pore helix and the selectivity filter. Loss of function of RYR1 may explain the mechanism of type 1 fiber uniformity as in the model of RYR1 inhibition by ryanodine, although further investigations are necessary to elucidate the precise pathomechanism for CNMDU1.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3 Location of mutation sites in the pore-forming region of RYR1 protein

Putative positions of the three missense mutations (filled circle) of RYR1 in congenital neuromuscular disease with uniform type 1 fiber found in this study are shown. The putative pore-forming segments from two RYR1 monomers are illustrated as TM3 (outer helix), TM4 (inner helix),³⁹ same as M8 and M10,²³ and the pore helix connected together.³⁹ The selectivity filter is indicated. R4861H and A4894P are located at each end of TM3 and the pore helix, whereas F4921T is situated at the beginning of TM4.

Acknowledgment

The authors thank Drs. Mieko Yoshioka (Section of Pediatric Neurology, Kobe City Pediatric and General Rehabilitation Center for the Challenged), Tatsuro Nobutoki (Department of Pediatrics, National Mie Hospital), Yoshinori Kohno (Department of Neonatology, Gifu Prefectural Gifu Hospital), Yasuyuki Suzuki (Division of Anesthesia, Department of Anesthesia and ICU, National Center for Child Health and Development), Yoshihiro Maegaki (Division of Child Neurology, Institute of Neurological Sciences, Faculty of Medicine, Tottori University), and Mariko Kodama (Department of Pediatrics, National Rehabilitation Center for Disabled Children) for clinical information on the patients; Drs. May Malicdan, Mina Astejada, and Sherine Shalaby (NCNP) for their critical comments on the manuscript; and Ms. Kumiko Murayama and Megumu Ogawa (NCNP) for their technical assistance. They also thank Prof. Shigeru Tsuchiya and Drs. Kazuie Iinuma, Kazuhiro Haginoya, and Mitsutoshi Munakata (Department of Pediatrics, Tohoku University School of Medicine) for their guidance and support. The ryanodine receptor antibody developed by Judith Airey and John Sutko was obtained from the Developmental Studies Hybridoma Bank, developed under the auspices of the NICHD, and maintained by Department of Biologic Sciences, the University of Iowa, Iowa City.