150th ENMC International Workshop: Core Myopathies, 9–11th March 2007, Naarden, The Netherlands

1. Introduction and overview 

Eighteen clinicians and basic scientists from 6 countries convened from the 9th to the 11th of March 2007 in Naarden, The Netherlands for the 150th ENMC sponsored Workshop on Core Myopathies. Members of the ENMC Consortium who attended the 150th ENMC workshop are indicated below: Robert Dirksen, Brigitte Estournet-Mathiaud, Ana Ferreiro, Susan Hamilton, Heinz Jungbluth, Isabelle Marty, Gerhard Meissner, Nicole Monnier, Francesco Muntoni, Ichizo Nishino, Feliciano Protasi, Ros Quinlivan, Caroline Sewry, Volker Straub, Susan Treves, Thomas Voit and Francesco Zorzato.

The previous ENMC workshops on Central Core Disease (CCD) in January 2001 and on Multi-minicore Disease (MmD) in May 2000 and November 2002 had lead to collaborations between the participating groups, resulting over the ensuing years in a rapid advance in the understanding of these congenital myopathies.

Following identification of mutations in the selenoprotein N (SEPN1) gene as the cause of the most instantly recognizable classic form of MmD [1], more recently other and clinically widely diverse subgroups of the condition have been associated with recessive mutations in the skeletal muscle ryanodine receptor (RYR1) gene [2], [3], [4], [5], [6], encoding the principal sarcoplasmic reticulum (SR) calcium release channel with a crucial role in excitation–contraction coupling (EC coupling). Because of the marked clinical and histopathologic overlap between RYR1-related recessive MmD and dominantly inherited CCD due to mutations in the same gene, clinical, histopathologic and genetic features of both conditions were covered in the present workshop on core myopathies. With the genetic background of MmD and CCD now largely established, a strong emphasis of this workshop was placed on the physiology and pathophysiology of intracellular calcium metabolism and of EC coupling associated with mutations in the RYR1 and SEPN1 genes, and animal models of those conditions which may provide a basis for future therapeutic interventions.

After the welcoming address by Pascale Guicheney on behalf of Kate Bushby, Research Director of the ENMC, the workshop was introduced by Ana Ferreiro and Heinz Jungbluth, who gave an overview over developments since the most recent workshop on MmD in November 2002.

Mutations in the skeletal muscle ryanodine receptor (RYR1) gene, initially associated with the dominantly inherited malignant hyperthermia susceptibility (MHS) trait and the congenital myopathy central core disease (CCD), have now been associated with a wide range of congenital myopathy phenotypes including subgroups of Multi-minicore Disease (MmD) [1], [3], [4], [5], [6] and centronuclear myopathy (CNM) [7]. Whilst typical dominantly inherited CCD has been associated with a fairly mild phenotype characterized by hip girdle weakness and absence of overt bulbar, extraocular and respiratory involvement, more severe presentations and diverse clinical features including arthrogryposis, external ophthalmoplegia, substantial bulbar involvement and respiratory compromise have been recently reported and predominantly attributed to recessively inherited RYR1 mutations. Corresponding to the more generalized clinical muscle involvement, recessive RYR1 mutations are associated with a more diffuse increase in signal intensity on muscle MRI of the lower limb [5] compared to the highly selective pattern observed in patients with dominant RYR1 mutations [8]. Whilst dominant RYR1 mutations associated with a typical CCD phenotype localize almost exclusively to the RYR1 C-terminus, predominantly exons 100–101, recessive RYR1 mutations appear widespread throughout the RYR1 coding sequence but more extensive data are currently only emerging. More recently, expression of heterozygous mutations on a haploinsufficient background due to epigenetic allele silencing of the RYR1 gene has been reported as a novel mechanism in the pathogenesis of core myopathies [9]. Whilst the pathogenesis of MH and CCD has been extensively investigated, functional studies on MmD-related RYR1 mutants are currently still limited but suggest a wider range of underlying mechanisms.

2. Histopathologic spectrum and differential diagnosis of core myopathies 

Caroline Sewry gave an overview of the pathological changes in muscle biopsies from 28 cases with mutations in the RYR1 gene and 9 cases with mutations in the SEPN1 gene, based on the experience at the Hammersmith Hospital Neuromuscular Unit and Robert Jones & Agnes Hunt Hospital, Oswestry. Caroline Sewry emphasised that cores per se are not specific and can be seen to varying degrees in several neuromuscular disorders, including muscular dystrophies and neurogenic disorders, where they sometimes appear as targets with a pronounced rim. Seventeen of the RYR1 cases had autosomal dominant inheritance (11 de novo), 5 cases had recessive inheritance, and an additional 6 cases had haploinsufficiency with monoallelic RyR1 expression in muscle. The SEPN1 cases had homozygous or heterozygous recessive mutations.

Large classical cores lacking oxidative enzyme stains and extending an appreciable length down the fibre were present in several of the dominant RYR1-related cases, particularly those with mutations in the ‘hot spot’ RYR1 C-terminal region; the position of these cores may be central, peripheral or eccentric, even within the same section. The spectrum associated with the same mutation may be wide, as demonstrated in three individuals from the same family: at 4 months of age the female proband showed no cores, whilst her 3-year-old brother showed large classical cores and the mother unevenness of stain. The absence of cores may relate to age and this needs to be considered when assessing biopsies. Recessive RYR1 mutations seemed to be more frequently associated with focal small multiple cores (minicores) or large focal cores that may extend across the width of the fibre, the latter appearance often seen in cases with ophthalmoplegia; the distinction between a ‘minicore’ or ‘central core’ is therefore not always clear and the term ‘core myopathy’ is now often used by the Hammersmith and Oswestry groups.

Ultrastructurally, cores in most RYR1 cases are of the structured type, with moderate preservation of myofibrillar structure; in occasional cases, however, they may be mainly of the unstructured type lacking ATPase activity. Cores lack phosphorylase, glycogen and mitochondria, and the area without the latter may be more extensive than the area of disrupted myofibrils. Some cores may be rimmed by glycogen, mitochondria and various proteins and/or show accumulation of proteins, including desmin, γ-filamin, syncoilin, myotilin, SERCA2, calsequestrin, dihydropyridine receptor and RyR1 protein. Phalloidin binding often highlights the hypercontraction of the myofibrils rather than an accumulation of F-actin. The accumulation of proteins and rimming of the cores varies between cores, even within one biopsy and no consistent patterns were observed in this series of cases.

In addition to core structures, most RYR1 cases show pronounced slow/type 1 fibre type predominance or uniformity, with the majority of fibres expressing slow myosin and only a few fibres co-expressing slow and fast myosin isoforms. Variation in fibre size is often not marked in RYR1 cases but very small fibres (often <5μm) expressing neonatal myosin are sometimes seen. The presence of adipose tissue is common and may be extensive; in some cases it may also be associated with an increase in fibrous tissue and resemble a congenital muscular dystrophy, but there is usually no fibre necrosis. Sampling of muscles with varying degrees of differential involvement may result in variable amounts of adipose and fibrous tissue in a biopsy.

Internal nuclei, some of which are central or multiple, are also a common feature; the presence of central nuclei can resemble myotubular/centronuclear myopathy and the RYR1 gene should be considered when this pathology is present, in addition to the MTM1, DNM2 and DM1 genes typically associated with this appearance. Internal nuclei are also common at myotendinous junctions and fibres with multiple splits resembling these areas are common in RYR1 cases, although they are not specific. Some cases may show clusters of nemaline rods in occasional fibres but these are not marked features. No cases in this series with proven mutations resembled the core-rod myopathy cases reported in the literature.

In comparison, cases with SEPN1 mutations often show less marked pathology with preservation of the two-fibre pattern, although moderate type 1 predominance can occur. Unevenness of oxidative enzyme stains or focal cores are present in both fibre types. Internal nuclei may also occur but are not usually as pronounced as in RYR1 cases. Similarly, a mild or moderate increase in adipose and endomysial connective tissue may occur but these are not as extensive as in RYR1 cases. The focal cores in SEPN1-related cases are not rimmed by accumulation of proteins or glycogen, but accumulation of some proteins may occur in some fibres, for example desmin and/or myotilin.

Ana Ferreiro presented morphological and immunohistochemical data from the core myopathy biopsies analysed at the Institut de Myologie (Paris). In RYR1-related core myopathies, most of the findings were fully consistent with those observed in the Hammersmith and Oswestry series described above. Well-delimited or rimmed cores, more often peripheral and multiple than central and unique, running most of the length of type 1 fibres, were highly characteristic of dominant CCD due to heterozygous RYR1 mutations, and have never been observed in classical, recessive MmD due to SEPN1 mutations. Conversely, short, poorly defined minicores were associated with dominant or recessive RYR1 mutations as well as with recessive SEPN1 mutations. Furthermore, age-related morphological evolution from minicores to cores has been documented on sequential deltoid biopsies from the same patient with a homozygous RYR1 mutation [2]. These findings underline that, in RYR1-related myopathies, there is a morphological continuum between minicores and cores, and support the global denomination of “core myopathies” for this group of congenital muscle disorders. The form of MmD with ophthalmoplegia associated with homozygous RYR1 mutations often showed cores which were short on the longitudinal axis but had a large transversal diameter (spanning most of the transverse fibre section) with abundant central nuclei [4].

A retrospective analysis of 42 muscle biopsies from patients with homozygous or compound heterozygous mutations of the SEPN1 gene confirmed that several categories of morphological lesions can be associated with SEPN1-related myopathy. Both fibre types were usually present; type 1 predominance and relative hypotrophy (fibre type disproportion) were observed in 97% cases. Most samples (83.3%) showed multiple minicores in type 1 and type 2 fibres, either isolated (45%), associated with dystrophic findings (21%), or with Mallory body-like inclusions (MBs, 17%). When present, dystrophic lesions were generally mild. MBs (hyaline inclusions containing desmin, with three components on EM) affected usually less than 10% of the fibres. There was no correlation between the morphological pattern and the type or localization of SEPN1 mutations. To some extent, morphological patterns correlated with the muscle biopsied: minicores and dystrophic lesions were present in all the muscles sampled, but a dystrophic pattern was more common in axial muscles.

Several proteins such as desmin, α B crystalline or γ filamin accumulate in cores, minicores and targets, in a non-specific way. To search for distinctive markers, the immunolocalization of six proteins of the Ca2+-release complex was analysed in the biopsies from 12 genetically-characterized core myopathy cases. In 7 cases with RYR1 mutations (6 CCD, 1MmD), RyR1 was depleted from the cores; in contrast, the other proteins of the sarcoplasmic reticulum (calsequestrin, SERCA1/2 and triadin) and the T-tubule (dihydropyridine receptor-α1 subunit) were accumulated within or around the lesions, suggesting an original modification of the Ca2+-release complex protein arrangement. Conversely, all Ca2+-related proteins were distributed normally in 5MmD cases with SEPN1 mutations. These results provide an appropriate tool to orientate the differential and molecular diagnosis of core myopathies, and suggest that different pathophysiological mechanisms lead to core formation in SEPN1- and in RYR1-related core myopathies.

Ichizo Nishino presented the pathological spectrum of three RYR1-related conditions in a large cohort of Japanese patients: malignant hyperthermia susceptibility (MHS), central core disease (CCD) [10] and congenital neuromuscular disease with uniform type 1 fibres (CNMDU1). In the Japanese cohort both CCD with typical pathological features and CNMDU1 were associated with C-terminal mutations; in contrast, all cases with atypical cores and the majority of patients with MHS had mutations in non-C-terminal regions. In addition, isolated single fibre analysis in two patients with typical CCD with C-terminal mutations did not reveal increased calcium-induced calcium release (CICR), suggesting that, albeit limited number of cases, some CCD causing C-terminal mutations may not be associated with MHS. Interestingly, histopathologic changes in one of the few C-terminal mutations associated with MHS, p.A4894T, were characterized by a normal mosaic pattern of fibre types and core-like structures in only a few fibres; in contrast, p.A4894P was associated with CNMDU1, suggesting that substitution to a different amino acid, even at the same position, may result in a different phenotype.

3. Clinical phenotype of core myopathies with or without RYR1 mutations 

Ros Quinlivan presented clinical features of nine patients from five families with Multi-minicore Disease but no confirmed SEPN1 or RYR1 mutation. Clinical features in those families included: arthrogryposis with distal involvement in the upper limbs, adducted thumbs or camptodactyly, and ptosis; cleft palate was present in two unrelated children. In two families rigid spine was a feature associated with restrictive respiratory insufficiency, one child has had scoliosis surgery. Muscle weakness was mild predominantly affecting truncal and pelvic girdle muscles. Two children had mild educational difficulties and one had epilepsy. One consanguineous family presented with a different phenotype affecting the mother and two children. Clinical features included: ophthalmoplegia, ptosis and scoliosis associated with generalized muscle atrophy and poor feeding. Electromyography revealed myopathic features. All of the patients have normal CK, there was no cardiac involvement and none of the patients to date required non-invasive ventilation. Those findings suggest either non-exonic mutations in the SEPN1/RYR1 genes or genetically distinct phenocopies of patients with mutations in above genes.

Heinz Jungbluth and Francesco Muntoni presented 3 children from two unrelated families with clinical and histopathologic features of a RYR1-related congenital myopathy but without confirmed RYR1 mutation. Heinz Jungbluth reported a 7-year-old boy who presented from birth with hypotonia, bulbar involvement and arthrogryposis following a pregnancy complicated by oligohydramnios and reduced fetal movements; his further course was characterized by developmental delay and slow progression. On examination, distribution of weakness and wasting pronounced in the axial and shoulder girdle muscles was similar to that observed in patients with recessive RYR1 mutations. Muscle biopsy showed marked type 1 predominance, increase in internal nuclei and unevenness of oxidative stain. Although no mutation could be detected on sequencing of the entire RYR1 coding sequence, RyR1 protein was markedly reduced on Western blotting.

Volker Straub presented patients from the neuromuscular group in Newcastle that showed clinical symptoms and muscle biopsy findings suggestive of CCD or MmD but without mutations in the SEPN1 gene or mutations in exons 95, 100, 101 and 102 of the RYR1 gene. In addition to weakness of the axial muscles several patients showed clinical symptoms that are not classically associated with CCD or MmD. A now 12-year-old boy with characteristic findings of CCD in his muscle biopsy, congenital onset of weakness and respiratory insufficiency showed marked distal hyperlaxity that resembled patients with Ullrich congenital muscular dystrophy (UCMD); however, no abnormalities of collagen VI expression in the patient’s muscle biopsy or in his fibroblasts were detected. It was felt that there is a clear clinical overlap between patients with CCD, MmD and UCMD. A 24-year-old lady with cores in her biopsy, generalized weakness and rigidity of her spine presented with slurred speech that has previously been described in patients with myofibrillar myopathies; although the biopsy did not show vacuoles or desmin accumulation there may well be an overlap between patients with core diseases and myofibrillar myopathies. A 35-year-old patient with congenital onset of weakness and cores in his biopsy (confirmed on electron microscopy) presented with external ophthalmoplegia and excessive sweating (hyperhydrosis); he also experienced difficulties with temperature regulation. It is not clear whether mutations in the RYR1 gene can affect the autonomous nervous system. Additional features in other patients with a CCD/MmD phenotype but no mutation in either the SEPN1 gene or the common exons of the RYR1 gene were learning difficulties; all of these patients had a normal serum CK activity

4. Genetic diagnosis of RYR1-related core myopathies 

Nicole Monnier presented genetic findings in a French cohort of 229 unrelated patients presenting with core myopathies ranging from congenital onset with severe phenotype to milder classical CCD. Muscle biopsies from those patients were characterized by core lesions showing variable localization, size, length and number within the muscle fibres and by frequent association with centrally located nuclei and predominance of type I fibres. RYR1-related diseases including malignant hyperthermia (MH) with cores, exertional heat stress syndrome and centronuclear myopathies were excluded. Most cases were sporadic (156) while the remaining patients had a family history of dominant (59) or recessive (13) core myopathy. Search for mutations in the RYR1 gene was performed by C-terminal screening of exons 92 to 106 and, when a muscle biopsy was available, by cDNA sequencing in 52 cases.

A single dominant mutation was identified in 54 cases, two recessive mutations each in 27 cases and one mutation of unknown significance in 20 cases. Furthermore, no mutation was identified after cDNA sequencing in 9 cases.

RYR1 mutations identified in dominant core myopathies (n=61) were missense mutations and in-frame deletions localized in the C-terminal domain of the protein, mostly affecting transmembrane and luminal domains of the calcium channel. Sixty percent of those mutations were recurrent and 26% were neomutations.

In contrast, RYR1 mutations identified in recessive core myopathies (n=54) were distributed along the entire coding sequence of the gene. Seventy seven percent were missense and in-frame deletions while 23% were truncating mutations leading to loss of protein (Tr). Three cases were compound heterozygous for 2 Tr mutations associated with strong RyR1 depletion. All patients presented with a severe neonatal phenotype. Eight cases were compound heterozygous for 1 Tr mutation and 1 missense mutation resulting in hemizygous expression of the mutated RyR1 protein. The variability observed in phenotype severity ranging from mild to severe was likely to be related to the nature of the missense mutation. The remaining 14 patients were homozygous (n=2) or compound heterozygous (n=12) for missense mutations that included a striking high frequency of MH mutations. We have also identified a recessive case of severe central core myopathy in a dominant CCD family.

Francesco Muntoni presented genetic findings in a large cohort of core myopathy patients from the United Kingdom [9]. In order to characterize the spectrum of congenital myopathies associated with RYR1 mutations, his group investigated a cohort of 44 patients from 28 families with clinical and/or histopathologic features suggestive of RYR1 involvement, and identified 25 RYR1 mutations, 9 of them novel, including 12 dominant and 13 recessive mutations. With only one exception, dominant mutations were associated with a CCD phenotype, prominent cores, and predominantly occurred in RYR1 C-terminal exons 101 and 102, whereas the 13 recessive RYR1 mutations were distributed evenly along the entire RYR1 gene and were associated with a wide range of clinico-pathological phenotypes. Protein expression studies in 9 cases suggested a correlation between specific mutations, RyR1 protein levels and resulting phenotype: in particular, whilst patients with dominant or some recessive mutations and typical CCD phenotypes appeared to have normal RyR1 expression, individuals with more generalized weakness, multi-minicores and external ophthalmoplegia had a severe depletion of the RyR1 protein. The phenomenon of severe protein depletion was observed in patients compound heterozygous for a recessive mutation and an apparently normal but silenced allele, providing evidence for the pathogenic role of allele silencing at the RYR1 locus when associated with recessive RYR1 mutations. These data indicate complex genotype–phenotype correlations associated with RYR1 mutations differentially affecting assembly and function of the RyR1 calcium release channel.

Heinz Jungbluth presented findings in a cohort of patients with a core myopathy associated with monoallelic expression of a heterozygous RYR1 missense mutation due to epigenetic allele silencing of the second RYR1 allele [6]. These patients had consistent clinical features characterized by external ophthalmoplegia and predominant axial and proximal weakness; muscle biopsies were characterized by marked type 1 uniformity and core-like structures often spanning the entire fibre diameter on longitudinal sections. Haplotyping studies in those families were consistent with recessive inheritance, however, sequencing of the entire RYR1 coding sequence in patients demonstrated a single RYR1 missense mutation invariably inherited from an unaffected father. Whilst these mutations were expressed heterozygously on the genomic DNA level, studies on cDNA derived from skeletal muscle tissue showed monoallelic expression of the mutated RYR1 allele, consistently due to maternal allele silencing and suggestive of genomic imprinting. Genomic imprinting is an epigenetic phenomenon resulting in allele-specific silencing according to parental origin; this may be tissue-specific, developmentally regulated and also polymorphic in the general population. Tissue-specific allele silencing was implicated by failing to identify monoallelic RYR1 expression in skin fibroblasts or lymphocytes from the same patients; further studies on normal fetal tissues suggested tissue-specific RYR1 allele silencing in skeletal muscle, brain, spinal cord, eye and intestine in 10% of a normal population which was not present in adult skeletal muscle. Reactivation of the silenced maternal allele by the DNA methyltransferase inhibitor 5′-aza-deoxycytidine (5-azaC) in cultured skeletal muscle myoblasts from patients implicated DNA hypermethylation as the underlying mechanism, however, bisulfite sequencing failed to show the differentially methylated region (DMR) within the 5′ region of the RYR1 gene, suggesting a control region further afield. These results suggested that RYR1 is polymorphically silenced in a tissue-specific manner during human fetal development, and that in a proportion of recessive core myopathy patients, the silencing of one RYR1 allele contributes to unmask the phenotype in the presence of recessive paternal mutations. Whilst consistent silencing of the maternal allele in the original cohort of patients strongly implicated imprinting as the mechanism underlying RYR1 allele silencing, more recent identification of a patient with a maternally inherited RYR1 mutation and silencing of the paternal allele indicates the existence of alternative mechanisms with similar effect.

5. Calcium handling and excitation–contraction coupling: Proteins involved and animal models 

Feliciano Protasi presented his work on spatial relationships between the key proteins of excitation–contraction EC coupling, the skeletal muscle ryanodine receptor (RyR1) and the dihydropyridine receptor (DHPR), in skeletal muscle. In muscle fibres, an extremely well organized system of tubules and vesicles, collectively named sarcotubular system, is able to finely control the cytoplasmic Ca2+ concentration. The sarcotubular system is formed by two separate systems of membranes: exterior membranes (sarcolemma and its invaginations, the transverse-tubules, T-tubules) and internal membranes (the sarcoplasmic reticulum). In adult muscle, SR and T-tubules form highly specialized intracellular junctions called Calcium Release Units (CRUs) or triads. CRUs are formed by three membranous elements: a central T-tubule and two apposed terminal cisternae of the SR. At the level of triads, the electrical signal coming from the neuromuscular junction is transduced into a release of Ca2+ from the SR by a mechanism called excitation–contraction (EC) coupling [11].

Ryanodine receptors (RyRs) and dihydropyridine receptors (DHPRs), are the two major proteins involved in EC coupling. RyRs are composed of four identical subunits that form the Ca2+ release channels of the SR [12]. They are specifically located at the junctional face of the SR terminal cisternae and their large cytoplasmic domain, i.e. the foot, appear in electron microscopy (EM) as evenly spaced densities in the junctional gap between T-tubule and SR membrane. RyRs are closely associated with each other forming ordered arrays in the SR junctional domains. DHPRs are L-type voltage-gated Ca2+ channels of exterior membranes, which are responsible for initiating EC coupling by sensing the change in membrane polarity of the sarcolemma/T-tubule. Like RyR1s, DHPRs form ordered arrays in skeletal muscle, specifically positioned into group of four receptors, the tetrads, which form a small square. Identification of tetrads with DHPRs was proven in dysgenic muscle (α1sDHPR−/−) where tetrads are missing [12]. Tetrads are structurally linked to the subunits of alternate RyRs, so that each DHPR is located immediately above each of the RyR subunits; this highly ordered tetrads/RyRs association represent the structural base that allows Ca2+ independent (or mechanical) EC coupling in skeletal muscle fibres. Tetrads are specifically formed by the association of α1sDHPR subunits (the skeletal isoform of the DHPR) with the subunits of RyRs type 1, the predominant RyR isoform expressed in skeletal muscle [13], [14]. α1sDHPR is unable to form tetrads with other RyR isoforms such as RyR2 and 3, indicating that only the specific interaction of α1sDHPR with RyR1 allows tetrads formation and support skeletal type EC coupling [15], [16].

In muscle, rapid changes in intracellular [Ca2+] concentration, or Ca2+ transients, control the contraction and relaxation of myofibrils. However, prolonged elevation of intracellular [Ca2+] above 10μM is deleterious to the cell life and can activate apoptosis. There is though a narrow window of Ca2+ dysregulation that can cause disease in both cardiac (i.e. cardiac arrhythmias, heart failure) and skeletal (i.e. MH, CCD and MmD) muscle rather then apoptosis.

Francesco Zorzato presented his work on accessory proteins of the excitation–contraction coupling (EC coupling) machinery. In striated muscle, activation of contraction is initiated by membrane depolarization during an action potential, which causes release of Ca2+ stored in the sarcoplasmic reticulum (SR) in a process called excitation–contraction coupling (EC coupling). EC coupling occurs via a highly sophisticated supramolecular signaling complex at the junction between the SR and the transverse tubules, which transduces the action potential into a transient increase of the myoplasmic calcium concentration. The major components of the EC coupling supramolecular complex are the dihydropyridine receptor (DHPR), the ryanodine receptor (RyR), and calsequestrin (CS) which serve as voltage sensor, SR Ca2+ release channel, and SR Ca2+ storage protein, respectively. However, highly purified sarcoplasmic reticulum membrane fraction revealed that its protein composition is even more complex. In addition to DHPR, RyR and CS, there are other protein components including triadin, the histidine rich Ca2+ binding protein (HCP), 90kDa JFP, mitsugumins, junctophilin, junction, junctate, JP-45 and the 32kDa ADP/ATP carrier. On the basis of their localization in the SR membrane, these proteins are thought to be involved in keeping the functional integrity of the spatial organization of the SR membrane. Francesco Zorzato’s group have discovered JP-45, a protein which is made up by a cytoplasmic domain, a single transmembrane segment followed by an intralumenal domain enriched in positively charged amino acids. JP-45 interacts with calsequestrin and with the DHPR receptor via the cytoplasmic loop connecting repeat I to repeat II of the α1s subunit. Depletion of JP-45 by RNA silencing or by JP-45 gene disruption reduces DHPR charge movements and calcium release from the SR. Because JP-45 modulates EC coupling by altering the functional expression of the DHPR receptor, it is a candidate gene to consider in clinical conditions characterized by a decrease of muscle strength.

Isabelle Marty presented her work on triadin, a protein associated with the skeletal muscle ryanodine receptor (RyR1) in the calcium release complex. Her group has identified four triadin isoforms in rat skeletal muscle, Trisk 95, Trisk 51, Trisk 49 and Trisk 32 and could demonstrate that adenoviral induced overexpression of Trisk 95 in rat skeletal muscle primary cultures induced blocking of excitation–contraction coupling, and inability of membrane depolarization to induce intracellular calcium release in absence of external calcium. In contrast, adenoviral induced overexpression of Trisk 51, also associated with RyR and involved in the calcium release complex, did not induce similar modifications when overexpressed, suggesting that Trisk 95 and Trisk 51 have different functions despite similar cellular localization. Trisk 95 and Trisk 51 are also expressed in human skeletal muscle. Studying expression of RyR, a huge reduction in RyR levels without equivalent modifications in the other proteins of the calcium release complex (DHPR, triadin) was observed in few CCD patients; as observed for Trisk 95 overexpression in rat myotubes, this could lead to a modification in the stoichiometry of the proteins within the calcium release complex and to a dysfunction of the complex possibly due to an excess of triadin compared to RyR. Trisk 49 and Trisk 32 are not associated with the calcium release complex, and do not co-localize with RyR, however, Trisk 32 is associated with IP3R as demonstrated using co-immunoprecipitation. Both Trisk 32 and Trisk 49 could have an anchoring function, connecting the sarcoplasmic reticulum to adjacent structures in order to maintain the organization of the sarcoplasmic reticulum in the muscle cell during contraction and relaxation.

Susan Hamilton presented her work on a mouse model with a mutation (Y522S) which, in humans, produces Malignant Hyperthermia (MH) with cores. The mutant mice are MH susceptible (MHS) and develop a myopathy with aging, but, unlike humans with the corresponding mutation, they do not develop cores in their muscle nor do they show type I fibre type predominance. The enhanced susceptibility to MH responses in these mice is initiated by a Ca2+ leak from the sarcoplasmic reticulum arising from the mutation affecting the RyR1 receptor: increased cytosolic Ca2+ initiates the production of reactive oxygen species, leading to pronounced oxidative stress in the muscle; redox modification of the mutant RyR1 further sensitizes RyR1 to activation, creating a feedforward cycle that increases the probability of an MH response and damages muscle. Antioxidants such as N-acetylcysteine partially reverse these effects.

6. Pathophysiology of RYR1-related core myopathies 

Gerhard Meissner presented the work of his group on single channel properties of recombinant CCD and MmD mutant skeletal muscle ryanodine (RyR1) receptors. Excitation–contraction (EC) coupling in skeletal muscle involves activation of ryanodine receptors (RyR1s), resulting in the rapid release of Ca2+ from the sarcoplasmic reticulum (SR). Previous work has shown that SR Ca2+ release is impaired by point mutations in RyR1 that cause Central Core Disease (CCD) and Multi-minicore Disease (MmD). Gerhard Meissner’s group studied the consequences of these mutations on RyR1 function, following their expression in human embryonic kidney (HEK293) cells and incorporation in lipid bilayers. Five CCD mutants in the C-terminal pore region of RyR1 and one N-terminal MmD mutant all showed variably decreased K+ conductances, loss or greatly decreased Ca2+ conductance, and loss of Ca2+-dependent gating. RyR1 is a large channel complex composed of four 565kDa subunits. Co-expression of wild type and mutant RyR1 subunits resulted in single channel activities that exhibited variable Ca2+ sensitivity, and intermediate K+ and Ca2+ conductances, which suggests that the tetrameric RyR1 complex was composed of wild type and mutant subunits. These data indicate that homozygous CCD and MmD RyR1 mutations abolish or greatly reduce SR Ca2+ release during EC coupling. These results further suggested that SR Ca2+ release is attenuated in patients heterozygous for these mutations.

Susan Treves presented her work on ryanodine receptor type 1 (RyR1) expression in cells of the immune system. For many years it was thought that the RyR1 receptor is mainly expressed in skeletal muscle, where it mediates Ca2+ release from the sarcoplasmic reticulum allowing muscle contraction to occur. In recent years however, more detailed investigations have revealed that the selective tissue distribution of the RyR1 is in fact more complex. Several investigators have shown that circulating leukocytes [17], leukocyte-derived cell lines [18] and immature mouse dendritic cells [19] express different RyR transcripts. Her group has shown that Epstein Barr Virus (EBV) immortalized B-lymphocytes express the transcript, the protein and the functional RyR1 Ca2+-release channel [20]; they have also demonstrated that immortalized B-lymphocytes from patients carrying RYR1 mutations linked to Malignant Hyperthermia and Central Core Disease are a good experimental model to study the functional impact of such mutations, offering several advantages over other methods such as ease at obtaining cell lines, no necessity to make large constructs and transfect cells, availability of a large number of cells which can be cultured, assayed and stored [21], [22], [23]. In recent years research of Susan Treves’ group has focused on the role of the RyR1 receptor in immune cells, particularly in human dendritic cells, which also express the RyR1 isoform. Dendritic cells are antigen presenting cells characterized by their ability to migrate to secondary lymphoid organs where they initiate primary immune responses. Immature DCs (iDCs) act as sentinels in peripheral tissues, continuously sampling the antigenic environment. Upon Toll-like receptor engagement by microbial products or tissue debris, iDCs undergo maturation and become the most potent antigen presenting cells. The mechanism and molecules involved in the early steps of Ca2+ release in dendritic cells have not yet been defined. Susan Treves presented data of her group supporting the expression of a functional RyR1 in an intracellular compartment (most likely the endoplasmic reticulum) of human dendritic cells; the RyR1 of human dendritic cells can be activated pharmacologically by typical agonists such as caffeine and 4-chloro-m-cresol, resulting in a calcium transient. They also found that RyR1 induced Ca2+ release together with the activation of Toll-like receptors by suboptimal concentrations of microbial stimuli, provide synergistic signals resulting in dendritic cell maturation and stimulation of T cell functions. Furthermore, they showed that the initial intracellular signaling cascade activated by RyR1 is different from that induced by activation of Toll-like receptors. Based on these findings they propose that under physiological conditions, especially when low “suboptimal” amounts of Toll-like receptor ligands are present, ryanodine receptor-mediated events cooperate in bringing about dendritic cell maturation.

Robert Dirksen presented investigations from his laboratory that focused on the cellular and molecular mechanisms by which CCD mutations in RyR1 reduce calcium release during EC coupling. Effects of different CCD mutations in RyR1 on steady-state intracellular calcium homeostasis and voltage-gated calcium release during EC coupling were characterized following expression in myotubes derived from RyR1-null (dyspedic) mice. Expression of RyR1 proteins harbouring CCD mutations from either N-terminal (R163C, I403M, and Y522S) or central (e.g. R2163H, V2168M, and R2435H/L) MH/CCD RyR1 regions resulted in variable degrees of increased SR calcium leak, as reflected in an increase in resting calcium, a reduction in releasable SR calcium content, and a reduction in the magnitude of depolarization-induced calcium release. In addition, the sensitivity of voltage- and caffeine-induced SR calcium release was also increased in myotubes expressing either N-terminal or C-terminal CCD mutations in RyR1. Similar results were observed in myotubes derived from knock-in mice heterozygous for the Y522S mutation. On the other hand, expression in dyspedic myotubes of CCD mutations in the pore-lining region of RyR1 (e.g. G4891R, R4893W, I4898T, G4899E/R, A4906V, and R4914G) resulted in a reduction in depolarization-induced calcium release without a significant change in resting calcium, SR calcium content, or release channel sensitivity to activation by either voltage or caffeine. Thus, CCD mutations in the RyR1 pore region reduced depolarization-induced calcium release from a full-complement SR calcium store, a process termed EC uncoupling. Since these pore-lining CCD mutations in RyR1 were also shown by Gerhard Meissner’s group to markedly reduce calcium conductance and calcium-dependent gating, the biophysical mechanism of EC uncoupling for these mutants results from a calcium permeation defect. To confirm this calcium permeation defect mechanism, the influence of the I4898T pore-lining mutation on increased SR calcium leak caused by the Y522S N-terminal mutation was determined. Indeed, incorporation of the I4898T pore mutation into Y522S channels normalized both resting calcium levels and SR calcium content. Thus, the EC uncoupling pore-lining mutation blocked calcium leak through Y522S channels. These results indicate that muscle weakness in CCD results, at least in part, from two distinct cellular and molecular mechanisms (enhanced SR calcium leak and calcium permeation defective EC uncoupling) by which mutations in RyR1 reduce calcium release during EC coupling.

7. RYR1-related versus SEPN1-related core myopathies 

Ana Ferreiro presented a comparative analysis of the phenotypical and pathophysiological links and boundaries between RYR1-related and SEPN1-related core myopathies. From the phenotypical point of view, the typical presentations of both forms of core myopathies are distinct and recognizable; whilst recessive transmission can be associated with either SEPN1 or RYR1 abnormalities, dominant inheritance indicates RYR1 involvement and has never been observed in SEPN1-related myopathies. Nevertheless, a percentage of RYR1-mutated cases, particularly those with recessive mutations, show a severe clinical phenotype with respiratory involvement and scoliosis which can be difficult to differentiate from SEPN1-related forms. The presence of one or several of the following clinical findings strongly suggests a RYR1-related myopathy: congenital contractures, neonatal feeding or respiratory difficulties, marked ptosis or facial diplegia, ophthalmoplegia (which can be a late sign, appearing in early adulthood), dorsal kyphosis (sometimes congenital) with or without scoliosis, predominant pelvic weakness and/or “tubular” thighs without focal amyotrophy. Morphologically, the presence of type 1 fibre uniformity and of long or very large cores, particularly if associated with minicores, rods and/or abundant central nuclei, is also strongly suggestive of RYR1 involvement. In contrast, the following signs point towards a SEPN1-related myopathy: delayed and/or poor head control in patients who walked at a normal age, predominant neck and trunk weakness and amyotrophy, scoliosis with dorsal lordosis and lateral trunk shift, severe diaphragmatic weakness, nocturnal hypoventilation, nasal speech and/or inner thigh amyotrophy (“bracket” thighs). Severe, non-ambulant CCD cases can require assisted ventilation, but the need for respiratory support in an ambulant patient is typically associated with SEPN1 mutations. Morphologically, SEPN1-related myopathy can present with minicores, mild dystrophic lesions, Mallory bodies or congenital fibre type disproportion (CFTD), but typically, well-delineated cores running along a significant extent of the longitudinal fibre axis have never been described in this entity. Spinal rigidity is typically but not exclusively associated with SEPN1-related myopathies, and can be observed in other early-onset muscle disorders such as collagenopathies and Emery–Dreifuss muscular dystrophy, and even in rare cases with recessive RYR1 mutations.

The phenotypical link between SEPN1- and RYR1-associated myopathies might be partly related to common pathophysiological mechanisms. Ana Ferreiro presented the studies currently in progress in her laboratory to determine the physiological role of SelN, and proposed the hypothesis that there might be a functional link between RyR1 and SelN, both localized in the membrane of the sarcoplasmic reticulum. Oxidative status and Ca2+ handling were analysed comparatively using cultured fibroblasts and muscle cells from patients expressing different RYR1 or SEPN1 mutations. Preliminary results demonstrate that this material represent a useful cell model of the disease, and suggest a role of SelN in maintaining the redox status of the cells. Further pathophysiological studies are currently in progress.

8. Conclusions and future perspectives 

The 150th International ENMC workshop brought together clinicians and basic scientists working on the core myopathies and the selenoprotein N (SEPN1) and the skeletal muscle ryanodine receptor (RYR1) genes associated with this heterogeneous group of conditions.

Although in some cases distinct clinical and histopathological features may already suggest a specific genetic background, this may be less straightforward in others, and the group discussed a number of clinico-pathological markers which are more likely to point to involvement of either the SEPN1 or the RYR1 gene (Table 1). In addition, it was discussed that the presence of core structures on muscle biopsy in the absence of muscle weakness is insufficient to constitute a diagnosis of either CCD or MmD. Also, metabolic and genetically determined conditions with cores as an additional finding on muscle biopsy but clinical features distinct from SEPN1- or RYR1-related core myopathies ought to be taken into account; this group currently includes type III glycogenosis [25], short-chain acyl-CoA dehydrogenase (SCAD) deficiency [26] and other myopathies due to mutations in the skeletal muscle α-actin (ACTA1) gene [27], the cofilin-2 (CFLN2) [28] and titin (TTN) gene [29].

Table 1. Clinical and histopathologic features which may aid the distinction between SEPN1-related and RYR1-related core myopathies

	Feature	SEPN1-related MmD	RYR1-related MmD
	Histopathology
	Type 1 predominance/uniformity	+	+++
	CFTD	++	+
	Increase internal nuclei	+	+++
	Multiple large cores (“multicores”)	+	+++
	Numerous small cores (“minicores”)	+++	+
	Clinical
	Congenital contractures	−	+
	Neonatal respiratory involvement	−	+
	Neonatal bulbar involvement	−	+
	Extraocular involvement	−	++
	Bulbar involvement	+	+
	Respiratory involvement	+++	+
	Scoliosis	+++	++
	Dorsal kyphosis	+	++
	Malignant hyperthermia susceptibility	−	+

− = feature not reported, + = feature reported, ++ = common feature, +++ = very common feature. CFTD = congenital fibre type disproportion.

Mutations in the RYR1 gene are now emerging as the most common cause of the core myopathies; whilst RYR1 mutations were initially associated with malignant hyperthermia (MH) and classic central core disease (CCD), recent findings indicate that RYR1 mutations are associated with a much wider range of phenotypes including MmD, CNM and more subtle histopathologic abnormalities including type 1 predominance or uniformity. Several large studies presented at this workshop indicated clear genotype–phenotype correlations, with CCD-related dominant RYR1 mutations predominantly localizing to the RYR1 C-terminus, in contrast to the even distribution of recessive RYR1 mutations giving rise to a MmD phenotype. In addition, expression of heterozygous RYR1 mutations on a haploinsufficient background, either due to epigenetic allele silencing or instability of the RyR1 protein associated with truncating mutations, has emerged as an important mechanism in core myopathies with a more severe phenotype and/or external ophthalmoplegia. Several presentations focusing on the increasing number of proteins with a recognized role in triad assembly and maintenance suggested possible candidate genes for the substantial number of core myopathy cases not due to SEPN1 or RYR1 mutations.

Functional studies of mutant RyR1 receptors in vitro and in currently available mouse mutants suggested an increased SR calcium leak and E–C uncoupling due to a permeation defect affecting the RyR1 pore region as the two principal pathogenetic mechanisms in core myopathies; currently available mouse mutants offer a suitable model to assess the effect of specific pharmacological compounds on these mechanisms. Data on the role of RyR1 in dendritic cells presented at this workshop indicated a wider role of the receptor in non-muscle cells which may expand in future.

9. Workshop participants