If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Twenty-one active participants from Austria, Brazil, Canada, Finland, France, Germany, Israel, Italy, Japan, the UK and the USA met in Naarden on 12–14, April 2002. This was the fourth meeting held under the auspices of the ENMC on the subject of the limb-girdle muscular dystrophies (LGMD). Following on the previous meetings of the consortium [
] which had discussed the identification of the genes and proteins involved in different types of LGMD and their associated phenotypes, the specific remit of this meeting was to address issues of disease causation. The huge heterogeneity now known to exist in LGMD (Table 1) and the large number of genes and proteins involved made it necessary to concentrate on the disorders involving proteins other than those of the dystrophin–sarcoglycan complex. The meeting discussed the recent identification of three new genes and proteins involved in different types of limb-girdle muscular dystrophy (LGMD2H, LGMD2I and LGMD2J), and went on to look more specifically at the different groups of proteins associated with LGMD which are known to interact with each other or work in similar pathways. A significant problem which remains in our understanding of these diseases is the variability of presentation or severity which may be associated with disease caused by single genes or even with single mutations (Table 2). Strategies to address this variability on a collaborative basis in patients and in animal models of disease were discussed.
Table 1Current state of knowledge on the genes and proteins involved in LGMD, with key references
Identification of mutations in the gene encoding lamins A/C in autosomal dominant limb-girdle muscular dystrophy with atrioventricular conduction disturbances (LGMD1B).
A common missense mutation in the adhalin gene in three unrelated Brazilian families with a relatively mild form of autosomal recessive limb-girdle muscular dystrophy.
Linkage analysis in autosomal recessive limb-girdle muscular dystrophy (AR LGMD) maps a sixth form to 5q33-34 (LGMDF) and indicates that there is at least one more subtype of AR LGMD.
A gene for autosomal recessive limb-girdle muscular dystrophy in Manitoba Hutterites maps to chromosome region 9q31-33: evidence for another limb-girdle muscular dystrophy locus.
Mutations in the fukutin-related protein gene (FKRP) identify limb girdle muscular dystrophy 2I as a milder allelic variant of congenital muscular dystrophy MDC1C.
Professor Wrogemann (Winnipeg, Canada) reviewed the form of LGMD previously described in the Hutterite population of Manitoba and Saskatchewan. This disorder is characterized by a relatively mild and slowly progressive disease and more than 60 affected individuals have been identified. The disorder had been linked previously to chromosome 9, distal to the locus for Fukuyama muscular dystrophy [
], but not all Hutterite patients show linkage to this locus (see below, LGMD21). Four genes were identified in a critical region of 560 kb, and all four were sequenced. A missense change (1459G>A D487N) was identified in exon 2 of TRIM 32 which segregated perfectly with the disease and which was not found in 200 control chromosomes. TRIM32 is a putative E3 ubiquitin ligase which is the 72 kDa protein expressed in numerous tissues, including heart and skeletal muscle. The entire coding sequence is contained in a single exon. The proposed mutation is at a highly conserved residue within the third NHL domain of the protein: however, with only a single mutation reported in this founder population there is an urgent need to confirm its pathogenic nature through the identification of other affected families or through a functional test for the gene product.
E3 ubiquitin ligases have not previously been implicated in the pathogenesis of muscular dystrophies, though they have been associated with other diseases including a form of Parkinson's disease. Typically their function involves tagging proteins for degradation in the proteasome, so that a target protein or proteins may be ligated to ubiquitin and tagged for degradation by the proteasome pathway. A mutation in such a gene could lead to misidentification of the target protein, or, in the absence of ubiquitin ligase, the accumulation of a target protein could potentially lead to muscle disease. To date there has been no evidence for protein accumulation (on light microscopy) in biopsies from these patients, though further studies may shed light on these issues.
2.2 LGMD2I and FKRP
Professor Muntoni (London, UK) described the identification of the gene for fukutin related protein (FKRP). The gene was identified via its homology (in the predicted active site only) to fukutin, the gene mutated in Fukuyama muscular dystrophy. The protein of 495 amino acids is predicted to contain a conserved DxD motif and a type II membrane spanning domain identical to that seen in many Golgi resident glycosyltransferases. The gene has four exons, with the entire coding sequence contained in exon 4. Highest expression is seen in skeletal muscle, placenta and heart.
Mutations in the FKRP gene were first identified in a form of congenital muscular dystrophy (MDC1C) with a secondary reduction in laminin alpha 2 labelling on muscle sections. These children have very severe muscle weakness with progressive respiratory failure and cardiomyopathy. Typically these children never achieve independent ambulation. They show prominent wasting of the shoulder girdle musculature together with calf and other muscle hypertrophy. Macroglossia, in a couple of cases severe enough to require surgery, has been reported in several of these severely affected cases. As is the case with fukutin mutations, no individual carrying two null-type FKRP alleles have to date been reported; the latter state may be incompatible with survival. In addition to a reduction in laminin alpha 2 immunolabelling, these children show markedly reduced alpha dystroglycan staining in muscle and a reduction in its molecular weight on immunoblotting. This may indicate a glycosylation defect which could be a part of the molecular pathogenesis of this disease, in common with various other types of muscular dystrophy secondary to defects in glycosylation (for example, muscle eye brain disease and Fukuyama muscular dystrophy).
Professor Muntoni also reported that in addition to its involvement in MDC1C, mutations in the FKRP gene are known to be responsible for the form of LGMD which had previously been mapped to chromosome 19 (LGMD2I). In these milder cases a single mutation (C826A) is present either homozygously or heterozygously in the vast majority of patients. This common mutation does not seem to be associated with a single conserved haplotype so it is likely to have arisen recurrently on a number of different genetic backgrounds.
Professor Voit (Essen, Germany) presented a series of childhood onset cases with LGMD2I in whom the disease was essentially of DMD or BMD like severity. Prominent clinical features in the group included gross hyperlordosis and shoulder girdle weakness, and macroglossia. IQ was normal.
Professor Bushby (Newcastle upon Tyne, UK) described 15 cases with LGMD2I ranging in age from 11 to 56 years who had been followed up for a period of up to 11 years. This has turned out to be the single most common cause of LGMD in the Newcastle muscle clinic population and is probably still underdiagnosed. Twelve patients were homozygous for the common C826A mutation in FKRP and had presented mainly in adulthood. The three other cases, two of whom presented in childhood, were compound heterozygotes for the common mutation and different nonsense mutations. In ten of the 15 cases, the major clinical differential diagnosis was dystrophinopathy because of the high serum creatine kinase (CK) (at least 10×normal levels in all patients at presentation), the similar pattern of muscle involvement and the common presence of calf and other muscle hypertrophy. Significant medical complications were seen in a high proportion of the patients in this group. Follow up over time demonstrated a progressive drop in forced vital capacity accompanied later by signs of diaphragmatic weakness and nocturnal hypoventilation. Three patients to date have required home nocturnal ventilation. In contradistinction to the situation seen in many other types of muscular dystrophy, nocturnal respiratory failure was seen in these patients when they were still ambulant. Cardiomyopathy was another significant complication, affecting one-third of the patients to date. This cardiomyopathy appears to respond to conventional anti-failure medication.
Professor Wrogemann described the finding of genetic heterogeneity amongst the Manitoba Hutterite population, where the second LGMD gene involved is now known to be FKRP, with all patients homozygous for the same C826A mutation also seen in European patients. Amongst these patients, severe cardiomyopathy (one patient is awaiting cardiac transplantation) and, in two patients, malignant hyperthermia like reactions, one with severe rhabdomyolysis, have been reported.
Dr Anderson (Newcastle upon Tyne, UK) presented the findings on protein analysis of the LGMD2I patients studied in Newcastle. Secondary protein abnormalities were common in this group, and out of 13 biopsies studied, only one showed normal immunolabelling for the antibodies examined. In the other biopsies, the most common abnormality was loss of laminin alpha 2 labelling on immunoblotting, as originally reported in a subgroup of patients now known to have LGMD2I. In contrast to the situation with the children with MDC1C, secondary loss of laminin alpha 2 on sections was not frequently seen. Two biopsies had reduced immunolabelling of laminin beta 1 on sections and further two biopsies showed a secondary reduction in calpain 3 on immunoblotting. There did not appear to be any marked phenotypic differences between the patients with different patterns of secondary protein abnormalities. The findings in the LGMD patients were much more variable than in the MDC1C group where a consistent finding is a reduction in laminin alpha 2 immunolabelling on muscle sections. The importance of noting these secondary protein abnormalities and their potential significance in patients with otherwise undiagnosed disease was highlighted.
Overall, LGMD2I would appear to be a significant cause of LGMD, both in terms of the numbers of patients potentially affected and because of the significant management implications for both the cardiovascular and respiratory systems. The diagnosis can be suspected clinically by the presence of a proximal, predominantly pelvifemoral muscle weakness in the presence of a high serum CK and muscle hypertrophy. Muscle biopsy is important to exclude dystrophinopathy and secondary protein changes, especially involving the laminin chains, may also be a clue to the diagnosis, which ultimately rests on the demonstration of the causative mutation.
2.3 LGMD2J and titin
Dr Udd (Finland) presented the identification of two titin mutations in tibial muscular dystrophy families. In the Finnish population, tibial muscular dystrophy (an autosomal dominant disease) is relatively common. For many years, the presence of potentially homozygous patients has been recognized in this group. These homozygous patients have a proximal early onset LGMD with high serum CK and secondary reduction in calpain 3 labelling on immunoblotting. It has now been proven that these patients are homozygous for titin mutations. Given that the disease in heterozygotes may be very mild and that the phenotype is also different (tibial MD), it is conceivable that the disease in other apparently sporadic or recessive LGMD patients, especially those with a secondary loss of calpain 3 immunoreactivity, may be due to mutations in titin, with mild or even presymptomatic disease in parents having gone unnoticed. Titin is, however, a huge gene and mutation detection in titin would be a major challenge.
3. Moving towards an understanding of pathogenesis in LGMD
3.1 Dysferlinopathy
Dysferlin (the gene mutated in LGMD type 2B and Miyoshi myopathy) is a member of a protein family characterized by the presence of multiple C2 domains and a C terminus transmembrane domain. Another member of this protein family with expression in muscle is myoferlin.
Dr McNally (Chicago, USA) described the characteristics of the FER-1 like family, relating back to the identification of FER-1 itself in Caenorhabditis elegans. FER-1 mutants are known to have failure of fusion of membranous organelles during organogenesis. Experiments in muscle development have shown that myoferlin is expressed in prefusion myoblasts, earlier than dysferlin, which is expressed in post fusion myotubes.
Functional studies on the first C2 domain in myoferlin suggest that it is able to bind tritiated phospholipids in the presence of calcium. A mutation which has been reported in dysferlinopathy in the corresponding region of dysferlin was introduced into culture and abolished myoferlin's ability to bind phospholipids. These data build into the general theory that dysferlin and myoferlin may play a role in myotube regeneration and repair.
The possibility that myoferlin might be involved in a complicated form of hereditary spastic paraplegia (SPG9) was explored by Dr Bashir (Durham, UK). Myoferlin maps to the critical region for this disease and mutation analysis in the rare families with this phenotype is in progress. The future role of myoferlin and other FER-1 like genes (additional family members are likely to exist based on the presence of further homologous ESTs in the sequence databases) in other disease states may shed further light on pathogenic mechanisms in this group of proteins.
The use of animal models to explore pathogenesis in dysferlinopathy was developed by Professor Bittner (Vienna, Austria) and Professor Campbell (Iowa, USA). Problems of non-muscle pathology (including immunological disease and the development of tumours) in the SJL mouse, a naturally occurring model of dysferlinopathy led to studies by Professor Bittner to breed the SJL mutation onto a different genetic background. The strategy was to transfer the dysferlin mutation which occurs naturally in the SJL strain onto C57/BL10, a mouse strain known to be ‘resistant’ to muscle pathology, in whom a series of muscle relevant data exist and which are not prone to tumourogenesis or autoimmune disease. Studies of these mice confirm a dystrophic process at the pathological level. Like in SJL, pyruvate kinase levels are high, while CK levels are normal. This difference from the human disease is not explained. Pathological changes in the mice include internal nuclei, muscle fibre degeneration and regeneration and the presence of inflammatory cells. The presence of inflammatory cells is of interest given the inflammation which may be prominent in human biopsies from patients with dysferlinopathy (reviewed by Professor Argov, Jerusalem, Israel). MHC studies in these patient samples typically do not show upregulation, but the finding of inflammation as a primary part of the disease process in both mice and humans suggests that inflammation may be an integral part of the muscular dystrophy, but whether this is a bystander phenomenon or contributes to the pathophysiology is not clear. Experimental anti-inflammatory therapy could be tried in this model before a formal trial in patients is considered. On Evans blue injection, small groups of positive fibres or single positive fibres are seen. Pathological changes are seen in heart and diaphragm as well as in the limb musculature, and all of these changes increase with age. Functionally, these mice were rather more severely weak than the SJL mice with the same mutation. In a few heterozygous mice, especially male heterozygotes, there was a clearcut muscle pathology.
Findings in a dysferlin knock out model (Professor Campbell) remain preliminary, but initial studies suggest that these mice have a slowly progressive muscular dystrophy phenotype.
The definition of the skeletal muscle promoter of dysferlin was described by Dr Laval (Newcastle upon Tyne, UK). 5′ RACE was used to map the transcriptional start site to 912 bp upstream of the translational start codon in muscle and 2039 bp of sequence upstream of the transcriptional start site was generated by sequencing. This sequence contains clusters of MyoD and Sp1 binding sites. A CpG island (669 bp) is present within the 5′ UTR and extends into intron 1. Reporter assays show that the dysferlin promoter is complex, with elements upstream of exon 1 which have both positive and negative effects on expression. Apart from a proposed interaction with caveolin 3, the proteins interacting with dysferlin are not known. Dr Laval described a comprehensive screen of dysferlin in an overlapping series of baits in the yeast two-hybrid system aiming to identify proteins interacting with dysferlin. A single interacting partner was identified using a bait containing the two C terminal C2 domains. Expression of this small (25 kDa) novel skeletal muscle specific protein is induced during myogenesis. The only significant motifs are two ankyrin repeats which form the bulk of the protein, and which result in its homology to MYPT2, the targeting subunit of myosin phosphatase. It is therefore possible that this protein may act as a targeting subunit or a linker in a larger complex or interaction.
3.2 Caveolinopathy
The subject of caveolin 3 and its relationship to other types of muscular dystrophy was explored by Professor Minetti (Genoa). Various techniques are available to study caveolae in muscle tissue. Electron microscopy and in particular freeze fracture techniques have been especially valuable, and show a highly organized structure at the plasma membrane. Mutations in caveolin 3 have now been demonstrated in a number of different clinical situations including LGMD, hyperCKaemia, an unusual case of distal myopathy and rippling muscle disease. The same mutations have been described in association with variable phenotypes in the same families, though it is of note that not many large families with caveolin 3 mutations have been described. In the presence of a caveolin 3 mutation, caveolin immunolabelling is typically lost on muscle sections and blots.
Implying a broader role in the pathogenesis of muscular dystrophy than its involvement in this rare form of autosomal dominant LGMD might suggest, muscle from patients with Duchenne muscular dystrophy shows abnormal caveolae, and specifically an increased number of caveolae, with overexpression of caveolin 3. To elucidate better the role of caveolin 3 in the pathogenesis of muscular dystrophy, transgenic mice overexpressing caveolin 3 were created. These mice have a severe muscular dystrophy with weakness, dystrophic pathology and high CK. Western blot and immunocytochemical analyses reveal that caveolin 3 overexpressing mice contain virtually undetectable levels of dystrophin and dramatically reduced levels of beta dystroglycan (∼2–3 fold reduction) while overexpressing caveolin by ∼3–5 fold. These data clearly indicate that transgenic overexpression of caveolin 3 in skeletal muscle fibres induces a DMD-like phenotype. A possible explanation for this phenomenon, because caveolin 3 and dystrophin bind to the same extreme C-terminus site of beta dystroglycan, is that caveolin 3 may regulate the interaction of beta dystroglycan with dystrophin, by competitively recognizing the same site as dystrophin on beta dystroglycan.
The caveolin 3 deficient mouse has a reduction in the number of caveolae and alterations in the T tubule system. There is an apparent discrepancy between the severity of the loss of caveolae and the mild clinical effects of this deficiency. In LGMD1C patients there is a great reduction in the formation of caveolae at the membrane level and a more severe disruption than in mice of the T tubule at the subsarcolemmal level.
Although the sarcoglycan complex was not a major topic of discussion of the workshop, Kevin Campbell provided an overview of some of the major findings from animal models of sarcoglycanopathy. These animals have shed light on the relative importance of different complex members in skeletal, smooth and cardiac muscle, and the feasibility of gene therapy in this group of diseases.
3.3 Calpainopathy
Isabelle Richard (Paris, France) described an animal model of calpainopathy, and the work on this model performed to assess any functional deficits and to try and worsen the muscle phenotype experienced. Capn3 null mice have a variable pathology depending on the muscle examined. There is an increase in the level of apoptotic nuclei, which reflects the human situation. In vivo functional tests have not shown any difference between Capn3 deficient and wild type mice, but on ex vivo analysis, force generation was reduced in soleus muscle only. These data suggest that the deficiency in Capn3 deficient mice does not lead to a particular sensitivity to exercise as is the case with the mdx model of DMD. These results potentially reinforce the idea that calpain 3 is likely to be more important in signalling in the muscle cell than in a mechanical process. Breeding the Capn3 knockout onto different genetic backgrounds showed that the phenotype could be worsened by breeding onto the 129 Sv strain.
Dr Richard also described the approach to gene transfer in this animal model. Following the efficiency of gene transfer was not easy at the protein level, however, transgene expression was detected over a month at the RNA level. There were also problems assessing any effect on function because of the minor differences found in the knockout.
The effects of calpain 3 deficiency were assessed in the animal models by microarray analysis. Expression patterns were found to be altered in genes involved in inflammation, muscle regeneration and cell metabolism. There was some down-regulation of the components of the ubiquitin-proteasome system, possibly indicating a decrease in protein ubiquitination and the accumulation of altered proteins or dysregulation of the NFkB/IkBa pathway. This observation is pertinent in the light of the involvement of TRIM 32 in LGMD2H.
Melissa Spencer (Los Angeles, USA) described studies in calpain 3 transgenesis, with analysis of the feasibility of gene therapy and insights into the biological role of calpain 3.
Various transgenic lines were generated, calpain 3 overexpressors, and lines manipulated to be exon 6- or exon 15, two exons involved in alternative splicing in vivo. No deleterious phenotype is associated with calpain 3 overexpression, raising hopes for gene therapy. However, variable pathological and functional alterations were observed with the latter strains, suggesting that the fine tuning between all isoforms may require better comprehension before embarking into gene therapy trials.
The mice were also used to study calpain 3 interactions. Other proteins involved in muscular dystrophies were studied in the different mouse strains but no consistent alterations were observed. However, the muscle specific isoform of filamin 2 (a protein known to interact with the sarcoglycans) binds calpain 3 in vitro and exon 15-mice showed loss of filamin while the exon 6-mice showed some filamin accumulation, though this was not statistically significant. In vitro cleavage analyses showed that muscle specific filamin can be cleaved by calpain 3.
Filamins have multiple interactions with proteins that are involved in muscular dystrophies. The relevance of these various interrelationships is not yet understood, but disorganization of filamin structure is seen in a number of different muscle diseases including DMD, central core disease and multicore myopathy.
The molecular networks in which calpain 3 plays a part were reviewed further by Hiro Sorimachi (Tokyo, Japan). A common theme in all of the calpain 3 missense mutations studied in culture was a loss of proteolytic function rather than any structural effects. In addition, transgenic mice with a proteolytically inactive calpain 3 have a mild myopathy phenotype. The proposed mode of action of calpain 3 in normal muscles therefore is via signal transduction leading to regulation/suppression of both activity of ubiquitous calpain and degradation of muscle proteins. The autolysis of conventional calpains is important in regulating their protease activity, but it appears not to alter the structure/activity in the case of calpain 3. The autolysis of calpain 3 appears to take place in two steps, possibly relating to a change in substrate specificity and/or change in the localization of the protein within the cell. This proteolytic activity does not appear to be critical for muscle development, as studied in embryonic stem cells expressing protease inactive calpain 3, or also from the observation of normal looking muscles among presymptomatic LGMD2A patients harbouring two null-type calpain 3 alleles. Rather, it may be required for maintenance/defence of muscle against various stresses such as physical movement, heat and pH change.
The molecular networks around myofibrils are becoming very complex, with various proteins known to be involved in different forms of muscular dystrophy linked at the sarcomere or via common interacting proteins (Fig. 1).
Fig. 1Molecular network around p94/calpain 3 and connectin/titin. Reported and possible interactions/translocations of various proteins in skeletal muscle cells are summarised. p94/calpain 3 is shown to bind to N2A and the C-terminal regions of connectin/titin by the yeast two-hybrid (YTH) studies [
], calpastatin, which is the endogenous specific inhibitor protein for the conventional calpains, DM protein kinase, and HSP60, which is the chaperonin in mitochondria [43, unpublished data]. The N2A region of connectin/titin, and cytoskeletal proteins, fodrin/α-spectrin and α-filamin, are very good substrates for conventional calpains [41]. Connectin/titin is also shown by Y2H to specifically bind to T-cap/telethonin at the N-terminus [44], to α-actinin at the N-terminal Zr-repeat region [45], and to the MURF-1 at the M-line region very close to the titin kinase domain [46], which was reported to phosphorylate T-cap/telethonin [47]. α-actinin was reported to interact with ZASP/Cypher (Z-band alternatively spliced PDZ-motif protein), ALP (actinin-associated LIM protein), myotilin, actin, FATZ (filamin, actinin and telethonin binding protein) etc. [48]. T-cap/telethonin binds to FATZ and regulatory subunit of K+-channel, minK, suggesting connection between Z-line and sarcolemma [49]. Molecular links, α-actinin-actin-dystrophin and α-actinin-myotilin-α-filamin, also suggest connections between Z-line/myofibril and sarcolemma. On the other hand, MURF-1, which has two more homologues, MURF-2 and -3, composing homo/hetero-dimers, has typical RBCC (RING-B-box-Coiled Coil) structure, strongly suggesting that MURFs are transcription factors as well as E3 subunit of ubiquitination system, connecting myofibril to the nuclei [
Muscle-specific RING finger 1 interacts with titin to regulate sarcomeric M-line and thick filament structure and may have nuclear functions via its interaction with glucocorticoid modulatory element binding protein 1.
]. Unlike TRIM32, a possible function of MURF-1 as E3 subunit is for sumoylation, because MURF-1 was shown to bind to UBC9, which is SUMO (small ubiquitin like modifier)-specific E2 (ubiquitin conjugating enzyme) [50]. Physiological functions of SUMO system have not yet been completely understood, half-lives of some proteins such as p53 is proposed to be regulated by both ubiquitination and sumoylation [51]. The figure was kindly provided by Hiroyuki Sorimachi.
Developing the discussion of the link between calpain 3 and titin, the meeting went on to discuss the types of LGMD where mutations in sarcomeric proteins have been described. Mayana Zatz (Sao Paulo, Brazil) outlined the situation concerning the disease (LGMD2G) associated with mutations in the gene encoding telethonin. This disorder to date remains restricted to four Brazilian families. One nonsense mutation is present homozygously in three of the families and heterozygously (with another null mutation) in the other. Intrafamilial variability was marked in terms especially of severity and the presence or absence of calf hypertrophy, which might be asymmetrical. CK was elevated 6–30 fold. The presence of rimmed vacuoles was not consistent in the muscle biopsies examined. All other proteins involved in muscular dystrophies are normally expressed, and there is no gross disruption of the sarcomere on light or electron microscopy.
Michael Hauser (USA) described the autosomal dominant form of muscular dystrophy associated with myotilin mutations (LGMD1A), until recently restricted to a single North American family, but now known to be present in a second unrelated family as well where a different mutation has now been described. A distinctive clinical feature in these families is an unusual voice with markedly dysarthric speech. Rimmed vacuoles are present in muscle and electron microscopy in affected patients showed the presence of autophagic vacuoles and Z line streaming in the absence of filaments. On light microscopy, this was indistinguishable from nemaline rods. Immunogold labelling suggests that myotilin in affected patients can localize correctly to the Z line, though whether this represents the product of the normal or the mutated allele is not clear. An unusual secondary reduction in laminin gamma 1 chain expression was seen in affected patients.
The similarities in some of the muscle biopsy findings between LGMD1A and nemaline myopathy are intriguing given that missense mutations in alpha tropomyosin may cause nemaline myopathy and both it and myotilin bind alpha actinin. Although the reported myotilin mutations do not disrupt alpha actinin binding there remains the possibility of the loss of tethering of alpha actinin playing a role in pathogenesis in these patients. The similarity of the findings on muscle biopsy between these two disorders implies that a further source of LGMD1A patients might be amongst those diagnosed as late onset nemaline myopathy.
Linkage studies in LGMD1E are ongoing. Filamin 2 maps to chromosome 7, just outside the critical region.
Bjarne Udd (Finland) recapitulated the story of the autosomal dominant tibial muscular dystrophy present at a relatively high frequency in Finland. It has been suspected for many years that people homozygous for the TMD gene suffered from a LGMD. This was confirmed on linkage once the TMD gene was linked to chromosome 2, in the region of the titin gene. Eventually, mutations in the titin gene have been identified in TMD families and the patients with a LGMD phenotype in these families have been proven to be homozygous for this titin mutation. A titin mutation has also been found in the mdm mouse which is a mouse model of muscular dystrophy with patchy muscle pathology. Homozygous patients and the mdm mice share the characteristics of a secondary loss of calpain 3 labelling in skeletal muscle, with the mutations predicted to disrupt one of the calpain binding sites in titin. Other titin mutations have been described in cardiomyopathies. These different diseases may reflect the effects of mutations on different titin isoforms with specific tissue distributions.
There are a growing number of proteins with sarcomeric localization and function now known to be involved in forms of LGMD. These proteins have multiple interactions and many proteins are involved in their modification or degradation (Fig. 1, Table 2). Klaus Wrogemann reviewed the relationship between the proteins of the proteasome, muscular dystrophy and the maintenance of normal muscle. TRIM 32 and MURF-1 are E3 ubiquitin ligases of which there are hundreds. These proteins are responsible for the selectivity of protein attachment to the proteasome for degradation. MURF-1 binds to titin repeats adjacent to the catalytic kinase domain, and may be important in the modulation of titin kinase. A knock out mouse for MURF-1 is resistant to muscle atrophy under a variety of experimental conditions.
3.5 Variability in LGMD
A major part of the discussion at the workshop focused on the variable phenotype seen in association with many of the genes involved in LGMD. Mutations in these genes might cause recognizably different phenotypes, variability in the pattern of muscles involved or variability in severity (see Table 2).
Zohar Argov reviewed the variability seen in the Libyan Jewish population who have a single founder mutation in dysferlin. An almost exactly equal split could be seen at onset between the patients presenting with proximal or distal muscle involvement. An additional variability at presentation could be seen with a small number having a transient calf swelling. Progression in the groups with proximal and distal onset was similar so that it was impossible, years after onset, to determine whether onset had been proximal or distal. Variability in the rate of progression of the disease was observed independent of the mode of onset.
Klaus Wrogemann described similar findings of variable mode of onset in populations with founder dysferlin mutation (a First Nations family of Canada and a large Russian kindred). No co-segregation was seen with the proximal or distal phenotypes and selected candidate modifier genes, i.e. closely linked markers to calpain 3 or variants in caveolin 3 or toonin.
Andoni Urtizberea (Garches, France) presented the findings in the Cajun/Acadian populations of North America which have been studied in collaboration with Dr Gene Jackson (Detroit, USA). Thirty-five to forty-five percent of all cases of LGMD in this population are due to dysferlinopathy, with a common mutation in the Acadian cluster. The origin of this mutation is not yet known. As described in the other population clusters, presentation might be with predominantly proximal or distal muscle involvement, but the distinction between the patients presenting proximally or distally was impossible after many years of disease. In addition to some patients having presented with gastrocnemius weakness, others had presented with foot drop, leading to confusion in the initial diagnosis with Charcot Marie Tooth disease. In general, though there could be co-existence of proximal and distal phenotypes in the same family, the disease in this cluster was relatively homogeneous, with little intrafamilial variability and few exceptions to the ‘mainstream’ phenotype.
The results of genotype/phenotype correlations in the FKRP gene are still at an early stage, but the initial findings were reviewed by Professor Muntoni. The spectrum of severity seen in association with FKRP mutations is extremely broad. At the most severe end, children never achieve independent ambulation and develop complications of respiratory and cardiac failure in the first or second decade. At the other end of the spectrum, patients may be asymptomatic until their fifth or sixth decades. The common C826A mutation is found in the vast majority of the mildly affected patients. It is present in a single copy, together with another missense or stop mutation in the intermediate cases. The most severe cases never have a copy of the C826A mutation and are compound heterozygotes for two missense or one missense and one stop mutation. No patients have yet been reported with two frameshifting mutations.
Despite this apparently relatively simple genotype/phenotype correlations, there remain various unresolved issues in LGMD2I, especially the intrafamilial variability in severity or in the presence of other complications such as respiratory or cardiac failure. It is also clear that a number of missense mutations (which may occur anywhere in the gene) have a very severe effect on the function of the protein. Functional assays for individual mutations will help to resolve this issue.
LGMD2A represents approximately 30% of all LGMD in Brazil. Experience with this group was presented by Professor Zatz. Mutations are particularly common in six exons, facilitating mutation detection in this population. Four of them correspond to the protease domain of calpain 3, and one to the possible Ca2+-interacting loop of C2-like domain. Marked phenotypic variability exists in families homozygous for the same mutations. This variability is seen in age at onset, level of calpain 3 expression, and rate of progression of disease. There is a trend towards more severe progression of the disease in male patients, and possibly also in the African Brazilian population. The possibility of the interaction of mutations in different genes accounting for some cases of muscular dystrophy was raised by the fascinating findings in one family with five affected brothers. All had very high CK (50–100 fold increase) but all analyzed muscle proteins in two affected brothers were normal. The five sibs were heterozygous for one null calpain 3 mutation but in whom a second calpain 3 mutation as well as dystrophinopathy could be excluded by linkage analysis.
Bjarne Udd reviewed the variable situation with mutations in titin. The phenotypic variability in titinopathy responsible for whether the disease presents distally or proximally appears to relate to whether an individual is heterozygous for a titin mutation (most of whom will have a predominantly anterior tibial distal myopathy, though some variation of the phenotype is possible and was reviewed in detail at the ENMC distal myopathy workshop March 2002) or homozygous, in which case the disease appears to be a proximal LGMD phenotype. This phenotype has been only very rarely described and includes the development of proximal weakness before the age of 10 years with requirement for a wheelchair by age 25. These patients have a secondary absence of calpain 3.
The presence of caveolin 3 mutations leads to a very variable phenotype, reviewed by Dr Straub. In a family where a young boy presented with hyperCKaemia and progressed to show myalgia and toe walking, the father had a history and examination consistent with rippling muscle disease. An unusual distal myopathy has been reported in association with a caveolin 3 mutation also reported with other presentations. Very striking variability of disease in association with mutations in the same gene can also be seen in association with lamin A/C mutations (the cause of LGMD1B, ADEDMD, AREDMD, familial dilated cardiomyopathy, polyneuropathy and Dunnigan's partial lipodystrophy and mandibuloacral dysplasia).
All of these different examples of phenotypic variability suggest that modifying factors affecting the expression of the primary mutation are critical with most of these genes. While human studies can be very informative, there is no doubt that animal models offer a very attractive system for the identification of potential modifying factors because of the high degree of control that can be exerted over the genetic background compared to the human situation. Reginald Bittner described the use of animal models of dysferlinopathy to study this problem. With the SJL mutation bred onto the C57/Bl10 line a series of strategies were chosen to identify modifying factors which could either operate in a classically Mendelian fashion, via imprinting, or via mitochondrial inheritance. These comprehensive breeding experiments have highlighted 3–6 potential regions of the mouse genome for further study. The identification of these potential modifier loci has been possible because of the phenotypic variability which can be seen in these mice – in the variable involvement of quadriceps versus gastrocnemius, in the level of pyruvate kinase elevation, and in the different severity of disease in male and female mice.
4. Conclusions
With greater clarity in the groupings of proteins involved in producing either an autosomal dominant or recessive type of LGMD, a much clearer approach to diagnosis and to the phenotypic effects of these disorders has been possible, and this was reviewed in detail in the LGMD workshop held in 1999 [
]. Many questions remain about pathophysiology in these diseases but the recognition of some common themes and mechanisms of disease causation has been an important recent development in our understanding. Given the rarity of these disorders, continued collaboration will be necessary to develop our understanding still further. At the patient level, this will require the identification of further families with the less common forms of LGMD in order to be able to confirm or expand our ideas of disease causation. For families still unlinked to the known LGMD loci, knowledge of the heterogeneity of muscular dystrophy phenotypes means that all loci should be targeted as potentially disease causing, with possibly a ‘muscular dystrophy’ set of markers to augment those already available for LGMD. In order to ensure that all potential disease causing mutations are available to the entire scientific community, groups identifying mutations in the genes implicated in different forms of LGMD should contribute to the existing mutation databases such as the Leiden database at dmd.nl. The cataloguing of apparently silent polymorphisms is important both in the diagnostic setting for instance to avoid misdiagnosis and in the search for factors potentially responsible for phenotypic modification.
Additional areas for future collaboration will include the further use of animal models of the various types of LGMD, for example, by cross breeding experiments to be able to examine the effects of double heterozygosity for non-allelic mutations, and by identifying potential therapeutic agents.
This is an era of information generation and should also become an era of efficient sharing of information: the integration of functional genomics with clinical data is an area of rapid evolution, which will require the sharing in public databases of expression array data (of RNA first then of proteins), whether from animal or human muscles. Indeed, carefully designed DNA chip experiments, particularly those performed on animal model systems, will yield valuable information. Expression profile data may eventually end up being a defining part of a given muscular dystrophy phenotype. Given the large data volumes and costs involved, it will be necessary to pool our resources in the most efficient manner. Ideally, all the data should be deposited in a central, public database. This may furthermore imply the need to devise ways to share expensive infrastructures. But in the end, such an evolution is likely to foster new ways to examine the data and promote collaborations. It will allow the scientific community to perform comparative analyses and integrate data across the different types of muscular dystrophy, including LGMD. This is of extreme importance given the difficulty and complexity of the LGMD pathophysiological puzzle. Finally, this is also likely to speed up the discovery of new potential targets for the recognition of common or specific themes of disease causation and therapeutic agents.
Acknowledgements
This workshop was made possible thanks to the financial support of the European Neuromuscular Centre (ENMC) and ENMC main sponsors:
Association Française contre les Myopathies (France)
Deutsche Gesellschaft für Muselkranke (Germany)
Telethon Foundation (Italy)
Muscular Dystrophy Campaign (United Kingdom)
Muskelsvindfonden (Denmark)
Prinses Beatrix Fonds (The Netherlands)
Schweizerische Stiftung für die Erforschung der Muskelkrankheiten (Switzerland)
Vereinigung zur Erforschung von Muskelkrankheiten bei Kindern und Erwachsenen (Austria)
Vereniging Spierziekten Nederland (The Netherlands)
and ENMC associate member:
Muscular Dystrophy Association of Finland.
Appendix A. Participants
Dr Louise Anderson, Newcastle upon Tyne, UK
Dr Zohar Argov, Jerusalem 91120, Israel
Dr Rumaisa Bashir, Newcastle upon Tyne, UK
Professor Reginald Bittner, Vienna, Austria
Professor Kate Bushby, Newcastle upon Tyne, UK
Dr Kevin Campbell, Iowa City, USA
Professor Michel Fardeau, Paris, France
Dr Michael Hauser, Durham, USA
Dr Steven Laval, Newcastle upon Tyne, UK
Dr Elizabeth McNally, Chicago, USA
Dr Carlo Minetti, Genova, Italy
Professor Francesco Muntoni, London, UK
Dr Isabelle Richard, Evry, France
Dr Hiroyuki Sorimachi, Tokyo, Japan
Dr Melissa Spencer, Los Angeles, USA
Dr Volker Straub, Essen, Germany
Professor Andoni Urtizberea, ENMC
Dr Bjarne Udd, Vaasa, Finland
Professor Thomas Voit, Essen, Germany
Professor Klaus Wrogemann, Winnipeg, Canada
Professor Mayana Zatz, Sao Paulo, Brazil
References
Bushby K.
Report on the 12th ENMC sponsored international workshop – the ‘limb-girdle’ muscular dystrophies.
Identification of mutations in the gene encoding lamins A/C in autosomal dominant limb-girdle muscular dystrophy with atrioventricular conduction disturbances (LGMD1B).
A common missense mutation in the adhalin gene in three unrelated Brazilian families with a relatively mild form of autosomal recessive limb-girdle muscular dystrophy.
Linkage analysis in autosomal recessive limb-girdle muscular dystrophy (AR LGMD) maps a sixth form to 5q33-34 (LGMDF) and indicates that there is at least one more subtype of AR LGMD.
A gene for autosomal recessive limb-girdle muscular dystrophy in Manitoba Hutterites maps to chromosome region 9q31-33: evidence for another limb-girdle muscular dystrophy locus.
Mutations in the fukutin-related protein gene (FKRP) identify limb girdle muscular dystrophy 2I as a milder allelic variant of congenital muscular dystrophy MDC1C.
Muscle-specific RING finger 1 interacts with titin to regulate sarcomeric M-line and thick filament structure and may have nuclear functions via its interaction with glucocorticoid modulatory element binding protein 1.
00183-9&id=gr1.jpg){kind=link}