Analysis of the UK diagnostic strategy for limb girdle muscular dystrophy 2A
Institute of Human Genetics, Newcastle upon Tyne, UK
Correspondence to: Prof. K. M. D. Bushby, Institute of Human Genetics, International Centre for Life, Central Parkway, Newcastle upon Tyne, NE1 3BZ, UK E-mail: kate.bushby{at}ncl.ac.uk
 |
Summary |
|---|
 Top Summary IntroductionMethodsResultsDiscussionSupplementary materialReferences |
|---|
Diagnosis of limb girdle muscular dystrophy type 2A can be complex due to phenotypic variability, lack of precision of protein analysis in muscle biopsies, and absence of mutational hot spots in the CAPN3 gene. The aim of this study was to review clinical and biopsy data from a group of patients with known CAPN3 genetic status to validate and refine our current diagnostic strategy, which combines clinical information and protein analysis to direct gene testing. We analysed 85 patients in whom CAPN3 gene sequencing had been performed. Forty-two patients had two mutations, 15 a single mutation and in 28 no mutation was found.
We identified clinical features that clearly discriminated the LGMD2A patients. These were: presence of scapular winging, contractures and normal respiratory function. In addition, a typical pattern of muscle weakness on manual muscle testing could be confirmed. Interpretation of protein expression obtained by Western blot was complex and involved the analysis of a number of bands detected by two antibodies for calpain 3. Loss of all calpain 3 bands was 100% specific for LGMD2A, but this pattern was found in only 23%. Absence or reduction of the 
60 kDa bands was also highly specific for LGMD2A, while increased abundance was highly predictive of no mutations being found even where other bands were reduced, suggesting that this is the most sensitive marker of artefactual protein degradation. Twenty-three percent of the patients with two mutations had normal full-sized calpain 3 protein, consistent with the finding of mutations localized in parts of the gene likely or proven to be involved in autolytic activity. Clinical and biochemical findings in patients with only one mutation were similar to patients with two mutations, indicating that other gene analysis techniques should be used before excluding the diagnosis.
Our analysis confirms that our strategy is still valid to prioritize genetic testing in this complex group of patients, provided patients with normal protein but a suggestive clinical phenotype are not excluded from genetic testing.
Key Words: calpain 3; limb girdle muscular dystrophy type 2A; CAPN3 mutation; Western blotting
Abbreviations: LGMD, Limb girdle muscular dystrophy; MMT, Manual Muscle Testing; FVC, forced vital capacity
Received May 3, 2007. Revised September 30, 2007. Accepted October 1, 2007.
|
Introduction |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionSupplementary materialReferences |
|---|
Limb girdle muscular dystrophy (LGMD) is a heterogeneous genetically determined group of skeletal muscle disorders. As at least 18 genetically distinct subtypes of LGMD have been described, diagnosis in a patient with an ‘LGMD’ phenotype can be challenging. Determining the exact subtype of LGMD in a particular patient is important for genetic counselling and care as they carry different risks of complications (Bushby et al., 2007
). As part of our national diagnostic and advisory service for LGMD our strategy to classify LGMD is to use the clinical and protein phenotypes to direct gene testing (Anderson et al., 1998; Pollitt et al., 2001).
Recessive mutations in the CAPN3 gene (15q15.1–q21.1) are responsible for LGMD2A (Richard et al., 1995) one of the commoner types of LGMD worldwide (Richard et al., 1997; Minami et al., 1999; Chae et al., 2001; Zatz et al., 2003; Fanin et al., 2004; Piluso et al., 2005; Balci et al., 2006). Confirming the diagnosis of LGMD2A can be especially complex. A classic LGMD2A phenotype has been described (Fardeau et al., 1996a, b; Urtasun et al., 1998; Pollitt et al., 2001; Chae et al., 2001; de Paula et al., 2002) but wide intra- and inter-familial variability is reported (Urtasun et al., 1998; Fanin et al., 2004; Saenz et al., 2005). Identification of calpain 3 protein abnormalities can be performed with Western blot analysis utilizing specific antibodies (Anderson and Davison, 1999). However, calpain 3 is highly susceptible to degradation leading to loss of immunoreactivity of the protein and secondary calpain 3 deficiency can also be seen in other types of muscular dystrophy (LGMD2B/2I/LGMD2J/Tibial muscular dystrophy) (Anderson et al., 2000; Chae et al., 2001; Haravuori et al., 2001; Saenz et al., 2005). Conversely, a normal protein profile has been described in a subset of genetically confirmed LGMD2A patients (Anderson et al., 1998; Jia et al., 2001; Talim et al., 2001; Fanin et al., 2003, 2004, 2005; Saenz et al., 2005; Fanin et al., 2007) leading some authors to hypothesize that an additional step (the assessment of autolytic activity) should be added into the diagnostic algorithm (Fanin et al., 2003, 2007). At the histopathological level, it has been hypothesized that the presence of eosinophils on haematoxylin–eosin-stained muscle tissue might point towards the diagnosis of LGMD2A, after a subset of patients diagnosed with idiopathic eosinophilic myositis was shown to carry 2 CAPN3 mutations (Krahn et al., 2006b). Genetic analysis is in itself not straightforward: the lack of defined mutational hot spots makes analysis of a 24 exon gene spanning a region of 40 kb (Richard et al., 1995) laborious and expensive, and in 23% of patients believed to have LGMD2A only one mutation can be detected (Saenz et al., 2005). Finally, some patients may carry mutations in more than one gene known to cause muscular dystrophy (Muntoni et al., 2006).
Here we report a group of patients with known CAPN3 mutation status to underline factors able to discriminate between LGMD2A and non-LGMD2A patients, in order to provide further guidance for refinement of this diagnosis. Patients with only one CAPN3 gene mutation are also scrutinized to assess to what extent they resemble LGMD2A patients to judge the likelihood a second mutation might still be present, calling for further gene analysis using different techniques.
|
Methods |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionSupplementary materialReferences |
|---|
Patient selection criteria
Eighty-five patients (79 unrelated) in whom CAPN3 gene analysis was performed, seeking a specific diagnosis for LGMD, were included in this study (Table 1). They were referred to our specialized LGMD diagnostic service (based at the Institute of Human Genetics, University of Newcastle upon Tyne) for diagnostic testing from around the UK except one patient who was from Greece.
|
Clinical analysis
Fifty-two patients were clinically assessed in Newcastle upon Tyne and from 33 patients clinical information was requested from referring clinicians. Data were collected on clinical onset and disease progression according to a questionnaire first used in the ENMC workshop in 1995 (Bushby and Beckmann, 1995
) (Table 1). Functional severity was graded from stages 1–7 according to a scale proposed by Gardner-Medwin and Walton (1974; G-M&W scale) and follow-up information based on this scale was available for 10 patients. Muscle strength was assessed in the patients seen in Newcastle upon Tyne by Manual Muscle Testing (MMT) scored by a modified Medical Research Council 10-point scale as described previously (Brooke et al., 1981) (Table 1). A score was given for total muscle strength, for left and right-sided strength, for upper, lower, proximal and distal strength. Scores were also given for muscle groups around the shoulder, elbow, wrist, hip, knee and ankle joints. A group of patients for whom we had detailed clinical information was selected for clinical analysis (Table 1). As this information was much more easily obtainable, age at onset and age at loss of ambulation was analysed in all the patients.
Protein analysis
Muscle biopsies collected over an 8-year period were analysed. Optimized immunohistochemical (IHC) and multiplex Western blot (WB) protocols were used for the detection of calpain 3 as well as the other LGMD associated proteins as previously described (Anderson and Davison, 1999; Bornemann and Anderson, 2000; Johnson, 2001). Haematoxylin and eosin (H&E) staining of cryosections from tissue blocks utilized for Western blot analysis was used to assess general morphology and fat content prior to immunoblotting. Protein loading was guided by muscle/fat content and assessed from the density of the myosin heavy chain bands on a Coomassie blue stained gel. β-Dystroglycan (antibody used: NCL-b-DG) was used as an internal loading control to assess proper protein transfer. If tissue was limited, preference was given to Western blotting and IHC was not carried out. β-Spectrin and caveolin 3 labelling was used to evaluate sarcolemmal integrity prior to assessing the expression of the other membrane-associated proteins.
The two antibodies (NCL-CALP-2C4 directed against exon 1 and NCL-CALP-12A2 against exon 8) used on blots, produced 4 bands: a 94 kDa band corresponding to the full-sized protein (detected by both antibodies), a 30 kDa band (NCL-CALP-2C4) of a clear calpain ‘fragment’, and a set of rather diffuse calpain degradation bands of 60 kDa (NCL-CALP-12A2). An example of normal calpain on blot is shown in Fig. 1.
|
Semi-quantitative analysis
Grading was done by EG and RC blind to the previous reports and to the results of mutation testing. Protein amount on sections was graded as normal (4), patchy (3), slightly reduced (2), markedly reduced (1) and absent (0). The bands produced on blots were graded as stronger than normal (5), normal (4), slightly reduced (3), reduced (2), markedly reduced (1) and absent (0). The assessment of the bands was done relative to the abundance of other proteins on the multiplex blot, both in the sample from the same patient (internal controls) and adjacent samples (external controls) including normal muscle samples at different dilutions.
Sequence analysis
Sequence analysis was performed using bidirectional fluorescent sequencing of all 24 exons of the CAPN3 gene either in Newcastle upon Tyne (Institute of Human Genetics), in Wuerzburg (Central sequencing facility of the MD-NET, Institute of Human Genetics, Germany) or (in 1 patient) in Leiden (The Netherlands, Clinical Genetic Centre Leiden). The primers used for sequencing are listed on www.dmd.nl (Leiden Database).
Sequence variant classification
Mutations were designated as pathogenic if they were nonsense, clearly altered a splice-site or altered the reading frame, as well as previously reported pathogenic missense or in-frame mutations and unreported missense mutations leading to non-synonymous amino acid changes. No RT-PCR or in vitro splicing assays were performed. All splicing defects are therefore putative and their position in relation to the exon boundaries was taken into account to judge the likelihood of pathogenicity. 2/1/0-mut patients will be used as terms for patients in whom we detected 2 (LGMD2A), 1 or no mutations, respectively.
Statistical analysis
The Mann–Whitney U test, the Fisher's Exact Test for Count, the Wilcoxon signed-rank test, and sensitivity/specificity estimates were performed using SPSS 12.0 where appropriate. Statistical significance was established when P < 0.05.
|
Results |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionSupplementary materialReferences |
|---|
The numbers of patients for whom different levels of information were available are presented in Table 1.
Clinical analysis
Within the groups selected for clinical analysis, information on different parameters was not always available for each patient. When relevant, the numbers of patients considered for each clinical parameter are written between brackets.
Clinical history: 2mut-patients (LGMD2A)
A non-significantly earlier age at onset was found in females (12.1 years of age, range: 2–50; N = 14) versus males (17.0 years of age, range 2–20; N = 23). Seven patients were wheelchair confined at a mean age of 35.2 years, giving a mean disease evolution of 22.6 years before losing independent ambulation. Almost all patients for whom these data were available (17/24) developed first symptoms in the lower limb, only one patient solely in the upper limb, and in the remaining six patients first involvement was reported in both the upper and lower limb. Most patients had normal early motor development and only seven patients reported early difficulties with sport. There were no intellectual problems. No patient suffered from nocturnal hypoventilation. Three patients were deceased; one died from metastatic breast cancer at age 45 years, one patient from the consequences of diabetes mellitus aged 83 years and one from increasing problems with feeding and self care aged 67 years.
Apart from mean age at onset which was 23.4 years (range: 7–60; 5 males, 7 females) in 1mut-patients and 24.4 years (range: 2–49; 13 males, 11 females) in 0mut-patients, clinical progression in the groups with either 1 or no CAPN3 mutations was broadly similar to the LGMD2A group.
General clinical examination
The 2mut-patients walked with a waddling gait (sometimes tiptoeing) and had a broad-based stance with locked knees and a lumbar lordosis which in some cases was extreme. All patients showed scapular winging, which was usually symmetrical. Four patients developed a mild scoliosis, which did not require treatment, at the age of 15, 18, 18 and 24 years. They had an age at onset of 2, 6, 7 and 20 years respectively. Data on contractures was available for 28 patients. They were present in 19 patients and more commonly in males (11/14) than females (8/14). Contractures most commonly affected ankle dorsiflexion, finger flexion, elbow flexion and wrist flexion. In eight patients (four males, four females) contractures were classified as severe and surgical treatment (on the Achilles tendons) was recommended. Facial weakness was reported in three patients, but was very mild. Calf pseudohypertrophy was present in four patients (out of 18 on which data for this sign was available) affecting both genders equally. Heart function was normal on ECG and echocardiography in all patients, except two patients who showed atrial fibrillation and mildly impaired left ventricular function respectively. Respiratory function declined over time, but was still well preserved with overall values in late stages (up to 30 years after onset) of 80% predicted forced vital capacity (FVC) and no indication of diaphragmatic involvement. In only two cases values significantly lower than 80% were found: 44% 58 years after onset in one patient and 62% 13 years after onset in another.
1mut-patients did not differ from 2mut-patients in clinical presentation, including the well-preserved respiratory function.
Respiratory function, scapular winging and contractures were the three clinical parameters which were different between the 2mut-group and the 0mut-group. In the group of 0mut-patients faster deterioration in respiratory function was seen. Scapular winging and contractures were significantly more common in mutation-positive patients and contractures were more likely to be severe (P = 0.000 and p = 0.028, respectively).
The range of creatine kinase levels found in 2mut-patients was 193–13 000 (iu/l), in 1mut-patients this was 399–11 000 (iu/l) and in 0mut-patients 298–38 620 (iu/l) (not controlled for disease duration).
Table A in the Supplementary material presents clinical information from patients with either one or two mutations.
Muscle strength testing
In 2mut- and 0mut-patients a very symmetrical and proximal muscle weakness was observed more severely affecting the lower limb, although lower scores (controlling for disease duration) were seen in 2mut-patients, especially for proximal muscle strength.
Patients with LGMD2A showed a very characteristic pattern of muscle weakness (Table 2). This pattern seemed to persist over time (data not shown). Clear progression was seen in weakness of shoulder abduction, shoulder flexion, wrist extension, hip abduction, hip adduction and in ankle eversion.
|
1mut-patients presented with a pattern of weakness which was similar to the LGMD2A patients except that two patients showed almost equal strength for the upper and lower limb and one patient was only very mildly affected.
In the 0mut-group weaker knee flexion and hip extension (p = 0.030 and 0.010) compared to their antagonistic muscle groups were the only muscle weakness patterns that were similar to those in the 2mut-group. In the 0mut-group weaker ankle dorsiflexion than plantar flexion was also found (p = 0.009). 2mut- and 0mut-patients could be discriminated by the different muscle strength scores they obtained at assessment of eight muscle groups shown in Table 3. The 2mut-patients were more severely affected in all muscle groups except in wrist flexion. Looking closer at these differentiating muscle groups shoulder flexion and elbow flexion were weaker in LGMD2A patients at all stages of disease duration. Wrist flexion progression was seen in 0mut-patients, but in 2mut-patients this function was well preserved, which made differentiation easier over time. In wrist extension a similar pattern was seen, where only 2mut-patients showed progression. In hip adduction, hip flexion, hip extension and knee flexion, which were very severely affected groups from the start in 2mut-patients, fast progression in 0mut-patients patients was seen. This made differentiation at an early stage very possible but in later stages both groups were affected similarly with only a slightly wider range of muscle scores seen in 0mut-patients.
|
Functional testing: 2mut-patients
Although a wide range of functional grades was seen at different ages, where patients were followed over time, functional progression could be seen (Fig. 2). Both genders were affected similarly.
|
|
Protein analysis
WB results can be found in Tables 4 and 5. It is important to note the specific patterns of all the bands produced by the two antibodies used, as abnormal abundance of the 94 kDa bands (exons 1 and 8) on blot were seen in 10/17 0mut-patients and in 20/26 2mut-patients. While absence of all of the calpain 3 bands is highly specific for LGMD2A, the majority of 2mut-patients did not show this protein phenotype (23% sensitivity). The absence of only the 94 and the 30 kDa bands or the 30 and the
60 kDa bands was not able to improve sensitivity significantly but did maintain a high specificity. A reduction of the 60 kDa bands was highly specific for LGMD2A (94%) with a higher sensitivity (62%). In contrast, where the 60 kDa bands were stronger than normal, this was highly likely to be a non-LGMD2A patient. Normal abundance of the full-sized bands was found in 23% of 2mut-patients. Furthermore, normal abundance of any single band did not convincingly point towards the 0mut-group (data not shown). Important differential band patterns found in our comparative analysis are shown in Table 6. See Fig. 1 for examples of these patterns. Looking at the 30 and 60 kDa bands in relation to the 94 kDa bands in 2mut-patients, both were of the same abundance as the 94 kDa bands in 50%. However, the 30 kDa band was more often of weaker abundance than 1 or both 94 kDa bands (34.6%) than the 60 kDa bands (3.8%), whereas the latter were more often of stronger abundance (46.2%) than the 30 kDa band (11.5%).
|
|
Utilizing the LGMD2A (2mut-) specific patterns (Table 6) to assess to what extent 1mut-patients resemble the former, we have found that 4/6 blots from 1mut-patients showed patterns suggestive for the diagnosis of LGMD2A. These were: an absent 30 kDa band, a slight reduction/reduction of the
60 kDa bands and on two blots both 94 kDa bands were also absent. Of the remaining two blots lacking LGMD2A-specific patterns, the first showed a reduction of the 30 kDa band and both 94 kDa bands with normal abundance of the 60 kDa bands, but also a slightly reduced size of dystrophin C-terminal and rod domain. This male patient presented with bulky calves at the age of 60 years. Dystrophin testing for Becker MD has not yet been performed. The last patient had an almost normal calpain 3 pattern, except for a weaker abundance of the 30 kDa band than both 94 kDa bands. However, this patient carried a missense mutation proven to affect autolytic activity (Arg490Trp), which may have been responsible for the normal presence of the full-sized calpain 3 and therefore the diagnosis of LGMD2A could not be excluded.
Secondary findings
All other proteins analysed on IHC and WB were not of value in the discrimination between LGMD2A and non-LGMD2A patients (data not shown). Dysferlin membrane labelling was reduced (slightly-markedly; see Fig. 3A) in 8/14 2mut-patients (patient number: 1, 7, 15, 18, 33, 36, 44 and 45), all associated with normal dysferlin on blot. In 3/7 0mut-patients abnormal dysferlin membrane labelling (markedly reduced-absent) was found, all showing abnormal dysferlin on blot (markedly reduced-absent). One 0mut-patient for whom IHC was not available had absent dysferlin on blot. In both (2mut- and 0mut-) groups, cases with abnormal membrane labelling showed retention of cytoplasmic labelling (Fig. 3A).
|
The general histopathological pattern seen in the muscle biopsies analysed was compatible with a dystrophic process. In 5/17 patients (patient number: 6, 10, 36, 38 and 55) scattered eosinophils were found (Fig. 3B) 4, 1 and 13 years after onset with an age at onset of 15, 17 and 18 years, respectively. This information was not available for the other patients. No specific mutational or clinical pattern could be found in these cases. Also, in 2/10 0mut-patients scattered eosinophils were found 11 and 15 years after onset with an age at onset of 11 and 20 years, respectively.
Sequence analysis: 2mut-patients (40 unrelated)
Mutations identified are summarized in Table 4. Counting each family only once, 80 pathogenic mutations have been found in LGMD2A patients, of these 46 mutations were different and 19 were novel. Thirty-two mutations were single-base substitutions. Nine of these were predicted to alter splicing, two to produce a stop codon (nonsense mutation) and 21 an amino acid substitution (missense change). There were eight deletions, of which two were in and six out of frame, five insertions (all out-of-frame) and one insertion/deletion predicted to be out-of-frame.
Looking at mutation types overall, 44% (35/80) of the CAPN3 mutations are likely to result in a truncated protein and are thus likely to be inactivating. In addition one variant of unknown pathogenicity (reported on the Leiden Database, but unknown whether the patient had LGMD2A) and two polymorphisms (one reported as a polymorphism on the Leiden Database, the other was a silent mutation) were found. Distribution of the CAPN3 mutations and recurrent mutations can be found in Fig. 4.
|
In 26% (14/54) only one mutation could be detected. In the patients with either one or no mutations a variety of additional genetic tests have been performed (FKRP, lamin A/C, dysferlin, dystrophin, VCP and myofibrillar myopathy genes depending on clinical and other indications). The patients with a single mutation have not had any mutations detected in other genes. Amongst the 0mut-group for CAPN3, one patient has subsequently been shown to have BMD (the muscle biopsy from this patient was not interpretable for protein analysis) and two LGMD2B (WB was available for one patient, which showed absent dysferlin and secondarily slightly reduced calpain 3 with stronger
60 kDa bands).
Genotype-protein-phenotype correlations: 2mut-patients
Genotype-protein
Patients with two nonsense/out-of-frame/splice-site mutations showed a significantly lower protein amount (94 kDa band exon 1 + 94 kDa band exon 8/2) than patients with two missense/in-frame mutations (p = 0.04) and showed a lower protein amount with a borderline significance than patients with two missense/in-frame mutations and patients with one missense/in-frame mutation and one nonsense/out-of-frame/splice-site mutation together (p = 0.05). An overall significant rank correlation was also found (Spearman 
= 0.395; p = 0.041). Mean values of protein amount found in patients with decreasing number of nonsense/out-of-frame/splice-site mutations were 0.25, 1.58 and 2.30, respectively.
Genotype/protein-phenotype
Patients with two nonsense/out-of-frame/splice-site mutations showed a significantly earlier onset than patients with two missense/in-frame mutations (p = 0.01) and also when comparing with patients with two missense/in-frame mutations and patients with one missense/in-frame mutation and one nonsense/out-of-frame/splice-site mutation together (p = 0.04). An overall rank correlation could be found (Spearman = 0.352; p = 0.032). The mean age at onset in patients with decreasing number of nonsense/out-of-frame/splice-site mutations was 10.7, 16.4 and 16.83 years, respectively. There were not enough patients wheelchair confined to be able to analyse correlations. No correlations could be found between protein amount and age at onset.
Genotype–phenotype (functional scale)
Figure 2 shows the correlation for disease evolution according to genotype. The evolution curves from 10 patients suggested an earlier age at onset and faster progression in patients with two nonsense/splice-site/out-of-frame mutations compared to patients who carry at least one missense/in-frame mutation, although this did not achieve statistical significance. Looking at all 42 patients with two mutations no genotype functional-phenotype correlations could be found.
|
Discussion |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionSupplementary materialReferences |
|---|
We present here a retrospective analysis of clinical, muscle biopsy and molecular data in LGMD patients with two, one or no CAPN3 mutations. Our approach makes it possible to determine factors that specifically define the clinical and protein phenotype of LGMD2A patients prior to genetic testing which will increase the chance a mutation will be found in the first gene analysed. An earlier diagnosis allows precise genetic counselling, provides more possibilities to prevent and treat complications (e.g. joint contractures) and is frequently of psychological importance to the patient.
Clinical analysis
We found a clinical phenotype consistent with previous reports (Fardeau et al., 1996a, b; Urtasun et al., 1998; Pollitt et al., 2001; Chae et al., 2001; de Paula et al., 2002). Mean age at onset was slightly older with a mean age of 15.2 years compared to 13.7 (Zatz et al., 2003) and 13.8 years (Saenz et al., 2005), but age at onset and the evolution from onset to becoming wheelchair confined fall within a known range (Kawai et al., 1998; Pollitt et al., 2001; Zatz et al., 2003; Saenz et al., 2005). Although we have found an earlier age at onset in females (Pollitt et al., 2001), our study confirms that functional progression similarly affects both genders (Zatz et al., 2003; Fanin et al., 2004). Facial weakness has been found in our patients as before (Fardeau et al., 1996a, b; Kawai et al., 1998; Pollitt et al., 2001) and should not be an excluding factor.
Even though the group of 0mut-patients was very heterogeneous and in most cases diagnosis is still unknown, the comparison between patients shown to have two CAPN3 mutations and those in whom mutations have been excluded is useful to underline bona fide LGMD2A features. The importance of contractures in LGMD2A has been noted before (Pollitt et al., 2001), the preserved respiratory function in LGMD2A is in contradistinction to most other types of LGMD except LGMD2B and although scapular winging can also be seen in facioscapulohumeral muscular dystrophy or some cases of sarcoglycanopathy, both these diagnoses are relatively readily achieved via genetic or protein diagnosis, respectively. Therefore these features will prove very helpful in clinical practice.
The typical LGMD2A muscle weakness pattern proved to be very helpful in classification, especially the more severely affected hip adductors and elbow flexors. Increasing understanding of the balance of forces in joints is important to be more proactive in prevention and treatment of contractures. We have shown that eight specific muscle groups could be potential identifiers of LGMD2A patients. In our study, group scores for muscle strength were analysed, with each patient assessed only once. Further refinement might be achieved by looking at multiple assessments over time in individual patients.
Protein analysis
All Western blots analysed in this study have been performed in our laboratory using a standardized protocol, which has undergone only minor modifications over an 8-year period (Anderson et al., 1998). Using the clinical and genetic correlates possible through this review of patients for whom the CAPN3 genetic status is known, we have now refined our WB interpretation. Data on calpain 3 bands collected for each patient were always put in context by taking into account: (1) expression of other LGMD proteins that might indicate the reduction of calpain 3 is secondary; (2) overall condition of the muscle biopsy to indicate whether protein degradation may affect the intensity of different subsets of calpain 3 bands; (3) protein loading and efficiency of gel transfer which might affect the overall intensity of calpain 3 bands; (4) relative abundance of calpain 3 bands. In this context, we found WB features highly suggestive of LGMD2A. On the other hand, the presence of stronger bands at 60 kDa suggests calpain abnormalities are caused by artefactual tissue deterioration. Our current data also reinforces our view that to be able to assign a future role for immunodiagnostics in LGMD2A we must examine all calpain bands for specific LGMD2A patterns (within the context of other protein expression and tissue preservation) instead of only determining whether the full-sized protein is of abnormal abundance or not.
It is known that while some mutations lead to a calpain 3 reduction in muscle tissue others may not affect protein content but rather its functional inactivation (Anderson et al., 1998; Ono et al., 1998; Fanin et al., 2003, 2004, 2005, 2007). Autolysis is believed to be important for self-activation and regulation of the proteolytic function of calpain 3. In this study six patients had normal protein expression on WB, of whom three carried mutations proven to affect autolytic activity: Arg490Trp and Arg490Gln (Fanin et al., 2003, 2004, 2005, 2007). The protein phenotype is determined by the allele with the less-severe effect, as one missense mutant allele is sufficient to give rise to normal calpain 3 levels even when the second mutation is null (Lanzillo et al., 2006). Therefore the mutations: Asp419Glu (found in two brothers) and Asp419Gly may possibly affect autolytic activity in the three remaining patients. Asp419Glu and Asp419Gly both lie at the C-terminal end of domain III close to a region reported as a contact region between domain II and III, which appears to be involved in intramolecular domain interactions and the assembly and activation of calpain 3 (Jia et al., 2001; Fanin et al., 2003, 2007).
For patients with a normal calpain pattern on WB there could be value in using a functional in vitro assay examining the autolytic activity of calpain 3 (Fanin et al., 2003) as part of the diagnostic strategy. An autolytic assay is likely to be more valuable than a proteolytic assay which by contrast was only able to identify LGMD2A patients with abnormal calpain 3 on WB (Milic et al., 2007). This suggests that functions other than loss of proteolysis may have to be invoked to explain normal calpain 3 expression on immunoblotting and corroborates the strategy of WB expression to identify LGMD2A patients. We are aware of one patient, identified outside this study, who had a single CAPN3 mutation (Arg490Trp) and had also been shown to have a desmin mutation (Lys449Thr), which both have been reported previously to be pathogenic. This patient had desmin accumulation on his biopsy, as well as reduced full-sized calpain 3. His clinical presentation was very atypical for calpainopathy with only symptoms of stiffness and weakness mainly of the atrophic distal upper limb, marked scapular winging and no contractures. These features, and his very high Creatine Kinase level (>10 000 iu/l) are also very atypical for desminopathy. In this context the particular value of an approach which can add in specific clinical information to identify the patients in whom additional mutation analysis is necessary can be appreciated.
Although the 30 and 60 kDa bands are considered to be proteolytic products of the full-sized protein, the identity of these two fragments is still unclear. Muscle-breakdown experiments (Anderson et al., 1998) showed that the 60 and 30 kDa bands were already present before the experiment had started and that whereas the 30 kDa band disappeared in parallel with the 94 kDa band, the 60 kDa bands degraded further into smaller bands down to 45 kDa in a ladder of degradation products. Our findings support the currently assigned status to the 60 kDa bands of ‘degradation’ bands. However, the 30 kDa fragment behaved differently and could be a separate polypeptide product in its own right expressed from exons 1 to 5 as previously suggested (Anderson et al., 1998). We did not however detect any correlation between the detection of mutations in these five exons and a pattern of weaker 30 kDa than 94 kDa bands. The use of antibodies against calpain 3 on tissue sections which has been impossible so far in the routine diagnostic setting might enhance our knowledge on the identity/subcellular localization of the different fragments.
An association between calpain and dysferlin has been suggested before (Anderson et al., 2000; Fanin et al., 2001; Vainzof et al., 2001; Chrobakova et al., 2004; Huang et al., 2005) and could be responsible for the reduced dysferlin membrane labelling seen in 57% of our LGMD2A patients. Cytoplasmic labelling of dysferlin was found regardless of CAPN3 mutation status and correlated strongly with regeneration as illustrated utilizing neonatal myosin heavy chain and laminin 
5 labelling (both markers for regeneration) as described before (Fanin et al., 2001). Our results suggest that abnormalities of dysferlin on WB rather than IHC should be used to direct dysferlin gene testing and that abnormalities on IHC can be regarded as a frequent secondary finding in LGMD2A.
In this study, we saw only few eosinophils 1–15 years after onset in both LGMD2A and non-LGMD2A (LGMD2B and unknown diagnosis) patients suggesting that EM may be an unspecific transient feature of skeletal muscle degeneration and dystrophy rather than a distinct phenotype associated with CAPN3 mutations as was suggested before (Krahn et al., 2006b).
Genetic analysis: 2mut-patients
Forty-one percent of the mutations found here were novel, as expected from the previous studies (Chou et al., 1999; Minami et al., 1999; Richard et al., 1999; Fanin et al., 2004; Piluso et al., 2005; Saenz et al., 2005; Krahn et al., 2006a). 77.5% of the mutations were localized in exons 1, 4, 5, 10, 11 and 17. We believe most of our population is of British origin, which could make prioritization of these six exons in the UK worthwhile though the remaining mutations lie scattered in nine exons. The lack of frequently recurring mutations implies no role can be given to allele specific PCR.
In accordance with most previous reports (Richard et al., 1999; Chae et al., 2001, 2002; Fanin et al., 2004; Saenz et al., 2005; Milic et al., 2007) our data show a generally more severe and homogeneous clinical and protein pattern in patients with two nonsense/splice-site/out-of-frame mutations than patients with at least one missense/in-frame mutation, who showed a more variable pattern. The very severe protein phenotypes found here in some patients who carry missense mutations, particularly in exon 1, are illustrative of this variability.
Only one mutation was detected in 26% of the patients, comparable to a previous study (Saenz et al., 2005). The fact that most patients with only one CAPN3 mutation show a very similar clinical and protein phenotype to 2mut-patients makes it plausible to think these patients do carry a second mutation which has just not been detected. The existence of dominant mutations might be another explanation, though transmission of the disease in these families has not been reported. Therefore, our results indicate that 1mut-patients are most likely to have LGMD2A and a careful search for the second mutation with alternative detection methods able to detect large deletions or insertions (such as MLPA), sequencing of promoter regions or the use of cDNA as starting material (Chou et al., 1999; Chrobakova et al., 2004) would be indicated. The patient described where mutations have been found in two MD genes provides an important caveat, highlighting again the need for confirmation clinically that a LGMD2A diagnosis is likely.
In conclusion, this study has underlined specific clinical and biochemical LGMD2A features and shown the importance of muscle strength assessment in establishing the diagnosis. The presence of contractures, scapular winging, a typical pattern of muscle weakness and normal respiratory muscle strength are useful pointers to the diagnosis. Western blot analysis remains a useful tool for diagnosis, however, the overall pattern of calpain 3 expression needs to be looked at and not just the abundance of the full-sized band. The incorporation of a functional in vitro assay testing the autolytic activity of calpain 3 might further enhance our diagnostic capability. Genetic testing itself is not perfect, with the high rate of non-detection of the second mutation. A multidisciplinary approach to this common form of LGMD remains appropriate.
|
Supplementary material |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionSupplementary materialReferences |
|---|
Supplementary material is available at Brain online.
|
Acknowledgements |
|---|
The Diagnostic and Advisory Service for Rare Neuromuscular Diseases is funded by the NHS National Commissioning Group. The Newcastle Muscle Centre is supported by the Muscular Dystrophy Campaign. VS and the Wuerzburg laboratory are members of the German network on muscular dystrophies (MD-NET, 01GM0601) funded by the German ministry of education and research (BMBF, Bonn, Germany).
|
References |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionSupplementary materialReferences |
|---|
Anderson LV, Davison K. Multiplex western blotting system for the analysis of muscular dystrophy proteins. Am J Pathol (1999) 154:1017–22.
Anderson LV, Davison K, Moss JA, Richard I, Fardeau M, Tome FM, et al. Characterization of monoclonal antibodies to calpain 3 and protein expression in muscle from patients with limb-girdle muscular dystrophy type 2A. Am J Pathol (1998) 153:1169–79.
Anderson LV, Harrison RM, Pogue R, Vafiadaki E, Pollitt C, Davison K, et al. Secondary reduction in calpain 3 expression in patients with limb girdle muscular dystrophy type 2B and Miyoshi myopathy (primary dysferlinopathies). Neuromuscul Disord (2000) 10:553–9.[CrossRef][Web of Science][Medline]
Balci B, Aurino S, Haliloglu G, Talim B, Erdem S, Akcoren Z, et al. Calpain-3 mutations in Turkey. Eur J Pediatr (2006) 165:293–8.[CrossRef][Web of Science][Medline]
Bornemann A, Anderson LV. Diagnostic protein expression in human muscle biopsies. Brain Pathol (2000) 10:193–214.[Web of Science][Medline]
Brooke MH, Griggs RC, Mendell JR, Fenichel GM, Shumate JB, Pellegrino RJ. Clinical trial in Duchenne dystrophy. I. The design of the protocol. Muscle Nerve (1981) 4:186–97.[CrossRef][Web of Science][Medline]
Bushby K, Norwood F, Straub V. The limb-girdle muscular dystrophies – diagnostic strategies. Biochim Biophys Acta (2007) 1772:238–42.[Medline]
Bushby KM, Beckmann JS. The limb-girdle muscular dystrophies – proposal for a new nomenclature. Neuromuscul Disord (1995) 5:337–43.[CrossRef][Web of Science][Medline]
Canki-Klain N, Milic A, Kovac B, Trlaja A, Grgicevic D, Zurak N, et al. Prevalence of the 550delA mutation in calpainopathy (LGMD 2A) in Croatia. Am J Med Genet A (2004) 125:152–6.[Medline]
Chae J, Minami N, Jin Y, Nakagawa M, Murayama K, Igarashi F, et al. Calpain 3 gene mutations: genetic and clinico-pathologic findings in limb-girdle muscular dystrophy. Neuromuscul Disord (2001) 11:547–55.[CrossRef][Web of Science][Medline]
Chou FL, Angelini C, Daentl D, Garcia C, Greco C, Hausmanowa-Petrusewicz I, et al. Calpain III mutation analysis of a heterogeneous limb-girdle muscular dystrophy population. Neurology (1999) 52:1015–20.
Chrobakova T, Hermanova M, Kroupova I, Vondracek P, Marikova T, Mazanec R, et al. Mutations in Czech LGMD2A patients revealed by analysis of calpain3 mRNA and their phenotypic outcome. Neuromuscul Disord (2004) 14:659–65.[CrossRef][Web of Science][Medline]
de Paula F, Vainzof M, Passos-Bueno MR, de Cassia MPR, Matioli SR, L VBA, et al. Clinical variability in calpainopathy: what makes the difference? Eur J Hum Genet (2002) 10:825–32.[CrossRef][Web of Science][Medline]
Fanin M, Fulizio L, Nascimbeni AC, Spinazzi M, Piluso G, Ventriglia VM, et al. Molecular diagnosis in LGMD2A: mutation analysis or protein testing? Hum Mutat (2004) 24:52–62.[CrossRef][Web of Science][Medline]
Fanin M, Nascimbeni AC, Angelini C. Screening of calpain-3 autolytic activity in LGMD muscle: a functional map of CAPN3 gene mutations. J Med Genet (2007) 44:38–43.
Fanin M, Nascimbeni AC, Fulizio L, Angelini C. The frequency of limb girdle muscular dystrophy 2A in northeastern Italy. Neuromuscul Disord (2005) 15:218–24.[CrossRef][Web of Science][Medline]
Fanin M, Nascimbeni AC, Fulizio L, Trevisan CP, Meznaric-Petrusa M, Angelini C. Loss of calpain-3 autocatalytic activity in LGMD2A patients with normal protein expression. Am J Pathol (2003) 163:1929–36.
Fanin M, Pegoraro E, Matsuda-Asada C, Brown RH Jr, Angelini C. Calpain-3 and dysferlin protein screening in patients with limb-girdle dystrophy and myopathy. Neurology (2001) 56:660–5.
Fardeau M, Eymard B, Mignard C, Tome FM, Richard I, Beckmann JS. Chromosome 15-linked limb-girdle muscular dystrophy: clinical phenotypes in Reunion Island and French metropolitan communities. Neuromuscul Disord (1996a) 6:447–53.[CrossRef][Web of Science][Medline]
Fardeau M, Hillaire D, Mignard C, Feingold N, Feingold J, Mignard D, et al. Juvenile limb-girdle muscular dystrophy. Clinical, histopathological and genetic data from a small community living in the Reunion Island. Brain (1996b) 119(Pt 1):295–308.
Gardner-Medwin D, Walton JN. The clinical examination of the voluntary muscles. In: Disorders of voluntary muscles.—Walton JN, ed. (1974) 3rd. Edinburgh: Churchill Livingstone. 517–60.
Georgieva B, Todorova A, Tournev I, Mitev V, Plageras P, Kremensky I. 550delA mutation in the calpain 3 (CAPN3) gene: DMD/BMD, SMA, or LGMD2A – clinically misdiagnosed cases. Am J Med Genet A (2005) 136:399–400.[Medline]
Haravuori H, Vihola A, Straub V, Auranen M, Richard I, Marchand S, et al. Secondary calpain3 deficiency in 2q-linked muscular dystrophy: titin is the candidate gene. Neurology (2001) 56:869–77.
Huang Y, Verheesen P, Roussis A, Frankhuizen W, Ginjaar I, Haldane F, et al. Protein studies in dysferlinopathy patients using llama-derived antibody fragments selected by phage display. Eur J Hum Genet (2005) 13:721–30.[CrossRef][Web of Science][Medline]
Jia Z, Petrounevitch V, Wong A, Moldoveanu T, Davies PL, Elce JS, et al. Mutations in calpain 3 associated with limb girdle muscular dystrophy: analysis by molecular modeling and by mutation in m-calpain. Biophys J (2001) 80:2590–6.[Web of Science][Medline]
Johnson MA. Immunocytochemical analysis. In: Muscular dystrophy: methods and protocols.—Bushby KMD, Anderson LVB, eds. (2001) Totowa, NJ: Humana Press. 339–67.
Kawai H, Akaike M, Kunishige M, Inui T, Adachi K, Kimura C, et al. Clinical, pathological, and genetic features of limb-girdle muscular dystrophy type 2A with new calpain 3 gene mutations in seven patients from three Japanese families. Muscle Nerve (1998) 21:1493–501.[CrossRef][Web of Science][Medline]
Krahn M, Bernard R, Pecheux C, Hammouda el H, Eymard B, Lopez de Munain A, et al. Screening of the CAPN3 gene in patients with possible LGMD2A. Clin Genet (2006a) 69:444–9.[CrossRef][Web of Science][Medline]
Krahn M, Lopez de Munain A, Streichenberger N, Bernard R, Pecheux C, Testard H, et al. CAPN3 mutations in patients with idiopathic eosinophilic myositis. Ann Neurol (2006b) 59:905–11.[CrossRef][Web of Science][Medline]
Lanzillo R, Aurino S, Fanin M, Aguennoz M, Vitale F, Fiorillo C, et al. Early onset calpainopathy with normal non-functional calpain 3 level. Dev Med Child Neurol (2006) 48:304–6.[CrossRef][Web of Science][Medline]
Milic A, Canki-Klain N. Calpainopathy (LGMD2A) in Croatia: molecular and haplotype analysis. Croat Med J (2005) 46:657–63.[Web of Science][Medline]
Milic A, Daniele N, Lochmuller H, Mora M, Comi GP, Moggio M, et al. A third of LGMD2A biopsies have normal calpain 3 proteolytic activity as determined by an in vitro assay. Neuromuscul Disord (2007) 17:148–56.[CrossRef][Web of Science][Medline]
Minami N, Nishino I, Kobayashi O, Ikezoe K, Goto Y, Nonaka I. Mutations of calpain 3 gene in patients with sporadic limb-girdle muscular dystrophy in Japan. J Neurol Sci (1999) 171:31–7.[CrossRef][Web of Science][Medline]
Muntoni F, Bonne G, Goldfarb LG, Mercuri E, Piercy RJ, Burke M, et al. Disease severity in dominant Emery Dreifuss is increased by mutations in both emerin and desmin proteins. Brain (2006) 129:1260–8.
Ono Y, Shimada H, Sorimachi H, Richard I, Saido TC, Beckmann JS, et al. Functional defects of a muscle-specific calpain, p94, caused by mutations associated with limb-girdle muscular dystrophy type 2A. J Biol Chem (1998) 273:17073–8.
Piluso G, Politano L, Aurino S, Fanin M, Ricci E, Ventriglia VM, et al. Extensive scanning of the calpain-3 gene broadens the spectrum of LGMD2A phenotypes. J Med Genet (2005) 42:686–93.
Pogoda TV, Krakhmaleva IN, Lipatova NA, Shakhovskaya NI, Shishkin SS, Limborska SA. High incidence of 550delA mutation of CAPN3 in LGMD2 patients from Russia. Hum Mutat (2000) 15:295.[Medline]
Pollitt C, Anderson LV, Pogue R, Davison K, Pyle A, Bushby KM. The phenotype of calpainopathy: diagnosis based on a multidisciplinary approach. Neuromuscul Disord (2001) 11:287–96.[CrossRef][Web of Science][Medline]
Richard I, Brenguier L, Dincer P, Roudaut C, Bady B, Burgunder JM, et al. Multiple independent molecular etiology for limb-girdle muscular dystrophy type 2A patients from various geographical origins. Am J Hum Genet (1997) 60:1128–38.[Web of Science][Medline]
Richard I, Broux O, Allamand V, Fougerousse F, Chiannilkulchai N, Bourg N, et al. Mutations in the proteolytic enzyme calpain 3 cause limb-girdle muscular dystrophy type 2A. Cell (1995) 81:27–40.[CrossRef][Web of Science][Medline]
Richard I, Roudaut C, Saenz A, Pogue R, Grimbergen JE, Anderson LV, et al. Calpainopathy-a survey of mutations and polymorphisms. Am J Hum Genet (1999) 64:1524–40.[CrossRef][Web of Science][Medline]
Saenz A, Leturcq F, Cobo AM, Poza JJ, Ferrer X, Otaegui D, et al. LGMD2A: genotype-phenotype correlations based on a large mutational survey on the calpain 3 gene. Brain (2005) 128:732–42.
Talim B, Ognibene A, Mattioli E, Richard I, Anderson LV, Merlini L. Normal calpain expression in genetically confirmed limb-girdle muscular dystrophy type 2A. Neurology (2001) 56:692–3.
Urtasun M, Saenz A, Roudaut C, Poza JJ, Urtizberea JA, Cobo AM, et al. Limb-girdle muscular dystrophy in Guipuzcoa (Basque Country, Spain). Brain (1998) 121(Pt 9):1735–47.
Vainzof M, Anderson LV, McNally EM, Davis DB, Faulkner G, Valle G, et al. Dysferlin protein analysis in limb-girdle muscular dystrophies. J Mol Neurosci (2001) 17:71–80.[CrossRef][Web of Science][Medline]
Zatz M, de Paula F, Starling A, Vainzof M. The 10 autosomal recessive limb-girdle muscular dystrophies. Neuromuscul Disord (2003) 13:532–44.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  K. Bushby Diagnosis and management of the limb girdle muscular dystrophies Practical Neurology, December 1, 2009; 9(6): 314 - 323. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Kramerova, E. Kudryashova, B. Wu, S. Germain, K. Vandenborne, N. Romain, R. G. Haller, M. A. Verity, and M. J. Spencer Mitochondrial abnormalities, energy deficit and oxidative stress are features of calpain 3 deficiency in skeletal muscle Hum. Mol. Genet., September 1, 2009; 18(17): 3194 - 3205. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||