Dysferlin (DYSF) is a transmembrane protein linked to sarcolemmal repair mechanisms. Autosomal recessive mutations in the DYSF gene cause muscular dystrophies (MD): a limb girdle MD, the so-called LGMD 2B, with onset in the proximal lower limbs [1], and two distal MD, one initially affecting the gastrocnemius muscle during early adulthood, Miyoshi myopathy (MM) [2], and the other involving muscles of the anterior compartment (DMAT) [3]. Disease severity ranges from asymptomatic mutation carriers with elevated serum CK levels [4] and exercise intolerance [5] to severe functional disability [6]. Muscle weakness with pelvic girdle involvement on clinical examination distinguishes LGMD from MM. Other common phenotypes include a “proximo-distal” phenotype, characterized by simultaneous distal and proximal weakness onset [6]. Of note, a recent MRI study suggests that all patients have radiographic evidence of proximo-distal muscle involvement, regardless of the clinical phenotype [7].

The DYSF gene maps to chromosome 2p12-p14, contains 55 exons and is transcribed as an 8.5 Kb major transcript mainly expressed in skeletal and cardiac muscles. The protein product is a 230 kDa molecule involved in muscle fibre repair [8, 9]. More than 400 DYSF mutations have been described in the Leiden muscular dystrophy database (http://www.dmd.nl); however, a mutational hot spot has not been detected [10–12]. Ethnic clusters have been described in Jews from Libya [13] and the Caucasus region [14], as well as in the Italian [15] and Spanish [16] populations. Four mutations account for 60 % of all mutations in Japanese patients with MM [17]. Interestingly, the type of mutation does not correlate with phenotypic features. In fact, the same mutation has been found to be associated with a broad spectrum of inter—and intra-familial clinical phenotypes [18–22].

Within this range of clinical and genetic heterogeneity, we sought to determine frequently encountered DYSF gene mutations and phenotypes in the Swiss population and to uncover possible founder effects.

Evaluation of the present cohort of 6 families brings new genetic information about dysferlinopathy in Switzerland, and novel data regarding a potential founder effect in Central Switzerland. The fact that unrelated families with identical mutations come from the same region indicates that these mutations are common, and indeed, haplotype analysis of 2 families carrying the mutation c.3031 + 2T>C suggests a possible founder effect for this mutation in Switzerland. The limitations of this study are twofold: the analysis is limited by the small cohort size and 3 out of 13 patients had not been examined clinically.

From a clinical point of view, the onset age in early adulthood confirms previous findings [6, 7, 24]. In addition, late-onset [25] or congenital forms were not encountered [26]. Three out of 4 patients from Central Switzerland presented with the PD phenotype, while another patient with the same mutation was considered to represent the MM phenotype. Upper limb weakness was not a presenting feature in the current cohort. Some reports suggest that the proximo-distal phenotype occurs in up to one third of dysferlinopathy cases [6]. Of note, 9 out of 10 patients developed distal upper limb weakness later in the disease course, which has been frequently reported [27–29]. Muscle MRI is known to reveal fatty replacement of distal and proximal lower limb muscles in the early stages of dysferlinopathy, including patients with clinically isolated proximal or distal involvement [7, 30]. However, an adolescent patient with a MM phenotype (C-7) showed a normal pelvic and whole lower limb muscle MRI, similarly to a patient described with isolated hyperCKemia [31].

Interestingly, we found clinical phenotype variability within families and within unrelated carriers with similar mutations. Such variability has already been reported in Libyan Jews [13] and was found to be particularly relevant in a 6-generation, highly consanguineous family originating from an isolated village in Dagestan [32]. This observation may indicate that to date, unknown genetic or environmental modifiers play a role in influencing the clinical phenotypes of the mutated DYSF gene [22]. Accordingly, it would be interesting to study micro-RNAs known to be involved in many biological processes, including epigenetic changes. Other cardinal features of our cohort were very high CK levels, as previously described [7], the absence of cardiac impairment (nevertheless encountered in few patients [6, 25, 33, 34] and present in a knock out mouse model [35]), and the risk of developing respiratory failure late in the disease course [36, 37], as observed in one of our patients.

EMG showed myopathic changes and pathologic spontaneous activity. In one case, the biopsy revealed striking endomysial inflammation, in line with frequently observed inflammatory changes in dysferlinopathy and other muscular dystrophies [18, 38]. Immunosuppressive treatment, however, fails to improve muscle weakness [39]. It has been postulated that dysferlin may be linked to membrane repair and/or inflammatory activation [40], and in a few cases, splice site mutations disrupting dysferlin are known to produce inflammation [18]. Conversely, most of our patients with splice mutations showed no evidence of inflammation on muscle biopsy, indicating the lack of clear correlation between inflammation and genotype. On Western blot (WB) 3 cases showed a consistent reduction in dysferlin, as has been previously described [41].

In dysferlinopathies, genotype is not predictive of disease severity, and genotype-phenotype correlations are not clear-cut [15, 42]. However, correlation has been found between certain geographical areas and the age of onset, notably in Japanese and Italian MM patients [17], who develop symptoms earlier than patients with LGMD 2B [10]. The most severely affected patients in our study (Patients A-1 and A-2, described previously [12]) carried the truncating mutation c.2217C>A (p.Tyr739X) with a disease duration of 17 and 24 years, respectively, before the patients became wheelchair-bound. One of the older patients (C-8), carrying the c.3031 + 2T>C homozygous mutation, became wheelchair bound after a disease duration of 27 years, and patient D-9 showed relatively mild impairment 20 years after onset. However, in all other patients carrying this mutation, disease duration was shorter; thus precluding longitudinal comparisons. A longer follow-up is necessary to assess the onset and time-course of the various disabling symptoms.

Our study confirms the broad phenotypic heterogeneity, often accompanied by markedly elevated CK levels, associated with DYSF mutations in the Swiss population. Due to a potential founder effect in the past, two mutations appear to be endemic in central Switzerlans. These findings are important for genetic counselling and should facilitate targeted molecular diagnosis of dysferlin deficiency in patients of Swiss origin.