A mutation in the dimerization domain of filamin c causes a novel type of autosomal dominant myofibrillar myopathy

Abstract

Myofibrillar myopathy (MFM) is a human disease that is characterized by focal myofibrillar destruction and pathological cytoplasmic protein aggregations. In an extended German pedigree with a novel form of MFM characterized by clinical features of a limb-girdle myopathy and morphological features of MFM, we identified a co-segregating, heterozygous nonsense mutation (8130G-->A; W2710X) in the filamin c gene (FLNC) on chromosome 7q32.1. The mutation is the first found in FLNC and is localized in the dimerization domain of filamin c. Functional studies showed that, in the truncated mutant protein, this domain has a disturbed secondary structure that leads to the inability to dimerize properly. As a consequence of this malfunction, the muscle fibers of our patients display massive cytoplasmic aggregates containing filamin c and several Z-disk-associated and sarcolemmal proteins.

Figure 1

Immunohistochemical and ultrastructural analysis of aggregates in patient IV:3 from the pedigree with MFM. A, Several muscle fibers with structural changes in the trichrome-stained sections, showing intracytoplasmic accumulation of the Z-disk–associated proteins filamin c, myotilin, and desmin. Aggregates also show strong immunoreactivity for dystrophin and the sarcoglycans (as exemplified by γ-sarcoglycan), which are main components of the sarcolemmal dystrophin-glycoprotein complex (bar = 100 μm). B, Ultrastructural analysis of skeletal muscle, showing multiple fibers displaying intermyofibrillar granulofilamentous protein aggregates (original magnification: × 20,000).

Figure 2

A, Pedigree and genotypic analysis of the German family with autosomal dominant MFM segregating with FLNC. Marker haplotypes on chromosome 7q31.33-7q32.3 that are linked to MFM are indicated by blackened bars. Blackened and unblackened symbols represent clinically affected individuals and unaffected individuals, respectively; unblackened symbols with question marks represent individuals of undetermined disease status. Presence (+) or absence (−) of the FLNC mutation in those individuals available for mutation analysis is indicated. N/A = not available for genetic diagnosis. B, RFLP analysis of the 8130G→A mutation, by digestion of a PCR product with AluI. Undigested PCR product (120 bp) corresponds to the G allele, and digested PCR product (98 bp + unverifiable 22 bp) corresponds to the A allele. C, Sequence chromatograms demonstrating the heterozygous mutation (8130G→A; W2710X) found in affected family members (left) and the corresponding normal sequence (right).

Figure 3

Functional studies of the FLNC W2710X mutation. A, Circular dichroism spectrometry of filamin c domain 24 in the far UV range, performed with purified, bacterially expressed proteins. Note that the wild-type protein (wt) shows the typical β-sheet secondary structure, whereas the mutant-protein (mut) spectrum has lower amplitudes, indicating improper folding. deg = degrees; [Θ]_MRW = mean residue weight ellipticity. B, The same recombinant polypeptides as above were cross-linked using ethylene glycol bis(succinimidylsuccinate) (EGS). Proteins were separated by SDS-PAGE, and the gel was stained with Coomassie Brilliant Blue. Without cross-linking (−), only monomers (m) of the wild type and the truncated mutant polypeptide are visible. Note that, on cross-linking (+), the wild-type protein yields dimers (d), whereas the mutant protein yields aggregates (a) of higher molecular mass.