The C2 domains of dysferlin: roles in membrane localization, Ca2+ signalling and sarcolemmal repair.

Dysferlin is an integral membrane protein of the transverse tubules of skeletal muscle that is mutated or absent in limb girdle muscular dystrophy 2B and Miyoshi myopathy. Here we examine the role of dysferlin's seven C2 domains, C2A through C2G, in membrane repair and Ca2+ release, as well as in targeting dysferlin to the transverse tubules of skeletal muscle. We report that deletion of either domain C2A or C2B inhibits membrane repair completely, whereas deletion of C2C, C2D, C2E, C2F or C2G causes partial loss of membrane repair that is exacerbated in the absence of extracellular Ca2+ . Deletion of C2C, C2D, C2E, C2F or C2G also causes significant changes in Ca2+ release, measured as the amplitude of the Ca2+ transient before or after hypo-osmotic shock and the appearance of Ca2+ waves. Most deletants accumulate in endoplasmic reticulum. Only the C2A domain can be deleted without affecting dysferlin trafficking to transverse tubules, but Dysf-ΔC2A fails to support normal Ca2+ signalling after hypo-osmotic shock. Our data suggest that (i) every C2 domain contributes to repair; (ii) all C2 domains except C2B regulate Ca2+ signalling; (iii) transverse tubule localization is insufficient for normal Ca2+ signalling; and (iv) Ca2+ dependence of repair is mediated by C2C through C2G. Thus, dysferlin's C2 domains have distinct functions in Ca2+ signalling and sarcolemmal membrane repair and may play distinct roles in skeletal muscle. KEY POINTS: Dysferlin, a transmembrane protein containing seven C2 domains, C2A through C2G, concentrates in transverse tubules of skeletal muscle, where it stabilizes voltage-induced Ca2+ transients and participates in sarcolemmal membrane repair. Each of dysferlin's C2 domains except C2B regulate Ca2+ signalling. Localization of dysferlin variants to the transverse tubules is not sufficient to support normal Ca2+ signalling or membrane repair. Each of dysferlin's C2 domains contributes to sarcolemmal membrane repair. The Ca2+ dependence of membrane repair is mediated by C2C through C2G. Dysferlin's C2 domains therefore have distinct functions in Ca2+ signalling and sarcolemmal membrane repair.

Nanodysferlins support membrane repair and binding to TRIM72/MG53 but do not localize to t-tubules or stabilize Ca2+ signaling.

Mutations in the DYSF gene, encoding the protein dysferlin, lead to several forms of muscular dystrophy. In healthy skeletal muscle, dysferlin concentrates in the transverse tubules and is involved in repairing the sarcolemma and stabilizing Ca2+ signaling after membrane disruption. The DYSF gene encodes 7-8 C2 domains, several Fer and Dysf domains, and a C-terminal transmembrane sequence. Because its coding sequence is too large to package in adeno-associated virus, the full-length sequence is not amenable to current gene delivery methods. Thus, we have examined smaller versions of dysferlin, termed "nanodysferlins," designed to eliminate several C2 domains, specifically C2 domains D, E, and F; B, D, and E; and B, D, E, and F. We also generated a variant by replacing eight amino acids in C2G in the nanodysferlin missing domains D through F. We electroporated dysferlin-null A/J mouse myofibers with Venus fusion constructs of these variants, or as untagged nanodysferlins together with GFP, to mark transfected fibers We found that, although these nanodysferlins failed to concentrate in transverse tubules, three of them supported membrane repair after laser wounding while all four bound the membrane repair protein, TRIM72/MG53, similar to WT dysferlin. By contrast, they failed to suppress Ca2+ waves after myofibers were injured by mild hypoosmotic shock. Our results suggest that the internal C2 domains of dysferlin are required for normal t-tubule localization and Ca2+ signaling and that membrane repair does not require these C2 domains.

Elevated Ca2+ at the triad junction underlies dysregulation of Ca2+ signaling in dysferlin-null skeletal muscle.

Dysferlin-null A/J myofibers generate abnormal Ca2+ transients that are slightly reduced in amplitude compared to controls. These are further reduced in amplitude by hypoosmotic shock and often appear as Ca2+ waves (Lukyanenko et al., J. Physiol., 2017). Ca2+ waves are typically associated with Ca2+-induced Ca2+ release, or CICR, which can be myopathic. We tested the ability of a permeable Ca2+ chelator, BAPTA-AM, to inhibit CICR in injured dysferlin-null fibers and found that 10-50 nM BAPTA-AM suppressed all Ca2+ waves. The same concentrations of BAPTA-AM increased the amplitude of the Ca2+ transient in A/J fibers to wild type levels and protected transients against the loss of amplitude after hypoosmotic shock, as also seen in wild type fibers. Incubation with 10 nM BAPTA-AM led to intracellular BAPTA concentrations of ∼60 nM, as estimated with its fluorescent analog, Fluo-4AM. This should be sufficient to restore intracellular Ca2+ to levels seen in wild type muscle. Fluo-4AM was ∼10-fold less effective than BAPTA-AM, however, consistent with its lower affinity for Ca2+. EGTA, which has an affinity for Ca2+ similar to BAPTA, but with much slower kinetics of binding, was even less potent when introduced as the -AM derivative. By contrast, a dysferlin variant with GCaMP6fu in place of its C2A domain accumulated at triad junctions, like wild type dysferlin, and suppressed all abnormal Ca2+ signaling. GCaMP6fu introduced as a Venus chimera did not accumulate at junctions and failed to suppress abnormal Ca2+ signaling. Our results suggest that leak of Ca2+ into the triad junctional cleft underlies dysregulation of Ca2+ signaling in dysferlin-null myofibers, and that dysferlin's C2A domain suppresses abnormal Ca2+ signaling and protects muscle against injury by binding Ca2+ in the cleft.