Molecular bandages: inside-out, outside-in repair of cellular membranes. Focus on “Myoferlin is critical for endocytosis in endothelial cells”

the plasma membrane separates the external and internal cellular environments. Many processes that injure the plasma membrane are a prelude to cell death (i.e., necrosis, apoptosis, and autophagy). Therefore, to maintain homeostasis and survival, cells must have mechanisms of membrane repair that limit intrusion of the external environment to the interior of the cell. At the heart of this repair machinery are highly regulated and coordinated endocytic and exocytic processes.

The article published in American Journal of Physiology-Cell Physiology by Bernatchez et al. (5) focuses on the role of myoferlin, caveolin-1, and dynamin in receptor-mediated and injury-induced endocytosis. Myoferlin is a 230-kDa transmembrane protein that is expressed primarily in cardiac and skeletal muscle. The study by Bernatchez et al. shows that myoferlin regulates caveolae/lipid raft and clathrin-mediated endocytosis but the greater effect is on the former process. Though a role for these two endocytic processes has been established for receptor trafficking, the intriguing interplay of myoferlin, caveolin-1, and dynamin in endocytosis-induced membrane repair is of note and worth highlighting for readers.

Membrane repair following injury was initially thought to be a passive event that was mediated by resealing of the lipid bilayer (15). However, this idea was later expanded to suggest that large disruptions (>1 μm) of the plasma membrane undergo “patch repair” where Ca²⁺ influx through membrane lesions triggers exocytosis of cytoplasmic vesicles that fuse with the injured membrane (6). Akin to synaptic vesicle fusion that releases neurotransmitters, the early insight that calcium-regulated exocytosis was involved in membrane repair provided a useful working hypothesis. Subsequent investigations then turned to identifying which intracellular vesicles were used to repair the damaged plasma membrane. These vesicles required three characteristics: 1) regulation by Ca²⁺, 2) capability of assembly to create a large area of patch, and 3) ability to fuse with the plasma membrane. Lysosomes were shown to fit all three criteria. Lysosomal membranes contain synaptotagmin VII (Syt VII), a calcium sensor (8), and, notably, mice deficient in Syt VII have defective membrane repair and show muscle fiber invasion by leukocytes, which together result in myopathy (7). The Ca²⁺ trigger induces fusion of lysosomes to form large membrane regions (3) that can then fuse with the plasma membrane (22). Inhibition of lysosome exocytosis, by modulating Syt VII activity or causing lysosomal aggregation, interferes with membrane resealing (21). The process of sensing injury, activating lysosomes, and mediating membrane resealing is rapid (occurring within seconds) and involves lysosomes in close proximity to the plasma membrane (17). Collectively, these observations suggest that exocytosis is highly regulated and critical to membrane repair.

Later studies established that endocytosis is of equal importance and may work in concert with exocytosis for membrane repair. It was initially shown that membrane injury not due to mechanical injury (i.e., exposure to toxins that create pores) results in Ca²⁺-dependent repair of membranes (16); however, although exocytosis was induced, an equally rapid and robust endocytic response that internalized the pore to maintain membrane integrity was also observed. The important implication of the current work is the identification of a molecular complex consisting of myoferlin, caveolin-1, and dynamin that may be integral to membrane repair (5).

Myoferlin is akin to other ferlin proteins in having C2 domains (which may serve as Ca²⁺ sensors) (10, 11) and a COOH-terminal membrane-spanning domain, but most of its structure resides in the cytoplasm. Myoferlin associates with both the plasma membrane and the nuclear membrane and can also be found in the nucleoplasm. Myoferlin is present in high amounts in regenerating muscle; mice deficient in myoferlin experience defective myogenesis and muscle regeneration (10, 12). Myoferlin was initially thought to be involved in plasma membrane dynamics because proteins in the ferlin family are homologous to fer-1, a spermatogenesis factor in Caenorhabditis elegans that mediates spermatid vesicle/plasma membrane fusion (2, 4). Bernatchez et al. confirm the interaction of myoferlin with the plasma membrane (5), but, in addition, they show interaction of myoferlin with caveolin-1 and localization in caveolae. This interaction and localization are necessary for membrane repair because small interfering RNA knockdown of either myoferlin or caveolin-1 leads to an equal degree of loss of membrane resealing following injury.

Caveolin-1 is a structural component of caveolae, which are specialized, lipid-rich microdomains that coordinate a variety of functional events (20). The budding (i.e., endocytosis) of caveolae from the plasma membrane requires dynamins, which are GTPases that are involved in various cellular processes. Dynamins self-assemble and oligomerize at the necks of plasma membrane caveolae, thereby resulting in caveolar budding and retention of dynamin in the membrane (9, 19). The trigger for this budding has remained elusive; however, on the basis of the molecular interactions proposed in the current study, we speculate that cellular stress, as sensed by myoferlin via Ca²⁺ influx, may be key to localized regulation of caveolin-dynamin dynamics.

Mutation or knockdown of caveolin-3, a muscle-specific caveolin, results in myopathies (1, 13, 25). Dysferlin (a member of the ferlin family with a function similar to myoferlin) is dependent on caveolin-3 expression for its retention in the membrane; knockdown of caveolin-3 results in mislocalized dysferlin and its rapid internalization (14). Perhaps the retention of dysferlin in the plasma membrane via caveolin-3 is a means to localize and anchor this “sensor of injury” to membranes and to facilitate rapid protective response. In this regard, it is interesting to note that cardiac myocyte-specific overexpression of caveolin-3 protects the heart from ischemia-reperfusion injury (which is known to disrupt membranes and lead to intracellular influx of Ca²⁺) (18, 24). Importantly, overexpression of caveolin-3 leads to the preservation of the ultrastructure of sarcolemmal membranes and intracellular organelles, mimicking the protection induced by sublethal ischemia before lethal hypoxic stress (24). Although the mechanism is unknown, multiple cycles of sublethal ischemia have been shown to preserve myocardial membrane and intracellular ultrastructure (18). With respect to membrane repair, a similar observation has been made: a second membrane disruption at the same site of original injury repairs more rapidly, an effect that occurs via endocytosis (23). Such results suggest that multiple exposures to injury enhance the efficiency of endocytosis and perhaps the maintenance or repair of membrane integrity. The study by Bernatchez et al. shows that increased myoferlin in a reconstituted system is sufficient to increase endocytosis independent of injury. The findings, however, lead to several questions. For example, does caveolin expression represent a control point for regulating the efficiency of endocytosis? Do membranes that have greater expression of caveolins and caveolae have increased expression and activity of ferlins and dynamins at the cell membrane? Are budded caveolae the “raw material” for sealing damaged plasma membranes? Can ferlins, caveolins, and dynamins be targeted as possible therapeutics for myopathic disease processes?

The current study defines three components (i.e., myoferlin, caveolin, and dynamin) of a molecular bandage that may be critical to the integrity of cellular membrane and may provide a means to regulate a variety of disease processes. Involvement of other elements, such as membrane tension and the cytoskeleton, may also contribute to membrane repair. A challenge for the future is to define the temporal nature of endocytic and exocytic processes and if the interaction of myoferlin, caveolins, and dynamins and their localization in caveolae represents a refinement or a paradigm shift (Fig. 1) in terms of membrane repair following injury.

Fig. 1.

Schematic of the classic model and a potential new model of membrane repair. A: in the classic model of plasma membrane repair, membrane injury leads to Ca²⁺ influx, which induces exocytic and endocytic repair pathways. In the exocytic pathway, lysosomes contribute membrane to patch the injury site. In the endocytic pathway, vesicles bud and remove the wound site from the plasma membrane. B: in the theorized new model, the exocytic repair pathway remains intact but the endocytic repair pathway takes a more prevalent role. In this model, injury-induced Ca²⁺ influx is sensed by myoferlin localized at the plasma membrane in caveolae and initiates dynamin-dependent endocytosis. The budded caveolae, due to their close proximity to the plasma membrane, may then patch the injured area, possibly by entering into an exocytic repair mode.

GRANTS

This work was supported by grants from the American Heart Association (060039N) and National Institutes of Health (HL-091071 and HL-066941).