Nebulin, a major player in muscle health and disease

Abstract

Nebulin is a giant 600- to 900-kDa filamentous protein that is an integral component of the skeletal muscle thin filament. Its functions have remained largely nebulous because of its large size and the difficulty in extracting nebulin in a native state from muscle. Recent improvements in the field, especially the development of knockout mouse models deficient in nebulin (NEB-KO mice), indicate now that nebulin performs a surprisingly wide range of functions. In addition to a major role in thin-filament length specification, nebulin also functions in the regulation of muscle contraction, as indicated by the findings that muscle fibers deficient in nebulin have a higher tension cost, and develop less force due to reduced myofilament calcium sensitivity and altered crossbridge cycling kinetics. In addition, the function of nebulin extends to a role in calcium homeostasis. These novel functions indicate that nebulin might have evolved in vertebrate skeletal muscles to develop high levels of muscle force efficiently. Finally, the NEB-KO mouse models also highlight the role of nebulin in the assembly and alignment of the Z disks. Notably, rapid progress in understanding the roles of nebulin in vivo provides clinically important insights into how nebulin deficiency in patients with nemaline myopathy contributes to debilitating muscle weakness.—Labeit, S., Ottenheijm, C. A. C., Granzier, H. Nebulin, a major player in muscle health and disease.

Whereas the contractile proteins myosin and actin have been intensely characterized since the beginning of the past century, nebulin was discovered comparatively late because of its giant size that requires unconventional low-porosity gels that resolve megadalton-sized proteins (1, 2). Such gel systems detect in skeletal muscle extracts a protein of ∼600–900 kDa (Fig. 1A) that was named nebulin because of its nebulous nature (1). Initial data on the sarcomeric layout of nebulin were derived from epitope mapping of nebulin-specific monoclonal antibodies. These early studies (1) demonstrated that the giant nebulin polypeptide forms a filament-like structure that is coextensive with actin filaments in situ (see Fig. 1B). Molecular insights into the nature of the association of nebulin with the thin filament came from subsequent cDNA cloning and sequencing studies (3, 4). This approach showed that ∼90% of the primary structure of nebulin is composed of ∼35-residue α-helical domains (Fig. 1B). Using recombinant proteins consisting of multiple nebulin repeat domains, it has been demonstrated that under physiological conditions, nebulin repeats bind actin with high affinity (5). Each of these repeats has a central SDxxYK consensus sequence that binds F-actin. Individual domains are organized into 7 domain super-repeats that match the 7-actin monomer repeat unit of the actin filament, further establishing a close association between nebulin and F-actin. The sequence of the nebulin N terminus is found near the pointed end of the thin filament; the C terminus of nebulin contains a serine-rich domain and an SRC homology 3 (SH3) domain (3, 4) that is embedded in the Z-disk region of the sarcomere. Overall, the large central region of nebulin is best understood and can be considered an integral component of the skeletal muscle thin filament.

Figure 1.

Nebulin is a large sarcomeric protein that is coextensive with the actin filament of the skeletal muscle sarcomere. A) Gel electrophoresis of human and mouse soleus muscle extracts reveals nebulin as a 770-kDa (human) and 730-kDa (mouse) protein. The extremely large titin protein is also shown, as is the small myosin heavy chain (MHC). B) Top: layout of nebulin in the skeletal muscle sarcomere. Bottom: modular organization of nebulin protein. Binding sites of nebulin-binding proteins are also shown: myopalladin (ref. 62); titin (ref. 63); the two reported binding CapZ sites, CapZ (1) (ref. 63) and CapZ (2) (ref. 43); desmin (refs. 15, 16); and tropomodulin (ref. 35).

RECENT RESULTS

The organization of nebulin is highly modular both at the protein and genomic levels (Fig. 1B). The nebulin gene contains a large number of exons [183 exons in humans (6) and 166 in mice (7)], the majority of which are ∼105-bp exons with splice donor/acceptor positions that map to the center of SDxxYK motifs. This modular organization at the gene level allows the generation of splice variants by exon shuffling. Knowledge of the nebulin gene structure has given rise to full gene covering arrays where each exon is represented by specific probes. This makes it possible to determine the specific set of nebulin exons transcribed in a certain muscle tissue as a function of developmental stage or disease. Consistent with earlier RT-PCR studies (3, 6), exon microarray analysis revealed nebulin splice isoform diversity in different tissue types and during postnatal muscle differentiation (8). Future studies are needed to establish how the splice isoform diversity is regulated, for example how the nebulin transcriptome adapts dynamically during the development of diseases. The array studies are likely to guide such studies, by refining maps for the differentially spliced nebulin exons, and the tissue contexts where they are skipped or included.

The nebulin field has been held back by the absence of methods to isolate and purify full-length nebulin, but recent progress has made this now possible (9, 10), which is expected to lead to major new insights. Knowledge of the nebulin gene structure has also made it possible to design genetic strategies for the targeted inactivation of the murine nebulin gene by homologous recombination. Witt et al. (11) deleted the 5′-UTR TATA box as well as the start ATG, which resulted in the loss of nebulin transcription. Similarly, Chen and colleagues (12) inactivated the murine nebulin gene. These gene-targeting strategies have resulted in mice with skeletal muscles that are essentially nebulin-free, making it possible to test the long-debated physiological roles of nebulin. The first studies focused on the role of nebulin in specifying thin-filament length. Witt et al. (11) performed an electron microscopy study using a technique that decorates the thin filament with phalloidin-gold and reported that thin-filament lengths in wild-type tibialis cranialis muscle are a constant ∼1.2 μm, but in nebulin-deficient muscle, the average length is ∼0.8 μm, with a range from ∼0.4 to 1.2 μm (Fig. 2A, B). Work by Bang et al. (12) on a similar nebulin-knockout (NEB-KO) model, using confocal microscopy on 1-d-old mice, reported thin-filament lengths in wild-type muscles of ∼1.15–1.3 μm (depending on muscle type), and in the absence of nebulin, an average length of ∼1.0 μm in all muscle types. The researchers did not report a thin-filament length gradient, unlike Witt et al. (11), which might be due to the different ages of mice that were used (∼10 days in ref. 11 and 1 d in ref. 12), or to the fact that Bang et al. (12) used an optical method that might not be ideal for detecting length gradients. However, the two studies on nebulin-deficient muscle agree that nebulin plays a critical role in the specification of thin-filament length and that in its absence the average thin-filament length is shorter.

Figure 2.

Nebulin specifies thin-filament length and the level of maximal active force. A) Top: example of sarcomere labeled with phalloidin-gold. Middle: explanation of how distance of gold bead to Z disk was measured. Bottom: histogram of results for wild-type and NEB-KO sarcomeres. Histogram suggests a constant thin-filament length in wild-type muscle and shorter and nonuniform length in NEB-KO muscle. Illustration based on Witt et al. (11). B) Schematic of half sarcomere showing thin-filament lengths in wild-type (top) and NEB-KO muscle (bottom). C) Comparison of measured force-sarcomere length relation of wild-type and nebulin-deficient NEB-KO muscle fibers. Illustration based on Ottenheijm et al. (27). D) Schematic of 2-state crossbridge cycle and effect of nebulin on rate constant of transition from non-force-generating crossbridge to force-generating crossbridge f_app and on the reverse rate constant g_app. Nebulin increases f_app and slows g_app, which has the net effect that the fraction of all available crossbridges that generate force will be increased.

Studies on NEB-KO mice have also demonstrated a role for nebulin in establishing lateral connectivity of myofibrils. Myofibrils from nebulin-free skeletal muscle lose their tight lateral register on stretch (13), results similar to previous studies on desmin-KO mice that showed that removal of desmin enhances the susceptibility of muscle to lateral myofibrillar displacements (14). Since previous studies localized binding sites for desmin in the nebulin C-terminal region within the periphery of the Z disk (15, 16), it is likely that nebulin together with desmin forms an intermyofibrillar connecting complex (for further details, see refs. 13, 17). Note that the Z-disk functions of nebulin might be performed in cardiac muscle by nebulette, an ∼100-kDa nebulin-like protein that contains multiple actin-binding repeats and that shares extensive similarity with the C-terminal region of nebulin (18, 19). Nebulette is far too short to reach to the thin-filament tip, and the functions of nebulette that overlap those of nebulin must be restricted to near-Z-disk functions, including establishing lateral connectivity of myofibrils.

Nebulin-free muscle fibers not only differ in thin-filament length but also have altered contractile properties(Fig. 2C). Studies on nebulin-deficient fibers indicate a severely reduced isometric force production (11, 12, 20–22). The mechanistic basis includes a reduced force-generating thick-thin-filament overlap zone, because of shortened thin filaments (Fig. 2). Interestingly, nebulin-free muscles also have changes in the regulation of contraction. Muscle contraction is driven by the interaction between the myosin-based crossbridges and actin, and the level of force a muscle generates is proportional to the number of crossbridges in the force-generating state. It is generally accepted that this interaction between actin and myosin is regulated through a steric hindrance mechanism in which tropomyosin and troponin control the conversion between interaction-permissive and nonpermissive states (23). In addition, recently provided evidence suggest that nebulin also contributes to the regulation of crossbridge cycling kinetics by altering the crossbridge attachment and detachment kinetics (12, 21). During the crossbridge cycle, unbound non-force-generating crossbridges move to an actin-bound force-generating state, followed by ATP-driven crossbridge release back to the non-force-generating state. This transition between force- and non-force-generating crossbridge states can be described by two apparent rate constants: one for crossbridge attachment (f_app), and one for crossbridge detachment (g_app) (24).These two rate constants determine the fraction of force-generating crossbridges during activation, and a change in one or both will affect this fraction and thus force production. g_app is directly proportional to the ATP consumption rate normalized to tension generation (i.e., tension cost) and can therefore be estimated from the simultaneous measurement of ATP consumption rate and tension in activated muscle fibers (24). Tension-cost studies on nebulin-deficient muscle (21) revealed significantly higher tension cost in nebulin-deficient muscle, thus indicating a faster g_app when nebulin is absent. Likewise, studies on another NEB-KO mouse model (20) reported higher velocity of unloaded shortening in nebulin-deficient muscle, also suggesting that g_app is higher when nebulin is absent. These findings are consistent with much earlier results by Root and Wang (25), who used in vitro motility assays to show that nebulin fragments reduce the sliding velocity of F-actin over myosin.

The rate constant of force redevelopment (K_tr), determined by imposing a rapid release-restretch protocol on an activated fiber and then measuring force redevelopment (24), is proportional to f_app + g_app, and the fraction of force-generating crossbridges to f_app/K_tr. K_tr experiments revealed that force redevelopment is slower in nebulin-deficient muscle (20, 21). Thus, the decrease in K_tr of nebulin-deficient muscle, together with the finding that g_app is increased, implies that f_app is reduced and that the reduction is larger than the increase in g_app (for details, see ref. 21). Combined, these findings lead to the conclusion that the fraction of force-generating crossbridges [f_app/(f_app+g_app)] is reduced in nebulin-deficient muscle. Thus, recent studies suggest that nebulin increases the rate of crossbridge attachment and reduces the rate of crossbridge detachment, and that as a result, the number of force-generating crossbridges is increased. These findings appear schematically in Fig. 2D. The mechanism by which nebulin affects crossbridge cycling needs further investigation. Previous work (26) has shown that nebulin associates with the actin N terminus in subdomain 1, where the myosin crossbridge also binds. Thus, the presence of nebulin at or near the S1 binding site might enhance the binding of crossbridges and slow their detachment. A recent intriguing study by Wang and colleagues (9) on single-nebulin molecules suggests that nebulin exerts a compressive force on the thin filament, and this could potentially also be a mechanism by which nebulin affects thin-filament activation and crossbridge cycling kinetics.

Chandra et al. (21) estimated that the effect of nebulin on crossbridge kinetics enhances the force-generating capacity of a muscle by ∼50% and increases the economy of contraction by ∼35%. These estimations are in line with findings reported by Bang et al. (20). Clearly, nebulin is a major factor in determining the level of force and the energetic cost of force production in skeletal muscle. Consistent with this role of nebulin in the regulation of crossbridge cycling kinetics, recent studies on muscle fibers from patients with nemaline myopathy (NM) with severely reduced nebulin protein levels revealed that, in addition to altered thin-filament length (27), changes in crossbridge cycling kinetics contribute to the muscle weakness observed in these patients (28).

Chandra et al. (21) also measured active force at a range of calcium levels and the obtained force-pCa relations were markedly shifted to the right in nebulin-deficient muscle fibers, with a 0.16-U reduction in pCa₅₀ (pCa that gives the half-maximal force level), suggesting that nebulin participates in the regulation of muscle contraction. This finding is supported by earlier in vitro studies with a recombinant protein consisting of 5 nebulin repeats (29), which revealed that nebulin increases the affinity of the tropomyosin-troponin complex for F-actin. Interestingly, the studies by Witt et al. (11) and Bang et al. (12) on NEB-KO fibers found no difference in calcium sensitivity. It is possible that the difference in sarcomere length between the studies provides an explanation. The two studies that did not detect a difference in calcium sensitivity were performed at long sarcomere lengths [∼2.5 μm (11) and ∼2.6 μm (12)], whereas the study that did show a difference (21) was performed at ∼2.0 μm. The implication is that nebulin plays a role in the length dependence of activation with a much larger ΔpCa₅₀ in the NEB-KO fibers than in wild-type fibers. Clearly, this issue needs to be investigated further, including the mechanism by which nebulin increases calcium sensitivity at short sarcomere length, such as the effect of nebulin on the troponinin/tropomyosin system and on crossbridge-based thin-filament activation. It is well known that as sarcomere length increases, muscle becomes more calcium sensitive. This length dependence of activation is most prominent in cardiac muscle (and is thought to underlie the Frank-Starling law of the heart) but is much less pronounced in skeletal muscle (23). The presence of nebulin provides an explanation for why skeletal muscle has less length dependence: The presence of nebulin increases calcium sensitivity at short length. Thus nebulin is an important player in a wide range of skeletal muscle characteristics. The appearance of nebulin in vertebrates (30, 31) can be considered to have evolved because of the need for the generation of high levels of force in an energetically efficient manner.

Changes in calcium homeostasis have also been noted in nebulin-deficient muscle. A striking up-regulation of sarcolipin (SLN), an inhibitor of SERCA, occurs in NEB-KO mice (11, 22, 32). This up-regulation might be viewed as an adaptation in NEB-KO mice as an attempt to increase cytosolic calcium levels and counteract reduced myofilament calcium sensitivity. The mechanism by which nebulin deficiency up-regulates SLN requires further future studies, as does the functional role of SLN up-regulation, for example by crossing SLN-KO mice and NEB-KO mice. Taken together, the above discussed studies clearly indicate that nebulin is not merely involved in thin-filament length regulation but also acts as a regulator of muscle contraction.

CONTROVERSY AND RESEARCH CHALLENGES

Nebulin as a thin-filament length ruler?

Earlier studies revealed that the electrophoretic mobility of nebulin from different muscle types correlates with thin-filament length (33, 34), findings that resulted in the widely held view that nebulin functions as a thin-filament length ruler. In vitro interaction studies identified a binding site for the F-actin capping protein tropomodulin (Tmod) at the nebulin N-terminal end, providing an elegant mechanism for how nebulin could function as a molecular ruler (35). Tmod1 is the Tmod gene family member that is most abundantly expressed in striated muscles, and both cell culture-based studies and biochemical studies had provided evidence for a critical Tmod1 role in F-actin pointed-end termination. For example, microinjection of antibodies against Tmod1 into cultured chick cardiac myocytes resulted in a marked lengthening of thin filaments at their pointed end, supporting the hypothesis that Tmod is required to restrict actin filament length in vivo (36). However, various recent findings have challenged this elegant mechanism. For example, targeted inactivation of Tmod1 in ES cells demonstrated that Tmod1 is not essential for thin-filament capping (37), and a recent immunofluorescence microscopy study determined that the N termini of nebulin and Tmod1 do not precisely colocalize in rabbit skeletal muscles (26, 38). Although these results need to be confirmed by higher-resolution methods, they do challenge the nebulin-ruler hypothesis and suggest a nebulin-independent mechanism controls actin filaments that have grown beyond the length of nebulin.

The finding that skeletal muscles of NEB-KO mice have thin filaments that are irregular in length and that are on average much shorter than in wild-type mice support that in vivo nebulin does participate in thin-filament length regulation (see Fig. 2A, B). However, the exact mechanism, as well as the specific contribution of nebulin, is debated. Earlier studies in cardiac myocytes in cell culture showed a lengthening of thin filaments following RNAi knockdown of nebulin (39). In contrast, experiments in mice show that following deletion of nebulin, thin filaments are shorter, which suggests that nebulin stabilizes thin filaments, which is needed for the assembly of long actin filaments (11, 12). Recent studies with genetically engineered mini-nebulin constructs in muscle fibers in culture also support the idea that nebulin stabilizes thin filaments and thereby affects their lengths (40). Thus, nebulin is emerging as a thin-filament length stabilizer that, in skeletal muscle, ensures the assembly of much longer actin filaments than are found in other cell types where nebulin is absent (such as red blood cells, where actin filaments are only 20–50 nm). It is possible that skeletal muscle has a nebulin-independent mechanism that controls actin filaments that have grown beyond the length of nebulin, involving perhaps other members of the Tmod family or other capping proteins. For example, recently leiomodin has been localized at pointed ends of thin filaments (41). It is also important to point out that in cardiac muscle, where full-length nebulin is absent (or present at very low levels only; ref. 11), thin-filament length varies less (42) than in the nebulin-deficient skeletal muscle (11), which suggests that cardiac muscle has nebulin-independent mechanisms for thin-filament length regulation. In summary, recent data suggest that in skeletal muscle, nebulin specifies the minimal thin-filament length and that a nebulin-independent mechanism extends the thin filament beyond the thin-filament length. Clearly, additional work is needed to more firmly establish the role of nebulin in thin-filament length specification.

Nebulin as regulator of Z-disk structure?

To understand the layout of the C-terminal region of nebulin in the Z disk, immunoelectron microscopy (IEM) has been used on human soleus muscle; the nebulin C terminus has been labeled with the nebulin-specific anti-SH3 antibody and the more N-terminal M177–181 domains with anti-neb177–181 (19). Results showed that the nebulin SH3 domain is located ∼25 nm inside the Z disk and that the repeats M176 to M181 are located near the edge of the Z disk. These results are consistent with two distinct models of the layout of nebulin in the Z disk (for details, see ref, 19). In the first model, nebulin penetrates the Z disk by only ∼25 nm, and nebulin filaments from neighboring sarcomeres do not overlap (no-overlap model). In the second model, nebulin penetrates the Z disk by ∼100 nm, and this creates a ∼75-nm-wide nebulin overlap zone in the center of the Z disk (overlap model). It has been argued that the no-overlap model is more likely to be correct (19), but definite experimental evidence for either model is lacking. In a more recently proposed model, nebulin fully penetrates the Z disk and cross-links thin filaments from adjacent sarcomeres (43). A drawback of this model is that it predicts that the SH3 epitope is further from the center of the Z disk than the M177–181 epitope, which is opposite of what has been measured. Clearly, further work is needed to map the layout of nebulin in the Z disk.

Sequence comparisons of rabbit nebulin cDNA clones have demonstrated that the Z-disk region of nebulin is expressed differentially (19). Several different isoforms have been identified that result from the skipping of various combinations of 7 Z-disk domains (19), which leads to the suggestion that nebulin is one of several proteins that are important for Z-disk width regulation (19). Consistent with this notion, the ultrastructural characterization of the Z disk in NEB-KO mice revealed that a subset of sarcomeres have extremely wide Z disks (11). In addition, inclusion-like bodies of misassembled Z disks are detected that resemble the rod bodies that are a hallmark of NM. Thus, the work on the NEB-KO mice suggests that nebulin plays an important role in Z-disk assembly and maintenance. Further support was obtained by a recent study on nebulin in a range of muscle types during postnatal development of the mouse in which transcript studies were performed with a mouse nebulin exon microarray (8). During postnatal development of the soleus muscle, major changes in splicing were detected in the Z-disk region of nebulin. Three differentially spliced Z-disk exons were up-regulated during postnatal development of soleus muscle, and this correlated with a significant increase in Z-disk width. The increase in Z-disk width of the soleus muscle might reflect the increasing stress exerted on this muscle due to the rapid increase in body weight during postnatal development. These findings support the view that nebulin plays an important role in Z-disk width regulation. The mechanism by which nebulin might regulate Z-disk structure is unknown. In vitro studies have identified interactions of the 25-kDa C-terminal region of nebulin with CapZ, α-actinin, and a unique sequence of titin that, in turn, interacts with filamin (11). This C-terminal end portion is composed of an SH3 domain similar to members of the Lasp gene family implicated in the regulation of cytoskeletal actin assemblies (44, 45), and of a serine-rich domain with putative phosphorylation sites. Taken together, the nebulin C-terminus is likely to play a role in Z-disk assembly, but the mechanisms involved and the relationship to conserved roles shared with Lasp family members in stress-fiber assemblies need to be established. Future studies will need to clarify how nebulin participates in the regulation of Z-disk assembly and how its absence promotes the formation of nemaline myopathy rod-like bodies.

MEDICAL GENETICS OF NEBULIN AND THERAPEUTIC AVENUES FOR NM

Dysfunctions of the skeletal muscle thin filament have emerged as an important cause of skeletal myopathies. NM is the most common of these diseases, with an estimated prevalence of 0.002% (46, 47). Genetically, NM is heterogeneous, and to date 6 genes have been identified as causing NM: α-tropomyosin-3 and β-tropomyosin (TPM3 and TPM2), NEB, actin α 1 (ACTA1), troponin T type 1 (TNNT1), and cofilin-2 (CFL2) (48–54). An unbiased linkage analysis of 45 NM families from 10 different countries implicated nebulin mutations as the cause of disease in 41 of these families (53). The typical form of NM presents with early onset in infancy, has a nonprogressive or slowly progressive course, and is caused by mutations in the nebulin gene (52, 53). These mutations are often missense mutations, but in addition, a 2502-bp deletion causing a 33-residue in-frame deletion of exon 55 was identified as a prominent recessive disease-causing allele, initially in the Ashkenazi Jewish population, and more recently also in non-Jewish patients (55–57). Finally, the nebulin gene has also been identified as a genetic cause of core-rod myopathy (58) and distal myopathies (50).

Recent studies on muscle fibers from patients with NM due to the deletion of nebulin exon 55 showed that these patients have severely reduced nebulin protein levels and that they show remarkable phenotypic similarities to fibers from NEB-KO mice, i.e., shorter and nonuniform thin-filament lengths and significantly impaired force-generating capacity (27). Thus, loss of thin-filament length regulation appears to be an important contributor to muscle weakness in patients with NM. In addition to altered thin-filament length, changes in crossbridge cycling kinetics and reduced calcium sensitivity of force production contribute to the muscle weakness observed in these patients (28), in line with the role of nebulin in these processes (21). Thus, the role of nebulin in thin-filament length regulation and contraction deduced from work on NEB-KO mice provides a mechanism to explain for the first time severe muscle weakness in patients with NM. These functional insights might prove very useful for developing treatment strategies. For example, analogous to current clinical studies that address whether contractility during heart failure can be augmented by actomyosin-activating small molecules (for review, see ref. 59), a rationale for similar approaches in NM now exists.

The highly variable disease presentation as well as the large coding size and exon number of the nebulin gene presents a significant challenge for diagnosis and treatment. The application of next-generation sequencing technology will allow a faster determination of recessive missense mutations, compound heterozygote genotypes, deletions, stop codons, or frame shifts that are the causative defects in the NEB gene, and these insights should provide a better basis for treatment options in the future. For example, aminoglyosides with stop-codon-suppressing activity are emerging as potential drugs for the treatment of neuromuscular diseases that are caused by termination codons (for PCT124 for stop-codon override, see ref. 60), and these approaches might be applicable for NM treatment. In addition, frame shifts or stop codons might be removable by splice correction therapies in the more distant future (61). Testing of such experimental strategies will benefit from the recently obtained insights into the roles of nebulin in muscle function and the availability of murine models that mimic NM. Thus, the recent rapid progress that has made the roles of nebulin much less nebulous now provides a rationale for therapeutic strategies for nebulin-based NM.