Paul T. Martin

Long-term enhancement of skeletal muscle mass and strength by single gene administration of myostatin inhibitors

Increasing the size and strength of muscles represents a promising therapeutic strategy for musculoskeletal disorders, and interest has focused on myostatin, a negative regulator of muscle growth. Various myostatin inhibitor approaches have been identified and tested in models of muscle disease with varying efficacies, depending on the age at which myostatin inhibition occurs. Here, we describe a one-time gene administration of myostatin-inhibitor-proteins to enhance muscle mass and strength in normal and dystrophic mouse models for >2 years, even when delivered in aged animals. These results demonstrate a promising therapeutic strategy that warrants consideration for clinical trials in human muscle diseases.

Overexpression of the cytotoxic t cell GalNAc transferase in skeletal muscle inhibits muscular dystrophy in mdx mice

Duchenne muscular dystrophy (DMD) is a congenital X-linked myopathy caused by lack of dystrophin protein expression. In DMD, the expression of many dystrophin-associated proteins (DAPs) is reduced along the sarcolemmal membrane, but the same proteins remain concentrated at the neuromuscular junction where utrophin, a dystrophin homologue, is expressed [Matsumura, K., Ervasti, J. M., Ohlendieck, K., Kahl, K. D. & Campbell, K. (1992) Nature (London) 360, 588–591]. This outcome has led to the concept that ectopic expression of a “synaptic scaffold” of DAPs and utrophin along myofibers might compensate for the molecular defects in DMD. Here we show that transgenic overexpression of the synaptic CT GalNAc transferase in the skeletal muscles of mdx animals (mdx/CT) increases the expression of utrophin and many DAPs, including dystroglycans, sarcoglycans, and dystrobrevins, along myofibers. Protein expression of utrophin and DAPs was equal to or above that of wild-type mice. In addition, α-dystroglycan was glycosylated with the CT carbohydrate antigen in mdx/CT but not in mdx muscles. mdx/CT mice have little or no evidence of muscular dystrophy by several standard measures; Serum creatine kinase levels, percentage of centrally located myofiber nuclei, and variance in myofiber diameter in mdx/CT muscles were dramatically reduced compared with mdx mice. These data suggest that ectopic expression of the CT GalNAc transferase creates a functional dystrophin-related complex along myofibers in the absence of dystrophin and should be considered as a target for therapeutic intervention in DMD.

Glycobiology of the neuromuscular junction

Most molecules that are present at synapses are glycosylated with carbohydrates, and some carbohydrate structures are themselves uniquely synaptic in their localization. Thus, proteins or lipids at the synapse may bear distinct carbohydrates that alter their localization or function. Here, I will review the evidence that there are unique synaptic carbohydrates at the neuromuscular junction. Then, I will review the evidence that such carbohydrates can affect the function of synaptic proteins, with particular attention to agrin, dystroglycan, and the neural cell adhesion molecule (NCAM). Finally, I will review recent data that demonstrates a role for one carbohydrate structure, the cytotoxic T cell (CT) antigen, in neuromuscular development. These studies suggest that glycosylation is an important modification to consider in studies of synapse formation and function.

Distinct structures and functions of related pre- and postsynaptic carbohydrates at the mammalian neuromuscular junction.

Carbohydrates that terminate in beta-linked N-acetylgalactosamine (betaGalNAc) residues are concentrated in the postsynaptic apparatus of the skeletal neuromuscular junction and have been implicated in the differentiation of the postsynaptic membrane. We now report that distinct synapse-specific betaGalNAc-containing carbohydrates are associated with motor nerve terminals. Two monoclonal antibodies that recognize distinct betaGalNAc-containing epitopes, CT1 and CT2, both stain synaptic sites on skeletal muscle fibers. However, CT1 selectively stains nerve terminal, whereas CT2 selectively stains the postsynaptic apparatus. Likewise, CT1 and CT2 selectively stain motoneuron-like and muscle cell lines, respectively. Using the cell lines, we identify distinct CT1- and CT2-reactive glycolipids and glycoproteins. Finally, we show that GalNAc modulates the adhesion of motoneuron-like cells to recombinant fragments of a synaptic cleft component, laminin beta2. Together, these results show that pre- as well as postsynaptic membranes bear and are affected by distinct but related synapse-specific carbohydrates.

Inhibition of dystroglycan cleavage causes muscular dystrophy in transgenic mice.

Dystroglycan (DG) is an essential component of the dystrophin-glycoprotein complex, a molecular scaffold that links the extracellular matrix to the actin cytoskeleton. Dystroglycan protein is post-translationally cleaved into alpha dystroglycan, a highly glycosylated peripheral membrane protein, and beta dystroglycan, a transmembrane protein. Despite clear evidence of the importance of dystroglycan and its associated proteins in muscular dystrophy, the purpose of dystroglycan proteolysis is unclear. By introducing a point mutation at the normal site of proteolysis (serine 654 to alanine, DGS654A), we have created a dystroglycan protein that is severely inhibited in its cleavage. Transgenic expression of DGS654A in mouse skeletal muscles inhibited the expression of endogenously cleaved dystroglycan, while overexpression of wild type dystroglycan by similar amounts did not. DGS654A animals had increased serum creatine kinase activity and most muscles had increased numbers of central nuclei. Overexpression of wild type dystroglycan, by contrast, caused no dystrophy by these measures. Dystrophy in DGS654A muscles correlated with reduced binding of antibodies that recognize glycosylated forms of alpha dystroglycan. Lastly, neuromuscular junctions in DGS654A muscles were aberrant in structure. These data show that aberrant processing of the dystroglycan polypeptide causes muscular dystrophy and suggest that dystroglycan processing is important for the proper glycosylation of alpha dystroglycan.

Muscular dystrophy with truncated dystrophin in a family of Japanese Spitz dogs

Muscular dystrophy was diagnosed in seven male Japanese Spitz dogs with clinical signs of slowly progressive exercise intolerance, generalized weakness, myalgia, difficulty chewing and dysphagia. Serum creatine kinase (CK) concentrations were markedly elevated. Histopathology showed degeneration and regeneration of muscle, consistent with a dystrophic phenotype. Immunohistochemical staining for dystrophin and related proteins showed no staining with a monoclonal antibody against the rod domain of dystrophin but near-normal staining with an antibody against the C terminus. Immunoblot analysis in two affected dogs showed a truncated dystrophin protein of approximately 70-80 kDa. The severity of disease showed that this fragment was not large enough to protect from the dystrophic process.

N-acetyllactosamine and the CT carbohydrate antigen mediate agrin-dependent activation of MuSK and acetylcholine receptor clustering in skeletal muscle.

Galbeta1,3GalNAc and Galbeta1,4GIcNAc are the subterminal saccharide structures present on the CT carbohydrate antigen GalNAcbeta1,4[NeuAcalpha2,3]-Galbeta1-(3GalNAc or 4GIcNAc)-R, which is localized at the mammalian neuromuscular junction. Here we show that Galbeta1,3GalNAc, Galbeta1,4GIcNAc, and the CT carbohydrate antigen affect postsynaptic assembly in cultured muscle cells. Treatment of C2C12 myotubes with benzyl-O-alpha-GalNAc or neuraminidase increased peanut agglutinin (PNA) expression and AChR clustering. Induction of AChR clustering was blocked by PNA and by muscle agrin. Addition of Galbeta1,4GIcNAc or Galbeta1,3GalNAc increased AChR clustering in myotubes and muscle-specific kinase (MUSK) autophosphorylation in vitro, while NeuAcalpha2,3Galbeta1,4GIcNAc and Galbeta1,4GIc did not. Neural agrin activated MuSK in vitro if the lactosamine-containing mucin domain was present, and this activation was blocked in large part by Galbeta1,3GalNAc and Galbeta1,4GIcNAc. Agrin fragments and MuSK bound to these disaccharides with differing specificities. Overexpression of the CT carbohydrate antigen also increased AChR clustering and MuSK autophosphorylation in the presence of neural agrin. These data suggest a model in which different portions of the CT carbohydrate structure contribute to agrin-dependent signal transduction.

Definition of pre- and postsynaptic forms of the CT carbohydrate antigen at the neuromuscular junction: ubiquitous expression of the CT antigens and the CT GalNAc transferase in mouse tissues

At the rodent neuromuscular junction, the synaptic expression of the CT carbohydrate antigens is defined by the binding of two monoclonal antibodies, CT1 and CT2. CT1 preferentially stains the presynaptic membrane, while CT2 preferentially stains the postsynaptic apparatus. Here we show that the differential subsynaptic distribution of these antigens is due to a preference of CT1 for structures containing N-acetyl neuraminic acid (NeuAc) and a preference of CT2 for structures containing N-glycolyl neuraminic acid (NeuGc). This was found to be the case both in binding to cultured myotubes, where NeuAc/NeuGc levels were manipulated by feeding acetylated N-acetyl mannosamine precursors, and in binding to purified GM2 ganglioside containing either NeuAc or NeuGc. At human neuromuscular junctions, where the enzymatic machinery to make NeuGc is absent [Proc. Natl. Acac. Sci. USA 95 (1998) 11751], CT1 and GM2(NeuAc) antibodies stained, while CT2 did not. Thus, the N-glycolyl modification of sialic acid helps to define the differential distribution of the CT antigens at the rodent neuromuscular junction, and this difference is lost in humans. In addition, sulfatase and 9-O-acetylesterase treatment of cells or tissues increased the amount of CT1 and CT2 antibody binding, with sulfatase differentially unmasking CT antigen expression on particular glycoproteins. Despite its uniquely synaptic localization in skeletal muscle, the CT antigens and the CT GalNAc transferase are ubiquitously expressed in other mouse tissues, including brain, spinal cord, and peripheral nerve. One of the proteins that can be co-purified with a CT-reactive glycoprotein is alpha dystroglycan. These data better define the sub-synaptic structures of the CT carbohydrate antigens at the neuromuscular junction and demonstrate their ubiquitous presence in mouse tissues, including the brain.

Modulation of agrin binding and activity by the CT and related carbohydrate antigens.

Agrin is a nerve-derived signal that is essential for the proper organization of postsynaptic acetylcholine receptors (AChRs) at the vertebrate neuromuscular junction. It is likely that carbohydrates play a significant role in regulating agrin activity, as agrin binds multiple glycan structures and is itself a highly glycosylated protein. Here we provide support for this contention by showing that agrin can be modified with the CT antigen, a carbohydrate structure expressed at the neuromuscular junction, and by describing the resulting changes in agrin binding to neoglycoconjugates and cultured myotubes, as well as changes in agrin-dependent AChR clustering. Glycosylation of agrin with the CT antigen required the mucin domain and the dystroglycan/heparin-binding domain. The presence of the mucin domain lowered agrin binding to several N-acetyllactosaminyl-containing saccharides and C2 myotubes and lowered agrin activity in AChR clustering. Glycosylation of agrin with the CT antigen, by contrast, increased agrin binding to myotubes and potentiated its AChR clustering activity at subsaturating concentrations. Last, sialylated and nonsialylated variants of N-acetyllactosamine differentially modulated AChR clustering and agrin activity, and these changes correlated with the ability of MuSK, an agrin-stimulated kinase, to bind to these structures. These experiments demonstrate that the glycosylation state of agrin affects its activity and suggest a role for the CT antigen in modulating agrin function.

Transgenic Overexpression of Dystroglycan Does Not Inhibit Muscular Dystrophy in mdx Mice

Recently, there have been a number of studies demonstrating that overexpression of molecules in skeletal muscle can inhibit or ameliorate aspects of muscular dystrophy in the mdx mouse, a model for Duchenne muscular dystrophy. Several such studies involve molecules that increase the expression of dystroglycan, an important component of the dystrophin-glycoprotein complex. To test whether dystroglycan itself inhibits muscular dystrophy in mdx mice, we created dystroglycan transgenic mdx mice (DG/mdx). The alpha and beta chains of dystroglycan were highly overexpressed along the sarcolemmal membrane in most DG/mdx muscles. Increased dystroglycan expression, however, did not correlate with increased expression of utrophin or sarcoglycans, but rather caused their decreased expression. In addition, the percentage of centrally located myofiber nuclei and the level of serum creatine kinase activity were not decreased in DG/mdx mice relative to mdx animals. Therefore, dystroglycan overexpression does not cause the concomitant overexpression of a utrophin-glycoprotein complex in mdx muscles and has no effect on the development of muscle pathology associated with muscular dystrophy.