Second International Workshop for Glycosylation Defects in Muscular Dystrophies, 11–12 November, 2010, Charlotte, USA

Abstract

The second International Workshop for Glycosylation Defects in Muscular Dystrophies took place on November 11 and 12, 2010 in Charlotte, North Carolina, USA. The meeting was hosted by the Carolinas Medical Center with financial support from the Carolinas Muscular Dystrophy Research Endowment at the Carolinas HealthCare Foundation, the Muscular Dystrophy Association and funds raised by the “Jeans, Genes & Geniuses” event organized by Jane and Luther Lockwood. Since conducting the first workshop in May 2008, significant progress has been made in a subset of muscular dystrophies associated with defects in alpha‐dystroglycan (α‐DG) glycosylation. New findings on α‐DG glycosylation and creation of novel animal models have expanded our understanding of the disease mechanism. The 2010 workshop focused on the following topics; (i) functional glycosylation of α‐DG; (ii) animal models; and (iii) novel experimental therapies. The workshop brought together a total of 22 internationally renowned scientists and clinicians from US, UK, Denmark and Japan with active research and expertise in these areas. Overall, the workshop provided a unique opportunity to discuss the significance of recent progress, facilitate international collaboration, and identify new approaches to treat the disease.

INTRODUCTION

The second International Workshop for Glycosylation Defects in Muscular Dystrophies (IWGDMD) took place on November 11 and 12, 2010 in Charlotte, North Carolina, USA. The meeting was hosted by the Carolinas Medical Center (CMC) with financial support from the Carolinas Muscular Dystrophy Research Endowment at the Carolinas HealthCare Foundation, the Muscular Dystrophy Association (MDA) and funds raised by the “Jeans, Genes & Geniuses” event organized by Jane and Luther Lockwood. Since conducting the first workshop in May 2008, significant progress has been made in many areas of muscular dystrophy, in particularly those associated with defects in alpha‐dystroglycan (α‐DG) glycosylation. New findings on α‐DG glycosylation have expanded our understanding of the disease mechanism. Several novel animal models have also been created which should accelerate the development of future therapies for this group of disorders. Building upon the success of the last meeting, the 2010 workshop focused on the following topics: (i) functional glycosylation of α‐DG; (ii) animal models; and (iii) novel experimental therapies. The workshop brought together a total of 22 internationally renowned scientists and clinicians from US, UK, Denmark and Japan with active research and expertise in these areas. In addition, the conference was joined by participants from Carolinas Medical Center, University of North Carolina at Charlotte, University of North Carolina at Chapel Hill, Emory University, Genzyme Corporation, Children's National Medical Center D.C., MDA and CURE CMD. Overall, the workshop provided a unique opportunity to discuss the significance of recent progress, facilitate international collaboration and identify new approaches to treat the disease.

BACKGROUND

α‐DG is a critical component of the multimeric dystrophin‐associated glycoprotein complex (DGC) in muscle and also maintains essential cellular functions in other tissues, such as brain and peripheral nerve (5). A subset of muscular dystrophies is associated with defective α‐DG glycosylation, and commonly referred to as dystroglycanopathies (29). The aberrantly modified α‐DG is defective in binding to extracellular matrix ligands including laminin, agrin and perlecan 13, 28. However, the precise mechanism leading to muscle weakness and cell death is not fully understood. Dystroglycanopathies often show a wide spectrum of clinical variations, ranging from mild limb‐girdle muscular dystrophy type 2I (LGMD2I) to severe muscle‐eye‐brain disease (MEB) and Walker‐Warburg syndrome (WWS) 9, 27. To date, mutations in at least seven genes have been identified to cause these diseases; they are fukutin‐related protein (FKRP), fukutin, POMT1, POMT2, POMGnT1, LARGE and DPM3 7, 8, 20, 21, 24, 36, 41. Consistent with the observed defect in α‐DG glycosylation, all these genes encode for either putative or known glycosyltransferases that facilitate the O‐mannosyl‐linked glycan addition on α‐DG (25).

REPORT SUMMARY

Functional glycosylation of α‐DG

Herbert Bonkovsky, Vice President of Research at CMC, delivered the opening remarks for the conference and welcomed the delegates to Charlotte and Carolinas Medical Center. The first session, chaired by Kevin Campbell, provided an overview of α‐DG glycosylation and how defects in this process play a critical role in the pathogenesis of the dystroglycanopathies. Speakers described a variety of biochemical and cellular approaches to studying α‐DG function.

Pamela Stanley started the workshop with a detailed summary of our current understanding of α‐DG glycosylation. In particular, Dr Stanley focused on the modification of α‐DG by the putative glycosyltransferase LARGE. Using a battery of glycosylation‐defective CHO cell mutants and mutational analysis of LARGE, her research has shown that complex N‐glycans, mucin O‐glycans and O‐mannose glycans on α‐DG are all substrates of LARGE. The combined data suggested that LARGE modifies α‐DG at an N‐acetylhexosamine residue to initiate glycan polymer synthesis 2, 31. With a similar panel of CHO cell mutants, Xiaohua Wu demonstrated that LARGE induces functional glycosylation of α‐DG in a DPM2/O‐mannosylation‐dependent manner (16). In addition, LARGE appeared to compete with galactosyltransferases to target GlcNAc terminals for α‐DG functional modifications.

Minoru Fukuda further demonstrated in the prostate carcinoma PC3 and breast carcinoma T47D cells that the function of α‐DG also depends on the activity of the glycosyltransferase β3GnT1 through forming a complex with LARGE (4). He provided evidence that the laminin‐binding glycans on α‐DG function as a tumor suppressor by antagonizing the ERK/MAP kinase cascade induced by integrins. During discussion, many researchers gave support to the current thinking of up‐regulating LARGE as potential therapy for dystroglycanopathies. However, several speakers cautioned that there could be other substrates for LARGE despite the fact that α‐DG is clearly the major player in the pathogenesis.

In the second half of the session, Takako Moriguchi presented her recent findings on a novel type of α‐DG modification. Using nuclear magnetic resonance (NMR) and mass spectrometry, Dr Moriguchi identified a phosphorylated O‐glycan linked to the threonine residue in the mucin‐like domain of α‐DG (42). Additional studies revealed that this phosphorylated glycan is required for the functional interaction of α‐DG with laminin. These results indicated that there could be additional genes involved in the regulation of α‐DG glycosylation.

Tamao Endo concluded the morning session with an extensive review of the O‐mannosylation pathway and his zebrafish model. Given that the O‐mannosyltransferases POMT1 and POMT2 are highly conserved in zebrafish, the organism is well suited for functional studies of protein O‐mannosylation. According to the data, POMT knockdown morphants displayed loss of α‐DG glycosylation and severe muscle abnormalities at the early stage of development (3). Because of the ease of breeding and observing phenotypes in zebrafish, Dr Endo reasoned that the model could be useful to screen for chemicals to up‐regulate α‐DG glycosylation.

Animal models

The afternoon session was chaired by Derek Blake and centered on the use of different mouse models to investigate dystroglycanopathies. Kevin Campbell described several unique models of conditional knockout mice that exhibited cell type‐specific patterns of dystroglycan deletion in the brain (33). His findings provided genetic evidence to support the distinctive roles of glial and neuronal dystroglycan in forebrain development and synaptic plasticity, respectively.

To investigate the role of O‐mannosylation in brain development, Huaiyu Hu's laboratory generated different models of POMT2‐null mice. Dr Hu's group has previously reported a POMGnT1‐null mouse model with developmental defects in muscle, eye and brain (23). His preliminary studies showed that mice with a targeted disruption of POMT2 were embryonic lethal. Meanwhile, the brain‐specific deletion of POMT2 resulted in the hypoglycosylation of α‐DG. The mutant mice developed different degrees of severity in brain abnormalities, depending on the timing of deletion during embryonic development. On the other hand, meninges‐specific deletion of POMT2 did not affect brain development in the animals. One of their significant findings was that disruption of basement membrane appeared to precede over‐migration of neocortical neurons.

Tatsushi Toda discussed his latest findings on the fukutin‐deficient Hp/– knockin mouse model. The mutant mice were shown to exhibit no obvious pathological phenotypes. When compared with the more severe LARGE^myd mice, Dr Toda observed that the difference in severity is generally correlated with the levels of α‐DG glycosylation. These data further implied that a small amount of glycosylated α‐DG is sufficient to maintain normal muscle function in the Hp/– knockin mice (17). Dr Toda then continued with a discussion of his recent work on pikachurin, a newly identified ligand of α‐DG in the retina. In common with other known ligands of α‐DG, pikachurin binding also requires LARGE‐dependent modification of α‐DG (18).

Susan Brown used her previously developed FKRP‐Neo^Y307N knockdown mice (1) to further investigate the role of FKRP during development of the brain and eyes. Reduced expression of FKRP and the resulting hypoglycosylation of α‐DG were shown to have influenced the expression of different laminin isoforms, and also disrupted their pattern of deposition in the cortex of this mouse. These observations suggest profound alterations in the organization of the matrix which may have implications for future therapeutic intervention in these disorders. Because the FKRP‐Neo^Y307N mice die around the time of birth caused by central nervous system (CNS) involvement, a strategy was presented for restoring FKRP expression in the CNS, thereby improving survival and allowing the muscle pathology associated with a reduction in FKRP to be evaluated.

Qi Lu took a different approach with regard to developing a FKRP mouse model. By generating different missense and nonsense mutations in the mouse fkrp gene, Lu's laboratory created a series of FKRP mutant strains with variable phenotypes that recapitulate the wide clinical spectrums of dystroglycanopathy patients (10). One of the strains, FKRP‐neo‐P448L knockin, was the first reported case of a viable FKRP mouse model. Moreover, the severity of the diseases in different strains is correlated to the levels of functional glycosylation of α‐DG. Dr Lu stated that these mice will be valuable for testing experimental therapies and candidate drugs.

Clinical trial readiness

Anne Rutkowski of CURE CMD spoke briefly on several important initiatives to raise awareness and support clinical trial readiness for Congenital Muscular Dystrophies (CMD). These initiatives include CMD International Registry, CMD BioBank, natural history studies and the identification of relevant biomarkers. Through a combination of workshop, partnership and grant awards, CURE CMD aims to bridge the gap between preclinical development and clinical trials in order to identify treatments for the CMDs.

Paul Muhlrad, of the Muscular Dystrophy Association, described the important role of MDA in advancing muscular dystrophy research and highlighted research projects of current MDA grantees studying dystroglycanopathies, many of whom were present at this workshop. Dr Muhlrad also announced the inauguration of the biennial MDA National Scientific Conferences in March 2011. The first‐day meeting was concluded in the evening with welcome addresses by James McDeavitt, Senior Vice President at CMC, and Jane B. McColl, whose diagnosis has inspired the founding of the McColl‐Lockwood Laboratory for Muscular Dystrophy Research in CMC.

Novel experimental therapies

The second day's session was chaired by Benjamin Brooks of Neuromuscular/Amyotrophic Lateral Sclerosis‐Muscular Dystrophy Association (ALS MDA) Center in CMC and devoted to development of strategies for translating basic research into useful treatments. Speakers reviewed the latest progress in treatments for dystroglycanopathies. Some of the topics covered in the session were the up‐regulation of glycosyltransferases, protein therapeutics, protein folding, anti‐fibrosis, exercise training, small molecule screening and gene replacement therapy.

The morning session began with a presentation by Susan Sparks on the challenges in molecular characterization of LGMD patients. Among the major obstacles to proper diagnosis of LGMD is the heterogeneity of the diseases and lack of standardized protocols. Many research centers are actively working to standardize clinical evaluations. One such center is the Cooperative International Neuromuscular Research Group (CINRG) (11). There was also a general feeling that the identification of appropriate serum markers will be essential for proper diagnosis as well as monitoring disease progression in future clinical trials. In her final remarks, Dr Sparks emphasized the importance of cardiac involvement which should not be overlooked in patients, especially those associated with FKRP mutations (37).

The possibility of up‐regulating glycosyltransferases as a potential therapy was given careful consideration at the conference. LARGE has been the focus of discussion, caused by its ability to restore α‐DG glycosylation in cell lines derived from patients with different forms of dystroglycanopathies as initially shown by Campbell's laboratory (6). In this workshop, Dr Toda presented in vivo data to support adenovirus‐mediated transfer of either fukutin or LARGE, which is able to restore α‐DG function in the fukutin‐deficient Hp/– mice and the POMGnT1‐deficient mice (17).

On the other hand, Paul Martin proposed a different target for up‐regulation, namely cytotoxic T cell GalNAc transferase (Galgt2), a glycosyltransferase that can also modify α‐DG. Previously, his group reported that the transgenic overexpression of Galgt2 inhibited the development of muscle pathology in mdx, laminin‐deficient CMD and sarcoglycan‐deficient LGMD mouse models (39). Using the gene therapy approach, Dr Martin demonstrated that adeno‐associated virus (AAV)‐mediated transfer of either Galgt2 or human micro‐dystrophin can achieve similar therapeutic effects in mdx mice (26). The long‐term therapeutic benefits of micro‐dystrophin, however, could be undermined by potential T‐cell responses to micro‐dystrophin, a foreign protein in Duchenne muscular dystrophy (DMD) patients. Here, human GALGT2 may have an advantage, as all DMD patients express this gene. Interestingly, gene transfer of Galgt2 failed to correct muscle defects in the dystroglycan‐deficient mice, implying that the therapeutic function of Galgt2 is limited to its effect on the glycosylation of α‐DG.

A number of researchers are exploring the potential application of protein therapeutics. Dr Campbell's recent data revealed that when recombinant α‐DG was locally injected into LARGE^myd muscle, the protein was incorporated into the muscle fiber and consequently restored membrane integrity (14). However, the expression of glycosylated α‐DG was only maintained for short period of time, thus raising concern regarding sustainability of the treatment. Meanwhile, James Ervasti is investigating the therapeutic potential of a recombinant micro‐utrophin modified with the cell‐penetrating TAT protein transduction domain. His recent work showed that the modified protein can effectively integrate into the DGC complex and improve the phenotypes of mdx mice via short‐term systemic administration (34). Dr Ervasti then discussed the physical properties of dystrophin and utrophin proteins. Surprisingly, his analysis suggested that many missense mutations in the actin‐binding regions of dystrophin had a more profound effect in thermal denaturation and aggregation than actin‐binding (15). As a result, it was argued that micro‐utrophin might be a better candidate than micro‐dystrophin in gene/protein replacement therapy because of the higher thermal stability, thus less chance for the micro‐utrophin to misfold and trigger immune response.

In a similar manner, Derek Blake evaluated the protein folding and stability properties of FKRP and their roles in LGMD2I. Dr Blake found that many FKRP mutations caused the protein to misfold and form high molecular weight aggregates in a heterologous cell expression system. Interestingly, this defect was chemically reversible and can be corrected by reducing agent. Using mass spectrometry and RNAi‐mediated knockdown strategy, his preliminary studies identified several chaperones that may participate in the folding and trafficking of FKRP. Both Dr Blake and Dr Ervasti agreed that strategy or drug aimed to stabilize protein folding might offer clinical benefits and improve the outcomes of any gene therapy approach.

Kathryn Wagner focused on treating fibrosis in dystroglycanopathy. Fibrosis and inflammation are common cellular responses observed in many forms of muscular dystrophies. Dr Wagner previously showed that myostatin, a negative regulator of muscle growth has pro‐fibrotic property in mdx mice (22). When senescent mdx mice were treated with soluble activin type IIB receptor, the putative receptor of myostatin, her preliminary studies indicated that this treatment can also reverse pre‐existing muscle fibrosis and promote muscle growth and regeneration. Various new myostatin inhibitors currently in development or clinical trials—for example, neutralizing antibodies, peptibody and activin type IIB receptor—were discussed.

In the second half of the session, two researchers outlined the potential benefits of exercise training. Previous studies directed by John Vissing provided evidence that aerobic training may benefit LGMD2I patients (35). He is currently studying the effect of strength training in LGMD2 and becker muscular dystrophy (BMD) patients. Preliminary studies suggested that strength training over a period of 12 weeks is safe but appears less effective than aerobic training. In collaborations with other researchers in UK and France, Dr Vissing is also conducting an ongoing 2‐year magnetic resonance imaging‐based natural history study of LGMD2I patients to non‐invasively quantify muscle pathology. The goal is to develop a useful tool for monitoring response to treatment in the future. In a similar pilot study, Mohammed Sanjak concluded that supported treadmill ambulation training (STAT) is safe and well tolerated in amyotrophic lateral sclerosis (ALS) patients (32). The training was also found to improve their work capacity and gait function. Dr Sanjak proposed that this type of exercise training may be applicable to LGMD2I patients who share a number of clinical characteristics with ALS patients.

Louis Kunkel presented his latest research on two zebrafish models of muscular dystrophy and their application in drug target screening. The FKRP‐deficiency fish was generated by knockdown approach using morpholino (19). The sapje fish was the result of a genetic mutation in the zebrafish dystrophin gene. Both zebrafish models provide a convenient way to analyze the phenotypes because their skeletal muscles exhibit a very characteristic structural abnormality readily detectable by birefringence of polarized light under the microscope. Using this assay, Dr Kunkel has identified seven candidate compounds that can prevent or reverse the abnormal muscle pathology in the sapje fish without restoring dystrophin expression. A similar strategy is now being applied to the FKRP‐deficient zebrafish.

The last two presentations were related to AAV‐mediated gene replacement therapy. Xiao Xiao first provided a brief summary of several promising gene therapy experiments in other muscular dystrophy animal models, such as sarcoglycan‐deficient Bio14.6 hamsters and laminin‐deficient dy^w mice. Given the main concern over the specificity of vector in gene therapy, Dr Xiao devised different schemes to modify AAV vectors aimed to reduce non‐specific delivery to liver and improve target delivery to striated muscle (40). His preliminary studies showed widespread FKRP gene transfer in muscles of FKRP mutant mice after systemic delivery by intraperitoneal injection. Current efforts are underway to evaluate functional improvement in the treated animals. In addition, Dr Xiao has recently created a FKRP knockdown mouse model using RNA interference technology (38). FKRP expression was suppressed for longer than 10 months post‐injection and the mice displayed late‐onset phenotypes resembling the mild LGMD2I.

Jeffrey Chamberlain then reviewed other major obstacles facing gene replacement therapy. One of the challenges was the limited carrying capacity of the AAV vector. Over the years, Dr Chamberlain has developed different truncated versions of dystrophin gene that can be carried by AAV, yet retain sufficient functional capacity (30). Ongoing studies in the DMD dog model demonstrated efficient gene transfer to downstream muscles from the injected veins and arteries but not whole body gene transfer. T‐cell mediated immunological responses against the AAV capsid or the transgene also raised a major concern (12). However, it was possible to sustain long‐term dystrophin expression in AAV‐treated dogs when immune response was blocked by transient immune suppression. These preliminary results suggested that a combination of strategically designed vectors, delivered via intravascular AAV delivery coupled with transient immune suppression, could lead to an effective gene replacement therapy for numerous types of muscular dystrophy.

In concluding the workshop, Dr Lu reminded that much progress had been made since the last workshop in 2008. The number of animal models available in the dystroglycanopathy field is steadily increasing. From knockin/knockdown mice to zebrafish, these models will be extremely useful for dissecting the mechanism of disease and developing treatments. With the recent discovery of new types of glycan structures on α‐DG, we have increased our understanding of α‐DG glycosylation. As a result of this new knowledge, researchers are increasingly moving toward more innovative experimental therapies. Finally, Dr Lu emphasized the continued need for wider and closer collaborations among the laboratories to further our understanding of the disease mechanism and development of effective treatments for dystroglycanopathies.

THE WORKSHOP ORGANIZERS

Yiumo Michael Chan, PhD (Carolinas Medical Center, US), Kevin Campbell, PhD (University of Iowa, US), Qi Lu, MD, PhD (Carolinas Medical Center, US), Jeannie Maggio (Carolinas Medical Center, US), Derek Blake, PhD (Cardiff University, UK).

LIST OF SPEAKERS

• (i) Derek Blake, PhD, Cardiff University, UK
• (ii) Susan Brown, PhD, The Royal Veterinary College, UK
• (iii) Kevin Campbell, PhD, University of Iowa, USA
• (iv) Jeffrey Chamberlain, PhD, University of Washington, USA
• (v) Tamao Endo, PhD, Tokyo Metropolitan Institute of Gerontology, Japan
• (vi) Jim Ervasti, PhD, University of Minnesota, USA
• (vii) Minoru Fukuda, PhD, Burnham Institute, USA
• (viii) 

  Huaiyu Hu, PhD, SUNY Upstate, USA

• (ix) Louis Kunkel, PhD, Children's Hospital Boston, USA
• (x) Qi Lu, MD, PhD, Carolinas Medical Center, USA
• (xi) Paul Martin, PhD, Ohio State University, USA
• (xii) Takako Moriguchi, PhD, University of Iowa, USA
• (xiii) 

  Paul Muhlrad, PhD, Muscular Dystrophy Association, USA

• (xiv) Anne Rutkowski, MD, CURE CMD, USA
• (xv) Mohammed Sanjak, PhD, PT, Carolinas Medical Center, USA
• (xvi) Susan Sparks, MD, PhD, Carolinas Medical Center, USA
• (xvii) 

  Pamela Stanley, PhD, Albert Einstein College of Medicine, USA

• (xviii) 

  Tatsushi Toda, MD, PhD, Kobe University, Japan

• (xix) John Vissing, MD, DMSci, University of Copenhagen, Denmark
• (xx) Kathryn Wagner, MD, PhD, Johns Hopkins University, USA
• (xxi) Xiaohua Wu, MD, PhD, Carolinas Medical Center, USA
• (xxii) 

  Xiao Xiao, PhD, University of North Carolina, Chapel Hill, USA

WEBSITE

For a full description of the workshop and program, please visit http://www.carolinasmedicalcenter.org/LGMD

MCCOLL‐LOCKWOOD LABORATORY FOR MUSCULAR DYSTROPHY RESEARCH

The McColl‐Lockwood Laboratory was established in 2005 with a generous philanthropic research endowment jointly created by the McColl and Lockwood families and Carolinas HealthCare Foundation. The main focus of the laboratory is translational research for limb‐girdle muscular dystrophy type 2I and Duchenne muscular dystrophy.