Workshop on Therapeutic Approaches for Duchenne Muscular Dystrophy - Presentation Summaries

Workshop on Therapeutic Approaches for Duchenne Muscular Dystrophy

Presentation Summaries

May 15-16, 2000
Natcher Conference Center
National Institutes of Health, Bethesda, Maryland

Table of Contents

Session 1: Mutation Detection/Diagnostics
Session 2: Pathogenesis
Session 3A: Gene Therapy Approaches-Delivery
Session 3B: Gene Therapy Approaches-Repair and Modulation
Session 4: Cell Therapy
Session 5: Pharmacologic Therapy
Session 6: Resources/Future Directions

Session 1:Mutation Detection/Diagnostics

Introduction and Overview of Duchenne Muscular Dystrophy Diagnostics.
Thomas Prior, Ph.D. (Topic Leader)
Duchenne muscular dystrophy (DMD) diagnostics have greatly improved during the previous 15 years. Previously, the only test was for creatine kinase (CK) activity, which is not a good test for carriers. Now, with the cloning of the gene, diagnostics have greatly improved. However, issues remain. Issues for discussion include newborn screening; direct diagnostic protocols; indirect diagnostic protocols; role of dystrophin testing; new mutations; germ-line mosaicism; duplication testing; point-mutation sequencing; and protocols for manifesting carriers.

Newborn Creatine Kinase Screening for DMD
Guenter Scheuerbrandt, Ph.D.
A newborn screening program for using a quantitative bioluminescence assay of CK activity was set up in Germany 26 years ago. As of now, blood spots from about 465,500 boys, mostly 4 to 6 weeks old, have been tested. About 200 boys with high CK activities were found and the diagnosis of DMD or Becker muscular dystrophy was made. Early detection provides the family of a DMD infant to receive genetic counseling, and as far as is known, no second affected boys were born in families that had access to information about DMD after the first boy was found to be affected. The program also provides information to families.

Utilization of Dinucleotide Repeat Polymorphisms in DMD Diagnostics
Raymond Fenwick, Ph.D.
With the cloning of the DMD and development of diagnostic probes, studies found that about 60% of affected members have deletion mutations in the DMD gene, and about 5% have duplications. However, the mutations could not be defined in about 35%. Later, depletion-prone regions of the DMD gene were found to be highly polymorphic for (dC-dA) repeats. This dinucleotide repeat polymorphism was used to develop a test for carrier identification and prenatal diagnosis. Advantages include being able to test fewer family members to determine carrier status, and to control for errors in laboratory testing.

Carrier Testing Strategies in DMD
Peter Ray, Ph.D.
Female relatives of a DMD patient want to know their carrier status. More than 80% of boys with DMD have sporadic cases, making carrier status determination a challenge. A method for carrier status determination that uses multiplexed fluorescent dosage analysis was developed. The method, which has been used to detect deletions and duplications, is useful in most instances for detecting carriers. Women who are determined to be carriers can undergo prenatal diagnosis with 100% accuracy for detecting an affected fetus.

Current Molecular Diagnostics in DMD
Eric Hoffman, Ph.D.
In DMD patients who do not have deletions or duplications in the dystrophin gene, it is unclear if their genes have mutations. Therefore, muscle biopsies were obtained from 15 DMD patients and 21 Becker muscular dystrophy patients, RNAs were isolated, and cDNAs were prepared and sequenced. Various mutations were noted, such as missense mutations and stop-codon mutations. Mutations were absent in three DMD and seven Becker muscular dystrophy patients, giving results suggesting ambiguity and complexity. For example, the absence of mutations in three DMD patients suggested that they might not have had mutations in the dystrophin gene or even DMD itself, even with unambiguous clinical findings and an absence of dystrophin. Therefore, more sequencing studies are needed. In other work, the experience of about two dozen cases involving obtaining fetal muscle biopsies for diagnostic purposes were successful.

Discussion

Participants agreed that muscle biopsies remain essential for diagnosing DMD, but molecular diagnostics are also needed to provide more information.

A participant stated that molecular diagnoses, such as a report on the extent of protein expression or existence of mutation, should not provide a clinical diagnosis, such as the nature of the disease phenotype. Only clinicians should report the disease phenotype. Other participants disagreed.

Some participants noted that neonatal testing for DMD in the absence of available therapy is deemed by some to be unethical. Other participants replied with the offsetting argument that neonatal testing has the possibility of preventing the birth of a second boy with DMD.

Session 2: Pathogenesis

Cardiomyopathy in Dystrophin and Sarcoglycan Muscular Dystrophy
Elizabeth McNally, M.D., Ph.D (Topic Leader)
Life-threatening cardiomyopathy accompanies disorders of the dystrophin-glycoprotein complex, and needs to be considered in diagnostics and therapy.

In mice, cardiomyopathy accompanies deletions of dystrophin or of gamma- or delta-sarcoglycan. Histopathologic features in the skeletal muscles (large focal areas of degeneration and regeneration) are similar in these mice, but the features are milder in the dystrophin-deficient mdx mouse than in the others. Similarly, histopathologic features in the heart (large focal areas of degeneration) are also milder in the mdx mouse. Regeneration does not occur in cardiac tissue. Also, perivascular fibrosis is a prominent finding in cardiac tissue and skeletal muscle in these mice.

The origin of cardiomyopathy in these mice might arise from defects that disrupt the sarcoglycan and sarcospan complex in smooth muscle that in turn leads to defects in cardiac muscle. Beta-sarcoglycan and delta-sarcoglycan are expressed in smooth, cardiac, and skeletal muscles, but gamma-sarcoglycan is expressed only in cardiac and skeletal muscles. Thus, a deficiency of gamma-sarcoglycan cannot perturb smooth muscle. A cardiomyopathic, gamma-sarcoglycan deficient heart shows abundant expression of delta-sarcoglycan in the smooth muscle vasculature. This indicates that delta-sarcoglycan expression in the smooth muscle vasculature occurs and is present in the face of cardiomyopathy. Therefore, smooth muscle disruption of the sarcoglycan complex is not a necessary part of the development of the pathogenesis of the cardiomyopathy. The fibrotic features seen around the vessels are an important correlate, but may develop as a secondary consequence of degeneration ongoing in the muscle. This might be also important for other types of cardiomyopathy.

Cytoskeletal Interactions in Dystrophin in Skeletal Muscle
James Ervasti, M.D.
Dystrophin is highly enriched in structures called costameres, which are somewhat analogous to focal adhesions in non-muscle cells. The main function of costameres appears to be linking the force-generating contractile apparatus of muscle to the plasma membrane and out beyond the extracellular matrix. Costameres are believed to coordinate organized shortening of muscle cells that are arranged in parallel in the muscle membrane.

In-vitro studies suggested that dystrophin interacts mainly with F-actin. In the absence of dystrophin, such as in DMD patients and in the mdx mouse, the pattern of costameric staining is abnormal. However, it is difficult to study costameric structure in vivo because of the large extent of myofibrillar actin. To overcome this difficulty, Dr. Ervasti and coworkers used an established technique that turns the sarcolemmal membrane of a muscle cell inside out. The membrane is then stained and examined.

Staining patterns in hindlimb mouse muscle cells showed that dystrophin and F-actin are co-localized in a costameric pattern. However, muscle cells from mdx mice do not have costameric actin filaments. These and other results suggest a new function for dystrophin in stabilizing costameric actin, which might influence the macromolecular organization or the mechanical function of the costamere.

Overexpresion of Alpha 7 Integrin in Alleviating Muscular Dystrophy in Transgenic Mice
Stephen J. Kaufman, Ph.D.
Alpha 7 integrin links the cytoskeleton and extracellular matrix in skeletal and cardiac muscle. It plays a role in structure, function, and signal transduction in muscle, and mutations in it play a role in muscle diseases. Patients with DMD have increased concentrations of alpha 7 integrin. This raises the possibility that increased alpha 7 integrin is compensatory for the loss of dystrophin. To test this possibility, Dr. Kaufman and coworkers generated transgenic mice that overexpressed alpha 7 integrin in a background deficient in dystrophin and in utrophin. Mice with this background have severe muscular dystrophy. The enhanced expression of alpha 7 integrin in the transgenic mice markedly alleviated the development of muscular dystrophy. Thus, enhanced expression of alpha 7 integrin can compensate for the absence of dystrophin in mice.

Signal Transduction Pathway (JNK1) in the Pathogenesis of Dystrophic Cardiac and Skeletal Muscle
Lynn A. Megeney, Ph.D.
The c-jun N-terminal kinase (JNK1) is the only studied signal transduction factor found to be constitutively activated in cardiac and skeletal muscles from two dystrophic mouse models, mdx and myd:MyoD-/-. JNK1 has already been implicated in inducing apoptosis in nerve tissues.

Dr. Megeney and coworkers found that use of a kinase to upregulate JNK1 in cardiac and skeletal muscle cell lines resulted in defects in myotube formation and integrity that were similar to the dystrophic phenotype. These cellular effects were concurrent with JNK1-mediated phosphorylation and inhibition of an NF-AT transcription factor. This factor has been implicated as a key component of muscle hypertrophy. Blocking JNK-1 activity partly rescued the cells.

Representation differentiation analysis in the cell lines and mouse models showed increases in the expression other genes, including three genes implicated as pro-atrophic in the central nervous system: ß-amyloid precursor protein, which is involved in Alzheimer's disease; ciliary neurotrophic factor receptor; and neuronal leucine-rich repeat protein.

Proteomic methods are being used to find the cascade from the upregulation of JNK1 to the absence of dystrophin. Interleukin-2 was found to be upregulated.

The Role of Immune Cells in Dystrophin-Deficient Muscle
James Tidball, Ph.D.
Morphometric analysis is being done by quantification of histopathologic state of muscle sections. This enables studies of the effects of experimental treatments of mdx mice. Promising treatments can then be tested for their effect on extending the lifespan of the short-lived mdx/utrophin-deficient mice.

Depletion of CD8(+) cytotoxic T lymphocytes or of helper T cells caused reduction in histopathologic characteristics in the mdx mouse. Depletion of CD8(+) cytotoxic T lymphocytes resulted in an 80% increase in the lifespan of mdx/utrophin-deficient mice. This provided evidence that the immune system promotes dystrophic histopathology.

CD8(+) and helper T cells require the action of accessory cells, the professional antigen presenting cells, to react to the muscle tissue. A population of muscle macrophages was found to act as the professional antigen presenting cells. Depletion of these macrophages resulted in a substantial reduction in dystrophic histopathology.

These results demonstrate the possibility that immune interventions offer treatment opportunities for the dystrophinopathies.

Expression Profiling in DMD
Eric Hoffman, Ph.D.
Global gene expression changes were studied in muscle biopsies from patients with dystrophin or alpha-sarcoglycan deficiencies. Affymetrix expression profiling methods were used. Two hypotheses were tested: (1) A specific series of gene transcription changes is correlated with the deteriorating pathology of muscular dystrophies; and (2) gene expression changes unique to dystrophin and alpha-sarcoglycan deficiencies can be used to identify novel interacting proteins.

Compared with samples from unaffected controls, samples from patients with dystrophin or alpha-sarcoglycan deficiencies had changes in expression of 133 genes, including 59 upregulated and 74 downregulated. Changes in expression were in genes controlling degeneration and regeneration, inflammatory response, and metabolism.

Expression of about four genes was highly significantly different between samples from patients with dystrophin deficiencies and those with alpha-sarcoglycan deficiencies. In dystrophin deficiency, expression changes were noted in two genes. One of these genes encodes ERK6, a kinase whose major substrate is syntrophin, which is a component of the dystrophin-glycoprotein complex. The other gene encodes a protein phosphatase, and is subject to further study.

Session 3A: Gene Therapy Approaches

A. Delivery (viral mediated; naked DNA)

Introduction and Overview
Jeffery Chamberlain, Ph.D.(Topic Leader)
Expression of the major form of dystrophin is sufficient for preventing or alleviating dystrophy in mdx and other mice. Expression of the other forms is not necessary. Expression of utrophin prevents dystrophy in mdx mice. This work needs to be expanded so that it can be applied to alleviating dystrophy in patients.

Issues that need to be considered are deciding what needs to be delivered into muscle; selecting the expression cassette and the best vectors; measuring the response of the immune system; delivering genes into all affected cells; and assessing the outcome.

Selection of the Optimal Transgene for Gene Therapy of DMD
Dominic J. Wells, Ph.D.
DMD patients express little or no normal dystrophin, and it is unlikely that most dystrophin epitopes have been presented to their immune system. This raises the concern that introduction of dystrophin into the muscles of DMD patients will lead to immune rejection of corrected fibers.

Plasmid-based intramuscular gene transfer into mice was used to study immune responses to dystrophin and the restoration of the dystrophin fibers. Plasmid DNA was used to avoid the introduction of any other proteins, which would complicate the analysis of transgene immune responses.

Administration of human dystrophin cDNA into skeletal muscles of mdx mice caused an immune response leading to loss of corrected fibers. This did not occur in upon administration into immunoincompetent mdx/nude mice. In contrast, administration of mouse dystrophin cDNA did not cause an immune response in mdx mice.

Dr. Wells believes that the mdx mice were tolerant to mouse dystrophin because of the presence of rare dystrophin-positive revertant fibers. Therefore, DMD patients who express some revertant fibers might be tolerant to the entire dystrophin molecule. This possibility is being tested.

Gutted Adenoviral Vectors for Gene Therapy of Muscular Dystrophy
Jeffery Chamberlain, Ph.D.
Gutted (helper-dependent) advenovirus vectors are being developed for gene therapy of DMD. Gutted viruses are deleted for all viral genes, creating a large cloning capacity (greater than 30 kilobases) that can accommodate the full-length dystrophin cDNA and appropriate gene regulatory elements. The lack of viral genes reduces the potential for a cell-mediated immune response against the vector.

Dystrophin-expressing gutted vectors can introduce a substantial amount of dystrophin into muscles of mice, and can reverse muscular dystrophy functional deficits in adult and old mdx mice. Therefore, Dr. Chamberlain and coworkers are optimizing methods for large-scale production.

Advenoviral Vectors: Humoral and Cellular Immune Responses in Dystrophin Delivery to Skeletal Muscle
Paula Clemens, M.D.
A high-capacity, fully deleted adenovirus vector expressing a full-length mouse dystrophin cDNA driven by a muscle-specific promoter was injected into muscle in mice. Muscles were collected for as long as 1 year after injection into newborn mice and 8 weeks after injection into adult mice. Dystrophin concentration was analyzed by Western blot analysis.

Dystrophin expression was stable until 20 weeks in newborn mice, and then declined. Dystrophin expression remained stable through 4 weeks in adult mice, but declined by 8 weeks. Studies are in progress to explain the decline. For example, mice injected as newborns do not have indications of antibodies to adenovirus, but mice injected as adults do. The numbers of newborn mice that have dystrophin antibody increased with increasing time after injection. Adult mice always had antibody to dystrophin after injection.

High-capacity adenovirus vectors elicit less immune response to adenovirus than do first-generation adenovirus vectors.

Intravascular Delivery of Naked DNA Molecules
Jon Wolff, M.D.
The efficiency of direct injection of naked plasmid DNA into muscle results in low expression, but the efficiency of intravascular injection of naked plasmid DNA into liver results in high expression. Dr. Wolff and coworkers recently provided evidence supporting their hypothesis that uptake of naked DNA is by a receptor-mediated process and not by injury. Anti-DNA antibodies against naked DNA have not been observed even after repeated DNA injections. The injections have been well tolerated by animals. The approach appears promising for delivering dystrophin DNA to humans.

Systemic Gene Delivery and Quantitation of Sarcolemmal Protection
Hansell Stedman, M.D.
A small-animal model system and a large-animal model system were developed for use in the study of vector transport from the central (aortic) circulation. Temporary use of extracorporeal circulatory support in conjunction with invasive hemodynamic monitoring enabled stabilization of animals undergoing cardiac and/or total circulatory arrest with forced extravasation of macromolecular complexes from the central circulation. This approach holds promise for use in widespread vector delivery to striated muscles throughout the body.

It is anticipated that intramuscular injection of vector in early clinical trials will result in sharp gradients between transduced and untransduced fibers, complicating the study of efficacy. Based on recent studies of the time course of protein degradation after synchronized injury of muscle fibers, a rationale can be presented for a systematic, automated analysis of sarcolemmal protection conferred by recombinant proteins. This approach should provide the sensitivity and specificity needed to establish the relationship between the extent of transgene expression achieved and the extent of protection occurred.

Trial of Adeno-associated Virus-mediated Gene Transfer into Humans
Jerry Mendell, M.D.
Dr. Mendell described gene transfer of alpha-sarcoglycan in an adeno-associated virus vector into two individuals who had alpha-sarcoglycan disease, and showed a videotape of the procedure. Injections were made into the short extensor muscles of the toes. One side of the body received vector and the other received a sham injection. Six weeks later, biopsies were obtained from the injection site.

Use of the short extensor muscles of the toes might have disadvantages. The muscle in one individual receiving vector looked robust. It is a residual muscle that is not used often. In contrast, dystrophic limb muscle looks pale. In addition, it is difficult to obtain strength data from studies of the short extensor muscles. Therefore, the short extensor muscles might be useful for initial toxicity studies. Later, a limb muscle would be used for efficacy studies.

Discussion

Participants discussed concerns about gene therapy, such as preparation purity, immunologic responses, and optimal promoters.

Serge Braun, Pharm. D., Ph.D., described a French phase I clinical trial of injections of full-length dystrophin plasmid including a cytomegalovirus promoter into muscles of nine individuals who have DMD or Becker muscular dystrophy. The subjects are about 15 years old. The aim of the study is to look, first, for expression and then, to evaluate immune response to the transgene.

Session 3B: Gene Therapy Approaches

B. Gene repair and modulation of expression

Chimeraplast-mediated Gene Correction--In-vivo/In-vitro Studies
Thomas Rando, M.D., Ph.D.(Topic Leader)
Chimeraplasts are chimeric RNA/DNA oligonucleotides. Chimeraplast-mediated gene correction is currently ideally suited for doing single-base exchanges in genomic DNA to correct a mutation in the genome. A chimeraplast has a sequence that is identical to the genomic sequence, except for a single mismatch at the mutation point. Evidence shows that when pairing occurs, cellular DNA mismatch repair mechanisms correct the mismatch, leading to restoration of a wild-type gene sequence. Advantages of this approach include that the target gene remains regulated by its own mechanisms and that exogenous DNA is not introduced.

Dr. Rando and coworkers have been injecting a chimeraplast into muscles of mdx mice to correct the point mutation in the dystrophin gene. Immunohistochemical analysis revealed the presence of dystrophin-positive clusters around the injection site, and immunoblot analysis showed that dystrophin from mdx-injected muscles was full-length. Staining with exon-specific antibodies showed that these fibers were not revertants. RT-PCR demonstrated the presence of wild-type sequence. Dystrophin and wild-type sequences were not observed after injection of control chimeraplasts. Attempts will be made to use chimeraplasts to correct other kinds of mutations. Cellular uptake of chimeraplasts remains a technical challenge. Systemic delivery into muscles will probably pose the major limitation.

Antisense Oligonucleotide-mediated Exon Skipping -- In-vitro Studies
J. George Dickson, Ph.D.
Revertant dystrophin in DMD might arise in part from exon skipping that bypasses a mutation. The mdx gene has a nonsense point mutation in exon 23. Transfection of cultured primary mdx myoblasts with antisense oligoribonucleotides corresponding to the 3'-splice site in intron 22 resulted in expression of dystrophin. Direct sequencing of RT-PCR products showed precise splicing of exon 22 to exon 30, skipping the mutant exon and creating a novel in-frame dystrophin transcript. These results might offer a therapeutic approach to DMD.

Antisense Oligonucleotide-mediated Exon Skipping -- In-vivo Studies
Steve Wilton, Ph.D.
Antisense oligoribonucleotides directed to the 5'-splice site of intron 23 in the mdx gene were transfected into cultured mdx primary myoblasts. Exon 23 skipping was 100%. However, transfection of the same oligoribonucleotides into immortalized cell lines was not as efficient. Use of newer oligoribonucleotides increased the efficiency to almost 100%. In preliminary experiments, intraperitoneal injection of antisense oligoribonucleotides into mice resulted in low extent of exon-23 skipping in the liver. Further work will consider issues of delivery and toxicity.

Potentiation of Uptake of Oligonucleotides Using Peptide Translocation Domains
Keith W. Jones, Ph.D.
HIV-1 encodes a transcription factor, transcriptional activator protein (TAT), which has an arginine-rich, 11-amino acid portion, the peptide translocation domain. Recent publications showed that this and other arginine-rich peptide translocation domains are very effective at linking to various cargoes and getting into many cell types. Translocation appears to a quick process that is energy- and receptor-independent, although some results suggest otherwise. The translocation motif does not appear to have a specific sequence or stereoselectivity. Dr. Jones and coworkers showed that for TAT to be effective, it must be covalently linked to its cargo.

Dr. Jones and coworkers aim to identify novel translocation domains and then use these to deliver nucleic acids, including oligonucleotides for imaging, chimeraplasts for target validation, and plasmids for various purposes. Dr. Jones and coworkers seek to develop a linkage between the translocation domain and the cargo such that when the cargo is delivered to its target, the linkage will break.

Some of the remaining questions about peptide translocation domains concern the mechanism of translocation, which is unknown, the limitations of cargo, and the possibility of immunogenicity.

Aminoglycoside-based Therapies for Modification of Translation: Studies In Vitro and In Vivo
H. Lee Sweeney, Ph.D.
Some cases of DMD result from stop codons that cause premature termination of translation of dystrophin mRNA. This is also the instance in the mdx mouse. Perhaps administration of aminoglycoside drugs, which derive their antibiotic property from suppressing premature termination, would be beneficial.

Dr. Sweeney and coworkers found that gentamicin suppresses premature termination in primary cultures of mdx myotubes and in muscle from mdx mice. A full complex was restored in muscle. This approach has potential for treating patients. Although adverse events including ototoxicity and nephrotoxicity are concerns, these effects can be managed. This approach would enable measurement of dystrophin in the patients. Trials are in progress or being planned.

Aminoglycoside-based Therapies for Modification of Translation: Discussion of Clinical Trials
Kenneth Fischbeck, M.D.
A pilot study at NIH tested if gentamicin could restore dystrophin concentrations in patients with nonsense mutations in the dystrophin gene. The primary objective was to see if doses used for antibiotic treatment would restore dystrophin. A secondary objective was to test for increased muscle strength. CK concentrations were also measured.

Among the four participants, who ranged in age from 6 to 18 years, three had DMD and one had Becker muscular dystrophy. Dystrophin was measured muscle samples, and strength evaluation was done by quantitative muscle testing and by dynamometry. Overall, muscle strength did not change by a statistically significant amount. Decline in serum CK concentration was statistically significant. Toxicity was not observed. Dystrophin analysis is in progress.

Aminoglycoside-based Therapies for Modification of Translation: Discussion of Clinical Trials
Jerry Mendell, M.D.
A trial of gentamicin for treating dystrophin-deficient patients and sarcoglycan-deficient patients is in progress. The primary efficacy measure is dystrophin expression, and secondary efficacy measures are changes in strength and in CK concentration. CK concentrations can vary widely during the day, so it is important to collect serum samples at the same time of day.

Session 4: Cell Therapy

Novel Stem Cells in Skeletal Muscle
Louis Kunkel, Ph.D.(Topic Leader)
Staining muscle cells with a vital dye revealed a subpopulation of cells that stained less brightly than the main population. These cells were obtained from normal male mice and then injected into female mdx mice. Cells in muscles of the recipients expressed dystrophin, and the presence of the Y chromosome in these cells confirmed that they originated from the donor male mice. The next step is to obtain marker characteristics of the donor cells.

Muscle Derived Stem Cells in Ex-vivo Gene Transfer to Enhance Muscle Regeneration in DMD
Johnny Huard, Ph.D.
Trials of myoblast transplantation for DMD were unsuccessful because of three limitation: immune response, poor migratory capacity, and poor survival of injected cells. Therefore, Dr. Huard and coworkers tried to isolate a population of muscle cells that would overcome these limitations. Collagen-coated flasks were used to separate cells according to their adherence. Enrichment of desmin-positive cells from mouse primary muscle cell culture also enriched a cell population containing CD34 and Bcl-2. Clonal isolation from this population yielded a putative mesenchymal stem cell, MC13, which is capable to differentiating both into myogenic and osteogenic lineage in vitro and in vivo. The intramuscular injection, and, more importantly, the intravenous injection of MC13 cells genetically engineered to express dystrophin resulted in the enhancement of muscle regeneration and the partial restoration of dystrophin in mdx mice. These results suggest that muscle-derived stem cells have been isolated that can be used for ex-vivo gene transfer approaches to enhance muscle regeneration to deliver dystrophin into dystrophic muscles. Also, CD34 cells are present in muscles from boys with DMD. This offers the possibility of ex-vivo gene therapy in which these cells are taken a patient, genetically engineered to produce dystrophin, and then returned to the patient.

Markers of Quiescent Satellite Cells in Normal and Dystrophic Mouse Muscle, and Development of a Model to Compare Human and Mouse Myogenic Cells
Terence Partridge, Ph.D.
None of the markers of quiescent satellite cells reliably distinguish between stages of disease or among ages of an animal. M-cadherin is a good marker of satellite cells. Antibodies to M-cadherin reliably detect satellite cells, and M-cadherin is specific for satellite cells. CD34 is also a good marker quiescent satellite cells in isolated fiber preparations, but is not useful for histologic preparations because it is present in the vasculature and in inflammatory cells. However, some unidentified cells are in the satellite cell position but lack the markers.

In separate but related work, human myogenic cells are being compared with mouse myogenic cells.

Signaling Factors of Satellite Cells in Quiescence, Activation, and Proliferation in Intact Skeletal Muscle Fiber Preparations
Bradley Olwin, Ph.D.
Satellite cells, the primary source of nuclei used for skeletal muscle hypertrophy and repair, are maintained in a quiescent and undifferentiated state between the skeletal muscle fiber and the basement lamina. Dr. Olwin and coworkers are studying factors responsible for activation and proliferation of quiescent satellite cells.

Single-cell RT-PCR showed that c-met (hepatocyte growth factor receptor, HGFR) and fibroblast growth factors 1 and 4 (FGF-1 and FGF-4) are expressed in quiescent satellite cells. Intracellular mediators of FGFs and HGF include the mitogen-activated protein kinase (MAPK). Consistent with this observation, active extracellular signal-regulated kinase 1,2 (ERK1,2, a form of MAPK) was observed in activated satellite cells of intact myofiber cultures.

To study the roles of the MAPK, Dr. Olwin and coworkers added specific inhibitors of ERK1,2 and p38 (a form of MAPK). Inhibition of ERK1,2 had no effect on the activation and proliferation of satellite cells. Inhibition of p38 resulted in the loss of satellite cells from intact fibers. This was similar to the result noted after treatment with sodium chlorate, an inhibitor of proteoglycan sulfation.

Heparin sulfate plays an important role in regulating growth factor action, and HGF and FGFs bind tightly to heparin sulfate and require it in addition to their respective receptors to transduce intracellular signals. Therefore, Dr. Olwin and coworkers asked if heparin sulfate plays an important role in satellite cell function. They predicted that addition of sodium chlorate would block signal transduction. The results showed within 24 hours of sodium chlorate treatment, satellite cells were no longer present on intact myofibers. These effects were partly reversed upon addition of sulfate or of heparin, suggesting that heparin sulfate proteoglycans are critical for satellite cell maintenance or activation.

The ERK1,2 pathway appears to be involved in proliferation, and does not affect differentiation of skeletal muscle cells. In cultured MM14 skeletal muscle cells, inhibition of ERK1,2 or of p38 blocked proliferation. Both ERK1,2 and p38 are stimulated by FGF. FGF is required for MM14 cell proliferation. Without FGF, the MM14 cells irreversibly differentiate. However, p38 is also required for differentiation. Removal of FGF inhibition of p38 from cultures of MM14 cells resulted in cells that neither proliferated nor differentiated.

Time- and Expression-dependent Restoration of Dystrophin to Dystrophin-deficient Myotubules
Judy Anderson, Ph.D.
Nuclear domains of dystrophin have been observed in myoblast transplantation experiments and in muscle fibers after somatic reversion in human DMD and in mdx mouse muscular dystrophy. However, the domain features of dystrophin are not well established because of the limitations and complexity of in-vivo studies. Therefore, Dr. Anderson and coworkers used a mouse satellite co-culture system to form hybrid myotubes from fusion between different ratios of dystrophin-negative (mdx) and normal myoblasts, and examined the dynamic evolution of restoring dystrophin expression in dystrophin-deficient myotubes.

Nuclear domains of dystrophin were clearly identified in hybrid myotubes. The occurrence of domain events was affected by the ratio of normal to dystrophin-deficient satellite cells. A higher amount of dystrophin expression produced fewer nuclear-domain containing myotubes and more myotubes that were homogeneous for dystrophin along their length. The domain feature of dystrophin expression was transitory, and the diffusion of dystrophin protein was asymmetrical from the source nuclei. Also, diffusion could restore the dystrophic cytoskeleton to normal as the domains enlarged, as demonstrated by the restoration of beta-dystroglycan, the clustering of large acetylcholine receptors, and the co-localization of rapsyn (an acetylcholine receptor-associated protein) with dystrophin. However, muscle-specific kinase is not co-localized and extends beyond the regions in which dystrophin is localized.

For obtaining as many satellite cells or stem cells as possible from a muscle, it is important to know how to get them in the best way. Activated satellite cells have to decrease their adhesion to the fiber to be isolated. Activation systems in other cells entail this use of shear-induced release of nitric oxide (NO). Therefore, Dr. Anderson and coworkers decided to find out if NO mediates satellite cell activation is through shear-induction or another way. They found that NO mediates satellite cell activation in inured leg muscle in a mouse. Inhibition of nitric oxide synthase (NOS) activity prevented immediate injury-induced myogenic cell release and delayed the hypertrophy of satellite cells in that muscle. These results suggest that NO release mediates satellite cell activation, possibly by shear-induced rapid increases in NOS activity.

In other work, Dr. Anderson and coworkers showed that treatment with the glucocorticoid drug deflazacort reduces the central nucleation ratio in mdx mice. Addition of a NO donor or of an NOS inhibitor to the deflazacort caused various changes in the central nucleation ratio. Therefore, deflazacort might be acting through NOS expression and activity.

Molecular Mechanisms Regulating the Specification and Developmental Program of Muscle Stem Cells
Michael Rudnicki, Ph.D.
Dr. Rudnicki and coworkers have been identifying genes whose expression is specific to the satellite cell lineage. One of the genes, Pax-7, encodes a protein that is a member of the Pax family of nuclear transcription factors, which regulate the development of diverse cell lineages during embryogenesis. Pax-7 is expressed in quiescent satellite cells.

Pax-7 knockout mice, generated by another laboratory, have histopathologic characteristics consistent with a satellite cell deficit. Dr. Rudnicki and co-workers were unable to derive myoblasts from these mice. Myoblasts are readily derived from wild-type mice. Cultures of intact myofibers from Pax-7 mice did not have cells coming off the cultured fibers. The results suggested either that much of the satellite cell lineage is not there or that the satellite cells cannot be activated.

To distinguish between these possibilities, Dr. Rudnicki and coworkers studied Pax-7 muscles by electron microscopy. Although satellite cells are typically found in wild-type muscle cells, they were not found in the Pax-7 muscle cells.

However, use of fluorescence-activated cell sorting showed that CD4(+) single positive (SP) cells were present in the same amount as in wild-type muscles, about 1.8%. Presumably these cells represented pluripotent stem cells in muscle. Cultures of satellite cells from wild-type muscle generated hemopoietic cells, myoblasts, adipose cells, and fibroblasts. Cultures of satellite cells from Pax-7 generated a 10-fold increase in the proportion of hemapoietic cells and no muscle cells.

The results supported the idea that SP cells are the progenitors of the myogenic lineage. The results showed that SP cells are a distinct population from satellite cells. The results also show that Pax-7 functions to specify the development of a satellite cell compartment in muscle, and that Pax-7 functions by restricting the potential of a pluripotent progenitor. In the absence of Pax-7, the SP cells have increased hemopoietic potential. The biological implications are the possibility that satellite cells are continually derived throughout life from the SP compartment, and that self-renewal within the satellite cell compartment is limited and perhaps transient. A therapeutic possibility is to do ex-vivo gene therapy on pluripotent progenitors by forcing expression of Pax-7, making them become muscle.

Session 5: Pharmacologic Therapy

(A) Transcriptional Regulation of the Utrophin Gene
(B) Promoter Screening Against Small Compound Libraries
Kay E. Davies, D.Phil.(Topic Leader)
Considerable evidence supports the conclusion that sufficient amounts of utrophin can replace dystrophin in mice. A limitation concerns the extent of upregulation of utrophin that would be needed to treat DMD in a patient, because humans have larger muscles than mice. One of the first issues concerns toxicity. Dr. Davies and coworkers injected into mdx mice a utrophin transgene under the control of a ubiquitous promoter which led to the overexpression of utrophin in many tissues. They found no evidence of toxicity in any tissue studied.

DMD patients are born dystrophin-deficient, so the question arises about the effect of delivering utrophin at birth. Dr. Davies and coworkers found that delivery of an adenovirus-vector containing utrophin into newborn dystrophin/eutrophin-deficient double knockout mice prevented expression of the dystrophic phenotype. The next question concerns how often such a drug would have to be administered to maintain therapeutic benefit. Work with a transgene under tetracycline control is being done to answer this question.

One potential way to increase utrophin expression in muscle is through transcriptional upregulation of the endogenous utrophin gene. The first recognized utrophin gene promoter, promoter A, includes a conserved promoter element, the N-box, which is responsible for the synaptic localization of utrophin gene expression. Therefore, Dr. Davies and coworkers subjected a construct encompassing promoter A to a high-throughput screen of 160,000 compounds to find those that upregulate the promoter. Ten compounds were found. This promising result is being followed up.

However, promoters of other genes also include an N-box, leading to the concern that upregulating promoter A would also upregulate promoters of other genes. Therefore, Dr. Davies and coworkers hypothesized that the utrophin gene contains other promoters. Using 5'RACE, they found a completely independent promoter, promoter B, which does not have an N-box or a TATA sequence. Transcription directed by this promoter resulted in the accumulation of as much mRNA as does transcription directed by promoter A. Dr. Davies and coworkers are looking for the presence of other promoters in the utrophin gene.

Regulatory Events in the Upregulation of Utrophin, and its Relevance to Pharmacologic Interventions
Bernard J. Jasmin, Ph.D.
Utrophin and utrophin mRNA accumulate at the neuromuscular junction, but utrophin mRNA is also detected in extra-synaptic regions of muscle cells. Dr. Jasmin and coworkers characterized some of the key factors that regulate the synaptic accumulation of utrophin mRNA. Two nerve-derived trophic factors, agrin and ARIA/heregulin, bind to their own receptors to trigger a signaling cascade that culminates in the phosphorylation of ets-related transcription factors GABP alpha and beta. Once phosphorylated, these transcription factors bind to the N-box motif of the utrophin gene, activating its transcription in myonuclei located in the vicinity of the neuromuscular synapse.

Dr. Jasmin and coworkers also tried to identify other mechanisms that maintain the pronounced accumulation of utrophin mRNA at the neuromuscular junction. The half-life of utrophin mRNA in cultured muscle cells is about 16 to 20 hours, similar to the half-life of dystrophin mRNA that was reported by another group. The half-life of utrophin mRNA did not change during differentiation of the cells from myoblasts to myotubes. Also, utrophin mRNA has to be in the correct intracellular location to be translated efficiently. Results of cellular fractionation experiments suggested that most utrophin mRNA was in the cytoskeletal-bound polysomes that are attached to the actin microfilament network.

Dr. Jasmin and coworkers identified a 150-base region in the 3'-end of utrophin mRNA that controls its stability and targeting, offering a different approach to upregulating utrophin gene expression. In a series of experiments, Dr. Jasmin and coworkers started to characterize the interaction between the 3'-untranslated region of utrophin mRNA and cytoplasmic binding proteins. UV cross-linking experiments and RNA gel shift assays suggested the presence of cytoplasmic factors that bind to the 3'-untranslated region of utrophin mRNA. The proteins are being characterized.

Dependence of Utrophin Assembly/Expression on Alpha-Syntrophin
Stanley Froehner, Ph.D.
The dystrophin complex maintains the integrity of the membrane at the neuromuscular junction, but also might be a scaffold upon which signaling proteins are assembled. One of the proteins that is bound directly to dystrophin is syntrophin, a modular protein comprised entirely of interaction motifs, including PDZ (protein domain named for PSD-95, discs large, ZO-1). PDZ domains might be important for bringing signaling proteins to the sarcolemma and synapses. The presence in the dystrophin-glycoprotein complex of kinases such as ERK6 also suggests that the complex is a signaling scaffold.

Dr. Froehner and coworkers generated alpha-syntrophin knockout mice, and found that these mice are not dystrophic. However, the microscopic examination showed that their neuromuscular junctions are unusual and aberrant. Unexpectedly, neuronal NOS was found to be lost from the membrane. Therefore, alpha-syntrophin is required for targeting of neuronal NOS to the membrane. A surprising result was that utrophin is almost completely absent from the neuromuscular junctions.

One explanation of why knocking out alpha-syntrophin affected the association of eutrophin is that alpha-syntrophin is important for the assembly of the complex. Another possibility is that alpha-syntrophin regulates the expression of utrophin. Dr. Froehner and coworkers are testing these ideas. Furthermore, syntrophin might have a role in treating DMD.

The Use of Protease Inhibitors as Therapeutic Agents in DMD and Related Disorders
Alfred Stracher, M.D.
Extensive proteolytic degradation occurs in muscle wasting diseases such as muscular dystrophy. Therefore, Dr. Stracher and coworkers hypothesized that inhibiting the proteases responsible for the proteolysis would delay proteolytic degradation and slow the deterioration in these diseases.

In any ischemic or traumatic process that injures the cell membrane, the cell is more permeable to calcium, which flows into the cell and activates calpain, a selective protease. Activation of calpain leads to widespread degradation leading to cell death. Dr. Stracher and coworkers hypothesized that inhibiting calpain activation would prevent subsequent degradation. They focused on using leupeptin, a tripeptide that is reasonably selective as a calpain inhibitor.

Intramuscular injection of leupeptin at the wound site led to restoration of muscle and nerve in monkeys that had a wounded nerve. Oral administration into similarly wounded monkeys restored the size of muscle at the injury site, and increased muscle size at uninjured site. Intramuscular injection of leupeptin inhibited almost totally restored myofiber diameter in mdx mice.

Leupeptin was linked to carnitine, a naturally occurring compound that accumulates selectively in skeletal and cardiac muscle. This modified leupeptin inhibits intracellular calpain activity about 15-fold more effectively than unmodified leupeptin.

Trials in patients are being planned. The applicability of protease inhibitors for treating other neurodegenerative disorders is being considered.

Results of High Dose Oral Prednisone in Boys with Duchenne Muscular Dystrophy
Anne Connolly, M.D.
Daily administration of prednisone (0.75 mg/kg) improves strength in boys with DMD. Adverse events, including linear growth arrest, obesity, and cushingoid features almost always occur between 6 months to 1 year, leading to reduction or elimination of the dose in about half of the boys, including those who have benefited. Reasoning that adverse events of prednisone arise from immunosuppression, Dr. Connolly and coworkers reasoned that changing the dosing might lead to fewer adverse events. Dosing was changed to oral 10 mg/kg per week split between two daily consecutive mornings. The results showed that participants treated this way demonstrated improved quantitative strength. In contrast, participants on daily prednisone and untreated participants showed no change. Linear growth was maintained in all of the two-morning participants. Improvements were noted in participants 7 years and older, who often do not benefit from prednisone treatment. Excessive weight gain was not noted. Some participants were irritable during treatment days, and some had abdominal pain.

Discussion of Pharmacological Treatment in DMD
Jerry Mendell, M.D.
In 1978, a group of clinical investigators started to define the natural history of DMD. At age 7.5 years, patients start to lose strength. Before the gene encoding dystrophin was discovered, DMD patients were classified into three groups. One group could no longer walk by age 12, a second group could no longer walk by age 15, and a third group could walk after age 15.

The first trials, which used leucine and calcium channel blockers, were unsuccessful. In planning the next trials, the investigators recalled the observation that muscle from DMD patients have a characteristic that is typical of inflammatory myopathies, infiltration of non-necrotic muscle fibers by CD8(+) cytotoxic T lymphocytes. Also, necrotic fibers activate the complement cascade. These observations prompted therapeutic trials with prednisone, which demonstrated efficacy. Surprisingly, prednisone had little or no effect on CK activity, and paradoxically, prednisone increased muscle mass. In every other setting, prednisone decreases muscle mass. Controlled comparisons are needed to define the timing of prednisone dosing so as to minimize the occurrence of adverse events.

In view of the need to use a drug associated with fewer adverse events, Dr. Mendell and co-workers compared deflazacort with prednisone and placebo. Deflazacort had similar efficacy and caused fewer adverse events. Unfortunately, deflazacort is not approved for use in the United States, and pharmaceutical companies regard pursuing approval as unprofitable. However, patient networks obtain it from other countries.

Trials of other immunosuppressants, such as azathioprine and mycophenolate, are an area of potential investigation. Trails with cyclosporine yielded modest results. Dr. Mendell and coworkers recently completed a trial with oxandrolone, an androgenic steroid favored by athletes. Presumably, oxandrolone has powerful effects in increasing muscle strength associated with exercise. The result showed that oxandrolone is not efficacious.

Given prednisone's unfavorable adverse events profile, clinicians disagree about when to start it. When patients stop taking prednisone, their strength returns to that which would be expected from the natural history of untreated DMD. Thus, some clinicians prefer to start prednisone as early as possible, whereas others prefer to wait as long as possible.

Session 6: Resources/Future Directions

H. Lee Sweeney, Ph.D.(Topic Leader)

Comparative International Neuromuscular Research Group
Diana Escolar, M.D.
The Comparative International Neuromuscular Research Group is a 20-site collaboration that intends to do translational research that takes hits from a mdx mouse drug screen and considers them for human trials. Each of the 20 sites has a neurologist, a clinical coordinator, and a physical therapist. The collaboration developed a quantitative test and avoids paper by entering data to a Web site for analysis by the statistical center at Children's National Medical Center, Washington DC. The investigators are open to suggestions about trials.

Assessment with NMR Imaging
H. Lee Sweeney, Ph.D.
Muscle-strength testing is the only outcome to be tested in drug trials. However, in gene therapy trials, expression of the transgene also needs to be examined.

Dr. Sweeney and coworkers are interested in developing noninvasive methods to measure therapeutic outcome. Use of NMR might be useful. NMR marker genes are useful in mouse studies because these genes encode a marker that follows time and distribution of expression. However, NMR marker genes might not be useful in human trials because of the need to use the engineered marker gene itself.

NMR imaging can easily assess muscle volume. Uptake of albumin/gadalinium conjugates could be measured. Generating T2 maps offers a noninvasive way of measuring efficacy of gene transfer, or, in gentamicin trails, of determining if the muscle has returned to normal phenotype.

Preclinical Assay of Sarcolemmal Integrity in Intact mdx Mice
John G. Quinlan, M.D.
Evan blue dye (EBD) binds to albumin, and is nontoxic. When a muscle membrane breaks down with sarcolemmal destruction, the dye enters the fibers, which are stained orange-red against a green background noted by fluorescence microscopy. To assess muscle membrane integrity, Dr. Quinlan and coworkers injected EBD injected intraperitoneally into mdx mice, and stressed their muscles by running the mice on a treadmill. Mice were killed, and the muscles were examined by histopathologically. Preliminary results showed the presence of more EBD-positive fibers in muscles from runners than from non-runners. This approach might have application as a screening technique.

Gene Therapy: One Biotechnology Viewpoint.
R. Michael Blaese, M.D.
Large pharmaceutical companies want to make at least $500 million during the peak marketing year of drugs that they are marketing. Therefore, they are not interested in developing drugs to treat rare diseases. Investors in small companies want to sell their companies to large companies, so the investors are interested in trying to develop something that interests large companies.

Indications for orphan drugs are often perceived as a threat by large companies. Some large companies do not want their drugs tested clinically for rare disease indications because the investigation might result in unfavorable findings. However, drugs to treat rare diseases can be profitable.

The death of a participant in a gene therapy trial for ornithine transcarbamylase deficiency resulted in various consequences. Protocols using adenovirus vectors were suspended, although some have resumed. The extent of record-keeping and review has increased, and the cost of vector production doubled.

Technical issues, such as ectopic gene expression in using hemopoietic stem cells for reconstitution, are of concern. A requirement is the use of appropriate regulatory elements. Because of the limited capacity of vectors, cDNAs must be used, which raises other questions about regulation and limits the number of splice variants. Random integration remains a problem. Various technical issues also concern delivery of vectors.

Discussion

Bill Moore, M.S., Research Program Coordinator of the Muscular Dystrophy Association (MDA), discussed the MDA's concerns about continuing its limb-girdle muscular dystrophy trial when difficulties arose at the University of Pennsylvania's gene therapy trial. Operation of the adeno-associated virus vector production facility that the MDA was using was suspended, and the MDA sought another source. The MDA plans to use the vector facility at the University of Florida, and is developing a protocol to submit for FDA approval. The MDA hopes that the trial resumes by the end of 2000.

Participants discussed the use of other animal models besides the mdx mouse. One participant noted that dogs vary, which makes it unfeasible to use them for preclinical trials. Another participant noted that dystrophic pigs might make a good model. They are the right size, are familiar to surgeons, and do not engender the emotional attachments that dogs do in people. However, the large size of a pig necessitates a large amount of drug for testing efficacy. Another participant believed that non-human primates should be used for tests of safety and efficacy.

A participant noted that the incidence of DMD in the United States is about 100,000 new cases per year, which means that DMD is not an orphan disease. Pharmaceutical companies should therefore be interested in developing drugs for DMD. Furthermore, developing drugs for DMD could be made more attractive to pharmaceutical companies as part of an idea to develop drugs for muscle weakness.

A participant urged support for basic research in muscle biology. Basic research is the basis of therapeutics. The participant recalled that MDA-sponsored fellowships prompted investigators to go into muscular dystrophy research. Perhaps the NIH should consider sponsoring a similar program. Other participants discussed other aspects of funding, such as stating that research support be broad-based. Perhaps NASA and other institutes at NIH besides NINDS and NIAMS could participate. NIH could issue a request for applications (RFA), which can stimulate research in a particular area. The climate for funding is favorable, but this could change. A participant noted that despite the improved funding climate, the extent of support is still only 25% of those who apply for it, which is discouraging to young scientists. This issue needs to be considered.

A participant suggested that NIH should consider supporting center applications for muscular dystrophy research. However, Dr. Spinella noted that NINDS already supports center applications, so it is up to investigators to apply. Dr. Lymn stated that NIAMS generally supports centers by issuing an RFA. NIAMS's support for centers changes year by year. Also, NIAMS has muscle-research training grants that include neurologic research. Dr. Spinella and Dr. Lymn clarified that indirect costs of a grant proposal are not included in the limits that may be applied for, correcting a misimpression.

A participant suggested other approaches to expanding research, such as an RFA that would include rare diseases with more common diseases. For example, am RFA about muscle wasting, which would include atrophy, dystrophy, aging, and space flight, could be jointly sponsored by different institutes. This would assure adequate funding and could lead to cooperation among institutes in NIH.

Applicants from outside the United States may apply for funding, as long as the application states that the work is not being done within the United States.

Research in dystrophin-like genes could be expanded to include simpler organisms, such as the worm, Caenorhabditis elegans.