Journal of Muscle Research and Cell Motility 13, 377-380 (1992) SYMPOSIUM
REPORT
Current status of research on the Xp21 myopathies The rapid advances in knowledge of the dystrophin gene and its function make it important to appraise from time-to-time the developments in the basic sciences for their clinical implications. To this end the Italian National Academy (Accademia Nazionale dei Lincei) and the British Council sponsored, with help from the Italian Muscular Dystrophy Association, a meeting on 20 and 21 February (1992) in Rome at the Palazzo Corsini. Although principally an Anglo-Italian occasion the programme was strengthened by invited papers from leading workers in the field from other countries.
Genes and their products The introductory talks by D. J. Blake presenting the work of the Oxford group and D. Yaffe from the group at the Weizman Institute (Rehovot, Israel) concentrated on the genes concerned in the synthesis of dystrophin and dystropbin related proteins. From the two genes coding for these proteins seven different gene products have so far been identified. The dystrophin gene located on the X chromosome encodes about 74 exons and gives rise to at least five different proteins. The muscle isoform is the predominant form with lower amounts in cardiac muscle. In brain neurones are smaller amounts of an isoform that has a different first exon from that of the muscle protein and is produced by transcription using an alternative promoter in a tissue specific manner. There is also evidence of another isoform associated with the Purkinje cells of the cerebellum. In addition to these proteins of approximately 400 kDa the normal dystrophin gene also gives rise to smaller proteins. Rat Schwannoma and human glioma cells do not produce full length mRNA in detectable amounts but express a 4.8 kb mRNA that is transcribed from a cDNA derived from the 3' end of the dystrophin gene but with a unique 5' untranslated region absent from the 14 kb transcript. The Oxford group have named the gene product from the 4.8 kb mRNA transcript, apo-dystrophin-1, and that from a 6.5 kb transcript identified only in a rat C6 glioma cell line, apodystrophin-2. The transcript for the apo-dystrophin-1 was first discovered in liver by the Weizman group who consider it to be the major product of the dystrophin gene in non-muscle tissues. It is a protein of 71 kDa consisting of the two C-terminal domains of dystrophin and is localized in the membrane. The Weizman group have now studied the cDNA corresponding to the 71 kDa protein in some detail to determine the mechanism of its formation from the dystrophin gene and have 0142-4319 9
1992 Chapman & Hall
developed translation vectors to produce the protein in the quantities required for its study. Although the 71 kDa protein and apo-dystrophin-1 are apparently identical there is a discrepancy in the size of the mRNA reported by the two groups. Progress on the study of the 13 kb mRNA encoded on chromosome 6 was reported by the Oxford group. They suggest that the gene product originally known as dystrophin related protein should be named 'utrophin'. As an interim measure there are some advantages in giving the members of this group of proteins unique individual names. Nevertheless it might be wiser to leave the final decision on nomenclature until such time as we have more details of the complexity of the dystrophin group of proteins, the interrelations between the members and their functions. Most of the cDNA of utrophin has been cloned and the evidence available indicates that the gene probably spans about 2000 kb of chromosome 6 with the genomic organization similar to that of dystrophin. The C-terminal domain has 70% homology with dystrophin but there is only 20-30% in the spectrin domain. It is expressed in larger amounts of fetal tissue and has been located at the myoneural junction. In normal muscle utrophin is not present in the sarcolemma but appears to be up-regulated in Duchenne muscular dystrophy. The fact that all seven transcripts of the X chromosome and chromosome 6 located genes so far identified have considerable homology in the C-terminal domains, and possibly elsewhere, is an important consideration in selecting probes for clinical diagnosis. Nucleic acid and antibody probes must be absolutely specific for the regions of the gene, its transcript and product, that are unique to the isoform which is under investigation. It seems likely that this was not always the case in the past.
Structure and function of dystrophin The session on the cytoskeleton and the functional role of dystrophin was introduced by a general review from P. C. Marchisio of the interactions between the extracellular matrix and the cytoskeleton. At this stage it is not known whether dystrophin is associated with the adhesion plaques, podosomes, to which particular attention was paid in this talk. The evidence currently available indicates that the dystrophin molecule exists in a form that is about 120-30 nm long. In the course of summarizing the Japanese contributions to research on Xp21 myopathies S. Ebashi described electron microscope studies on the isolated protein from K. Maruyama's group
378 (Chiba, Japan) that suggested it could exist as an antiparallel tetramer of that length. A similar value for the molecular length in situ was reported by M. Cullen (from the Newcastle group) using gold conjugated antibodies to specified regions of the dystrophin molecule. By using antibodies directly labelled with gold it has been confirmed that the spectrin domain of dystrophin lies 15 nm from the membrane to which the C-terminal is attached. The Newcastle group also reported occasional staining of other membranes and fine filaments within the cell but the significance of these findings is not clear. Dystrophin could not be detected in the interculated discs of cardiac muscle nor in the attached plaques of smooth muscle. The pattern of association of dystrophin with the proteins of the cytoskeletal-membrane is slowly beginning to merge. S. V. Perry described proton NMR studies that have enabled two N-terminal sites of dystrophin involved in interaction with actin to be identified. The corresponding binding sites on actin have also been characterized and are probably the same as those that can bind a-actinin and fl-spectrin. Unlike the latter protein the spectrin domain of dystrophin does not contain an actin-binding site suggesting that this property is confined to the N-terminal region of dystrophin. Evidence for the binding by dystrophin of actin was also presented by G. Salviati who reported that talin, but not calcium, could bind to this protein. Membrane-bound dystrophin is a substrate for a variety of protein kinases but not for endogenous tyrosine kinase. The dystrophin gene is remarkable in the range and location of the deletions it exhibits in the Xp21 myopathies. Detailed studies of the deletions using cDNA probes and antibodies with specificities for defined regions of dystrophin can provide information about the properties of the gene and hint at those regions of the molecule of particular functional significance. A study reported by G. A. Danieli on data collected by diagnostic centres in Europe indicates that there were approximately three times as many deletions occuring distal to exon 40 than in the region proximal to this gene. There were suggestions from the data that preferential breakpoints might lie in introns 43, 44, 49, 50 and 53, with a bias to a particular breakpoint being apparent in different populations. Louise Nicholson directed attention to the observation that dystrophin can often be detected in Duchenne patients whose dystrophin genes exhibit flame shift deletions. These would be expected to lead to the synthesis of severely truncated proteins that are rapidly broken down in the muscle. Nevertheless 60% of the Duchenne patients exhibited clear dystrophin-positive fibres, although such fibres were usually less than 1% of the total. A phenomen referred to as 'reversion'. Using a monoclonal antibody to the spectrin domain the dystrophin content of biopsy muscle samples from patients with Xp21 myopathies was estimated to range 7-73% of the normal. These values were in good correlation with the severity of the clinical conditions. It was suggested
SYMPOSIUM REPORT that dystrophin synthesis could occur in Duchenne patients by 'exon skipping'. If indeed the Duchenne muscle cell has this capacity to make a slightly shortened dystrophin molecule with alleviation of the clinical condition the possibility of its exploitation should be examined.
Approach to therapy The current interest in myoblast transfer as a therapy in the Xp21 myopathies has focussed attention on muscle precursor cells in general. An additional stimulus to research into muscle regeneration and development is the remarkable fact that the mdx mouse after a period of impaired muscle function in early life is able to overcome the absence of dystrophin and function almost normally. This is widely considered to be due to the highly developed capacity of mouse muscle to regenerate. As myogenic cells isolated from embryonic, fetal and adult tissues express different phenotypes in vitro it is clearly important to decide which type of cell is most suitable for myoblast transfer. The possibility that the differential expression of the recently discovered myogenic regulatory genes might be responsible for the phenotypic differences observed between myoblasts was examined by G. Cossu. He showed that embryonic, fetal and adult (satellite) myoblasts expressed the myo D1 and myogenin gene products in similar amounts. In contrast primordial myoblasts in somites undergo terminal differentiation without expression of these gene products. No differential effect of growth factors on the proliferation of embryonic and fetal myoblasts could be demonstrated. In this context it is of interest that myotubes can be shown to release a factor in the medium that supports myoblast proliferation. In many ways the regenerative capacity of dy mouse muscle is more comparable to that of human Duchenne muscle than is that of the mdx mouse mutant. By using antibodies to embryonic myofibrillar protein isoforms S. Schiaffino demonstrated that the regenerative capacity after injury produced by cold and myotoxins was much reduced in the muscle of old dy dystrophic mice compared to that obtained with young animals. Similar effects of age on the regenerative capacity of muscle from Duchenne patients have been reported in the literature. Although it has been clearly demonstrated that myoblast transfer can lead to production of muscle fibres containing dystrophin in a muscle from which this protein is normally absent, there are still many problems in using this procedure as an effective therapy. T. Patridge from the Charing Cross Medical School (London, UK) discussed investigations aimed to make the procedure more efficient. After injection of myoblasts into mdx mice the dystrophin-positive myotubes tend to be localized around the injection site and often there is a patchy distribution of dystrophin in the membrane of the new fibres. Indeed it might be argued that for the myotube to be restored to normal function it requires an even
SYMPOSIUM REPORT distribution of dystrophin along the membrane. Most of the injected myoblasts fuse to form myotubes within the muscle bundle into which they are injected although occasionally migration into neighbouring muscle bundles could be detected. Pre-X-irradiation, which destroys the existing satellite cells, makes dystrophic muscle much more susceptible to colonization by injected normal myogenic cells. The problems of immune rejection would be overcome if autologous myoblasts could be used. This would require more knowledge than we currently possess to enable small biopsy samples to be grown up to large amounts of suitable myoblasts genetically modified to contain the normal dystrophin gene. Such an aim is a long way from achievement but progress was reported in the production from a transgenic mouse of a myogenic cell line that exhibited the required level of control of proliferation and differentiation. Stimulated by the work of Wolf et al. (Science 24, 1465-8 (1990)) the possibility of introducing dystrophin into deficient cells by the direct introduction of DNA has aroused much interest. The dystrophin cDNA is approaching the upper limits of size for this approach but nevertheless F. S. Walsh reported that fibroblasts and an mdx mouse myoblast cell line could be transfected and the dystrophin synthesized in these cells was localized in the membrane. Cells were transfected with a full-length construct of dystrophin cDNA supplied with a promotor and appropriate RNA processing signals. Take up of the construct was about 1% and so far there has not been much success in increasing efficiency. Transgenic mdx mice have been produced using the minigene which is a mutated form of the dystrophin gene found in a patient with a very mild form of Becker muscular dystrophy. The product of this minigene is a shortened form of dystrophin with the spectrin domain missing and a molecular weight of about 200 000. The muscle cells of these animals showed a mosaic distribution of the mutant dystrophin which in total corresponded to about 20% of that present in normal muscle. In those cells containing the mutant dystrophin it was located in the sarcolemma. The muscle of the transgenic mice showed reduced central nucleation and serum creatine phosphokinase compared to control mdx mice at comparable ages. The difficulties enountered in introducing substantial amounts of the dystrophin gene into the deficient mdx mouse muscle suggest that myoblast transfer is not yet a satisfactory therapy for the Xp21 myopathies. The results of the carefully controlled trial, probably the best that has so far been carried out anywhere, described by G. Karparti, reinforced this point. The effects of injecting myoblasts, prepared from the fathers' muscles, into biceps muscle were compared with the sham-injected contralateral control in eight immunosuppressed boys with Duchenne muscular dystrophy. The eficacy as ascertained by dystrophin production and isometric force production was poor and the results of doubtful
379 significance. It is of interest that no glycogen phosphorylase could be induced by myoblast transfer in two patients with McArdle's disease. Why is the dystrophin gene so large.'/ The first part of the final afternoon was devoted to a discussion on the implications of the rapidly expanding knowledge of the dystrophin group of proteins and the genes that control their synthesis. One striking feature of the genes is their large size with many introns, some of which are many kilobases long. It was pointed out by T. Monaco that this feature gave the potential of a diversity of gene products, evidence for which is rapidly accumulating. If these products have different functions the implication is that the effects of mutation of the dystrophin gene may not depend simply on the absence of one product (i.e. dystrophin itself). In this context clearly we have much to learn when we compare the effect of the absence of dystrophin on the Duchenne patients and the mdx mouse. Another consequence of the large number of exons may be to increase the possibility of 'exon-hopping'. In this way flame shift mutants could synthesize slightly shortened dystrophin molecules thus explaining the phenomenon of reversion. The fact that the intron sequences of dystrophin, which make up 99% of the gene, are strongly preserved was pointed out by D. Yaffe. This suggests that they possess a role and one can not help speculating whether these regions code for some important function or product. Prospects for the future In the closing session V. Dubowitz and G. Karparti discussed the perspectives for clinical applications. The results of clinical trials so far carried out on myoblast transfer have been disappointing. Nevertheless the fact that in less than a decade the gene responsible for the disease has been identified and procedures devised for its introduction into intact deficient muscle enabling clinical trials to be conducted, is indeed remarkable progress. One of the problems is that both myoblast transfer and somatic cell transfection are extremely inefficient processes as currently carried out and therefore not suitable for clinical treatment. It is likely that research in progress will increase the efficiency of these processes to acceptable levels. Also the other important problem of producing from small muscle samples large quantities of myoblasts capable of inducing myotube formation will also be solved as we learn more about the processes that control myoblast proliferation and differentiation. Nevertheless once efficient myoblast transfer has been achieved, the question of the immune response needs to be solved. In the case of the trials described by G. Karparti there were no problems in this respect for the boys were immunosuppressed with cyclophosphamide. Finally for complete cure the gene must be introduced into all
380 muscles in the human body of which there are over 300 located with varying degrees of accessibility. It should also be pointed out that dystrophin isoforms are present in cells other than muscle. Myoblast transfer will not deal with the problem of replacing these missing isoforms, the function of which we currently have no knowledge. In my opinion the immediate prospects for somatic gene transfer as a treatment for the Xp21 myopathies are very limited unless we get some dramatic breakthrough in this field. Genetic manipulation of the germline is a much more elegant solution to the problem. The work with transgenic mdx mice indicates that it can work although much more research is required in this area before there is any real prospect of applying it to humans. Social and ethical considerations, however, will determine whether this approach is acceptable. Understandably most of the research directed to the treatment of the Xp21 myopathies aims at developing a procedure to provide the deficient cell with dystrophin. The genetic and clinical evidence strongly suggests that this is the correct approach. There are, however, a number of indications that the Xp21 dystrophies may be the consequence of factors other than the simple absence of dystrophin. These include the dramatic difference in consequence of the absence of dystrophin in the human compared with the mouse, the association of the dystrophin level with the amount of the 156 kDa glycoprotein present in the membrane system (Ervasti et al., Nature 345, 315-9 (1990)) and the complexity of the dystrophin and dystrophin-related gene systems and their products. Few would dispute that the most effective treatment would be to replace the mutant form with the normal gene but despite the tremendous advances being made in the field this goal is a long way ahead. In the
SYMPOSIUM REPORT meantime it is important to increase efforts to understand the role of dystrophin and those components apparently functionally associated with it in the normal cell. This knowledge should enable the prospects for development of an effective pharmacological approach to be evaluated. For example from an understanding of the cellular mechanisms that enable the mdx mouse to dispense with dystrophin with little impairment of muscle cell function it might be possible to stimulate or initiate this process in the human. It has been known for some time that the properties of the muscle cell membrane are modified as a consequence of mutation of the dystrophin gene. Nevertheless although evidence is accumulating as to the relation of its protein products with the membrane system, the precise molecular associations have yet to be defined. A recent important step in this direction has been the demonstration that the 156 kDa glycoprotein may be involved in linking dystrophin to laminin (IbraghimovBeskrovnaya et al., Nature 355,696-702 (1992)). As our knowledge of the interactions between dystrophin and membrane and cytoskeletal components increases so do the prospects of developing an effective pharmacological intervention. Although such intervention would not cure the condition, hopefully it would alleviate the clinical consequences of the mutant gene until treatment involving its efficient replacement in somatic cells can be developed. S. V. Perry Department of Physiology Medical School University of Birmingham Birmingham B 15 2 TT UK
REPORT
Current status of research on the Xp21 myopathies The rapid advances in knowledge of the dystrophin gene and its function make it important to appraise from time-to-time the developments in the basic sciences for their clinical implications. To this end the Italian National Academy (Accademia Nazionale dei Lincei) and the British Council sponsored, with help from the Italian Muscular Dystrophy Association, a meeting on 20 and 21 February (1992) in Rome at the Palazzo Corsini. Although principally an Anglo-Italian occasion the programme was strengthened by invited papers from leading workers in the field from other countries.
Genes and their products The introductory talks by D. J. Blake presenting the work of the Oxford group and D. Yaffe from the group at the Weizman Institute (Rehovot, Israel) concentrated on the genes concerned in the synthesis of dystrophin and dystropbin related proteins. From the two genes coding for these proteins seven different gene products have so far been identified. The dystrophin gene located on the X chromosome encodes about 74 exons and gives rise to at least five different proteins. The muscle isoform is the predominant form with lower amounts in cardiac muscle. In brain neurones are smaller amounts of an isoform that has a different first exon from that of the muscle protein and is produced by transcription using an alternative promoter in a tissue specific manner. There is also evidence of another isoform associated with the Purkinje cells of the cerebellum. In addition to these proteins of approximately 400 kDa the normal dystrophin gene also gives rise to smaller proteins. Rat Schwannoma and human glioma cells do not produce full length mRNA in detectable amounts but express a 4.8 kb mRNA that is transcribed from a cDNA derived from the 3' end of the dystrophin gene but with a unique 5' untranslated region absent from the 14 kb transcript. The Oxford group have named the gene product from the 4.8 kb mRNA transcript, apo-dystrophin-1, and that from a 6.5 kb transcript identified only in a rat C6 glioma cell line, apodystrophin-2. The transcript for the apo-dystrophin-1 was first discovered in liver by the Weizman group who consider it to be the major product of the dystrophin gene in non-muscle tissues. It is a protein of 71 kDa consisting of the two C-terminal domains of dystrophin and is localized in the membrane. The Weizman group have now studied the cDNA corresponding to the 71 kDa protein in some detail to determine the mechanism of its formation from the dystrophin gene and have 0142-4319 9
1992 Chapman & Hall
developed translation vectors to produce the protein in the quantities required for its study. Although the 71 kDa protein and apo-dystrophin-1 are apparently identical there is a discrepancy in the size of the mRNA reported by the two groups. Progress on the study of the 13 kb mRNA encoded on chromosome 6 was reported by the Oxford group. They suggest that the gene product originally known as dystrophin related protein should be named 'utrophin'. As an interim measure there are some advantages in giving the members of this group of proteins unique individual names. Nevertheless it might be wiser to leave the final decision on nomenclature until such time as we have more details of the complexity of the dystrophin group of proteins, the interrelations between the members and their functions. Most of the cDNA of utrophin has been cloned and the evidence available indicates that the gene probably spans about 2000 kb of chromosome 6 with the genomic organization similar to that of dystrophin. The C-terminal domain has 70% homology with dystrophin but there is only 20-30% in the spectrin domain. It is expressed in larger amounts of fetal tissue and has been located at the myoneural junction. In normal muscle utrophin is not present in the sarcolemma but appears to be up-regulated in Duchenne muscular dystrophy. The fact that all seven transcripts of the X chromosome and chromosome 6 located genes so far identified have considerable homology in the C-terminal domains, and possibly elsewhere, is an important consideration in selecting probes for clinical diagnosis. Nucleic acid and antibody probes must be absolutely specific for the regions of the gene, its transcript and product, that are unique to the isoform which is under investigation. It seems likely that this was not always the case in the past.
Structure and function of dystrophin The session on the cytoskeleton and the functional role of dystrophin was introduced by a general review from P. C. Marchisio of the interactions between the extracellular matrix and the cytoskeleton. At this stage it is not known whether dystrophin is associated with the adhesion plaques, podosomes, to which particular attention was paid in this talk. The evidence currently available indicates that the dystrophin molecule exists in a form that is about 120-30 nm long. In the course of summarizing the Japanese contributions to research on Xp21 myopathies S. Ebashi described electron microscope studies on the isolated protein from K. Maruyama's group
378 (Chiba, Japan) that suggested it could exist as an antiparallel tetramer of that length. A similar value for the molecular length in situ was reported by M. Cullen (from the Newcastle group) using gold conjugated antibodies to specified regions of the dystrophin molecule. By using antibodies directly labelled with gold it has been confirmed that the spectrin domain of dystrophin lies 15 nm from the membrane to which the C-terminal is attached. The Newcastle group also reported occasional staining of other membranes and fine filaments within the cell but the significance of these findings is not clear. Dystrophin could not be detected in the interculated discs of cardiac muscle nor in the attached plaques of smooth muscle. The pattern of association of dystrophin with the proteins of the cytoskeletal-membrane is slowly beginning to merge. S. V. Perry described proton NMR studies that have enabled two N-terminal sites of dystrophin involved in interaction with actin to be identified. The corresponding binding sites on actin have also been characterized and are probably the same as those that can bind a-actinin and fl-spectrin. Unlike the latter protein the spectrin domain of dystrophin does not contain an actin-binding site suggesting that this property is confined to the N-terminal region of dystrophin. Evidence for the binding by dystrophin of actin was also presented by G. Salviati who reported that talin, but not calcium, could bind to this protein. Membrane-bound dystrophin is a substrate for a variety of protein kinases but not for endogenous tyrosine kinase. The dystrophin gene is remarkable in the range and location of the deletions it exhibits in the Xp21 myopathies. Detailed studies of the deletions using cDNA probes and antibodies with specificities for defined regions of dystrophin can provide information about the properties of the gene and hint at those regions of the molecule of particular functional significance. A study reported by G. A. Danieli on data collected by diagnostic centres in Europe indicates that there were approximately three times as many deletions occuring distal to exon 40 than in the region proximal to this gene. There were suggestions from the data that preferential breakpoints might lie in introns 43, 44, 49, 50 and 53, with a bias to a particular breakpoint being apparent in different populations. Louise Nicholson directed attention to the observation that dystrophin can often be detected in Duchenne patients whose dystrophin genes exhibit flame shift deletions. These would be expected to lead to the synthesis of severely truncated proteins that are rapidly broken down in the muscle. Nevertheless 60% of the Duchenne patients exhibited clear dystrophin-positive fibres, although such fibres were usually less than 1% of the total. A phenomen referred to as 'reversion'. Using a monoclonal antibody to the spectrin domain the dystrophin content of biopsy muscle samples from patients with Xp21 myopathies was estimated to range 7-73% of the normal. These values were in good correlation with the severity of the clinical conditions. It was suggested
SYMPOSIUM REPORT that dystrophin synthesis could occur in Duchenne patients by 'exon skipping'. If indeed the Duchenne muscle cell has this capacity to make a slightly shortened dystrophin molecule with alleviation of the clinical condition the possibility of its exploitation should be examined.
Approach to therapy The current interest in myoblast transfer as a therapy in the Xp21 myopathies has focussed attention on muscle precursor cells in general. An additional stimulus to research into muscle regeneration and development is the remarkable fact that the mdx mouse after a period of impaired muscle function in early life is able to overcome the absence of dystrophin and function almost normally. This is widely considered to be due to the highly developed capacity of mouse muscle to regenerate. As myogenic cells isolated from embryonic, fetal and adult tissues express different phenotypes in vitro it is clearly important to decide which type of cell is most suitable for myoblast transfer. The possibility that the differential expression of the recently discovered myogenic regulatory genes might be responsible for the phenotypic differences observed between myoblasts was examined by G. Cossu. He showed that embryonic, fetal and adult (satellite) myoblasts expressed the myo D1 and myogenin gene products in similar amounts. In contrast primordial myoblasts in somites undergo terminal differentiation without expression of these gene products. No differential effect of growth factors on the proliferation of embryonic and fetal myoblasts could be demonstrated. In this context it is of interest that myotubes can be shown to release a factor in the medium that supports myoblast proliferation. In many ways the regenerative capacity of dy mouse muscle is more comparable to that of human Duchenne muscle than is that of the mdx mouse mutant. By using antibodies to embryonic myofibrillar protein isoforms S. Schiaffino demonstrated that the regenerative capacity after injury produced by cold and myotoxins was much reduced in the muscle of old dy dystrophic mice compared to that obtained with young animals. Similar effects of age on the regenerative capacity of muscle from Duchenne patients have been reported in the literature. Although it has been clearly demonstrated that myoblast transfer can lead to production of muscle fibres containing dystrophin in a muscle from which this protein is normally absent, there are still many problems in using this procedure as an effective therapy. T. Patridge from the Charing Cross Medical School (London, UK) discussed investigations aimed to make the procedure more efficient. After injection of myoblasts into mdx mice the dystrophin-positive myotubes tend to be localized around the injection site and often there is a patchy distribution of dystrophin in the membrane of the new fibres. Indeed it might be argued that for the myotube to be restored to normal function it requires an even
SYMPOSIUM REPORT distribution of dystrophin along the membrane. Most of the injected myoblasts fuse to form myotubes within the muscle bundle into which they are injected although occasionally migration into neighbouring muscle bundles could be detected. Pre-X-irradiation, which destroys the existing satellite cells, makes dystrophic muscle much more susceptible to colonization by injected normal myogenic cells. The problems of immune rejection would be overcome if autologous myoblasts could be used. This would require more knowledge than we currently possess to enable small biopsy samples to be grown up to large amounts of suitable myoblasts genetically modified to contain the normal dystrophin gene. Such an aim is a long way from achievement but progress was reported in the production from a transgenic mouse of a myogenic cell line that exhibited the required level of control of proliferation and differentiation. Stimulated by the work of Wolf et al. (Science 24, 1465-8 (1990)) the possibility of introducing dystrophin into deficient cells by the direct introduction of DNA has aroused much interest. The dystrophin cDNA is approaching the upper limits of size for this approach but nevertheless F. S. Walsh reported that fibroblasts and an mdx mouse myoblast cell line could be transfected and the dystrophin synthesized in these cells was localized in the membrane. Cells were transfected with a full-length construct of dystrophin cDNA supplied with a promotor and appropriate RNA processing signals. Take up of the construct was about 1% and so far there has not been much success in increasing efficiency. Transgenic mdx mice have been produced using the minigene which is a mutated form of the dystrophin gene found in a patient with a very mild form of Becker muscular dystrophy. The product of this minigene is a shortened form of dystrophin with the spectrin domain missing and a molecular weight of about 200 000. The muscle cells of these animals showed a mosaic distribution of the mutant dystrophin which in total corresponded to about 20% of that present in normal muscle. In those cells containing the mutant dystrophin it was located in the sarcolemma. The muscle of the transgenic mice showed reduced central nucleation and serum creatine phosphokinase compared to control mdx mice at comparable ages. The difficulties enountered in introducing substantial amounts of the dystrophin gene into the deficient mdx mouse muscle suggest that myoblast transfer is not yet a satisfactory therapy for the Xp21 myopathies. The results of the carefully controlled trial, probably the best that has so far been carried out anywhere, described by G. Karparti, reinforced this point. The effects of injecting myoblasts, prepared from the fathers' muscles, into biceps muscle were compared with the sham-injected contralateral control in eight immunosuppressed boys with Duchenne muscular dystrophy. The eficacy as ascertained by dystrophin production and isometric force production was poor and the results of doubtful
379 significance. It is of interest that no glycogen phosphorylase could be induced by myoblast transfer in two patients with McArdle's disease. Why is the dystrophin gene so large.'/ The first part of the final afternoon was devoted to a discussion on the implications of the rapidly expanding knowledge of the dystrophin group of proteins and the genes that control their synthesis. One striking feature of the genes is their large size with many introns, some of which are many kilobases long. It was pointed out by T. Monaco that this feature gave the potential of a diversity of gene products, evidence for which is rapidly accumulating. If these products have different functions the implication is that the effects of mutation of the dystrophin gene may not depend simply on the absence of one product (i.e. dystrophin itself). In this context clearly we have much to learn when we compare the effect of the absence of dystrophin on the Duchenne patients and the mdx mouse. Another consequence of the large number of exons may be to increase the possibility of 'exon-hopping'. In this way flame shift mutants could synthesize slightly shortened dystrophin molecules thus explaining the phenomenon of reversion. The fact that the intron sequences of dystrophin, which make up 99% of the gene, are strongly preserved was pointed out by D. Yaffe. This suggests that they possess a role and one can not help speculating whether these regions code for some important function or product. Prospects for the future In the closing session V. Dubowitz and G. Karparti discussed the perspectives for clinical applications. The results of clinical trials so far carried out on myoblast transfer have been disappointing. Nevertheless the fact that in less than a decade the gene responsible for the disease has been identified and procedures devised for its introduction into intact deficient muscle enabling clinical trials to be conducted, is indeed remarkable progress. One of the problems is that both myoblast transfer and somatic cell transfection are extremely inefficient processes as currently carried out and therefore not suitable for clinical treatment. It is likely that research in progress will increase the efficiency of these processes to acceptable levels. Also the other important problem of producing from small muscle samples large quantities of myoblasts capable of inducing myotube formation will also be solved as we learn more about the processes that control myoblast proliferation and differentiation. Nevertheless once efficient myoblast transfer has been achieved, the question of the immune response needs to be solved. In the case of the trials described by G. Karparti there were no problems in this respect for the boys were immunosuppressed with cyclophosphamide. Finally for complete cure the gene must be introduced into all
380 muscles in the human body of which there are over 300 located with varying degrees of accessibility. It should also be pointed out that dystrophin isoforms are present in cells other than muscle. Myoblast transfer will not deal with the problem of replacing these missing isoforms, the function of which we currently have no knowledge. In my opinion the immediate prospects for somatic gene transfer as a treatment for the Xp21 myopathies are very limited unless we get some dramatic breakthrough in this field. Genetic manipulation of the germline is a much more elegant solution to the problem. The work with transgenic mdx mice indicates that it can work although much more research is required in this area before there is any real prospect of applying it to humans. Social and ethical considerations, however, will determine whether this approach is acceptable. Understandably most of the research directed to the treatment of the Xp21 myopathies aims at developing a procedure to provide the deficient cell with dystrophin. The genetic and clinical evidence strongly suggests that this is the correct approach. There are, however, a number of indications that the Xp21 dystrophies may be the consequence of factors other than the simple absence of dystrophin. These include the dramatic difference in consequence of the absence of dystrophin in the human compared with the mouse, the association of the dystrophin level with the amount of the 156 kDa glycoprotein present in the membrane system (Ervasti et al., Nature 345, 315-9 (1990)) and the complexity of the dystrophin and dystrophin-related gene systems and their products. Few would dispute that the most effective treatment would be to replace the mutant form with the normal gene but despite the tremendous advances being made in the field this goal is a long way ahead. In the
SYMPOSIUM REPORT meantime it is important to increase efforts to understand the role of dystrophin and those components apparently functionally associated with it in the normal cell. This knowledge should enable the prospects for development of an effective pharmacological approach to be evaluated. For example from an understanding of the cellular mechanisms that enable the mdx mouse to dispense with dystrophin with little impairment of muscle cell function it might be possible to stimulate or initiate this process in the human. It has been known for some time that the properties of the muscle cell membrane are modified as a consequence of mutation of the dystrophin gene. Nevertheless although evidence is accumulating as to the relation of its protein products with the membrane system, the precise molecular associations have yet to be defined. A recent important step in this direction has been the demonstration that the 156 kDa glycoprotein may be involved in linking dystrophin to laminin (IbraghimovBeskrovnaya et al., Nature 355,696-702 (1992)). As our knowledge of the interactions between dystrophin and membrane and cytoskeletal components increases so do the prospects of developing an effective pharmacological intervention. Although such intervention would not cure the condition, hopefully it would alleviate the clinical consequences of the mutant gene until treatment involving its efficient replacement in somatic cells can be developed. S. V. Perry Department of Physiology Medical School University of Birmingham Birmingham B 15 2 TT UK
