It is reasonable to believe that the transfer, the transportation, of sufficient quantities of the intact dystrophin gene into the nuclei of dystrophic muscle cells would cure the disease if the genetic information of the new genes is used by the protein synthesizing ribosomes of the cell to produce sufficiently large quantities of functional dystrophin which then migrates to its normal place under the cell membrane where it is correctly incorporated into the intricate dystrophin-glycoprotein complex of the costameres.

   The dystrophic mouse: Most of the experiments with this aim have been performed with one kind of laboratory animal, the dystrophic mdx mouse, which has a point mutation at base pair 3,185 in exon 23 of its dystrophin gene. This mutation has changed a CAA codon, which signifies the amino acid glutamine, to a TAA codon, which is a stop sign, so that the synthesis of dystrophin is interrupted prematurely. Thus, the mouse has no functional dystrophin in its muscles. However, these mice do not lose their muscles because they do not develop fibrosis, a proliferation of connective tissue, like Duchenne boys do, so that the degeneration caused by the disease does not overtake the regeneration.
   Although gene transfer can be studied with them, one should keep in mind that any results obtained with these mice cannot be regarded as immediately applicable to children as long as they are not confirmed by clinical studies with Duchenne patients. A child is not a big mouse!

   The dystrophic dog: Some experiments are performed with dystrophic dogs, the golden retriever muscular dystrophy (GRMD) dogs which, in contrast to the mice, have a muscular dystrophy similar to the dystrophy of Duchenne boys. They are really handicapped and hard to raise and to manage. Their dystrophin gene has a point mutation in the splice receptor site of intron 6 which leads to the deletion of exon 7 in the mRNA, to a frame shift, and a premature stop codon.

   Gene transfer with adeno viruses: In order to transfer a gene, a transporter, a gene vector, is needed. One way to transfer genetic material into living cells is to pack it into viruses, which consist of a string of their own genes enveloped in a protein shell. They attach themselves to special receptor proteins on the surface of a cell, inject their genes through the cell membrane and then use the synthesizing apparatus of the infected cell to reproduce themselves.
   Mainly two kinds of viruses are used as gene vectors in Duchenne research, the adeno virus and the ten times smaller adeno-associated virus. The vectors derived from them cannot be further multiplied inside the target cell because almost all of their own genes have been eliminated. The most advantageous vector seems to be the gutted, practically empty adenovirus, which does not contain any genes of its own, and thus has room for up to 36,000 foreign genetic letters, sufficient for the entire cDNA, i.e., all exons with the information for the complete dystrophin protein and for additional control sequences. With these gutted viruses new dystrophin was produced in up to 30 % of the muscle fibers of a dystrophic mouse followed by an improvement of muscle function in young as well as in older mice (Lochmüller, Munich; Chamberlain, Seattle).

   Two genes in one virus: In new experiments two mouse dystrophin cDNAs were packed into each gutted virus, that means, two times all 79 exons without the introns, and in addition two more gene sequences, the very strong enhancer of the cytomegalovirus and the beta-actin promoter. These gene vectors were injected into the tibialis anterior muscle of newborn mdx mice and into 4-6 weeks old juvenile mice. After 30 days, 42 % of the cells in the treated muscle of the newborn mice had new dystrophin which was still detectable after half a year. In the juvenile mice, 24 % of the muscle cells had new dystrophin after 30 days, which, however, was reduced to half the amount after 6 months. Only the juvenile mice showed small immune problems. At present, these new vector constructions are the most effective ones for the treatment of Duchenne muscular dystrophy by gene therapy (Karpati, Montreal).

   Gene transfer with adeno-associated viruses: These small viruses can only transport genetic material that is not longer than about 5,000 base pairs, about one third of the entire dystrophin cDNA. Their advantage is that they transfer the gene more effectively than the normal adenoviruses. The disadvantage is that the dystrophin cDNA to be transferred has to be shortened considerably to fit into this small vector. Patients with Becker muscular dystrophy, that progresses much slower than Duchenne dystrophy, mostly have such shortened dystrophin in their muscles. Therefore, a transfer of one of these Becker mini genes would not completely cure Duchenne muscular dystrophy but only transform it into the benign Becker form.
   In order to determine which of the four regions of the normal dystrophin protein - the two end regions, the cystein-rich, or the central rod regions - are important and which are not, extensive experiments were performed with dystrophic mice which were genetically modified so that they had one of nine different shortened dystrophins in their muscles.
   It had already been known that some of the rodlike central region can be removed without loss of function. The detailed analysis of the three-dimensional structure revealed now which particular parts of the rod structure could be deleted without significantly losing the muscle protecting function. One of the shortened dystrophins, which was about half as long as the normal one, had practically the same properties as the unaltered protein. Delivery of the genes of this and some of the other shortened dystrophins with adeno associated viruses into the muscles of the mdx mice prevented and partially reversed the dystrophic symptoms. These results demonstrated that specific modifications of the dystrophin gene can generate novel proteins that are significantly smaller but more functional than the naturally shortened dystrophins of patients with Becker muscular dystrophy (Chamberlain, Seattle).
   If about 30 % of the normal amount of unchanged dystrophin re-appears, measurements on the diaphragm of the mice indicated that the muscle function also improved. The transfer of the mini gene also led to an improvement of muscle function. In addition, the gene transfer showed better results in younger animals, because there were fewer problems of immune rejection, and, after a single injection, the newly synthesized dystrophin remained in the muscles longer than in adult animals. After a single treatment of newborn mice, newly formed dystrophin could still be found after one year.
   For a future application in humans, this means that the patients would have to be treated as soon as possible after birth when their muscles are still largely intact (Clemens, Pittsburgh).

   Gene transfer via the blood stream: In the experiments so far described, solutions of the viruses that contain the dystrophin gene were injected directly into the muscles. In order to reach all muscles, also those of the heart and the lungs, experiments are being performed to develop a systemic treatment, i.e. the injection of the viruses into the blood stream. To ensure that the new dystrophin is only synthesized in muscle cells, another genetic sequence had to be added to the dystrophin gene in the virus, the CK promoter, which normally activates the gene of the muscle protein creatine kinase. This promoter would assure that the new dystrophin gene is activated only in muscle cells and not elsewhere.
   To investigate whether such strategies would function in a living animal, transgenic mdx mice were created which have an additional gene construction inserted in their chromosomes. This consisted of an often used Becker gene which, with 6,300 base pairs, is only half as long as the normal gene, and which was preceded by a CK promoter of 1,354 base pairs. Shortened dystrophin was then present during the entire life time of the mice of up to 24 months in sufficient quantities to improve all measurable clinical symptoms.
   The new protein appeared only in the muscles of the transgenic mice, significantly more in the fast muscles than in the slow muscles. The fast muscles, which can work immediately but not consistently, obtain their energy through glycolysis, through the fast degradation of glucose without the use of oxygen. The energy for the slow and consistently working muscles is provided by the enzymes of the respiratory chain, the slow "burning" of organic substances with oxygen. As the fast muscles are destroyed first by the muscular dystrophy, the predominant appearance of new dystrophin in these fast muscles is able to ameliorate just this early consequence of the disease. Experiments have also shown that as few as 20 to 30 % of the normal quantity of dystrophin have a significant therapeutic effect.
   For these experiments with transgenic mice, the Becker dystrophin gene together with the CK promoter was introduced into the germ cells by artificial insemination and genetic manipulation, so that it was inherited to the progeny of the mice. This is a technique which obviously cannot be performed with humans. To cure the sick children, one would have to transport the gene with its promoter by gene transfer with a vector soon after birth, if possible, into all muscles (Lochmüller, Munich).

   Amplification of targeted gene transfer: The viruses attach themselves to specific structures on the cell membrane, to the coxsackie and adenovirus receptors, from where they penetrate the muscle cell through the membrane. These receptors, however, are more numerous on developing and regenerating muscle cells. They are downregulated, less numerous, on mature muscle cells which no longer divide. Therefore, gene transfer with adenovirus vectors is more efficient in dividing muscle cells.
   To overcome this handicap, transgenic mice were produced which expressed these receptors in large amounts on the surface of mature muscle fibers. This led to an up to 10 times more efficient transfer of the dystrophin gene (Holland, Montreal; Lochmüller, Munich).
   As a multiplication of the of the virus receptors on the muscle cells of a patient could only be achieved by additional gene technical methods with all their risks, helper proteins were produced. In this case, they were monoclonal antibodies which consisted of only one kind of molecule that could bind with one end highly specifically on other, more numerous receptors. With their other end, these immune proteins could bind to modified coat proteins of the adeno viruses. To achieve this binding, additional genetic information for a chain of 33 amino acids from a bacterium was introduced into the adeno viruses by genetic manipulation.
   These adeno viruses could now attach themselves via the antibody bridge to integrin and nerve receptors (NCAM) which were much more numerous than the adeno receptors. This method allowed up to 77 times more adeno viruses to pass through the membrane into the muscle cell than before. With this new technique, which can also be used against other diseases, only the gene was transported which produces the easily detectable protein beta-galactosidase. Experiments to transfer the dystrophin gene are being prepared (Kochanek, Cologne).

   Transfer of naked genes: For this technique, the genetic material DNA to be transferred is not built into viruses but into plasmids, small circular DNA structures without protein, that exist in bacteria where they mostly give rise to resistance against antibiotics. The advantage of this kind of gene transfer is that the plasmids do not contain any protein, only genetic material, naked DNA, so that no immune reaction develops against the vector material.
   Experiments were made with plasmids containing marker or reporter genes for proteins that can be detected in the muscle tissue after staining or by light production so that the success of a transfer experiment can easily be monitored.
   When relatively large volumes of a solution of these plasmids were injected under pressure into the arteries of the limbs of rats and rhesus monkeys, the reporter genes beta-galactosidase and luciferase were transferred into up to 20 % of the muscle fibers after one single injection and into up to 40 % after repeated injections. The pressure was created by blocking the venous outflow from the limb for a short time. Experiments to transfer the dystrophin gene are currently being performed in mdx mice, monkeys, and GRMD dogs (Wolff, Madison; Braun, Strasbourg). In France, clinical studies with Duchenne boys have been started which are described in the section "Clinical studies".

   Experiments to overcome immunological problems: As plasmids contain only DNA but not any proteins, gene transfer with such naked DNA allows the examination of potential problems with immune responses to newly created dystrophin without the interference of other proteins which are newly produced with the other methods of gene transfer using viral and cell vectors.
   If the human dystrophin gene is introduced into the muscles of mdx mice, an immune response develops against the foreign dystrophin. This does not happen after the transfer of the dystrophin gene of mice, although normal dystrophin is not present in the muscles of these dystrophic mice. The lack of an immune response to mouse dystrophin may be due to the presence of the other forms of dystrophin. These results suggest that immune responses to dystrophin gene transfer may not be a particular problem in Duchenne patients, especially in those with point mutations that do not interfere with the synthesis of the other forms of dystrophin. Current experimental work is focussed on determining which of the other dystrophins must be present to allow immune tolerance to the appearance of new muscle type dystrophin after gene transfer (Wells, London).

   Stem cells from skeletal muscles: On the surface of the muscle cells are satellite cells, also called myogenic cells or myoblasts which, after an activation through several intermediate steps, can form new muscles or repair injured muscles. To determine whether such "back-up" cells are entirely or partly stem cells, satellite cells from the skeletal muscles of newborn mice were isolated. It was found that they consist of three different kinds, mainly the so-called EP and LP myogenic cells, which are already quite specialized. The third kind are muscle derived stem cells, MDSC. They are very rare, only one among 100,000 satellite cells is one such MDS cell. But these rare cells are pluripotent, that is, they can develop into muscle cells as well as into cells of the nerves and of the blood vessels for new muscles. And they can be cultured for a long time.
   These MDS cells were multiplied in the laboratory, and then 400,000 of them in 25 microliters (a fortieth of a cubic centimeter) of liquid injected in one single portion into one muscle of living mdx mice. In order to see in what way these cells are different from the normal myogenic cells, the more specialized EP cells were also injected into mdx mice under the same conditions.
   In the mice which were treated with the EP cells, 130 muscle cells at the injection site contained new dystrophin after 30 days, which however, after 90 days, had largely disappeared. In these mice, immune rejection by lymphocites was observed which was probably responsible for the slow disappearance of the new dystrophin.
   In contrast to these results, 1,500 muscle cells at the injection site of the mice treated with MDS cells contained new dystrophin, about 10 times as many as after the injection of EP cells. Even after 90 days, in practically all of these cells, the new dystrophin was still present. In this case, there was no immune rejection although the receptor mice had not received any immune suppressive drugs.
   As the muscle stem cells used in these experiments were isolated from newborn normal mice, this technique will probably not be applicable without modification with Duchenne children. They could, however, influence the recently resumed myoblast studies (Huard, Pittsburgh; Wernig, Bonn).

   Stem cells from bone marrow, muscles and skin: At the end of the last decade, the first experiments with stem cells from bone marrow and with muscle satellite cells were performed. In all experiments, the stem cells were obtained from normal male mice with normal dystrophin genes and then injected into the tail vein of female mdx mice. The female mice were homozygous, that is, the dystrophin genes on both of their X chromosomes were mutated so that they could not produce any own dystrophin. The fate of the injected cells in the female mice could be followed by the detection of the male Y chromosome. The female mice had to be lethally irradiated with X rays to avoid an immune reaction. The transfer of the stem cells regenerated the bone marrow of the mice.
   First, 10 to 50 million cells from the untreated bone marrow were injected. Three months later, dystrophin could be detected in up to 10 % of the muscle cells of the mice. Some of the dystrophin-positive cells contained Y chromosomes as proof that the bone marrow cells from the normal male mice came via the blood stream and had fused with the muscle cells of the mdx mice. They had brought with them the information for the normal dystrophin which was then used for the production of functional dystrophin. It was important to prove that the new dystrophin did not appear by reversion, that is, by spontaneous exon skipping which occurs to a small extent in mdx mice and also in Duchenne boys.
   To find out which part of the bone marrow cells has the stem cell properties, a small side population, SP, could be separated from the original cell mixture by the FACS method, fluorescence-activated cell sorting. As these SP cells had stem cell properties, they were used for further experiments.
   Three months after the injection of only 2,000 to 5,000 of these SP bone marrow cells, up to 4 % of the muscle cells of the mdx mice had functional dystrophin. In a similar way, SP muscle stem cells were isolated from muscle satellite cells, and 7,000 to 20,000 of them were injected into the blood stream of female mdx mice. After one month, 5 to 9 % of their muscle cells contained normal dystrophin and some also had Y chromosomes.
   Now, in 2002/2003, these experiments were repeated and extended to stem cells which were obtained from skin. The advantages of an isolation from skin are that this tissue is more easily accessible than muscle tissue, and that the acceptor animals no longer have to be irradiated.
   A small side population of cells was also obtained with the FACS technique from skin cells which had the same surface properties, markers, as the SP muscle stem cells. The percentages of SP cells were now 0.1 % from bone marrow cells, 0.7 % from muscle cells, and 1.2 % from skin cells. Three months after the injection of 6.000 to 50.000 of the SP skin cells, 2.3 % of the muscle cells of the female mice contained new dystrophin and also some Y chromosomes.
   This experiment had proved that stem cells can be isolated from skin, and that, after application via the blood stream, they can form new muscle cells with normal dystrophin. This technique is not yet sufficiently effective to be of therapeutic significance. But after optimization, a Duchenne therapy might possibly be based on it (Kunkel, Boston).

   Activation of muscle stem cells: Stem cells with specific surface structures which are located in intact muscles, did not develop into new muscle cells, they were not myogenic. But if the muscles were injured and the muscle cells regenerated themselves, then the number of muscle stem cells increased up to ten times. These cells were now myogenic, they developed into myoblasts and new muscle cells, as experiments with cell cultures and mice have shown. This activation was initiated by Wnt proteins, a family of signal proteins which apparently are produced by injured muscle cells and which also play a role during embryonic development (Rudnicki, Ottawa).
   Some Wnt proteins have been isolated and characterized now. They are about 400 amino acids long and contain palmitic acid, a fatty acid which seems to be important for the transmittance of molecular signals (Nusse, Stanford).

   Mesoangioblasts, stem cells from blood vessels: One half million of newly detected stem cells from blood vessels of fetal normal mice were delivered by one single injection into the artery of one hind leg of mice. These mice were missing one protein of the dystrophin complex, the alpha sarcoglycan which causes one of the many limb-girdle muscular dystrophies in humans. The injected cells migrated into the blood capillaries, passed through their cell walls, and from there into all muscles of the injected leg, especially into regenerating fibers.
   For at least three months after the injection, new sarcoglycan was again present in almost normal quantities. The same was true for many of the other proteins of the dystrophin complexes which had also been missing. Also, there were no immune reactions. After three consecutive injections every 40 days, not only was the gene defect almost completely corrected, but also the muscle force in the treated leg was practically normalized again. Here as well, no immune problems were encountered.
   In order to check whether this technique could possibly be used for a gene therapy, the mesoangioblasts from "sick" mice were isolated, and then the missing gene for alpha sarcoglycan was transferred with retroviruses into these defective stem cells. The systemic application of these ex-vivo treated cells led to the same positive results as found with the normal cells. However, retroviruses which insert themselves with the transported therapeutic gene into the chromosomes, should not be used in humans because of their risk of causing cancer.
   For these experiments, the blood-vessel derived stem cells were isolated from unborn mice. But if it were also possible to isolate them from boys with Duchenne muscular dystrophy, these unexpected positive results would mean that a number of problems connected with the present gene transfer experiments could be avoided as, e.g., low effectivity, immune rejection, and the requirement of many injections into all muscles which can be reached.
   Intact dystrophin genes would have to be transferred into these patient-derived cells by an ex-vivo procedure with the known vectors, then multiplied in the laboratory, and finally re-injected into the most important arteries of the child. Possibly, the treatment would have to be repeated after several months, therefore it is important that these cells are not rejected by the immune system. And one of the most important advantages would be that, with this technique, all muscles could be reached, also the cardiac and respiratory muscles. It would be a systemic treatment with relatively few injections compared to other gene therapy approaches (Cossu, Milano).

   A stem cell experiment of nature: At the age of one year, a Duchenne boy had received a bone marrow transplantation from his father because of another disease. As he could still walk at 14 years in spite of the reading-frame-shifting deletion of exon 45, it was assumed that stem cells from the bone marrow had provided the information for new dystrophin. Investigation of new bioptic material, however, have now shown that the milder symptoms of the disease are due only to a small extent to the transplant, but mainly to a spontaneous additional deletion of exon 44, so that, in the mRNA, exon 46 follows directly after exon 43. This exon skipping normalizes the reading frame. Dystrophin can again be produced but is shorter and therefore causes the symptoms of a Becker dystrophy (see paragraph on exon skipping). Apparently, after 13 years, a bone marrow transplantation can contribute to some extent to a Duchenne therapy, however not sufficiently so to significantly change the disease (Kunkel, Boston).

   Cell therapy with myoblasts: Muscles have their own stem cells, the satellite cells or myoblasts, which, during the development or repair of muscles, fuse together to form myotubes and then long muscle fibers. As the word "myoblast transfer" has been misused, one now often prefers to say myogenic cells instead of myoblasts.
   In the years 1990 and 1991, extensive clinical studies with Duchenne boys were performed with myoblasts. This myoblast transfer technique had shown positive results in mdx-mice. The cells used contained the normal dystrophin gene because they were derived from a healthy donor, mostly from the father of the patient. They were applied in multiple injections at 0.5 centimeters distance into some muscles of Duchenne boys with the expectation that muscle cells with normal dystrophin would form. However, these experiments were not successful, because the transplanted cells did not migrate sufficiently inside the muscle, because there were immunological problems, and because almost all of the injected myogenic cells had died after a short time (Karpati, Montreal, and others).

   New experiments to transplant myoblasts: But work on this technique continues in order to determine the reason why far less than 1 % of the transplanted myoblasts survived in the dystrophic muscles. Researchers are now trying to characterize these rare active cells and to find out how they can be isolated from the inactive cells. They are therefore looking for molecular signals, special substances in the muscle cells, which could activate the myoblasts (Partridge, London).
   In earlier experiments, only two known drugs, cyclosporin A and cyclophosphamide were used to suppress the immune reactions in Duchenne patients. Now, other substances have been investigated in studies with monkeys. This has shown that the immune inhibitor FK506 alone or in combination with the inhibitor MMF avoids the rejection problems much better for several months. A greater application density with injection distances of only 1 millimeter and a higher number of transplanted cells contributed to the fact that in monkeys, up to 67 % of the muscle cells had taken up the myoblasts, they became hybrid muscle cells (Tremblay, Québec City).
   A clinical study with the modified technique has been started in Canada (see section "Clinical trials). Also in other laboratories, work is being done to improve this cell therapy technique. E.g., it was found that m144, a protein of the immune system, avoids the immediate death of the myoblasts after transplantation into mouse muscles (Hodgetts, Crawley, Australia).

   Experiments for an ex-vivo gene therapy: To avoid immune problems completely, experiments were performed to isolate the myoblasts from the patient himself and then, in cell culture, to transfer an intact dystrophin gene into the cells, before they are re-injected again. In preliminary experiments an electroperforation technique was used to transfer the gene for a fluorescent marker protein into myoblasts in cell culture. This technique transiently permeabilized the membranes with a single electrical pulse, at e.g. 400 volts across a distance of 4 mm. Under optimal conditions, the marker gene could be transferred into up to 70 % of the myoblasts where it produced the fluorescent protein. The cells maintained their ability to fuse into myotubes, the next stage of muscle development (Bernheim, Geneva).

   Repairing the gene in the GRMD dog: Attempts to repair point mutations at the gene level are made by using short chimeric oligonucleotides, which contain RNA on one strand and DNA on the other. The DNA strand of the chimeric oligonucleotide is perfectly complementary to the correct gene sequence at the point mutation site of the gene, while the RNA portion is perfectly complementary to the mutant sequence. This leads to paired quadruplex structures, fourfold strands, which are capable of correcting such a small mutation by activating the biological DNA repair mechanisms of the cell.
   With this technique, a DNA and RNA oligonucleotide complementary to the splice-site mutation in the dystrophin gene of the dystrophic GRMD dog was injected into a 6-week old affected dog. The treated muscle showed: (1) evidence of restoration of the exon which was missing because of the mutation, (2) restoration of the protein region encoded by the missing exon within the full-length dystrophin protein which was found localized correctly to the muscle membrane, and (3), most importantly, demonstration that the gene on the X chromosome was corrected. The repair of the mutation was sustained for almost a year in this dog (Bartlett, Bethesda).

   Gene repair in the mdx mouse: In a similar experiment, the point mutation in exon 23 of myoblasts from mdx mice was repaired in-vitro, and these myoblasts then fused to myotubes which produced normal full-length dystrophin. Two weeks after one single injection of the oligonucleotides into the muscles of mdx mice, up to 2 % of the fibers around the injection site contained new dystrophin which was not revertant dystrophin, i.e. normal dystrophin after a spontaneous suppression of the mutation. This amount of new dystrophin remained stable for at least 10 weeks (Rando, Palo Alto).

   Exon skipping: With this technique one tries to change a Duchenne mutation into a Becker mutation. This can be done by inducing the splicing mechanism, which cuts out the introns from the pre-messenger RNA, to also eliminate one specific exon after a point mutation or a deletion had shifted the reading frame and thus caused a premature stop codon. The aim of this approach is to restore the disturbed reading frame.
   The gene with its mutation is not altered by exon skipping, but the messenger RNA, mRNA, no longer contains the information of the skipped exon. As the mRNA is shorter than normal, the dystrophin protein is also shorter, it contains fewer amino acids. If the missing amino acids are part of the central region of the dystrophin, they are often not essential, and the resulting shorter protein can still perform its stabilizing role of the muscle cell membrane. The result would be the change of the severe Duchenne symptoms into the much milder symptoms of Becker muscular dystrophy.

   Elimination of the mutation of the mdx mouse: This strategy was used to by-pass the single base-pair defect, the point mutation in exon 23 of the mdx mouse. An antisense oligoribonucleotide, consisting of 20 base pairs which were complementary to the RNA sequence of the pre-mRNA at the border region of exon 23 to intron 23, induced the splicing process to disregard the exon containing the mutation. The genetic information of exon 23, which codes for 71 amino acids in the rod domain, was thus skipped during the reading process.
   This and the other investigated antisense-oligoribonucleotides were chemically modified, e.g., by protecting the normally free and sensitive OH-groups of the ribose units of the RNA by methyl groups (-CH₃). About 5 micrograms (millionths of a gram) of these stabilized potential "gene drugs" were injected together with the detergent F127 into the leg muscles of living mdx mice. After two to four weeks, up to 20 % of the muscle fibers contained almost the normal amount of slightly shortened dystrophin. And, together with the other components of the dystrophin complex, it was located correctly on the cell membrane. The muscle force was significantly improved but not completely normalized. A repeated treatment increased the number of dystrophin-positive muscle fibers without the development of an immune rejection (Wilton, Perth, Partridge, London).

   Exon skipping in the mRNA of the human dystrophin: Exon 45 of the dystrophin gene is the single most frequently deleted exon in boys with Duchenne dystrophy. This causes a frame shift in the mRNA and a premature stop codon leading to a truncated and non-functional dystrophin protein which is subsequently degraded in the muscle cells. However, if both exons 45 and 46 are missing simultaneously, the reading frame is not disturbed, not shifted, resulting in a shorter than normal dystrophin with 108 non-essential amino acids from the middle part of the protein missing. Patients with this specific deletion have the milder symptoms of Becker muscular dystrophy.
   In-vitro experiments to specifically delete exon 46 from the dystrophin pre-mRNA in myotubes from mdx mice were successfully performed with four different antisense oligoribonucleotides complementary to a splicing regulatory sequence within exon 46. With these oligonucleotides, an exon recognition sequence (ERS) or an exonic splice enhancer (ESE) is blocked, structures which are necessary for the splicing process, the joining of the exons in the mRNA.
   Then, similar in-vitro experiments were performed with several antisense oligonucleotides specific to the analogous splice sites of the human exon 46. With one of them, consisting of 19 base pairs, it was possible to delete exon 46 from about 15 % of the dystrophin pre-mRNA in myotubes obtained from two Duchenne patients who had a deletion of exon 45. (The last page of this report shows the molecular details of this experiment.)
   This percentage of shortened mRNA without exons 45 and 46 led to normal quantities of a shortened dystrophin in at least 75 % of the myotubes. After 16 hours, this new dystrophin could be detected in the cells, after 48 hours it had moved to the cell membrane and stayed there for at least one week. The re-appearance of the dystrophin also led to the restoration of the dystrophin complex in and under the muscle cell membrane. In the meantime, the reading frame could be corrected in in-vitro muscle preparations from six other patients who had different deletions and also a point mutation.
   This technique is very specific, resulting only in the removal of the one targeted exon. With the same technique, it was possible to skip18 other exons in muscle cell cultures. Thus, this very promising strategy, proven in test-tube experiments, may possibly later convert the Duchenne mutations into Becker mutations of more than 65 % of patients with deletions.
   Ongoing studies investigate different methods for the transfer of the of the antisense oligoribonucleotides into a living organism. To this end, experiments with living mice are performed which, instead of their own dystrophin gene, have the human gene in their muscles which, in addition, has been changed by creating "human" deletions. Clinical trials on Duchenne boys with these methods can only be contemplated after these studies with "humanized" mice have given positive results.
   As the structure of the dystrophin gene is known in all details, it is already possible to predict which exon would have to be skipped in order to restore the reading frame after a defined deletion or point mutation. At this time, such predictions can only be purely theoretical. It is not certain, whether the results obtained in cell culture or with mice, will be the same in Duchenne boys, and it is not certain either, whether, in an individual case, the restored but shortened dystrophin will really lead to the symptoms of a Becker muscular dystrophy (van Ommen, van Deutekom, Leiden).
   Splice sites are specific RNA sequences at the borders of exons and introns which are essential for the correct removal of the non-coding intron sequences from the pre-mRNA. This pre-messenger RNA is the first product of an active gene. After removal of the intron sequences by splicing, it becomes messenger RNA, mRNA, that moves to the ribosomes where it acts as the information transmitter for protein synthesis.

   Exon skipping with internally produced oligonucleotides: For a new method of skipping exons, the antisense oligoribonucleotides do not have to be injected but are synthesized in the cell nucleus, where they are needed, after transfer of their gene. The intron sequences are cut out from the pre-mRNA in the cell nucleus by spliceosomes. They are complex structures consisting of several proteins and small RNAs, small nuclear = snRNAs, which recognize the exon-intron borders and which join the exons after splicing precisely and without shifting the reading frame.
   Some of these very short RNAs are the U7-snRNAs. They bind to splice sites at special recognition sequences in the pre-mRNA, block the splicing of specific exons, and thus ensure that several proteins of different length are produced by one gene. The U7-snRNAs normally regulate the splicing processes of the mRNA of histones. Histones are proteins necessary for packaging the DNA in the chromosomes.
   For this new method, the U7-snRNAs were genetically modified so that they no longer bound to the histone pre-mRNA but only to the splice sites in the region of exon 23 of the pre-mRNA of the mouse dystrophin. To achieve this, a gene coding for the modified U7-snRNA together with a creatine kinase enhancer was packed into plasmids as vectors and then, in a laboratory experiment, transferred into isolated myoblasts of the mouse. These cells then developed in the culture dish into muscle cells.
   In the nuclei of the transgenic myoblasts, the transferred gene produced modified U7-snRNAs. They had new recognition sequences and bound now in front and behind exon 23 of the dystrophin pre-mRNA to sites important for the splicing process. Thus, these splice sites were blocked and exon 23 was removed together with the introns during the splicing of the pre-mRNA (exon skipping). As the point mutation of the mdx mice is localized in exon 23, the removal of this exon had the consequence that the mdx muscle cells, which could not make any normal dystrophin, now produced a slightly shortened dystrophin protein which was localized at its correct site underneath the cell membrane.
   The aim of this fundamental in-vitro gene experiment - outside of the living mouse - was to prove that the muscle cells themselves can produce the therapeutic antisense oligribonucleotides.
   For a human application, U7-snRNAs adapted to the individual mutation of the patient would have to be used, whose gene must then be transported with a gene therapeutic vector into the cell nuclei of the muscle fibers. An alternative would be an ex-vivo transfer transporting the U7-snRNA genes into satellite cells or other muscle stem cells and then injecting them into the blood stream or directly into the muscles. As the U7-snRNA genes are very short, this transfer would probably be easier than the transfer of the cDNA for the entire or the shortened dystrophin as is being tried with other methods (Weis, Bern; Lochmüller, Munich).

   Homologous recombination: The point mutation in the dystrophin gene of mdx mice could be repaired in 15 to 20 % of isolated myoblasts by the addition of a DNA string of 603 base pairs whose sequence is the same as the sequence before and after the mutation site of exon 23 of the mice but not containing the C-to-T exchange of the mutation. Part of the new gene segment was exchanged against the corresponding, the homologous, in exon 23. This short fragment homologous recombination, SFHR, technique was then applied to isolated myoblasts from a Duchenne patient with a deletion of exon 13 (Kapsa, Melbourne).

   Ignoring a premature stop codon by antibiotics: About 5 % of Duchenne boys have a point mutation in their dystrophin gene which changed an amino acid code word into one of the three stop codons, TGA, TAG and TAA, after which the synthesis of dystrophin stops prematurely.
   Gentamicin is an antibiotic that causes the RNA translation mechanism in the ribosomes to ignore such a premature stop codon, i.e. to read through it. The normal stop codons, which are protected by a special three-dimensional structure, will, however, be respected as before. In mdx-mice, up to 20 % of the normal amount of new and functional dystrophin has been obtained in this way. Gentamicin has the advantage of being a well known drug whose use as a possible therapy for Duchenne muscular dystrophy would not need long approval procedures (Sweeney, Philadelphia).
   In order to confirm these positive results, two other studies with mdx mice were performed which showed, however, that under similar conditions after a treatment with gentamicin no new dystrophin could be detected (Karpati, Montreal; Lochmüller, Munich).
   Two clinical studies with gentamicin have been performed on Duchenne boys but have also not led to new dystrophin. Possibly the 14-day trial period was insufficient. Therefore, another and longer lasting clinical trial with 36 patients is now being performed (Mendell, Columbus).

Upregulation of the utrophin gene: Utrophin is a protein with a structure and function very similar to dystrophin. In humans, its gene is located on chromosome 6, it has 75 exons and is about one million base pairs long. The utrophin protein is about 7 % shorter than dystrophin. It is present in many body tissues, also in muscle, but there it is concentrated in regions where the motor nerves contact the muscle membrane. Before birth, the utrophin concentration in muscle is much higher than afterwards. This protein, though it is present only in small amounts, makes the Duchenne symptoms less severe than they would be if utrophin were also missing. In fact, mdx-mice whose utrophin gene was knocked out experimentally, which thus have neither dystrophin nor utrophin, have Duchenne-like symptoms and die early in contrast to "normal" mdx mice whose muscles hardly degenerate.
   Experiments with mice have shown, that utrophin, if it is present in larger amounts, can replace dystrophin. The mice used were transgenic mice who contained utrophin mini genes in their germ line, introduced by a technique that cannot be used in humans. Other transgenic mice were raised which produced utrophin only when they were given the antibiotic tetracyclin in their drinking water. The increased amount of utrophin prevented the development of Duchenne symptoms, and this effect was more pronounced in newborn than in 10 or 30 day old mice.
   For a possible Duchenne therapy, another strategy is followed, namely to increase the normally low amount of utrophin by upregulation of the activity of its gene. To achieve this, an activating substance is needed, which could well be a known drug, or some other chemical or a naturally occurring substance.
   During the last years, synthetic chemistry has developed methods to automatically produce thousands of partly unknown substances. Many of these substances are being tested in the laboratory, also automatically, on cell cultures from mdx mice for their ability to activate the gene of luciferase, which is preceded by the two promoters of the utrophin gene. The light producing enzyme luciferase from fireflies is easier to detect than utrophin. Every hit, i.e., every substance which shows at least a low activity in these preliminary tests, is further modified, and then, all these similar substances are tested first on muscle cell cultures, and, if they react positively, in living mdx mice, too, to see whether they can also upregulate the utrophin gene. Only after one or several convincingly active substances are found, would it be possible to start clinical studies with Duchenne patients in a few years (Davies, Oxford).
   Such an activator of the utrophin gene could, if it is a small molecule, be applied via the blood stream from where it would reach all the muscles. And the immune system would recognize an additional amount of utrophin as a substance of its own because it is already present in small amounts in Duchenne boys. Therefore, no immune rejection should develop.
   However, something that upregulates the utrophin gene, could also do the same with other genes. Thus, before such an activator is tested in children, it should be made certain that it does not produce any undesirable side effects.

   Other experiments to increase utrophin: The transfer of a shortened utrophin gene with adenoviruses into dystrophic dogs led to utrophin which could take over the function of the missing dystrophin. --- Glucocorticoides, among them prednisone, can upregulate the utrophin gene to some extent. --- Such corticoides have also been isolated from Chinese plant medicines which traditionally are used against muscular dystrophy. --- Low grade chronic inflammation in muscles of mdx mice leads to a marked upregulation of utrophin on the membrane outside of the contacta with the motor nerves. --- The amino acid L-arginine can increase the amount of utrophin in mdx mice and alleviate significantly the dystrophic symptoms. The enzyme nitric oxide synthase uses arginine for the production of the biologically active gas nitric oxide. But nitric oxide is also active in other biological processes. Arginine, therefore, cannot be used for a Duchenne therapy without further investigations. --- The small protein heregulin with which the motor nerves stimulate muscle development can also significantly alleviate the symptoms of mdx mice. --- One end of the utrophin mRNA is not translated into protein, but binds to structures at the contact regions of the nerves, thus restricting utrophin to these sites. If this binding could be prevented by a drug, it might become possible to distribute utrophin more evenly over the entire muscle membrane so that it can better replace dystrophin. --- Another form of utrophin, the very similar B-utrophin was recently identified in blood vessels. But only the "normal" A-utrophin is present in muscles and can partially compensate for dystrophin after upregulation of its gene.

Activities of thousands of muscle genes: In order to measure simultaneously the activities of very many genes in one tissue sample in a single experiment, gene arrays are used. As the sequences of practically all human genes are known, short segments of thousands of genes can be produced automatically and applied by a robot in a certain pattern to a quartz chip a few square centimeters in size. If biochemically produced DNA copies of the mRNAs of all active genes are applied as an analysis sample to the chip, light points appear at those chip sites where there are complementary DNA sequences of these active genes. The light intensity of these points is automatically measured which then makes it possible to determine which genes are active and to what extent and which are not active.
   With this expression profiling, several thousand genes in muscle samples were tested which came from normal and mdx mice, from healthy and Duchenne boys, from transgenic mice which neither had their own dystrophin nor utrophin, and also from other mice which instead had human dystrophin in their muscles.
   The results showed that the absence of dystrophin causes the increase and the decrease of the activities of many muscle genes. A whole series of genes which are responsible for the energy production in muscle cells were less active in mice without dystrophin and utrophin, i.e. their muscles had an energy crisis which contributed to the degeneration of their dystrophic muscles. On the other hand, many genes necessary for the development and repair of muscles were increased in their activities, in some cases by a hundred times, i.e., they were upregulated by the disease process. Similar results were obtained with human muscle samples.
   Further experiments showed that in mice many other genes were upregulated, genes which are involved in the development of surface structures, in the production of signal factors for protein synthesis, in the intensification of immune reactions, and in other processes responsible for the muscle dystrophic symptoms of mdx mice. In Duchenne patients, these changes were less pronounced, and in transgenic mice with human dystrophin, these gene activities were normal, because they no longer had a muscular dystrophy.
   After these first results were obtained a few years ago, many more experiments with this new technique were performed to answer other questions of Duchenne research. Their and future results will help to understand the complex relationship between the many parts of the muscular architecture and its changes if one of its most important components, dystrophin, is absent. This will open new avenues for the development of a Duchenne therapy (van Ommen, Leiden; Kunkel, Boston; Hoffman, Washington, and others).

   A dystrophic worm: Caenorhabditis elegans is a 0.9 millimeter long transparent worm which is used extensively by gene researchers because all its 19,733 genes are known and also all its 959 body cells, 95 of which are muscle cells. Its muscles have a dystrophin similar to the dystrophin of humans, which can also have mutations causing dystrophic symptoms.
   Individual genes were inactivated and also activated so that the dystrophic symptoms could be studied which were caused by the missing or increased gene activities. E.g., it could be shown that the upregulation of dystrobrevin, a very short form of dystrophin, slowed down the muscle degeneration significantly. Further investigations will contribute to the clarification of still unknown molecular relationships during the development of muscular dystrophy (Ségalat, Lyon).

   Integrins and synthrophins: Integrins are a family of proteins which are located in the muscle cell membrane. They are necessary for the fusion of myoblasts to myotubes and the development of myotubes to mature muscle cells. They also participate in the propagation of signals from cell to cell.
   In mice without dystrophin and utrophin, the amount of one of these integrins was increased about twofold by gene transfer. This extended the life expectancy of the mice by threefold, and their dystrophic symptoms ameliorated significantly (Kaufman, Urbana).
   The five known syntrophins are proteins that mediate the interaction of signalling proteins with the dystrophin and utrophin complexes at the muscle cell membrane. Understanding these effects in more detail may have consequences for a therapy with utrophin (Froehner, Seattle).