Mesoangioblast composition

Abstract

A mesoangioblast composition is disclosed, in which the mesoangioblasts (MABs) are engineered to express a truncated agrin protein.

Description

[0001] MESOANGIOBLAST COMPOSITION

[0002] Technical Field of the Invention

[0003] The invention relates generally to ex vivo gene therapy, and in particular, to compositions comprising genetically engineered mesoangioblasts, methods of their production, and use in the treatment of muscular dystrophy disorders, including

[0004] Duchenne Muscular Dystrophy (DMD) and Becker Muscular Dystrophy (BMD).

[0005] Background of the Invention

[0006] Muscular dystrophies (MD) are genetic diseases affecting skeletal and often cardiac muscle. They differ greatly in age of onset, severity and muscle groups predominantly affected. All lack an efficacious therapy, and steroids represent the only treatment able to delay the progression of the disease, though with severe collateral effects.

[0007] Duchenne Muscular Dystrophy (DMD) is the most common and one of the most severe muscular dystrophies, affecting approximately one in four thousand newly born children. It is characterized by progressive wasting of skeletal and cardiac muscle, leading to a variable but progressive muscle weakness that limits the patient’s motility and, in later years affects cardiac and respiratory functions. DMD is caused by different mutations of the dystrophin gene, located on the X chromosome. In 90% of cases, mutations lead to a change in the mRNA reading frame that prevents dystrophin protein production. In-frame deletions lead to a shorter but partially functional dystrophin, associated with milder Becker Muscular Dystrophy (BMD).

[0008] Dystrophin and the proteins associated with the dystrophin-associated glycoprotein complex (e.g., sarcoglycans) have a critical role in muscle cell interaction with the basal lamina and provide elastic resistance to the sarcolemma during contraction. In the absence of dystrophin and associated proteins, the membrane is more easily damaged, leading to calcium influx. This influx can be easily measured by dyes such as Fura2, that become fluorescent when bound to calcium ions. Depending on the duration of the influx this may appear as spikes or as more sustained elevation of fluorescence. If not regulated, calcium influx leads to hypercontraction, proteolysis and fibre degeneration. Muscle fibre degeneration is followed by regeneration carried out by satellite cells, resident myogenic stem / progenitor cells and, to a minor extent, interstitial cells such as pericytes. In humans, adult myogenic cells have limited self-renewal potency and in DMD the continuous regeneration cycles eventually lead to depletion of the cell population. Muscle degeneration is accompanied by chronic inflammation that progressively leads to accumulation of dense connective and adipose tissues that replace muscle fibres making any therapy ineffective at this stage.

[0009] Dystrophin was cloned more than 30 years ago, but there is still no cure for DMD.

[0010] Many therapeutic approaches have entered the clinical arena, but none has yet reached significant and long -lasting clinical efficacy.

[0011] There is, therefore, a considerable clinical need for an improved therapeutic approaches for treating DMD and related muscular dystrophy disorders.

[0012] Summary of the Invention The inventors have found that mesoangioblasts (MABs) expressing truncated agrin protein block calcium spikes in vivo in a mouse model of muscular dystrophy.

[0013] Thus, in accordance with a first aspect of the present invention, there is provided a mesoangioblast (MABs) composition, wherein the MABs are engineered to express a truncated agrin protein.

[0014] In addition, the inventors have found that the expression of a small nuclear RNA (snRNA) in MABs induces exon-skipping in dystrophin pre-mRNA transcripts. Significantly, the inventors have surprisingly found that when used in combination, the strategy provides a synergistic therapeutic effect leading to abundant dystrophin production, in a therapeutically effective range.

[0015] Thus, in some embodiments, the MABs are engineered to additionally express an snRNA capable of inducing exon-skipping in a dystrophin pre-mRNA transcript. Thus, in these embodiments, the MABs are engineered to express:

[0016] (i) a truncated agrin protein; and

[0017] (ii) a small nuclear RNA (snRNA) capable of inducing exon-skipping in a dystrophin pre-mRNA transcript. Mesoangioblasts MABs are progenitor cells, initially identified in the dorsal aorta of developing embryos. They are also present in post-natal skeletal muscle as a vessel-associated cell population. MABs have similar properties to and likely represent the in vitro derivative of adult pericytes. They are capable of differentiating into several mesodermal lineages when derived from embryonic tissues, including skeletal and cardiac muscle, bone and cartilage. Human skeletal muscle derived MABs are defined by the expression of the pericyte markers NG2 proteoglycan and alkaline phosphatase (ALP) and absence of endothelial markers. This enabled their prospective isolation from freshly dissociated ALP+ cells. Their differentiation potency is restricted to skeletal and smooth muscle in vitro and in vivo.

[0018] Advantageously, MABs can be genetically manipulated in vitro and, following intraarterial injection, they efficiently home to skeletal muscle. Such cells have been shown to promote regeneration in both murine and canine DMD models. Analysis of MAB- transplanted muscles showed that these cells also contribute, though to a limited extent, to the satellite cell pool, which represents an important aspect for ensuring long-term muscle homeostasis. In addition, in vitro studies demonstrated that human MABs can also act as immunomodulatory cells, through an Indoleamine 2,3- dioxygenase (IDO) and prostaglandin E-2 (PGE)-mediated suppression of T cell proliferation.

[0019] In some embodiments, the MABs for use in the invention may be identified by the expression of NG2 proteoglycan and alkaline phosphatase (ALP) and the absence of expression of endothelial markers such as CD31, CD34, and / or CD45.

[0020] In some embodiments, the MABs may be identified by the expression of the following markers: CD31-, CD34-, CD45-, CD62L-, CDio6“, CD117-, CD133-, CD146L CD49b+, CD13+, and CD44L In some embodiments, the MABs may be identified by the further expression at least one of the following: VCAM-1 (vascular cell adhesion molecule), CD36, CD44, b , b5, bi, b2 integrins, alpha integrins (some ai, a.5 and a6), LFA-1 (leukocyte factor antigen), IL-1R (interleukin-i receptor), SDF-R (stromal derived factor receptor), and / or Cadherins.

[0021] MABs Production Previous studies using MABs expanded in vitro have shown that these cells differentiate mainly into skeletal and smooth muscle, and most importantly, can cross the vessel wall and become distributed systemically for treating diffuse forms of muscular dystrophy such as DMD. Unfortunately, despite the systemic infusion of large numbers of cells (over a billion in total), the level of engraftment has been found to be low (<i%) and though some donor derived dystrophin maybe detected as a result of this approach, the efficacy in restoring muscle function is modest.

[0022] In some embodiments, the MABs can differentiate in vitro into skeletal muscle in suitable culture conditions and in the absence of inducing agents.

[0023] Detailed protocols for isolation and culture of mesoangioblasts have been reported.

[0024] Briefly, a method for isolating MABs from a tissue sample previously obtained from a donor comprises the steps of mincing mechanically the tissue, followed by culture of the resulting fragments, without proteolytic enzymes, until cells outgrow from the fragments. The outgrown cells are then cultured in a mammalian cell growth medium such as DMEM (or Megacell or MyoCultTN-SF) including suitable supplements.

[0025] In some embodiments, the donor from which the MABs are derived may be a disease- affected subject, such as a subject having, or suspected of having, a muscular dystrophy or other myopathy, eventually characterised by aberrant calcium homeostasis as defined herein.

[0026] In some embodiments, the donor from which the MABs are derived maybe a subject having a muscular dystrophy disorder. The muscular dystrophy disorder may be a muscular dystrophy disorder characterised by aberrant calcium homeostasis.

[0027] In some embodiments, the donor from which the MABs are derived may be a healthy subject.

[0028] In some embodiments, the process of obtaining MABs from a donor does not form part of the disclosed methods.

[0029] The MABs may be engineered (genetically modified) to express exogenous sequences encoding a truncated agrin protein and optionally also an snRNA capable of inducing exon-skipping in a dystrophin pre-mRNA transcript. Agrin

[0030] Muscle contraction is triggered by a nerve impulse traveling on the axon as a wave of depolarization and leading to the release of a chemical message (acetylcholine) at the synapse. The connection between nerves and muscle occurs via a highly complex synaptic structure called the neuromuscular junction (NMJ).

[0031] Agrin is an important component of the NMJ and consists of a core protein of ~220 kDa with two glycosaminoglycan side chains that carry heparan sulfate and / or chondroitin sulfate. These increase the molecular weight of agrin to greater than 400 kDa. Agrin binds several proteins on the surface of muscle fibres, including dystroglycan on the membrane and laminin in the basal lamina, and these interactions help to stabilise the NMJ. Agrin deficient mice die at birth due to an inability to form NMJs. However, the role of agrin in the NMJ in the pathology of muscular dystrophies has not been significantly investigated because there is no known involvement of this structure in most muscular dystrophies.

[0032] The present inventors have for the first time surprisingly found that the addition of an agrin protein to cultures of DMD muscle fibres abolishes cytosolic calcium spikes.

[0033] As shown in Figure 17D, agrin is a multidomain protein and includes:

[0034] 9 follistatin-like module (FS) domains;

[0035] 2 laminin-EGF-like (LE) domains;

[0036] 2 serine / threonine-rich (S / T) domains; 1 sperm protein, enterokinase and agrin (SEA) domain;

[0037] 4 epidermal growth factor-like (EG1-4) domains; and

[0038] 3 laminin globular (LG1-3) domains.

[0039] References to a “native agrin”, “native agrin protein”, “native agrin sequence”, and the like, refer to any of the agrin isoforms, and may thus refer to the human agrin isoform 1 sequence having Database accession number NP_ooi2922O4, or to the human agrin isoform 2 sequence having Database accession number NP_94O978, or to the human agrin isoform 3 sequence having Database accession number NP_ooi35i656. The inventors have surprisingly found that truncated agrin proteins may advantageously be used in the disclosed approach in preference to native agrin. Agrin is a large molecule and the truncated agrin has been found to provide significantly greater effects due to an increased capacity to diffuse freely throughout (inside or outside) the muscle fibres. A “truncated agrin protein” refers to a portion or derivative of the native agrin protein that is smaller in size than the native protein, for example, comprising fewer amino acids and thus a lower molecular weight.

[0040] The truncated agrin protein defined herein preferably, when expressed, additionally includes, for example at the N terminus, a secretion signal peptide. An example of a suitable signal peptide is the peptide consisting of residues 1-29 of the human agrin isoform 1 sequence having Database accession number NP_ooi2922O4.

[0041] A truncated agrin protein may comprise or consist of a single portion of the native agrin protein, or a combination of a plurality of portions of the native agrin protein. Thus, a truncated agrin protein, in some examples, may comprise or consist of one particular domain of the native agrin protein, or a combination of a number of domains that is fewer that the total number found in the native agrin protein, and that may occur consecutively or non-consecutively in the native agrin protein.

[0042] Thus, in some embodiments, the truncated agrin protein may be a truncated agrin protein having a length, for example, based on the number of amino acid residues, that is less than 70%, 60%, 50%, 40%, 30%, or 20% of the length of a native agrin protein. Preferably, the truncated agrin protein may have a length that is less than 40%, or less than 30%, more preferably less than 25%, of the length of a native agrin protein.

[0043] In some embodiments, the truncated agrin protein maybe a truncated agrin protein having a length, for example, based on the number of amino acid residues, that is greater than 3%, preferably greater than 5%, of the length of a native agrin protein.

[0044] In some embodiments, the truncated agrin protein maybe a truncated agrin protein having a length, for example, based on the number of amino acid residues, that is 4- 40%, preferably 6-30% or 7-25%, and most preferably 10-20% of the length of a native agrin protein. In some embodiments, the truncated agrin protein maybe a truncated agrin protein comprising a portion of a native agrin protein. Such a portion may have a length, for example, based on the number of amino acid residues, that is 4-40%, preferably 6-30% or 7-25%, and most preferably 10-20% of the length of a native agrin protein.

[0045] In some embodiments, the truncated agrin protein may comprise or consist of one or a plurality of domains that is derived from a domain present in a native agrin protein. Each of the one or more domains of the truncated agrin protein that is derived from a domain present in a native agrin protein may have at least 90%, 95%, or 99% sequence identity to a corresponding domain present in a native agrin protein.

[0046] In some embodiments, the truncated agrin protein may comprise or consist of one or a plurality of domains, each of which is derived from, or present in, a native agrin protein. Thus, in some embodiments, the truncated agrin protein may comprise or consist of at least 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 domains that are derived from, or present in, a native agrin protein.

[0047] In some embodiments, the truncated agrin protein may comprise or consist of no more than 15, 12, 10, or 8 domains that are derived from, or present in, a native agrin protein. Preferably, the truncated agrin protein may comprise or consist of no more than 7 domains that are derived from, or present in, a native agrin protein, such as 5, 4, 3, 2, or 1 domain.

[0048] In some embodiments, the truncated agrin protein may comprise or consist of at least 1, 2, or 3 domains that are derived from, or present in, a native agrin protein.

[0049] Preferably, the truncated agrin protein may comprise or consist of at least 4 domains that are derived from, or present in, a native agrin protein.

[0050] In some embodiments, the truncated agrin protein may comprise or consist of between 1 and 12 domains that are derived from, or present in, a native agrin protein, preferably between 2 and 10, between 3 and 9, or between 4 and 8 domains that are derived from, or present in, a native agrin protein. Most preferably, the truncated agrin protein may comprise or consist of 5, 6, or 7 domains that are derived from, or present in, a native agrin protein. Thus, in some embodiments, the truncated agrin protein may be a truncated agrin protein having a length based on the number of amino acid residues, that is 5-40% of the length of a native agrin protein, wherein the truncated agrin protein comprises or consists of at most 7 domains that are derived from, or present in, a native agrin protein.

[0051] In some embodiments, the truncated agrin protein may comprise or consist of one or a plurality of domains that are present sequentially in the native agrin protein. In some embodiments, the truncated agrin protein may comprise or consist of one or a plurality of domains that are present in the C terminal portion of the native agrin protein.

[0052] In some embodiments, the truncated agrin protein may comprise or consist of one or a plurality of domains that are present sequentially in the C terminal portion of the native agrin protein.

[0053] In some embodiments, the truncated agrin protein maybe a truncated agrin protein having a length based on the number of amino acid residues, that is 5-40% of the length of a native agrin protein, wherein the truncated agrin protein comprises or consists of at most 7 domains that are derived from, or present in, a native agrin protein, wherein the domains are present sequentially in the native agrin protein.

[0054] In some embodiments, the truncated agrin protein may comprise or consist of a plurality of domains that are present non-sequentially in the native agrin protein.

[0055] In some embodiments, the truncated agrin protein may comprise or consist of a plurality of domains, some or all of which may be ordered in the order that they are present in a native agrin protein.

[0056] In some embodiments, the order of the domains in the truncated protein may be different to the order of the domains in the native agrin protein.

[0057] In some embodiments, the truncated agrin protein may comprise or consist of a plurality of domains that are present sequentially in the native agrin protein, in which one or more of the domains is duplicated. Thus, in some embodiments, the truncated agrin protein may be a truncated agrin protein having a length based on the number of amino acid residues, that is 5-40% of the length of a native agrin protein, wherein the truncated agrin protein comprises or consists of at most 7 domains that are derived from, or present in, a native agrin protein, wherein the domains are present sequentially in the native agrin protein, and ordered in the order that they are present in a native agrin protein.

[0058] Different splice variants of native agrin exist, the most abundant of which is a secreted protein containing an N-terminal agrin (NtA) domain (see Figure 17D). Agrin is produced in the soma of the neurons, transported down the axon and released from the axon ending of the motor nerve into the synaptic cleft of the NMJ.

[0059] When added to cultured muscle cells, agrin induces the aggregation of acetylcholine receptors (AChRs) and several additional components that are concentrated at the neuromuscular junction (NMJ) in vivo.

[0060] Some truncated agrin proteins have been found to retain this AChR-aggregating activity. Truncated agrin proteins are identified on the basis of the size and position of the fragment. For example, the inventors have found that a 95 kDa C-terminal fragment of agrin has been shown to be particularly advantageous for expression in MABs and for use in the disclosed approach. This fragment is referred to as the “C95” agrin fragment, and corresponds to the LG1-3 agrin domains and the EG1-4 agrin domains (see Figure 17D).

[0061] Thus, in some embodiments, the truncated agrin protein may be a truncated agrin protein having a length based on the number of amino acid residues, that is less than about 40% of the length of a native agrin protein, wherein the truncated agrin protein comprises or consists of at most 7 domains that are derived from, or present in, a native agrin protein, wherein domains comprise or consist of the LG1-3 agrin domains and the

[0062] EG1-4 agrin domains of a native agrin protein. Preferably, the domains are ordered in the sequence that they are present in the native agrin protein.

[0063] In some embodiments, the truncated agrin protein may have at least 90% sequence identity to the C95 agrin fragment. Thus, in some embodiments, the truncated agrin protein may consist of or comprise the C95 agrin fragment.

[0064] The C95 agrin protein may have an amino acid sequence consisting of or comprising residues 1329-2068 of the human agrin isoform 1 sequence.

[0065] In some embodiments, the truncated agrin protein may comprise the C95 agrin fragment, further including 1-5 additional agrin domains, such as an additional 4, 3, 2, or 1 agrin domains.

[0066] In some embodiments, the C95 agrin protein may have an amino acid sequence consisting of residues 1329-2068, and may further comprise up to about 100 additional amino acids. For example, in some embodiments, the C95 agrin protein may have an amino acid sequence consisting of residues 1329-2068 of the human agrin isoform 1 plus, at most, an additional 50, 40, 30, 20, or 10 amino acids N-terminal to this sequence. In addition, or alternatively, in some embodiments, the C95 agrin protein may have an amino acid sequence consisting of residues 1329-2068 of the human agrin isoform 1 plus, at most, an additional 50, 40, 30, 20, or 10 amino acids C-terminal to this sequence.

[0067] Two sites within the C terminal portion of agrin are modified by alternative mRNA splicing:

[0068] An insert “A” (called y in rat), in the LG2 domain (as shown in Figure 17D), has been found to have the sequence Lys-Ser-Arg-Lys identically conserved across a number of vertebrate species. The presence or absence of this KSRK insert is indicated with the suffix “A4” or “Ao”, respectively.

[0069] Inserts at site “B” (called z in rat) , in the LG3 domain (as shown in Figure 17D), are less well conserved, and comprise 8, 11, or 19 amino acids. The presence or absence of an insert at site B is indicated with the suffix “Bo”, “B8”, “B11”, or “B19”, as appropriate.

[0070] The sequence of these inserts affects the AChR-aggregating activity of agrin: while recombinant rat, chick, and ray agrin A4B8 induces AChR aggregation at picomolar concentrations, no activity of soluble agrin A0B0 is detected. This is true of native agrin and also various truncated forms. For example, the C95 A4B8 agrin fragment has been found to be capable of inducing AChR aggregation but the C95 A0B0 fragment is not. Likewise, the “C45” A4B8 agrin fragment (consisting of the LG2-3 and the EG3-4 domains), and the “C21” B8 agrin fragment (a 21 kDa fragment containing only site B8 and the LG3 domain) have been found to induce AChR aggregation. Agrin binding to the nonintegrin receptor o-dystroglycan, a peripheral membrane protein of the dystrophin-glycop rotein complex, is not required for agrin-induced AChR aggregation, and the AChR-aggregating activity of agrin isoforms do not correlate with their affinity to a-dystroglycan. The agrin A0B0 isoform, which does not bind heparin and is inactive in AChR aggregation, binds to a-dystroglycan with high affinity. In contrast, the agrin A4B8 isoform, which binds heparin and is active in AChR aggregation, binds to a-dystroglycan with up to ten times lower affinity than the agrin A0B0 isoform.

[0071] The present inventors have found that, when expressed in MABs, the agrin A0B0 isoform is able to able to abolish Ca++flux and is thus surprisingly advantageous for use in the disclosed approach.

[0072] Thus, in some embodiments, the truncated agrin protein is capable of binding to a- dystroglycan.

[0073] In some embodiments, the truncated agrin protein is capable of binding to a- dystroglycan with a high affinity. Methods of determining the binding affinity between truncated agrin proteins and a-dystroglycan will be known to the skilled person. In some embodiments, binding to a-dystroglycan with a high affinity means that the truncated agrin protein is capable of binding to a-dystroglycan with a binding affinity greater than 50% of the binding affinity of native agrin protein under comparable conditions.

[0074] In some embodiments, the truncated agrin protein is not capable of inducing AChR aggregation. Methods of determining AChR aggregation will be known to the skilled person. In some embodiments, inducing AChR aggregation means that the truncated agrin protein induces AChR aggregation at a level of less than 50% of the level of AChR aggregation induced by native agrin protein under comparable conditions. In some embodiments, the truncated agrin protein is not capable of inducing AChR phosphorylation. Methods of determining AChR phosphorylation will be known to the skilled person. In some embodiments, inducing AChR phosphorylation means that the truncated agrin protein induces AChR phosphorylation at a level of less than 50% of the level of AChR phosphorylation induced by native agrin protein under comparable conditions.

[0075] In some embodiments, the truncated agrin protein is an A0B0 agrin isoform.

[0076] In some embodiments, the truncated agrin protein comprises or consists of a C terminal portion of the native agrin protein having a molecular weight of less than 120 KDa. In some embodiments, the truncated agrin protein is an A0B0 agrin isoform and comprises or consists of a C terminal portion of the native agrin protein having a molecular weight of less than 120 KDa.

[0077] In some embodiments, the truncated agrin protein comprises or consists of a C terminal portion of the native agrin protein having a molecular weight of less than too

[0078] KDa. In some embodiments, the truncated agrin protein is an A0B0 agrin isoform and comprises or consists of a C terminal portion of the native agrin protein having a molecular weight of less than too KDa. Thus, in some embodiments, the truncated agrin protein may be a truncated agrin protein having a length, based on the number of amino acid residues, that is less than about 40% of the length of a native agrin protein, wherein the truncated agrin protein comprises or consists of the LG1-3 agrin domains and the EG1-4 agrin domains of a native agrin protein, wherein the truncated agrin protein is a A0B0 isoform. Preferably, the domains are ordered in the sequence that they are present in the native agrin protein.

[0079] In some embodiments, the truncated agrin protein comprises or consists of a C terminal portion of the native agrin protein having a molecular weight of greater than 20 KDa. In some embodiments, the truncated agrin protein is an A0B0 agrin isoform and comprises or consists of a C terminal portion of the native agrin protein having a molecular weight of greater than 20 KDa.

[0080] In some embodiments, the truncated agrin protein comprises or consists of a C terminal portion of the native agrin protein having a molecular weight of 21-95 KDa. In some embodiments, the truncated agrin protein is an A0B0 agrin isoform and comprises or consists of a C terminal portion of the native agrin protein having a molecular weight of 21-95 KDa.

[0081] In some embodiments, the truncated agrin protein comprises or consists of the C95 A0B0 agrin fragment.

[0082] In some embodiments, the truncated agrin protein may consist of or comprise a truncated agrin protein referred to as “mini-agrin” (MAG). MAG is derived from the muscle agrin isoform, and is capable of binding to a- dystroglycan with a high affinity. MAG has a domain structure consisting of the NtA domain, one FS module (FS), and the C95 agrin fragment.

[0083] MAG has been found to be particularly advantageous for use in the disclosed approach, due to ease of expression in MABs, and due to having significant residual cellular function.

[0084] Advantageously, MAG has also been found by the inventors to be highly diffusible along muscle fibres.

[0085] In some embodiments, the truncated agrin protein comprises or consists of MAG.

[0086] In some embodiments, the truncated agrin protein comprises or consists a MAG A0B0 truncated agrin protein.

[0087] The truncated agrin protein may be modified, for example, to provide improved in vivo stability. Thus, in some embodiments, the agrin protein may comprise one or a plurality of mutations which provide improved stability, such as protease resistance.

[0088] For example, in some embodiments, the agrin protein may comprise a mutation to remove a protease cleavage site. Thus, in some embodiments, the agrin protein may comprise no protease cleavage sites.

[0089] The inventors have found, however, that protease resistance does not provide any advantages for the therapeutic use of the disclosed MABs compositions. Thus, the truncated agrin protein advantageously does not require the inclusion of any site specific mutations. Thus, in preferred embodiments, the agrin protein does not comprise a site-specific mutation, such as for example, to remove a protease cleavage site.

[0090] There are at least three potential heparan sulfate (HS) attachment sites within the primary structure of native agrin, but it is thought that only two of these actually carry HS chains when the protein is expressed in vivo. Agrin molecules lacking any HS chains have been found by the inventors to abolish cytosolic calcium spikes when added to cultures of DMD muscle fibres. Thus, in some embodiments, the truncated agrin protein may include at most 1, or more preferably, no HS chains. Truncated agrin proteins having a reduced number, or no HS chains have been found to have advantageously improved diffusion properties.

[0091] Indeed, any functional truncated agrin protein may be expressed in the disclosed MABs. Thus, in some embodiments, the truncated agrin protein is a functional truncated agrin protein.

[0092] Various tests may be used to determine whether a truncated agrin protein is a functional truncated agrin protein for the purposes of the present disclosure.

[0093] A “functional truncated agrin” protein refers generally to a truncated agrin protein having biological activity, and in particular, having sufficient biological activity to reduce the progressive degradation of muscle tissue that is otherwise characteristic of muscular dystrophy. For example, the inventors have found that agrin proteins are capable of significantly altering the level of proteins associated with calcium homeostasis, such as Cavi.i, which reduces one or more signs or symptoms associated with a muscular dystrophy disorder, such as a muscular dystrophy disorder characterised by aberrant calcium homeostasis. For example, a suitable test may involve determining the ability of the truncated agrin protein, when used at any concentration in the micromolar range, to reduce or abolish cytosolic calcium spikes in cultures of DMD muscle fibres. Suitable methods for testing this property will be known to the skilled person and are disclosed herein. Any truncated agrin protein that is capable of abolishing cytosolic calcium spikes when expressed in cultures of DMD muscle fibres is considered to be a functional truncated agrin protein for the purposes of the present disclosure. “Abolishing” in this context refers to reducing the level of intracellular calcium to less than 10% of the maximal level measured using a calcium indicator such as fura2 in a calcium imaging assay. Likewise, any truncated agrin protein that is capable of reducing cytosolic calcium spikes when expressed in cultures of DMD muscle fibres is considered to be a functional truncated agrin protein for the purposes of the present disclosure.

[0094] “Reducing” in this context refers to reducing the level of intracellular calcium to less than 50%, less than 40%, less than 30%, or less than 20% of the maximal level measured using a calcium indicator such as fura2 in a calcium imaging assay. In some embodiments, the truncated agrin protein is capable of reducing or abolishing cytosolic calcium spikes when expressed in cultures of DMD muscle fibres. Methods of detecting cytosolic calcium spikes in cultures of DMD muscle fibres will be known to the skilled person. In some embodiments, reducing or abolishing cytosolic calcium spikes when expressed in cultures of DMD muscle fibres means that in the presence of the truncated agrin protein the level of cytosolic calcium spikes in cultures of DMD muscle are at a level of less than 50% of the level detected in the absence of the expression of the truncated agrin protein.

[0095] An additional, or alternative test, may involve determining the ability of the truncated agrin protein, when used at any concentration in the micromolar range, to increase, restore, or substantially restore the level of expression, in DMD muscle fibres or in vitro models of such fibres, of one or more specific proteins that are known to have reduced or absent expression in such fibres. Suitable proteins include proteins involved in Ca++ homeostasis such as, for example, Cavi.i, CACNA1S, SCN4A, RyRi, and / or TRPC1. Suitable methods for determining the level of expression, including for determining whether the one or more proteins shows reduced or absent expression in the DMD muscle fibres and whether the level of expression is increased, restored, or substantially restored by the truncated agrin protein, will be known to the skilled person and are disclosed herein. Any truncated agrin protein that is capable of restoring the level of expression of the one or more proteins is considered to be a functional truncated agrin protein for the purposes of the present disclosure. “Restoring” in this context refers to increasing the level of expression of the one or more proteins to more than 90% of the level of expression in healthy cultured muscle fibres. Likewise, any truncated agrin protein that is capable of increasing or substantially restoring the level of expression of the one or more proteins is considered to be a functional truncated agrin protein for the purposes of the present disclosure. “Substantially restoring” in this context refers to increasing the level of expression of the one or more proteins to more than 50%, more than 60%, more than 70%, or more than 80% of the level of expression in healthy cultured muscle fibres. An alteration in expression of a protein, such as Cavi.i, refers to an increase or decrease in the level of the protein that is detectable in a biological sample (such as a sample from a subject at risk of having, or having, muscular dystrophy) relative to a control (such as a sample from a subject without a muscular dystrophy) or relative to a reference value known to be indicative of the level of protein in the absence of the disease. An “alteration” in expression includes an increase in expression (upregulation) or a decrease in expression (down-regulation).

[0096] In some embodiments, the truncated agrin protein is capable of inducing and / or increasing the expression of proteins associated with calcium homeostasis, such as Cavi.i. Methods of determining the level expression of proteins associated with calcium homeostasis, such as Cavi.i, and thereby whether expression of the protein is induced or increased, will be known to the skilled person. In some embodiments, increasing the expression of proteins associated with calcium homeostasis, such as Cavi.i. means that, when expressed in cultures of DMD muscle fibres, the truncated agrin protein induces expression of the subject protein at a level of greater than 50% more than the level of expression of the corresponding protein in the absence of the expression of the truncated agrin protein.

[0097] An example of a vector that may be used to express a truncated agrin protein in MABs is shown in Figure 18.

[0098] Small nuclear RNA

[0099] In the past, oligonucleotide-based exon skipping therapy failed to reach efficacy in a phase III trial possibly because of the inability of these large molecules to diffuse through specific barriers such as collagen-rich fibrotic tissue and muscle basal lamina. In contrast, cell therapy, even if technically more demanding, promises a longer lasting, even permanent effect, since muscle fibres last in principle for a lifetime, and this would avoid life-long drug administration. However, the present inventors have developed a new strategy combining the advantage of exon skipping with that of cell-based therapy, thereby synergistically reducing the respective drawbacks of the two separate approaches. Cell-mediated exon skipping exploits the multinucleated nature of myofibers so that one single cell, once fused with a regenerating muscle fibre, will produce a high level of the snRNA that will enter and cross-correct neighbouring resident dystrophic nuclei. It has been found that this approach is capable of amplifying dystrophin production by an order of magnitude and, hence, significantly increasing the therapeutic effect. Thus, this cell-mediated exon skipping approach exploits the multinucleated nature of myofibers to achieve cross-correction of resident dystrophic nuclei by the snRNA.

[0100] The inventors observed that co-culture of genetically corrected human myogenic cells with their dystrophic counterparts at a ratio of either 1:10 or 1:30 leads to dystrophin production at a level several folds higher than predicted by simple dilution, thus reaching what is considered a therapeutic level. This is due to diffusion of RNA, such as U7 snRNA, to neighbouring dystrophic resident nuclei.

[0101] The present inventors have surprisingly found that this cell-mediated exon skipping approach may be advantageously used in combination with the delivery of a truncated agrin protein, using a MABs composition. The resulting engineered MABs have surprisingly been found to delay muscle fibre cell death by an amount that exceeds the level of effect that might have been anticipated based on the results seen when either approach is used alone.

[0102] In use, both the truncated agrin protein and snRNA have advantageously been found to diffuse along (the outside and inside, respectively) the muscle fibres, to thereby provide therapeutic intervention at all relevant locations throughout the muscle fibres, both internally (snRNA) and externally (truncated agrin protein).

[0103] Thus, in some embodiments, the composition may comprise MABs engineered to express a truncated agrin protein and a snRNA capable of inducing exon skipping in a dystrophin pre-mRNA transcript. Methods of determining exon skipping activity will be known to the skilled person. For example, exon skipping activity may be measured by analysing total RNA isolated by reverse transcriptase polymerase chain reaction (RT-PCR) using dystrophin-specific primers flanking the targeted exon.

[0104] The snRNA is capable of inducing exon-skipping in a dystrophin pre-mRNA transcript, and any snRNA that is capable of increasing, restoring, or substantially restoring, the level of dystrophin in muscle fibres maybe used.

[0105] Thus, in some embodiments, the MABs maybe engineered to express an snRNA that increases the level of dystrophin in muscle fibres. Suitable methods for determining the level of dystrophin in muscle fibres will be known to the skilled person.

[0106] In some embodiments, the snRNA maybe a modified uridine-rich 7 (U7) snRNA.

[0107] In some embodiments, the snRNA comprises a sequence of bases that specifically hybridizes to a region of the dystrophin pre-mRNA transcript to thereby induce the skipping of one or a plurality of dystrophin exons.

[0108] In some embodiments, the snRNA comprises a sequence of bases that specifically hybridizes to a region of the dystrophin pre-mRNA transcript to thereby induce the skipping of one or more of dystrophin exons 50-55 of the dystrophin gene.

[0109] In some embodiments, the snRNA comprises a sequence of bases that specifically hybridizes to a region of the dystrophin pre-mRNA transcript to thereby induce the skipping of dystrophin exon 51.

[0110] In some embodiments, the snRNA sequence may consist of or comprise the RNA sequence of SEQ ID NO: 1:

[0111] CCUCUGUGAUUUUAUAACUUGAUUCAAGGAAGAUGGCAUUUCU

[0112] [SEQ ID NO: 1]

[0113] In some embodiments, the snRNA sequence may consist of or comprise the RNA sequence of SEQ ID NO: 1, further comprising at most two nucleotide substitutions, preferably further comprising at most one nucleotide substitution. In some embodiments, the snRNA sequence may consist of or comprise the RNA sequence of SEQ ID NO: 1, further comprising at most two nucleotide additions, preferably further comprising at most one nucleotide addition. In some embodiments, the snRNA sequence may consist of or comprise the RNA sequence of SEQ ID NO: 1, further comprising at most two nucleotide deletions, preferably further comprising at most one nucleotide deletion.

[0114] In some embodiments, the snRNA sequence may consist of the RNA sequence of SEQ ID NO: 1.

[0115] An example of a lentiviral vector that may be used to express the snRNA of SEQ ID NO: 1 in MABs is shown in Figure 5. In some embodiments, MABs cells are engineered to express both a truncated agrin protein and the small nuclear RNA capable of inducing exon skipping in the human dystrophin gene. An advantage of this approach is that substantially all the MABs cells in the composition deliver both therapeutic agents for diffusion throughout the muscle fibre.

[0116] In some embodiments, a single lentivector may be used to express in the MABs both a truncated agrin protein and the small nuclear RNA capable of inducing exon skipping in the human dystrophin gene. In some embodiments, separate lentivectors may be used to express in the MABs a truncated agrin protein and the small nuclear RNA capable of inducing exon skipping in the human dystrophin gene.

[0117] In some embodiments, the snRNA comprises a sequence of bases that specifically hybridizes to a region of the dystrophin gene transcript to thereby induce the skipping of one or more of exons 2-7 of the dystrophin gene.

[0118] In some embodiments, the snRNA comprises a sequence of bases that specifically hybridizes to a region of the dystrophin gene transcript to thereby induce the skipping of one or more of exons 11-18 of the dystrophin gene. In some embodiments, the snRNA comprises a sequence of bases that specifically hybridizes to a region of the dystrophin gene transcript to thereby induce the skipping of one or more of exons 43-48 of the dystrophin gene. In some embodiments, the snRNA comprises a sequence of bases that specifically hybridizes to a region of the dystrophin gene transcript to thereby induce the skipping of one or more of exons 45-53 of the dystrophin gene.

[0119] In some embodiments, the MABs may be engineered to express a plurality of different snRNAs. Thus, in some embodiments, the MABs may be engineered to express a truncated agrin protein and a plurality of different snRNA. In some embodiments, each of the different snRNA maybe capable of inducing skipping of a different dystrophin. In some embodiments, each of the different snRNA may be capable of inducing skipping of the same dystrophin exon.

[0120] With five different vectors each designed for multiple exon skipping, targeting the most common mutations of the dystrophin gene, approximately 60% of the DMD patient population maybe treated, as shown in Table I. Table I:

[0121] Thus, in some embodiments, the MABs maybe engineered to express a plurality of different snRNs, wherein: at least one snRNA comprises a sequence of bases that specifically hybridizes to a region of the dystrophin gene transcript to thereby induce the skipping of one or more of exons 45-53 of the dystrophin gene; at least one snRNA comprises a sequence of bases that specifically hybridizes to a region of the dystrophin gene transcript to thereby induce the skipping of one or more of exons 2-7 of the dystrophin gene; at least one snRNA comprises a sequence of bases that specifically hybridizes to a region of the dystrophin gene transcript to thereby induce the skipping of one or more of exons 11-18 of the dystrophin gene; at least one snRNA comprises a sequence of bases that specifically hybridizes to a region of the dystrophin gene transcript to thereby induce the skipping of one or more of exons 43-48 of the dystrophin gene; at least one snRNA comprises a sequence of bases that specifically hybridizes to a region of the dystrophin gene transcript to thereby induce the skipping of one or more of exons 45-53 of the dystrophin gene; and at least one snRNA comprises a sequence of bases that specifically hybridizes to a region of the dystrophin gene transcript to thereby induce the skipping of one or more of exons 50-55 of the dystrophin gene.

[0122] Method of Manufacture In accordance with a second aspect of the present invention, there is provided a method of producing the MABs composition of the first aspect, the method comprising transducing MABs with a vector for expressing a truncated agrin protein.

[0123] In some embodiments, the method may comprise transducing MABs with a first vector for expressing a truncated agrin protein and a second vector for expressing a small nuclear RNA (snRNA) capable of inducing exon-skipping in a dystrophin pre-mRNA transcript.

[0124] In some embodiments, the method may comprise transducing MABs with a single vector for expressing both a truncated agrin protein and a small nuclear RNA (snRNA) capable of inducing exon-skipping in a dystrophin pre-mRNA transcript.

[0125] In some embodiments, the first and / or second vector may comprise a viral vector. In some embodiments, the first and / or second vector may comprise a retroviral vector.

[0126] In some embodiments, the first and / or second vector may comprise a lentiviral vector.

[0127] In some embodiments, the viral vector may be modified to delete any non-essential sequences. In some embodiments, the MABs may be expanded in vitro and transduced with the “trans-skip” lentiviral vector. The “trans-skip” lentiviral vector is produced using a widely used HIV-i based transient packaging system, which encodes minimal packaging functions and is devoid of all HIV accessory genes except REV, which is expressed separately by the packaging construct. This system incorporates excellent biosafety features and the self-inactivating lentiviral vector encodes no viral genes.

[0128] The “trans-skip” lentiviral vector is replication defective as it does not encode any viral genes and is only packaged in the presence of minimal viral proteins in trans under transient transfection conditions. The backbone is a maximally gutted vector, which encodes no viral genes and carries only minimal essential cis-acting sequences. In addition to the removal of the U3 region from the 5’ LTR, the U3 region from the 3’ LTR is also deleted. Since the 3’ LTR is naturally copied to the 5’ end following transduction in target cells, the resulting newly generated 5’ LTR lacking the U3 region is effectively inactivated. This self-inactivation ensures its inability to be mobilised following transduction, even in the extremely unlikely presence of spurious HIV helper packaging functions in target cells. The 5’ to 3’ configuration of the vector is as follows:

[0129] • A chimeric 5’ LTR encompassing the RSV enhancer / promoter replacing the HIV U3 region and substituting the transcriptional dependence on Tat. • The vector-packaging signal.

[0130] • Rev responsive element (RRE).

[0131] • HIV central polypurine tract (cppt) and termination sequence (cts).

[0132] • EFia, a human specific ubiquitous efficient promoter to drive transgene expression. • The U7 small nuclear RNA of interest, for example, U7 snRNA engineered to skip exon 51 (#51) of the dystrophin gene

[0133] • W RE: Woodchuck Hepatitis Virus Regulatory Element

[0134] • SVqo / Ori: Simian Virus 40 Origin of replication

[0135] • pB322 / 0ri: Plasmid BR322 Origin of replication. • The 3’ LTR has a major deletion in the U3 region, removing a 400-nucleotide sequence encompassing all the major determinants responsible for regulating the HIV- 1 LTR promoter activity, including the TATA box. In the absence of this sequence, the template used to generate both copies of the LTR in the integrated provirus is lost, hence resulting in vector inactivation. In some embodiments, the MABs are obtained, such as previously obtained, from a donor and the method comprises expanding the MABs in vitro and transducing the cells with a lentiviral vector. In some embodiments, the MABs are obtained, such as previously obtained, from a donor and may be expanded in vitro and transduced with the “trans-skip” lentiviral vector encoding a small nuclear RNA engineered to skip exon 51 of the dystrophin gene.

[0136] In some embodiments, the MABs may comprise autologous human mesoangioblasts (hMABs), expanded in vitro and transduced with a lentiviral vector.

[0137] In some embodiments, the MABs may comprise autologous human mesoangioblasts (hMABs), expanded in vitro and transduced with the “trans-skip” lentiviral vector encoding a small nuclear RNA engineered to skip exon 51 of the dystrophin gene.

[0138] The MABs may be autologous MABs stably transduced to express a small nuclear RNA, to correct the primary transcript of dystrophin by inducing skipping of an exon, such as exon 51. Thus, the MABs may be “Trans-skip” lentiviral vector gene modified autologous MABs.

[0139] In some embodiments, the MABs may be obtained from the Extensor Digitorum Brevis (EBD) muscle biopsy of a DMD patient. In some embodiments, the method of manufacture of the MABs may comprise one or more of the following steps:

[0140] 1. obtaining EBD muscle sample from a DMD patient, or using an EBD muscle sample previously obtained from a DMD patient;

[0141] 2. identification of suitable minced muscle fragments, such as by microscopy; 3. plating of fragments onto collagen I coated dishes;

[0142] 4. harvesting of MABs (outgrown cells);

[0143] 5. transduction of MABs; and

[0144] 6. expansion and characterisation of MABs. A detailed example of a suitable method of producing transduced MABs is provided below. The muscle biopsy is washed and dissected under an inverted microscope and unwanted connective and adipose tissues are disposed. The cleaned muscle is minced into small fragments, transferred to collagen coated dishes and cultured overnight in the minimum volume of medium necessary to cover the dish surface. This facilitates adhesion of the fragments to the dish(es). The next day additional medium is gently added and the fragments cultured for up to 15 days, with periodic monitoring of possible cell outgrowth. Once outgrown cells are apparent and cover medium-large areas of the dish (usually between 5 and 15 days), cells are collected by mild proteolytic treatment. The collected MABs are cultured in specific culture media that favours their proliferation (such as Myocult-SF from STEMCELL Technologies). When the MABs have reached stages of sub-confluence (e.g. 2 / 3 confluence), they are transduced with the Trans-Skip lentivector. Transduced cells are analysed for the efficiency of transduction, measured as average Vector Copy Number / cell. For example, a QPCR assay may be used with a primer / probe set able to recognise GAPDH (Housekeeping) and WPRE (DNA sequence delivered from viral vector). Transduction is considered acceptable, for example, when > 1 VCN.

[0145] Transduced MABs, are then cultured in a suitable medium, such as Myocult-SF from STEMCELL Technologies. The cells actively proliferate in Myocult-SF with an average population doubling time of 1 day and can be sub-cultured for at least 40 passages before reaching senescence.

[0146] The MABs may be cryopreserved, for example, using a DMSO-free, GMP grade freezing medium (e.g. Cryo-SFM) at a cell concentration between >ixio7and <5xio7cells / ml. Vials are placed firstly in a GMP controlled -8o°C freezer for < 24 hours, then placed into GMP controlled Vapour Phase Liquid Nitrogen storage.

[0147] When transduction is considered acceptable (i.e. > 1 VCN), cells are characterised, for example, for cell number and viability, proliferation rate, cell surface marker phenotype, tumourigenesis, viral integration, and / or karyotype, using standard protocols. Cells may also be tested for potency, for example, by analysing dystrophin production in vitro (as % of the level expressed in healthy muscle cells). In vitro differentiation and hence dystrophin production may underestimate the activity in vivo. For example, mouse MABs do not differentiate in vitro but do so in vivo. Also, human MABs from different donors may show little differentiation potency in vitro but they do better in vivo.

[0148] “Non target” cells, i.e. other cell types present in the composition may be checked by FACS analysis (>90% of cells express mesenchymal markers such as CD44 and CD13).

[0149] In some embodiments, the cell preparation may contain a variable proportion of satellite cells (CD56+). This is not detrimental, however, because these cells also make muscle and provide therapeutic benefit.

[0150] In preferred embodiments, there are no endothelial or hematopoietic cells (CD31, CD34, CD45 <5%) in the cell preparation. In preferred embodiments, the culture medium selects for MABs from an initially heterogeneous cell population.

[0151] Inflammatory cells present in the initial biopsy (lymphocytes or macrophages) do not survive under the culture conditions used.

[0152] In some embodiments, each dose of the composition may comprise between 1 million and too million cells, such as between 5 million and 50 million cells, or between 10 million and 30 million cells.

[0153] In some embodiments, the viral titre of the MABs in the composition may be between 10 x to6and 10 x to8, such about 10 x 107.

[0154] In some embodiments, the Vector Copy Number of the MABs in the composition may be between 1 and 5.

[0155] In some embodiments, the viral vector may be arranged to directly enter into a cell.

[0156] In some embodiments, a delivery system such as a liposome-based delivery system may be used to transfect the target MABs.

[0157] Therapeutic Use and Conditions That May be Treated In accordance with a third aspect of the present invention, there is provided the MABs composition of the first aspect, or produced by the method of second aspect, for use as a medicament. In accordance with a fourth aspect of the present invention, there is provided the MABs composition of the first aspect, or produced by the method of second aspect, for use in treating a muscular dystrophy disorder.

[0158] In some embodiments, the muscular dystrophy disorder maybe DMD or BMD. Thus, the composition may be for use in treating DMD or BMD.

[0159] In some embodiments, the muscular dystrophy disorder may be characterised by aberrant calcium homeostasis. In some embodiments, the MABs composition may be for use in reducing or preventing calcium spikes in muscle fibres.

[0160] Thus, in some embodiments, the MABs composition may be for use in treating DMD or BMD by reducing or preventing calcium spikes in muscle fibres.

[0161] In some embodiments, the donor from which the MABs are derived may be a disease- affected subject, such as a subject having, or at risk of having, a muscular dystrophy disorder, such as a muscular dystrophy disorder characterised by aberrant calcium homeostasis as defined herein.

[0162] In some embodiments, the donor from which the MABs are derived maybe a subject having a muscular dystrophy disorder. In some embodiments, the MABs composition maybe suitable for autologous administration for use in the treatment of the subject’s disease.

[0163] In some embodiments, the MABs composition maybe for use autologously in the treatment of a muscular dystrophy disorder.

[0164] In some embodiments, the donor from which the MABs are derived may be a healthy subject. In some embodiments, the process of obtaining MABs from a donor does not form part of the disclosed methods.

[0165] The MABs may be engineered (genetically modified) to express exogenous sequences encoding a truncated agrin protein and an snRNA capable of inducing exon-skipping in a dystrophin pre-mRNA transcript.

[0166] The composition of the first aspect provides several advantages when used as a medicament, such as when used in treating a muscular dystrophy disorder.

[0167] For example, the inventors have found that agrin delivered to muscle fibres in culture using the disclosed methods abolishes cytosolic calcium spikes in DMD muscle cells characterised by aberrant calcium homeostasis. Agrin has been surprisingly found to restore the physiological levels of the Ca++sensor Cavi.i, which binds to and regulates the Ryanodine receptor 1 (RyRi), thereby preventing unregulated Ca++release into the cytoplasm. When delivered using the disclosed MABs, the truncated agrin protein has been found to provide these effects more effectively than native agrin, as a result of an improved capacity to diffuse along multinucleated myofibres, due to its smaller size than native agrin and the absence of negatively charged heparan sulphate chains.

[0168] Moreover, in contrast to other therapeutic approaches, cell therapies using the composition of the first aspect provide a long-lasting, even substantially permanent effect, because in principle muscle fibres last for a lifetime. Thus, this approach provides the advantages of avoiding lifelong drug administration, and immune suppression, in the case of heterologous methods.

[0169] In any event, the MABs delivery, especially in embodiments involving autologous MABs, may be repeated multiple times without the appearance of a specific immune response (which invariably happens with AAV therapies).

[0170] In addition, the disclosed approach provides the further advantage that there is no requirement for the truncated agrin protein to be modified for increased stability, such as mutated for protease resistance, which may be the case in therapeutic approaches comprising the delivery of a therapeutic protein composition. Cell therapy for muscular dystrophy has met with limited success, mainly due to the poor engraftment of donor cells, especially in fibrotic muscle at an advanced stage of the disease. To overcome this limitation, the present inventors have developed a cell- mediated exon skipping that exploits the multinucleated nature of myofibres to achieve cross-correction of resident, dystrophic nuclei by small nuclear RNA such as the U7 snRNA engineered to induce the skipping of one or more exons, such as exon 51, of the dystrophin gene.

[0171] This approach combines the advantages of cell-based therapy and exon skipping while reducing their respective drawbacks. The inventors have found that co-culture of genetically corrected human myogenic cells with their dystrophic counterparts leads to dystrophin production at a level several folds higher than previously possible, at a level that is considered a therapeutic level. This advantageously high level of expression is due to diffusion of U7 snRNA to neighbouring dystrophic resident nuclei. Under the same conditions wild-type MABs (co-cultured with a 10- or 30- fold excess of DMD cells) only slightly increase the level of dystrophin expression. When transplanted into immune-deficient mdx mice carrying a mutation of exon 51, genetically corrected human myogenic cells produce dystrophin at higher level than WT cells, well in the therapeutic range, and lead to force recovery even with an engraftment of only 3-5%. This level of dystrophin production promises clinical efficacy with the use of the disclosed cell therapy.

[0172] Aberrant calcium homeostasis is characterised by transient (spikes) or sustained elevation in the cytosolic intracellular calcium concentration, identified, for example, using a calcium indicator such as fura2 in a calcium imaging assay. These calcium spikes were previously thought to occur to elicit muscle contraction upon neural stimulation in physiological conditions or because membrane damage that may lead to hyper contractures. The inventors have now surprisingly found, however, that spikes in the cytosolic intracellular calcium concentration precede sarcomere assembly, innervation and muscle contraction.

[0173] Homeostasis is the state of steady internal, physical, and chemical conditions maintained by living systems. References to calcium homeostasis refer to intracellular calcium homeostasis. Intracellular free Ca++concentration ([Ca++]i) widely varies in different conditions; however, the cytoplasmic concentration of Ca++under resting conditions is much lower than Ca++in the extracellular milieu. Specific intracellular organelles (known as Ca++stores), the mitochondria and the sarcoplasmic reticulum however, accumulate Ca++and maintain a higher concentration of Ca++than the cytoplasm. Abnormal or dysregulated calcium homeostasis is caused by disturbances of Ca++channels, exchangers, calcium ion pumps, calcium ion transport channels and calcium ion binding proteins, often as a consequence of a damage to the membranes such as those caused by absence of dystrophin.

[0174] Calcium spikes refer to transient, repeated fluctuations in the relative concentration of [Ca++]i which, in the context described, occur spontaneously in muscle cells, i.e. in the absence of muscle innervation. Spikes occur as the result of a release of Ca++from intracellular calcium stores such as the sarcoplasmic reticulum, since blocking release from these stores using inhibitors of voltage dependent Ca++channels nimodipine or nitrendipine, substantially reduces their amplitude. As the inventors have shown in the examples, calcium spikes do not occur in healthy, untreated, non-innervated myotubes.

[0175] Moreover, since calcium spikes are transient, they are distinct from other types of calcium influx, occurring, for example, as a result of damage to (e.g. loss of integrity of) the sarcoplasmic reticulum and / or plasma membrane. Methods of identifying and measuring calcium spikes will be known to the skilled person and are described herein. For example, suitable approaches include the microscopic imaging method discussed in the examples, in which intracellular calcium status is measured using the green-fluorescence calcium indicator Fluo-4 AM. Calcium spikes may be observed using live cell calcium indicators. Calcium indicators include, but are not limited to, chemical calcium indicators. Chemical calcium indicators are small molecules that bind calcium ions via chelation. Typically, binding of a Ca++to an indicator molecule leads to either an increase in quantum yield of fluorescence, or an emission / excitation wavelength shift. One example of a chemical calcium indicator is Fura-2. This molecule possesses an excitation spectrum at 380 nm (emission at 500 nm) as a Ca++-free form. When the molecule binds to Ca++the excitation is shifted to 340 nm (with the same emission wavelength at 500 nm). Thus, the increase in [Ca++] excites the molecule in the Ca++-binding form at 340 nm, along with a decrease in fluorescence intensity at 380 nm from the free form of the dye. Monitoring the fluorescence intensity over time using an excitation wavelength of ~38o nm will, therefore, enable the visualisation of calcium spikes which occur as Fura-2 switches between the Ca++-free form and the Ca++-binding form as a result of changes in [Ca++].

[0176] Dystrophinopathy refers to diseases (Duchenne and Becker muscular dystrophy) due to mutations in the DMD gene, which encodes for the dystrophin protein found in muscle.

[0177] As can be seen in the examples, the inventors have shown that lack of dystrophin protein leads to reduced expression of Cavi.i, a calcium sensor component of the Dihydropyridine Receptor (DHPR) complex, known to bind to and regulate the Ryanodine Receptor 1 (RyRi). In the absence of dystrophin, reduced expression of Cavi.i deregulates DHPR thus leading to de-regulated or spontaneous Ca++fluxes.

[0178] In some embodiments, the muscular dystrophy is either Duchenne muscular dystrophy (DMD) or Becker muscular dystrophy (BMD). In some embodiments, the muscular dystrophy is another muscular dystrophy caused by a membrane protein mutation resulting in aberrant (i.e. de-regulated) calcium homeostasis. A non-exhaustive list include of such muscular dystrophies includes several Limb Girdle Muscular Dystrophies (e.g. 2A, 2D, 2E 2l, 2L) and myotonic dystrophy,

[0179] In some embodiments, a therapeutically effective amount of the composition of the first aspect is used, that restores calcium homeostasis, thereby treating one or more signs or symptoms associated with the disorder. A therapeutically effective amount refers to an amount effective for lessening, ameliorating, eliminating, preventing, or inhibiting at least one symptom of the disorder, and may be empirically determined. In various embodiments of the present disclosure, a “therapeutically effective amount” is an amount sufficient to achieve a statistically significant promotion of either tissue or cell regeneration, such as muscle cell regeneration, or delay in cell death, compared to a control.

[0180] Thus, in some embodiments, the therapeutically effective amount is an amount sufficient to achieve a statistically significant delay in muscle fibre death, compared to a control. In some embodiments, the therapeutically effective amount is an amount sufficient to increase the expression of Cavi.i.

[0181] In some embodiments, the therapeutically effective amount is an amount sufficient to prevent or delay muscle fibre death. Methods of measuring the prevention or delay in muscle fibre death will be known to the skilled person. These include, for example, methods involving measuring the level of creatine kinase (an enzyme released by dying muscle fibres) in blood serum, and methods comprising assessing the presence of necrotic fibres by hematoxylin and eosin (H&E) staining.

[0182] The delivery of the composition is not limited to any specific form of administration but pertains generally to the delivery of MABs engineered to express an agrin protein and a snRNA, for use in treating or alleviating the symptoms of the disclosed disorders, delivered by any suitable means.

[0183] In some embodiments, the use as a medicament comprises the intra-arterial and / or intra-muscular administration of the composition of the first aspect.

[0184] Composition for use as a medicament In some embodiments, the composition of the first aspect, or for use in accordance with the third or fourth aspect, may comprise a pharmaceutically acceptable excipient.

[0185] In some embodiments, the composition may be in the form of a cell suspension comprising a liquid.

[0186] In some embodiments, the composition may be in the form of an injectable cell suspension.

[0187] In some embodiments, the composition may also comprise one or more excipients which are suitable for use in cellular therapies. Such excipients will be known to persons skilled in the art.

[0188] In some embodiments, the composition for administration may comprise a Sodium Chloride buffer comprising Human Serum Albumin (HSA). For example, in some embodiments, the composition for administration may comprise 0.9% w / v Sodium

[0189] Chloride Injection BP containing 1% HSA. Method of Treatment

[0190] In accordance with a fifth aspect of the present invention, there is provided a method of treating a muscular dystrophy disorder, the method comprising administering a therapeutically effective amount of a composition of the first aspect, or a produced by the method of the second aspect, to a subject in need thereof.

[0191] The muscular dystrophy disorder maybe a muscular dystrophy disorder characterised by aberrant calcium homeostasis.

[0192] All the features described herein (including any accompanying claims, abstract and drawings), and / or all the steps of any method or process so disclosed, maybe combined with any of the above aspects in combination, except combinations where at least some of such features and / or steps are mutually exclusive.

[0193] For a better understanding of the invention, and to show how embodiments of the same may be carried into effect, reference will now be made, by way of example, to the accompanying Figures, in which: Figure 1 shows cell-mediated exon skipping induced by lentiviral vector U7#51T2AGFP. A. RT-PCR analysis of transcripts from WT muscle (1), or DMD TIM cells (2), or DMD-GFP cells (3) or DMD-U7 cells (4). Note the band corresponding to the shorter transcript (skipping exons 48- 51) is only detectable in 4. GAPDH and MyHC are loading controls. B. Sequencing of the cDNA from the band C in Figure 1A (4), showing the expected sequence for exon 47-52. C. FACS analysis of DMD TIM cells transduced with the vectors expressing either GFP (T2AGFP) or U7 snRNA (U7#51T2AGFP). FACS analysis of a mixture (1:10 or 1:30) of transduced and nontransduced cells. D. Double IF with antibodies against dystrophin (red) and GFP (green) of DMD-U7 cells, co-cultured and differentiated with 10- or 30- fold excess of DMD TIM cells. Note expression of dystrophin in virtually all myotubes at 1:10 dilution and in most myotubes at 1:30. DAPI stains nuclei in Blue. E. qRT-PCR analysing the expression of dystrophin and GFP in DMD-U7 cells, co-cultured and differentiated with 10- or 30-fold excess of DMD TIM cells. Note the expected progressive reduction of GFP (green bars) expression at 1:10 and 1:30 dilution. In contrast dystrophin expression (blue bars) largely exceeds the expected value. Expression of MyHC (red bars) is also reported as internal control (*P<o.O5; **P<o.oi). Figure 2 shows intracellular diffusion of snRNA and Dystrophin protein expression. A. Double in situ hybridization of a differentiated co-culture of DMD-U7 cells, with a 30- fold excess of DMD TIM cells. GFP mRNA appears green and is localised around few (transduced) nuclei (Green arrows) while the U7 snRNA (red dots) diffuses throughout the cytoplasm of multinucleated myotubes (Red arrows). B. WB analysis of protein extracts from differentiated cultures of TIM cells. Samples are from left to right: 1: WT TIM cells; 2: WT TIM cells co-cultured with a 10-fold excess of DMD TIM cells; 3: WT TIM cells co-cultured with a 30-fold excess of DMD TIM cells; 4: DMD TIM cells; 5: U7#51T2AGFP transduced DMD TIM cells; 6: U7#51T2AGFP transduced DMD TIM cells co-cultured with a 10-fold excess of DMD TIM cells; 7: U7#51T2AGFP transduced DMD TIM cells co-cultured with a 30 fold excess of DMD TIM cells. Top lane: antiDystrophin antibody; medium lane: anti-MyHC antibody; bottom lane: anti-GAPDH antibody. Densitometric scans are shown on the right (WTTIM cells: blue bars; U7#51T2AGFP transduced DMD TIM cells: green bars. C. WB analysis of protein extracts from differentiated cultures of hMABs. Samples are from left to right: 1: DMD hMABs; 2: DMD-U7 cells; 3: DMD-U7 cells co-cultured with a 10-fold excess of DMD hMABs; 4: WTTIM cells co-cultured with a 10-fold excess of DMD hMABs; 5: WT hMABs cells. Top lane: anti-Dystrophin antibody; medium lane: anti-MyHC antibody; bottom lane: anti-GAPDH antibody. Densitometric scans are shown on the right WT TIM cells: blue bars; DMDU7 cells: green bars. D. WB analysis of protein extracts from differentiated cultures of DMD MyoD converted Fbs. Samples are from left to right: 1: DMD MyoD-converted Fbs; 2: DMD-U7 cells; 3: DMD-U7 cells co-cultured with a 10 fold excess of DMD Myod-converted Fbs; 4:DMD-U7 cells co-cultured with a 30 fold excess of DMD Myod-converted Fbs 5: WT TIM cells; 6: WT TIM cells co-cultured with a 10 fold excess of DMD Myod-converted Fbs; 7: WT TIM cells co-cultured with a 30 fold excess of DMD Myod-converted Fbs; Top lane: anti-Dystrophin antibody; medium lane: anti-MyHC antibody; bottom lane: anti-GAPDH antibody. Densitometric scans are shown on the right WT TIM cells: blue bars; DMD-U7 cells: green bars.

[0194] Figure 3 shows in vivo Dystrophin protein expression induced by lentiviral vector U7#51T2AGFP. A. WB analysis of protein extracts from the TA of NSG-mdx-Asi or NSG WT mice. Samples are from left to right: 1: NSG WT muscle; 2: NSG-mdx-Asi muscle transplanted with 5x10sDMD-U7 cells; 3: NSG-mdx-Asi muscle transplanted with 5XIO5WT TIM cells; 4: NSG-mdx-Asi muscle non transplanted. Top lane: anti-

[0195] Dystrophin antibody; medium lane: anti-MyHC antibody; bottom lane: anti-GAPDH antibody. Densitometric scans are shown on the right WT TIM cells: blue bars; DMDU7 cells: green bar; NSG WT muscle: white bar. B. WB analysis of protein extracts from the TA of NSG-mdx-Asi or NSG WT mice. Samples are from left to right: 1: NSG WT muscle; 2: NSG-mdx-Asi muscle transplanted with 5x105 U7 hMABs; 3: NSG-mdx-Asi muscle transplanted with 5x10sWT hMABs; 4: NSG-mdx-Asi muscle non transplanted. Top lane: anti-Dystrophin antibody; medium lane: anti-MyHC antibody; bottom lane: anti-GAPDH antibody. Densitometric scans are shown on the right WT hMABs: blue bars; U7 hMABs: green bars; NSG WT muscle: white bar. C. Double IF with antibodies against Dystrophin (green) and anti-human Lamin AC (red) of transverse sections of TA of NSG-mdx-Asi mice. Top lane TA transplanted with 5x10sWT TIM cells. Bottom lane TA transplanted with 5x10sDMD-U7 cells. Lower magnification of dystrophin staining is shown in the lower panels DAPI stains nuclei in Blue. Quantification of dystrophin positive fibres is shown in the inset (**P<o.oi). D. Double IF with antibodies Alpha-SG (green) and nNOS (red) of transverse sections of TA of NSG-mdx-Asi mice transplanted with sxio5DMD-U7 cells. Top lane TA transplanted with 5x10sWT TIM cells. Bottom lane TA transplanted with 5x10sDMD- U7 cells. DAPI stains nuclei in Blue.

[0196] Figure 4 shows intracellular diffusion of snRNA in vivo and improving of motility after transplantation. A. Double in situ hybridization of transverse sections of TA of NSG-mdx-Asi mice transplanted with sxio5DMD-U7 cells. LAM A / C appears green and is indicated the human transplanted cells while the U7 snRNA (red dots) diffuses throughout the cytoplasm of multinucleated myotubes. B. WB analysis of protein extracts from the TA of NSG-mdx-Asi: samples are from left to right: 1: NSG-mdx-Asi muscle transplanted with 5x10sDMD-U7 cells; 2: NSG-mdx-Asi muscle transplanted with 2.5XIO5DMD-U7 cells; 3: NSG-mdx-Asi muscle transplanted with 1.25x10sDMD- U7 cells. 4: NSG-mdx-Asi muscle transplanted with 0.625x10sDMD-U7 cells.5: NSG- mdx-Asi muscle not transplanted. Top Lane: anti-Dystrophin antibody; medium lane: anti-LAM A / C antibody; bottom lane: anti-GAPDH antibody. Densitometric scans are shown on the right LAM A / C: blue bars; Dys: green bar. C. Analysis of the motility assay Mice are from left to right; NSG WT: white Bar; NSGmdx-Asi; black bar; NSG- mdx-Asi transplanted with 5x10sDMD-U7 cells; green bar; NSG-mdx-Asi transplanted with 5x10sWT TIM cells (*P<o.os). Figure 5 shows a map of lentiviral vector U7#51T2AGFP and sequence of the snRNA (SEQ ID NO. 2). On the left the scheme of the U7#51T2AGFP derived from pCDH.EFK.MCH.T2A.GFP is shown, and on the right, the sequence of the U7 snRNA in red is underlined the antisense sequence (SEQ ID NO. 1).

[0197] Figure 6 shows cell-mediated exon skipping induced by lentiviral vector U7#51T2AGFP. A. Double IF with antibodies against dystrophin (red) and GFP (green) of DMD-U7 cells, co-cultured and differentiated with 10- or 30- fold excess of DMD TIM cells. Note expression of dystrophin in virtually all myotubes at 1:10 dilution and in most myotubes at 1:30. DAPI stains nuclei in Blue. B. Double IF with antibodies against dystrophin (red) and GFP (green) of DMD-GFP cells, co-cultured and differentiated with 10- or 30- fold excess of DMD TIM cells. There is no expression of dystrophin. DAPI stains nuclei in Blue. C. Analysis of DMD TIM cells transduced with the vectors expressing either GFP (T2AGFP) or U7 snRNA (U7#51T2AGFP). Analysis of a mixture (1:10 or 1:30) of transduced and non-transduced cells. Figure 7 shows phenotypical characterization of NSG-mdx-Asi mice before and after motility assay. A. Phenotypical characterization of NSG-mdx-Asi mice by Immunohistochemistry staining of Diaphragm; Heart; TA. Top Lane: Haematoxylin and Eosin; bottom lane: Trichrome Masson. B. Haematoxylin and Eosin staining of TA of NSG- mdx- si mice after 4 consecutive days of spontaneous motility. Samples are from left to right: 1: NSG WT muscle; 2: NSG-mdx-Asi muscle; 3: NSG-mdx-Asi muscle transplanted with 5x105 DMD-U7 cells; 4: NSG-mdx-Asi muscle transplanted 5x105 TIM WT cells.

[0198] Figure 8 shows transplantation of DMD-U7 cells in a Matrigel plug under the back skin of mdx / SCID mouse. A. IF with antibodies against dystrophin (red) of DMD-U7 cells, co-cultured and differentiated with 10- fold excess of DMD TIM cells in a matrigel plug implanted under the dorsal skin of mdx / SCID mice. B. WB analysis of protein extracts from matrigel plug implanted under the dorsal skin of mdx / SCID. Samples are from left to right: DMD: DMD TIM cells; 1: DMD U7 cells; 1:10: DMD U7 cells co- cultured with a 10-fold excess of DMD TIM cells. Top Lane: anti-Dystrophin antibody; medium lane: anti-MyHC antibody; bottom lane: anti-GAPDH antibody.

[0199] Figure 9 shows fluorescence intensity (FI) analysis of live-cells calcium level in differentiated myotubes. Myotubes were loaded with Flou-4 AM calcium dye (excited at 388 nm). Images were acquired from live time-lapse videos of myotubes at 37°C and 5%

[0200] CO2: each graph represents a five seconds time frame of each video. (A, B, C) untreated, non-innervated myotubes. (E, F, G) myotubes treated with to pM of calcium channel blocker nimodipine (NIM). (H, I, G) nimodipine-treated myotubes after addition of caffeine C) that induces calcium release from the SR (K, L, M) myotubes treated with to pM of calcium channel blocker nitrendipine (NIT). (D) Shows the effect of agrin on DMD myotubes.

[0201] Figure 10 shows fluorescence intensity (FI) analysis of live-cells calcium imaging of innervated myotubes. (A, B, C) analysis of FI of myotubes co-cultured with mouse embryonic MNs. (D) 3D co-cultures of myotubes (stained in green by anti-myosin heavy chain antibody, and motoneurons, stained in red with anti neuro-filament antibody (upper panels) and with rhodamine labelled bungarotoxin (BTX) in lower panel, showing the formation of neuro-muscular junctions.

[0202] Figure 11 shows RT-qPCR showing relative levels of expression of various calcium regulatory genes in human immortalized myotubes. The relative expression level of mRNA (Log) of each gene was normalized to the reference gene (GAPDH) and then normalized to the expression of the same gene in healthy cells (indicated as 1). In each experiment, the expression of MyHC (I) was assessed to eliminate the effect of possible variation of differentiation on genes expression. Values are mean ± SEM, P<o.O5. Two- tailed P values were calculated between indicated columns. n=3. (J) western blot analysis of Cavi.i expression in healthy, DMD and DMD-U7 myotubes. (K) Immunofluorescent analysis of Cavi.i and dystrophin in the tibialis anterior of an mdx- NGS mouse transplanted with healthy human myoblasts. Figure 12 shows RT-qPCR showing relative levels of expression of various calcium regulatory genes in agrin-treated myotubes. Experimental conditions were as described in Fig 3. Values are mean ± SEM, P<o.O5. Two-tailed P values were calculated between indicated columns. n=3. (H.I) A Western blot analysis of Cavi.i (H) and alpha dystroglyan (DG) in agrin-treated DMD myotubes.

[0203] Figure 13 shows a schematic representation of calcium leaking due to the absence of dystrophin in early differentiated myotubes and its abolishment by agrin.

[0204] Figure 14 shows immunostaining of myotubes analysed by fluorescence microscope. The cells were plated on collagen I coated plates and then pushed to differentiate for ten days. They were then fixed and stained with anti-MyHC antibody (MyHC-green) to assess the level of differentiation, and anti-dystrophin (DYS-red) antibody for expressed dystrophin, and DAPI for staining the nucleus. Scale bar: 50 pM.

[0205] Figure 15 shows TEM analysis of early differentiated myotubes. (A, B, C) Differentiation and sarcomere formation in healthy, DMD, and DMD-U7 myotubes. (D, E, F) Level of differentiation and sarcomere maturation in healthy, DMD, and DMD-U7 myotubes co-cultured with mouse embryonic MNs. Scale bar: 0.5-2 pM.

[0206] Figure 16 shows fluorescence intensity (FI) analysis of live-cells calcium imaging of human MABs-derived myotubes. Non-innervated, early differentiated MABs-derived myotubes were loaded with Flou-4 AM calcium dye (excited at 388 nm). Images were acquired from live time-lapse videos of the myotubes at 37°C and 5% CO2: each graph represents a five second time frame of each video. (A) IF analysis of healthy myotubes. (B) IF analysis of DMD myotubes. (C) IF analysis of DMD-U7 myotubes.

[0207] Figure 17 shows the effect of Agrin (full molecule) in blocking Ca++ spikes (A), and its effect on the expression of Cavi.i (B). (C) shows immunostaining of dystrophic muscle fibres for dystrophin and / or Cavi.i. Upon transplantation in dystrophic muscles in vivo, fibers that express dystrophin also express Cavi.i. (D) shows the domain structure of the Agrin molecule and the domains at the C terminus are highlighted. Two truncated Agrin proteins are shown; C95 A4B8 (top), and C95 A0B0 (bottom). The C95 A0B0 Agrin fragment is able to abolish Ca+-i- flux, but the C95 A4B8 fragment is not.

[0208] Figure 18 shows a map of a vector suitable for expressing a truncated mouse agrin protein (SEQ ID No: 3).

[0209] Figure 19 shows DMD myogenic cells transduced with the lentivector expressing mini-agrin that is shown diagrammatically in Figure 18. (A) Shows cells expressing the m-Cherry transduction marker, and (B) shows the same cells, loaded with Fura-2, showing Ca++transients in some cells. (C) shows that Ca++ spikes in human DMD cells disappear after transduction with the LV expressing mini-agrin (for comparison, see, for example, Figure 9).

[0210] Figure 20 shows, by immunoflouresence, that Cavi.i (red) is expressed in healthy mucle cells (WT, panel A) but not DMD muscle cells (panel B). Expression of Cavi.i in

[0211] DMD muscle cells is restored by transduction with the Trans-skip LV expressing U7 snRNA (panel C), or with the LV expressing mini-agrin that is shown diagrammatically in Figure 18 (panel D).

[0212] Examples The inventors have found that cells in culture, including both immortalized DMD myoblasts and primary DMD MABs, show the occurrence of transient Ca++spikes that are not observed in healthy cells.

[0213] The inventors have found that these Ca++spikes are abolished by genetic correction leading to the re-expression of dystrophin, and by co-culture with mouse embryonic motoneurons. Surprisingly, the inventors also found that this effect can be replaced by agrin alone. Agrin stabilises the sensor Cav.1.1 that regulates the Ryanodine Receptor complex and abolishes Ca++spikes. These findings reveal a novel and early mechanism of Ca++dysregulation in DMD muscle cells. This mechanism offers a novel therapeutic strategy, in which truncated agrin may be used in the treatment of muscular dystrophy conditions characterised by aberrant calcium homeostasis, and in particular, the presence of calcium spikes.

[0214] The inventors used either Telomerase-Immortalised Myogenic cells (TIM) with frameshift mutation in exon 51, MyoD converted fibroblasts (22) or primary human hMABs from a DMD patient. The inventors generated a new mouse model that harbours a skippable mutation of exon 51. Myogenic cells were transduced with a lentiviral vector expressing the U7 snRNA, able to induce skipping of exon 51, to diffuse along the myofiber and permanently correct also dystrophic resident nuclei in the same area. The inventors show that this strategy multiplies the therapeutic effect both in vitro, by co-culturing DMD genetically corrected and uncorrected cells, and in newly regenerated muscle fibres in the in vivo DMD model, leading to abundant dystrophin production, in a therapeutically effective range. Materials and Methods

[0215] Cells

[0216] Myogenic cells

[0217] The telomerase-immortalised myogenic (TIM) cells used in examples 1-5 were maintained in 80% DMEM / RPMI199 at 4:1 ratio (Sigma), 20% FBS, 25 mg / mL Dexam (Gibco), 10 mg / mL gentamicin (Sigma), 25 mg / mL fetuin (Gibco) and lx insulin- transferrinselenium-X (Gibco), supplemented with 5 pg / mL hFGF and 500 pg / mL recombinant hEGF. hMABs were maintained in Megacell DMEM, 5 % FBS, 0.2% P- mercaptoethanol (Gibco), 1 % LGlut (Gibco), 1% Pen / Strep (Gibco) and 1% Non- essential amino acids (Sigma). MyoD-ER(T) Fibroblast were maintained in DMEM (Sigma), 20% FBS, 1% L-Glut (Gibco), 1% Pen / Strep (Gibco), imM sodium pyruvate (Gibco). TIM cells were differentiated on collagen (Gibco)-coated culture dishes using DMEM (Sigma), 10 mg / mL gentamicin (Sigma), lx insulin-transferrin-selenium-X (Gibco) while hMABs and MyoDER(T) Fibroblast were differentiated on collagen- coated culture dishes using DMEM (Sigma), 4% HS, 1% L-Glut (Gibco) and 1% Pen / Strep (Gibco). Both media were supplemented with 1.25 mM Forskolin (Sigma) to enhance myotube fusion.

[0218] The cell models used in examples 6-10 were human immortalized WT and DMD muscle cells produced in Vincent Mouly Lab (Institute of Myology, Paris). DMD cells are characterized with Exon 51 specific mutation. The DMD-U7 cells were generated in our lab through transducing DMD cells with a lentiviral vector expressing a small nuclear

[0219] RNA (U7 snRNA) engineered to skip exon 51. The human immortalized muscle cells were cultured in a mixture of Dulbecco’s modified Eagle’s medium (DMEM; SIGMA, USA; D5796-500ML) supplemented with 20% M199 media (199; Gibco, USA; 22340- 020), 20% fetal bovine serum (FBS; SIGMA, USA; F9665-500ML), 50 pg / mL Gentamicin (Gent; SIGMA, USA; G1272-10ML), 25 pg / mL Fetuin (FET; Gibco, USA; PSB10052), Human Fibroblast Growth Factor (hFGF) (50 ng / ml; Gibco, USA; 17105- 041), Human Epidermal Growth Factor (hEGF) (5 ng / ml; Gibco, USA; 2129284), Dexamethasone (0.2 pg / ml ; SIGMA, USA; D4902-100MG), and Insulin (5 pg / ml; Gibco, USA; 51500-056 10ML). The medium was kept at 4°C and used in future experiments.

[0220] Mouse primary neural cells (MNs)

[0221] Murine embryonic spinal cords were isolated from E12.5 embryos. The isolation procedures were conducted and processed under sterilized conditions using a dissecting microscope (ZEISS AX10.V16). Isolated spinal cords were promptly chopped into small pieces and digested in HBBS (1X; Gibco, USA; 14175-053) containing Dispase (2.4 U / ml; Gibco, USA; 17105-041) in a specific concentration for 25 minutes at 37°C in a shaking water bath. The cell suspension was then centrifuged for 10 minutes at 28oxg. A specific culture medium was used to maintain and culture isolated MNs. The medium was prepared as follow: the DMEM / F-12 medium (1X; Gibco, USA; 21331-020) was supplemented with 2% B-27 (1X; Gibco, USA; A35828-01), 1% L-GLUT (1X; Gibco, USA; 15140-122), 1% Penicillin / Streptomycin (P / S) (1X; Gibco, USA; 15140- 122), and 10 ug / ml Mouse Neuronal Growth Factor (1X; Gibco, USA; 17105-041). Later, the medium was kept at 4°C and used in future experiments. 3D culture of Human myotubes & mouse embryonic MNs

[0222] Human muscle cells were prepared in growth media and seeded on a layer of collagen- coated plates and incubated for overnight at 37 °C and 5% C02. Growth medium was replaced with differentiation media prepared with high glucose DMEM medium supplemented with insulin plus gentamicin and trace amount of serum to enhance cell differentiation into multinucleated myotubes. On day 6thof differentiation, extracted

[0223] MNs were mixed with matrigel prepared in MNs growth medium and added on top of myotubes.

[0224] Lentivectors The lentiviral vector U7#51T2AGFP derived from pCDH.EFK.MCH.T2A.GFP, (described as T2AGFP) was used, allowing co-expression of the reporter gene GFP and snRNA under the same promoter Human Elongation Factor-1 alpha (EF-ia).

[0225] Treatments Nimodipine (Sigma, USA, N-149) was prepared by dissolution in DMSO and stored at - 20 °C. Stock solution (10 mM) was diluted with DMSO to get a working concentration of 10 pM. Nitrendipine (Sigma, USA, N-149) was prepared by dissolution in DMSO and stored at -20 °C. Stock solution (10 mM) was diluted with DMSO to get a working concentration of 10 pM. Caffeine (Sigma, USA, C-0750) was prepared by dissolution in dH20 and stored at 4 °C. Stock solution (100 mM) was diluted with dH20 to get a working concentration of 10 mM. EGF powder (5 ng / ml; Gibco, USA; 2129284) was reconstituted with PBS at 50 ng / ml and stored as aliquots at -20 °C. The stock solution was freshly prepared to get a final concentration of 5 ng / ml. FGF powder (5 ng / ml;

[0226] Gibco, USA; 2129284) was reconstituted with PBS at 500 ng / ml and stored as aliquots at -20 °C. The stock solution was freshly prepared to get a final concentration of 50 ng / ml. Human recombinant agrin powder (50 pg; R&D system, USA; 6624-AG) was reconstituted with PBS at 300 pg / ml and stored as aliquots at -20 °C.

[0227] Immunofluorescence Staining Immunocytochemistry (ICC) in Examples 1-5: cells were stained using the following primary antibodies: anti-MyHC MF-20 (Development Studies Hybridoma Bank); anti- Dys MANDRA17 (Development Studies Hybridoma Bank); anti-GFP (Abeam). Secondary antibodies used were: anti-mouse Alexa Fluor 546 and antichicken Alexa Fluor 488 (Abeam). Nuclei were stained with DAPI (Sigma).

[0228] For immunofluorescence staining of injected and not injected TA muscles. Tissues were dissected, frozen in liquid nitrogen-cooled isopentane and cut on cryostat (LEICA CM 1850, Leica, Germany) into transverse and longitudinal cryostat sections. Primary antibody used, in addition to those mentioned above, were: anti-dystrophin DMD (Sigma), anti a-sarcoglycan (Atlas antibodies), antinNOS (R&D System) and anti-LAM A / C (Invitrogen). Secondary antibodies used were: anti-mouse Alexa Fluor 488, anti- rabbit Alexa Fluor 546 (Abeam). Nuclei were stained with DAPI (Sigma).

[0229] Immunocytochemistry (ICC) in Examples 6-10: cells were plated and differentiated on either collagen I coated 24 multi-well plates (Thermo Fischer Scientific, USA; 142475), washed with PBS 2x10 minutes and fixed in 4% paraformaldehyde (PFA) in PBS for 20 minutes at 4°C followed by washing with PBS 3x5 minutes at room temperature (RT) and incubation with 10% fetal bovine serum (FBS) prepared in PBS for 30 minutes at RT. Cells were then permeabilized with 1% BSA in 0.2% Triton in PBS for 30 minutes at RT followed by incubation overnight at 4 °C with primary antibodies prepared in the permeabilization buffer. The next day, cells were washed 3x10 minutes with 1% BSA in 0.2% Triton in PBS and incubated for 1 hour at 4 °C with the secondary antibody specific to the primary antibody host species. Images were acquired using a (ZEISS AX10) inverted microscope and processed and analysed using (Zeiss Zen Imaging software). Immunohistochemistry (IHC) in Examples 1-5: Haematoxylin and Eosin staining was performed according to the instruction provided by Sigma Aldrich (Haematoxylin H9627 and Eosin 230251, Sigma). Masson’s Tricromic was performed according to the instruction provided for Iron Stain kit (HT20 Sigma). Sirius Red staining was performed according to the instruction provided Piero Sirius red stain kit (Ab 150681, Abeam).

[0230] Immunohistochemistry (IHC) in Examples 6-10: for histological studies, tibialis anterior muscles were dissected and immediately frozen in liquid nitrogen in preparation for tissue sectioning. The samples were placed in optimum cutting temperature compound (O.C.T) (Thermo Fischer Scientific, USA; LAMB / OCT) and frozen in isopentane cooled in liquid nitrogen. 10-micron cross-sections were obtained at -18 °C using Lieca cryostat (Leica, USA; CM-1950). Sections were collected on gelatine-coated slides and incubated overnight at 4 °C with primary antibodies Mouse anti-CaVi.i (1:200; Thermo Fischer Scientific, MA3-920), Rabbit anti-Dystrophin (1:1000; Merckmillipore, 574777). The next day, sections were washed with PBS, then incubated for 1 hour at 4 °C with the secondary antibody specific to the primary antibody host species. Nuclei were counterstained with Hoechst stain (1:1000; Thermo Fischer Scientific). Images were acquired using a (ZEISS AX10) inverted microscope and processed and analysed using (Zeiss Zen Imaging software). Antibodies

[0231] In Examples 6-10: the following Primary and secondary antibodies were used at the following concentration: Mouse anti-TUJi (1:500; Biolegend, MMS-435P), Mouse anti- MyHC (1:3; Developmental Studies Hybridoma Bank MF20), a-Bungarotoxin (BTX, Sigma-Aldrich, T0195), Mouse anti-CaVi.i (1:200; Thermo Fischer Scientific, MA3- 920), Goat anti-Mouse (1:500; Thermo Fischer Scientific), Goat anti-Rabbit (1:500;

[0232] Thermo Fischer Scientific). Nuclei were counterstained with Hoechst stain (1:1000;

[0233] Thermo Fischer Scientific).

[0234] In Situ The In Situ assay was performed according to the instruction provided with ACDbio RNAscope Multiplex Fluorescent Reagent Kit. The signal was amplified using specific probe for GFP, LAM A / C and the U7 snRNA detection, specifical design for us by ACDbio. Microscopic imaging

[0235] Live-Cells Ca++imaging: Intracellular calcium status was measured using the greenfluorescence calcium indicator Fluo-4 AM (Invitrogen, USA, F14201), an improved version of Fluo-2 and Fluo-3. After medium removal, differentiated myotubes were loaded with 2 pM of Fluo-4 AM prepared in DMSO and diluted with fresh differentiation media. The cells were incubated for 1 hour at 37 °C followed by washing with differentiation media for 2x5 minutes. Cells were incubated at 37 °C during imaging and images were acquired using (ZEISS AX10) inverted fluorescence microscope at 4.2 frames / second with 488 nm laser line excitation. The analysis was performed using ImageJ software by measuring the fluorescence intensity of acquired images over seconds and normalized the values to fluorescence intensity at o times.

[0236] Calculations and figures were done using GraphPad Prism (8.4.2). Transmission electron microscope (TEM): Healthy, DMD, and DMD-U7 myoblasts were platted on collagen I coated ACLAR embedding film for overnight at 37 °C and 5% C02. Growth medium was then replaced with the differentiation medium for promoting cells fusion and myotubes formation. Ten days later, cells were fixed with a fixative solution contains 4% PFA, 2.5% GA, 0.1m MEPES. Samples were processed at the Bioimaging Core facility at the University of Manchester.

[0237] RT-PCR and qRT-PCR In Examples 1-5: total RNA was extracted using TRIzol (Invitrogen) according to the manufacturer’s instruction. RNA was reverse transcribed using cDNA preparation kit (Thermoscientific). For RT-PCR reaction the following primers were used: hGAPDH, hMyHC, and hDYS. For qRT-PCR reaction the following primers were used: hGAPDH, hMyHC, GFP, hDYS 47 / 52, and hDYS 53. Sequences of the primer pairs are outlined in Table 1. The cDNA for sequencing was extracted according to the instruction provided with QIAquick gel extraction kit (Qiagen), and sequence was performed by GeneWiz.

[0238] Table 1: Primers (forward and reverse) used with qPCR analysis in Examples 1-5.

[0239] In Examples 6-10: RNA was extracted from differentiated myotubes. After medium removal, cells were washed twice with 1X PBS at room temperature. After that, 500 pl of TRIzol reagent (Life Technologies, USA, 15596018) was added for cells harvesting and homogenisation, following steps are stated in the manufacture’s protocol. Extracted RNA was re-suspended in nuclease-free water (Invitrogen, USA, 10977-035) followed by immediate cooling on ice. Eluted RNA was then proceeded for quantity and purity assessment or stored at -8o°C. To synthesize cDNA, starting with genomic digestion, a standardized amount of 1 pg RNA per sample was treated with DNase BUFFER (B43, Thermo Fisher Scientific, USA), DNase (EN0525, Thermo Fisher Scientific, USA), and nuclease-free water (Invitrogen, USA, 10977-035). Samples were then incubated at 37°C for 30 minutes, followed by adding 1 pl of 0.5 M EDTA (R1021, Thermo Fisher Scientific, USA) and 10 minutes incubation at 65°C. in the following steps, retro-transcription was performed by adding random hexamer primers (SO142, Thermo Fisher Scientific, USA) plus nuclease-free water (Invitrogen, USA, 10977-035) to the digested product followed by 5 minutes at 65°C. A 12 pl of the product was added to the PCR reaction tubes contains 5X Reaction buffer (5X RT) (00557616, Thermo Fisher Scientific, USA), RNase ribolock RNase Inhibitor (EOO381, Thermo Fisher Scientific, USA), dNTP mix (R0191, Thermo Fisher Scientific, USA), RevertAid Reverse Transcriptase (200 V / pl) (EP0442, Thermo Fisher Scientific). Samples were then amplified using (Mastercycler, epgradient S).

[0240] RT-qPCRs were carried out using FastStart Essential DNA Master Mix (06924204001, Roche, Swiss) in a Roche LightCycler 96 system. Samples were prepared by adding 10 ng of cDNA to the SYBER Green mater mix and 10 pM of forward and reverse primers (Table 2). Absolute quantification of each target was performed using a standard curved as a reference in Roche Light Cycler software version 1.5. Table 2: Primers (forward and reverse) used with qPCR analysis in Examples 6-10.

[0241] Western Blot

[0242] In Examples 1-5: cells were lysed in RIPA buffer (lomM TRIS, 100 mM NaCl, imM EDTA, 1% Triton, 10% Glycerol, 0.1% SDS and 1% protease inhibitor). Protein concentration was determined with the Bio-Rad Protein Assay. Proteins were separated in 4-15% gradient, pre-casted SDS PAGE and then transferred onto a nitrocellulose membrane using standard protocol. The blots were incubated with the following antibodies: anti-MyHC MF-20 (Development Studies Hybridoma Bank); anti-Dys MANDRA17 (Development Studies Hybridoma Bank); anti-GAPDH (Abeam) and anti- LAM A / C (Invitrogen). Proteins were visualized by an enhance chemiluminescence method (Thermoscientific) according to the manufacturer’s instructions.

[0243] In Examples 6-10: cells were washed once with cold PBS supplemented with cocktail protease inhibitor (78446, Thermo Fisher Scientific, USA), and then dislodged using cells scrapers. The cell suspension was then centrifuged for 20 minutes at 5000 rpm and a temperature of 4°C. The Cells pellet was resuspended with a RIPA buffer containing (too Mm TRIS, lOoMm NaCl, 1 Mm ETDA, 1% Triton, 10% Glycerol, 0.1% SDS), supplemented with cocktail protease inhibitor, and followed by 10 minutes centrifugation at 5000 rpm. A 30-50 ng of total proteins were loaded on 4-12% ready- made gels (Novex, NPO322BOX). Both 10X NuPAGE sample reducing agent (Invitrogen, NP0004) and 4X NuPAGE loading buffer (Invitrogen, NP0006) were added to samples and heated at 95°C for 10 minutes. Gels were run in a 1X NuPAGE SDS running buffer (Invitrogen, NP0001) mixed with NuPAGE antioxidant (Invitrogen, NP0005). Proteins were transferred onto nitrocellulose membrane in the presence of 1X transfer buffer (Invitrogen, NP0006-1). To avoid non-specific protein binding, the membrane was incubated with 5% non-fat dry milk (Marval) prepared in 1X TBST for 1 hour at room temperature. It was then incubated with the desired antibody for overnight at 4°C. In the next day, the blot was washed two times every 15 minutes with 1X TBST. Next, the blot was incubated with the appropriate HRP-conjugated secondary antibody (DAKO, P0448) for 1 hour at room temperature. In Vivo

[0244] In Examples 1-5: NGS mice carrying a skippable mutation of exon 51 (NOD.Cg- Prkdcscid IlsrgtmiWjl Tg(HLA-A / H2-D / B2M)iDvs / SzJ) obtained from Jackson Laboratory (USA). Mice were intramuscularly injected into Tibialis Anterior muscle (TA) and / or Gastrocnemius (G) and / or Quadriceps (Q) with either 5x105 TIM cells or hMABs; contro-lateral muscles were used as not injected control ones. Mice were culled after 30 days since injection. Scid mice carrying a skippable mutation of exon 23 (C57BL6 / ioScSn.Cg-PrkdcscidDMDmdx / J) were obtained from Jackson Laboratory (USA). Mice were subcutaneously injected in the back with 5x10sTIM cells mixed in a Matrigel (Corning) plug. The use of animals in this study was authorized (license PDB0CF0C2); mice were fed ad libitum and allowed continuous access to tap water.

[0245] In Examples 6-10: the female pregnant CD1 mice and NGS-DMD mice carrying a skippable mutation of exon 51 (NOD.Cg-PrkdcSC!dIl2rgil"1Wi' g (HLA-A / H2- D / B2M)IDVS / SZJ) were obtained from Jackson Laboratory (USA) and housed in the University of Manchester animal facility. All experiments and procedures were approved by the University of Manchester Animal Welfare and Ethical Review Body (AWERB). All Experiments were conducted under the University of Manchester Animal Care guidelines and in accordance with UK Home Office Regulations (ASPA 1986) and under the project licence PDB0C0.

[0246] Eight months NGS-DMD mice were injected with WT human cells. The NGS-DMD mice were prepared and injected at the Tibialis Anterior and they were checked regularly. The number of injected cells were assessed and optimised in previous pilot studies in our lab.

[0247] Motility assay

[0248] 16 males NOD.Cg-Prkdcscid I12rgtmiWjl Tg(HLA-A / H2-D / B2M)iDvs / SzJ split into four groups (4 WT mice, 4 DMD untreated mice, 4 DMD mice transplanted with WT TIM cells, 4 DMD mice transplanted with DMD-U7 cells). Mice were intramuscularly injected into the TA, Gas, Quad, both the lower limbs with 0.5x1c)6. After 30 days mice were put alone for 5 days in a home cage containing a running -wheel and they will be let run and move naturally. Their activity was monitored continuously.

[0249] Statistical Analysis In Examples 1-5: all the in vitro experiments were repeated three times, each in triplicate, to ensure to measure statistically significant differences upon different experimental treatments. However, for cross correction experiments, the amount of protein expressed was quantified by densitometric analysis of WB and compared the different samples by t-test testing with GraphPad Prism 9.0 software. In vivo experiments were conducted on experimental groups of 6 animals. This represents the best compromise between the 3R guidelines and the need to have reproducibility and consistency in the results. All the in vivo data were analysed with GraphPad Prism 9.0 software. For comparison and analysis between the different animals, t-test testing was used.

[0250] In Examples 6-10: all figures and statistical analysis were performed using Excel software (Microsoft) or GraphPad Prism (8.4.2). Data were expressed as the mean ± standard deviation (SD). Statistical analysis was performed, and the differences were considered significant if the p-value was <0.05.

[0251] Example 1 - Cell-mediated exon skipping induced by lentiviral vector U7# HT2AGFP. To work with a long term expandable myogenic cell line, isolate from a DMD patient and to avoid inter-patient variability, the inventors chose to work with a line of immortalised myogenic cells carrying a deletion at exon 48-50 (A48-50). The cells (referred to as TIM) were immortalised with telomerase and CDK4, can be indefinitely expanded and even at late passages maintain reproducible and robust myogenic differentiation. The inventors also used a similarly immortalised WT type control line, with similarly high differentiation potency. The inventors confirmed results also in primary MABs from a DMD patient and in DMD fibroblasts, myogenically converted with an Adenovector expressing MyoD. To correct the genetic defect via exon skipping, the inventors built a 30generation, SIN lentiviral vector (LV) expressing, under the transcriptional control of the human EFi-a ubiquitous promoter, the U7 small nuclear RNA, engineered to skip the acceptor and donor splice sites of exon 51, thus causing restoration of the correct reading frame. To facilitate the detection of transduced cells the vector also expresses GFP after a T2A site. A control vector, only expressing GFP was also produced. Fig. 1C and Fig. 6C show that at a low MOI (0.8) approximately 8o% of cells were transduced by the U7#51T2AGFP LV and more than 90% by the control GFP LV, as revealed by FACS analysis (Fig. 1C). Co-culture of transduced and non-transduced cells at a ratio of 1:10 and 1:30 led to a progressive reduction of fluorescence (Fig. 6C).

[0252] The inventors then tested the ability of the U7#51T2AGFP LV to induce cell mediated exon skipping and thus a shorter version of dystrophin. Total RNA was extracted from DMD differentiated myotubes, either corrected or not with the LV and the generated cDNA was used to perform a RT-PCR. A human muscle biopsy was utilised as a positive control to show the size of the full-length Dystrophin (goobp). A band of approximately 4oobp was observed in all samples from DMD TIM myotubes. A second dystrophin band of approximately 25obp (size of the shorter Dystrophin) was visible only in the cells transduced with the LV showing the dystrophin mRNA produced through exonskipping by the U7#51T2AGFP (Fig. 1A). Sequencing of the 25obp band showed that it corresponds to A47-52 (Fig. 1B). This result confirms the ability of U7#51T2AGFP to induce exon-skipping.

[0253] To test the ability of A47-52 dystrophin mRNA to produce a protein, immunofluorescence (IF) staining was carried out to visualize the dystrophin protein in DMD TIM cells and genetically corrected DMD TIM cells (herein referred to as DMD-

[0254] U7). To test whether the U7 snRNA would diffuse to neighbouring DMD nuclei and thereby also correct the dystrophin produced in these nuclei, DMD-U7 cells were cocultured with DMD TIM cells at a 1:10 and 1:30 ratio, mimicking an in vitro cell engraftment of 10% and 3% respectively (Fig. 1C, Fig. 6C). Figure 1D shows merged images of the 1:10 and 1:30 co-cultures where the expression of Dystrophin (red) extends well beyond the areas expressing GFP (green) in the cytoplasm of multinucleated myotubes as indicated by arrows. Individual fluorescence channels are shown in Fig. 6A-B. This result demonstrates the cross correction of dystrophin by U7#51T2AGFP due to its diffusion to neighbouring dystrophic nuclei.

[0255] The inventors next tested whether snRNA can diffuse and induce exon skipping in the neighbouring dystrophic nuclei by qRT-PCR on DMD-U7 cells and DMD TIM cells cocultured at a 1:10 and 1:30 ratio (Fig. 1E). MyHC expression was used to quantify the level of the differentiation and no significant difference was noted across all the conditions (Fig. 1E), confirming the IF analysis (Fig. 6A-B). GFP expression was directly compared with the dystrophin expression to evaluate the level of dystrophin produced by cross-correction. GFP expression decreased in proportion with the ratio of transduced vs non-transduced cells. In contrast, the 1:10 and 1:30 co-cultures showed higher expression of dystrophin in comparison to GFP. Dystrophin expression of almost 40% (4-fold increase) and 20% (about 7-fold increase) was observed in cells co- cultured at 1:10 ratio and 1:30 ratio, respectively, (Fig. 1E), confirming the ability of the snRNA produced by U7#51T2AGFP to induce cross correction of neighbouring nuclei, thus amplifying dystrophin production.

[0256] Example 2 - Intra-cellular diffusion of snRNA produced by U7#51T2AGFP. To demonstrate the possible intra-cytoplasmic diffusion of the U7 snRNA from the original transduced nucleus into neighbouring dystrophic nuclei of the same myotube, the inventors conducted double in situ hybridization with specific mRNA probes. One probe was directed against the U7 snRNA (Red) and the other against GFP mRNA (Green). Whereas the GFP mRNA was localised in the cytoplasm surrounding a single, donor nucleus (green arrows), the snRNA diffused along most of the available cytoplasm in the same myotube (red arrows) and was observed where no GFP mRNA expressing nuclei could be detected (Fig. 2A). These results demonstrate the ability of the U7#51T2AGFP snRNA to diffuse along the myotube and to enter also in the neighbouring nuclei which had not been genetically corrected.

[0257] Example - Dystrophin protein expression induced by U7 snRNA in vitro.

[0258] To quantify the protein expression produced by nuclear cross correction, the inventors performed a Western Blot assay (WB). Proteins were extracted from co-cultures of DMD-U7 and DMD TIM cells at 1:10 and 1:30 ratio. As a control, WT TIM cells were co-cultured with DMD TIM cells under the same conditions. The level of differentiation was quantified by analysing Myosin Heavy Chain (MyHC) that showed no significant difference in expression among the different samples (Fig. 2B). As expected, DMD TIM cells did not produce any dystrophin protein whereas dystrophin protein expression was detected in the WT TIM cells and, to a lower level, in DMD-U7 cells. However, at 1:10 dilution dystrophin expression was already higher in co-cultures of DMD TIM with

[0259] DMD-U7 cells than in co-cultures of DMD TIM with WT TIM cells. At 1:30 dilution, dystrophin was still clearly detectable in co-cultures of DMD TIM and DMD-U7 cells but not in co-cultures with WT TIM cells. The quantification of the WB (densitometry) showed a robust dystrophin expression in DMD TIM cells co-cultured with DMD-U7 cells at 1:10 and 1:30 ratio (50% and 20% of WT TIM cells) in comparison with the co- culture WT TIM cells and DMD TIM cells at the same ratio, showing only a 1.2-fold increase.

[0260] To determine whether exon skipping cross-correction via snRNA is effective also with other types of myogenic cells, the inventors co-cultured DMD-U7 and WT TIM cells with two different DMD myogenic cell types, either Mesoangioblasts (DMDhMABs) or MyoD-converted Fibroblasts (DMDMyoDERTFbs) derived from DMD patients. Dystrophin protein expression was detected in all the different co-cultured samples (Fig. 2C, 2D), always with the same trend. The quantification of the protein expression indicated an increase in the dystrophin protein expression with DMD-U7 cells in comparison with the WT TIM cells in all the co-cultures. This result confirms that the phenomenon is not specific to immortalised cells, and in all conditions dystrophin protein production is not exclusively due to the WT nuclei but is produced also in neighbouring dystrophic nuclei corrected by the U7 snRNA produced by genetically corrected nuclei.

[0261] These data demonstrated that the snRNA-based strategy enhances the production of dystrophin protein through cross-correction of dystrophic neighbouring nuclei contained in the same myotube.

[0262] Example 4 - Dystrophin protein expression induced by U7 snRNA in vivo.

[0263] The inventors next analysed the efficiency of the exon-skipping strategy in vivo, where transplanted human cells will fuse with resident dystrophic myogenic cells to form hybrid regenerating myofibers. Also, mouse cells could be corrected as also acceptor and donor splice sites had been humanised. In a preliminary experiment, the inventors repeated the in vitro dilution experiments by co-culturing DMD-U7 and DMD TIM cells at 1:10 ratio in a Matrigel plug implanted under the dorsal skin of mdx / SCID mice (n=3). 100% DMD-U7 cells were used as control in a similar Matrigel plug implanted under the dorsal skin on the contra-lateral site of the same mouse (Fig. 8A). After a month, the inventors recovered plugs containing multinucleated myofibers only in two mice. IF analysis showed extensive dystrophin expression, even in the sample at 1:10 dilution, while WB analysis showed virtually the same amount of dystrophin protein detected both in the plug containing only DMD-U7 cells and in those containing only 1 / 10 of DMD-U7 and 9 / 10 of DMD TIM cells (Fig. 8B). Currently used mdx mice carry a mutation in exon 23 that is not amenable to cell- mediated exon skipping by a snRNA engineered to skip exon 51. For this reason, the inventors generated a NSG mouse carrying a skippable mutation of exon 51. The mice (defined for simplicity NSG-mdx-Asi) are completely immune deficient and develop a muscular dystrophy like the classic mdx, with a peak of degeneration about the first month of age and a progressive but relatively mild dystrophic phenotype that worsens after one year of age (Fig. 7A).

[0264] The inventors injected the same number (5x10scells) of either DMD-U7 or WT TIM cells in NSG-mdx-A5i. In other animals, the inventors also injected either DMD-U7 hMABs or WT hMABs. To evaluate Dystrophin correction at the protein level, the inventors performed WB analysis of the injected Tibialis Anterior (TA), 1 months after a single intramuscular injection. WB analysis showed restoration of dystrophin expression, both with TIM cells and hMABs (Fig. 3A and 3B), in the skeletal muscle. As shown in Fig. 3A the TA injected with DMD-U7 cells showed approximately 60% of protein level detected in WT muscle while the TA injected with WT TIM showed protein restoration but a lower-level in comparison with DMD-U7 cells (approximately 20%) suggesting the ability of the snRNA to induce expression also in the neighbouring nuclei which had not genetically corrected. A similar result was observed in the TA injected with DMD-U7 hMABs and WT hMABs (Fig. 3B) respectively with around 50% and 10% of dystrophin expression. 30% of normal level of dystrophin is estimated to provide therapeutic benefits for DMD patients. IF of muscle sections from injected TA with either WT TIM or DMD-U7 cells confirmed recovery of dystrophin (Fig. 3C) a- sargoclycan and nNOS (Fig. 3D) expression.

[0265] To test whether the very high level of dystrophin expression was due to the genetic correction of resident nuclei, mediated by intra-cytoplasmic diffusion of the U7 snRNA from the nucleus of the original transduced-injected cell, the inventors conducted double in situ hybridization with specific mRNA probes, one directed against the snRNA (Red) and the other against human LAM A / C (Green). The snRNA diffused along most of the available cytoplasm in the same myofiber also where no LAM A / C mRNA expressing nuclei could be detected (Fig. 4A). These results demonstrate the ability of the U7#51T2AGFP snRNA to diffuse along the myofiber and to enter the neighbouring nuclei which had not been genetically corrected, thus explaining the efficacy of a single cell injection in restoring the protein expression of dystrophin, up to around 60% of WT levels, a result only achievable so far with injection of AAV. The inventors reasoned that, in view of the mechanism discussed above, lowering the dose of cells injected should lead to a less than proportional decrease in the amount of dystrophin produced. To test this hypothesis, the inventors injected in the TA of NSG- mdx-Asi mice 5, 2.5 and 1.25 and 0.62 x to5cells. After one month, the inventors performed a WB analysis of the extracted proteins. Fig. 4B shows that even at the lowest cell dose (12.5% of the maximum), about 50% of the maximum dose of protein was detected as normalised by the analysis of LAM A / C that detects the number of human cells present.

[0266] Example - Motility assay.

[0267] The inventors also evaluated whether the increased production of dystrophin would restore mouse motility. The spontaneous motility of NSG WT mice and NSG-mdx-Asi mice, some of which had transplanted with either WT TIM or DMD TIM or DMDU7 cells, was monitored in real time for 4 consecutive days. NSG-mdx- 5i mice showed a 50% reduction of motility when compared with NSG WT mice (Fig. 3C); the same mice, transplanted with DMD-U7 cells, showed a recovery of spontaneous motility up to 70% of WT animals while mice transplanted with WT TIM cells only showed a very small recovery (Fig. 3C).

[0268] Example 6 - Lack of dystrophin causes Ca2+leaking in early differentiated DMD myotubes.

[0269] The inventors found that non-innervated, early differentiated DMD myotubes, but not healthy myotubes, show repetitive spikes as revealed by the Ca++indicator Fura-2. (Fig. 9A, 9B). Interestingly, the DMD-U7 myotubes showed no detectable Ca++leaking (Fig.

[0270] 9C), thus proving that the expression of dystrophin is necessary and sufficient to prevent the occurrence of Ca++spikes. Moroever, DMD mesoangioblast-derived myotubes were also found to show the existence of transient Ca++spikes that are not observed in healthy mesoangioblast-derived myotubes or in myotubes derived from DMD, genetically corrected myotubes (Fig. 16, A-C).

[0271] Example 7 - Calcium regulatory genes are downregulated in DMD myotubes.

[0272] To determine whether the alteration of Ca++homeostasis maybe caused by changes in the expression of genes encoding for ion channels or sensors, the inventors analysed the level of the relative mRNA expression using quantitative PCR analysis. The inventors observed significant reduction of calcium voltage-gated channel subunit alpha 1S (CACNAiS) that encodes the skeletal muscle L-type, alpha i subunit of the DHPR (Cavi.i), and sodium voltage-gated channel alpha subunit 4 (SCN4A) genes in DMD myotubes (Fig. 11A-B). In contrast, these genes were expressed in DMD-U7 myotubes to a level comparable with healthy cells (Fig. 11A-B). On the other hand, RyRi and transient receptor potential cation channel subfamily C (TRPC1) were slightly upregulated in DMD cells with no significant changes in DMD-U7 myotubes compared with the healthy myotubes (Fig. 11C-D). Finally, the inventors found that beta I and A subunits of the AChR were reduced in DMD myotubes. The L-type Voltage Gate Calcium Channel (VGCC) Cavi.i isoform is located on the sarcolemma and T-tubules where it interacts and regulates the RyRi; it is exclusively expressed and highly regulated in skeletal muscle. To test whether Cavi.i contributes to Ca++leaking, the inventors treated DMD myotubes with specific inhibitors such as nimodipine and nitrendipine that abolished the Ca++spikes. (Fig. 9F-L). On the other hand, directly stimulating calcium release from RyRi showed no significant alterations in DMD compared with healthy cells (Fig. 9H-J).

[0273] Example 8 - Ca++spikes do not induce muscle contraction. The effect of motoneurons.

[0274] The inventors observed that under these culture conditions, the Ca++spikes were not followed by muscle contraction and hypothesised that this was due to the absence of a developed and mature contractile apparatus. To test this, the inventors performed transmission electron microscopy (TEM) on healthy, DMD and DMD-U7 myotubes (Fig. 15A-C). In all cases, the inventors observed initial assembly of thick and thin filaments and early forming Z lines which explains why the inventors did not see any twitching myotube, despite Ca++spikes. The inventors then reasoned that co-cultures with motoneurons may promote myotube maturation and the onset of a contractile activity. Indeed, Bungarotoxin stained the formation of neuro-muscular junctions (NMJ) (Fig. 10D) and TEM showed that sarcomerogenesis was advanced but not yet to the point that myotubes may contract. Surprisingly, the inventors observed the disappearance of the Ca++spikes in innervated DMD myotubes, under conditions where motoneurons were not stimulated. No Ca++spikes appeared in healthy or DMD-U7 myotubes (Fig. 10A-C). The fact that Ca++spikes disappeared in the absence of motoneuron stimulation, suggested the possibility that molecules released at the NMJ level may be responsible for the effect observed. Example o -Agrin abolishes Ca++spikes and restores normal expression of calcium regulatory proteins in DMD myotubes.

[0275] Agrin is involved in the formation of the NMJ. Surprisingly, when added to cultures of DMD myotubes, agrin, at any concentration tested in the pM range, abolished Ca++spikes in DMD myotubes suggesting that the effect of muscle innervation occurs via the release of agrin (Fig. 9D). To understand the mechanism of agrin action, the inventors investigated whether it effects the expression and / or the stability of proteins involved in Ca++homeostasis. The inventors found that the expression of CACNA1S and SCN4A SCN4A was restored to normal level in agrin-treated DMD myotubes (Fig. 12A-B). The inventors also found that both RyRi and TRPCi were expressed at normal levels in agrin-treated DMD myotubes (Fig. 12C-D). These results show that agrin has a role in regulating the expression of Ca++homeostasis genes in skeletal muscle, thus explaining its effect on Ca++spikes. Example 10 - Either dystrophin or agrin are essential for Cavi.i expression.

[0276] Western blot analysis showed that Cavi.i expression was significantly reduced in DMD myotubes with restored expression in DMD-U7 myotubes (Fig. nJ). These findings are consistent with what has been observed at the mRNA level suggesting that the absence of dystrophin impacted negatively on the structure and expression of Cavi.i at very early stages of differentiation. Surprisingly, the expression of Cavi.i was restored in DMD myotubes treated with agrin (Fig. 12H).

[0277] To confirm the role of dystrophin in regulating Cavi.i expression also in vivo, the inventors analysed the Tibialis Anterior (TA) of NSG-mdx mice, previously transplanted with healthy human myoblasts (Fig. 14). The inventors found that only dystrophin expressing myofibers show strong expression of Cavi.i which is barely detectable in DMD non-corrected myofibres (Fig. 11K). These results show that the expression of Cavi.i is either directly or indirectly linked to the expression of dystrophin in early as well as in mature muscle fibres.

[0278] Example 11 - CQ AOBO truncated agrin protein abolishes Ca++flux

[0279] Figure 17A-C shows that the expression of full length agrin is capable of blocking Ca++spikes and upregulating Cavi.i in vitro and in vivo. Upon transplantation in dystrophic mucles in vivo, fibers that express dystrophin also express Cavi.i (Figure 17C). The domain structure of the agrin protein is shown in Firgure 17D. The domains at the C terminus of the agrin protein are highlighted to distinguish the A / y and B / z insertion sites. As shown in Figure 17E, expression of the C95 A0B0 truncated agrin protein fragment was found to abolish Ca++flux (left hand panel), but this effect was not observed upon expression of the corresponding C95 A4B8 truncated agrin protein (right hand panel).

[0280] Example 12 - Mini-agrin protein expression abolishes Ca++flux As shown in Figure 9, the inventors found that non-innervated, early differentiated DMD myotubes, but not healthy myotubes, show repetitive spikes as revealed by the Ca++indicator Fura-2. (Fig. 9A, 9B).

[0281] The inventors found that Ca++spikes disappear after transduction of DMD myogenic cells with a lentivector expressing mini-agrin (Fig. 19). This is consistent with the finding in Example 11 that C95A0B0 mini-agrin prevents calcium spikes. DMD myotubes transfected with mini-agrin showed no detectable Ca++leaking (Fig. 19B).

[0282] The absence of Ca++spikes in the cells transduced with an integrating lentiviral vector indicates that correction may be permanent and suggests that expressing mini-agrin in

[0283] DMD cells can permanently prevent transient Ca++spikes.

[0284] Example 13 - Mini-agrin protein expression or U snRNA induces Cavi.i expression The inventors observed reduced expression of Cavi.i in DMD myotubes (Fig. 11A-B), and found that the expression of full length agrin is capable of up regulating Cavi.i in vitro and in vivo, as described in Example 11.

[0285] Consistent with these findings, the inventors also found that expression of Cavi.i in

[0286] DMD muscle cells is restored by transduction with a lentivector expressing U7 snRNA (Fig. 20C), or by transduction with a lentivector expressing mini-agrin (Fig. 20D).

[0287] Discussion

[0288] The inventors observed the occurrence of frequent and repeated Ca++spikes in different types of DMD myogenic cells but not in their healthy counterparts. The relationship of these spikes with the absence of dystrophin had not previously been investigated or reported. The spikes were observed to occur at an early stage of differentiation, when sarcomerogenesis is incomplete, and thus the cells cannot contract. This allowed the phenomenon to be studied independently from any mechanical effect consequent to contraction. Inhibitors of voltage dependent Ca++channels such as nimodipine or nitrendipine, abolished the spikes, indicating that the DHPR-RyRi complex are involved in this leakage when dystrophin is not present on the muscle membrane.

[0289] Indeed, when cell-mediated exon skipping restored dystrophin expression, Ca++spikes disappeared.

[0290] To promote myotube maturation in culture, the inventors developed a 3D co-culture system with embryonic mouse motoneurons that are physically separated from myotubes by a Matrigel layer but project axons that contact the myotube surface and induce assembly of an Ach receptor as revealed by bungarotoxin staining. Under these conditions, and in the absence of any stimulation of the neurons, the Ca++spikes disappeared. The inventors therefore hypothesised that the disappearance of Ca++spikes is due to the secretory and not the electrical activity of motoneurons. The inventors therefore investigated agrin, which is involved in regulating the assembly of the Ach receptor and provides a link with the basal lamina in congenital muscular dystrophy MDC1A due to a mutation in laminin 211. Investigating membrane proteins whose expression could be reduced in the absence of dystrophin and restored in DMD-U7 cells, the inventors observed that the Ca++sensor CaVi.i was dramatically reduced in DMD cells and expressed at the same if not higher level than healthy cells after genetic correction. CaVi.i is the only isoform expressed in skeletal muscle and it is part of the DHPR complex that in turn regulates Ca++fluxes through RyRi. The inventors found that the exposure of DMD myotubes to agrin dramatically increases the expression of CaVi.i which correlated with the disappearance of Ca++spikes. Therefore, both dystrophin and agrin prevent Ca++spikes and regulate expression of the Ca++sensor CaVi.i. Based on the present findings, the inventors propose a model that is illustrated in Fig. 13. The present findings suggest that agrin, like dystrophin, stabilizes dystroglycan on the membrane, which in turn directly or indirectly maintains CaVi.i in the complex. DHPR and RyRi form a complex that regulates Ca++fluxes from internal stores. In the absence of dystrophin, reduced expression of CaVi.i deregulates DHPR thus leading to deregulated Ca++fluxes. Re-expression of dystrophin or addition of agrin lead to reexpression of CaVi.i and abolish Ca++spikes. Of note, when healthy human myogenic cells are injected into the muscle of an NSG-DMD mouse, only dystrophin expressing fibres express CaVi.i.

[0291] In studies in which mesoangioblasts are used as donor cells to fuse with regenerating muscle fibres and produce the U7 small nuclear RNA to induce dystrophin exon skipping, the inventors found that the U7 snRNA diffuses along the muscle fibres and enters neighbouring nuclei, thus amplifying dystrophin production by several fold, and thereby providing the therapeutic effect. In view of this clear and surprising finding that U7 snRNA delivered by mesoangioblast cell therapy can diffuse along the muscle fibres, the inventors have identified the use of mesoangioblasts as an advantageous vehicle in a cell therapy approach for the delivery of agrin for use in treating muscular dystrophies characterised by the presence of calcium spikes, and specifically to reduce or prevent calcium spikes in these conditions. Agrin is expressed at the NMJ, but when administered by mesoangioblast cell therapy will diffuse along the muscle fibres to induce the expression of CaVi.i and abolish Ca++ spikes.

[0292] Agrin is a large molecule and its heparan sulphate chains, with a strong negative charge may limit its diffusion throughout the muscle fibres. As mini-agrin has been shown to perform this function and is a small and diffusible molecule, functional, truncated agrin proteins provide an even more potent tool to abolish Ca++leakage than agrin.

Claims

Claims1. A mesoangioblast composition, wherein the mesoangioblasts (MABs) are engineered to express a truncated agrin protein.

2. A composition as claimed in claim 1, wherein the donor from which the MABs are derived is:(a) a subject having a muscular dystrophy disorder; or(b) a healthy subject.

3. A composition as claimed in claim 2, wherein the muscular dystrophy disorder is a muscular dystrophy disorder characterised by aberrant calcium homeostasis.

4. A composition as claimed in any of the preceding claims, wherein the truncated agrin protein is a truncated agrin protein having a length, based on the number of amino acid residues, that is 4-40% of the length of a native agrin protein.

5. A composition as claimed in any of the preceding claims, wherein the truncated agrin protein is a truncated agrin protein comprising or consisting of between 1 and 12 domains, each of which is derived from a domain that is present in a native agrin protein.

6. A composition as claimed in claim 5, wherein each of the one or more domains of the truncated agrin protein has at least 90% sequence identity to a corresponding domain present in a native agrin protein.

7. A composition as claimed in claim 5 or 6, wherein each of the plurality of domains is derived from a domain that is present in the C terminal portion of the native agrin protein.

8. A composition as claimed in any of the preceding claims, wherein the truncated agrin protein is an A0B0 agrin isoform.

9. A composition as claimed in any of the preceding claims, wherein the truncated agrin protein:(a) is capable of binding to a-dystroglycan;(b) is not capable of inducing aggregation and / or phosphorylation of the acetylcholine receptor (AChR); and / or(c) is capable of reducing or abolishing cytosolic calcium spikes when expressed in cultures of DMD muscle fibres. to. A composition as claimed in any of the preceding claims, wherein the MABs are engineered to additionally express a small nuclear RNA (snRNA) capable of inducing exon-skipping in a dystrophin pre-mRNA transcript.

11. A composition as claimed in claim io, wherein the snRNA is a modified uridine- rich 7 (U7) snRNA.

12. A composition as claimed in claim 11, wherein the snRNA sequence consists of or comprises the RNA sequence of SEQ ID NO: 1.

13. A composition as claimed in any of claims 10-12, wherein the snRNA comprises a sequence of bases that specifically hybridizes to a region of the dystrophin pre-mRNA transcript to thereby induce the skipping of one or a plurality of dystrophin exons.

14. A composition as claimed in claim 13, wherein the snRNA induces the skipping of dystrophin exon 51.

15. A composition as claimed in any of claims 10-14, wherein the MABs are engineered to express a plurality of different snRNAs.

16. A composition as claimed in any of claims 1-15, for use as a medicament.

17. A composition as claimed in claim 16, for use in treating a muscular dystrophy disorder.

18. A composition for use as claimed in claim 16 or 17, wherein the composition is for use autologously.

19. A composition for use as claimed in any of claims 16-18, wherein the composition is in the form of an injectable cell suspension.

20. A method of producing the composition of any of claims 1-19, the method comprising transducing MABs in vitro with a vector for expressing a truncated agrin protein.

21. A method as claimed in claim 20, wherein the vector is arranged to express both a truncated agrin protein and a small nuclear RNA (snRNA) capable of inducing exonskipping in a dystrophin pre-mRNA transcript.

22. A method as claimed in claim 20, wherein the method further comprises transducing the MABs in vitro with a second vector for expressing a small nuclear RNA (snRNA) capable of inducing exon-skipping in a dystrophin pre-mRNA transcript.

23. A method as claimed in any of claims 20-22, wherein the first and / or second vector comprises a lentiviral vector.

24. A method as claimed in any of claims 20-23, wherein the MABs are previously obtained from a donor and the method comprises expanding the MABs in vitro and transducing the cells with a lentiviral vector, wherein the donor from which the MABs are derived is:(a) a subject having a muscular dystrophy disorder; or(b) a healthy subject.

25. A method of treating a muscular dystrophy disorder, the method comprising administering a therapeutically effective amount of a composition as claimed in any of claims 1-15 to a subject in need thereof.

Images