Field of Invention

The present invention relates to delivery of gene therapies into muscle via the central nervous system (CNS).

Background to the Invention

The delivery of gene therapies to skeletal muscle has application not only in treating diseases directly affecting muscle but also to create a production site for transgene products that can act to treat disease systemically.

Gene transfer to skeletal muscle may be accomplished by both viral and non-viral transfer vectors. Methods of non-viral transfer of DNA and mRNA include the use of cationic liposomes or other polycations to form nanoparticles that enter cells by endocytosis and the use of plasmid DNA injected or infused into muscle. Similarly, antisense oligonucleotides can be injected into muscle to induce RNAse HI or exon-skipping. Viral vectors have been the most commonly used form of gene transfer to skeletal muscle and, of these, the recombinant adeno-associated viruses (rAAVs) have proved to be the most popular. This is because of their safety, non-pathogenic nature, high transfection ability and their long-term gene expression. It is also of note that the transfection efficiency of AAV can be enhanced whilst reducing their immunogenicity if they are associated with exosomes and delivered as exo-AAV (Hudry E, et al ). In addition to delivering transgenes for the production of therapeutic proteins rAAV can be used as a delivery platform for synthetic micro RNA (mRNA) in order to achieve long-term knockdown of gene expression in diseases caused by toxic protein production. The bacterial clustered regularly interspersed short palindromic repeats (CRISPR)/CRISPR associated protein 9 (Cas9) system proves to be a promising method of gene editing which may be delivered with viral vectors to muscle and has already shown efficacy in the treatment of experimental Duchenne muscular dystrophy.

A particular advantage of delivering gene therapies to skeletal muscle is that once transfected it can potentially provide a site of long-term transgene expression because the natural turn-over of skeletal muscle is thought to be in excess of 10 years. Thus, in many instances only a single gene therapy treatment would be required. This is in contrast to targeting other sites such as the liver where the lifespan of hepatocytes is only 5 months and so one would expect loss of transgene expression over a similar period, leading to a need for repeated treatments.

There are however a number of challenges in delivering gene therapy to skeletal muscle. A single injection into muscle will provide a heterogeneous distribution of vector covering a few cm ³ of tissue at most. Even with multiple injections into skeletal muscle, adequate coverage to treat diseases of muscle is not feasible due to the body wide distribution and large bulk of skeletal muscle which occupies 30- 40% of body mass. Similarly, for the treatment of systemic diseases with gene therapy using skeletal muscle as a site of therapeutic protein production, the small volume of muscle that can be transfected by injection is unlikely to achieve therapeutic levels of the protein within the circulation (Mueller C. et al ). An alternative method of delivering gene therapy to larger volumes of skeletal muscle called hydrodynamic limb vein injection has been described by Hagstrom JE (Hagstrom JE, et al.). Here a limb circulation is transiently blocked by a tourniquet and the isolated limb infused with the vector. This method can increase the extent of muscle transfection, but transfection is inefficient and patchy, and depends upon the proximity of individual muscle fibres to vessels and the duration of exposure, which is limited by the tourniquet time. Again, for diseases widely affecting skeletal muscle this method would still provide limited coverage. To achieve widespread transfection of skeletal muscle gene therapy currently requires systemic delivery.

Systemic (parenteral) delivery of gene therapy requires high dose infusions and a large proportion of the infused vector will be lost in major organs such as the liver and spleen. It has been estimated that the viral load required to achieve therapeutic transfection of muscle when delivered systemically is in the order of 2 x 10 ¹⁴ viral genomes /Kg. This dose of an AAV9 variant (AAVhu68) carrying the survival motor neuron (SMN) transgene delivered intravenously to Non-Human Primates and piglets has been reported as inducing both liver failure and dorsal root ganglia degeneration (Hinderer C et al. 2018). Systemic administration of gene therapy not only carries risks of toxicity from off-target tissue transfection but also has the potential of early loss of gene expression from transfection of tissues with a high cell turnover. In addition there are concerns about the possible risk of inadvertent gene transfer to the gonads with systemic (parenteral) delivery and germ-line transmission of the transgene. Nevertheless the major risk of systemic delivery of gene therapy is that it can provoke B and T-cell immune responses to the viral capsid and in some cases to the transgene and result in loss of transgene expression (Mendell JR, et al 2010; Mays LE, Wilson JM, 2011). Establishing ways to minimise the immune response to the vector will therefore be critical to the success of these therapies.

It is evident that there is a need for an improved method to deliver gene therapy to skeletal muscle whilst minimising systemic exposure and an immune response.

Summary of the Invention

In a first aspect the present invention provides a gene therapy vector for use in the treatment of a disease, wherein the gene therapy vector comprises a therapeutic agent and is to be administered into the sub-motor cortical white matter or motor cortex of the brain in order to deliver the therapeutic agent to skeletal muscles. The inventor has demonstrated that delivery of a gene therapy vector to the motor cortex resulted in the transfection of secondary motor neurons in the spinal cord and, unexpectedly, in the widespread transfection of skeletal muscle. This demonstrates that the transgene cassette crosses the neuromuscular junction (NMJ). The magnitude of the transgene transfection in muscle can also be increased with repeated infusions into the motor cortex. This technique may facilitate effective delivery of gene therapies targeting skeletal muscle, not only for the treatment of diseases affecting muscle but also to create widely distributed, long-term production sites for therapeutic proteins or other transgene products to treat diseases systemically. Confining transgene expression to skeletal muscle (outside of the CNS) will reduce the likelihood of off-target transfection toxicity including the inadvertent gene transfer to the gonads with potential germ-line transmission of the transgene.

The CNS has relative immune privilege due to a number of factors, including the presence of the blood-brain barrier, a dearth of professional antigen-presenting cells (APCs), a lack of draining lymphatics (except in the meninges), and limited access of activated T-cells in the steady state. Consequently, delivering gene therapy vectors into the CNS as a means of transfecting skeletal muscle will greatly reduce the likelihood of generating B or T-cell immune responses to the vector, such as a viral capsid. The transgene, travelling trans-synaptically into secondary motor neurons and thence across the NMJ directly into myocytes will have little exposure to APCs further reducing the likelihood of provoking an immune response. By targeting topographical regions of the motor cortex this method of delivery can harnesses the anterograde axonal transport mechanisms of primary and secondary motor neurons to deliver transgenes to the sarcolemma of specific muscle groups. In contrast to systemic delivery the dose of vector required to achieve therapeutic levels of transgene expression in muscle by motor cortical delivery is substantially less. Gene therapies are expensive to manufacture at GMP grade and so the significantly reduced doses required to treat patients will also make such treatments more affordable.

The present invention provides a method of delivering high dose gene therapies with low immunogenicity to skeletal muscle via transduced motor neurons to create widely distributed, long-term production sites for transgene products that can act to treat disease systemically. This method has potential application in the delivery of gene therapies to provide a continuous source of replacement enzymes for the Lysosomal storage diseases, Lipoprotein lipase deficiency, Alpha-1 antitrypsin deficiency, and Severe Combined Immunodeficiency (SCID) that will be distributed through the circulation. Other applications include the long-term manufacture of Clotting factors VIII and IX in Haemophilia A and B, respectively. In addition the invention could facilitate effective gene transfection of skeletal muscle for the long-term manufacture and delivery of transgene products such as antibodies, for the treatment of a broad range of diseases.

Description

As explained above, current gene therapies for muscle diseases or systemic disease treatable by a transgene product manufactured in skeletal muscle rely on delivering the gene therapy directly to affected muscles. However, this typically requires each muscle to be individually targeted by injection, which makes it difficult to effectively treat diseases affecting widespread muscle groups. Surprisingly, the present inventor has demonstrated that infusing a gene therapy vector comprising a therapeutic agent into the brain results in transfection of primary motor neurons, anterograde transport and trans-synaptic transfection of secondary motor neurons and transport of the therapeutic agent across the motor endplate to transfect muscle. The present invention therefore allows a gene therapy vector to be administered into the brain to target skeletal muscles. In more detail, the gene therapy vector may be administered into sub-motor cortex white matter or the motor cortex of the brain and from there the therapeutic agent can be delivered to skeletal muscle fibres.

Suitable gene therapy vectors for use in the present invention include a virus, an exosome, a plasmid, a liposome or a polycation nanoparticle. In preferred embodiments of the invention the viral vector is selected from all serotypes of Adeno Associated Virus (AAV) or an AAV hybrid, Herpes simplex virus, Baculovirus, Rabies Virus or Human Immunodeficiency Virus (HIV). Preferably the viral vector is AAV2/5 and most preferably is AAV9. Optionally, the viral vector may be encapsulated in an exosome.

The AAV serotypes most studied in the CNS have been serotypes 1, 2, 5, 6, 8, 9, and recombinant human (rh)10. The recombinant genome of a given serotype can be packaged into the capsid of another serotype (e.g., rAAV2/5 contains the AAV2 recombinant genome packaged in the capsid proteins encoded by the cap gene of AAV5). The AAV serotypes have been characterised in animal models and can be selected for targeted gene therapy based on their specific tissue tropisms, bio distribution and transduction efficiency. Transfection of AAV serotypes 2 and 6 are neuronally restricted whereas serotypes 7, 9 and 5 transfect both neurons and astrocytes. Axonal transport of the AAV transgene capsid, either anterograde or retrograde, varies between serotypes and can be both target and dose dependent. AAV6 typically shows retrograde axonal transport, whereas AAVs 1, 2, 5, 8, and 9 have been shown to have either anterograde or retrograde axonal transport depending on the neuronal type transfected and/ or the dose of vector delivered to the CNS tissue.

In addition to axonal transport, dose dependent trans-synaptic transfection of neurons has been reported with infusion of AAV5-CAG-GFP (lxlO ¹³ vector genomes per millilitre (vg/ml)) into the thalamus resulting in transfection of primary and secondary motor neurons (Samaranch L et al ) but not to muscle. Currently there are no reports of viral vector transfection of muscle from the CNS.

The inventor has demonstrated that widespread transfection of skeletal muscle can be achieved by infusing effective doses of the anterogradely transported vectors (AAV2/5) and AAV9 into the pre-motor and motor cortex of rats and sheep and demonstrated transgene expression in muscle. Transfected primary motor neurons in layer 5 of the motor cortex anterogradely transport the transgene to secondary motor neurons in the brainstem and spinal cord via the corticobulbar and corticospinal tracts, respectively. Transfection of secondary motor neurons in the anterior horn cells of the spinal cord will be via a single synapse via the corticospinal tracts but for transfection via the rubrospinal tract this will be predominantly via an interneuron and therefore via 2 synapses. Delivery of the transgene to the anterior horn cells by both the corticospinal and rubrospinal tracts maximises transfection of secondary motor neurons, facilitating anterograde transport of the transgene to and across the neuromuscular junction and sarcolemma to transfect muscle.

Combinations of gene therapy vectors could be administered. For example, two different gene therapy vectors each comprising the same therapeutic agent or different therapeutic agents may be administered sequentially or in combination (i.e., simultaneously). Sequential administration of two different gene therapy vectors may be advantageous in the event that the first vector triggers or is at risk of triggering an immune response. Using a different vector for a further administration of the therapeutic agent can avoid the immune response from being triggered. Additionally or alternatively, two or more different therapeutic agents may be administered using the same gene therapy vector. The two or more therapeutic agents may be delivered sequentially or in combination (i.e., simultaneously).

The disease may be a muscle disease or a systemic disease treatable by a transgene product manufactured in skeletal muscle. The skeletal muscles can be used as a production site for transgene products that can act to treat diseases systemically. Skeletal muscle fibres are terminally differentiated cells with a life expectancy in excess of 10 years and therefore provide a potential site of long term transgene expression.

The ability to directly target skeletal muscle with gene therapy throughout the body may facilitate the long-term delivery of therapeutic doses of transgene products to correct known gene deficits in muscle disease. Diseases of muscle that are potentially treatable with gene therapy include muscular dystrophies, congenital myopathies, motor neurone diseases and lysosomal storage diseases. Over the past 20 years there have been considerable advances in the development of gene therapies for a number of these diseases some of which are now entering clinical trials.

The muscular dystrophies (MD) are a group of muscle diseases that result in the progressive loss of muscle power and the breakdown of skeletal muscles over time. There are nine main categories of muscular dystrophy that contain more than thirty specific types. The most common type is Duchenne muscular dystrophy (DMD), which typically affects males beginning around the age of four. Other types include Becker muscular dystrophy, facioscapulohumeral muscular dystrophy (FSHD), limb-girdle muscular dystrophy (LGMD), Congenital Muscular Dystrophy (CMD) and myotonic dystrophy. Muscular dystrophies are caused by genetic mutations involved in making muscle proteins and may be X-linked recessive, autosomal recessive, or autosomal dominant.

Currently there is no cure for muscular dystrophy. Physical therapy, braces, and corrective surgery may help with some symptoms. Assisted ventilation may be required in those with weakness of breathing muscles. Medications used include steroids to slow muscle degeneration, anticonvulsants to control seizures and immunosuppressants to delay damage to dying muscle cells. Outcomes depend on the specific type of disorder.

Duchenne muscular dystrophy (DMD), which represents about half of all cases of muscular dystrophy, affects about one in 5,000 males at birth. The average life expectancy is 26 years. It is caused by a mutation in the gene for the protein dystrophin at locus Xp21, located on the short arm of the X chromosome. Dystrophin is responsible for connecting the cytoskeleton of each muscle fibre to the underlying basal lamina (extracellular matrix), through a protein complex containing many subunits. The absence of dystrophin permits excess calcium to penetrate the sarcolemma (the cell membrane). Alterations in calcium and signalling pathways result in sarcolemma damage and necrosis of the muscle fibres which are ultimately replaced with adipose and connective tissue.

Becker's muscular dystrophy is related to Duchenne muscular dystrophy in that both result from a mutation in the dystrophin gene. It occurs in approximately 1.5 to 6 in 100,000 male births, making it much less common than Duchenne muscular dystrophy. Symptoms usually appear in men at about the ages of 8-25, but may sometimes begin later. Limb-girdle muscular dystrophies (LGMD) comprise 30% of all progressive muscular dystrophies. They comprise a group of diseases that cause weakness and wasting of the proximal limb muscles and can affect girls and boys with different types produced by different gene mutations. LGMD 2D is the most common form of LGMD and is caused by a mutation in the a-sarcoglycan gene. LGMD2B is caused by mutations in the dysferlin gene which is a protein involved in repair of the sarcolemma membrane. Several gene therapy approaches have been used in pre-clinical models of Duchenne muscular dystrophy. These include methods to restore dystrophin expression by delivering antisense oligonucleotide (AON) cassettes to induce pre- mRNA exon-skipping, delivering plasmids encoding dystrophin or by delivering fragments of the dystrophin gene (micro-dystrophin (pDys) genes) packaged in adeno-associated viral vectors (AAV). The pDys genes re-assemble in the muscle cell after delivery because the full-length transgene is otherwise too large to package in the AAV capsid. In a recent study using the mini-gene dual vectors (AAV9) injected into the extensor carpi ulnaris muscle of a 12-month-old affected dog at the dose of 2 x 10 ¹³ viral genome particles/vector/muscle, widespread mini-dystrophin expression was observed 2 months after gene transfer in the muscle. There was also a reduction in muscle degeneration and fibrosis and improved myofiber size distribution (Kodippili K et al ). An alternative gene therapy approach to Duchene muscular dystrophy is by targeted correction of dystrophin gene mutations with CRISPR/Cas9. This has been shown to restore dystrophin expression in the skeletal muscles of mdx ^4cv mice (Nelson CE, et a/.).

Motor neuron diseases include amyotrophic lateral sclerosis (ALS), hereditary spastic paraplegia (HSP), primary lateral sclerosis (PLS), progressive muscular atrophy (PMA), progressive bulbar palsy (PBP), pseudobulbar palsy or a spinal muscular atrophy (SMA). The motor neurone diseases, including ALS and SMA, have been shown to have intrinsic pathology in skeletal muscle in addition to pathology in motor neurons that result in muscle atrophy independent of denervation. In animal models of ALS with mutant superoxide dismutase 1 enzyme (SOD1), muscle atrophy is associated with increased oxidative stress, mitochondrial dysfunction and bioenergetic disturbances. Increased levels of mutant SOD1 in muscle supress the P13K/AKT (pro cell survival) pathway and promotes apoptosis in myocytes. In SMA low levels of full-length survival motor neuron (SMN) protein causes the motor neuron disease. However, changes in muscle often proceed those in the neurons and lead to atrophy with a molecular profile distinct from that of denervated muscle. Therapies for motor neuron disease that address the motor neurone loss alone will therefore be unlikely to prevent progressive muscular atrophy and disability. As with the treatment of the muscular dystrophies, gene therapy shows promise in motor neurone disease but infusions of a vector carrying a transgene into affected muscle is impractical due to the large bulk and wide distribution of muscle affected. Additionally, systemic delivery of viral vectors, even if they can cross the blood brain barrier, carry the risk of provoking immune responses with loss of transgene expression or adverse effects due to body wide, non-targeted transgene expression.

Lysosomal storage diseases are a group of about 50 rare inherited metabolic disorders resulting from defects in lysosomal function. Although they are systemic diseases the degree of specific organ or tissue involvement depends on the particular disorder. Lysosomal storage diseases with significant skeletal muscle pathology include acid maltase deficiency (Pompe disease), Batten-Kufs' disease, Fabry disease and mannosidosis. Fabry disease is an X-linked lysosomal storage disease caused by a mutation in a-galactosidase A (a-Gal A). Recent clinical trials of gene therapy for Fabry disease have been conducted using ex-vivo gene therapy in which haematopoietic stem cells were transfected with a-Gal A and reintroduced into patients. To date this has not resulted in significant amelioration of symptoms due to low levels of expression.

Where transgene expression is desired only in the skeletal muscle (e.g., in order to prevent CNS toxicity from the transgene product) muscle-specific transgene promoters may be used. Consequently, when the vector transfects primary motor neurons and/or red nucleus neurons in the CNS, the transgene will not be expressed in the motor pathways, which will act as conduits to distribute the transgene widely into skeletal muscle where it will be expressed exclusively. Examples of muscle specific promoters include muscle creatine kinase (MCK) promoters.

As mentioned above, the skeletal muscles can be used as a production site for transgene products that can act to treat diseases systemically. Systemic diseases treatable by a transgene product manufactured in skeletal muscle include haemophilia, an enzyme deficiency syndrome, a chronic infection, a cancer or an autoimmune disease. The systemic disease may be a spinal cord injury from trauma or multiple sclerosis (MS). In particular the therapeutic agent may be directed to treat the consequences of such an injury, such as muscle wastage and/or to facilitate neuro-restoration .

Gene therapy trials of intramuscular delivery of the factor IX transgene with AAV2 for the treatment of Haemophilia B have shown persistence of transduced fibres after 10 years, however, expression never reached therapeutic levels. This led to a shift to hepatic gene transfer with AAV2, and more recently with AAV8, achieving up to 5% of normal circulating levels of Factor IX in some. However, this form of therapy has been associated with neutralising anti-capsid antibodies, delayed liver damage and loss of expression from a capsid-specific memory CD8+ T cell response. Gene therapy for Haemophilia A with delivery of Factor VIII is more challenging than for Haemophilia B because the circulating levels of Factor VIII need to be significantly higher than Factor IX to achieve therapeutic benefit. Consequently the doses of vector that would need administered systemically, or into the liver, would not be achievable currently due to likely immune responses.

Antibody gene therapy may be applicable in the treatment of cancer, autoimmune, inflammatory and infectious diseases. Gene transfer of monoclonal antibody (mAb)-encoding nucleotide sequences into skeletal muscle, after infusion of the gene transfer vector into the motor cortex and/or the red nucleus may facilitate long-term production of the therapeutic mAb protein. In addition to full-length IgG, expressed mAb products may include antibody-protein fusion products such as immunoadhesins, bispecifics and fragments (e.g. antigen binding fragment (Fab), single-chain variable fragment (scFv) and single-domain antibodies). Due to the abundant blood supply in skeletal muscle it provides an efficient transport system for secreted mAbs to enter the circulation. Examples of chronic infectious diseases treatable with transgene expression of mAbs in skeletal muscle include the treatment of HIV with human anti-HIV mAb (Nat. Biotechnol. 2014:32, 397), Hepatitis C with anti-Hepatitis C mAb (de Jong YP, et al. 2014), Hepatitis B with anti-Hepatitis B mAb (Kitaguchi K, et al. 2005), and malaria with anti-P. falciparum CSP mAb (Deal C, et al. 2014). Examples of potential cancer treatments with transgene expression of mAbs in skeletal muscle include the delivery of anti-Herceptin 2 (anti-HER2) antibodies for the treatment of metastatic breast cancer (Kim H, et al. 2016) , anti-DR5 (death receptor 5) to induce tumour cell apoptosis in a range of tumours including liver and bowel cancer (Lv F, et at. 2011), the delivery of anti-epidermal growth factor receptor (EGFR) antibody (Ho DT, et at. 2008), and the delivery of anti-vascular endothelial growth factor mAb (Watanabe M, et at. 2008) to treat a range of cancers. Examples of potential treatments for autoimmune / chronic inflammatory disease treatable with transgene expression of mAbs in skeletal muscle include therapies that target pro-inflammatory cytokines such as anti-tumour necrosis factor-alpha (anti-TNF-a), anti-interleukin-1 (anti-IL-1) and anti-interleukin-6 receptor(anti IL-6R) antibodies (Hashizume et al. 2015).

Enzyme deficiency syndromes include lipoprotein lipase deficiency (LPLD), alpha-1 antitrypsin (AAT) deficiency or Severe Combined Immunodeficiency (SCID).

LPLD is associated with chylomicronaemia, hypertriglyceridaemia, metabolic complications and life-threatening pancreatitis. Clinical trials of delivering AAV1- LPL (alipogene tiparvovec) by intramuscular injection in addition to a low-fat diet was associated with a transient effect on fasting plasma triglyceride levels, low level but sustained gene expression in muscle with a reduced level of pancreatitis. All patients developed anti-AAVl antibodies and 9/14 developed a T-cell response to AAV1 (Gaudet D, et al. 2013).

AAT deficiency is associated with emphysema secondary to insufficient protection of the lung from neutrophil proteases. The disorder affects 1 in 1,500 to 3,500 individuals with European ancestry. Trials of intramuscular injection of AAVI carrying the AAT transgene have failed to achieve therapeutic plasma levels of AAT, in part because of capsid-specific neutralising antibodies and CD4+ and CD8+ T cell responses against the AAVI capsid (Brantly ML, et a/. 2009).

While the gene therapy vector may be directly infused into the motor cortex (e.g. using micro-catheters or cannulae), such infusion can be technically challenging due to the convoluted morphology of this region of the brain. Accordingly, in some embodiments the method may comprise administering a gene therapy vector into white matter of the brain. More preferably, the gene therapy vector is administered below the motor cortex at a junction of cortical gray matter and underlying white matter as described in WO2014/184576. In other words, in preferred embodiments of the invention the gene therapy vector is to be administered into subcortical white matter ventral or inferior to the motor cortex. Surprisingly, administering the gene therapy vector into the white matter below the motor cortex at a junction of cortical gray matter and underlying white matter results in transfection of primary motor neurons, anterograde transport and trans-synaptic transfection of secondary motor neurons and transport across the motor endplate to transfect muscle. This is requires the vector to cross at least one synapse in addition to the motor end plate and has not previously been described. The importance of this unexpected finding is that a vector can be delivered to a defined topographical region of the motor cortex by direct cortical or subcortical infusion and this will result in transport of the vector to pre-defined muscle groups. Infusion of vector into the medial motor cortex will result in anterograde transport to muscles in the contralateral lower limb and trunk. Infusion of vector to the central aspect of the motor cortex will be transported to muscles in the contralateral upper limb and hand and infusions to the lateral motor cortex will be transported to the facial and jaw muscles, the tongue and pharynx.

The gene therapy vector could be administered into any or all of the motor nuclei of the brain, such as the motor thalamus, the basal ganglia, the motor nuclei of the cranial nerves or the red nucleus. However, due to the relatively small volumes of the nuclei, without also targeting the motor cortex or subcortical white matter it is likely to be difficult to infuse sufficient amounts of the vector to achieve a therapeutic dose of the therapeutic agent in the skeletal muscle. Additionally, targeting nuclei such as the thalamus can increase the risk of the vector leaking into other non-motor nuclei of the brain and potentially resulting in sensory, cognitive or behavioural adverse effects.

In preferred embodiments of the invention the gene therapy vector is administered by convection enhanced delivery (CED). Convection enhanced delivery is well known in the art. It means the delivery of a pharmaceutical, or other composition, to the brain under a positive pressure gradient via a narrow catheter, usually having an inner diameter of less than 500pm, more usually less than 250pm. Administration of the pharmaceutical or other composition by CED leads to a greater volume of distribution than is typically achieved by intracerebral injection or infusion. Appropriate catheters are described in WO03/077785. Preferably the catheter is a reflux resistant catheter, such as a stepped catheter or recessed step catheter, e.g. as described in Gill et at 2013 (incorporated herein by reference). It may also comprise a port for connecting the catheters to a delivery device. Suitable ports are described in W02008/062173 and WO2011/098769.

The gene therapy vector is preferably for administration via at least one convection enhanced delivery catheter, especially an intraparenchymal catheter. More preferably it is for delivery via at least two, at least three or four or more such catheters. In embodiments of the invention the gene therapy vector is to be administered by at least one, preferably at least two convention enhanced delivery catheters. One or more catheters may be chronically implanted into a patient allowing repeat infusions of the therapeutic agent. The reason for repeated dosing may be that the number of viral particles required to transfect muscle to achieve a therapeutic effect cannot be delivered to the motor cortex in a single administration. This is because concentrations of vector greater than lx 10 ¹⁴ viral particles per ml are difficult to manufacture and the amount of infused vector that transfects primary motor neurons in layer 5 will be small. Many hundreds or thousands of viruses will need to transfect each primary motor neuron and of these but a small fraction will reach the secondary motor neurons and motor end plates. When several hundreds or thousands of AAV capsids are endocytosed into the soma of a primary motor neuron many, such as the AAV5 and AAV9 serotypes will undergo fast anterograde axonal transport at a rate of several mm per day effectively clearing the soma. Repeated delivery of a vector to the motor cortex via the sub-cortical white matter over several consecutive days, for example, will thereby enable a therapeutic dose to be achieved in the muscle. The region of the motor cortex supplying the lower limbs and trunk is relatively small when compared with the representation of the upper limb and hand region and the facial region particularly when one considers the muscle bulk that each supplies. Hence a treatment strategy to achieve a more homogenous distribution of the therapy throughout the body may involve differential dosing to each region with greater doses being delivered to the lower limb and trunk region. It may take several months for a vector to reach its target depending on the rate of axonal transport and so a determination of the therapeutic effect will not be possible until this time has passed and the transgene expression has reached a steady state in the muscle. If there has been inadequate dosing then chronically implanted catheters may be used for topping up treatment without the need for repeated surgical implantation. Similarly repeated dosing may be required if there is a fall off of effect. The therapeutic agent may be infused in the brain via an array of at least two or at least three catheters. For example, three or more catheters may be implanted in a fan-shaped array, e.g. on an antero-posterior trajectory under the frontal cortex with their entry points and the frontal pole and each catheter tip positioned beneath the motor cortex. The array may comprise one or more further catheters for infusing the therapeutic agent into the sub-cortical white matter of the orbito- fontal cortex with a fronto-polar entry point. Approaches from a posterior trajectory are also possible. Preferably, the array comprises at least one central catheter for administration of a biologically inert fluid, i.e. a fluid which does not initiate a response or interact when introduced to biological tissue. For example, the catheter for infusing the biologically inert fluid may be placed in the frontal white matter in an antero-posterior direction with an entry point at the frontal pole and distal end deep to the target cortex.

When infusing the therapeutic agent via an array of at least two or at least three catheters comprising at least one central catheter for administration of the biologically inert fluid, the biologically inert fluid may be infused through the central catheter prior to, simultaneously with and/or subsequent to infusion of the therapeutic agent. Alternatively, the therapeutic agent and biologically inert fluid may be infused sequentially through the same catheter.

In these circumstances the distribution of the infused therapy can be shaped and driven preferentially into the cortical grey matter by the judicious placement of one or more catheters deep to those infusing the therapy and co-infusing the biologically inert fluid, such as artificial CSF, which will counteract the flow into the white matter. This strategy will limit the total dose of the therapeutic agent that needs to be administered to achieve coverage of the target area and thereby reduce the likelihood of toxicity. Where patients have particular patterns of muscle atrophy or where defined groups of muscles are typically involved (e.g. as in facioscapulohumeral muscular dystrophy or limb girdle muscular dystrophy) the vector can be delivered specifically to the region of the motor and/or premotor cortex that topographically represents the muscle groups.

The biologically inert fluid may be selected from one or more of phosphate buffered saline (PBS) or artificial cerebrospinal fluid (aCSF). In preferred embodiments of the invention the biologically inert fluid is aCSF. Artificial cerebrospinal fluid (aCSF) as used in the present invention may comprise ionic constituents. Preferably the aCSF comprises NaCI at a similar concentration to that found in natural CSF, that is to say the concentration is preferably within 15%, more preferably within 10% of the concentration in natural CSF. Preferably the aCSF comprises NaHC03 at a similar concentration to that found in natural CSF, that is to say the concentration is preferably within 15%, more preferably within 10% of the concentration in natural CSF. Preferably the aCSF comprises KCI at a similar concentration to that found in natural CSF, that is to say the concentration is preferably within 15%, more preferably within 10% of the concentration in natural CSF. Preferably the aCSF comprises NaH2P04 at a similar concentration to that found in natural CSF, that is to say the concentration is preferably within 15%, more preferably within 10% of the concentration in natural CSF. Preferably the aCSF comprises MgCI2 at a similar concentration to that found in natural CSF, that is to say the concentration is preferably within 15%, more preferably within 10% of the concentration in natural CSF. The aCSF can comprise glucose at a similar concentration to that found in natural CSF, that is to say the concentration is within 15%, preferably within 10% of the concentration in natural CSF. Alternatively, the aCSF does not comprise glucose. Most preferably, the aCSF does not comprise proteins.

The gene therapy vector and/or therapeutic agent may be administered in the form of a pharmaceutical composition, which is preferably sterile and may comprise one or more pharmaceutically acceptable carriers or excipients. Suitable carriers and excipients will be familiar to the skilled person and may be optimised in line with the intended route of delivery. For example, suitable pharmaceutical compositions may include buffers, binders, preservatives, thickeners or antioxidants.

The gene therapy vector may be administered at a concentration of about 1 x 10 ¹⁰ viral particles per ml to about 5 x 10 ¹⁵ viral particles per ml, preferably about 1 x 10 ¹³ viral particles per ml to about 1 x 10 ¹⁴ viral particles per ml. Flow rates for CED will typically range from 3pl/minute to IOmI/minute. In the scenario described above where a sub motor-cortical catheter delivers the gene therapy vector and a more deeply placed catheter in the white matter delivers a biologically inert fluid to assist in driving it into the cortex, the flow rate in the latter catheter may be higher than in the former. To achieve sufficient transfection of the secondary motor neurons and muscle, the concentration of vector required to transfect pyramidal cells in layer 5 of the motor cortex (after sub-motor cortical infusion) is preferably in the order of 1 x

10 ¹² to 1 x 10 ¹⁵ vg/ml. Lower concentrations are less likely to achieve sufficient transfection of secondary motor neurons and trans-NMJ transfection of muscle to achieve a therapeutic effect. The volume of the human motor cortex is approximately 12cm ³ . When a vector is infused into the brain grey matter the volume of distribution per volume of infusion (Vd/Vi) is ~4: 1. Consequently, the motor cortex could be adequately covered with an infusion of 3ml (confined to the grey matter). If the vector is delivered to the motor cortex in both hemispheres at a concentration of 4 x 10 ¹³ vg/ml then the total viral load which could potentially be transported to muscle would be up to 24 x 10 ¹³ vg. By contrast, if one delivers a viral vector to the motor thalamus (the ventral anterior and ventro-lateral thalamic nuclei) of each hemisphere as described by Samarach L et al 2017, which in a human occupies a volume of ~2cm ³ , then at a concentration of 4 x

10 ¹³ vg/ml, the total viral load will be less than 10% of that achievable by the motor cortical target. It is also of note that less than 50% of the transgene expressed in the thalamus after infusion of AAV5-GFP will be expressed in the motor cortex so that achieving an adequate dose in muscle by this route of delivery is improbable.

The gene therapy vector is preferably for administration by infusion for between 4 and 24 hours, especially for at least 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 16 hours and/or for less than 23, 22, 21, 20, 19, 18, 17, 16 , 15, 14, 13, 12, 11, 10, 9 or 8 hours. It is preferably for infusion for around 8 hours. In embodiments of the invention the gene therapy vector is for repeated daily infusions over a period of up to 6 days.

Whether or not the gene therapy vector is for administration for a number of consecutive days or for regular administration over a number of days, it may independently or additionally be for administration weekly, fortnightly, monthly, every six, eight, twelve or fifteen or more weeks. For example, a cycle of two or three days of infusions may be repeated every three months. Alternatively, it may be for administration in a series of cycles of infusions, with 6, 7, 8, 9, 10, 11 or 12 months between the end of a first cycle of infusions and the next cycle of infusions. The interval between repeat administrations of the gene therapy vector may be determined by the rate of axonal transfer from the soma of the pyramidal cells in the motor cortex. In other words, a further dose is preferably not administered until the preceding dose has substantially cleared the soma of the primary motor neurons.

Gene therapy, as is well known, is the use of genetic material to modulate or add to genes in an individual's cells in order to treat disease. The genetic material to be introduced, i.e. the transgene, may be any appropriate genetic material, including DNA, RNA, small interfering RNA (siRNA), microRNA, short hairpin RNA (shRNA) or antisense oligonucleotide (AON) cassette. In embodiments of the invention the transgene may comprise CRISPR/Cas9. The transgene may be used to treat the disease in any known manner, such as gene replacement, gene knockdown, gene editing, pro-survival gene therapy and cell suicide therapy. In embodiments of the invention the transgene may be a monoclonal antibody encoding nucleotide sequence, which may express full-length IgG, antibody- protein fusion products, such as immunoadhesins or bispecifics, or antibody fragments, such as antigen-binding fragments (Fab), single-chain variable fragments (scFv) or single domain antibodies. Encoded monoconal antibodies for use in the invention may include anti-HIV, anti-Hep C, anti-Hep B, anti-P. falciparum CSP, anti-Herceptin 2, anti-death receptor 5, anti-epidermal growth factor receptor, anti-tumour necrosis factor-alpha, anti-interleukin-6 receptor or anti-interleukin-1.

Preferred transgenes for use in the invention include one or more of dystrophin, functional fragments of dystrophin (often referred to as "mini-dystrophin"), myotubularin (MTM1), glial cell-derived neurotrophic factor (GDNF), Wnt7a, alpha-sarcoglycan, superoxide dismutase 1 (SOD1), alpha-galactosidase A (a- GalA), Factor IX, Factor VIII or survival motor neuron protein (SMN).

Examples include the delivery of dystrophin in DMD, a-sarcoglycan in LGMD2D, dysferline +/- follistatin in LGMD2B, a-GalA in Fabry disease, SOD-1 in ALS and SMN in SMA. It may also facilitate the delivery of transgenes for a number of secreted factors or growth factors that have been shown experimentally to participate in muscle repair and regeneration that would otherwise cause toxicity if delivered systemically due to off-target transfection. These could be delivered by the method described either singly or in combination and include hepatocyte growth factor (HGF), fibroblast growth factors (FGFs), transforming growth factor- 3s (TGF-3s), insulin-like growth factors (IGFs) and Wnt7a. Similarly neurotrophins such as brain-derived neurotrophic factor (BDNF) and glial cell-line derived neurotrophic factor GDNF are known to play an important role in the maintenance and repair of motor neurons and in the plasticity of the neuromuscular junction (NMJ).

In muscle diseases and neuromuscular diseases, augmenting motor neuron innervation and the efficiency of neural transmission to surviving muscle fibres would be a useful therapeutic strategy but, importantly, the neurotrophin GDNF has an anti-inflammatory role by reducing tumour necrosis factor-a (TNF-a) and interleukin-ΐb (IL-Ib) and reduces leukocyte infiltration (Liu GX, et at 2014) . In the muscular dystrophies such as DMD and LGMD2D, TNF-a and IL-Ib play a significant role in the initiation and perpetuation of muscle pathology and reducing the expression of these cytokines has been shown experimentally to improve disease severity and symptom control (Benny Klimek ME, et al. 2016). Delivering GDNF by systemic gene therapy however has been shown to cause significant off- target transfection and toxicity (Thomsen GM, et al. 2017) and so the described method of targeted delivery to muscle should avoid these complications.

The subject is preferably a mammal, more preferably a primate, especially a human and may be a paediatric or geriatric patient.

The present invention additionally provides methods for treating a muscle disease, the methods comprising administering gene therapy vector to patient in need therefore, wherein the gene therapy vector comprises a therapeutic agent and the gene therapy vector and therapeutic are administered into the sub motor cortical white matter, or motor cortex of the brain via CED in accordance with the embodiments of the invention described above.

Brief Description of the Drawing

The invention will now be described in detail, by way of example only, with reference to the figures.

Figure 1. Targeting the hind-limb topographical region in the rat motor cortex: Atlas of the motor cortex topography in a rat showing targeted infusion sites in the hind-limb region. (Images based on data Fonoff et al 2009 and Yu et al 2017). Figure 2. Human GDNF in skeletal muscle following infusion of AAV5-CMV-hGDNF in rat motor cortex: Representative images of immunostaining for hGDNF within the hind limb muscle contralateral to the infused motor cortex, showing positive detection in the treated (A) versus no detection in the control (D) animals. Additional images show the internal controls, without primary and secondary antibodies, for treated (B&C) and control (E&F).

Figure 3. Human GDNF concentration in skeletal muscle in the hind limb contralateral to the infused motor cortex. The mean (+/- standard deviation) of GDNF concentration obtained for the test group (64.98+/-15.14 pg/ml) was higher than for the control (36.03+/-17.38 pg/ml). Wilcoxon test revealed statistically significant differences in the concentration between the two groups (W=815, rank sum=52, p=0.0018).

Figure 4. The concentration of hGDNF (pg/ml of total protein) in skeletal muscle measured using Sandwich ELISA in all 4 limbs in one animal following the infusion of AAV5-CMV-hGDNF in the topographical location of the right lower limb in left motor cortex. A Kruskal-Wallis H test showed that there was not a statistically significant difference in human GNDF above baseline between the different locations, c2(2) = 2.55, p = 0.466, with a mean rank GDNF amount of 52.14pg/ml for LHS forelimb, 57.13pg/ml for RHS forelimb, 49.50pg/ml for LHS hindlimb and 76.16pg/ml for RHS hindlimb.

Figure 5 shows (A) the sheep model used; (B) a schematic illustration of the position of the motor cortex in the brain (blue); (C) the position of catheters used for a test infusion of 500pL of gadolinium to the sheep motor cortex (yellow catheter = aCSF; white catheter = contrast, i.e. gadolinium); (D) a post-infusion sagittal view of gadolinium distribution; (E) a post-infusion axial view of gadolinium distribution; and (F) a post-infusion coronal view of gadolinium distribution.

Figure 6 shows a mid-sagittal T2 image of sheep brain with 3D reconstruction of the volume of distribution of 500pl of infused gadolinium in the motor cortex (blue). The sub-cortical catheter that delivers the gadolinium is shown in yellow. Also shown are the deep white matter catheter (white) and the trans-cutaneous port located in the occipital bone. The volume of distribution / volume of infusion ratio (Vd/Vi) = 4: 1. Figure 7A shows a sagittal section of sheep brain post-infusion of AAV9-CMV- Mcherry to the motor cortex. MCherry fluorescence indicates transfection of primary motor neurons (pyramidal cells) and anterograde transport in the cortico- spinal tracts.

Figure 7B shows a transverse section of sheep medulla post-infusion of AAV9- CMV-Mcherry to the motor cortex. MCherry fluorescence indicates transfection of secondary motor neurons in the Hypoglossal nucleus as well as in the Inferior Olivary nucleus.

Figure 8 shows a schematic transverse section of the cervical spinal cord (A). Anterograde transport of MCherry is seen in the corticospinal and reticulospinal tracts and in the anterior horn cells of the sheep (B).

Figure 9 shows sections through the cheek muscle indicating expression of MCherry in muscle fibres contralateral to the infused motor cortex.

Figure 10 shows a section through the sheep tongue indicating expression of MCherry in muscle fibres contralateral to the infused motor cortex.

Figure 11 shows the presence of hGDNF in muscle fibres of the sheep's tongue.

Examples Example 1: Delivery of AAV5-GDNF to muscle following infusion into the motor cortex of the rat

Method:

Juvenile male Wistar rats (Harlan, UK) weighing 250-275g were group-housed in Techniplast 1500U cages with irradiated lignocel bedding and sawdust (International Product Supplies Ltd UK). The study room was illuminated by fluorescent light set to give a cycle of 12 hours light and 12 hours dark and was air-conditioned. The ambient temperature was held between 17 and 22 °C. All animal work was performed in accordance with the UK Animal Scientific Procedures Act 1986 and was covered by both project and personal licences that were issued by the Home Office. Animal licences were reviewed and approved by the University of Bristol Ethics Committee (project licence PA95E951). All efforts were made to minimise animal use and suffering.

Motor Cortex infusion and distribution analysis:

Each rat was anaesthetised with 2% inhaled isoflurane in oxygen in an anaesthetic chamber and then placed in a stereotactic frame (David Kopf Instruments, CA, USA). Anaesthesia was maintained with inhaled 2% isoflurane/oxygen. The scalp fur was clipped and skin cleaned using alcoholic chlorhexidine. Following a midline incision from glabella to occiput to expose the skull a 5 mm burr hole was made over the left motor cortex. Infusions into the motor cortex were performed using a custom-made, fused silica cannula with an outer diameter of 0.22mm and inner diameter of 0.15mm. The cannula was attached to a 1ml syringe (Hamilton, Switzerland) connected to a rate-controlled micro infusion pump (World Precision Instruments Inc. USA). The cannula was delivered to 3 targets, 1 mm posterior to the bregma, which were 1 mm, 1.5mm and 2mm lateral to the midline. A further target was at the bregma, 1.5 mm lateral to the midline. These were in the topographical location of the right hind limb in the left motor cortex as defined cortical functional mapping studies (Figure 1). Infusions of 1.5 pL of AAV5-CMV-hGDNF (5.4 X 10 ¹³ VG/ml) or vehicle alone (artificial CSF) were made at each target, 1mm below the cortical surface.

Bone wax was applied where needed for bone haemostasis and the wound was closed with absorbable sutures (4/0 Vicryl Rapide®). Intramuscular analgesia was used postoperatively (buprenorphine i.m, 30 pg/kg) and the animals were returned to their housing when recovered. The rats were euthanized after 8 weeks.

For immunohistochemical analysis (IHC), rats were transcardially perfused with 4% paraformaldehyde. Brains were removed and placed in 4% paraformaldehyde for 24 hours, then cryoprotected in 30% sucrose. For the Enzyme-Linked Immuno-Sorbent Assay (ELISA), brains and muscle were explanted and rapidly frozen at -80°C until required, in addition muscle specimens were post fixed from fresh frozen (as per table 1).

Immunohistochemistry methods:

Sections from the rat muscle was cut into 25 pm thick sections using a Leica CM 1850 cryostat (Leica Microsystems, Germany) at -20 °C. For immunohistochemistry, fixed sections were mounted on gelatine-subbed slides. Once dry, the sections were washed with PBS for 5 minutes x 3. The antigen retrieval was completed within a microwave, sections were added to sodium citrate buffer pH6, for the following times; 3 minutes full power, 1 minute rest, 2 minutes full power, rest 30 minutes. Sections were then washed PBS with 0.1% Trition-X-100 for 5 minutes x 3. Sections were blocked in 0.1% triton-x-100 in PBS containing 10% normal goat serum (Sigma Aldrich, UK). Sections were incubated for 1 hour at room temperature. Sections were then washed with 0.1% triton-x-100 in PBS for 5 minutes. Following washing, sections were incubated in Rabbit Anti-GDNF primary antibody (Abeam) 1:200, in order to determine the presence and distribution of hGDNF. The next day, primary antibody was removed and sections were washed with 0.1% triton-x-100 in PBS for 5 minutes x 3. Goat anti-rabbit IgG antibody (H+L), peroxidase 1:300 (Vector Laboratories), for 1 hour at room temperature. Then washed with PBS for 5 minutes x 3. Sections were dehydrated using an ethanol gradient, sections were then mounted using CleariumTM (Lecia, UK) before viewing. Images were captured using a Leica DM5500 microscope digital camera (Leica Microsystems, Germany). hGDNF Sandwich ELISA:

From 3 treated and 3 control animals, 3 muscle punches were taken from each of their 4 limbs. Each muscle punch weighing approximately 200 mg was placed into lml lysis buffer consisting of Pearce Ripa buffer (Thermo fisher) 8i protease inhibitor cocktail solution (Sigma) at 1: 100 dilution. The samples were then homogenised in a in Precellys 24 homogeniser (2 x 15 seconds at 6,000 rpm). Samples were centrifuged at 10,000 xg for 10 minutes, and the supernatant collected. Following this, hGDNF levels were measured using a commercially available hGDNF DuoSet (Bio Techne, UK). A standard curve using hGDNF was completed and values equivalent to the hGDNF DuoSet Immunoassay system standard were obtained. Absorbance was measured using a plate reader (FLUOstar OPTIMA from BMG Labtech,UK) at 450nm. Results were expressed as pg of total human GDNF per ml of tissue or total protein (Pierce BCA protein assay kit from Thermo Scientific (Loughborough, UK).

Results:

8 weeks following infusion of AAV5-CMV-hGDNF (3.24 x 10 ¹¹ VG) into the hind limb topographical region of the rat motor cortex, hGDNF was detected in the skeletal muscle of the contralateral hind limb. This was shown with positive immunostaining for hGDNF in the contralateral hind limb that was not seen in the control group (Figure 2, A - F).

GDNF sandwich ELISA of muscle samples from the hind limb, contralateral to the infused motor cortex in the treated group (n=3) showed significantly higher concentrations of hGDNF than in the control group (n=3) (Figure 3).

In a further experiment we demonstrate that 8 weeks following an infusion of AAV5-CMV-hGDNF into the hind limb topographical region of the rats left motor cortex the concentration of hGDNF in the right hind limb was higher than in the other limbs (Figure 4).

This study demonstrates that infusion of AAV5-CMV-hGDNF at a concentration of 5.4 X 10 ¹³ VG/ml into the hind limb topographical region of the rat motor cortex will result in the expression of hGDNF in skeletal muscle in the contralateral hind limb. This indicates that when delivered at a sufficient concentration the transgene and, or transgene products can not only cross a single synapse i.e. from primary to secondary motor neurons but will also cross the neuromuscular junction. This has not been previously reported and has important therapeutic implications.

Example 2: Bio-distribution of AAV9-MCherry and AAV9-hGDNF delivered to the motor cortex of a large animal model (adult Romney Marsh sheep)

Method: A Romney Marsh adult sheep (Fig 5) was anaesthetised and its head fixed in a custom made head fixation device (Renishaw Ovine fixation and fiducial platform). T1 and T2 volumetric MRI images were acquired of the sheep's head with a radio opaque fiducial fixed to the head frame. The trajectories of 4 long-term implantable micro-catheters (Neuroinfuse™, Renishaw pic) were planned from the MRI volumes using customised Neuroinspire ™ (Renishaw pic) surgical planning software. The sheep's motor cortex occupies the medial aspect of the frontal lobe and 2 posterior to anterior catheter trajectories were planned for each hemisphere. These included a subcortical and a deep white matter catheter in a parallel orientation in each hemisphere with entry points in the parietal region (Fig 5).

With reference to the fiducials visible on the MRI image volumes the surgical plan was co-registered with a CRW stereotactic frame (Integra. USA) that was attached to the sheep head fixation device. This then facilitated implantation of the catheters under stereotactic guidance. The catheters were connected to independent channels in a bone anchored, septum sealed, trans-cutaneous port via a manifold (Renishaw drug delivery system™. Renishaw pic.). When the scalp wound is closed the port emerges though the skin over the occipital bone where it is implanted. The port facilitates repeated infusions of therapies into the brain parenchyma over many months or years if required. This is achieved by fixing an administration set to the port and by turning an actuator, 4 needles within the set are driven through the port's septum and into individual channels thereby connecting external pumps, which deliver the therapy, to specific catheters.

Prior to delivering AAV to the sheep motor cortex a test infusion of an equivalent volume of gadolinium (500pL to each motor cortex) was first conducted to establish its volume of distribution. The infusions were carried with the sheep under general anaesthesia in an MRI scanner. Due to the convoluted morphology of the motor cortex, delivery of the gadolinium (and subsequently the vector) was achieved by infusing it into the subcortical white matter whilst simultaneously infusing artificial cerebrospinal fluid (aCSF) into the deep white matter, both at a flow-rate of 5pL per minute. The pressure gradient created by infusing aCSF into the white matter drives the gadolinium or vector into the cortex. When the latter had been infused a further 500pL of aCSF was delivered down each of the sub cortical and deep white matter catheters at 5pL per minute to further drive the gadolinium into the cortical grey matter (Fig 6). Having confirmed satisfactory motor cortical coverage with the test infusion of gadolinium the sheep was recovered and 1 week later underwent infusions of AAV9-CMV-MCherry (500mI_ of 2.2x 10 ¹³ VG/ml) into the left motor cortex and AAV9-CMV-hGDNF (500mI_ of 2.64x 10 ¹³ VG/ml) into the right motor cortex using the above described method.

The sheep was housed in a pen with other sheep for 6 weeks prior to being euthanized and perfusion fixated. The brain, spinal cord, tongue, cheek muscle and limbs were cryopreserved and retained for histological analysis.

Histological Methods : The fixed brain was sectioned at 100 pm increments in the sagittal plane and from the level of the cerebral peduncles the pons, medulla and spinal cord were sectioned axially at 40 pm increments. The tongue was sectioned in the sagittal plane and the cheek sectioned in the coronal plane.

MCherry fluorescence was visualised using florescence microscopy and hGDNF quantified with anti-hGDNF immunohistochemistry with diaminobenzidine staining.

Every 25 ^th section was processed for immunohistochemistry and H&E staining.

Results:

1. Bio-distribution of MCherry florescent protein transgene in the brain, spinal cord and skeletal muscle following infusion of AAV9-MCherry into the left motor cortex.

Brain:

MCherry fluorescence was observed throughout layer 5 of the motor cortex with pyramidal cells and the cortico-spinal and cortico-bulbar tracts being strongly fluorescent (Fig. 7A, 3). Less intense fluorescence was observed throughout all layers of the cortex, being expressed in both neurons and glia.

MCherry fluorescence was observed in the hypoglossal nucleus neurons (Fig. 7B) confirming transgene delivery to the origin of cranial nerves relevant to bulbar function, and the inferior olivary nucleus, which has a vital role in co-ordinating motor pathways from the spinal cord and cerebellum, regulating motor control. Spinal cord:

Axial sections through the cervical spinal cord show anterograde transport of MCherry in the anterior and lateral corticospinal tracts as well as the ventral white commissure. MCherry is also seen in the medial and lateral reticulospinal tracts and anterior horn cells indicating trans-synaptic transfection of AAV9 (Fig. 8).

Skeletal muscle:

Sections through the cheek (Fig. 9) and tongue (Fig. 10) contralateral to the infused left motor cortex demonstrate the expression of MCherry in muscle fibres. MCherry is not a secretory protein which indicates anterograde transfection of AAV9 across the neuromuscular junction. This has not previously been described and was unexpected.

2. Biodistribution of GDNF transgene in the brain, spinal cord and skeletal muscle following infusion of AAV9-hGDNF into the right motor cortex :

Immuno-histochemical staining of sections of the sheep's tongue showed the presence of hGDNF in muscle fibres. (Fig. 11).

Conclusion:

The present study unexpectedly demonstrates that it is possible to transfect skeletal muscle by infusing a gene therapy vector and cargo into sub-motor cortical white matter of the brain. This finding opens up exciting new options for treating muscle diseases and/or systemic diseases treatable by a transgene product manufactured in skeletal muscle.

References Benny Klimek ME, et al. PLoS One. 2016 May 23;ll(5):e0155944

Brantly ML, et al. Proc. Natl Acad. Sci. USA. 2009; 106(38): 16363-8

Deal C, et al. Proc. Natl. Acad. Sci. USA. 2014; 111(34): 12528-12532 de Jong YP, et al. Sci Transl Med. 2014;6(254):254ral29

Fonoff ET, et al. Behav Brain Res. 2009;202(1): 138-41. Gaudet D, et al. Gene Ther. (2013); 20(4):361-9

Gill T, et al. J Neurosci Methods. 2013; 219(1): 1-9

Guo et a/. Nature Communications, 2017; 8:861

Hagstrom JE, et al. Mol. Ther. 2004; 10(2):386-98

Hashizume et a/. Int Rev Immunol. 2015 May;34(3):265-79

Hinderer C et al. Human Gene Therapy, 2018 Jan; 29(3):285-296

Ho DT, et al. Cancer Gene Ther. 2008; 15(9): 553-62

Hudry E, et al. Gene Ther. 2016 April; 23(4):380-92

Kim H, et al. Cancer Gene Ther. 2016; 23:341-347

Kitaguchi K, et al. Int J Mol Med. 2005: 16(4):683-8

Kodippili K et al, Hum Gene Ther. 2017 Aug 04

Liu GX, et a/: Int Immunopharmacol. 2014 Mar;19(l): 145-52

Lv F, et al. Cancer Lett. 2011; 302(2): 119-27

Mays LE, Wilson JM; Mol Ther. 2011; 19(1): 16-27

Mendell JR, et a/; New Engl J Med. 2010; 363(15): 1429-37

Mueller C. et al, Mol Ther. 2017 Jun;25(6): 1387-1394

Nat. Biotechnol. 2014:32, 397

Nelson CE, et al. Science, 2016; 351(6271):403-7

Samaranch L et a/: Gene Therapy 2017; 24:253-261 Suraweera et al. Frontiers in Oncology, 2018; 8:92

Thomsen GM, et al. Gene Ther. 2017; Apr;24(4):245-252

Watanabe M, et al, Hum Gene Ther. 2008; Mar;19(3):300-10

Yu Z, et al. J Neurosci Methods. 2017;291:69-82