|Home | About | Journals | Submit | Contact Us | Français|
Current treatment options for patients with monogenetic congenital myopathies (MCM) ameliorate the symptoms of the disorder without resolving the underlying cause. However, therapies are being developed where the mutated or deficient gene target is replaced. Thousands of clinical trials have been undertaken relating to gene therapy, with around 9% focused on monogenetic diseases such as Duchenne muscular dystrophy (DMD) and limb girdle muscular dystrophy (LGMD)1. Preclinical findings in animal models have been promising, as illustrated by studies of a potential treatment for X-linked myotubular myopathy (XLMTM) in canine and murine models2. We will therefore discuss the prospective applications and approaches of gene replacement therapy, using these disorders as examples.
Both limb girdle muscular dystrophy type 2C and Duchenne muscular dystrophy are part of a subclass of myopathies known as dystrophies, diseases where muscle degeneration is accompanied by replacement with fatty or connective tissue. DMD is caused by X chromosome linked genetic mutations leading to the absence of membrane-anchored dystrophin protein, the centerpiece of the large dystroglycan complex that plays a pivotal role in sarcolemma stability during muscle contraction3(Table 1). The symptoms are visible as early as 2–3 years of age, a progressive decrease in striated muscle function, starting from proximal muscle such as legs and pelvis and eventually involve the whole body. Most patients are wheelchair-dependent starting from early teen. The average life expectancy is around 25 years (http://www.nlm.nih.gov/medlineplus/ency/article/000705.htm), with respiratory failure and cardiac complications the highest causes of mortality.
Congenital centronuclear myopathies are inherited muscle diseases where the nucleus is located in the center of the muscle fiber instead of the periphery. X-linked myotubular myopathy (XLMTM) is the most common centronuclear myopathy, affecting an estimated 1 in 50 000 male births (Table 1).4,5 The disease is due to a mutation on the long-arm of the X chromosome, usually inherited by hemizygous boys from an asymptomatic carrier mother.6 This mutation causes a deficiency of the protein myotubularin7. Myotubularin has been identified as a phosphoinositol phosphatase and may be critical to normal excitation-contraction coupling and remodeling of the sarcoplasmic reticulum in muscle.8 When XLMTM patients are first born, they typically exhibit hypotonia and may be blue due to respiratory insufficiency.6 The disease is often fatal in the first year of life and long-term survivors may require ventilatory support.9 Affected boys are particularly susceptible to infection and respiratory dysfunction is the leading cause of death.10
Although there are differences between the symptomatic presentations of these diseases, there are some shared difficulties to consider when design gene therapies. High vector titers may be required to reach an effective dose,11 increasing the chance of adverse effects in patients. In addition, the need to treat respiratory muscles as well as the heart in DMD and XLMTM may complicate delivery. Improvements in delivery methods12 and in vector characterization to increase efficiency may address this problem.13 Vector modification may also ensure more efficient delivery to the muscle and improve safety by reducing off-target delivery to organs like the liver.11,14 Tissue-specific promoters is another strategy to secure tissue-specific transgene expression. Immune response is a major concern, particularly in genetically-null patients who may have antibodies against the gene product produced by the treatment15 and immunosuppression before and during treatment may have to be considered. There are also challenges specific to each disease. For example, the large size of the dystrophin gene limits the choice of vector to be used in treatment. Overexpression of γ-sarcoglycan in LGMD patients may exacerbate the condition.16 Significant wasting in XLMTM patients leaves very little muscle to treat and, due to the young age of the patients, selecting an appropriate and reproducible outcome measure may prove difficult. We will be discussing new developments that address these concerns, including modifications of the vector and the combination of gene therapy with other approaches.
Gene therapy is defined as the introduction of nucleic acids, including DNA, RNA and their analogues into cells of living organism to treat diseases31. This occurs through modified expression of genes of interest to trigger alterations of certain biological functions. Gene therapy targets living cells, primarily because cell’s intrinsic gene expression machinery is indispensible to mediate the production of therapeutic molecules, including protein, shRNA and microRNA.
Since classical gene therapy acts on native tissues, the abundance of target cells largely determines the effect of gene therapy. This is especially true in congenital myopathies. In the advanced stage of diseases, such as DMD and XLMTM, surviving myocytes are so limited that even if the function of individual myofibers were fully restored, there would be no appreciable functional improvement on tissue level. The advent of stem cell technology, especially the discovery of induced pluripotent stem cells (iPSCs)32,33, has the potential to overcome this hurdle. Pluripotent stem cells may be able to replenish tissue loss through their indefinite self-replicating potential and capacity to be converted into nearly all cell types within the body. The advantages of combining stem cell therapy with gene therapy have been demonstrated in several animal studies, in which vectors were administered ex vivo and modified donor cells were later engrafted into native tissue34–36.
Various gene therapy strategies target gene expression and regulatory network at different levels. For example, genetic sequence can be permanently inserted into genome for long-term expression, using retrovirus or lentivirus31. With the development of genome modification tools37 such as clustered regularly interspersed short palindromic repeats (CRISPR) enzymes and Transcription activator-like effector nuclease (TALEN), the technical barrier of in situ editing eukaryotic genomic DNA has been substantially lowered. These techniques hold the potential to seamlessly restore the genome to a disease-free state without the introduction of foreign genetic elements, eliminating the widely shared concern of increased tumorigenic risk. Alternatively, the therapeutic genetic sequence can be designed to persist within cells as a stable episome, allowing maintenance of long-term expression without genomic integration31.
Introduced nucleic acids can be further divided into protein-coding and non-coding sequence. While protein-coding sequences serve as templates for protein production, non-coding nucleic acids function to modulate epigenetic processes controlling gene expression. A good example is RNA interference, in which microRNA (miRNA)38 or small interfering RNA (siRNA)38 bind to mRNA molecules to halt protein translation and mediate mRNA degradation.
Based on the mechanism of action, gene therapy can be categorized into gene addition, gene correction or gene subtraction. In gene addition, exogenous genetic sequence is introduced into cells. Gene correction alters the diseased loci. Attenuating or silencing the expression of single or a network of genes is known as gene subtraction.
The root causes of diseases dictate the gene therapy strategy, which differs dramatically among diseases (Table 2). Genomic mutations sometimes interrupt the reading frame of the protein-coding genes, leading to the absence of proteins, as occurs with dystrophin39 and myotubularin40 in DMD and XLMTM. Consequently the focus of gene therapy has been devoted to supplement a protein-coding sequence to replace the defective gene. Alternately, the mRNA splicing event can be modified to bypass the mutated region, restoring the reading frame with a truncated but functional protein product41. In other cases, where the disease is caused by pathogenic overexpression, it is imperative to silence the gene expression42. Once such example is facioscapulomumeral muscular dystrophy (FSHD), which is caused by the overexpression of the myopathic DUX4 gene18.
The goal of gene therapy in treating congenital myopathies is two-folds. For some monogenic diseases, the current technology is adequate to completely restore the genetic defect in somatic cells, representing a cure that is unachievable by any other methods. Recent findings in canine and mouse models of XLMTM are good examples, where a single treatment recovered animals with a severe monogenic disease to normal function2. For myopathies with complex genetic makeup, a more realistic goal is to delay disease progression and preserve muscle function in order to maintain life quality.
There are several advantages associated with gene therapies in comparison to conventional pharmacotherapies.
Ongoing clinical trails are mainly targeting diseases with defective dystrophoglycan complex, including DMD, BMD and LGMD. Local delivery of full-length dystrophin plasmid to patients’ radialis muscle results in detection of dystrophin mRNA and protein in 6 out of 9 patients45. Encouraged by this study, various gene therapy strategies have been employed in clinical trails aiming to restore dystrophin expression, including read-through agents (Ataluren/PTC124, Arbekacin Sulfate/NPC14 and Gentamicin), exon skipping oligonucleotides (AVI-4658, Drisapersen, Pro044/045/053, SRP4045/4043 and NS-065/NCNP-01) and virus mediated delivery of protein encoding genes, such as minidystrophin and Follistin. Mendell et al. reported delivery of truncated minidystrophin gene (NCT00428935) elicited dystrophin specific T cells, without direct visualization of the protein in muscle, suggesting immune response to be a major hurdle for successful gene therapy15. The Phase 2a study of the read-through agent ataluren reported 61% patients demonstrated increase of dystrophin expression after a course of 28 days treatment (NCT00264888)46. Intramuscular injection of the exon skipping agent AVI-4658 (NCT00159250) resulted in 17% increase of the mean dystrophin signal, reaching 22% to 32% of the healthy control47. For LGMD, AAV packaged alpha and gamma sarcoglycan are now being tested in clinical trails. Local injection of AAV1.tMCK.hSGCA (NCT00494195, NCT01976091) led to persistent α-sarcoglycan expression up to 6 months and augment muscle fiber size43,48. A trial with AAV1-γ-sarcoglycan vector in LGMD 2C patients has completed though result has not yet been disclosed (NCT01344798). The ongoing clinical trials are summarized in the table 3.
A critical element of gene therapy is the vector, the vehicle that facilitates the transfer of genetic material. Due to the size and negative charge of ribonucleic acid, shuttles are needed to carry the cargo across biological barriers to reach target cells. The vector toolbox is composed of viral vectors and non-viral vectors. Viruses have naturally evolved sophisticated machinery to target cells with high efficiency, making them the ideal tool for gene transfer. A summary of the advantages and disadvantages of commonly used viral vectors is outlined in Table 4.
AAV is the preferred viral vector for many ongoing clinical trials, largely due to the fact that AAV efficiently targets post-mitotic parenchymal cells, such as neurons and skeletal myofibers, which are usually inpermissive to other vectors50. Unlike integrating viral vectors, AAV exists as epichromosome, reducing the likelihood of insertional mutagenesis while maintain long-term transgene expression. Another advantage of AAV is that the vector genome can be pseudotyped with alternative capsids. The viral capsid largely determines tissue tropism, gene transfer efficacy and the vector’s dose-dependent toxicity. For example AAV 1, 6, 9 transduce muscle with high efficiency51. This array of available serotypes allows the vector to be tailored to different applications. Though many natural AAV variants have been identified, efforts are being devoted to engineer synthetic viral capsids to address specific clinical challenges. Mutating tyrosine residues (Y445F and Y731F) in the AAV6 capsid can improve the skeletal muscle gene transfer52. Moreover, retention of AAV vector in liver has been a problem of systemic infusion. Vectors that “de-target” liver have been created through randomly mutating the surface-exposing region of AAV9 capsid (AAV9.45 and AAV9.61)53 or by engineering a chimeric capsid with AAV2 and AAV8 (AAV2i8)54. Both strategies redirected the vector away from liver while maintaining high transduction efficiency to skeletal muscle.
Viral vectors have been widely used in ongoing gene therapy clinical trials1. Despite their differences, however, the use of any viral-based vector carries with it certain risks, including tumorigenicity, immunogenicity and limited cargo space. Most of these risks can be minimized by non-viral vectors. On the other hand, the main drawbacks of non-viral vectors, including the capacity to cross various biological barriers and stability, have been largely addressed by late breakthroughs. For example, material science has provided various lipid-based and polymer-based DNA vectors, many of which have been tested in clinical trials55. Nucleic acid chemistry evolution has resulted in the development of nucleic acid analogues, such as 2′-O-methyl-phosphorothioate (2′OMe) and morpholino phosphorodiamidate oligonucleotide (PMO). While retaining the same nucleo-base to enable Watson-Crick base pairs with natural nucleotides, ribonuclease-resistant moieties have replaced natural ribose ring and backbone phosphodiester linkage. These modifications lead to increased molecular stability against enzyme degradation. Direct infusion of analogue oligonucleotides is associated with efficient targeting 56–59. 2′OMe and PMO have been employed as antisense oligonucleotide (AON) to mediate exon skipping in DMD treatment (Table 5). Both Drisapersen (2′OMePS AON to exon 51) and Eteplirsen (PMO morpholino to exon 51) have also demonstrated restored dystrophin expression and even mild clinical improvement as measured by 6 minute walk60,61.
The success of gene therapy is largely determined by the efficient delivery of vectors to target tissues directly determines the success. Effective delivery approaches include direct injection, locoregional perfusion and systemic delivery.
Direct injection into the target tissue is commonly used to determine the efficacy of a potential new therapy2,62–64. This method ensures that the desired organ receives the necessary therapeutic dose and restricts the treatment to that organ, reducing off-target effects. However, distribution may be limited, even within the injected muscle65. For example, XLMTM dogs treated by AAV8-MTM1 injection into the cranial tibialis of the hindlimb show improvements in the strength, size and histopathology of the treated limb2 but show a continued progression of the disease including muscle weakness, impaired ambulation and early death2,66. This limited distribution is therefore disadvantageous when the disease affects multiple systems or the entire body, as is often the case with congenital myopathies. Also, direct treatment of organs like the heart or diaphragm may require invasive surgery or complicated techniques67,68, which can be difficult in chronically ill-patients.
Locoregional perfusion, where a limb is isolated before intravascular infusion under high pressure, is another approach to gene therapy delivery. Despite the use of high pressure, it is a safe, relatively painless option for human patients12 and has been used successfully in animal models, including the XLMTM dogs, where there was widespread clinical improvement2,69,70. However, vector titers may be reduced due to the low permeability of the vascular endothelium71 and, like direct injection, isolated locoregional perfusion does not address treatment of the cardiorespiratory system.
In systemic delivery, a potential gene therapy vector is introduced to the entire body. This is of particular importance in congenital myopathies where cardiorespiratory failure is the leading cause of death but there are effects of the disease throughout the body.72,73 As with locoregional perfusion, the vascular endothelium could hinder distribution to the skeletal muscle and cardiorespiratory systems most affected by disease.71,74 Systemic dosing also increases the likelihood of off-target gene delivery, which may require the use of additional safety measures like tissue-specific promoters. Finally, higher doses may be required for systemic delivery, due to circulating antibodies and filtration by the liver. However, modifications of the vector or immunosuppression can address these problems.75–77
Reliable model systems are indispensible for the critical transition from bench to bedside. It is required by the FDA the potency and toxicity of vectors be validated on several levels of preclinical models, before entering clinical trials. The following section summarizes model systems at different level from cell culture to animal models.
Cell culture models can be used to test the potency and cellular toxicity of a biological therapy. There are several advantages, such as intricate manipulation of experiment conditions and considerably lower costs. Early phase vector validations are usually carried out in relevant cell cultures. Immortalized mammalian human embryonic kidney (HEK) 293 cells and C2 myoblasts were employed to validate the vector expressing a truncated dystrophin protein78. Engineered HEK293 cells over-expressing pathogenic DUX4 gene has been utilized to confirm the efficacy of RNAi vector42. Cells derived from diseased tissues, such as myoblasts isolated from animal models or human patients, have been utilized to test potency for vectors developed for DMD79, Pompe disease80 and myotonic dystrophy44.
Due to physiological differences between animal models and human patients, it is not rare therapies that are effective in diseased model animals failed to show benefits in clinical trials81, including DMD82. Moreover, mutations may vary among patients, which requires vector personalization, such as oligonucleotides for exon skipping therapy83. Consequently a personalized human cell culture system is highly desirable for potency testing.
The emergence of human iPSC offers a novel cell culture platform to meet both these criteria. Unlimited disease relevant cells with specific patient mutations could be generated. More importantly, these cells demonstrate disease-associated phenotypes reflecting disease severity84. As a result, correction of these disease-associated phenotypes can be measured as efficacy. Moreover, testing vectors on these personalized cells will not only demonstrate the presence of the transgene product85, but also enables quantified measurement of biological activities, as required by the FDA for gene therapy products.
While in vitro models allow for the testing of potential therapies in human cells, the ability to adequately recreate the complexities of the human body is limited. As such, preclinical testing in an animal model remains the gold standard for investigating the efficacy and potential toxicity of a putative treatment.
Non-mammalian models such as zebrafish have been invaluable to the understanding of disease mechanisms within a complex organism, particularly in monogenetic neuromuscular disorders.86,87 However physiological and phenotypic dissimilarities with humans limit their translational power88,89 (Table 5).
Small mammalian models like rodents are often used in preclinical assessments of efficacy and toxicity. With extensive research into study of the murine genome,90 mice in particular are a powerful tool in the study of monogenetic disorders and knockdown or knockout mouse models have been developed to study musculoskeletal disease.
However the small size of the mouse as well as anatomical and phenotypic differences make them less than ideal candidates for the preclinical assessment of gene therapies.91 Therefore, a larger animal model like a dog, where organ size more closely approximates that of humans, may be more suitable. Methods of functional assessment developed for use in the clinic have also been successfully adapted for use in dogs.22 Many naturally-occurring musculoskeletal diseases, similar to those seen in human patients, have been identified and characterised92 (Table 5) facilitating the use of that dog model for the pre-clinical assessment of potential treatments.2,93,94
AAV8-mediated delivery of myotubularin in the XLMTM dog provides a successful example of gene therapy in preclinical practice. Due to the gene’s small size, a full-length canine MTM1 cDNA was carried by the muscle-tropic AAV8. This was packaged with a human desmin promoter to ensure muscle specific expression. Dogs were treated by direct injection and by locoregional hindlimb perfusion, where systemic effects were observed, likely due to leakage. Treated dogs have near normal muscle strength, with the marked improvements in respiratory function and increased survival for dogs treated intravascularly. Indeed, these animals continue to survive more than two years after treatment and continue to thrive and breed. The size of the dog has allowed for the inclusion of many at other relevant assessments at multiple time points including neurological scoring, MRI, EMG and gait testing as we investigate potential outcome measures for translation of this therapy to human patients.
Funding: American Heart Association fellowship to X.G., Muscular Dystrophy Association to M.K.C, Association Française contre les Myopathies to M.K.C.; NIH grants R21 AR064503 and R01 HL115001 to M.K.C.; Joshua Frase Foundation to M.K.C.; Where There’s a Will There’s a Cure to DM and MKC; Peter Khuri Myopathy Research Foundation to M.K.C.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.