Progress in gene therapy of dystrophic heart disease

Abstract

The heart is frequently afflicted in muscular dystrophy. In severe cases, cardiac lesion may directly result in death. Over the years, pharmacological and/or surgical interventions have been the mainstay to alleviate cardiac symptoms in muscular dystrophy patients. Although these traditional modalities remain useful, the emerging field of gene therapy has now provided an unprecedented opportunity to transform our thinking/approach in the treatment of dystrophic heart disease. In fact, the premise is already in place for genetic correction. Gene mutations have been identified and animal models are available for several types of muscular dystrophy. Most importantly, innovative strategies have been developed to effectively deliver therapeutic genes to the heart. Dystrophin-deficient Duchenne cardiomyopathy is associated with Duchenne muscular dystrophy (DMD), the most common lethal muscular dystrophy. Considering its high incidence, there has been a considerable interest and significant input in the development of Duchenne cardiomyopathy gene therapy. Using Duchenne cardiomyopathy as an example, here we illustrate the struggles and successes experienced in the burgeoning field of dystrophic heart disease gene therapy. In light of abundant and highly promising data with the adeno-associated virus (AAV) vector, we have specially emphasized on AAV-mediated gene therapy. Besides DMD, we have also discussed gene therapy for treating cardiac diseases in other muscular dystrophies such as limb-girdle muscular dystrophy.

INTRODUCTION

Muscular dystrophy refers to a heterogeneous group of inherited muscle diseases. Patients exhibit progressive muscle wasting and force decline. Although the molecular pathogenesis of muscular dystrophy has not been fully elucidated, it is generally agreed that mechanical injury may play an important role. Contraction-induced cell shape change distinguishes muscle cells from other body cells. Deformation creates tremendous stress on the muscle cell membrane. In the absence of an appropriate protective mechanism, the sarcolemma is broken. Membrane disruption leads to myofiber degeneration, necrosis, fibrosis and eventually muscular dystrophy.

Compared with skeletal muscle, the demand for a stable sarcolemma is even higher for the myocardium because of the constant pumping activity of the heart. As a result, cardiac involvement is regularly seen in muscular dystrophy. Cardiomyopathy and/or heart arrhythmia are typical clinical presentations of dystrophic heart disease. Lethal consequences include heart failure and sudden cardiac arrest. Symptom-based therapies have been practiced for years to attenuate cardiac complications in muscular dystrophy patients. Yet, palliative treatments offer only limited clinical benefit. The ultimate cure for dystrophic heart disease will require genetic correction of the defective genes.

The identification of the disease gene, and development of gene transfer vector and animal models have established a strong foundation for exploring gene therapy. Genetic mutations in many types of muscular dystrophy are now defined. Interestingly, a lot of these mutations involve components of dystrophin-associated glycoprotein complex (DGC). The DGC is a multiprotein structure that links cytoskeleton F-actin (γ-actin) with laminin in the extracellular matrix. It mainly consists of dystrophin, α- and β-dystroglycan, and α-, β-, γ- and δ-sarcoglycan. Additional components of the DGC include sarcospan, syntrophin, dystrobrevin and neuronal nitric oxide synthase.¹ Numerous lines of evidence suggest that the DGC may function as a molecular shock absorber and protect the sarcolemma from mechanical strain-induced damages.² Deficiency of the DGC compromises the cytoskeleton–extracellular matrix connection and destabilizes the sarcolemma.^3,4

An ever-increasing armory of viral and non-viral vectors has been developed for heart gene transfer.⁵ Unfortunately, most are either inefficient or are associated with strong inflammation/immune reaction. The only exception is adeno-associated virus (AAV). AAV is a non-enveloped single-strand DNA virus.⁶ Wild-type AAV contains an ~5-kb genome. Except for the ~145-bp inverted terminal repeat (ITR) at the ends, the entire viral genome can be replaced with a therapeutic expression cassette. Early experimentation with AAV serotype-2 and -5 showed persistent cardiac transduction following direct myocardial injection, trans-coronary delivery or intra-cavity injection.^7–9 Now, a broad range of naturally occurring and synthetic AAV capsids have been engineered for cardiac targeting.⁵ Impressively, these newly generated AAV serotypes (such as AAV-6, 8 and 9) result in robust gene expression in the heart even if delivered through a peripheral vein.^10–14

Establishment of reliable animal models is another crucial factor. It provides a stage for pre-clinical testing. Several rodent and canine models have been carefully characterized. They recapitulate representative clinical features of dystrophic heart diseases. Some commonly used models are dystrophin-null mdx mice and dogs and cadiomyopathic hamsters.^15–19

The majority of current gene therapy studies are focused on dystrophin-deficient cardiomyopathy. Some studies have also been conducted for δ-sarcoglycan-null cardiomyopathy. For this reason, we have emphasized these topics in this review article. The experience learned from these studies will provide valuable guidance for gene therapy of cardiac diseases in other types of muscular dystrophy.

DUCHENNE CARDIOMYOPATHY GENE THERAPY

Duchenne cardiomyopathy refers to cardiac diseases seen in Duchenne muscular dystrophy (DMD), Becker muscular dystrophy (BMD) and X-linked dilated cardiomyopathy (XLDC). The common genetic underpinning is dystrophin deficiency. DMD is the most common and severe form of muscular dystrophy. In DMD dystrophin is completely lost in all striated muscles. BMD refers to a mild form caused by partial loss of dystrophin (quantitative or qualitative). In XLDC, dystrophin is selectively removed from the heart. Dilated cardiomyopathy is the characteristic clinical manifestation in Duchenne cardiomyopathy. Patients may die from heart failure and/or ventricular arrhythmia. As dystrophin gene mutation is the molecular culprit in Duchenne cardiomyopathy, replacement or repair of the defective dystrophin gene has been considered as a promising cure.

Gene replacement with synthetic dystrophin genes

Gene delivery to dystrophin-deficient heart

Several AAV serotypes have been tested in mdx mice, a widely used mouse model. The first successful gene delivery to the mdx heart was achieved with AAV-5.⁸ Virus was directly injected into the cardiac chamber in hypothermia-shocked neonatal mdx mice. Although meaningful as a proof-of-principle, the invasive protocol used in the study is unlikely going to translate to human patients. Recently, several groups evaluated AAV-6 and AAV-9. Impressive whole heart transduction was obtained following systemic gene delivery. Importantly, these serotypes were not only effective in neonatal and young adult mdx mice but they also worked in the heart of aged mdx mice.^20–26

Dystrophin gene abbreviation

The full-length dystrophin protein is encoded in an ~14-kb mRNA. It contains four structural domains, including the N-terminal, rod, cysteine-rich and C-terminal domains. The rod domain is further divided into 24 spectrin-like repeats and four intervening hinges. The size of the dystrophin gene presents a great challenge to AAV-mediated gene transfer.

Much effort has been devoted to the development of minimized synthetic dystrophin genes.^27,28 The goal is to generate abbreviated dystrophin variants that are small enough for viral vector packaging but yet retain the critical functional domains. Mini-dystrophin and micro-dystrophin are two categories currently under investigation. The minigene is ~6–8 kb and it carries ≥50% of the coding sequence. In mini-dystrophin, part of the rod domain is removed. The microgene is ~4 kb and it contains only about one-third of the coding sequence. Micro-dystrophin loses a significant portion of the rod domain and the C-terminal domain. In terms of AAV gene transfer, the biggest difference is that the microgene can fit into a single AAV virion but the minigene requires the dual vector system for delivery.²⁹

Gene therapy of Duchenne cardiomyopathy with mini-dystrophin

Several pairs of dual AAV vectors have been developed to express mini-dystrophin.^23,30–32 In dual vector system, a complete minigene cassette (including the promoter, minigene and polyadenylation signal) is engineered into two independent AAV viruses. Mini-dystrophin expression is achieved after co-infection and inter-viral recombination. The first working set of the minigene dual vectors was built with the trans-splicing strategy.³² In this system, minigene expression is reconstituted via ITR-mediated end-to-end vector genome joining and subsequent splicing using pre-engineered splicing signals.^33,34 Two independent sets of over-lapping mini-dystrophin vectors were reported recently.^23,30 In the over-lapping approach, the minigene is split into two overlapping fragments and separately packaged. Reconstitution is achieved through the shared region of homology.³⁵ The hybrid dual vectors are a new addition.³¹ It allows reconstitution through either ITR-mediated end joining (trans-splicing) or transgene sequence-mediated homologous recombination (over-lapping). Several minigene hybrid vectors have been described.^30,34 Currently, mini-dystrophin dual AAV vectors have only been tested for treating DMD skeletal muscle disease. Their therapeutic efficacy for Duchenne cardiomyopathy remains to be determined.

The possibility of using mini-dystrophin to treat Duchenne cardiomyopathy has only been investigated in transgenic mdx mice.³⁶ Cardiac specific mini-dystrophin expression was achieved in mdx mice using the alpha-myosin heavy chain promoter.³⁷ Mini-dystrophin completely prevented sarcolemmal leakage and myocardial fibrosis.³⁶ At 22 months of age, several ECG parameters (including the PR interval, QT interval and cardiomyopathy index) were normalized. Cardiac catheter assay also revealed significant improvement. End-systolic volume, maximal pressure, dP/dt max and ejection fraction were indistinguishable from those of normal mice. Nevertheless, dP/dt min, stroke volume and cardiac output remained suboptimal.³⁶

Gene therapy of Duchenne cardiomyopathy with micro-dystrophin

Until now, most Duchenne cardiomyopathy gene therapy studies were conducted with the AAV micro-dystrophin vector.^{8,21,22,24,26,38} The first evidence supporting microgene therapy in the heart was provided by Yue et al.⁸ The authors delivered an AAV-5 microgene vector to newborn mdx heart. Micro-dystrophin restored DGC expression and improved sarcolemmal integrity in cardiomyocytes in vivo.⁸ Subsequent electrocardiographic study provided evidence of physiological benefits in neonatal mdx mice.²⁴ A single intravenous injection of the AAV-9 microgene vector transduced the entire heart in neonatal mdx mice. The heart rate, PR interval and QT interval were normalized up to four months after treatment.²⁴

Two studies examined AAV micro-dystrophin therapy in young adult (4–10-week-old) mdx mice.^22,26 One study (Shin et al.) used AAV-9 and the other (Townsend et al.) used AAV-6. Robust myocardial transduction was observed following systemic gene transfer in both studies. Shin et al.²⁶ observed a significant improvement of the heart rate and PR interval for up to 6 months. They also noticed a reduction of cardiac fibrosis and normalization of fractional shortening on echocardiography.²⁶ ECG was not performed in Townsend et al. study.²² However, hemodynamic assay at 10 weeks post AAV therapy showed an increased end-diastolic volume. Further, AAV microgene treated mice were protected from dobutamine stress-induced acute heart failure.²²

One caveat of above studies is that newborn or young adult mdx mice do not show overt signs of dilated cardiomyopathy.¹⁷ Further, male and female mdx mice exhibit different cardiac phenotypes and this difference has not been appreciated.¹⁶ Characteristic dilated cardiomyopathy has only been demonstrated in very old female mdx mice.^16,17 A recent study by Bostick et al.²¹ evaluated AAV microgene therapy in aged female mdx mice. AAV-9 microgene vector was delivered to 19-months-old female mdx mice. Histopathology and cardiac function were examined four months later. Despite ongoing cardiac disease, impressive AAV transduction was observed throughout the entire heart. Myocardial fibrosis was significantly reduced in treated mdx mice. Importantly, beneficial ECG and hemodynamic changes were observed and dobutamine-induced cardiac death was prevented in all treated mice.²¹ In summary, evidence so far strongly supports further development of Duchenne cardiomyopathy gene replacement therapy with minimized dystrophins.

Repair dystrophin gene through exon skipping

Several thousands of mutations have been reported in DMD patients.^39,40 A common theme of these mutations is that they all disrupt the open reading frame.^41,42 Exon skipping is designed to modulate splicing with antisense oligonucleotides (AON). Removal of mutated and/or adjacent exon(s) restores the open reading frame. The resulting transcript yields an internally truncated, but likely functional protein. There has been tremendous progress in DMD exon skipping in last few years.⁴³ Positive results were obtained from several clinical trials in treating skeletal muscle disease.^44–47

A long-standing hurdle of exon-skipping is its inefficiency in the heart.⁴⁸ Minimal dystrophin expression was observed in the mdx heart even after repeated delivery for 1 year.^49,50 It is now clear that the problem is the poor uptake of AON by cardiomyocytes.⁴³ Increasing AON dose to the extreme level seems to have resulted in some improvement.⁵¹ However, such an approach is clinically unsustainable.⁵² Another approach to enhance AON delivery is via chemical modification. Using AON that is conjugated to cell penetrating peptides, non-peptide polymers or nanoparticles, several groups have now demonstrated widespread myocardial exon skipping in mdx mice.^53–62 Interestingly, peptide-conjugated AON did not induce cardiac exon skipping in dystrophin/utrophin double knockout mice, a much severe DMD model.⁶³ Future studies are needed to clarify this difference.

Cardiac function has been evaluated in a few exon-skipping studies that used peptide conjugated AON. Wu et al.⁵³ treated 4-months-old mdx mice for 3 weeks. Treatment significantly protected mdx mice from dobutamine-induced cardiac dysfunction and death.⁵³ Jearawiriyapaisarn et al.⁵⁹ also noticed beneficial echocardiographic changes in young adult mdx mice. As the heart of young adult mdx mice is only mildly affected, future studies in a symptomatic model (such as aged female mdx mice) may provide more rigorous support for exon skipping mediated Duchenne cardiomyopathy gene therapy. Other issues that also need the attention are the potential toxicity and immunogenicity of modified AONs.^52,64 Addressing these issues may pave the way for future clinical translation.

Alternative gene therapy strategies for Duchenne cardiomyopathy

As the primary cause of Duchenne cardiomyopathy is dystrophin deficiency, it is logical to emphasize on strategies that aim at restoring dystrophin expression (mini- and micro-dystrophin replacement or gene repair with exon skipping). However, there are some limitations. First is the immune response to viral vector and/or newly expressed dystrophin. Recent studies in dystrophin-deficient dogs and human DMD patients suggest that untoward immune rejection is a major obstacle.^65–68 Another issue is the incomplete recovery of cardiac function.³⁶ Little is known about the structure–function relationship of dystrophin in the heart. Cardiac-specific domain(s) have been suspected but never been clarified.^69,70 In the absence of this knowledge, it is very unlikely that the current gene replacement/repair strategies will fully meet the need. As a matter of fact, one mini-dystrophin gene that is known to normalize skeletal muscle force did not completely restore heart function.³⁶ Exploring alternative gene therapy strategies may enhance dystrophin-based therapies.

Utrophin for Duchenne cardiomyopathy gene therapy

Utrophin is a homolog of dystrophin.⁷¹ Despite some important differences between two proteins,^72,73 many lines of evidence suggest that utrophin may greatly improve muscle function in dystrophin-deficient subjects.⁷⁴ Several studies also suggest that utrophin may have helped preserving heart function in young adult mdx mice and heterozygous mdx mice.^17,75 As utrophin is already expressed in DMD patients, exogenous utrophin expression from a gene therapy vector will likely be tolerated by our immune system. Similar to dystrophin, the large size of the utrophin gene also presents a significant challenge to gene delivery. Development of minimized synthetic utrophin genes may overcome this hurdle.^76,77

Sarcoplasmic reticulum calcium ATPase 2a (SERCA2a)

SERCA2a is an enzyme that pumps cytosolic calcium back to the sarcoplasmic reticulum in the heart. There are several advantages in developing SERCA2a-based gene therapy for Duchenne cardiomyopathy. As an endogenous gene, it posses the same immune advantage as utrophin (less immunogenic). Unlike utrophin, the SERCA2a gene readily fits into a single AAV vector.⁷⁸ Accumulated evidence suggests that calcium overloading may represent a major pathogenic mechanism in Duchenne cardiomyopathy.⁷⁹ Forced expression of SERCA2a via AAV gene transfer may restore calcium homeostasis and improve cardiac contractility. Only one study has tested the therapeutic efficacy of SERCA2a in the mdx heart. Shin et al.⁷⁹ delivered AAV SERCA2a vector to 12-m-old mdx mice and examined ECG eight months later. Significant improvement was observed in treated mice.⁷⁹

Recent progresses have lent more support to SERCA2a gene therapy. First, AAV SERCA2a gene therapy has shown encouraging results in patients with advanced heart failure.^78,80,81 Second, it was found that SERCA2a is highly regulated by a protein called small ubiquitin-related modifier 1 (SUMO1).⁸² AAV-mediated SUMO1 overexpression offers another opportunity to elevate SERCA2a expression/ activity and hence improve heart function.⁸²

Other possible candidate genes for Duchenne cardiomyopathy gene therapy

Besides utrophin and SERCA2a, many other endogenous genes may also be worth considering. Some examples include AAV-mediated vascular endothelial growth factor expression and AON-mediated microRNA-208 knockdown.^83,84 These strategies have been shown to treat other types of dilated cardiomyopathy. Although not tested yet for Duchenne dilated cardiomyopathy, they may be worth pursuing either alone or in combination with dystrophin replacement/repair therapy.

Additional considerations in Duchenne cardiomyopathy gene therapy

Therapeutic threshold

Dystrophin is lost in all cardiomyocytes in Duchenne cardiomyopathy. However, in a gene therapy scenario we may not be able to correct every single cell in the heart. It is thus important to determine the minimal level of gene transfer needed for the treatment. Heterozygous mdx carriers display a mosaic pattern of dystrophin expression in approximately 50% of cardiomyocytes. Interestingly, these mice are completely protected from cardiomyopathy. ^17,85 At 21 months of age, there are minimal signs of histological lesions. Further, cardiac function (ECG and hemodynamics) is normalized. ¹⁷ Human DMD carriers also show 50% mosaic dystrophin expression. Importantly, the majority does not develop cardiomyopathy. ^86–89 Taken together, it seems reasonable to set a 50% transduction rate as the therapeutic threshold.

Currently, there is no concrete data on whether expression below 50% can protect the heart. Interestingly, a recent exon skipping study revealed a quite surprising finding.⁵⁰ The investigators induced <5% dystrophin positive cells in the heart in adult mdx mice. No change was found in cardiac histology and baseline function. However, an improved response was noticed under dobutamine stress.⁵⁰

Therapeutic time window

Therapeutic time window is an important consideration. Dystrophin expression has been detected during embryonic development in both rodent and human in both skeletal muscle and the heart.^90–93 One study in mdx mice even suggests that dystrophin deficiency may severely impair myogenesis.⁹⁴ However, this may not necessarily mean the therapeutic window will be missed if gene therapy is not initiated during fetal development. Several lines of evidence suggest that embryonic dystrophin deficiency does not have immediate developmental consequences. For example, neonatal mdx mice appear normal.⁹⁵ Similarly, dystrophic symptoms have not been observed in newborn DMD patients.⁹⁶ It is possible that predominant utrophin expression during embryogenesis may have compensated for the loss of dystrophin.^97–100

Transgenic study with the alpha-myosin heavy chain promoter also suggests that prenatal therapy is not absolutely required for heart disease amelioration.³⁶ This promoter is not active in the ventricle during fetal development.^37,101 Yet, cardiomyopathy is significantly mitigated in mini-dystrophin transgenic mice that are under transcriptional regulation of the alpha-myosin heavy chain promoter.³⁶ Taken together, additional studies are needed to determine whether absence of dystrophin leads to irreversible developmental defects. In the meantime, considering exciting preclinical data already published, it is certainly worth to further pursue postnatal gene therapy. On the other side, with the advance in prenatal and newborn screening, attempts on fetal and/or neonatal gene transfer should also be encouraged. Such an approach may allow immediate therapy following early diagnosis.

One interesting concern on fetal and/or neonatal gene transfer is the loss of the episomal AAV vector genome during postnatal heart growth. Although a valid concern, it may not be a serious problem. Cardiomyocyte regeneration is essentially terminated after the first week of birth.^102–104 Postnatal heart size increase is mainly due to hypertrophy rather than hyperplasia.¹⁰⁴

Impact of skeletal muscle disease and smooth muscle dystrophin deficiency on Duchenne cardiomyopathy gene therapy

Different from diseases that only affect the heart, there is often concurrent involvement of skeletal muscle in muscular dystrophy-associated heart diseases. It has been hotly debated whether severity of skeletal muscle disease influences the outcome of cardiomyopathy and whether treating skeletal muscle disease benefits or harms the heart.^105,106

Several mouse studies have revealed a direct correlation between skeletal muscle disease and cardiomyopathy. Genetic inactivation of skeletal muscle-specific myogenic transcription factor myoD not only worsens skeletal muscle disease, it also induces a severe cardiac phenotype in young adult mdx mice.^106,107 Recently, Crisp et al.¹⁰⁸ published a very intriguing result. The authors performed skeletal muscle specific exon skipping in dystrophin-null mice. Although dystrophin was not restored in the heart, the authors reported normalization of heart function by magnetic resonance imaging assay (unfortunately, no histological data were presented).¹⁰⁸ The authors concluded that there is no need to treat the heart.¹⁰⁸ In other words, cardiac dysfunction will recover automatically when skeletal muscle disease is under control.

This is quite surprising as a similar exon skipping study by a different group showed just the opposite result.⁴⁹ Using histology examination, Malerba et al.⁴⁹ found a more severe cardiac pathology in mdx mice that were only treated for skeletal muscle disease by exon skipping. Along the same line, Townsend et al.¹⁰⁹ demonstrated that transgenic rescue of skeletal muscle disease alone worsens heart function in young adult mdx mice. In summary, studies from Malerba et al. and Townsend et al. suggest that the heart has to be treated. Curing only skeletal muscle may aggravate heart disease. Clinical data from XLDC also suggests that we may indeed need to treat the heart. Dystrophin expression is compromised in the heart but not skeletal muscle in XLDC.¹¹⁰ Despite a lack of apparent skeletal muscle disease, XLDC patients develop severe dilated cardiomyopathy.¹¹⁰ Apparently, additional studies are needed to resolve the conflict. In the mean time, considering all the pros and cons, the primary focus for Duchenne cardiomyopathy gene therapy should set to restore dystrophin expression in the heart, or to treat both skeletal muscle and the heart simultaneously.

Another interesting consideration is the coronary vasculature. Under normal condition, dystrophin is also expressed in vascular smooth muscle.^91,111,112 Loss of dystrophin induces remarkable biomechanical changes in the vasculature.^113–115 Recent studies further suggest that selective restoration of dystrophin in smooth muscle or increasing the vasculature alone may reduce DMD skeletal muscle disease.^116,117 However, therapeutic effect of dystrophin expression in the coronary vasculature has not been determined. As systemic AAV gene delivery has been shown to transduce smooth muscle,¹⁰ it is possible that the success of current AAV micro-dystrophin gene therapy may partially attribute to the restoration of dystrophin in the coronary vasculature.^21,22,24,26 Nevertheless, direct experimental evidence is lacking. Future studies are needed to either support or reject this intriguing hypothesis.

GENE THERAPY OF CARDIOMYOPATHY IN LIMB-GIRDLE MUSCULAR DYSTROPHY 2F (LGMD2F)

LGMD2F is an extremely rare form of muscular dystrophy.¹¹⁸ It is caused by δ-sarcoglycan deficiency. Although most patients show skeletal muscle dysfunction and late-onset cardiomyopathy, a subgroup of patients develop severe dilated cardiomyopathy without obvious skeletal muscle illness. Despite the scarce incidence, δ-sarcoglycan deficient cardiomyopathy has been frequently used in proof-of-principle gene therapy studies. The reason for this is two-fold. The animal models are readily available and the δ-sarcoglycan gene is fairly small. Naturally occurring cardiomyopathic hamsters have been around for several decades.¹¹⁹ These hamsters carry a deletion mutation in the δ-sarcoglycan gene. They exhibit hypertrophic (BIO 14.6 strain) or dilated (TO-2 strain) cardiomyopathy.¹⁸ In addition to the hamster model, two independent lines of δ-sarcoglycan knockout mice have also been generated.^120,121 The small size of the gene is another advantage. It allows easy packaging into any viral vector including AAV.

Early studies used adenovirus to express the δ-sarcoglycan gene.^122,123 Efficient myocardial transduction is achieved but expression does not last long.¹²³ This obstacle is overcome with AAV vectors. Persistent δ-sarcoglycan expression was first obtained in the hearts of TO-2 and BIO14.6 hamsters using AAV-2. A drawback of AAV-2 is that it requires highly invasive technique (such as intramyocardial injection and ex vivo coronary perfusion) for gene delivery.^124–126 Systemic gene delivery is clinically more appealing. This strategy was initially tested with an AAV-8 δ-sarcoglycan vector.¹²⁷ Efficient myocardial transduction was observed following intraperiotoneal and intravenous injection in neonatal and young adult TO-2 hamsters. Histopathology (myocardial fibrosis, calcification, and necrosis) was completely eliminated and heart function was normalized on echocardiography examination.¹²⁷ Recently, studies from several independent groups have further confirmed this exciting result in BIO14.6 hamsters and δ-sarcoglycan-null mice with AAV-1, 8 and 9.^128–130

Besides δ-sarcoglycan gene replacement, investigators have also explored other alternative gene therapy strategies. In one study, Hoshijima et al. tested a phospholamban inhibiting AAV-2 vector. Disturbance of calcium homeostasis has an important role in the pathogenesis of LGMD2F.¹¹⁸ Phospholamban ablation has been shown to improve sarcoplasmic reticulum calcium cycling in mouse cardiomyopathy models.¹³¹ Encouragingly, trans-coronary AAV-2 gene therapy enhanced heart function in young adult BIO14.6 hamsters.⁹ Thus, aiming at downstream events may open another promising avenue to treat δ-sarcoglycan-deficient cardiomyopathy.

Similar to DMD, δ-sarcoglycan deficiency also compromises skeletal muscle. Two studies tested whether gene transfer to the heart alone is sufficient for treating cardiomyopathy in LGMD2F. Goehringer et al.¹²⁹ achieved heart-specific expression with a cardiac muscle specific promoter. Yang et al.¹³² generated a relatively myocardium-specific AAV capsid using DNA shuffling technique. Interestingly, heart restricted expression significantly reduced myocardial fibrosis and improved cardiac function in δ-sarcoglycan-null mice and TO-2 hamsters.^129,132

Untoward immune response to newly expressed dystrophin is a serious concern in DMD gene therapy.^65–68 However, immune rejection has not been encountered in AAV or even adenovirus-mediated δ-sarcoglycan gene delivery in rodent models.^{122,124,125,127–130} Whether this will hold up in human patients remains to be seen. Interestingly, a recent AAV α-sarcoglycan clinical trial has revealed a quite safe immune profile. No T cell response to α-sarcoglycan was detected.^133,134 Nevertheless, caution should be taken when we proceed to human trial. Collectively, animal study results suggest that gene therapy may represent a viable approach to treat cardiomyopathy in LGMD2F.

GENE THERAPY OF CARDIAC COMPLICATIONS IN OTHER TYPES OF MUSCULAR DYSTROPHY

Cardiac complications are universally observed in muscular dystrophies. ^135,136 However, besides dystrophin-deficient cardiomyopathy and δ-sarcoglycan-deficient cardiomyopathy, gene therapy has been barely explored to treat heart diseases in other types of muscular dystrophy. With the promising pre-clinical results in Duchenne and LGMD2F cardiomyopathy, it is expected that there will be a growing interest in testing gene therapy in other dystrophic heart diseases. Several points should be considered in the design of these studies. First is the nature of the mutation. Many muscular dystrophies are caused by dominant mutations. In these cases, the strategy should be focused on the elimination of the abnormal allele (for example, via RNA interference) rather than gene supplementation.^137,138 Another issue is the clinical presentation. In Emery–Dreifuss muscular dystrophy and myotonic dystrophy, arrhythmia rather than dilated cardiomyopathy is more frequently encountered.¹³⁶ Innovative means are needed to tailor gene therapy protocols to meet different clinical situations. Third, despite the lack of heart-focused gene therapy studies, there have been considerable progresses in gene therapy of skeletal muscle diseases in many muscular dystrophies. The lessons learned from these studies should be incorporated in the development of dystrophic heart disease gene therapy.

SUMMARY AND PERSPECTIVE

With the ability to reverse the fundamental genetic defect, gene therapy has demonstrated immense potential to cure dystrophic heart disease. A strong foundation has been established from treating cardiac complications in rodent models of Duchenne cardiomyopathy and LGMD2F. With the ongoing research activities, we have no doubt that more powerful gene delivery vectors (such as an AAV vector with reduced immunogenicity/improved transduction efficiency) will emerge in the coming years. However, despite the rosy picture, there remain significant obstacles for an unequivocal clinical success in the near future. For LGMD2F, because of the extreme low incidence it will be a great challenge to assemble enough patients for a standard clinical trial. For DMD, reproducing murine results in dystrophic dogs may represent a more realistic goal.²⁸