Prevention of exercised induced cardiomyopathy following Pip-PMO treatment in dystrophic mdx mice

Duchenne muscular dystrophy (DMD) is a fatal neuromuscular disorder caused by mutations in the Dmd gene. In addition to skeletal muscle wasting, DMD patients develop cardiomyopathy, which significantly contributes to mortality. Antisense oligonucleotides (AOs) are a promising DMD therapy, restoring functional dystrophin protein by exon skipping. However, a major limitation with current AOs is the absence of dystrophin correction in heart. Pip peptide-AOs demonstrate high activity in cardiac muscle. To determine their therapeutic value, dystrophic mdx mice were subject to forced exercise to model the DMD cardiac phenotype. Repeated peptide-AO treatments resulted in high levels of cardiac dystrophin protein, which prevented the exercised induced progression of cardiomyopathy, normalising heart size as well as stabilising other cardiac parameters. Treated mice also exhibited significantly reduced cardiac fibrosis and improved sarcolemmal integrity. This work demonstrates that high levels of cardiac dystrophin restored by Pip peptide-AOs prevents further deterioration of cardiomyopathy and pathology following exercise in dystrophic DMD mice.

D uchenne muscular dystrophy (DMD) is a severe, debilitating, muscle wasting disorder caused by mutations that disrupt the reading frame of dystrophin mRNA, thus preventing production of this essential structural protein 1 . Cardiomyopathy is prominent in DMD patients and manifests with left ventricular dilation and hypertrophy, decreased fractional shortening and electrocardiogram (ECG) abnormalities [2][3][4][5] . As the disease progresses, dilated cardiomyopathy (DCM) develops, a condition prevalent in all DMD patients by adulthood 3 . While respiratory complications are still the major cause of death amongst DMD patients 6 , cardiomyopathy contributes significantly to premature death 7 . A number of cardiac pharmacotherapies are in clinical use 8 , however, no therapy is currently capable of correcting the underlying genetic defect and restoring dystrophin protein in the heart.
A multitude of experimental therapies have been investigated in an effort to restore [9][10][11] or substitute 12,13 for dystrophin. Antisense oligonucleotides (AOs) act to modulate splicing of dystrophin pre-mRNA to generate a truncated, but functional dystrophin protein [14][15][16] . 29 O-Methyl phosphorothioate (2OMePS) and phosphorodiamidate morpholino oligomer (PMO) AO chemistries have shown great promise in recent clinical trials [17][18][19][20] , with the most recent study reporting encouraging improvements in the 6 minute walk test following continuous treatment with Drisapersen (2OMePS) for 48 weeks 21 . Despite this, both chemistries have demonstrated both limited and variable ability to restore dystrophin, with negligible activity in heart 22,23 .
A powerful approach to improve AO potency is the conjugation of cell penetrating peptides to uncharged AOs, such as PMO [24][25][26][27] . While early generation peptide-PMOs were weakly active in heart, they were shown to have an influence on heart function in mdx mice displaying a mild cardiac phenotype 27,28 . Recently we have developed a novel series of highly active peptides known as Pip's (PMO internalising peptides) for PMO delivery 29 . This series of peptides comprise a hydrophobic central core with 2 flanking arginine-rich domains. This combines the improved skeletal muscle delivery capacity observed with the arginine rich B-peptide, with a hydrophobic sequence that dramatically enhances efficacy to create a new innovative series of peptides 30 . Previous studies showed that the hydrophobic region is critical for the delivery to the heart tissue as other arginine-rich peptides with a shorter hydrophobic region failed to achieve this. Progressive evolution of this peptide series through structure-activity studies has identified novel peptides with dramatically enhanced cardiac dystrophin restoration [30][31][32] . Pip6-PMO is currently the most potent class of peptide-PMO, demonstrating high levels of dystrophin protein restoration following a single low intravenous dose, most notably in cardiac muscle.
Here we sought to determine the benefit on cardiac function and pathology offered by Pip6-PMO conjugates which restore high levels of dystrophin in heart. To do this we modelled the treatment of DMD cardiomyopathy by repeat treating dystrophic mdx mice in which a moderate cardiomyopathy was induced by forced exercise over a twelve week period. Using stringent statistical measures, our results demonstrate that restoring high levels of dystrophin in heart with Pip6-PMO was sufficient to prevent exercise-induced progression of cardiomyopathy and to improve multiple indices of cardiac pathology.

Results
Repeat doses of Pip6f-PMO restore high levels of dystrophin protein in mdx mouse hearts. The widely used dystrophic mdx mouse model has a natural mutation in exon 23 of the Dmd gene which generates a premature termination codon. AOs directed towards exon 23 induce exon skipping and allow restoration of a shorter, but functional dystrophin protein. The most recent screen of Pip6-PMO conjugates revealed 2 lead candidates, Pip6a-and Pip6f-PMO, which were capable of restoring high dystrophin protein levels following a single administration. Pip6f-PMO was chosen for the purpose of this study and a single lower dose of 10 mg/kg, delivered intravenously (IV), was performed to confirm effective dystrophin restoration in heart. Figure 1 displays dystrophin protein restoration in mdx heart muscle fibres as measured by immunohistochemical staining (Fig. 1A) and western blotting ( Fig. 1B and Supplementary Fig. 1). The single, low dose treatment resulted in 5% dystrophin protein restoration in heart relative to C57BL/10 control. Therefore a longer treatment regimen was initiated whereby 12 week old mdx mice received IV administrations of 10 mg/kg on 4 consecutive days followed by 5 doses every 2 weeks ( Supplementary  Fig. 2). Immunohistochemical staining of the treated mdx hearts revealed widespread and homogenously distributed dystrophin protein restoration (Fig. 1C). This is indicated by the higher magnification inserts showing right ventricle (RV) wall ( Fig. 1 C, i), outer left ventricle (LV) wall from apex ( Fig. 1 C, ii) to base ( Fig. 1 C iv and v) and inner myocardial dystrophin protein restoration ( Fig. 1 C, iii). Representative images of dystrophin restoration in treated hearts relative to the C57BL/10 and mdx controls are also shown (Fig. 1D). Quantification of dystrophin staining was determined relative to a laminin co-stain. The dystrophin intensity values of multiple regions of interest, relative to the corresponding intensity value of laminin, were normalised to C57BL/10. This generated a recovery score of 46% in the hearts of the treated cohort (Fig. 1E). In addition, RT-qPCR revealed high levels of dystrophin transcript (32% of Dmd transcripts lacking exon 23, see Supplementary Fig. 3 B; RT-PCR also showed high D23 exon skipping, see Fig. 1F), and western blotting quantification indicated 28% protein restoration in heart tissue ( Fig. 1G and Supplementary Fig. 3 C).
Dystrophin was also restored in the diaphragm and other accessory respiratory muscles, intercostal (IM) and sternomastoid (SM) as well as tibialis anterior (TA) of the lower limb. Again immunohistological staining demonstrated extensive dystrophin protein restoration ( Supplementary Fig. 3 A) with exceptionally high recovery scores between 74-104% ( Supplementary Fig. 4). The RT-qPCR ( Supplementary Fig. 3 B) and western blot quantification (Supplementary Fig. 3 C&6) exhibited similar results with very high dystro-phin transcript levels (77-86%) and between 58-101% restoration of dystrophin protein in the diaphragm, TA, IM and SM muscles. RT-PCR also showed complete skipping of exon 23 in these muscles ( Supplementary Fig. 5). This demonstrates that the repeated lowdose Pip6f-PMO administration protocol successfully restored high levels of dystrophin in skeletal and cardiac muscles.
Exercised mdx mice exhibit dilated cardiomyopathy. The cardiac phenotype in 6 month old mdx mice is mild, but may be aggravated by exercise 33,34 . A forced exercise regimen was undertaken in control wild type C57BL/10 mice (n 5 10) and mdx mice both untreated (n 5 9) and treated with Pip6f-PMO (n 5 10). From 12 weeks of age, mice were exercised on a treadmill 3 times every 2 weeks for 45 minutes ( Supplementary Fig. 2). 10 mdx and 10 C57BL/10 unexercised mice were used as controls to monitor the effects of exercise alone. Mdx mice failed to run consistently, as they stopped intermittently and required constant encouragement (Supplementary Video 1). Indeed, 3 untreated mdx mice were removed from the study as they exhibited severe aversions to running, thereby reducing the sample size from 12 to 9 mice, which demonstrates their poor running capacity. C57BL/10 mice were markedly more active than their mdx counterparts (Supplementary Video 2).
All groups of mice underwent cine-MRI at 24 weeks of age, 5 days after the last exercise event in the exercised groups. A range of cardiac parameters were determined and all normalised to body weight with the exception of ejection fraction (EF) ( Table 1). Data was analysed using stringent statistical tests (two-way ANOVA and one-way ANOVA with Tukey post-hoc and Games-Howell post-hoc). As previously described, most LV parameters were unchanged between the unexercised mdx and C57BL/10 cohorts at 24 weeks of age (with the exception of LV cardiac output; LV cardiac output (CO)) 28,35 . The exercise regimen was developed to moderately increase the workload on the heart without causing damage to cardiac tissue. The exercise had a significant effect on the untreated mdx hearts for a number of cardiac measurements (increased average mass, end diastolic volume and end systolic volume with consequential decrease in LV EF; interaction effect between exercise and mouse groups determined by twoway ANOVA -see Table 1). These changes suggest reduced contractility of the heart and the onset of cardiac hypertrophy. This is complemented by Fig. 2 A which illustrates thicker LV walls of the mdx exercised cohort in scans 4-6 at systole. The heart rate was also elevated, with resultant increase in CO, suggesting compensatory mechanisms to increase cardiac contractility (Table 1).
These results are complemented by the one-way ANOVA analysis whereby the mdx and C57BL/10 exercised cohorts were directly compared (mdx mice: LV ESV larger, with reduced LV EF, RV EDV and ESV larger with consequential decrease in EF; Table 1).

Pip6f-PMO administration prevents onset of cardiomyopathy.
While exercise induced a strong heart phenotype in the mdx mice, treatment with Pip6f-PMO stabilised cardiac function, as demonstrated by cine-MRI. Moreover, in contrast to the untreated exercised mdx cohort which required much encouragement, the Pip6f-PMO group ran very well (Supplementary Video 3). The most notable cine-MRI result was the significantly larger average mass of the heart for the exercised mdx cohort compared to the Pip6f-PMO treated group (P , 0.05; Fig. 2A and Table 1). All LV and most of the RV cardiac parameters of the Pip6f-PMO treated cohort were comparable to the C57BL/10 cohort (Table 1). In addition, the Pip6f-PMO treated cohort and the unexercised mdx cohort (Table 1)   immunohistochemical staining in heart following single 10 mg/kg Pip6f-PMO IV administration in mdx mice. (B) Dystrophin western blot of treated mdx hearts relative to 5% C57BL/10 following a single 10 mg/kg Pip6f-PMO administration. Approximately 5% dystrophin restoration observed. All samples run under the same experimental conditions and on the same SDS gel. Dotted line indicates where image was cropped. (C) Mice then underwent 4 daily injections followed by additional administrations every two weeks. Dystrophin immunohistochemical staining of treated heart with inserts indicating higher magnification of designated areas namely the right ventricle (RV) wall (C, i), outer left ventricle (LV) wall at apex (C, ii), inner myocardium (C, iii), and base (C iv and v) of heart. (D) Representative images of immunohistochemical staining of dystrophin protein in exercised C57BL/10, mdx and Pip6f-treated mdx mouse hearts. (E) Quantification of immunohistochemical staining in hearts (from D) following multiple administrations. Dystrophin expression is determined relative to laminin co-stain. The scatter plots show the normalised relative intensity values for each region of interest. Statistical significance was determined using ANOVA followed by Games-Howell post-hoc test to correct for variance heterogeneity (***5 P , 0.001, **5 P , 0.01, *5 P , 0.05). (F) Representative images of reverse-transcriptase (RT) PCR illustrating D23 splicing in heart Pip6f-PMO treated cohort. All samples run under the same experimental conditions and on the same agarose gel. RT-qPCR was also performed (see Supplementary  Fig. 3 B) which resulted in 32.3% Dmd transcripts lacking exon 23, following normalisation to exon 20-21 (SEM 3.1). (G) Representative images of western blots for Pip6f-PMO treated heart. For western blots, 10 mg of protein was loaded and dystrophin (dys) was quantified relative to vinculin loading control. All samples run under the same experimental conditions and on the same SDS gel. Dotted line indicates where image was cropped. For full agarose and SDS gels see Supplementary Fig. 1, 5&6. www.nature.com/scientificreports prevented hypertrophy and prevented exercise induced deterioration of cardiac function in mdx hearts.
Reduction of cardiac damage markers and heart pathology in Pip6f-PMO treated mice. It was anticipated that the exerciseinduced increase in cardiac workload combined with the absence of dystrophin in the mdx heart would result in an increase in tissue biomarkers of cardiac injury or damage. Two such markers, Nppa and Nox4, were measured 5 days following exercise in mdx mouse hearts. Atrial natriuretic peptide (Nppa) is secreted by the atria in response to haemodynamic overload or stress 36 . Its release results in the reduction of blood pressure and cardiac hypertrophy and is therefore considered a compensatory mechanism 36 but also serves as a diagnostic measure of heart failure 37 . Quantitative assessment (RT-qPCR) revealed elevated expression of this gene in the untreated exercised mdx group which was significantly higher than in the C57BL/10 cohort (P , 0.05; Fig. 3A). The expression of Nppa in the Pip6f-PMO treated cohort was partially normalised.
NADPH oxidase 4 (Nox4) has been associated with increases in oxidative stress 38 and has previously been described in mdx mouse studies to be associated with cardiac dysfunction 39 . Nox4 expression in the mdx exercised cohort was markedly elevated compared to C57BL/10 mice (P , 0.001; Fig. 3B). Nox4 levels in the Pip6f-treated cohort remained elevated, but to a lesser extent than in the mdx group (P , 0.01).
MicroRNA (miRNA) 21 is another biomarker of cardiac injury, specifically, it has been directly linked to myocardial fibrosis during hypertrophy 40 . Mdx mice showed an increase in miR-21 expression over wild type mice suggesting an increase in interstitial fibrosis (P , 0.001; Fig. 4A). Treatment with Pip6f-PMO partially normalised miR-21 expression (P , 0.05) over the untreated counterpart. The quantification of Masson's trichrome staining, which stains for collagen deposition, confirmed this observation of significantly reduced interstitial fibrosis in Pip6f-PMO treated mdx hearts (Fig. 4B). Again the untreated mdx hearts exhibited greater collagen deposition compared to C57BL/10 (P , 0.01) and Pip6f-PMO (P , 0.001) cohorts. Images of cardiac fibrosis strongly indicated the greater collagen deposition in untreated mdx hearts that was significantly reduced by treatment (Fig. 4D). Moreover, infiltration of Evans blue dye in the hearts of Pip6f-treated mice was also significantly reduced compared to untreated counterparts (P , 0.05; Fig. 4C). This showed that Pip6f-PMO treatment prevented sarcolemmal damage following exercise. These results denote a significant reduction in cardiac pathology in the treated mouse cohort.
Blood analyses of markers for toxicity were performed to determine the safety of the Pip6f-PMO treatment regimen. Plasma toxicity markers, including markers of hepatic and renal function, were unaffected by Pip6f-PMO treatment ( Supplementary Fig. 7).

Discussion
While AO exon skipping clinical trials appear promising for DMD 17,18,20,41,42 , the high incidence of cardiomyopathy in these patients, with associated high levels of mortality, indicate that restoring dystrophin protein in heart to modulate the cardiac phenotype is of high priority. Here we evaluated the therapeutic value of a highly potent peptide-PMO, Pip6f-PMO, which we previously demonstrated to be active in heart 30 . Using a forced exercise model we show that repeat treatment with Pip6f-PMO induced high levels of dystrophin restoration in heart which prevented cardiac hypertrophy and stabilised other cardiac parameters as determined by cine-MRI. Moreover, we also show that treatment ameliorates cardiac pathology as evidenced by reduced collagen and interstitial fibrosis and improved sarcolemmal integrity. These results suggest that high levels of dystrophin protein restoration in heart early in the course of DMD disease progression might have a significant clinical benefit.
A major limitation of the mdx mouse model of DMD is the mild cardiomyopathy which typically manifests only with advanced age 35 . Whilst previous peptide-PMO studies have shown dystrophin restoration in heart and some influence on cardiac function parameters, they either required dobutamine administration to stress the heart of young mdx mice 27 or observed improvements predominantly in the RV of 6 month old mice 28 . Dobutamine administration is commonly used to exacerbate cardiac dysfunction in mdx mice, however this acute insult does not model the effect of progressive macro-structural changes to the myocardium typical of a degenerative cardiomyopathy. In order to induce a cardiomyopathy which recapitulates the DMD patient phenotype, mice were subjected to a moderate exercise protocol over a 12 week period. This exercise regimen was based on the 'Treat NMD protocol and guidelines' for exercising mdx mice (http://www.treat-nmd.eu). As this protocol has only previously been utilised in assessing skeletal muscle pathology, the exercise regimen was modified. The established exercise regimen was designed so as to cause minimal damage to the heart while moderately increasing the cardiac workload in order to advance the stage of cardiomyopathy incrementally. This 12 week time-course was sufficient to yield significant functional changes in the untreated mdx heart (Fig. 2 and Table 1). Treadmill running was favoured over voluntary running, since wild-type mice are known to run further distances than mdx mice 33,43,44 .
Cine-MRI is considered the optimal technique for cardiac function assessment in small animals 45 . Cine-MRI demonstrated that moderate forced exercise was capable of inducing dilated cardiomyopathy in mdx mice, characterised by an enlarged heart and reduced contractility. This condition, dilated cardiomyopathy, is exhibited in all DMD patients by the age of 18 years 3,46 . However, treatment with Pip6f-PMO prevented these events. The low dose administration regimen of Pip6f-PMO induced high dystrophin protein restoration In addition Masson's trichrome staining is reduced and the representative images indicate less fibrosis then the untreated mdx cohort. Statistical significance was determined using ANOVA followed by Tukey post-hoc test (***5 P , 0.001, **5 P , 0.01, *5 P , 0.05). Nppa (A) and Nox4 (B) in heart tissue normalised to C57BL/10 cohort. The untreated mdx exercised cohort reveals elevated expression of these injury markers whereas there is partial normalisation for the Pip6f-PMO cohort. Statistical significance was determined using ANOVA followed by Tukey posthoc test (***5 P , 0.001, **5 P , 0.01, *5 P , 0.05).
www.nature.com/scientificreports SCIENTIFIC REPORTS | 5 : 8986 | DOI: 10.1038/srep08986 levels in skeletal muscle and approximately 28% dystrophin protein restoration in the heart. This level of dystrophin prevented hypertrophy, with left and right ventricular volumes in the normal range. In addition, LV CO and RV CO were stable. This study indicates that approximately 28% dystrophin restoration is sufficient to prevent the onset of exercised-induced dilated cardiomyopathy and maintain cardiac function.
In addition to MRI changes, the untreated exercised mdx cohort exhibited elevated markers of oxidative damage and injury, and fibrosis (collagen deposition) and sarcolemmal damage in the heart. This compares well with previous studies that show clear myocardial changes following exercise in mdx mice 33,34,47 . In contrast, the Pip6f-PMO treatment prevented injury in the mdx heart. It further reduced cardiac pathology as shown by significant reductions in cardiac fibrosis and sarcolemmal damage which demonstrated the protective capacity of a highly effective dystrophin restoration treatment. Cardiac fibrosis generally correlates with the decline in cardiac function 48,49 and therefore the marked reduction of myocardial fibrosis would likely have had a beneficial effect on cardiac function.
Our data suggests that restoring high levels of dystrophin protein in heart is likely to prevent the progression of cardiomyopathy in DMD patients. This supports the idea that DMD patients should be treated as early as possible. It has been reported that mild molecular abnormalities in mdx heart manifest as early as 10-12 weeks of age and therefore it is likely that our treatment was not initiated prior to the onset of disease 50,51 . It is possible that an even greater degree of cardio-protection would have been observed if Pip6f-PMO treatment had been initiated earlier, in younger mdx mice. In addition, this work demonstrates 'proof of principle' for the utilisation of Pip6conjugated AOs for the correction of cardiac structural protein defects. Three genes encoding cardiac sarcomeric proteins, MYH7, MYBPC3 and TNNT2 have been implicated in over 70% of all familial hypertrophic cardiomyopathies 52 and indeed exon skipping technology has already been demonstrated as a possible therapy for Mbpc3 mutations 53 . Thus Pip-PMO compounds could have wider application in the treatment of many familial cardiac defects.

Methods
Synthesis of Peptide-PMO Conjugates. Pip6f Ac-(RXRRBRRXRFQILYRXRBR-XRB)-COOH, where X is aminohexanoic acid and B is b-alanine) was synthesized by standard solid phase Fmoc chemistry and purified by HPLC. The PMO sequence (5'-GGCCAAACCTCGGCTTACCTGAAAT-3') was purchased from Gene Tools LLC. Pip6f was conjugated to PMO through an amide linkage at the 3' end of the PMO, followed by purification by HPLC as previously described 30 . The final product was analysed by MALDI-TOF MS and HPLC. Peptide-PMO conjugates were dissolved in sterile water and filtered through a 0.22 mm cellulose acetate membrane before use.
Animals and Intravenous Injections. Male C57BL/10 mice were attained from Harlan Laboratories (Oxford, UK) and mdx mice were bred in house at the Biomedical Sciences Unit (University of Oxford, UK). All procedures were authorized and approved by the University of Oxford ethics committee and UK Home Office (project licence 30/2907). Procedures were carried out in the Biomedical Sciences Unit, University of Oxford, in accordance with 'Laboratory Animal Handbooks NO.14; The Design of Animal Experiments (2010)'. For the treated cohort, 12 week old male mdx mice were injected via the tail vein with Pip6f-PMO prepared in 0.9% saline solution at a dose of 10 mg/kg. Mice were injected daily for 4 days followed by additional administrations every two weeks; 9 doses in total.
Exercise Regimen. This exercise regimen was modified from a Treat-NMD protocol (http://www.treat-nmd.eu/downloads/file/sops/dmd/MDX/DMD_M.2.1.001.pdf). The Exer 3/6 treadmill (Columbus Instruments, USA) was used for the exercised cohorts. Exercise commenced at 12 weeks of age and mice were exercised 3 times every 2 weeks for a 12 week period. Mice were allowed 2 minutes for familiarisation. The exercise regimen was started at 5 m/min and gradually increased in 1 m/min increments to 10 m/min for the first 2 exercise days. For the following 6 exercise sessions, mice were run for 45 minutes with the speed of the treadmill incrementally increased to 12 m/min. The mice were run at 12 m/min for 45 minutes for the remaining exercise sessions.
Cine-MRI. All mice underwent cardiac cine-MRI at 24 weeks of age, as previously described 35 . Mice were anaesthetised and placed in the supine position into a purpose built cradle. ECG electrodes were inserted into the forepaws of the mouse and the respiration loop was taped across the abdomen. Once the mouse was secured, with a stable ECG measurement, the cradle was lowered into a vertical-bore, 11.7 T MR system (Magnex Scientific, Oxon, UK) with a 40 mm birdcage coil (Rapid Biomedical, Wurzburg, Germany). Images were acquired using a Bruker console running Paravision 2.1.1 (Bruker Medical, Ettlingen, Germany). The entire left and right ventricles were imaged by taking a contiguous stack of cine images in 1 mm increments. Images were analysed using ImageJ software (NIH Image, Bethesda, MD). The epicardial and endocardial borders were outlined using the ImageJ freehand tool at end-diastole and end-systole.
Following MRI, mice were sacrificed by CO 2 inhalation, and muscles and other tissues harvested and snap-frozen in cooled isopentane before storage at 280uC.
Immunohistochemistry and Quantification. Transverse sections of tissue samples were sectioned (8 mm thick) and co-stained with rabbit-anti-dystrophin (Abcam) and rat anti-laminin (Sigma) as previously described 30 . Dystrophin was quantified relative to laminin co-stain using ImagePro software (MediaCybernetics) and normalised to C57CL/10. Briefly, 4 images of dystrophin and the corresponding laminin fields were taken for each section. Following the Arechavala-Gomeza approach 54 , 10 regions of interest were randomly placed on the laminin image which was overlaid on the corresponding dystrophin image to attain the minimum and maximum fluorescence intensity (Image-Pro, Media Cybernetics, Inc.). Data was normalised relative to C57BL/10. The percentage recovery score was calculated using the following equation: (dystrophin recovery of treated mdx mice-dystrophin recovery of untreated mdx mice)/(dystrophin recovery of C57BL/10 mice-dystrophin recovery of untreated mdx mice); as described on the TREAT-NMD website (http:// www.treat-nmd.eu/downloads/file/sops/dmd/MDX/DMD_M.1.1_001.pdf).
RT-PCR and RT-qPCR for Mouse Tissues. Total RNA was extracted using TRIzol reagent (Invitrogen) following manufacturer's instructions. For the RT-PCR reaction, 400 ng of RNA template was used in a 50 ml reverse transcription reaction using One Step RT-PCR Kit (QIAGEN) and gene specific primers as previously described 30 . Two microlitres of cDNA was amplified in a 50 ml nested PCR (QIAGEN PCR kit).
For RT-qPCR, 0.5-1 mg of RNA was reverse transcribed using a High Capacity cDNA Synthesis kit (Applied Biosystems). Gene specific primers sets ( miRNA qPCR was performed as previously described 55 . To quantify mature miRNA abundance, small RNA TaqMan Assays (Applied Biosystems) were used according to manufacturer's instructions. As described previously 56 , 5 ng of total RNA was reverse transcribed using the TaqMan miRNA Reverse Transcription Kit (Applied Biosystems). 1.33 ml of RT reaction were used in each qPCR and miR-16 was served as endogenous normalizer miRNA.
Protein Extraction and Western Blot. Samples were homogenised and quantified by Bradford assay (Sigma) as previously described 30 . Ten micrograms of untreated and treated mdx protein, and 5 mg and 2.5 mg of C57BL/10 protein (positive control) were loaded onto 3-8% Tris-Acetate gels (Invitrogen). Protein was blotted onto PVDF membrane and probed for dystrophin using DYS1 (Novocastra) and loading control, vinculin (Sigma). Primary antibodies were detected using IRDye 800 CW goat-anti mouse IgG (Licor). Western blots were imaged (LiCOR Biosciences) and analysed using the Odyssey imaging system. Western blots were quantified by calculating dystrophin protein expression relative to vinculin loading control.
Masson's Trichrome Staining. Longitudinal sections of heart samples were cut (8 mm thick) and stained with a Masson's trichrome kit following the manufacturer's instructions (Sigma). Whole sections were imaged and reassembled. Fibrosis was quantified as a percentage of surface area stained with Masson's trichrome measured using ImageJ.
Evans Blue Dye Administration. Between 3 and 4 mice from each cohort were injected intra-peritoneally with 1% Evans blue dye (10 ml/g of mouse; Sigma) and heart tissues were harvested 20 hours later. Hearts were sectioned (8 mm), placed in acetone for 10 minutes, washed with PBS and cover slipped. Evans blue infiltration was visualised using a Leitz DM RBE fluorescent microscope (Leica) and the entire section was imaged using Axiovision Rel 4.7 Software (Zeiss). The images were collated so that the entire section could be observed and the surface area of Evans blue staining was quantified using the threshold function of ImageJ software.
Clinical Biochemistry. Plasma samples were extracted from the jugular vein of mdx mice immediately after sacrifice by CO 2 inhalation. Analysis of toxicity biomarkers was performed by a clinical pathology lab, Mary Lyon Centre, MRC, Harwell, UK.
www.nature.com/scientificreports SCIENTIFIC REPORTS | 5 : 8986 | DOI: 10.1038/srep08986 Statistical Analysis. All values reported are mean 6 standard error of the mean (SEM). Regarding cine-MRI data, each cohort was directly compared using Students t-test. microRNA RT-qPCR of murine heart tissues. Statistical significance was calculated with IBM SPSS Statistics (Version 22.0) using one-way ANOVA followed by Tukey post-hoc test (TU), or Games-Howell post-hoc test to correct for variance heterogeneity. Variance heterogeneity was assessed by Levene's tests. Two-way ANOVA was also performed to determine the interaction effect between exercise and the mouse groups (ie. mdx and C57BL/10). Principal component analysis was conducted on a correlation matrix using the function prcomp() in R (http://www.r-project.org/).