Serum proteomic profiling reveals fragments of MYOM3 as potential biomarkers for monitoring the outcome of therapeutic interventions in muscular dystrophies

ORIGINAL ARTICLE Serum proteomic profiling reveals fragments of MYOM3 as potential biomarkers for monitoring the outcome of therapeutic interventions in muscular dystrophies Jérémy Rouillon1,†, Jérôme Poupiot1,†, Aleksandar Zocevic1,†, Fatima Amor1, Thibaut Léger2, Camille Garcia2, Jean-Michel Camadro2, Brenda Wong3, Robin Pinilla1, Jérémie Cosette1, Anna M.L. Coenen-Stass4, Graham Mcclorey4, Thomas C. Roberts4,5, Matthew J.A. Wood4, Laurent Servais6, Bjarne Udd7, Thomas Voit8,9, Isabelle Richard10 and Fedor Svinartchouk1,* 1 Généthon, Evry, France, 2 Mass spectrometry Laboratory, Institut Jacques Monod, UMR 7592, University Paris Diderot, CNRS, Sorbonne Paris Cité, F-75205 Paris, France, 3 Division of Pediatric Neurology, Cincinnati Children’s Hospital Medical Center, Cincinnati, OH, USA, 4 Department of Physiology, Anatomy and Genetics Oxford, Oxford, OX1 3QX, UK, 5 Department of Molecular and Experimental Medicine, The Scripps Research Institute, La Jolla, CA, USA, 6 Service of Clinical Trials and Databases, Institut de Myologie, Paris, France, 7 Folkhälsan Institute of Genetics and Department of Medical Genetics, Haartman Institute, University of Helsinki, Helsinki, Finland, 8 UPMC Inserm, UMRS 974, CNRS FRE 3617, Paris, France, 9 Université Pierre et Marie CurieParis 6, Institut de Myologie, GH Pitié-Salpêtrière, Paris, France and 10 Genethon, CNRS UMR 8587, Evry, France *To whom correspondence should be addressed at: Généthon, 1 bis rue de l′Internationale, 91002, Evry, France. Tel: +33 169472535; Fax: +33 169472838; Email: [email protected] Abstract Therapy-responsive biomarkers are an important and unmet need in the muscular dystrophy field where new treatments are currently in clinical trials. By using a comprehensive high-resolution mass spectrometry approach and western blot validation, we found that two fragments of the myofibrillar structural protein myomesin-3 (MYOM3) are abnormally present in sera of Duchenne muscular dystrophy (DMD) patients, limb-girdlemuscular dystrophy type 2D (LGMD2D) andtheir respective animal models. Levels of MYOM3 fragments were assayed in therapeutic model systems: (1) restoration of dystrophin expression by antisense oligonucleotide-mediated exon-skipping in mdx mice and (2) stable restoration of α-sarcoglycan expression in KO-SGCA mice by systemicinjectionofaviralvector.FollowingadministrationofthetherapeuticagentsMYOM3wasrestoredtowardwild-typelevels. IntheLGMDmodel,wheredifferentdosesofvectorwereused,MYOM3restorationwasdose-dependent.MYOM3fragmentsshowed lower inter-individual variability compared with the commonly used creatine kinase assay, and correlated better with the restoration of the dystrophin-associated protein complex and muscle force. These data suggest that the MYOM3 fragments hold promise for minimally invasive assessment of experimental therapies for DMD and other neuromuscular disorders. † J. Rouillon, J. Poupiot and A. Zocevic contributed equally to this work. Received: April 24, 2015. Revised and Accepted: June 4, 2015 © The Author 2015. Published by Oxford University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. Human Molecular Genetics, 2015, Vol. 24, No. 17 4916–4932 doi: 10.1093/hmg/ddv214 Advance Access Publication Date: 9 June 2015 Original Article 4916 at Tampere University Library. Department of Health Sciences on September 27, 2016http://hmg.oxfordjournals.org/Downloaded from Introduction The dystrophin-associated protein complex (DAPC) consists of several transmembrane and intracellular scaffolding elements implicated in maintaining the structure and morphology of vertebrate muscle fibres. Loss-of-function mutations in genes encoding these proteins give rise to different forms of muscular dystrophy. The absence of functional dystrophin or sarcoglycans in the DAPC is accompanied by a strong destabilization of the complex at the sarcolemma (1). As a consequence, muscle fibres become more sensitive to mechanical damage leading to muscle degeneration, chronic inflammation and an increase in fibrosis—hallmarks of the dystrophic phenotype (2). The most prevalent and severe disease is Duchenne muscular dystrophy (DMD), an X-linked disorder caused by mutations in the dystrophin gene, with a world-wide incidence of 1/5000 male newborns. DMD patients usually lose the ability to walk around the ageof12anddieintheirthirdorfourthdecadeduetocardiorespiratory complications (3). Deficiencies in the sarcoglycan genes are usually less severe but can also be accompanied by cardiac problems (4,5). Recently, substantial progress in the development of therapeutic approaches for the treatment of muscular dystrophies has been accomplished. Therapies for DMD based on the delivery of minidystrophin (6) or antisense oligonucleotide-mediated exon-skipping (7–10) are in pre-clinical evaluation or in phase I–III clinical trials. The small-molecule compound Ataluren (11,12) has recently obtained conditional marketing authorization for the treatment of DMD. Furthermore, a long-term, sustained restoration of α-sarcoglycan (Sgca) and γ-sarcoglycan (Sgcg) expression was observed following intramuscular gene transfer to muscles of patients with limb-girdle muscular dystrophy types 2D (LGMD2D) (13)and2C(14)respectively.With recent progress in pharmacoor gene-therapy for muscular dystrophies there is a growing need for minimally invasive biomarkers that can be used to assess and monitor the efficacy of therapy. Indeed, in order to evaluate the efficiency of a treatment during animal studies, researchers have unlimited access to different types of biopsies or necropsies. In contrast, trials in humans impose ethical restrictions requiring minimally invasive methods to assess and monitor the efficacy of therapy. Current methods include functional evaluation scales to measure patients’status (15–17), measurement of the level of fatty infiltration by magnetic resonance imaging (MRI) (18)andquantification of serum microRNAs (19–21) or urinary proteins (22). The biomarker most commonly used for DMD is serum creatine kinase (CK), which leaks into the blood stream upon muscle damage. However, CK demonstrates variations due to physical activity, muscle injury, cramping, toxic agents or age (23,24). Thus, although serum CK measurement is a useful diagnostic biomarker (25), it is not appropriate to predict the course of disease, severity of pathology or to monitor the efficacy of treatment. Variations in the composition of serum proteome are considered a promising source of biomarkers (26). In the present study, serum samples from DMD patients and healthy controls were compared using a comprehensive high-resolution mass spectrometry approach and several tens of proteins with altered levels were revealed by label-free protein quantification analysis. Among these proteins, the myofibrillar structural protein myomesin-3 (MYOM3), which was more abundant in DMD patient sera than in healthy controls, was chosen for detailed analysis. MYOM3 was present in sera as two internal fragments of 100 and 130 kDa rather than as an intact protein. Importantly, these fragments demonstrated lower inter-individual variations compared to CK. High levels of these MYOM3 fragments were also detected in sera from LGMD2D patients, as well as in animal models of DMD and several limb-girdle muscular dystrophies. In the dystrophin-deficient mdx mouse, these fragments were more reliable for the early detection of the disease and less sensitive to physical exercise when compared to CK. MYOM3 fragments were also superior when compared to CK for the monitoring the restoration of the DAPC and correlated to the rescue of physical force after gene therapy treatment of LGMD2D mouse model. Taken together, our data suggest that MYOM3 fragments are biomarkers for the detection, evaluation and treatment monitoring of DMD, LGMD2D and potentially for other forms of muscular dystrophy associated with increased turnover of sarcomeric proteins. Results Detection of serum proteins with altered levels in DMD patients by mass spectrometry Serum samples from 39 DMD patients and 38 control subjects collected in USA as part of the Advanced Diagnostics for New Therapeutic Approaches (ADNA) project (http://www.institut-merieux. com/projetssante_adna.php) were analysed using a mass spectrometry approach. To reduce the number of LC-MS/MS analyses, the samples were organized into four groups (G1: young DMDfrom3to10yearsold;G2:olderDMDfrom12to20years old; G3: young controls from 3 to 10 years old and G4: older controls from 12 to 20 years old) subdivided in a total of 12 pools according to the patient’sage(Table1). Each pool included sera from at least four individuals where serum of each individual was equally represented. In order to ensure deep proteome coverage, the pools were immunodepleted for the 12 major serum proteins. Mass spectrometry analysis of serum samples of all 12 subgroups enabled the identification a total of 3329 unique peptides matching 378 proteins (with a false discovery rate less than 0.01). Among those, 69% of protein identification calls (260 proteins) were based on spectra from two or more peptides. To reveal Table 1. Schema of samples assembling into groups and subgroups. DMD G1-1 to G1-4: serum from young DMD patients of 3 to 10 years old; DMD G2-1 to G2-2: serum from DMD patients of 12 to 20 years old. Healthy controls G3-1 to G3-4 and G4-1 to G4-2: age-matched healthy controls to the young and older DMD patients respectively. Numbers below each pool indicate the interval of age and the number of patients (in brackets) DMD G1–1G1–2G1–3G1–4G2–1G2–2 Age (number) 3–4 (7) 4–6 (11) 6–7 (4) 7–10 (4) 12–16 (6) 16–20 (7) Healthy controls G3–1G3–2G3–3G3–4G4–1G4–2 Age (number) 3–4 (5) 4–6 (6) 6–7 (5) 7–10 (5) 12–16 (10) 16–20 (7) Human Molecular Genetics, 2015, Vol. 24, No. 17 |4917 at Tampere University Library. Department of Health Sciences on September 27, 2016http://hmg.oxfordjournals.org/Downloaded from proteins differentially present in sera from DMD and healthy individuals, the data were analysed by a label-free quantification approach using the following parameters: number of peptides ≥2; Mascot protein score (http://www.matrixscience.com/) ≥50 and fold change ≥2. The analysis of young DMD patients with their age-matched controls (G1 versus G3 groups) revealed 24 proteins more abundant in DMD and 13 in healthy subjects (Table 2). The top 10 proteins with the lowest P-value were more abundant in DMD patients and either involved in muscle energy metabolism (pyruvate kinase PKM, L-lactate dehydrogenase B chain, CK-M, alanine aminotransferase 1, β-enolase, carbonic anhydrase 3, fructose-bisphosphate aldolase A), in sarcomere organization (myomesin-3, myosin-7) or costamere organization (vinculin). Comparison of older DMD patients with their age-matched controls (G2 versus G4) using the same parameters resulted in only nine altered proteins: five proteins more abundant in DMD (CK-M, adiponectin, fructose-bisphosphate aldolase A, L-lactate dehydrogenase B chain and haemoglobin β) and four in healthy subjects (gelsolin, phosphatidylcholine-sterol acyltransferase, cadherin-13 and cartilage acidic protein 1) (Table 3). Only four of these proteins (CK-M, fructose-bisphosphate aldolase A, L-lactate dehydrogenase B chain and haemoglobin β) were differentially abundant in both DMD age groups according to the mass spectrometry analysis. Importantly, the expression ratios for these four proteins in DMD versus healthy controls were substantially lower in older DMD patients as compared to the young DMD group (19.5; 3.3; 2.2 and 2.4 folds in older DMD versus 39.8; 14.2; 5.4 and 3.6 times in young, respectively). The decrease in the number of proteins with altered levels and in magnitude of their fold changes is most probably due to the drastic decrease of muscle mass in older DMD patients and relative immobility of these patients. Interestingly, label-free analysis of young and older DMD patients (G1 versus G2) revealed eight secreted proteins that increased in abundance with patient age (dopamine β-hydroxylase: 3-fold, adiponectin: 3-fold, serum amyloid P-component: 3-fold, insulin-like growth factor-binding protein complex acid labile Table 2. List of proteins with altered levels between G1 and G3 groups (young DMD and age-matched healthy controls) classified by the decrease in the ratio DMD/healthy (fold change) No. accession Description Localization Peptides Score ANOVA (P-value) Fold change MYG_HUMAN Myoglobin Cytoplasm 4 195 2.7e-03 234.8 MYOM2_HUMAN MYOM2 Myofibril 10 390 9.8e-05 100.1 MYOM3_HUMAN MYOM3 Myofibril 11 491 1.5e-05 49.7 TPIS_HUMAN Triosephosphate isomerase Cytoplasm 3 128 2.3e-03 48.4 AATC_HUMAN Aspartate aminotransferase Cytoplasm 3 75 4.7e-04 45.7 KCRM_HUMAN CK-M Cytoplasm 15 849 2.9e-05 39.8 MYH7_HUMAN Myosin-7 Myofibril 11 520 2.2e-05 38.3 ENOB_HUMAN β-enolase Cytoplasm 4 178 7.4e-05 34.8 G6PI_HUMAN Glucose-6-phosphate isomerase Cytoplasm/Secreted 4 130 1.6e-03 29.5 CAH3_HUMAN Carbonic anhydrase 3 Cytoplasm 5 182 8.6e-05 23.9 FLNC_HUMAN Filamin-C Myofibril 4 145 4.3e-04 19.4 ALAT1_HUMAN Alanine aminotransferase 1 Cytoplasm 4 127 3.0e-05 15.6 ALDOA_HUMAN Fructose-bisphosphate aldolase A Cytoplasm 15 729 9.3e-05 14.2 KPYM_HUMAN Pyruvate kinase PKM Cytoplasm 16 845 1.1e-05 12.8 TITIN_HUMAN Titin Myofibril 14 495 1.9e-03 10.8 VINC_HUMAN Vinculin Cytoplasm/Membrane 2 74 7.2e-05 10.3 PYGM_HUMAN Glycogen phosphorylase, muscle form Cytoplasm 8 257 6.1e-04 9.9 LDHA_HUMAN L-lactate dehydrogenase A chain Cytoplasm 8 378 9.1e-04 9.5 HPT_HUMAN Haptoglobin Secreted 29 1867 1.5e-04 7.6 HBD_HUMAN Haemoglobin subunit δCytoplasm 3 100 5.1e-03 6.2 LDHB_HUMAN L-lactate dehydrogenase B Cytoplasm 10 598 2.4e-05 5.4 HBB_HUMAN Haemoglobin subunit βCytoplasm 7 552 8.0e-03 3.6 HBA_HUMAN Haemoglobin subunit αCytoplasm 7 407 5.3e-03 3.4 TPM2_HUMAN Tropomyosin βchain Myofibril 5 170 2.0e-02 2.6 VASN_HUMAN Vasorin Membrane 4 135 4.0e-02 0.5 ALS_HUMAN Insulin-like growth factor-binding protein complex Secreted 22 1096 1.0e-02 0.5 PHLD_HUMAN Phosphatidylinositol-glycan-specific phospholipase D Secreted 9 533 4.7e-03 0.5 CHL1_HUMAN Neural cell adhesion molecule L1-like protein Membrane/Secreted 2 66 3.0e-02 0.5 COL11_HUMAN Collectin-11 Secreted 2 72 2.6e-03 0.4 CADH5_HUMAN Cadherin-5 Membrane 6 220 2.0e-03 0.4 CD109_HUMAN CD109 antigen Membrane 2 59 3.0e-02 0.4 LBP_HUMAN Lipopolysaccharide-binding protein Secreted 7 386 5.0e-03 0.4 CRAC1_HUMAN Cartilage acidic protein 1 Secreted 6 223 2.0e-02 0.4 C4BPB_HUMAN C4b-binding protein Secreted 4 207 2.0e-02 0.4 CNDP1_HUMAN β-Ala-His dipeptidase Secreted 8 294 4.2e-03 0.3 DPP4_HUMAN Dipeptidyl peptidase 4 Membrane/Secreted 5 162 5.5e-03 0.3 CETP_HUMAN Cholesteryl ester transfer protein Secreted 7 296 8.2e-04 0.2 Top 10 proteins with the lowest P-value are in bold. All shown proteins passed thresholds of peptide numbers ≥2, a score≥50, a fold change ≥2 and a P-value ≤0.05. Peptides: number of peptides identified for a given protein. Score: Mascot protein score. 4918 |Human Molecular Genetics, 2015, Vol. 24, No. 17 at Tampere University Library. Department of Health Sciences on September 27, 2016http://hmg.oxfordjournals.org/Downloaded from subunit: 3-fold, β-Ala-His dipeptidase: 5-fold, insulin-like growth factor I: 5-fold). Levels of MYOM3 demonstrate less inter-individual variations compared to CK in DMD patients Elevated levels of cytosolic proteins such as CK in the blood are now widely used as the first stage of DMD diagnosis (24,27). Therefore, it was appealing to compare serum levels of CK with one of the myofibrillar structural proteins found in the present study. Serum biomarkers are often presented by truncated fragment of the full-length proteins, thus making difficult finding of commercial antibodies for the specific fragment (28). In our hands, of the eleven antibodies tested against the three myofibrillar structural proteins (MYOM2, MYOM3 and Myosin 7) with the highest high fold change and lowest P-value between DMD and healthy controls (Table 2)onlyantibodyagainstMYOM3 was efficientinimmunoblotanalysis.MYOM3(UniProtKB# Q5VTT5), a protein of 1437 amino acids (162.2 kDa), is a member of a family of closely related structural proteins detected at the M-band of the sarcomere in striated skeletal muscles: MYOM1, MYOM2 (or M protein) and MYOM3. These proteins are involved in sarcomere stability and resistance during intense or sustained stretching (29). Western blot analysis of serum from DMD patients with an anti-MYOM3 antibody targeting amino acids 887–1178 of the protein revealed the presence of two bands of 100 and 130 kDa respectively (Fig. 1). To identify the position of these fragments within the protein, they were purified by immunoprecipitation, separated by SDS-PAGE and analysed by mass spectrometry separately (Table 4). The obtained peptide coverages of the fragments (total 55 peptides covering amino acid 254–1331 for the upper fragment and 38 peptides covering the sequence from amino acid 476–1331 for the lower fragment) suggest that both fragments have similar C-terminal end but different N-terminus. Minimum molecular weight of the fragments based on the positions of the most N-terminal and C-terminal identified peptides is equal to 121 kDa for the upper fragment (1077 aa) and 96 kDa (855 aa) for the lower fragment, which fit well with the size of the fragments estimated by SDS-PAGE (130 and 100 kDa respectively). Importantly, fragments of the same size were barely detectable in sera from healthy subjects by Western blot analysis (Fig. 1), thus validating the mass spectrometry data. We next compared the levels of the MYOM3 fragments and CK in all 103 subjects from the US cohort. The serum expression levels of both MYOM3 fragments were determined by Western blot analysis and CK assessed by measuring its enzymatic activity (Fig. 2). In accordance with the mass spectrometry data, results showed that expression levels of both, CK and the MYOM3 fragments, were much higher in young DMD patients compared to the respective healthy controls (ratio DMD/Control: MYOM3 = 284 and CK = 193) (Fig. 2A and B). In older DMD patients, the expression levels of CK and MYOM3 fragments were respectively 14 Table 3. List of proteins with altered levels in serum samples between G2 and G4 groups (older DMD and age-matched healthy controls) classified by the decrease in the ratio DMD/healthy (fold change) No. accession Description Localization peptides Score ANOVA (P-value) Fold change KCRM_HUMAN CK-M Cytoplasm 3 96 1.0e-02 19.5 ADIPO_HUMAN Adiponectin Secreted 3 213 3.0e-02 4.4 ALDOA_HUMAN Fructose-bisphosphate aldolase A Cytoplasm 2 84 3.0e-02 3.3 HBB_HUMAN Haemoglobin subunit βCytoplasm 9 632 8.4e-03 2.4 LDHB_HUMAN L-lactate dehydrogenase B chain Cytoplasm 6 223 4.0e-02 2.2 GELS_HUMAN Gelsolin Cytoplasm 32 2287 1.0e-02 0.5 LCAT_HUMAN Phosphatidylcholine-sterol acyltransferase Secreted 6 295 1.0e-02 0.4 CAD13_HUMAN Cadherin-13 Membrane 2 86 5.8e-03 0.4 CRAC1_HUMAN Cartilage acidic protein 1 Secreted 3 105 1.0e-02 0.2 All shown proteins passed thresholds of peptide numbers ≥2, a score ≥50, a fold change ≥2andaP-value ≤0.05. Peptides: number of peptides identified for a given protein. Score: Mascot protein score. Figure 1. Western blot analysis of MYOM3 in pools of sera from subgroups of young (G1) and older (G2) DMD patients as well as young (G3) and older (G4) healthy subjects. (A) normal exposure; (B) boosted exposure. For explanation of groups and subgroups see Table 1. Fifty micrograms of serum proteins were loaded in each well. Human Molecular Genetics, 2015, Vol. 24, No. 17 |4919 at Tampere University Library. Department of Health Sciences on September 27, 2016http://hmg.oxfordjournals.org/Downloaded from Table 4. List of MYOM3 peptides identified by mass spectrometry in upper (∼130 kDa) and lower (∼100 kDa) bands from DMD patient serum # Position Sequence Ion score (Upper band) Ion score (Lower band) 1 254–263 DAGFDSEIFK 24 2 265–283 STFGPSVEFTSVLKPVFAR 44 3 320–326 KILYTDR 19 4 321–326 ILYTDR 12 5 338–346 EDEGLYMVR 57 6 347–354 VPSPFGPR 44 7 355–363 EQSTYVLVR 48 8 364–380 DAEAENPGAPGSPLNVR 54 9 401–411 GNPITAYTIER 40 10 461–474 ASELVVMGDHDAAR 54 11 476–486 KTEIPFDLGNK 64 53 12 477–486 TEIPFDLGNK 19 43 13 487–513 ITISTDAFEDTVTIPSPPTNVHASEIR 92 50 14 514–527 EAYVVLAWEEPSPR 65 55 15 530–538 APLTYSLEK 66 37 16 539–556 SVIGSGTWEAISSESPVR 86 57 17 560–567 FAVLDLEK 40 37 18 560–568 FAVLDLEKK 33 35 19 569–574 KSYVFR 32 20 577–596 AMNQYGLSDPSEPSEPIALR 69 58 21 597–612 GPPATLPPPAQVQAFR 79 42 22 613–638 DTQTSVSLTWDPVKDPELLGYYIYSR 33 18 23 639–657 KVGTSEWQTVNNKPIQGTR 27 23 24 640–657 VGTSEWQTVNNKPIQGTR 67 79 25 658–664 FTVPGLR 25 21 26 675–692 SVSEAGVGESSAATEPIR 96 63 27 714–724 NEMVIGWKPPK 38 26 28 731–757 ILGYFLDQHDSEELDWHAVNQQPIPTR 12 29 761–773 VSDLHEGHFYEFR 29 30 796–809 EWTMPQPGPPYDVR 43 26 31 863–870 VSDLQPGK 42 15 32 935–944 DYKGPLDPQR 55 23 33 956–993 VILKEPGLEDLGTYSVIVTDADEDISASHTLTEEELEK 25 34 960–993 EPGLEDLGTYSVIVTDADEDISASHTLTEEELEK 17 35 1008–1019 LISGWNIDILER 64 45 36 1024–1030 LWLEVEK 14 20 37 1031–1044 LSPAAELHLIFNNK 52 55 38 1045–1052 EIFSSPNR 25 20 39 1045–1053 EIFSSPNRK 26 28 40 1053–1058 KINFDR 26 22 41 1061–1075 GLVEVIIQNLSEEDK 52 42 1061–1086 GLVEVIIQNLSEEDKGSYTAQLQDGK 33 21 43 1089–1102 NQITLTLVDDDFDK 72 96 44 1089–1105 NQITLTLVDDDFDKLLR 60 23 45 1118–1129 QGPYFERPLQWK 16 14 46 1151–1157 FQWFFQR 22 23 47 1189–1218 AMVSDDRGEDDTILDLTGDALDAIFTELGR 88 32 48 1219–1228 IGALSATPLK 62 57 49 1229–1237 IQGTEEGIR 46 56 50 1244–1251 YYNVEYMK 40 30 51 1252–1257 TTWFHK 16 52 1269–1284 TGTTLDEIWLHILDPK 19 21 53 1291–1299 YTLEIAAGK 37 35 54 1303–1322 QLSTDLSGQAFEDAMAEHQR 40 38 55 1325–1331 TLAIIEK 17 24 Ion scores (Mascot MS/MS ion scores) are shown for each peptide identified in each band. The entire length of MYOM3 is 1437 aa. The obtained peptide coverage of the MYOM3 suggests that both fragments have similar C-terminal end but different N-terminus. Minimum molecular weights of the fragments based on the positions of the most N-terminal and C-terminal identified peptides are equal to 121 kDa for the upper fragment (1077 aa) and 96 kDa (855 aa) for the lower. Thus estimated MW of the fragments fit well with the positions of the fragments on the SDS-PAGE (130 and 100 kDa respectively). 4920 |Human Molecular Genetics, 2015, Vol. 24, No. 17 at Tampere University Library. Department of Health Sciences on September 27, 2016http://hmg.oxfordjournals.org/Downloaded from and 5 times lower than in young DMD patients. Of note, in two outlier patients of 16 and 20 years old the levels of the MYOM3 fragments and CK were lower than the maximal values in the respective healthy controls. The decrease of these proteins with patient’s age can be explained by the severe loss of total muscle mass due to the advanced stage of the disease. Importantly, even if both proteins were able to discriminate DMD patients and healthy controls, there were less inter-individual variations in MYOM3 fragment levels compared to CK levels. While the CK levels in the young patients varied from 9000 to 60 000 IU/L (mean 27 130 IU/L ± 13 130), the values for MYOM3 fragments remained between 11 a.u. and 24 a.u. (mean 19 a.u. ± 3). The low correlation observed between the levels of serum CK and MYOM3 fragments in the group of young patients (R 2 = 0.28) indicates that different physiological mechanisms may account for the secretion/stability of these proteins at this age (Fig. 2C). Conversely, these two biomarkers were well correlated in older patients (Fig. 2D). MYOM3 fragments are specifically present in sera from animal models of DMD The levels of MYOM3 fragments were quantified in two animal models of DMD: golden Retriever muscular dystrophy (GRMD) which has a severe phenotype similar to DMD patients (30)and dystrophin-deficient mdx mice. Western blot analysis of GRMD and mdx sera revealed the presence of two bands migrating at the same positions as human MYOM3 fragments (Fig. 3A and B). Importantly, the abundance of these fragments was 100 times higher than in the healthy control dogs. Whereas the level of the MYOM3 fragments in human DMD samples decreased with age, expression of these fragments was very similar in the serum of 2 and 18 months old ambulant GRMD. We hypothesize that the high MYOM3 levels in elder GRMD can be due to the ambulant state of the dogs. Age-independent expression of the MYOM3 fragments could be an advantage for utilization of this biomarker in gene therapy studies conducted in dogs. MYOM3 fragments are elevated in sera of LGMD2D patients and mouse models of LGMDs The presence of the MYOM3 fragments was also analysed in serum samples of three patients with α-sarcoglycanopathy (LGMD2D). Fragments of the same length (100 and 130 kDa) were detected at elevated levels in all these patients. Overall, the level of these fragments in LGMD2D patients was lower compared to their intensity in young DMD patients (Fig. 4, upper panel). The following mouse models of limb-girdle muscular dystrophies were included in this study: KO-Calpain 3 (models for LGMD2A) (31), KO-Dysferlin (models for LGMD2B) (32), KO-Sgcg (models for LGMD2C) (33) and KO-Sgca (models for LGMD2D) (34). These mouse models are congenic strains on the genetic background of the C57BL/6J mouse, which was included in the study as their wild-type (WT) control. Taking into consideration the muscle impairment and time of disease onset, these mouse models can be classified in terms of decreasing order of severity: Figure 2. Expression levels of serum MYOM3 fragments (A) and CK (B) in sera from the entire US cohort including 39 young and 17 older DMD patients as well as 29 young and 18 older healthy controls. To measure levels of the MYOM3 fragments, 50 μg of serum proteins were analysed by Western blot, then band intensities were quantified and expressed in arbitrary units (a.u). The CK enzyme activity in serum is expressed in international units per litre (IU/L). (C) Linear regression analysis between serum levels of the MYOM3 fragments and CK for young patients. (D) Linear regression analysis between serum levels of the MYOM3 fragments and CK for older DMD patients. The linearity of the response by Western blot for MYOM3 is demonstrated in Supplementary material, Figure S2. Human Molecular Genetics, 2015, Vol. 24, No. 17 |4921 at Tampere University Library. Department of Health Sciences on September 27, 2016http://hmg.oxfordjournals.org/Downloaded from KO-Sgca, KO-Sgcg, KO-dysf and KO-Capn3. Serum from these mouse models was collected at 1 and 6 months of age, corresponding to the early and advanced stages of the dystrophies, and the levels of the MYOM3 fragments were compared by Western blot. The highest levels of serum MYOM3 fragments were observed in the three mouse models with perturbations in the DAPC (Fig. 4, lower panel). In KO-Dysf mice, these fragments were barely detectable at 1 month of age and then increased at 6months,reflecting the aggravation of the disease at this age. MYOM3 fragments were hardly detectable in KO-Capn3 mice at any age. In mdx mice, the MYOM3 fragments are expressed early, demonstrate less inter-individual variability and are less sensitive to physical exercise compared to CK In order to identify the earliest time point when the serum MYOM3 fragments are detectable, we investigated sera from mdx mice of different ages (from birth to 1-year-old). The MYOM3 fragments were detected in mdx mice at birth, with a small decrease in their levels at 1 week of age and followed by a rise in abundance at 3 weeks (Fig. 5A). Importantly, the levels of these fragments in the age-matched control mice was lower at all ages tested (Fig. 5B). The kinetics of the MYOM3 fragment abundance in the serum of mdx mice correlates with the timing of an acute phase of muscle necrosis generally occurring at 3–4 weeks of age, followed by an apparent stabilization of the muscle phenotype (35). The kinetics of serum CK levels in mdx mice were different from that of the MYOM3 fragments during the first weeks of age. Consistent with previous studies (36,37) serum CK was elevated in newborn mice, but then became undetectable during the 1st and 2nd week of age (except for 1 mouse), rising again at 3 and 12 weeks followed by a stabilization (Fig. 5C). In healthy mice, serum CK was also slightly elevated in newborns and 12and 24-week-old animals (Fig. 5D). Importantly, less variation was observed in the levels of MYOM3 fragments in mice of the same age compared to the CK (maximum fold change two for MYOM3 versus 110 for CK). The difference in age-dependent expression patterns between serum CK and MYOM3 fragments in mdx mice is probably related to different mechanisms of bioprocessing of these proteins, especially during the early phases of disease. To assess the impact of physical exercise on the serum levels of the MYOM3 fragments and CK, WT and mdx mice were subjected to downhill running for 30 min. This exercise regimen is often used to increase muscle injury and worsen the mdx phenotype (38,39). Sera were collected 7 days before and 3, 24 and 48 h after exercise. Importantly, while in mdx mice CK concentration peaked at 3 h post-exercise (up to 10-fold increase) followed by a substantial decrease (Fig. 6C), physical exercise had relatively little impact on the serum levels of the MYOM3 fragments (less than 2-fold increase 48 h post-exercise) (Fig. 6A). Interestingly, in healthy mice, there was a slight increase in the levels of the MYOM3 fragments 24 and 48 h after exercise, even though the maximum level of the fragments in healthy mice was 50-fold less than in mdx mice (Fig. 6B). Serum CK levels were variable in Figure 3. MYOM3 fragments are specifically present in sera from animal models of DMD. (A) Western Blot analysis of serum from GRMD and healthy dogs. GRMD # 1–4: two months old; # 5–6: 18 months old. Healthy # 1–4: two months old; # 5–6: 18 months old dogs. DMD: control serum from DMD patient. (B) Western Blot analysis of serum from 6 months old mdx and WT mice. WT: C57/BL10 strain. Figure 4. Upper panel: Western blot analysis of the MYOM3 fragments in serum from 3 LGMD2D patients (#1 is 35, #2 is 23 and #3 is 24 years old). Serum from two DMD patients (group G1) and three healthy individuals (group G4) were used as controls. Lower panel: Western blot analysis of the MYOM3 fragments in serum from mouse models of different muscular dystrophies at 1 and 6 months of age. WT: C57BL/6J mouse; mdx (model for DMD); KO-Sgcg: model for LGMD2C; KO-Sgca: model for LGMD2D; KO-Dysf: model for LGMD2B; KO-Capn3: model for LGMD2A. 4922 |Human Molecular Genetics, 2015, Vol. 24, No. 17 at Tampere University Library. Department of Health Sciences on September 27, 2016http://hmg.oxfordjournals.org/Downloaded from healthy mice without noticeable correlation with physical exercise (Fig. 6D). Given that MYOM3 is predominantly expressed in slow and intermediate speed (type I and IIa) skeletal fibres (29) which are less affected in DMD relative to fast myofibres (40), it is possible that the difference in the kinetics of these biomarkers is partially due to the differential sensitivity of these muscle fibre types to exercise-induced damage. MYOM3 fragments enable monitoring of pharmacoand gene-therapy treatment efficacy The presence of the MYOM3 fragments in serum of DMD and LGMD2D patients and their respective mouse models prompted us to evaluate the utility of these biomarkers for monitoring the response to experimental therapies in mdx and KO-Sgca mice. Restoration of dystrophin expression in mdx mouse muscles was achieved by a single administration of an arginine-rich cellpenetrating peptide (CPP) conjugated to a phosphorodiamidate morpholino oligonucleotide (PMO) that efficiently induces skipping of exon 23 and restores dystrophin protein expression and muscle function (41,42). In order to evaluate the impact of the restoration of dystrophin expression on the serum levels of MYOM3 fragments and CK, quadriceps femoris muscles and blood samples from treated mdx were collected 2, 4 and 8 weeks post-injection. Quadriceps and blood samples from nontreated mdx andWTcontrol12-week-oldmiceweretakenas controls. In accordance with the previously published data (41–43), the restoration of the dystrophin expression and percentage of exon skipping in quadriceps were between 10 and 45% 2 weeks after injections, followed by a decrease at later time points (Fig. 7A and B). In a good agreement with the restoration of dystrophin levels, two weeks after injection the levels of the MYOM3 fragments in treated mdx mice substantially decreased (without reaching the level in the control mice) and then gradually increased over time (Fig. 7C). In contrast to the MYOM3 fragments, CK levels did not reflect restoration of dystrophin expression. Thus, 2 weeks after injection the level of serum CK was lower in treated mdx mice compared to WT control mice (Fig. 7D), while dystrophin expression did not exceed 50% at that time. Moreover, 8 weeks after the treatment, when the estimated level of dystrophin-positive fibres was around 10%, CK levels were higher in treated than in non-treated mdx mice. Different behaviour of the MYOM3 fragments and CK after partial restoration of dystrophin expression may reflect the capacity of these biomarkers to differentially reveal intracellular process such as microparticle turnover (44) or increased myofibrillar protein catabolism (45,46). To restore α-sarcoglycan expression in KO-Sgca mice, we used recombinant adeno-associated virus rAAV2/8 vector. Control C57BL/6J mice received an intravenous injection of PBS and four groups of KO-Sgca mice received intravenous injections of either PBS or low (1e11 vg), medium (5e11 vg) or high (1e12 vg) Figure 5. Levels of the MYOM3 fragments (A,B) and CK-M (C,D) in serum from healthy (B, D) and mdx (A, C) mice at different ages as estimated by Western blot analysis. Intensity of the bands (in arbitrary units, a.u.) on different gels was normalized by the respective bands of the positive control (50 μg of serum proteins from the same mdx mouse present on each gel). Fifty micrograms of serum proteins were used for the analysis. Age 0 corresponds to newborn mice. Estimation of the CK-M level by Western blot analysis correlated well with the CK activity (Supplementary material, Fig. S3). Human Molecular Genetics, 2015, Vol. 24, No. 17 |4923 at Tampere University Library. Department of Health Sciences on September 27, 2016http://hmg.oxfordjournals.org/Downloaded from doses of rAAV2/8 coding for hSGCA. Mice were monitored for 3 months after the treatment. The following assays were compared in order to define the most appropriate for the follow-up of the treatment: histological analysis of muscle biopsies (HPS staining and restoration of the sarcoglycan complex); total physical force 3 months after the treatment (1 week before animal sacrifice); biweekly analysis of serum CK and MYOM3 fragments levels. Histological analysis of the gastrocnemius muscles demonstrated restoration of the complex in 5–30% (mean 15.6 ± 8.4), 60– 100% (mean 79.2 ± 16.7) and 84–100% (mean 94.6 ± 8.8) of fibres after low, medium and high rAAV dose treatments, respectively (Fig. 8A and B). Importantly, by assessing the expression level of α-sarcoglycan (determined by immunostaining) the KO-Sgca, low, medium and WT mice could be clearly distinguished. However, no statistically significant difference was found between medium and high rAAV doses by this method. Importantly, this analysis is highly laborious, and the size of the biopsies makes itunsuitablefor the follow-up of the therapeutic effect in small animals. Similar to histological analysis, the conventional whole body tension (WBT) method is an end-point assay because mice become accustomed to the protocol (47). The WBT method was only able to discriminate two clusters of animals: (1) KO-Sgca mice injected with PBS or low dose of rAAV and (2) control C57BL/6J mice and KO-Sgca mice injected with medium or high doses of rAAV (Fig. 8C). A threshold 3000 IU/L of CK clearly separates KO-Sgca mice injected with PBS from all other experimental groups (Fig. 8D). Nevertheless, when applying the Student’stest(P-value threshold < 0.01), differences only between few time points/ injection doses appeared as statistically significant (Fig. 8F). Changing of the P-value threshold to <0.05 permits to distinguish more experimental groups of mice (Fig. 8F). Lower CK levels in all groups of mice at day 90 (one week after the total force measurements) (Fig. 8D) could be explained by the fact that an increase of CK levels after physical exertion is followed by a substantial decrease persisting for 2 weeks (48). Inter-individual variations of the MYOM3 fragment levels were lower compared to serum CK in the case of all experimentalgroups (Fig. 8D and E). In accordance with a previous study showing progressive development of muscular dystrophy in KO-Sgca mice (34) the levels of the MYOM3 fragments in the control mice injected with PBS increased gradually with age (Fig. 8E). Even the lowest dose of rAAV (1e11 vg) stabilized the MYOM3 fragment levels, while medium and high doses reduced MYOM3 fragment levels 5-fold and 8-fold, respectively. Due to the low inter-individual variability, measurement of the MYOM3 fragments enabled nearly all groups of mice to be distinguished with either of the thresholds (P< 0.01 or 0.05) at the majority of time points (Fig. 8E and F). Furthermore, MYOM3 fragment abundance was better correlated (R 2 = 0.71) with muscle force as measured by the escape test compared with CK (R 2 = 0.59) (Supplementary material, Fig. S1). Figure 6. Levels of the MYOM3 fragments (A,B) and CK-M (C,D) in serum from healthy (B, D) and mdx (A, C) mice at different time after physical exercise estimated by Western blot analysis. Band intensity on different gels was normalized by the respective bands of the positive control (50 μg of serum proteins from the same mdx mouse present on each gel). Fifty micrograms of mouse serum were used for the analysis. 4924 |Human Molecular Genetics, 2015, Vol. 24, No. 17 at Tampere University Library. Department of Health Sciences on September 27, 2016http://hmg.oxfordjournals.org/Downloaded from 37. Wooddell, C.I., Zhang, G., Griffin, J.B., Hegge, J.O., Huss, T. and Wolff, J.A. (2010) Use of Evans blue dye to compare limb muscles in exercised young and old mdx mice. Muscle Nerve,41, 487–499. 38. Brussee, V., Tardif, F. and Tremblay, J.P. (1997) Muscle fibers of mdx mice are more vulnerable to exercise than those of normal mice. Neuromuscul. Disord.,7, 487–492. 39. Vilquin, J.T., Brussee, V., Asselin, I., Kinoshita, I., Gingras, M. and Tremblay, J.P. (1998) Evidence of mdx mouse skeletal muscle fragility in vivo by eccentric running exercise. Muscle Nerve,21, 567–576. 40. Webster, C., Silberstein, L., Hays, A.P. and Blau, H.M. (1988) Fast muscle fibers are preferentially affected in Duchenne muscular dystrophy. Cell,52, 503–513. 41. Yin, H., Saleh, A.F., Betts, C., Camelliti, P., Seow, Y., Ashraf, S., Arzumanov, A., Hammond, S., Merritt, T., Gait, M.J. et al. (2011) Pip5 transduction peptides direct high efficiency oligonucleotide-mediated dystrophin exon skipping in heart and phenotypic correction in mdx mice. Mol. Ther.,19, 1295–1303. 42. Betts, C., Saleh, A.F., Arzumanov, A.A., Hammond, S.M., Godfrey, C., Coursindel, T., Gait, M.J. and Wood, M.J. (2012) Pip6-PMO, A New Generation of Peptide-oligonucleotide Conjugates With Improved Cardiac Exon Skipping Activity for DMD Treatment. Mol. Ther. Nucleic Acids,1, e38. 43. Jearawiriyapaisarn, N., Moulton, H.M., Sazani, P., Kole, R. and Willis, M.S. (2010) Long-term improvement in mdx cardiomyopathy after therapy with peptide-conjugated morpholino oligomers. Cardiovasc. Res.,85, 444–453. 44. Duguez, S., Duddy, W., Johnston, H., Laine, J., Le Bihan, M.C., Brown, K.J., Bigot, A., Hathout, Y., Butler-Browne, G. and Partridge, T. (2013) Dystrophin deficiency leads to disturbance of LAMP1-vesicle-associated protein secretion. Cell. Mol. Life Sci., 70, 2159–2174. 45. Mussini, E., Cornelio, F., Colombo, L., De Ponte, G., Giudici, G., Cotellessa, L. and Marcucci, F. (1984) Increased myofibrillar protein catabolism in duchenne muscular dystrophy measured by 3-methylhistidine excretion in the urine. Muscle Nerve,7, 388–391. 46. McKeran, R.O., Halliday, D. and Purkiss, P. (1977) Increased myofibrillar protein catabolism in Duchenne muscular dystrophy measured by 3-methylhistidine excretion in the urine. J. Neurol. Neurosurg. Psychiatry,40, 979–981. 47. Carlson, C.G., Rutter, J., Bledsoe, C., Singh, R., Hoff, H., Bruemmer, K., Sesti, J., Gatti, F., Berge, J. and McCarthy, L. (2010) A simple protocol for assessing inter-trial and inter-examiner reliability for two noninvasive measures of limb muscle strength. J. Neurosci. Methods,186, 226–230. 48. Kobayashi, Y.M., Rader, E.P., Crawford, R.W. and Campbell, K. P. (2012) Endpoint measures in the mdx mouse relevant for muscular dystrophy pre-clinical studies. Neuromuscul. Disord., 22,34–42. 49. Hathout, Y., Marathi, R.L., Rayavarapu, S., Zhang,A., Brown, K.J., Seol, H., Gordish-Dressman, H., Cirak, S., Bello, L., Nagaraju, K. et al. (2014) Discovery of serum protein biomarkers in the mdx mouse model and cross-species comparison to Duchenne muscular dystrophy patients. Hum. Mol. Genet.,23, 6458–6469. 50. Mendell, J.R. and Lloyd-Puryear, M. (2013) Report of MDA muscle disease symposium on newborn screening for Duchenne muscular dystrophy. Muscle Nerve,48,21–26. 51. Turner, D.C., Wallimann, T. and Eppenberger, H.M. (1973) A protein that binds specifically to the M-line of skeletal muscle is identified as the muscle form of creatine kinase. Proc. Natl. Acad. Sci. U. S. A.,70, 702–705. 52. Wallimann, T., Moser, H. and Eppenberger, H.M. (1983) Isoenzyme-specific localization of M-line bound creatine kinase in myogenic cells. J. Muscle Res. Cell Motil.,4, 429–441. 53. Stolz, M. and Wallimann, T. (1998) Myofibrillar interaction of cytosolic creatine kinase (CK) isoenzymes: allocation of N-terminal binding epitope in MM-CK and BB-CK. J. Cell Sci., 111(Pt 9), 1207–1216. 54. He, B., Tang, R.H., Weisleder, N., Xiao, B., Yuan, Z., Cai, C., Zhu, H., Lin, P., Qiao, C., Li, J. et al. (2012) Enhancing muscle membrane repair by gene delivery of MG53 ameliorates muscular dystrophy and heart failure in delta-Sarcoglycan-deficient hamsters. Mol. Ther.,20, 727–735. 55. Masters, C.J., Reid, S. and Don, M. (1987) Glycolysis–new concepts in an old pathway. Mol. Cell. Biochem.,76,3–14. 56. Gizak, A., Rakus, D. and Dzugaj, A. (2003) Immunohistochemical localization of human fructose-1,6-bisphosphatase in subcellular structures of myocytes. Histol. Histopathol.,18, 135–142. 57. Wang, Y., Kinzie, E., Berger, F.G., Lim, S.K. and Baumann, H. (2001) Haptoglobin, an inflammation-inducible plasma protein. Redox Rep.,6, 379–385. 58. Le Borgne, F., Guyot, S., Logerot, M., Beney, L., Gervais, P. and Demarquoy, J. (2012) Exploration of lipid metabolism in relation with plasma membrane properties of Duchenne muscular dystrophy cells: influence of L-carnitine. PloS one, 7, e49346. 59. Stapleton, D.I., Lau, X., Flores, M., Trieu, J., Gehrig, S.M., Chee, A., Naim, T., Lynch, G.S. and Koopman, R. (2014) Dysfunctional muscle and liver glycogen metabolism in mdx dystrophic mice. PloS one,9, e91514. 60. Agarkova, I. and Perriard, J.C. (2005) The M-band: an elastic web that crosslinks thick filaments in the center of the sarcomere. Trends Cell Biol.,15, 477–485. 61. Lange, S., Himmel, M., Auerbach, D., Agarkova, I., Hayess, K., Furst, D.O., Perriard, J.C. and Ehler, E. (2005) Dimerisation of myomesin: implications for the structure of the sarcomeric M-band. J. Mol. Biol.,345, 289–298. 62. Schoenauer, R., Bertoncini, P., Machaidze, G., Aebi, U., Perriard, J.C., Hegner, M. and Agarkova, I. (2005) Myomesin is a molecular spring with adaptable elasticity. J. Mol. Biol.,349, 367–379. 63. Alderton, J.M. and Steinhardt, R.A. (2000) How calcium influx through calcium leak channels is responsible for the elevated levels of calcium-dependent proteolysis in dystrophic myotubes. Trends Cardiovasc. Med.,10, 268–272. 64. Cai, C., Masumiya, H., Weisleder, N., Matsuda, N., Nishi, M., Hwang,M.,Ko,J.K.,Lin,P.,Thornton,A.,Zhao,X.et al. (2009) MG53 nucleates assembly of cell membrane repair machinery. Nat. Cell Biol.,11,56–64. 65. Carpenter, S. and Karpati, G. (1979) Duchenne muscular dystrophy: plasma membrane loss initiates muscle cell necrosis unless it is repaired. Brain,102, 147–161. 66. Lange, S., Xiang, F., Yakovenko, A., Vihola, A., Hackman, P., Rostkova, E., Kristensen, J., Brandmeier, B., Franzen, G., Hedberg, B. et al. (2005) The kinase domain of titin controls muscle gene expression and protein turnover. Science,308, 1599–1603. 67. Roberts, T.C., Blomberg, K.E., McClorey, G., El Andaloussi, S., Godfrey, C., Betts, C., Coursindel, T., Gait, M.J., Smith, C.I. and Wood, M.J. (2012) Expression analysis in multiple muscle groups and serum reveals complexity in the microRNA transcriptome of the mdx mouse with implications for therapy. Mol. Ther. Nucleic Acids,1, e39. Human Molecular Genetics, 2015, Vol. 24, No. 17 |4931 at Tampere University Library. Department of Health Sciences on September 27, 2016http://hmg.oxfordjournals.org/Downloaded from 68. Ayoglu, B., Chaouch, A., Lochmuller, H., Politano, L., Bertini, E., Spitali, P., Hiller, M., Niks, E.H., Gualandi, F., Ponten, F. et al. (2014) Affinity proteomics within rare diseases: a BIONMD study for blood biomarkers of muscular dystrophies. EMBO Mol. Med.,6, 918–936. 69. Chapman, V.M., Miller, D.R., Armstrong, D. and Caskey, C.T. (1989) Recovery of induced mutations for X chromosomelinked muscular dystrophy in mice. Proc. Natl. Acad. Sci. U. S. A., 86, 1292–1296. 70. Laure, L., Suel, L., Roudaut, C., Bourg, N., Ouali, A., Bartoli, M., Richard, I. and Daniele, N. (2009) Cardiac ankyrin repeat protein is a marker of skeletal muscle pathological remodelling. FEBS J.,276, 669–684. 71. Boisgerault, F., Gross, D.A., Ferrand, M., Poupiot, J., Darocha, S., Richard, I. and Galy, A. (2013) Prolonged gene expression in muscle is achieved without active immune tolerance using microrRNA 142.3p-regulated rAAV gene transfer. Hum. Gene Ther.,24, 393–405. 72. Smith, R.H., Levy, J.R. and Kotin, R.M. (2009) A simplified baculovirus-AAV expression vector system coupled with one-step affinity purification yields high-titer rAAV stocks from insect cells. Mol. Ther.,17, 1888–1896. 73. Carlson, C.G. and Makiejus, R.V. (1990) A noninvasive procedure to detect muscle weakness in the mdx mouse. Muscle Nerve,13, 480–484. 74. Fougerousse, F., Bartoli, M., Poupiot, J., Arandel, L., Durand, M., Guerchet, N., Gicquel, E., Danos, O. and Richard, I. (2007) Phenotypic correction of alpha-sarcoglycan deficiency by intra-arterial injection of a muscle-specific serotype 1 rAAV vector. Mol. Ther.,15,53–61. 4932 |Human Molecular Genetics, 2015, Vol. 24, No. 17 at Tampere University Library. Department of Health Sciences on September 27, 2016http://hmg.oxfordjournals.org/Downloaded from