© 2020. Published by The Company of Biologists Ltd | Disease Models & Mechanisms (2020) 12, dmm043281. doi:10.1242/dmm.043281 RESEARCH ARTICLE Absence of p.R50X Pygm read-through in McArdle disease cellular models Guillermo Tarrasó 1,2, Alberto Real-Martinez1,2, Marta Paré s3, Lı́dia Romero-Cortadellas3, Laura Puigros3, Laura Moya3, Noemı́ de Luna4,2, Astrid Brull5, Miguel Angel Martı́n6,2, Joaquin Arenas6,2, Alejandro Lucia6,7, Antoni L. Andreu1,2, Jordi Barquinero3, John Vissing8, Thomas O. Krag8 and Tomàs Pinó s1,2, * ABSTRACT INTRODUCTION McArdle disease is an autosomal recessive disorder caused by the McArdle disease (glycogen storage disease type V, OMIM® 232600) absence of muscle glycogen phosphorylase, which leads to blocked is an autosomal recessive disorder of glycogen metabolism that was muscle glycogen breakdown. We used three different cellular models first described in 1951 (Mc, 1951). This condition is one of the most to evaluate the efficiency of different read-through agents (including frequent metabolic myopathies, with an estimated prevalence in the amlexanox, Ataluren, RTC13 and G418) in McArdle disease. The Caucasian population of 1/139,000 and both sexes similarly affected first model consisted of HeLa cells transfected with two different (Santalla et al., 2017). McArdle disease is caused by pathogenic GFP-PYGM constructs presenting the Pygm p.R50X mutation (GFP- mutations in both copies of the gene (PYGM) encoding the muscle PYGM p.R50X and PYGM Ex1-GFP p.R50X). The second cellular isoform of glycogen phosphorylase (GP-M), generally leading to loss model was based on the creation of HEK293T cell lines stably of activity and inability to break down glycogen, making this source expressing the PYGM Ex1-GFP p.R50X construct. As these plasmids of energy unavailable to patients (Mc, 1951; Tsujino et al., 1995; encode murine Pygm cDNA without any intron sequence, their Quinlivan et al., 2010; Lucia et al., 2012; Bartram et al., 1993; transfection in cells would allow for analysis of the efficacy of read- Dimauro et al., 2002). Patients present with exercise intolerance, through agents with no concomitant nonsense-mediated decay usually in the form of reversible, acute crises of early exertional interference. The third model consisted of skeletal muscle cultures fatigue and contractures that can also be accompanied by derived from the McArdle mouse model (knock-in for the p.R50X rhabdomyolysis and myoglobinuria (‘dark urine’) (Santalla et al., mutation in the Pygm gene). We found no evidence of read-through 2017; Lucia et al., 2012). Although more than 150 pathogenic PYGM at detectable levels in any of the models evaluated. We performed a mutations have been reported since the first pathogenic variants were literature search and compared the premature termination codon described in 1993 (Bartram et al., 1993; Nogales-Gadea et al., 2015; context sequences with reported positive and negative read-through Tsujino et al., 1993), clearly the most prevalent among Caucasian induction, identifying a potential role for nucleotide positions −9, −8, patients is the nonsense p.R50X (c.148C>T) mutation, with an allele −3, −2, +13 and +14 (the first nucleotide of the stop codon is assigned frequency of ∼50% (Santalla et al., 2017; Quinlivan et al., 2010; as +1). The Pygm p.R50X mutation presents TGA as a stop codon, Lucia et al., 2012; Bruno et al., 2006; Bartram et al., 1994; G nucleotides at positions −1 and −9, and a C nucleotide at −3, which el-Schahawi et al., 1996; Martin et al., 2004; Aquaron et al., 2007; potentially generate a good context for read-through induction, Gurgel-Giannetti et al., 2013; Deschauer et al., 2007). Nonsense counteracted by the presence of C at −2 and its absence at +4. mutations like the p.R50X variant introduce premature termination codons (PTCs) into the protein-coding gene sequence, thereby KEY WORDS: McArdle disease, Metabolic myopathy, Premature leading to premature termination of translation and, as a consequence, termination codon, Read-through, Cellular models to the production of truncated proteins (Mendell and Dietz, 2001). PTCs are often associated with severe disease phenotypes [and account for ∼12% of all described genetic alterations causing human 1 Mitochondrial and Neuromuscular Disorders Unit, Vall d’Hebron Institut de inherited conditions (Mort et al., 2008)]. In McArdle disease, ∼35% Recerca, Universitat Autònoma de Barcelona, Barcelona 08035, Spain. 2Centro de of all mutations other than p.R50X generate PTCs (Aquaron et al., Investigació n Biomé dica en Red de Enfermedades Raras (CIBERER), Madrid 2007; Deschauer et al., 2007; Rubio et al., 2006, 2007; Nogales- Disease Models & Mechanisms 28029, Spain. 3Gene and Cell Therapy Unit, Vall d’Hebron Institut de Recerca, Universitat Autònoma de Barcelona, Barcelona 08035, Spain. 4Laboratori de Gadea et al., 2007). Furthermore, ∼75% of all patients (251 of 333) in Malalties Neuromusculars, Institut de Recerca Hospital de la Santa Creu i Sant Pau, Universitat Autònoma de Barcelona, Barcelona 08041, Spain. 5Sorbonne the Spanish registry of McArdle disease present a PTC in at least one Université , INSERM UMRS_974, Center of Research in Myology, 75013 Paris, of the two gene copies (Santalla et al., 2017). France. 6Mitochondrial and Neuromuscular Diseases Laboratory, 12 de Octubre There is currently no curative therapy for McArdle disease. Hospital Research Institute (i+12), Madrid 28041, Spain. 7Faculty of Sport Sciences, European University, Madrid 28670, Spain. 8Copenhagen Gene therapy has been evaluated in the spontaneously occurring Neuromuscular Center, Department of Neurology, Rigshospitalet, University of McArdle sheep model and recently in the knock-in (p.R50X Copenhagen, Copenhagen DK-2100, Denmark. homozygous) McArdle mouse model (McNamara et al., 2019; *Author for correspondence (tomas.pinos@vhir.org) Howell et al., 2008). In the sheep model, intramuscular application of two different vectors (adenovirus 5 vector and an adeno-associated L.R.-C., 0000-0003-1080-3774; L.M., 0000-0002-2630-7911; M.A.M., 0000- virus serotype 2) containing GP-M expression cassettes produced 0003-4741-772X; T.O.K., 0000-0001-9330-668X; T.P., 0000-0002-4379-0917 only local expression of functional GP-M, and the number of GP-M- This is an Open Access article distributed under the terms of the Creative Commons Attribution expressing fibers decreased with time (Howell et al., 2008). License (https://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution and reproduction in any medium provided that the original work is properly attributed. Furthermore, intraperitoneal injection of recombinant adeno- associated virus serotype 8 containing a functional copy of Pygm in Received 5 November 2019; Accepted 5 December 2019 the McArdle mouse model resulted in Pygm expression, improved 1 RESEARCH ARTICLE Disease Models & Mechanisms (2020) 12, dmm043281. doi:10.1242/dmm.043281 skeletal muscle architecture, reduced accumulation of glycogen and induce ‘read-through’ in PTCs, including 3-[5-(2-fluorophenyl)- restoration of voluntary running wheel activity, but a lack of 1,2,4-oxadiazol-3-yl]-benzoic acid (also known as PTC124, Ataluren improvement in the hanging wire ability (McNamara et al., 2019). or Translarna) (Welch et al., 2007), RTC13 and RTC14 (Du et al., Thus, further studies are required in this field before this approach 2009a) and amlexanox (Gonzalez-Hilarion et al., 2012), among can be proposed in patients. Several drug treatments have also been others. Thus, it was the purpose of our study to assess the efficacy of evaluated in vitro or in vivo. Valproate, an inhibitor of histone different RTAs in (1) transiently transfected cells with p.R50X deacetylation, was shown to induce a remarkable decrease in plasmid constructs, (2) cells stably expressing these constructs glycogen levels by increasing the expression of the brain isoform and (3) primary skeletal muscle cells derived from the McArdle of glycogen phosphorylase in skeletal muscle cultures derived from mouse model. the knock-in (p.R50X homozygous) McArdle mouse model (de Luna et al., 2015). Similarly, treatment of the McArdle sheep model RESULTS with valproate induced the re-expression of the brain/developmental Absence of read-through effect on transiently transfected glycogen phosphorylase in their muscle fibers, although this HeLa cells effect was not accompanied by an increase in exercise capacity We first evaluated the capacity of the different compounds to induce (Howell et al., 2015). However, the fact that PTC-causing mutations read-through of the p.R50X mutation in the Pygm gene in cells with are the most prevalent pathogenic variants in McArdle disease high levels of Pygm expression, in conditions in which the Pygm patients provides biological rationale to assess the effectiveness of the mRNA is not subjected to the nonsense-mediated decay (NMD) treatment with read-through agents (RTAs) such as aminoglycosides, process, a specialized mRNA surveillance mechanism that aims to which are known to induce the ribosome to bypass a PTC (Bidou reduce the synthesis of deleterious C-terminally truncated proteins et al., 2012). In this regard, a preliminary trial with short-term in eukaryotic organisms (Kayali et al., 2012). We used HeLa cells (10-day) gentamicin treatment in McArdle disease patients with a transfected with four different plasmids that contained the wild-type PTC failed to normalize 31P-magnetic resonance spectroscopy (WT) or mutated ( p.R50X) Pygm complementary DNA (cDNA) indicators of GP-M deficiency in muscle (Schroers et al., 2006). sequence (GFP-PYGM WT and p.R50X, and PYGM Ex1-GFP-WT However, read-through of the PYGM p.R50X mutation was and p.R50X) (Fig. 1A,B). In order to optimize transfection reported in non-muscle culture cell models (Chinese hamster ovary efficiency, different amounts of the GFP-PYGM WT plasmid cells) that had been transiently transfected with p.R50X-GFP were transfected in HeLa cells and GFP fluorescence was analyzed plasmid constructs and treated with the aminoglycoside G418 24 h after transfection. There was a positive correlation between (Birch et al., 2013). Other low-molecular RTAs were reported to the amount of GFP-PYGM fluorescence signal and the quantity of Fig. 1. GFP-PYGM plasmid constructs used to analyze p.R50X Pygm read-through. (A-C) Different GFP-PYGM constructs were transiently (A,B) and stably (C) transfected in HeLa and HEK293T cells, respectively. (A) The GFP-PYGM wild-type (WT) and p.R50X constructs included the GFP encoding sequence followed by the full- length mouse Pygm cDNA encoding sequence, and varied in the absence or presence of the p.R50X mutation. The XhoI restriction site (yellow) disrupted the GFP coding sequence corresponding to the last 20 amino acids of the GFP protein. The MluI (blue) restriction site was used to insert the Pygm cDNA into the pCI-neo multicloning site. (B) The PYGM Ex1-GFP WT and p.R50X constructs consisted of the mouse exon 1 Pygm cDNA encoding the first 50 amino acids of the GP-M protein followed by the GFP encoding sequence. MluI (blue) and NheI (orange) restriction sites were used to insert the Pygm sequence, while NheI and XhoI (yellow) restrictions sites were used to insert the GFP encoding sequence into the pCI-neo multicloning site. (C) Structure of pC-PYGM EX1-GFP WT (left) and pC-PYGM-EX1-GFP p.R50X (right). EF1α, elongation factor 1α; GFP, enhanced green fluorescent protein; IRES, internal ribosomal entry site; Bsd, blasticidin S resistance gene. Disease Models & Mechanisms 2 RESEARCH ARTICLE Disease Models & Mechanisms (2020) 12, dmm043281. doi:10.1242/dmm.043281 Disease Models & Mechanisms Fig. 2. Transfection efficiency analysis using different amounts of the GFP-PYGM plasmids. (A) GFP PYGM fluorescence analysis of HeLa cells transfected with different amounts of GFP-PYGM WT. Images were obtained 24 h after transfection. Scale bars: 75 μm. (B,C) Western blot analyses of HeLa cells transfected with different amounts of GFP-PYGM WT and p.R50X plasmids. In both B and C, measures were performed 72 h after transfection. Black arrows mark specific protein bands; the blue arrow marks unspecific protein band. (D) GFP-PYGM fluorescence details of cells transfected with GFP-PYGM WT and p.R50X plasmids. Images were obtained 24 h after transfection. Scale bars: 20 μm. (E) Cell growth studies of HeLa cells transfected with 1.5, 2 and 2.5 µg GFP-PYGM WT and p.R50X plasmids. Two independent experiments were performed in each condition. Error bars indicate s.d. between the two independent experiments. GFP-PYGM WT plasmid transfected in these cells (Fig. 2A). indicating a reduction in cell viability with higher amounts of Additionally, a moderate decrease in cell confluency was observed plasmid transfection (Fig. 2A). When western blot analyses were in HeLa cells transfected with 2.0 µg and 2.5 µg of GFP-PYGM WT performed, similar amounts of GFP-PYGM protein were observed plasmid compared to cells transfected with 0.5, 1.0 and 1.5 µg, in cells transfected with 0.5, 1, 1.5, 2 and 2.5 µg of GFP-PYGM WT 3 RESEARCH ARTICLE Disease Models & Mechanisms (2020) 12, dmm043281. doi:10.1242/dmm.043281 plasmid (Fig. 2B). However, a positive correlation was observed PYGM p.R50X-transfected cells when treated with 50, 100 and between GFP-PYGMp.R50X band intensity and the amount of 200 mg/ml G418 (Fig. 4C). An attempt to reproduce this specific transfected plasmid (Fig. 2C). Interestingly, HeLa cells transfected result failed, thus a read-through effect caused by the G418 with GFP-PYGM WT plasmid presented a cytosolic punctuate GFP- treatment could not be confirmed. PYGM localization, whereas those transfected with the GFP-PYGM p.R50X plasmid showed a diffuse GFP-PYGMp.R50X localization Absence of read-through effect on stably transfected (Fig. 2D). Overall, these results suggest that higher levels of WT and HEK293T cells GFP-PYGMp.R50X protein are obtained when 1.5, 2.0 and 2.5 µg of A wider panel of RTAs was used to induce read-through in human both plasmids are transfected, although they might imply a embryonic kidney 293T (HEK293T) cells stably expressing the WT reduction in cell viability. In this regard, cell growth was analyzed and the pC-PYGM Ex1-GFP WT and p.R50X constructs (Table 1). in HeLa cells 72 h after transfection with 1.5, 2.0 and 2.5 µg of In a set of preliminary experiments, we tested different ranges of the distinct plasmids, and, although no major differences were drug concentrations and measured cell viability by flow cytometry observed, a trend for higher cell growth in cells transfected with after labeling with 7-aminoactinomycin (7-AAD). Based on these 2.0 µg of plasmids was observed (Fig. 2E). Thus, as a compromise, results, we established the optimal dose of the drugs and determined 2 µg of plasmid was used in the subsequent experiments. Next, we that the optimal time for analysis was 48 h after drug exposure (data evaluated the potential toxicity of the different RTAs (amlexanox, not shown). None of the RTAs tested increased the mean Ataluren, RTC13 and G418) in HeLa cells. Thus, cells were treated fluorescence intensity (MFI) (Fig. 5) or the percentages of GFP+ for 72 h with either amlexanox, Ataluren, RTC13 or G418 at doses cells (data not shown) in cells carrying the mutated form of Pygm, that have previously been shown to induce read-through in different indicating the lack of read-through effect. cell cultures with no apparent toxicity [i.e. 10 µM for amlexanox (Gonzalez-Hilarion et al., 2012), PTC124 (Lentini et al., 2014; Du Primary cultures treated with read-through compounds et al., 2013; Kayali et al., 2012) or RTC13 (Du et al., 2013; Kayali present NMD et al., 2012; Nakamura et al., 2012), and 100 µg/ml for G418 We next evaluated whether amlexanox, Ataluren and G418 were (Howard et al., 1996; Heier and DiDonato, 2009; Friesen et al., able to suppress the NMD of Pygm mRNA in the skeletal muscle 2018; Floquet et al., 2011)]. No significant changes in cell growth cultures derived from the McArdle mouse model. In this case, no were observed in HeLa cells treated with amlexanox, Ataluren or significant increase in Pygm mRNA levels was observed for any of G418 compared to non-treated cells (Fig. 3). However, a progressive the drugs and doses tested (Fig. 6A-C). Therefore, these results reduction in cell number was observed between 24 h and 72 h in indicate that amlexanox, Ataluren and G418 are unable to suppress HeLa cells treated with RTC13 (Fig. 3). Consequently, RTC13 was NMD in this model. not further evaluated as a potential RTA in this cell model. Finally, when the read-through capacity for amlexanox, Ataluren and G418 Absence of read-through effect on primary cultures was tested in transfected cells, no read-through effect was observed In western blot analyses, the 94 kDa band corresponding to the full- in amlexanox- or Ataluren-treated cells, as the 118 kDa and 34 kDa length GP-M protein was observed in treated and untreated WT bands corresponding to the full-length translation of the GFP- myotubes (Fig. 6D-F; Fig. S1). By contrast, no specific bands were PYGM p.R50X and PYGM Ex1-GFP p.R50X were not detected detected in McArdle mouse myotubes for any dose of the four (Fig. 4A,B). However, a faint 118 kDa band was observed in GFP- compounds tested (amlexanox, Ataluren, RTC13 and G418) Disease Models & Mechanisms Fig. 3. Viability of HeLa cells after treatment with amlexanox, Ataluren, G418 or RTC13. Results were obtained after 24, 48 and 72 h of treatment with amlexanox (10 µM), Ataluren (10 µM), G418 (100 µg/ml) and RTC13 (10 µM). Scale bars: 200 µm unless indicated. 4 RESEARCH ARTICLE Disease Models & Mechanisms (2020) 12, dmm043281. doi:10.1242/dmm.043281 Fig. 4. Read-through analysis in transient transfected cells. (A-C) Read-through analysis in HeLa cells transfected with WT, p.R50X GFP-PYGM and PYGM- GFP plasmids after 72 h treatment with amlexanox (A), Ataluren (B) and G418 (C). The GFP-PYGM WT plasmid generates a fusion protein of 1079 amino acids and a molecular weight of 118 kDa, consisting of the N-terminal GFP protein (219 amino acids) fused to the full-length PYGM protein (18 amino acids from the 5′UTR+842- amino-acid full-length protein). In the absence of read- through induction, the GFP-PYGM p.R50X plasmid generates a fusion protein of 286 amino acids (GFP 219 amino acids+18 amino acids PYGM 5′UTR and 49 amino acids from PYGM coding sequence) and a molecular weight of 31 kDa. The PYGM-Ex1 GFP WT plasmid generates a fusion protein of 309 amino acids and molecular weight of 34 kDa, consisting of the first 50 amino acids of the GP-M protein fused to the 219 amino acids from the GFP coding sequence and the 40 unspecific extra amino acids (8 and 32 in the N- and C-terminus of GFP protein, respectively) derived from the cloning process. Finally, in the absence of read-through induction, the PYGM-Ex1 GFP p.R50X plasmid generates a protein consisting only of the first 49 amino acids (∼5 kDa) of the GP-M protein. Blue arrows in C indicate potential read-through obtained for the GFP- PYGM p.R50X construct after treatment with G418, which could not be replicated in further experiments. Disease Models & Mechanisms (Fig. 6D-F; Fig. S1). The absence of full-length PYGM protein in which nucleotide/s surrounding the PTC might influence the treated McArdle myotubes suggests the lack of read-through effect. read-through induction caused by the different compounds. Thus, These findings are consistent with the results from transiently and we analyzed the context sequence [15 nucleotides upstream (from stably transfected cells. −1 to −15) and downstream (from +3 to +18) from the PTC] from 30 different sequences with reported positive read-through Analysis of mouse Pygm p.R50X nucleotide context induction and 29 sequences with described negative read-through sequence induction. In sequences with positive read-through induction, TGA As the read-through efficiency is influenced by the stop codon was the most represented stop codon (66.6%), followed by TAG composition and its surrounding sequence (Howard et al., 2004), (30.0%) and TAA (3.3%) (Fig. 7A). In sequences with absence of we wanted to determine whether the absence of read-through read-through induction, there was a clear reduction in the percentage induction could be caused by the nature of the Pygm p.R50X PTC of sequences with TGA stop codon (55.1%) as well as a fourfold and its surrounding sequence. To this end, we needed to establish increase in the percentage of TAA stop codon (13.8%) (Fig. 7B). 5 RESEARCH ARTICLE Disease Models & Mechanisms (2020) 12, dmm043281. doi:10.1242/dmm.043281 Table 1. RTAs and doses used in each cell model Treatment McArdle disease cellular models RTAs Dose Duration (h) Transiently transfected cells Stably transfected cells Primary skeletal muscle cultures Amlexanox 0.5 µM 72 X X 1 µM 72 X X 5 µM 72 X 10 µM 72 X X 25 µM 72 X X 30 µM 48 X Ataluren 5 µM 72 X X 10 µM 72 X X 20 µM 72 X X 30 µM 48 X 35 µM 72 X X 50 µM 72 X X RTC13 2.5 µM 72 X 5 µM 72 X 10 µM 48 X 72 X 37.5 µM 72 X RTC14 10 µM 48 X G418 25 µg/ml 72 X X 50 µg/ml 48 X 72 X X 100 µg/ml 72 X X 200 µg/ml 72 X X 500 µg/ml 72 X X Azithromycin 50 µg/ml 48 X Erythromycin 100 µg/ml 48 X Gentamycin 100 µg/ml 48 X Neomycin 100 µg/ml 48 X RTA, read-through agent. Additionally, major differences between sequences with or without of the C nucleotide in position −2 was more than twofold higher in read-through induction were that 80% (24/30) of the sequences with sequences with no read-though induction (∼41%) than in those read-through induction presented a G or T nucleotide in position −9 showing induction (∼20%) (Fig. 7A,B). Furthermore, the presence versus 52% (15/29) in those with no induction; 73% (22/30) of the of C nucleotide in position +4, which promotes PTC read-through sequences with read-through induction presented an A or T (Dabrowski et al., 2015), was 33% and 21% in sequences with and nucleotide in position −8 versus 52% in those with no induction; without induction, respectively (Fig. 7A,B). The estimated and 50.0% (15/30) of the sequences with read-through induction consensus sequence for the 30 and 29 sequences with and without presented a C nucleotide in position −3 versus 34.5% (10/29) in induction, respectively, are shown in Fig. 7C and D. Compared to those showing no induction (Fig. 7A,B). By contrast, the frequency these analyzed sequences, mouse Pygm p.R50X sequence presented the TGA stop codon, as well as G at position −9, and C at −3 and −2, but did not present A/G or A at +13 or +14 (Fig. 7E). DISCUSSION In the present work, we report the absence of Pygm p.R50X read- through induction in McArdle disease cell cultures using a variety of RTAs. Three different cell culture models were used to evaluate the efficiency of these RTAs in McArdle disease. The first model Disease Models & Mechanisms consisted of HeLa cells transiently transfected with two different GFP-PYGM constructs presenting the p.R50X mutation (GFP- PYGM p.R50X and PYGM Ex1-GFP p.R50X). The rationale for the evaluation of the distinct RTAs in cells transfected with these two constructs was based on their capacity to generate GFP-PYGM fusion proteins of different molecular weight, depending on whether p.R50X read-through is induced or not. Additionally, as these two plasmids include mouse Pygm cDNA without any intron sequence, Fig. 5. Flow-cytometric analysis of the read-through effect 48 h after their transfection in cells would allow for the analysis of the efficacy treatment. Mean fluorescence intensity (MFI) in non-transfected cells and in of read-through induction with no concomitant NMD interference those stably expressing the WT and the p.R50X constructs. Azithromycin and geneticin, 50 μg/ml; erythromycin, gentamycin and neomycin, 100 μg/ml; (Pérez-Ortín et al., 2013). No read-through induction was observed RTC13 and RTC14, 10 μM; Ataluren and amlexanox, 30 μM. Experiments in HeLa cells transfected with either GFP-PYGM p.R50X or PYGM were performed in duplicate for all conditions. Error bars indicate s.d. from EX1-GFP p.R50X plasmids when treated with amlexanox, Ataluren duplicate experiments. or G418. These results are discordant with those obtained by Birch 6 RESEARCH ARTICLE Disease Models & Mechanisms (2020) 12, dmm043281. doi:10.1242/dmm.043281 Fig. 6. Nonsense-mediated decay (NMD) and read-through analysis in McArdle myotubes. (A-C) PYGM mRNA levels in WT and McArdle mouse myotubes after 72 h treatment with different concentrations of amlexanox (AML) (A), Ataluren (ATL) (B) and G418 (C), respectively. (D-F) GP-M protein levels in WT and McArdle mouse myotubes after 72 h treatment with different concentrations of AML (D), ATL (E) and G418 (F), respectively. Black arrows mark the specific PYGM protein band (94 kDa); blue arrows mark unspecific protein bands. et al. (2013), who reported that G418 induced read-through in 2015). In these cells, NMD of Pygm mRNA has been shown to Chinese ovary hamster cells transfected with the PYGM EX1-GFP occur, as Pygm mRNA levels of myotubes derived from McArdle p.R50X plasmid. The study did not proceed with RTC13 as it mice are ∼10% of those from their WT counterparts (de Luna et al., negatively affected the viability of HeLa cells over time. RTC13 is a 2015). Thus, this model is useful not only to evaluate the read- non-aminoglycoside low-molecular-mass compound identified in a through capacity of the different compounds, but also to assess their high-throughput screening of ∼34,000 compounds in the search for capacity for inhibiting the NMD mechanism. Although some reports read-through activity (Du et al., 2009a). Together with another read- have indicated that a moderate induction of the PTC read-through by through compound identified in the screening, RTC14, RTC13 has using RTAs may promote a stabilization of mutant transcripts and proven to suppress nonsense mutations in ataxia telangiectasia counteract the NMD process (Gonzalez-Hilarion et al., 2012; Floquet patient cell lines and in myotube cells from the mdx mouse model of et al., 2011; Bellais et al., 2010; Salvatori et al., 2009), we did not find Duchenne muscular dystrophy (Kayali et al., 2012; Du et al., a significant change in the Pygm mRNA levels in McArdle mouse 2009b). Remarkably, no major toxicity effects have been reported skeletal muscle cultures when treated with different doses of for RTC13 in any of the mammalian cells tested. amlexanox, Ataluren or G418. Consistent with these findings, we The second cell culture model used to evaluate Pygm p.R50X did not detect GP-M protein in any of the conditions tested. Thus, in read-through was based on HEK293T cells stably expressing our experimental conditions, the results obtained in these three modified versions of the mutated constructs. In this case, we used cellular models indicate an absence of p.R50X read-through Disease Models & Mechanisms the EF1A (also known as EEF1A1) promoter to minimize silencing induction capacity of the different RTAs tested – amlexanox, phenomena, and used flow cytometry to detect small changes in Ataluren, RTC13 and G418. In addition, the aminoglycosides fluorescence in large amounts of cells. Additionally, we have azithromycin, erythromycin, gentamycin, neomycin and RTC14 analyzed a wider panel of RTAs (including the aminoglycosides were unable to induce p.R50X read-through on HEK293T stably azithromycin, erythromycin, gentamycin and neomycin, as well as transfected cells. Previous research in McArdle patients showed that the non-aminoglycoside RTC14), but no read-through effect was 10-day administration of gentamicin failed to provoke full-length GP- detected in the experimental conditions used. M production or attenuation of disease manifestations [e.g. no The third cellular model evaluated the read-through activity in reduction in the levels of serum creatine kinase, a muscle damage skeletal muscle cultures derived from the McArdle mouse model. In biomarker (Schroers et al., 2006)]. contrast to what occurs with muscle cultures derived from affected The absence of Pygm p.R50X read-through induction in our humans, McArdle mouse skeletal muscle cultures do not present GP experimental conditions might be multifactorial. Our choice of read- protein and activity, and accumulate large amounts of glycogen through compounds was meant to reflect different types of deposits, and therefore represent a suitable in vitro model to analyze compounds, not to cover the entire and constantly expanding and evaluate potential treatments for the disease (de Luna et al., spectrum of compounds. Amlexanox, Ataluren, RTC13 and G418 7 RESEARCH ARTICLE Disease Models & Mechanisms (2020) 12, dmm043281. doi:10.1242/dmm.043281 Fig. 7. Analysis of PTC sequences with positive and negative read-through induction. (A) Multiple Alignment using Fast Fourier Transform (MAFFT) of 30 sequences with described positive read-through induction. (B) MAFFT alignment of 29 sequences with reported absence read-through induction. Orange arrows mark highly conserved nucleotides (>70% for the combination of two nucleotides and ≥50% for a single nucleotide) among the aligned sequences. (C) Consensus sequence obtained from the 30 sequences with described read-through induction. (D) Consensus sequence obtained from the 29 sequences with reported absence of read-through induction. (E) Mouse Pygm PTC context sequence. Potentially relevant nucleotides for read-through induction are squared. have proven to be effective in read-through induction in several compounds for read-through induction. It has been reported that the different PTCs both in vitro and in vivo (Gonzalez-Hilarion et al., stop codon sequence influences read-through induction; in this 2012; Howard et al., 1996; Heier and DiDonato, 2009; Floquet regard TGA>TAG>TAA is the order in which the different stop et al., 2011; Salvatori et al., 2009; Brasell et al., 2019; Yu et al., codons present higher read-through potential (Dabrowski et al., Disease Models & Mechanisms 2014; Tan et al., 2011; Wilschanski et al., 2011; Du et al., 2008; 2015). Additionally, experimental studies in a number of organisms Sánchez-Alcudia et al., 2012; Schwarz et al., 2015; Martorell et al., have shown that both downstream and upstream sequence context 2019). However, it is also important to mention that some criticism plays an essential role in determining the read-through potential of has been raised against Ataluren, its mode of action and its ability to stop codons. In both bacteria and eukaryotes, the base immediately induce PTC read-through (Chowdhury et al., 2018; McElroy et al., following the stop codon exerts the strongest influence on read- 2013; Auld et al., 2009; Dranchak et al., 2011). Adequate dosage through efficiency. For example, the level of basal TGA-C read- and incubation time may be factors for each of the tested through in mammalian cells (3-4%) was shown to be three to six compounds, but our choice of both relied on past published read- times higher than for the remaining TGA-tetranucleotides through induction in various in vitro/vivo conditions (Gonzalez- (Dabrowski et al., 2015). In yeast (Saccharomyces cerevisiae), at Hilarion et al., 2012; Lentini et al., 2014; Heier and DiDonato, least six nucleotides after the stop codon are a key determinant of 2009; Tan et al., 2011; Sánchez-Alcudia et al., 2012; Dranchak read-through efficiency (Namy et al., 2001). Regarding the PTC et al., 2011; Gómez-Grau et al., 2015). Finally, we evaluated the upstream sequence context, in experiments performed with mouse possibility that the nucleotide context sequence surrounding the fibroblasts (NIH3T3) and human HEK293T cell lines, the highest p.R50X mutation might have influenced the capacity of the different read-through occurrence was observed when the −1 position (the 8 RESEARCH ARTICLE Disease Models & Mechanisms (2020) 12, dmm043281. doi:10.1242/dmm.043281 first nucleotide of the stop codon is assigned as +1) was occupied by on six-well plates with Dulbecco’s modified Eagle medium (DMEM) high adenine or generally purine; in these models, uracil at the −1 glucose (Biowest, Nuaillé, France) plus 10% fetal bovine serum (FBS; position was always associated with the lowest read-through level Gibco™, Thermo Fisher Scientific, Waltham, USA), 100× 1% non- (Dabrowski et al., 2015). In this respect, our analysis of 30 PTCs and essential amino acids, 200 mM L-glutamine, 100 mM sodium pyruvate (all Biowest) and 100× penicillin/streptomycin (Pen-Strep solution, Biological their associated context sequences that presented positive read- Industries, Kibbutz Beit HaEmek, Israel), and incubated at 37°C, 5% CO2 through induction, in comparison to 29 different sequences with and 100% relative humidity. When the cells reached 70% confluence, negative read-through induction, suggest a potential role of different amounts of each construct (0, 0.5, 1 1.5, 2 and 2.5 µg) were added nucleotide positions −9, −8, −3, −2, +13 and +14. The mouse to different wells following the Lipofectamine 3000 Transfection Reagent Pygm p.R50X mutation presents TGA as a stop codon, G nucleotide (Invitrogen, Carlsbad, CA, USA) protocol. at positions −1 and −9, and C nucleotide at −3, which potentially generate a good context for read-through induction, counteracted by Generation of stable cell clones expressing the constructs the presence of C at −2 and its absence at +4. Similarly, human HEK293T cells were obtained from ATCC and tested for contamination. These PYGM p.R50X context sequence presents homology of ∼85% with cells were cultured in DMEM high glucose supplemented with 10% FBS, the mouse sequence and the potential read-through hot spots are 200 mM 1% L-glutamine (all from Biowest) and 100× penicillin/streptomycin conserved among both species. (Capricorn Scientific, Ebsdorfergrund, Germany) and incubated at 37°C, 5% Overall, amlexanox, Ataluren, RTC13 and G418 failed to induce CO2 and 100% relative humidity. Cells were seeded in a 24-well plate so that confluence at the time of transfection, 24 h after seeding, was 70% (0.5-10×105 p.R50X read-through in three McArdle disease cell culture models cells/well). Both plasmids (pC-PYGM Ex1-GFP WT and p.R50X) were containing the mouse Pygm sequence. It is likely that these would also transfected with Lipofectamine 3000 Transfection Reagent (Invitrogen), fail to induce read-through in the human sequence based on the following the manufacturer’s instructions. Non-transfected cells were used as a complete conservation of the read-through hotspots. However, further control. Forty-eight hours after transfection, cells were treated with Bsd studies confirming these results in the human PYGM p.R50X (Gibco™) at 15 µg/ml, a concentration that was found to be optimal for sequence, as well as using newer RTAs such as NB74, NB84, selection in preliminary experiments (data not shown). Bsd was added to the NB124, GJ071, GJ072, BZ6, BZ16 and Clitocine among others, could cultures on the day of seeding and replaced every 3-4 days with the cell provide a better understanding of the necessary molecular passages. Bsd-resistant cells expressing the WT construct were analyzed by requirements and spatial allowance in the ribosome, as it has been flow cytometry after 7 days of selection to determine GFP expression. In order described that as little as 2% of normal GP-M activity might be enough to obtain more homogeneity, at this point, single cells were isolated using a FACSAria™ I cell sorter (BD Biosciences, Franklin Lakes, NJ, USA) and to ameliorate the disease phenotype in patients (Vissing et al., 2009). individual clones were expanded in culture in the presence of Bsd. In the case of cells carrying the WT construct, the clones were generated from the GFP+ MATERIALS AND METHODS cell population. Non-transfected and transfected untreated cells were used as Plasmid constructs controls. Four constructs encoding the WT and mutant ( p.R50X) GP-M and enhanced green fluorescent protein (GFP) fusion proteins were a kind gift Skeletal muscle cultures from Dr Glenn E. Morris (Birch et al., 2013). The Pygm sequence used in WT and skeletal muscle cultures derived from the McArdle mouse model these constructs corresponded to the mouse sequence (NCBI Accession were obtained as previously reported (de Luna et al., 2015). Myoblasts were BC012961) (Birch et al., 2013), as mouse and human GP-M amino acid grown in Ham F10 medium (Biowest) supplemented with 20% FBS sequences have high identity (97%). Two of these consisted of N-terminal (Gibco™), 1% glutamine (Biowest), Pen-Strep solution (Biological GFP fused to WT and p.R50X GP-M full-length sequences inserted into a Industries) and 20 ng/ml basic fibroblast growth factor (PeproTech, pCI-neo vector, while the other two consisted of C-terminal fusions of GFP London, UK) on 10-cm dishes until 90% confluence. Then, cells were to the N-terminal 50 amino acids of WT or 49 amino acids of p.R50X GP-M transferred to six-well plates and grown in DMEM high glucose (Biowest) protein (Fig. 1A,B). The pCI-neo vector provides the cytomegalovirus supplemented with 2% horse serum (Gibco™), 1% glutamine (Biowest) and immediate early promoter ( pCMV). In order to check the sequence of the Pen-Strep solution (Biological Industries) for myotube differentiation. constructs and the presence/absence of the p.R50X mutations, the four plasmids were sequenced using the primer comprising the beginning of the Treatment with RTAs mouse Pygm coding sequence (5′-ATGTCCAGGCCTCTTTCAGACCAG- A variety of RTAs were tested in the different cellular models using a wide 3′), T7 promoter primer (5′-TAATACGACTCACTATAGG-3′) and range of concentrations (Table 1). Treatment was initiated 8 h after T3 promoter primer (5′-ATTAACCCTCACTAAAGGG-3′) surrounding transfection in HeLa cells, while in the skeletal muscle cultures it was the pCI-neo multicloning site. The GFP coding sequence from the started after 48 h in differentiation medium. In both cases, the treatment was four constructs was disrupted by the XhoI restriction site used in the maintained for 72 h [based on previous reports (Lentini et al., 2014; Du Disease Models & Mechanisms cloning process and, as a consequence, the last 20 amino acids from the GFP et al., 2013; Nakamura et al., 2012; Howard et al., 1996; Heier and protein were missing (from Ala219 to Arg238) (Fig. 1A,B). In addition, DiDonato, 2009; Friesen et al., 2018; Floquet et al., 2011)] and the medium the WT and p.R50X GFP-PYGM constructs included 18 amino acids was changed every day. In the case of the HEK293T stable cell clones, that resulted from the translation of the 5′UTR Pygm sequence included in treatment was initiated 24 h after seeding and maintained for 48 h. the construct because of XhoI and MluI restriction site usage during the cloning process (Fig. 1A). Finally, the WT and p.R50X PYGM-GFP mRNA analyses constructs included 32 amino acids between the disrupted GFP protein and Total RNA was obtained from cultured cells following the manufacturer’s the first stop codon of the pCI-neo vector (Fig. 1B). For the stable expression instructions for TRIzol (Invitrogen). RNA was treated with DNase I, of the constructs, the WT and p.R50X PYGM-GFP sequences were amplification grade (Invitrogen) to eliminate any traces of DNA. cDNA was subcloned into the pCMV/blasticidin S (Bsd) vector in which the expression synthesized from RNA using a high-capacity cDNA archive kit (Applied of the WT and p.R50X PYGM-GFP (C-terminal fusions) was driven by Biosystems, Foster City, CA, USA), which uses random primers. We used the EF1A promoter and contained the Bsd resistance gene as a selectable real-time polymerase chain reaction (PCR), with TaqMan fluorogenic marker (Fig. 1C). probes in a 7900 Real-Time PCR System (Applied Biosystems) to assess Pygm RNA levels (Mm00478582_m1). Results were normalized to Transient transfections peptidylprolyl isomerase A (cyclophilin A, Ppia) mRNA levels ( probe HeLa cells were obtained from American Type Culture Collection (ATCC; Mm02342430_g1) and quantified using 7900 SDS v2.4.1 software Manassas, VA, USA) and tested for contamination. These cells were grown (Applied Biosystems). 9 RESEARCH ARTICLE Disease Models & Mechanisms (2020) 12, dmm043281. doi:10.1242/dmm.043281 Western blot analysis Author contributions Protein extraction from transfected cells and skeletal muscle cultures was Conceptualization: G.T., A.L.A., J.B., J.V., T.O.K., T.P.; Methodology: G.T., A.R.-M., performed by adding extraction buffer (20 mM Tris-HCl pH 7.5, 150 mM M.P., L.R.-C., L.P., L.M., N.d.L., J.B., T.P.; Validation: G.T., A.L., J.V., T.P.; Formal analysis: G.T., M.P., L.R.-C., L.P., L.M., N.d.L., A.L., T.P.; Investigation: G.T., NaCl and 1% Triton X-100) to each well. Then, cells were sonicated for 5 s A.R.-M., T.P.; Data curation: G.T., A.R.-M.; Writing - original draft: T.P.; Writing - (one cycle 100% amplitude) with a Labsonic® M homogenizer (Sartorius review & editing: A.B., M.A.M., J.A., A.L., A.L.A., J.B., J.V., T.O.K., T.P.; Biotech, Göttingen, Germany). Equal amounts of extracted proteins (30 μg) Visualization: G.T., T.O.K.; Supervision: T.O.K., T.P.; Project administration: T.P.; were resolved on Criterion™ TGX™ 4-15% Precast Midi gels (Bio-Rad, Funding acquisition: T.P. Hercules, CA, USA) at 150 V for 90 min and blotted to a polyvinylidene difluoride (PVDF) membrane (Immun-Blot® PVDF membrane, Bio-Rad) Funding using a Trans-Blot® SD Semi-Dry Transfer Cell (Bio-Rad) at 20 V for This work was supported by Instituto de Salud Carlos III (PI16/01492, PI15/00172 50 min. Membranes were incubated in anti-GFP (A6455, Thermo Fisher and PI15/00558) and the European Regional Development Fund. Scientific; 1:2000) or anti-PYGM (19716-1-AP, Proteintech™, Rosemont, IL, USA; 1:2000) primary antibodies overnight at 4°C and in horseradish Supplementary information peroxidase (HRP)-conjugated goat anti-rabbit (Jackson Laboratories, Supplementary information available online at Baltimore Pike, PA, USA) secondary antibody for 3 h at room http://dmm.biologists.org/lookup/doi/10.1242/dmm.043281.supplemental temperature. Ponceau S staining (Sigma-Aldrich, St Louis, MO, USA) was used as a loading control for all the membranes. Membranes were References Aquaron, R., Bergé -Lefranc, J.-L., Pellissier, J.-F., Montfort, M.-F., Mayan, M., developed with Immobilon Western Chemiluminescent HRP Substrate Figarella-Branger, D., Coquet, M., Serratrice, G. and Pouget, J. (2007). (Merck-Millipore, Burlington, MA, USA) and images were obtained with Molecular characterization of myophosphorylase deficiency (McArdle disease) in an Odyssey® Fc Imaging System (LI-COR Biosciences, Lincoln, Nebraska, 34 patients from Southern France: identification of 10 new mutations. Absence of USA) and quantified with ImageJ, version 1.45 (National Institutes of genotype-phenotype correlation. Neuromuscul. Disord. 17, 235-241. doi:10.1016/ Health, Bethesda, MD, USA; http://imagej.nih.gov/ij). j.nmd.2006.12.014 Auld, D. S., Thorne, N., Maguire, W. F. and Inglese, J. (2009). Mechanism of PTC124 activity in cell-based luciferase assays of nonsense codon suppression. Flow-cytometric analysis of stable HEK293T cell clones treated Proc. Natl. Acad. Sci. USA 106, 3585-3590. doi:10.1073/pnas.0813345106 with RTA Bartram, C., Edwards, R. H., Clague, J. and Beynon, R. J. (1993). McArdle’s Transfected and non-transfected cells were seeded in a flat-bottom 96-well disease: a nonsense mutation in exon 1 of the muscle glycogen phosphorylase plate (8000 cells/well) in the presence of Bsd. After 24 h, treatment was gene explains some but not all cases. Hum. Mol. Genet. 2, 1291-1293. doi:10. 1093/hmg/2.8.1291 started using the RTAs described in Table 1. Cells were analyzed by flow Bartram, C., Edwards, R. H., Clague, J. and Beynon, R. J. (1994). McArdle’s cytometry 48 h after treatment initiation. Treated cells were labeled with the disease: a rare frameshift mutation in exon 1 of the muscle glycogen viability marker 7-aminoactinomycin D (7-AAD). Flow cytometry was phosphorylase gene. Biochim. Biophys. Acta 1226, 341-343. doi:10.1016/0925- performed with BD LSRFortessaTM (BD Biosciences) and results were 4439(94)90047-7 analyzed with FCS Express 4 Software (De Novo Software, Pasadena, CA, Bellais, S., Le Goff, C., Dagoneau, N., Munnich, A. and Cormier-Daire, V. (2010). USA). Experiments were performed in duplicate for all conditions. In vitro readthrough of termination codons by gentamycin in the Stü ve- Wiedemann Syndrome. Eur. J. Hum. Genet. 18, 130-132. doi:10.1038/ejhg. 2009.122 Literature search Bidou, L., Allamand, V., Rousset, J.-P. and Namy, O. (2012). Sense from In order to establish a consensus sequence for positive and negative read- nonsense: therapies for premature stop codon diseases. Trends Mol. Med. 18, through induction, a number of gene sequences with PTCs and treated with 679-688. doi:10.1016/j.molmed.2012.09.008 RTAs were analyzed and separated according to positive or negative Birch, K. E., Quinlivan, R. M. and Morris, G. E. (2013). Cell models for McArdle disease and aminoglycoside-induced read-through of a premature termination outcome (Du et al., 2013, 2009b, 2006; Kayali et al., 2012; Howard et al., codon. Neuromuscul. Disord. 23, 43-51. doi:10.1016/j.nmd.2012.06.348 1996; Floquet et al., 2011; Tan et al., 2011; Sánchez-Alcudia et al., 2012; Brasell, E. J., Chu, L., El Kares, R., Seo, J. H., Loesch, R., Iglesias, D. M. and Schwarz et al., 2015; Martorell et al., 2019; Dranchak et al., 2011; Bukowy- Goodyer, P. (2019). The aminoglycoside geneticin permits translational Bieryllo et al., 2016; Linde et al., 2007; Harada et al., 2018; Lincoln et al., readthrough of the CTNS W138X nonsense mutation in fibroblasts from 2018). All sequences corresponded to human gene sequences with the patients with nephropathic cystinosis. Pediatr. Nephrol. 34, 873-881. doi:10. exception of the mouse Cftr G542X (Du et al., 2009b, 2006) and Dmd 1007/s00467-018-4094-0 Bruno, C., Cassandrini, D., Martinuzzi, A., Toscano, A., Moggio, M., Morandi, Q995X (Kayali et al., 2012) gene sequences. For each gene, the surrounding L., Servidei, S., Mongini, T., Angelini, C., Musumeci, O. et al. (2006). McArdle PTC sequence including 15 nucleotides upstream and downstream from the disease: the mutation spectrum of PYGM in a large Italian cohort. Hum. Mutat. 27, reported PTC (−15 to +18; being +1 the first nucleotide of the PTC) was 718. doi:10.1002/humu.9434 obtained using the consensus coding sequence (CCDS) project database Bukowy-Bieryllo, Z., Dabrowski, M., Witt, M. and Zietkiewicz, E. (2016). (https://www.ncbi.nlm.nih.gov/CCDS/). Sequences were aligned with Aminoglycoside-stimulated readthrough of premature termination codons in Multiple Alignment using Fast Fourier Transform (MAFFT) (https:// selected genes involved in primary ciliary dyskinesia. RNA Biol. 13, 1041-1050. doi:10.1080/15476286.2016.1219832 www.ebi.ac.uk/Tools/msa/mafft/; output ClustalW). Aligned sequences Disease Models & Mechanisms Chowdhury, H. M., Siddiqui, M. A., Kanneganti, S., Sharmin, N., Chowdhury, were reformatted using MView (https://www.ebi.ac.uk/Tools/msa/mview/; M. W. and Nasim, M. T. (2018). Aminoglycoside-mediated promotion of parameters: width 80, coloring identity, color map D2 and group map none). translation readthrough occurs through a non-stochastic mechanism that Additionally, manual annotation was performed; nucleotides that were competes with translation termination. Hum. Mol. Genet. 27, 373-384. doi:10. represented ≥35% and ≥30% among the aligned sequences were marked in 1093/hmg/ddx409 blue and purple, respectively. Finally, to obtain the consensus sequences, Dabrowski, M., Bukowy-Bieryllo, Z. and Zietkiewicz, E. (2015). Translational readthrough potential of natural termination codons in eucaryotes–The impact of the combination of the WebLogo program (https://weblogo.berkeley.edu/ RNA sequence. RNA Biol. 12, 950-958. doi:10.1080/15476286.2015.1068497 logo.cgi) along with manual annotation was used; thus, for each position the de Luna, N., Brull, A., Guiu, J. M., Lucia, A., Martin, M. A., Arenas, J., Marti, R., consensus nucleotide was defined as a single nucleotide with a frequency Andreu, A. L. and Pinos, T. (2015). Sodium valproate increases the brain isoform ≥50%, or when the combination of two nucleotides generates a frequency of glycogen phosphorylase: looking for a compensation mechanism in McArdle ≥70% (only when each nucleotide had a frequency ≥30%). disease using a mouse primary skeletal-muscle culture in vitro. Dis. Model. Mech. 8, 467-472. doi:10.1242/dmm.020230 Deschauer, M., Morgenroth, A., Joshi, P. R., Glaser, D., Chinnery, P. F., Aasly, Acknowledgements J., Schreiber, H., Knape, M., Zierz, S. and Vorgerd, M. (2007). Analysis of We thank Drs Glenn Morris, Rosaline Quinlivan and Kathryn Birch for kindly spectrum and frequencies of mutations in McArdle disease. identification of 13 providing the WT and p.R50X GFP-PYGM plasmids. novel mutations. J. Neurol. 254, 797-802. doi:10.1007/s00415-006-0447-x Dimauro, S., Andreu, A. L., Bruno, C. and Hadjigeorgiou, G. M. (2002). Competing interests Myophosphorylase deficiency (glycogenosis type V; McArdle disease). Curr. Mol. The authors declare no competing or financial interests. Med. 2, 189-196. doi:10.2174/1566524024605770 10 RESEARCH ARTICLE Disease Models & Mechanisms (2020) 12, dmm043281. doi:10.1242/dmm.043281 Dranchak, P. K., Di Pietro, E., Snowden, A., Oesch, N., Braverman, N. E., Lincoln, V., Cogan, J., Hou, Y., Hirsch, M., Hao, M., Alexeev, V., De Luca, M., De Steinberg, S. J. and Hacia, J. G. (2011). Nonsense suppressor therapies rescue Rosa, L., Bauer, J. W., Woodley, D. T. et al. (2018). Gentamicin induces LAMB3 peroxisome lipid metabolism and assembly in cells from patients with specific nonsense mutation readthrough and restores functional laminin 332 in junctional PEX gene mutations. J. Cell. Biochem. 112, 1250-1258. doi:10.1002/jcb.22979 epidermolysis bullosa. Proc. Natl. Acad. Sci. USA 115, E6536-E6E45. doi:10. Du, M., Keeling, K. M., Fan, L., Liu, X., Kovaçs, T., Sorscher, E. and Bedwell, 1073/pnas.1803154115 D. M. (2006). Clinical doses of amikacin provide more effective suppression of the Linde, L., Boelz, S., Nissim-Rafinia, M., Oren, Y. S., Wilschanski, M., Yaacov, Y., human CFTR-G542X stop mutation than gentamicin in a transgenic CF mouse Virgilis, D., Neu-Yilik, G., Kulozik, A. E., Kerem, E. et al. (2007). Nonsense- model. J. Mol. Med. 84, 573-582. doi:10.1007/s00109-006-0045-5 mediated mRNA decay affects nonsense transcript levels and governs response Du, M., Liu, X., Welch, E. M., Hirawat, S., Peltz, S. W. and Bedwell, D. M. (2008). of cystic fibrosis patients to gentamicin. J. Clin. Invest. 117, 683-692. doi:10.1172/ PTC124 is an orally bioavailable compound that promotes suppression of the JCI28523 human CFTR-G542X nonsense allele in a CF mouse model. Proc. Natl. Acad. Lucia, A., Ruiz, J. R., Santalla, A., Nogales-Gadea, G., Rubio, J. C., Garcia- Sci. USA 105, 2064-2069. doi:10.1073/pnas.0711795105 Consuegra, I., Cabello, A., Pé rez, M., Teijeira, S., Vieitez, I. et al. (2012). Du, L., Damoiseaux, R., Nahas, S., Gao, K., Hu, H., Pollard, J. M., Goldstine, J., Genotypic and phenotypic features of McArdle disease: insights from the Spanish Jung, M. E., Henning, S. M., Bertoni, C. et al. (2009a). Nonaminoglycoside national registry. J. Neurol. Neurosurg. Psychiatry 83, 322-328. doi:10.1136/jnnp- compounds induce readthrough of nonsense mutations. J. Exp. Med. 206, 2011-301593 2285-2297. doi:10.1084/jem.20081940 Martin, M. A., Rubio, J. C., Wevers, R. A., Van Engelen, B. G., Steenbergen, Du, M., Keeling, K. M., Fan, L., Liu, X. and Bedwell, D. M. (2009b). Poly-L-aspartic G. C., Van Diggelen, O. P., Visser, M. D., Die-Smulders, C. D., Blazquez, A., acid enhances and prolongs gentamicin-mediated suppression of the CFTR- Andreu, A. L. et al. (2004). Molecular analysis of myophosphorylase deficiency in G542X mutation in a cystic fibrosis mouse model. J. Biol. Chem. 284, 6885-6892. Dutch patients with McArdle’s disease. Ann. Hum. Genet. 68, 17-22. doi:10.1046/ doi:10.1074/jbc.M806728200 j.1529-8817.2003.00067.x Du, L., Jung, M. E., Damoiseaux, R., Completo, G., Fike, F., Ku, J. M., Nahas, S., Martorell, L., Cortina, V., Parra, R., Barquinero, J. and Vidal, F. (2019). Variable Piao, C., Hu, H. and Gatti, R. A. (2013). A new series of small molecular weight readthrough responsiveness of nonsense mutations in hemophilia A. compounds induce read through of all three types of nonsense mutations in the Haematologica haematol.2018.212118. doi:10.3324/haematol.2018.212118 ATM gene. Mol. Ther. 21, 1653-1660. doi:10.1038/mt.2013.150 Mc, A. B. (1951). Myopathy due to a defect in muscle glycogen breakdown. Clin. Sci. El-Schahawi, M., Tsujino, S., Shanske, S. and Dimauro, S. (1996). Diagnosis of 10, 13-35. McArdle’s disease by molecular genetic analysis of blood. Neurology 47, Mcelroy, S. P., Nomura, T., Torrie, L. S., Warbrick, E., Gartner, U., Wood, G. and 579-580. doi:10.1212/WNL.47.2.579 Mclean, W. H. I. (2013). A lack of premature termination codon read-through Floquet, C., Deforges, J., Rousset, J. P. and Bidou, L. (2011). Rescue of non- efficacy of PTC124 (Ataluren) in a diverse array of reporter assays. PLoS Biol. 11, sense mutated p53 tumor suppressor gene by aminoglycosides. Nucleic Acids e1001593. doi:10.1371/journal.pbio.1001593 Res. 39, 3350-3362. doi:10.1093/nar/gkq1277 Mcnamara, E. L., Taylor, R. L., Clayton, J. S., Goullee, H., Dilworth, K. L., Pinos, Friesen, W. J., Johnson, B., Sierra, J., Zhuo, J., Vazirani, P., Xue, X., Tomizawa, Y., T., Brull, A., Alexander, I. E., Lisowski, L., Ravenscroft, G. et al. (2019). Baiazitov, R., Morrill, C., Ren, H. et al. (2018). The minor gentamicin complex Systemic AAV8-mediated delivery of a functional copy of muscle glycogen component, X2, is a potent premature stop codon readthrough molecule with phosphorylase (Pygm) ameliorates disease in a murine model of McArdle therapeutic potential. PLoS ONE 13, e0206158. doi:10.1371/journal.pone.0206158 disease. Hum. Mol. Genet. ddz214. doi:10.1093/hmg/ddz214 Gó mez-Grau, M., Garrido, E., Cozar, M., Rodriguez-Sureda, V., Domı́nguez, C., Mendell, J. T. and Dietz, H. C. (2001). When the message goes awry: disease- Arenas, C., Gatti, R. A., Cormand, B., Grinberg, D. and Vilageliu, L. (2015). producing mutations that influence mRNA content and performance. Cell 107, Evaluation of Aminoglycoside and Non-Aminoglycoside Compounds for Stop- 411-414. doi:10.1016/S0092-8674(01)00583-9 Codon Readthrough Therapy in Four Lysosomal Storage Diseases. PLoS ONE Mort, M., Ivanov, D., Cooper, D. N. and Chuzhanova, N. A. (2008). A meta- 10, e0135873. doi:10.1371/journal.pone.0135873 analysis of nonsense mutations causing human genetic disease. Hum. Mutat. 29, Gonzalez-Hilarion, S., Beghyn, T., Jia, J., Debreuck, N., Berte, G., Mamchaoui, 1037-1047. doi:10.1002/humu.20763 K., Mouly, V., Gruenert, D. C., Dé prez, B. and Lejeune, F. (2012). Rescue of Nakamura, K., Du, L., Tunuguntla, R., Fike, F., Cavalieri, S., Morio, T., Mizutani, nonsense mutations by amlexanox in human cells. Orphanet J. Rare Dis. 7, 58. S., Brusco, A. and Gatti, R. A. (2012). Functional characterization and targeted doi:10.1186/1750-1172-7-58 correction of ATM mutations identified in Japanese patients with ataxia- Gurgel-Giannetti, J., Nogales-Gadea, G., van der Linden, H., Jr, Bellard, T. M., telangiectasia. Hum. Mutat. 33, 198-208. doi:10.1002/humu.21632 Brasileiro Filho, G., Giannetti, A., de Castro Concentino, E. L. and Vainzof, Namy, O., Hatin, I. and Rousset, J. P. (2001). Impact of the six nucleotides M. V. (2013). Clinical and molecular characterization of McArdle’s disease in downstream of the stop codon on translation termination. EMBO Rep. 2, 787-793. Brazilian patients. Neuromolecular Med. 15, 470-475. doi:10.1007/s12017-013- doi:10.1093/embo-reports/kve176 8233-2 Nogales-Gadea, G., Arenas, J. and Andreu, A. L. (2007). Molecular genetics of Harada, N., Hatakeyama, A., Okuyama, M., Miyatake, Y., Nakagawa, T., Kuroda, M., McArdle’s disease. Curr. Neurol. Neurosci. Rep. 7, 84-92. doi:10.1007/s11910- Masumoto, S., Tsutsumi, R., Nakaya, Y. and Sakaue, H. (2018). Readthrough of 007-0026-2 ACTN3 577X nonsense mutation produces full-length α-actinin-3 protein. Biochem. Nogales-Gadea, G., Brull, A., Santalla, A., Andreu, A. L., Arenas, J., Martin, Biophys. Res. Commun. 502, 422-428. doi:10.1016/j.bbrc.2018.05.193 M. A., Lucia, A., De Luna, N. and Pinó s, T. (2015). McArdle Disease: Update of Heier, C. R. and DiDonato, C. J. (2009). Translational readthrough by the Reported Mutations and Polymorphisms in the PYGM Gene. Hum. Mutat. 36, aminoglycoside geneticin (G418) modulates SMN stability in vitro and improves 669-678. doi:10.1002/humu.22806 motor function in SMA mice in vivo. Hum. Mol. Genet. 18, 1310-1322. doi:10. Pé rez-Ortı́n, J. E., Alepuz, P., Chá vez, S. and Choder, M. (2013). Eukaryotic 1093/hmg/ddp030 mRNA decay: methodologies, pathways, and links to other stages of gene Howard, M., Frizzell, R. A. and Bedwell, D. M. (1996). Aminoglycoside antibiotics expression. J. Mol. Biol. 425, 3750-3775. doi:10.1016/j.jmb.2013.02.029 restore CFTR function by overcoming premature stop mutations. Nat. Med. 2, Quinlivan, R., Buckley, J., James, M., Twist, A., Ball, S., Duno, M., Vissing, J., 467-469. doi:10.1038/nm0496-467 Bruno, C., Cassandrini, D., Roberts, M. et al. (2010). McArdle disease: a clinical Howard, M. T., Anderson, C. B., Fass, U., Khatri, S., Gesteland, R. F., Atkins, review. J. Neurol. Neurosurg. Psychiatry 81, 1182-1188. doi:10.1136/jnnp.2009. Disease Models & Mechanisms J. F. and Flanigan, K. M. (2004). Readthrough of dystrophin stop codon 195040 mutations induced by aminoglycosides. Ann. Neurol. 55, 422-426. doi:10.1002/ Rubio, J. C., Lucia, A., Ferná ndez-Cadenas, I., Cabello, A., Blá zquez, A., ana.20052 Gá mez, J., Andreu, A. L., Martı́n, M. A. and Arenas, J. (2006). Novel mutation in Howell, J. M., Walker, K. R., Davies, L., Dunton, E., Everaardt, A., Laing, N. and the PYGM gene resulting in McArdle disease. Arch. Neurol. 63, 1782-1784. Karpati, G. (2008). Adenovirus and adeno-associated virus-mediated delivery of doi:10.1001/archneur.63.12.1782 human myophosphorylase cDNA and LacZ cDNA to muscle in the ovine model of Rubio, J. C., Garcia-Consuegra, I., Nogales-Gadea, G., Blazquez, A., Cabello, McArdle’s disease: expression and re-expression of glycogen phosphorylase. A., Lucia, A., andreu, A. L., Arenas, J. and Martin, M. A. (2007). A proposed Neuromuscul. Disord. 18, 248-258. doi:10.1016/j.nmd.2007.10.006 molecular diagnostic flowchart for myophosphorylase deficiency (McArdle Howell, J. M., Dunton, E., Creed, K. E., Quinlivan, R. and Sewry, C. (2015). disease) in blood samples from Spanish patients. Hum. Mutat. 28, 203-204. Investigating sodium valproate as a treatment for McArdle disease in sheep. doi:10.1002/humu.9474 Neuromuscul. Disord. 25, 111-119. doi:10.1016/j.nmd.2014.10.002 Salvatori, F., Breveglieri, G., Zuccato, C., Finotti, A., Bianchi, N., Borgatti, M., Kayali, R., Ku, J.-M., Khitrov, G., Jung, M. E., Prikhodko, O. and Bertoni, C. Feriotto, G., Destro, F., Canella, A., Brognara, E. et al. (2009). Production of (2012). Read-through compound 13 restores dystrophin expression and improves beta-globin and adult hemoglobin following G418 treatment of erythroid precursor muscle function in the mdx mouse model for Duchenne muscular dystrophy. Hum. cells from homozygous β(0)39 thalassemia patients. Am. J. Hematol. 84, 720-728. Mol. Genet. 21, 4007-4020. doi:10.1093/hmg/dds223 doi:10.1002/ajh.21539 Lentini, L., Melfi, R., Di Leonardo, A., Spinello, A., Barone, G., Pace, A., Sá nchez-Alcudia, R., Pé rez, B., Ugarte, M. and Desviat, L. R. (2012). Feasibility Palumbo Piccionello, A. and Pibiri, I. (2014). Toward a rationale for the PTC124 of nonsense mutation readthrough as a novel therapeutical approach in propionic (Ataluren) promoted readthrough of premature stop codons: a computational acidemia. Hum. Mutat. 33, 973-980. doi:10.1002/humu.22047 approach and GFP-reporter cell-based assay. Mol. Pharm. 11, 653-664. doi:10. Santalla, A., Nogales-Gadea, G., Encinar, A. B., Vieitez, I., Gonzá lez-Quintana, 1021/mp400230s A., Serrano-Lorenzo, P., Consuegra, I. G., Asensio, S., Ballester-Lopez, A., 11 RESEARCH ARTICLE Disease Models & Mechanisms (2020) 12, dmm043281. doi:10.1242/dmm.043281 Pintos-Morell, G. et al. (2017). Genotypic and phenotypic features of all Spanish Tsujino, S., Shanske, S., Nonaka, I. and Dimauro, S. (1995). The molecular patients with McArdle disease: a 2016 update. BMC Genomics 18, 819. doi:10. genetic basis of myophosphorylase deficiency (McArdle’s disease). Muscle 1186/s12864-017-4188-2 Nerve Suppl. 3, S23-S27. doi:10.1002/mus.880181407 Schroers, A., Kley, R. A., Stachon, A., Horvath, R., Lochmuller, H., Zange, J. Vissing, J., Duno, M., Schwartz, M. and Haller, R. G. (2009). Splice mutations and Vorgerd, M. (2006). Gentamicin treatment in McArdle disease: failure to preserve myophosphorylase activity that ameliorates the phenotype in McArdle correct myophosphorylase deficiency. Neurology 66, 285-286. doi:10.1212/01. disease. Brain 132, 1545-1552. doi:10.1093/brain/awp065 wnl.0000194212.31318.fc Welch, E. M., Barton, E. R., Zhuo, J., Tomizawa, Y., Friesen, W. J., Trifillis, P., Schwarz, N., Carr, A.-J., Lane, A., Moeller, F., Chen, L. L., Aguilà, M., Nommiste, Paushkin, S., Patel, M., Trotta, C. R., Hwang, S. et al. (2007). PTC124 targets B., Muthiah, M. N., Kanuga, N., Wolfrum, U. et al. (2015). Translational read- genetic disorders caused by nonsense mutations. Nature 447, 87-91. doi:10. through of the RP2 Arg120stop mutation in patient iPSC-derived retinal pigment 1038/nature05756 epithelium cells. Hum. Mol. Genet. 24, 972-986. doi:10.1093/hmg/ddu509 Wilschanski, M., Miller, L. L., Shoseyov, D., Blau, H., Rivlin, J., Aviram, M., Tan, L., Narayan, S. B., Chen, J., Meyers, G. D. and Bennett, M. J. (2011). Cohen, M., Armoni, S., Yaakov, Y., Pugatch, T. et al. (2011). Chronic ataluren PTC124 improves readthrough and increases enzymatic activity of the CPT1A (PTC124) treatment of nonsense mutation cystic fibrosis. Eur. Respir. J. 38, R160X nonsense mutation. J. Inherit. Metab. Dis. 34, 443-447. doi:10.1007/ 59-69. doi:10.1183/09031936.00120910 s10545-010-9265-5 Yu, H., Liu, X., Huang, J., Zhang, Y., Hu, R. and Pu, J. (2014). Comparison of read- Tsujino, S., Shanske, S. and Dimauro, S. (1993). Molecular genetic heterogeneity through effects of aminoglycosides and PTC124 on rescuing nonsense mutations of myophosphorylase deficiency (McArdle’s disease). N Engl. J. Med. 329, of HERG gene associated with long QT syndrome. Int. J. Mol. Med. 33, 729-735. 241-245. doi:10.1056/NEJM199307223290404 doi:10.3892/ijmm.2013.1601 Disease Models & Mechanisms 12