Inhibition of cyclooxygenase-1 by nonsteroidal anti-inflammatory drugs demethylates MeR2 enhancer and promotes Mbnl1 transcription in myogenic cells

Muscleblind-like 1 (MBNL1) is a ubiquitously expressed RNA-binding protein, which is highly expressed in skeletal muscle. Abnormally expanded CUG-repeats in the DMPK gene cause myotonic dystrophy type 1 (DM1) by sequestration of MBNL1 to nuclear RNA foci and by upregulation of another RNA-binding protein, CUG-binding protein 1 (CUGBP1). We previously reported that a nonsteroidal anti-inflammatory drug (NSAID), phenylbutazone, upregulates MBNL1 expression in DM1 mouse model by demethylation of MeR2, an enhancer element in Mbnl1 intron 1. NSAIDs inhibit cyclooxygenase (COX), which is comprised of COX-1 and COX-2 isoforms. In this study, we screened 29 NSAIDs in C2C12 myoblasts, and found that 13 NSAIDs enhanced Mbnl1 expression, where COX-1-selective NSAIDs upregulated Mbnl1 more than COX-2-selective NSAIDs. Consistently, knockdown of COX-1, but not of COX-2, upregulated MBNL1 expression in C2C12 myoblasts and myotubes, as well as in myotubes differentiated from DM1 patient-derived induced pluripotent stem cells (iPSCs). Luciferase assay showed that COX-1-knockdown augmented the MeR2 enhancer activity. Furthermore, bisulfite sequencing analysis demonstrated that COX-1-knockdown suppressed methylation of MeR2. These results suggest that COX-1 inhibition upregulates Mbnl1 transcription through demethylation of the MeR2 enhancer. Taken together, our study provides new insights into the transcriptional regulation of Mbnl1 by the COX-1-mediated pathway.

viral vector rescues disease-associated muscle hyperexcitability or myotonia in a transgenic mouse model for DM1 17,18 . Similarly, overexpression of human MBNL1 alleviated muscle degeneration in a fly model of DM1 19 . In Drosophila, Mbnl, the homolog of human MBNL1, harbors two enhancers located in intron 2, which drive specific gene expression in embryonic somatic muscle and in the central nervous system 20 . Enhancers homologous to the Drosophila enhancers in mammalian MBNL1 were not known. We previously identified an enhancer for Mbnl1 transcription, designated MeR2 (the methylated region 2), in intron 1 of mouse Mbnl1 21 . Our analysis revealed that demethylation of MeR2 enhances Mbnl1 expression in C2C12 myoblasts as well as in mouse skeletal muscle 21 .
Nonsteroidal anti-inflammatory drugs (NSAIDs) are widely used for inflammatory diseases including neurodegenerative and neuromuscular diseases [22][23][24] . NSAIDs inhibit cyclooxygenase (COX) that catalyzes the conversion of arachidonic acid to prostanoids. There are two COX isoforms: COX-1 is constitutively expressed in most tissues, and COX-2 is inducibly expressed in response to inflammation [25][26][27] . Selective inhibitors for each COX isoform have been developed, although most classical NSAIDs inhibit both COX isoforms 28 . We 21 and others 29 reported beneficial effects of NSAIDs on DM1. Screening of 400 preapproved drugs using a Drosophila model for DM1 disclosed 10 candidate drugs including two NSAIDs (ketoprofen and indomethacin) 29 . We previously reported that phenylbutazone, a non-selective NSAID, ameliorates muscle weakness and muscle pathology by enhancement of MBNL1 expression, as well as by inhibition of the interaction between MBNL1 and CUG RNA 21 . Phenylbutazone augments transcription of Mbnl1 mRNA by demethylating the MeR2 enhancer.
In the current study, we screened 29 NSAIDs for upregulation of Mbnl1 in C2C12 mouse myoblasts, and found that COX-1-selective NSAIDs were more effective than COX-2-selective NSAIDs. Knockdown of COX-1, but not of COX-2, enhanced demethylation of the MeR2 enhancer and increased Mbnl1 transcription. We show that Mbnl1 transcription is suppressed by COX-1-mediated pathway, and propose that COX-1-selective NSAIDs are potential drugs for DM1 patients.

Results
Drug screening for 29 NSAIDs to elevate Mbnl1 mRNA expression. We previously reported that phenylbutazone, a non-selective NSAID, enhances MBNL1 expression in myogenic cells 21 . We here screened 29 NSAIDs to examine a class effect of NSAIDs on enhancement of Mbnl1 expression. C2C12 myoblasts were incubated for 24 h with 100 μM each NSAID, followed by quantification of Mbnl1 mRNA using real-time RT-PCR. Our screening identified 13 drugs that increased Mbnl1 mRNA, in which phenylbutazone had the largest effect (Fig. 1). The second-ranked drug, ketoprofen, was previously reported to suppress CUG-induced lethality in Drosophila 20 . We also confirmed that the fourth-ranked drug, naproxen, elevated mRNA and protein levels of MBNL1 in C2C12 cells ( Supplementary Fig. S1b,c). Four out of six COX-1-selective NSAIDs and two out of seven COX-2-selective NSAIDs increased Mbnl1 mRNA (Fig. 1). Wilcoxon rank sum test showed that COX-1selective NSAIDs increased Mbnl1 mRNA more than COX-2-selective NSAIDs (p = 0.022). Additionally, two COX-2-selective NSAIDs, but no COX-1-selective NSAID, downregulated Mbnl1 mRNA. We also examined the effect of each NSAID at a lower concentration (10 μM) on Mbnl1 expression in C2C12 cells. We observed that the 6 top-ranked NSAIDs at 100 µM that upregulated Mbnl1 mRNA more than 1.4-folds were again the 6 top-ranked NSAIDs at 10 μM (red letters in Supplementary Fig. S1a). In addition, at 10 µM, a statistically significant increase of Mbnl1 mRNA was observed in two COX-1-selective NSAIDs, but in no COX-2-selective NSAID (red letters in Supplementary Fig. S1a). In contrast, at 10 µM, a statistically significant decrease of Mbnl1 mRNA was observed in no COX-1-selective NSAID, but in two COX-2-selective NSAIDs (blue letters in Supplementary Fig. S1a). Taken together, our analysis suggests that a substantial number of NSAIDs can increase MBNL1 expression, in which COX-1-mediated pathway, rather than COX-2-mediated pathway, plays a pivotal role. Mean and SD (n = 3 culture dishes) are indicated with individual values in red dots. *p < 0.05, **p < 0.01, and ***p < 0.001 compared to DMSO alone by Student's t-test with Bonferroni multiple comparison correction. dissect the regulation of MBNL1 expression through COX-1-mediated and COX-2-mediated pathways, we knocked down COX-1 and COX-2, and examined Mbnl1 mRNA levels and MBNL1 protein levels in C2C12 cells. We first confirmed that our siRNAs efficiently knocked down COX-1 and COX-2 ( Fig. 2a-d). We found that knockdown of COX-1 with two independent siRNAs (siCOX1-a and -b) upregulated expression of Mbnl1 mRNA approximately 1.3-folds while knockdown of COX-2 downregulated Mbnl1 mRNA in C2C12 myoblasts. In addition, double knockdown of COX-1 and COX-2 increased Mbnl1 mRNA in these cells (Fig. 2e,f). Western blotting analysis also showed elevation of MBNL1 protein expression in COX-1-knocked down cells (Fig. 2g,h). Since the expression level of MBNL1 is upregulated in differentiated myogenic cells 5,30,31 , we further analyzed the Mbnl1 mRNA and MBNL1 protein levels in differentiated C2C12 myotubes. Similar to the results of undifferentiated C2C12 myoblasts, knockdown of COX-1 increased Mbnl1 mRNA and MBNL1 protein in differentiated C2C12 myotubes (Fig. 2e-h).
We next examined the effect of knockdown of COX-1 and COX-2 on MBNL1 expression in DM1 patient-derived induced pluripotent stem cells (MyoD-DM1-iPSCs), which were differentiated into myotubes by the induction of MyoD expression (Fig. 3a,b) 32,33 . Similar to C2C12 myoblasts and myotubes (Fig. 2e,f), knockdown of COX-1 significantly upregulated expression of MBNL1 mRNA, while knockdown of COX-2 downregulated MBNL1 mRNA in myotubes differentiated from MyoD-DM1-iPSCs. Double knockdown of COX-1 and COX-2 increased MBNL1 mRNA in these cells (Fig. 3c,d). The upregulation of Mbnl1 mRNA by COX-1-knockdown was also observed in primary mouse myoblasts and KD3 human myoblast cells (Supplementary Figs. S2 and S3). Taken together, our analysis suggests that MBNL1 expression is suppressed through the COX-1-mediated pathway in myogenic cells.
The MeR2 enhancer in Mbnl1 is suppressed through the COX-1-mediated pathway. We previously identified a transcription enhancer, named MeR2, in Mbnl1 intron 1, and reported that the demethylation of the MeR2 enhancer upregulated Mbnl1 expression (Fig. 4a) 21 . To examine whether NSAIDs upregulate the activity of the MeR2 enhancer, we inserted the MeR2 enhancer upstream of the SV40 promoter and the firefly luciferase cDNA to make pGL3P-MeR2 (Fig. 4b). C2C12 myoblasts were introduced with pGL3P-MeR2 or pGL3P, and were treated with one of the 29 NSAIDs at 100 μM for 24 h. Luciferase assay revealed that 10 out of the 29 NSAIDs significantly increased the luciferase activity of pGL3P-MeR2 (Fig. 4c), while no drugs affected that of pGL3P (Fig. 4e). Furthermore, fold inductions of Mbnl1 mRNA expression by NSAIDs ( Fig. 1) were correlated well with those of luciferase activity of pGL3P-MeR2 (Pearson's correlation coefficient, r = 0.81, p < 0.001, Fig. 4d), but not those of pGL3P vector (Pearson's correlation coefficient, r = 0.10, p = 0.61, Fig. 4f). We also examined whether the COX-1-mediated pathway is specifically involved in the activity of the MeR2 enhancer. As predicted, knockdown of COX-1, but not of COX-2, upregulated the luciferase activity of pGL3P-MeR2 (Fig. 4g,h). Overall, our results suggest that the MeR2 enhancer is suppressed through the COX-1-mediated pathway.
To investigate the role of the MeR2 enhancer more directly, we deleted the genomic MeR2 region in C2C12 myoblasts by CRISPR/Cas9 system using two single-guide RNAs (sgRNAs) targeting the upstream and downstream sites of MeR2, respectively (Fig. 5a). We obtained one MeR2-knockout C2C12 cell line (MeR2-KO C2C12), where MeR2 was deleted in both alleles (Fig. 5b). We confirmed that knockdown of COX-1 with two independent siRNAs (siCOX1-a and -b) had no effect on Mbnl1 mRNA expression in MeR2-KO C2C12, but upregulated Mbnl1 mRNA expression in the control cell line (WT C2C12), which retained MeR2 (Fig. 5c). Thus, COX-1 requires the MeR2 enhancer to suppress Mbnl1 mRNA expression.
The MeR2 enhancer in Mbnl1 is methylated through COX-1-mediated pathway. We further analyzed COX-1-dependent regulation of MeR2-methylation in C2C12 myoblasts using bisulfite sequencing. We transfected C2C12 cells with siRNAs against COX-1 or COX-2 on myogenic differentiation day 0, and performed bisulfite sequencing analysis of MeR2 on differentiation day 3. Our analysis revealed that knockdown of COX-1 ( Fig. 6), but not of COX-2 ( Supplementary Fig. S4), significantly suppressed methylation of MeR2 in C2C12 cells. These results suggest that COX-1 inhibition suppresses methylation of MeR2 and enhances Mbnl1 transcription in C2C12 myoblasts.
The involvements of COX pathway in the regulation of DNA methylation have been previously reported. Treatment with celecoxib, a COX-2 specific inhibitor, suppressed DNA methylation in the promoter region of ESR1 encoding the estrogen receptor α in rat colon tumor cells 34 . Similarly, aspirin, a non-selective NSAID, lowered DNA methylation in the promoter region of CDH1 encoding cadherin 1 in human gastric mucosa cells 35 . Overexpression of prostaglandin E2 (PGE2), one of the products of COX enzyme, elevated expression of DNA methyltransferase 3α (DNMT3a), which is necessary for de novo DNA methylation 36 , and increased DNA methylation in the promotor regions of 13 genes in fetal lung fibroblast cells 37 . In addition, overexpression of COX-2 in hepatocytes downregulated expression of tet methylcytosine dioxygenase 1 (TET1), and induced The effect of knockdown of COX-1 and COX-2 on Mbnl1 mRNA and MBNL1 protein expression in C2C12 cells. Western blotting (a,b,g,h) and real-time RT-PCR analysis (c-f) were performed using undifferentiated C2C12 cells or C2C12 cells differentiated for 5 days. The cells were treated with siRNA against COX-1 (siCOX1-a or siCOX1-b), COX-2 (siCOX2-a or siCOX2-b), both COX-1 and COX-2 (siCOX1a + 2a or siCOX1b + 2b), or control siRNA (siCont). Undifferentiated C2C12 cells were examined 2 days (mRNA levels) or 3 days (protein levels) after the transfection. Differentiated C2C12 cells were examined on day 5 of differentiation. Expression levels of COX-1, COX-2, Mbnl1 mRNA and MBNL1 protein are normalized to those of Gapdh and GAPDH, respectively, and also to siCont-treated cells. Mean and SD (n = 3 and 6 culture dishes for real-time RT-PCR and Western blotting, respectively) are indicated with individual values in red dots. *p < 0.05, **p < 0.01 and ***p < 0.001 by Student's t-test with Bonferroni multiple comparison correction.  www.nature.com/scientificreports www.nature.com/scientificreports/ promoter hypermethylation of 68 genes 38 . TET1 is a member of ten-eleven translocation (TET) family proteins that promote DNA demethylation through oxidation of 5-methylcytosine 39 . We observed that knockdown of COX-1, but not of COX-2, upregulated mRNA expressions of all three members of TET (Tet1, Tet2, and Tet3) (Fig. 7). We consistently observed that inhibition of COX-1, but not of COX-2, demethylated MeR2 ( Fig. 6 and Supplementary Fig. S4). In contrast, knockdown of neither COX-1 nor COX-2 affected Dnmt3a mRNA expression (Fig. 7). These results suggest a role of TET in the COX-1-dependent suppression of MeR2-methylation in C2C12 cells, although additional molecules may also be involved in the suppression.
The promoter/enhancer-dependent transcription of Mbnl1 in human and mouse remains largely unknown. In contrast, in Drosophila, two enhancers located in Mbnl intron 2 specifically drive Mbnl expression in embryonic somatic muscle and the nerve cord. These enhancers are enriched in consensus binding sites for Mef2 transcription factor 20 , which is necessary for the expression of MBNL in Drosophila 7 . Although the sequences of the Drosophila Mbnl enhancers are not conserved in mammals, we found that MeR2 has three Mef2 binding sites according to JASPAR (http://jaspar.genereg.net/). As DNA methylation interferes with bindings of transcription factors to enhancers and promoters 40,41 , methylation of MeR2 may inhibit the binding of Mef2 to suppress transcription of Mbnl1.
MBNL1 regulates a large subset of alternative splicing transition 8,42 . Genes involved in skeletal muscle development and function, including Bin1, Clcn1, Ldb3, Serca1, and Tnnt3, are aberrantly spliced in Mbnl1-knockout mice as well as in DM1 patients, in which MBNL1 is sequestrated to the nuclear RNA foci with abnormally expanded CUG-repeats [43][44][45] . These missplicing events lead to muscle weakness 45 and myotonia 8 in DM1 skeletal muscle, where splicing isoforms are shifted from adult forms to immature forms 46,47 . A previous study showed that 2-fold upregulation of MBNL1 protein by transduction of a recombinant adeno-associated viral vector reversed missplicing of Clcn1, Ldb3, Serca1, and Tnnt3, resulting in a significant reduction in myotonia in a DM1 mouse model 17 . We also reported that ~1.5-folds upregulation of MBNL1 protein by phenylbutazone, a non-selective NSAID, reversed missplicing of Clcn1, Nfix, and Rpn2, and improved muscle weakness in a DM1 mouse model 21 . These results suggest that upregulation of MBNL1 effectively ameliorates aberrant splicing in DM1 muscle cells, in which the MBNL1-availability is severely impaired. In the current study, we found that COX-inhibitions increased MBNL1 ~1.5-folds ( Fig. 2 and Supplementary Fig. S1c), which is likely to sufficient to exert some beneficial effects in DM1 cells. In contrast, in normal C2C12 cells, we observed that knockdown of COX-1 did not obviously affect the splicing of Clcn1 as well as expressions of myogenic regulatory factors, MyoD and Myogenin ( Supplementary Fig. S5), suggesting that upregulation of MBNL1 by NSAIDs is unlikely to be toxic in normal cells, in which MBNL1 is abundantly available.
Enhancement of MBNL1 expression is one of the attractive therapeutic modalities for DM1 [17][18][19] . Our study shows that nearly a half of the screened NSAIDs enhanced Mbnl1 expression in myogenic cells, suggesting possible application of NSAIDs for DM1 patients. Indeed, phenylbutazone, a non-selective COX inhibitor, ameliorates muscle weakness and muscle pathology in DM1 mouse model 21 . The beneficial effects of NSAIDs on skeletal muscles are not limited to DM1. Chronic administration of NSAIDs, such as ibuprofen and acetaminophen, increases muscle volume and muscle strength in healthy aged adults 48 . Furthermore, several NSAIDs have been reported to prevent muscle damage and preserve the number and function of muscle cells in a mouse model of Duchenne muscular dystrophy 49,50 . NSAIDs have adverse effects on the gastrointestinal tract, the kidneys, and the cardiovascular system 51 . In addition, a recent study revealed that the long-term use of NSAIDs has an influence on the diversity of fecal microbiota 52 . These issues should be taken into account in clinical trials in the future. Although further analyses are required, we expect that COX-1-selective NSAIDs and possibly their derivatives will become one of therapeutic options for DM1 patients.

Methods
Classification of NSAIDs. NSAIDs were classified based on the relative inhibitory activities against COX-1 and COX-2. According to a previous report 53 , NSAIDs with the IC 50 ratio (COX-2 IC 50 /COX-1 IC 50 ) > 5 were classified as COX-1-selective inhibitors, and those with the IC 50 ratio < 0.2 were classified as COX-2-selective inhibitors. Although the IC 50 ratios of celecoxib and meloxicam were 0.70 and 0.37, respectively, these drugs were classified as COX-2-selective inhibitors according to the IC 80 ratios 53 . In addition, celecoxib and meloxicam have been deemed as COX-2 inhibitors in clinical and biochemical studies 54,55 . We thus classified celecoxib and meloxicam as COX-2-selective inhibitors. The other NSAIDs lacking the IC 50 ratio 53 were classified as follows: tenoxicam, acemetacin and oxprozin were classified as COX-1-selective NSAIDs according to the previous reports [56][57][58] . Acetaminophen was classified as a COX-2-selective NSAID 59 , and the remaining NSAIDs were classified as non-selective NSAIDs [60][61][62][63][64] . The classification of the NSAIDs is summarized in Table 1  C2C12 cell culture and drug screening. C2C12 myoblasts were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% (v/v) fetal bovine serum (FBS) and incubated at 37 °C with 5% CO 2 . When the cells were grown to 50-60% confluency, 10 μM or 100 μM of each NSAID (Prestwick Chemical Library) was added. After 24 h, cells were harvested and total RNA was extracted as described below.

Scientific RepoRtS |
(2020) 10:2558 | https://doi.org/10.1038/s41598-020-59517-y www.nature.com/scientificreports www.nature.com/scientificreports/ To examine the effect of naproxen (02190247, MP Biomedicals) on MBNL1 expression in C2C12 myotubes, C2C12 myoblasts were grown to 95% confluency in DMEM with 10% FBS, and then the medium was changed to DMEM with 2% horse serum to induce myotube differentiation. Naproxen was dissolved in dimethyl sulfoxide (DMSO) to make 100 and 400 mM solutions separately, and then added to the culture medium at final concentrations of 100 and 400 μM at the initiation of differentiation. DMSO (0.1%) was also added to the medium of control cells. After 5 days, gene expression of Mbnl1 mRNA was quantified using real-time RT-PCR, and the amount of MBNL1 protein was estimated by Western blotting. All samples were analyzed in triplicates.   -1 (siCOX1-a or siCOX1-b), COX-2 (siCOX2-a or siCOX2-b), both COX-1 and COX-2 (siCOX1a + 2a or siCOX1b + 2b), or control siRNA (siCont). Undifferentiated C2C12 cells were examined on 2 days after the transfection. Differentiated C2C12 cells were examined on differentiation day 5. Expression levels of Tet1, Tet2, Tet3, and Dnmt3a mRNA are normalized to that of Gapdh and also to siCont-treated cells. Mean and SD (n = 3 culture dishes) are indicated with individual values in red dots. *p < 0.05, **p < 0.01 and ***p < 0.001 compared to siCont by Student's t-test with Bonferroni multiple comparison correction.
Primary myoblast cell and KD3 myoblast cell culture. All experiments using mice were approved by the Animal Care and Use Committee of the Nagoya University and were performed in accordance with the relevant guidelines. The pectoralis major, gastrocnemius, and quadriceps were removed from ten neonatal mice at 9 days of age. Skeletal muscles were minced in sterile PBS under the microscope with sterile forceps. Following addition of an appropriate amount of trypsin, the minced muscles were incubated at 37 °C for 2 h to so that the muscles were fully digested to a slurry. Then two volumes of growth medium [DMEM with 20% HS, 10% FBS, 2% penicillin, 1% glutamine and 20 U/ml interferon γ (PeproTech)] were added, and the slurry was filtered through 440-and 70-μm nylon mesh (Corning Costar) to remove large pieces of tissue. After centrifugation of the slurry at 300 × g for 10 min, the pellet was suspended in the growth medium and centrifuged again at 8 × g for 1 min to precipitate blood cells. Then, the supernatant was filtered through 35-μm nylon mesh and the cells were plated in 2 ml of the growth medium in a 60-mm collagen-coated culture dish.

Splicing analysis and gene expression analysis. Total RNA was extracted by RNeasy Mini Kit (Qiagen)
according to the manufacturer's instructions, and reverse-transcribed to cDNA using random hexamer primers (ThermoFisher) and ReverTraAce (Toyobo). PCR amplifications were performed using GoTaq (Promega) for 30 cycles. The primer sequences used for RT-PCR are shown in Supplementary Table S2. The intensities of RT-PCR-amplified spliced products were quantified with the ImageJ program (http://imagej.nih.gov/ij/). We then estimated the ratio of exon inclusion by dividing the signal intensity of the upper band by the sum of signal intensities of two bands. All samples were analyzed in triplicates.
Real-time RT-PCR was performed with the LightCycler 480 (Roche Applied Science) using the TB Green Premix ExTaq II (Takara Bio). Gene expression levels were normalized by that of glyceraldehyde-3-phosphate dehydrogenase (Gapdh). PCR primers are shown in Supplementary Table S2. All real-time RT-PCR experiments were performed in triplicate. All samples were analyzed in triplicates.
Western blot analysis. Cells were washed with cold PBS twice, and were lysed with PLC buffer [50 mM HEPES pH 7.0, 150 mM NaCl, 10% (vol/vol) glycerol, 1% (vol/vol) TritonX-100, 1.5 mM MgCl 2 , 1 mM EGTA, 100 mM NaF, 10 mM NaPPi (sodium pyrophosphate)] with protease inhibitors (1 μg/μl aprotinin, 1 μg/μl leupeptin, and 1 mM PMSF). Then, the lysates were gently sonicated twice for 1 min using Handy Sonic (Tomy Seiko) at intensity levels 3 to 4, and were cleared with centrifugation for 15 min at 15,000 × g at 4 °C. The total protein concentrations of the lysates were measured using Pierce 660 nm Protein Assay Reagent. Cell lysates were boiled for 5 min in 2 × Laemmli buffer, separated on a 7.5% or 10% SDS-polyacrylamide gel, and transferred to a polyvinylidene fluoride membrane (Immobilon-P, Millipore). Then, the membrane was blocked for 1 h at room temperature in TBS-T with 5% skim milk. After washing with Tris-buffered saline containing 0.05% Tween 20 (TBS-T), the membrane was incubated overnight at 4 °C with primary antibodies listed in Supplementary Table S3. Next, the membrane was washed with TBS-T and incubated with secondary goat anti-mouse IgG (1:3000, LNA931V/ AG, GE Healthcare) or anti-rabbit IgG (1:3000, LNA934V/AE, GE Healthcare) antibody conjugated to horseradish peroxidase (HRP) for 1 h at room temperature. The bound antibodies were detected with ECL Western www.nature.com/scientificreports www.nature.com/scientificreports/ blotting detection reagents (GE Healthcare), and the signal intensities were quantified with the ImageJ program (http://imagej.nih.gov/ij/). All samples were analyzed at least in triplicates.
For knocking down of COX-1 or COX-2 in undifferentiated C2C12 myoblasts, primary myoblast cells and undifferentiated KD3 myoblasts, siRNAs were introduced when cells were grown to 50-60% confluency. For knocking down of COX-1 or COX-2 in differentiated C2C12 myotubes and differentiated KD3 myotubes, cells were first grown to 95% confluency and then medium was changed to differentiation medium and the cells were transfected with siRNAs. For knocking down of COX-1 or COX-2 in MyoD-DM1-iPSCs, siRNAs were introduced at 24 h after replating (Fig. 3a). After incubation for 2 additional days, new medium was added and siRNAs were introduced again. Total RNA or protein lysates were extracted at the indicated time points. mRNA expression was quantified by real-time RT-PCR and protein expression was analyzed by Western blotting as stated above.
C2C12 cells were transfected with sgRNA-1, sgRNA-2, and Cas9 enzyme (Integrated DNA Technologies) using Lipofectamine RNAiMax (ThermoFisher) according to the manufacturer's recommendations. Two days after transfection, cells were seeded at 0.5 cells/well and 0.2 cells/well in 96-well plates. After 10 days, all single cell clones were screened with PCR amplification of the genomic segment harboring MeR2 region, using a pair of primers, 5′-ATGGCTTCAGGGTATTCAGCC-3′ and 5′-GGCGAGGGATGAACAAAAGC-3′. The expected size of PCR amplicons of MeR2-KO clones was around 165 bp, while that of the MeR2-retained wild-type clones was around 1900 bp. Following the PCR amplification, the amplicons were cloned into pGEM-T vector (Promega) and sequenced to determine the deleted segment of the MeR2 (Fig. 4b). We obtained one single cell clone, in which MeR2 was deleted in both alleles (MeR2-KO C2C12). Another single cell clone retaining MeR2 on both alleles was used as a WT control (WT C2C12).
Transfection and luciferase assays. To make the luciferase vector harboring the MeR2 enhancer (pGL3P-MeR2), the mouse genomic region from positions 60,330,708 to 60,331,205 on chromosome 3 according to NC_000069.6 (GRCm37/mm9) was PCR-amplified with the following primers; 5′-CATGGTACCCAGTAATCTGAGTCC TGCTGTAGTAA-3′ and 5′-GATGCTAGCGAAGACATCTACCTATGCTAAAAGCA-3′, where KpnI and NheI restriction sites are underlined, respectively. Then the amplified DNA fragment was cloned into the pGL3-Promotor Vector (pGL3P, Promega) at KpnI and NheI restriction sites. The pGL3P-MeR2 and pGL3P were introduced into XL-10 gold competent cells (Agilent) and were propagated. To measure the transcriptional activity of MeR2 enhancer, C2C12 myoblasts were transiently transfected with pGL3P-MeR2 or pGL3P using Lipofectamine 3000 (ThermoFisher) according to the manufacturer's recommendations, and were incubated for 24 h. Then, 100 μM of each NSAID was added to the culture medium and cells were cultured for an additional 24 h. Alternatively, siRNAs were introduced into cells with Lipofectamin RNAiMax (ThermoFisher), and cells were cultured for an additional 48 h. The luciferase assay was performed using the Dual Luciferase Reporter Assay System kit (Promega). To normalize for the transfection efficiency and the cell number, pRL-CMV Renilla Luciferase Reporter Vector was cotransfected, and the firefly luciferase activities were normalized to the Renilla luciferase activity. All samples were analyzed at least in triplicates.