Human cardiomyocytes (hCMs) (BFN608007121, Bluefbio Biotechnology Development Co., Ltd., Shanghai, China) were plated on gelatin-coated plates (V900863, Sigma, Darmstadt, Germany), and then cultured in D10 medium: DMEM medium (C22400500BT, Gibco, Shanghai, China) supplemented with 10% fetal bovine serum (fsp500, Excell BIO, Shanghai, China), 1$\times$ Nonessential Amino Acid (N1250, Solarbio, Beijing, China), and 1$\times$ Penicillin-Streptomycin Solution (P1400, Solarbio). The culture medium was changed every two days. Cells were passaged until they reached 80–90% confluence and were then digested in a trypsin-EDTA solution (T1320, Solarbio) at a ratio of 1:2–1:3. Doxycycline hyclate (Dox, 2 µg/mL) (D9891, Sigma, St. Louis, MO, USA) was used to induce myogenin expression, and an equal volume of dimethyl sulfoxide (DMSO) was added to the culture medium of controls. To induce apoptosis, cells were exposed to 1 mM H${}_2$O${}_2$ for 4 hours. In the knockdown assay, cells were exposed to 500 µM H${}_2$O${}_2$ for 4 hours. All cell lines were validated by STR profiling and tested negative for mycoplasma. Cells were all cultured in a humidified incubator at 37 °C and 5% CO${}_2$.

The plasmid pCW-Cas9-Blast (83481, Addgene, Watertown, MA, USA) was modified by removing Cas9 cDNA and replacing the blasticidin S resistance gene with a puromycin resistance gene. Then human MYOG cDNA and an enhanced green fluorescent protein (EGFP) gene, which was linked by a T2A sequence, were subcloned into the plasmid. Lentivirus was prepared using a third-generation lentivirus packaging system as described elsewhere [18]. After 24 hours of transduction, the medium was replaced with fresh D10 medium supplemented with Dox (to induce myogenin expression) or DMSO (controls). After 2 days of induction, EGFP expression was observed by fluorescence microscopy, and the EGFP-positive cells were sorted by flow cytometry (FACSAria III, BD Biosciences, San Jose, CA, USA) and cultured further in D10 medium. After 5–7 days of incubation, single clones were picked at random for further expansion, to be used in the following experiments.

Cells were fixed with 4% paraformaldehyde at room temperature for 20 minutes and washed three times with 1$\times$PBS, then permeabilized with 1$\times$PBS containing 0.25% Triton X-100 at room temperature for 10 minutes. After incubating in the blocking buffer (1$\times$PBS with 10% goat serum) at 37 °C for 1 hour, cells were incubated with recombinant anti-myogenin antibody (ab124800, Abcam, Cambridge, MA, USA; 1:250 dilution) at 4 °C overnight. After washing three times with 1$\times$PBS containing 0.1% Triton X-100, cells were incubated with goat anti-rabbit IgG (H+L) cross-adsorbed secondary antibody (A-21428, Invitrogen, Eugene, OR, USA; 1:500 dilution) at 37 °C for 1 hour. Nuclei were labeled with Hoechst 33342 (C1022, Beyotime, Shanghai, China; final concentration: 5 µg/mL) in the dark for 5 minutes. Images were obtained using an inverted fluorescence microscope (MF52-N, Mshot, Guangzhou, China).

Total RNA was isolated from cells using the EZ-10 Total RNA Mini-Preps Kit (B618583-0100, Sangon Biotech, Shanghai, China) followed by treatment with RNase-Free DNase (B618253, Sangon Biotech). cDNA was synthesized using the Hifair®Ⅲ 1st Strand cDNA Synthesis Kit (11139ES60, YEASEN, Shanghai, China). Quantitative real-time PCR was performed using a Light Cycler® 480 (Roche) with TB Green® Fast qPCR Mix (RR430B, Takara, Beijing, China). Data were analyzed by the relative standard curve method and normalized to GAPDH. Quantitative PCR primers and their sequences, designed by Primer3 [20], are listed below (from 5${}^{\prime }$ to 3${}^{\prime }$):

MYOG-RT-F: GCCCAAGGTGGAGATCCT;

MYOG-RT-R: GGTCAGCCGTGAGCAGAT;

FOXA3-RT-F: TGGGCTCAGTGAAGATGGAG;

FOXA3-RT-R: GGGGATAGGGAGAGCTTAGAG;

ATOH8-RT-F: GAGATCAAAGCCCTGCAGC;

ATOH8-RT-R: TCGGCACTGTAGTCAAGGTC;

DUSP13-RT-F: CTTCCCCAACCGAGGCTT;

DUSP13-RT-R: CAGGGAGTAAGGACCAGCTC;

GAPDH-RT-F: CGCTCTCTGCTCCTCCTGTT;

GAPDH-RT-R: CCATGGTGTCTGAGCGATGT.

Apoptosis was detected using APC annexin V (640920, Biolegend, San Diego, CA, USA). Cells were seeded into 6-well plates with 5 $\times$ 10${}^5$ cells per well. After treatment, cells were digested by a trypsin-EDTA solution (T1320, Solarbio) and then washed twice with cold cell-staining buffer (420201, Biolegend). Cells were then resuspended in 100 µL annexin V binding buffer (422201, Biolegend) containing 5 µL APC annexin V, and incubated at room temperature for 15 minutes. After adding 400 µL of annexin V binding buffer to each tube, samples were analyzed on the NovoCyte D2040R (Agilent, Beijing, China).

Cells were washed twice with ice-cold 1$\times$PBS, then lysed using RIPA buffer (R0010, Solarbio, Beijing, China) supplemented with 1$\times$phosphatase inhibitor cocktail (B15001, Bimake, Houston, TX, USA) and 1$\times$protease inhibitor cocktail (B14001, Bimake, Houston, TX, USA). Lysates were scraped from the plates, then centrifuged at 12,000 $\times$g at 4 °C for 15 minutes. Protein concentration was determined using a modified Bradford protein assay kit (C503041, Sangon, Shanghai, China). Equal amounts of protein were separated by SDS-PAGE and transferred onto polyvinylidene fluoride membranes (ISEQ00010, Merck-Millipore, Darmstadt, Germany). The membranes were blocked with 5% nonfat milk at room temperature for 1 hour. Next, membranes were incubated with an appropriate dilution of primary antibodies at 4 °C overnight, followed by incubation with secondary antibodies. Protein expression levels were measured using the StarSignal Plus Chemiluminescent Assay Kit (E170, GeneStar, Beijing, China), and bands were observed using a gel detection system (Tanon 5200, Shanghai, China). The following antibodies were used [dilution, species, catalog number, company, city, country]: DUSP13-M1 (N-term) [1:1000, rabbit, AP8455a, Abcepta, Suzhou, China], GAPDH [1:2500, rabbit, AF1186, Beyotime], p38 MAPK (D13E1) [1:1000, rabbit, 8690T, CST, Danvers, MA, USA], Phospho-p38 MAPK (Thr180/Tyr182) (D3F9) [1:1000, rabbit, 4511T, CST, Danvers, MA, USA], Beta Actin [1:50,000, mouse, 66009-1, Proteintech, Rosemont, IL, USA], and HRP-conjugated affinipure goat anti-rabbit IgG (H+L) [1:5000, goat, SA00001-2, Proteintech, Rosemont, IL, USA].

Potential binding sites of myogenin in the DUSP13 promoter region (–2500 to 0 bp) were predicted by the JASPAR database [21]. Fragments with different lengths of DUSP13 promoter were amplified from the genomic DNA of hCMs via PCR. The mutant (Mut) sequence of the DUSP13 promoter was synthesized by Sangon Biotech. These DNA fragments were then subcloned into the pGL6-TA luciferase reporter vector (D2105, Beyotime). MYOG cDNA was subcloned into the expression vector pCMV-N-Myc-Puro (D2791, Beyotime). HEK293T cells were transfected with luciferase reporter vectors, Renilla pRLTK control reporter vector (D2760, Beyotime), and MYOG expression vector using the EZ-Trans transfection reagent (AC04L071, Life-iLab, Shanghai, China). After 48 hours of transfection, Dual-Lumi™ Luciferase Reporter Gene Assay Kit (RG089S, Beyotime) was used to detect luciferase activity. Each experiment was performed in triplicate. Results are shown as firefly luciferase activity normalized to Renilla luciferase activity. The PCR primers are listed below (from 5${}^{\prime }$ to 3${}^{\prime }$):

DUSP13-HindIII-R: GTAAGCTTGTCATTTCTCCTTCCCAG;

DUSP13-500-XhoI-F: TACTCGAGCCTGACCTCCGTAATTCAGAG;

DUSP13-1000-XhoI-F: AACTCGAGCGTCTCCCCATTTGACAT;

DUSP13-1350XhoI-F: AACTCGAGCGACTTCAATGGTTCTGA;

DUSP13-1500-XhoI-F: AACTCGAGCTGTATTTTCCTCACTCC;

DUSP13-2500-XhoI-F: AACTCGAGAGTTCGACAACCTCCATTTG.

ChIP assay was performed using the Pierce™ Magnetic ChIP Kit (26157, Thermo Scientific, Eugene, OR, USA). Cells were first treated with 2 µg/mL Dox for 72 hours to induce myogenin expression, then cross-linked with 1% formaldehyde for 10 minutes at room temperature and incubated with a final concentration of 0.2 M glycine for 5 minutes to terminate cross-linking. After washing with PBS, cells were collected and resuspended in Membrane Extraction Buffer on ice for 10 minutes. The cell lysates were resuspended in MNase Digestion Buffer Working Solution supplemented with MNase (10 U/µL) for chromatin digestion (37 °C, 15 minutes), and the digestion was terminated by adding MNase Stop Solution for 5 minutes on ice. Lysates were sonicated on ice using 3 W power for three 20 s-pulses (JY92-IIN, SCIENTZ, Ningbo, China). One-tenth of supernatant was used as Input. The remaining solution was divided into 3 fractions and incubated with antibodies overnight at 4 °C (Negative control IP: 1–2 µL of Rabbit IgG (included in the kit); Target-specific IP: 10 µg Myogenin Rabbit pAb (A17427, Abclonal, Wuhan, China). Magnetic beads (20 µL) were added to each IP sample and incubated at 4 °C for 2 hours. The beads were collected by a magnetic scaffold, washed three times with IP Wash Buffer I, once with IP Wash Buffer II, then mixed with IP Elution Buffer. Sodium chloride (5 M) and Proteinase K (20 mg/mL) were added to the supernatant to remove staining and protein cross-linking. DNA was purified using a DNA Clean-Up column. ChIP DNA was detected by qPCR using TB Green® Fast qPCR Mix (RR430A, TaKaRa). The primers specific for PBS3, PBS4, and PBS5 are as followed (from 5${}^{\prime }$ to 3${}^{\prime }$):

PBS3-F: TGTCAGACCCAGCATGTCCT;

PBS3-R: TGCAAGTCAGCTACCACCCT;

PBS4-F: CCTCTGCACCCTGGATCTCT;

PBS4-R: AAGGACATGCTGGGTCTGAC;

PBS5-F: GGTCTTGTCCACTCCAGAGGC;

PBS5-R: CCTCTGTGGTCACTCCTCTCT.

DUSP13A and DUSP13B cDNA were subcloned into the expression vector pcDNA3.1 (HG-VPI0001, HonorGene, Changsha, China). HCMs were seeded into 6-well plates with 2.5 $\times$ 10${}^5$ cells per well and cultured for 24 hours. The cells were transiently transfected with DUSP13 expression vectors or an empty vector using the TransIT®-LT1 transfection reagent (MIR2300, Mirus, Madison, WI, USA). After H${}_2$O${}_2$ treatment, cells were trypsinized, washed twice with cold cell-staining buffer (420201, Biolegend), and then stained with APC annexin V (640920, Biolegend, San Diego, CA, USA) and 7-AAD (420403, Biolegend). Samples were incubated with a staining reagent for 15 minutes in the dark at room temperature, then analyzed by NovoCyte D2040R (Agilent). The PCR primers are listed below (from 5${}^{\prime }$ to 3${}^{\prime }$):

DUSP13A-NheI-F: TATAGCTAGCACCATGGCTGAGACCTCTCTCC;

DUSP13A-BamhI-R: TATAGGATCCTCAGCTCTGGCCGGCAC;

DUSP13B-NheI-F: TAGCTAGCACCATGGACTCACTGCAGAAGCAG;

DUSP13B-BamhI-R: TATAGGATCCTCAGAACCGCCCCGTCTCC.

RNA-seq was performed by Novogene Co., Ltd., in three biological replicates for each group. 150-bp paired-end RNA-seq reads were obtained from the Illumina sequencing platform. First, adaptors and low-quality bases were trimmed by Trim Galore (https://www.bioinformatics.babraham.ac.uk/projects/trim_galore/). Then, gene expression was quantified by Kallisto [22] and the GENCODE database (v32). DESeq2 [23] was used to identify DEGs. The list of significant DEGs was defined at a false discovery rate (FDR) 0.05 and $|$log2(Fold Change)$|$ $>$1. Toppfun [24] was used to detect enriched gene ontology (GO) terms for each DEG list. Raw RNA-seq data have been submitted to GEO, and the accession number is pending.

According to the mRNA sequence of human DUSP13B gene (Accession number: NM_016364.3), we obtained three small interfering RNA (siRNA). The sequences are shown in Supplementary Table 1. All sequences were synthesized by Genomeditech Company (Shanghai, China) Co., Ltd. Knockdown efficiency was evaluated by western blotting and siRNA-1 was selected to perform the knockdown assay.

HCM${}^{\text{MYOG}}$ were seeded into 6-well plates with 2 $\times$ 10${}^5$ cells per well and cultured for 24 hours. The cells were transfected with siRNA-1 or negative control siRNA using the TransIT®-LT1 transfection reagent (MIR2300, Mirus, Madison, WI, USA). After 72 hours of siRNA transfection, the cells were treated with 500 µM H${}_2$O${}_2$ for 4 hours. After treatment, cells were trypsinized, washed twice with cold cell-staining buffer (420201, Biolegend), and then stained with APC annexin V (640920, Biolegend). Samples were incubated with a staining reagent for 15 minutes in the dark at room temperature, then analyzed by NovoCyte D2040R (Agilent).

Values are expressed as mean $\pm$ standard deviation. Statistical significance was calculated in SPSS Statistics (v.21, IBM, Armonk, NY, USA) using one-way ANOVA with Bonferroni correction, or Student’s t-test. p 0.05 was considered statistically significant.

To obtain stable cell lines expressing myogenin, an hCM cell line in which the expression of myogenin could be modulated by a Tet-On system (hCM${}^{\text{MYOG}}$) was generated by lentiviral expression. In this Tet-On system, Dox interacts with the reverse tetracycline-controlled transactivator and makes it bind to the TRE element to initiate the transcription of target genes. Hence, myogenin expression was dependent on supplementation with Dox in hCM${}^{\text{MYOG}}$. Real-time PCR and immunofluorescence both demonstrated markedly low expression of myogenin in uninduced hCM${}^{\text{MYOG}}$, and a 164.6-fold increase in expression after 2 days of Dox induction (Fig. 1b,c). Next, myogenin-induced apoptosis was examined in cardiomyocytes. Annexin V, an early apoptotic marker, was used to evaluate the apoptotic ratio in a flow cytometry assay. Neither a 24-hour nor a 48-hour Dox induction increased the apoptotic ratio compared with the control group (Fig. 1e). ROS-mediated damage was induced by H${}_2$O${}_2$, a commonly used reagent that induces cellular oxidative stress [25]. Apoptotic cells were detected at a rate of 7.4% in the cardiomyocytes exposed to H${}_2$O${}_2$ for 4 hours (Fig. 1d). Next, effects of myogenin overexpression on H${}_2$O${}_2$-induced apoptosis were tested. MYOG expression was induced for 0, 24, or 48 hours prior to 4-hour H${}_2$O${}_2$ treatments. Myogenin induction continued throughout the H${}_2$O${}_2$ treatments. All three experimental groups exhibited a significantly lower apoptotic ratio compared with the control group: 24-hour and 48-hour induction reduced the apoptotic ratio by 47.9% and 63.5%, respectively (Fig. 1d,f). These results suggest that the anti-apoptotic effect of myogenin in H${}_2$O${}_2$-treated cardiomyocytes is time-dependent.

Fig. 1.

Overexpression of myogenin decreased the rate of apoptosis after hydrogen peroxide (H${}_{\mathrm{2}}$O${}_{\mathrm{2}}$) treatment. (a) Schematic diagram illustrating the experimental design. After expansion from single clones, hCM${}^{\text{MYOG}}$ cells were cultured until green fluorescence disappeared, then myogenin expression was induced by Dox, or controls were supplemented with DMSO. (b) MYOG expression was significantly increased after a two-day induction (n = 3). (c) Representative immunofluorescence images of myogenin (red) in Dox- and DMSO-treated hCM${}^{\text{MYOG}}$. Cell nuclei were stained by 4${}^{\prime }$,6-diamidino-2-phenylindole (DAPI, blue). (d) Representative fluorescence-activated cell sorting (FACS) analysis of the apoptosis marker annexin V (conjugated with Allophycocyanin (APC)) in hCM${}^{\text{MYOG}}$ after various treatments. In the negative control group, DPBS buffer was used instead of the annexin V staining reagent. (e) The apoptotic ratio was unchanged after 24 or 48 hours of Dox induction. (f) The apoptotic ratio declined in myogenin-overexpressing hCM${}^{\text{MYOG}}$ after H${}_2$O${}_2$ treatment. The incubation times with Dox alone and in conjunction with H${}_2$O${}_2$-treatment are separated by a plus sign. **, p 0.001 compared to the no-induction control group. NS, no significant difference. Scale bar: 50 µm.

As a transcription factor, myogenin regulates the expression of a series of genes, including several key components of transcriptional machinery [26]. Some of the DEGs observed after long-term expression of myogenin may be due to its regulation of other transcription factors. Therefore it is necessary to monitor changes in gene expression profile during the early period of myogenin expression. To this end, hCM${}^{\text{MYOG}}$ were induced by Dox for 6, 12, 24, or 48 hours and compared to DMSO-treated controls (Fig. 1a). A subset of hCM${}^{\text{MYOG}}$ were also treated with H${}_2$O${}_2$ after a 48-hour Dox induction and compared to uninduced hCM${}^{\text{MYOG}}$ that were treated with H${}_2$O${}_2$. RNA-seq was used to obtain transcriptome data.

Only protein-coding genes with FDR 0.05 and $|$log2(fold change)$|$ $>$1 were considered. Under these criteria, 84 up-regulated and seven down-regulated genes were differentially expressed between 48-hour Dox induction and control groups (Fig. 2a). GO enrichment analysis of the up-regulated genes showed that “chemokine receptor binding” was one of the most significant molecular functions (Fig. 2b). The upregulated genes DUSP13 (25.99-fold increase), ATOH8 (5.94-fold increase), and FOXA3 (5.77-fold increase) were selected to validate the RNA-seq data using quantitative real-time PCR (Fig. 2d). Upregulation of DUSP13 was also verified at the protein level (Fig. 2e,f).

Fig. 2.

DUSP13 expression was elevated in myogenin-overexpressing hCM${}^{\text{𝐌𝐘𝐎𝐆}}$. (a) Volcano plot of differentially expressed genes between hCM${}^{\text{MYOG}}$ induced by Dox for 48 hours and control samples. Red and blue dots represent up- and down-regulated genes, respectively, in the 48 h Dox group. (b) Gene ontology (GO) enrichment analysis of the up-regulated genes in the 48 h Dox group (GO term: molecular function). (c) Heatmap of upregulated genes with continuously increased expression from 6- to 48-hour Dox induction. Genes are listed in descending order by values of log2 (fold change) from the 48 h Dox group. The rightmost column shows expression data for cells that were treated with Dox for 48 h and then with H${}_2$O${}_2$ for 4 h (as compared to cells that were treated with DMSO for 48 h, and then with H${}_2$O${}_2$ for 4 h). (d) Quantitative real-time PCR confirmed the mRNA expression levels of selected genes (FOXA3, ATOH8, and DUSP13) indicated by the RNA-seq data (n = 3). #, p 0.05; *, p 0.01; **, p 0.001. (e,f) Western blotting indicated that the protein level of DUSP13 was elevated in myogenin-overexpressing hCM${}^{\text{MYOG}}$. #, p 0.05; *, p 0.01; **, p 0.001, compared to DMSO-treated controls. (g,h) Western blotting indicated that the level of phosphorylated p38 MAPK was decreased significantly in myogenin-overexpressing hCM${}^{\text{MYOG}}$ under H${}_2$O${}_2$-treatment. #, p 0.05, compared to DMSO-treated controls.

Next, the DEGs that increased in expression from 6- to 48-hour induction were analyzed to identify candidate genes directly regulated by myogenin (Fig. 2c). Among these genes, ATOH8, ASCL2, and ID3 encoded transcription factors or transcriptional regulators, and the 6-hour induction only slightly increased their expression (1.42-fold, 1.13-fold, and 1.29-fold, respectively). By contrast, PHGR1, DUSP13, and VWA5B2, the top three up-regulated genes after 48-hour induction, exhibited about 2-fold increased expression after the 6-hour induction. There are few studies about the molecular functions of PHGR1 and VWA5B, while it has been reported that DUSP13 dephosphorylates JNK1 and p38$\alpha$ and decreases AP-1-dependent gene expression [27]. Considering that p38 phosphorylation is a key event in p38 MAPK-mediated apoptosis, it is possible that myogenin suppresses the p38 MAPK pathway by directly initiating DUSP13 expression. To evaluate this hypothesis, we first detected endogenous levels of phosphorylated p38 MAPK (p-p38) and p38 MAPK in hCM${}^{\text{MYOG}}$ with or without H${}_2$O${}_2$-treatment (Fig. 2g,h). The western blotting showed that hCM${}^{\text{MYOG}}$ with or without myogenin expression exhibited very low levels of p-p38 expression in the absence of H${}_2$O${}_2$, however, expression of myogenin decreased the phosphorylation level of p38 significantly after H${}_2$O${}_2$-treatment (for the ratio of p-p38/p38, Dox + H${}_2$O${}_2$ vs. DMSO + H${}_2$O${}_2$ is 0.846 vs. 0.568, p = 0.035). Next, we continued to test whether myogenin directly regulated the expression of DUSP13.

HEK 293T cells were used to examine whether myogenin bound to the promoter region of DUSP13. A 2500-bp sequence upstream of the 5${}^{\prime }$UTR of DUSP13 was analyzed in JASPAR database [21] to predict the binding sites of myogenin (Fig. 3). According to the distribution of putative binding sites, five DUSP13 promoter regions with different lengths (500, 1000, 1350, 1500, and 2500 bp) were cloned into a firefly luciferase reporter vector (PGL6). These vectors were named P500, P1000, P1350, P1500, and P2500, respectively. Then these firefly luciferase reporter vectors, Renilla luciferase reporter vectors, and MYOG expression vectors were co-transfected into HEK 293T cells. The dual-luciferase reporter assay showed that, compared to control cells transfected with the PGL6 vector, P1350 cells exhibited the strongest relative luciferase activity (40.5-fold). P500, P1000, and P1500 cells exhibited lower relative luciferase activity (20.2-fold, 20.3-fold, and 25.4-fold, respectively). Because P500 cells still showed strong relative luciferase activity, it was reasonable to speculate that the core binding sites of myogenin were located in the 500-bp sequence upstream of the 5${}^{\prime }$UTR of DUSP13. We then used multiple online tools, such as TFBIND, JASPAR, and Lasagna-Search 2.0, to predict potential binding sites in this 500-bp region. Five predicted binding sites (PBS1–5) were identified (Supplementary Table 2). Chromatin immunoprecipitation (ChIP) assay was performed to confirm myogenin binding to these PBSs. We observed substantial recruitment of myogenin to PBS3, PBS4, and PBS5 only upon Dox treatment (Fig. 3b, left panel). In the absence of myogenin, we observed no recruitment of myogenin to these PBSs (Fig. 3b, right panel). Negative results were obtained from ChIP assay for PBS1 and PBS2 (data not shown). These in vitro data demonstrated that myogenin could directly bind to the promoter region of DUSP13 and upregulate its expression.

Fig. 3.

Myogenin bound the DUSP13 promoter. (a) Different DUSP13 promoter regions (500, 1000, 1350, 1500, and 2500 bp) and a mutated promoter sequence (Mut p1350) were cloned into luciferase reporter vectors. Cells transfected with the PGL6 vector were used as the control group. Luciferase activity was measured 48 hours after transfection. p-values for the comparisons between each experimental group can be found in Supplementary Table 3. NS: no significant difference. (b) ChIP analysis of Dox-treated hCM${}^{\text{MYOG}}$ and DMSO-treated hCM${}^{\text{MYOG}}$, followed by qPCR analysis of myogenin occupancy at PBS3, PBS4, and PBS5. Each panel is representative of at least three independent experiments. #, p 0.05; *, p 0.01; NS, no significant difference.

To further demonstrate that myogenin inhibited H${}_2$O${}_2$-induced apoptosis by increasing the expression of DUSP13, cardiomyocytes were transiently transfected with DUSP13 expression vectors (including two isoforms of DUSP13: DUSP13A and DUSP13B) or an empty vector, then treated with H${}_2$O${}_2$ and analyzed by flow cytometry. Annexin V and 7-AAD were used to detect apoptotic and necrotic cells, respectively. Compared with empty-vector controls, cells expressing DUSP13A and DUSP13B had significantly lower rates of apoptosis and necrosis (Fig. 4). The rate of annexin V and 7-AAD double-positive cells in the DUSP13B group was reduced by 54.7% compared with the control group (8.6% vs. 19.0%, p 0.001), while the rate of double-positive cells in the DUSP13A group only decreased by 23.7% (14.5% vs. 19.0%, p 0.001; Fig. 4f), indicating that DUSP13B possessed stronger anti-apoptotic ability than DUSP13A. These results were consistent with a previous study showing that DUSP13B, not DUSP13A, had MAPK phosphatase activity in COS-7 and HEK 293 cells [27].

Fig. 4.

DUSP13 decreased the apoptotic and necrotic ratios of H${}_{\mathrm{2}}$O${}_{\mathrm{2}}$-treated cardiomyocytes. (a–e) Representative fluorescence-activated cell sorting analyses of the apoptotic marker annexin V (conjugated with APC) and necrotic marker 7-ADD in hCM${}^{\text{MYOG}}$ after various treatments. In the negative control group, DPBS buffer was used instead of the annexin V or 7-ADD staining reagents. Across groups, the proportions of (f) annexin V and 7-ADD double-positive cells, (g) annexin V positive cells, and (h) 7-ADD positive cells vary significantly. **, p 0.001.

We also knocked down DUSP13B by using siRNA in myogenin-expressing cardiomyocytes to confirm this conclusion. Three pairs of siRNA targeting DUSP13B gene were designed and synthesized. According to the knockdown efficiency evaluated by western blotting, siRNA-1 was selected to perform the knockdown assay. As shown in Supplementary Fig. 1, in the myogenin-expressing cardiomyocytes treated by H${}_2$O${}_2$, knockdown of DUSP13B significantly increased the apoptotic ratio compared with the controls (Supplementary Table 4), indicating that myogenin prevents H${}_2$O${}_2$-induced apoptosis through DUSP13B.

At the cellular level, the genes regulated by myogenin overexpression have not been studied systematically. It has been variously reported that overexpression of myogenin downregulates the expression of Pax7 [28], and induces the expression of Mef2c [29], MuRF1 [30], Atrogin-1 [30], MEGF10 [31], and Ncam1 [32]. Most of these genes were not identified in our RNA-seq data, possibly because (i) Many studies were performed in a mouse myoblast cell line (C2C12) that exhibits epigenetic characteristics distinct from human cardiomyocytes; (ii) The present study only considered the first 48 hours of myogenin expression and may have missed genes that require long-term induction. The expression profiles of myogenin-overexpressed human induced pluripotent stem cell-derived cardiomyocytes were described in our previous work [18], in which myogenin-overexpressed cardiomyocytes were treated with Ang II for 6 days. The RNA-seq data from these two studies did not completely agree. For example, DUSP13 did not appear in the DEGs of the original work [18], probably because the gene regulatory networks influenced by complex external factors (such as Ang II treatment and incubation time) eliminated any changes in its expression. These results demonstrate the necessity of time-course analysis of gene expression at the early stages of myogenin expression.

Based on the results from the dual-luciferase reporter assay, P1350 cells exhibited higher relative luciferase activity than P1000 cells, suggesting the existence of a certain motif in the –1350 bp to –1000 bp region that might enhance the transcriptional activity of myogenin. We speculated that this motif was the CAGCTG site in the –1244 to –1233 region. A mutated motif (CAGCTG to ACTGGC) was therefore generated, and its relative luciferase activity was found to be 46% lower when compared to unmutated P1350 cells, and was almost equal with P1000 cells (Fig. 3a). These data confirmed that the –1244 to –1233 region was a key motif that could increase the expression of a downstream gene (DUSP13). We further speculate that the reason that P1500 and P2500 cells had less activity than P1350 cells was due to the presence of repressor elements in the –1500 to –1350 bp or –2500 to –1500 bp regions. To identify these repressor elements, more delicate design experiments will be necessary to carry out in future studies. The lack of knowledge and the complexity of its promoter region are serious impediments to exploring the regulatory mechanisms of DUSP13 gene. For example, an online tool, Find Individual Motif Occurences (Version 5.5.4, FIMO, Seattle, WA, USA), was used to search the potential E-box motifs in the 2500-bp region, and in total 48 E-boxes were identified (Supplementary Table 5). What roles these motifs actually play is still undiscovered.

The p38 MAPK signaling pathway is involved in skeletal muscle development [33]. P38 directly phosphorylates MEF2A and MEF2C [34], and may induce MyoD through an indirect effect [33]. Although there is no evidence that p38 directly regulates myogenin, the p38 inhibitor SB203580 decreased the expression of myogenin and prevented myogenic differentiation [35]. The present study demonstrated that myogenin might reduce the phosphorylation of p38 by increasing expression of DUSP13, suggesting that there might be a feedback loop to regulate p38 MAPK signaling by myogenic regulatory factors. In addition, overexpression and knockdown assays illustrated the vital role of DUSP13 in the inhibition of apoptosis induced by H${}_2$O${}_2$. Considering the potential for therapeutic applications in heart diseases, DUSP13 might be a more appropriate candidate gene than MYOG. As a transcription factor, myogenin regulates many genes and would be more likely to have off-target effects. In contrast, DUSP13 has defined substrates, such as p38. Although a Japanese patent (JP5187875B2) claims that DUSP13 transgenic mice developed cardiac hypertrophy, which could be interpreted as a long-term effect of DUSP13 expression, we still propose that transient expression of DUSP13 might mitigate myocardial apoptosis induced by ROS in cardiac tissue.