Objective: To investigate the effect of DEAD-box helicase (DDX) 21 on myocardial ischemia-reperfusion (I/R) injury and its potential mechanisms. Methods: In vivo, adult male Bama pigs and C57BL/6J mice were used to establish a myocardial I/R injury model by ligating the left anterior descending coronary artery, with sham-operated groups set as controls. The expression of DDX21 in myocardium after I/R injury was assessed by quantitative real-time PCR (qRT-PCR), Western blot, and immunofluorescence staining. Following the establishment of the myocardial I/R injury model in mice, AAV9 vectors with cardiac-specific expression were injected in situ into the peri-infarct region (The I/R+DDX21 group, I/R+negative control (NC) group, I/R+sh-NC group and I/R+sh-DDX21 group were injected with AAV9:cTnT-DDX21, AAV9:cTnT-NC, AAV9:cTnT-sh-NC and AAV9:cTnT-sh-DDX21, respectively). Additionally, the I/R+A-485 group received intraperitoneal injections of the cAMP response element-binding protein (CREB) binding protein inhibitor A-485, while the I/R+PBS group was injected with an equivalent volume of phosphate-buffered saline (PBS) as the control. Echocardiography was performed on postoperative days 1 and 28 to evaluate cardiac function (left ventricular ejection fraction and fractional shortening). At 28 days post-surgery, mice were euthanized and heart tissues were harvested for histological sectioning. Myocardial fibrosis was evaluated using Masson's trichrome staining. In vitro, primary cardiomyocytes were isolated from neonatal day 1 C57BL/6J mice using enzymatic digestion method. Cardiomyocytes were transfected with plasmids or small interfering RNA (siRNA). The cardiomyocytes transfected with DDX21-siRNA were assigned to the siDDX21 group, those transfected with the DDX21 plasmid were assigned to the DDX21 group, and those transfected with the corresponding empty plasmid or siRNA were assigned to the NC group. Additionally, cardiomyocytes were treated with A-485 (A-485 group) or PBS (PBS group). An oxygen-glucose deprivation/reoxygenation (OGD/R) model was used to simulate cellular injury. Transcriptome sequencing was performed to identify downstream mechanisms of DDX21. Differential gene expression analysis was conducted using software such as DESeq2, and alternative splicing events in the mRNA transcriptome were analyzed using rMATS software. Mitochondrial superoxide, mitochondrial membrane potential, ATP content, and mitochondrial respiratory chain complex enzyme activity in cardiomyocytes were detected using immunofluorescence staining and commercial assay kits. The oxidative phosphorylation level of the cells was assessed by the Seahorse extracellular flux analyzer. Acetylated DDX21 levels were measured using co-immunoprecipitation and Western blot assays. Results: The expression levels of DDX21 in myocardium from the Bama pigs and mice in the I/R injury model were significantly higher than those in the sham group (all P<0.001). Echocardiographic results showed that at 28 days post-surgery, compared to the I/R+NC group, the I/R+DDX21 group exhibited higher left ventricular ejection fraction and fractional shortening, while the I/R+sh-DDX21 group showed lower values; Masson staining results demonstrated that, compared to the I/R+NC group, the myocardial fibrosis area in the I/R+DDX21 group was significantly reduced, whereas it was significantly increased in the I/R+sh-DDX21 group (all P<0.001). Transcriptomic sequencing results suggested that DDX21 may influence myocardial injury by regulating mitochondrial metabolic activity. In vitro, compared to the OGD/R+NC group, the OGD/R+DDX21 group exhibited lower mitochondrial superoxide levels, higher polymer/monomer ratio, maximal oxygen consumption, reserve capacity, and ATP content. In contrast, the OGD/R+siDDX21 group showed the opposite results, with reduced activity of mitochondrial respiratory chain complex V (all P<0.05). Mechanistically, rMATS software and other analyses indicated that knockdown of DDX21 affected the alternative 3' splicing sites of ATP5J precursor mRNA, inhibiting the splicing of certain exonic sequences. Overexpression of DDX21 upregulated both mRNA and protein levels of ATP5J. Co-immunoprecipitation experiments showed that, compared to the PBS group, acetylated DDX21 levels were reduced in the A-485 group. Further in vivo experiments showed that, compared to the I/R+PBS group, the I/R+A-485 group exhibited higher left ventricular ejection fraction and fractional shortening, and a lower proportion of left ventricular fibrosis (all P<0.001). Conclusions: DDX21 improves cardiomyocyte energy metabolism and alleviates I/R injury by regulating the alternative splicing of ATP5J. A-485 holds potential as a novel small molecule candidate for the treatment of myocardial injury.