Identification of FOXK2 mutations in pedigrees of congenital ptosis and associated myopathy.

(A) Pedigree illustrating isolated congenital ptosis with a heterozygous FOXK2 mutation (c.643 C > T:p.R215W). Photographs of affected individuals are included. Black-filled symbols denote affected members. ‘+’ indicates the wild-type FOXK2 allele, while ‘M’ represents the mutated allele. The black arrow points to the proband. Genotypes of some individuals remain unspecified due to the unavailability of genomic DNA. (B) H&E staining of eyelid muscle tissues from a control individual and the proband (II-5) in the pedigree described in (A). The zoomed area (scale bars: 50 µm) provides an enlarged view of the boxed region in the left image (scale bars: 100 µm). (C) IHC staining of MyHC in the eyelid muscle samples of a control individual and the proband (II-5) in the pedigree described in (A) (n = 3 area per group). The zoomed area (scale bars: 50 µm) offers an enlarged view of the boxed region in the left image (scale bars: 100 µm). Quantitative analysis was performed using ImageJ software, with results expressed by InDen/Area. P value: *P = 0.0448. (D) IHC staining of FOXK2 in the eyelid muscle samples of a control individual and the proband (II-5) in the pedigree described in (A). The zoomed area (scale bars: 50 µm) offers an enlarged view of the boxed region in the left image (scale bars: 100 µm). (E) Pedigrees depicting congenital myopathy associated with ptosis, carrying various heterozygous FOXK2 mutations (c.164 C > T:p.T55I; c.1231 C > G:p.P411A; c.1631 G > A:p.R544Q; c.1650 G > C:p.Q550H). ‘+’ denotes the wild-type FOXK2 allele, and ‘M’ represents the mutated allele. The black arrow identifies the proband. (F) Sanger sequencing diagrams of the FOXK2 mutation sites in healthy individuals and patients across five pedigrees. Green arrows in the upper panels indicate the mutation sites, highlighted in gray. (G) Diagram of the FOXK2 protein showing the locations of the mutations. (H) Species conservation analysis of FOXK2 amino acids p.T55, p.R215, p.P411, p.R544, and p.Q550. Data were analyzed by Student’s t test. All error bars represent mean ± standard deviation. H&E hematoxylin-eosin staining, FHA forkhead-associated domain, FOX forkhead DNA-binding domain, NLS nuclear localization sequence, IHC immunohistochemistry. Source data are available online for this figure.

foxk2 knockdown impaired skeletal muscle development of zebrafish.

(A) Structure of the zebrafish foxk2 gene. The zebrafish foxk2 gene was targeted using specific MOs to inhibit transcription at the start codon (ATG-MO) and proper splicing of exon 3 (I2E3-MO). (B) Quantitative of foxk2 expression level by qPCR. Larvae were collected at 72 hpf (n = 3 repeats). P value: ****P < 0.0001. (C) Percentage of larvae exhibiting defect or mortality. The control-MO group included 379 larvae, the ATG-MO group included 296 larvae, and the I2E3-MO group included 375 larvae. (D) Survival rate of larvae over a 72-hpf time course in control-MO, ATG-MO, and I2E3-MO groups. (E) Gross morphology of Tg(-1.9mylpfa:EGFP) larvae at 72-hpf, along with the quantification of curvature angle of larvae (n = 10 larvae for each group). The angle formed between the blue arrows indicates the degree of body curvature in the larvae. Scale bars: 500 µm. P value: ****P < 0.0001. (F) Body muscle morphology of Tg(-1.9mylpfa:EGFP) larvae at 72-hpf. Blue and yellow arrows denote normal and defective somites, respectively. The white asterisk indicates an abnormal trunk orientation. The inset in the upper right corner shows a magnified view of the boxed area (scale bars: 10 µm for the inset, 50 µm for the full image). (G) Cranial facial muscles of Tg(-1.9mylpfa:EGFP) larvae at 72-hpf. The inset in the upper right corner shows a magnified view of the boxed area (scale bars: 40 µm for the inset, 200 µm for the full image). (H–K) Statistical analysis of (H) movement distance (P value: ****P < 0.0001), (I) velocity (****P < 0.0001), (J) maximum acceleration (****P < 0.0001) and (K) mobility (percentage of actual moving distance to total distance) (***P = 0.0001; *P = 0.0231) in larvae at 5-dpf (n = 10 larvae for each froup). Data were analyzed by Student’s t test. All error bars indicate the mean ± standard deviation. MO morpholino, hpf hours post fertilization, qPCR quantitative real-time polymerase chain reaction, dpf days post fertilization. Source data are available online for this figure.

Foxk2 deficiency in muscle stem cells results in skeletal muscle dysplasia in mice.

(A) Representative images of Foxk2^fl/fl and Foxk2^fl/fl; Myod1-Cre littermates at 8 weeks, including their TA and Gas muscle. (B) Body weight of Foxk2^fl/fl and Foxk2^fl/fl; Myod1-Cre mice at 8 weeks (n = 7 mice for each group). P value: **P = 0.0020. (C) Ratio of Gas muscle weight to body weight in Foxk2^fl/fl and Foxk2^fl/fl; Myod1-Cre mice at 8 weeks (n = 6 mice for each group). P value: **P = 0.0047. (D) Hanging endurance time measured by the hanging endurance test in Foxk2^fl/fl and Foxk2^fl/fl; Myod1-Cre mice at 8 weeks (n = 6 mice for each group). P value: ***P = 0.0003. (E) H&E staining illustrating histological aspects and the percentage of central nuclei myofibers in skeletal muscles (Gas, TA, and EM) of Foxk2^fl/fl and Foxk2^fl/fl; Myod1-Cre littermates at postnatal 1 day, 4 weeks, and 8 weeks (n = 4 areas). Black arrows indicate central nuclei. The inset in the upper right corner (scale bars: 25 µm) is an enlarged view of the boxed area in the main image (scale bars: 100 µm), highlighting central nuclei myofibers. P value: 1 d: Gas, *P = 0.0389; TA, **P = 0.0019; EM, ns-P = 0.1708. 4w: Gas, **P = 0.0017; TA, ****P < 0.0001; EM, **P = 0.0031. 8w: Gas, **P = 0.0030; TA, ***P = 0.0006; EM, **P = 0.0026. (F) IHC staining of MyHC in the skeletal muscles (Gas, TA, and EM) of Foxk2^fl/fl and Foxk2^fl/fl; Myod1-Cre littermates at 8 weeks (n = 4 areas). MyHC was quantified by InDen/Area using ImageJ software. Scale bars: 100 µm for Gas and TA; 50 µm for EM. P value: Gas, *P = 0.0212; TA, **P = 0.0033; EM, **P = 0.0029. (G) IF staining of Laminin was the TA of Foxk2^fl/fl and Foxk2^fl/fl; Myod1-Cre littermates at 8 weeks. The distribution of cross-sectional area of myofibers was calculated (n = 3 areas). The zoomed area (scale bars: 100 µm) is an enlarged view of the boxed area in the main image (scale bars: 200 µm). P value: 500-1099, *P = 0.0401; 1100-1699, **P = 0.0049; 3500-4099, **P = 0.0014; 4100-4699, **P = 0.0041; >4700, **P = 0.0047. Data were analyzed by Student’s t test. All error bars indicate mean ± standard deviation. TA tibialis anterior, Gas gastrocnemius, d day, w week, EM eyelid muscle, H&E hematoxylin-eosin staining, IHC immunohistochemistry, IF immunofluorescence. Source data are available online for this figure.

Defective mitochondrial functions in FOXK2 deficiency models.

(A) Representative TEM images in TA sections of Foxk2^fl/fl and Foxk2^fl/fl; Myod1-Cre littermates at 8 weeks. Black arrows point to the damaged mitochondria. The image below (scale bars: 1 µm) is an enlarged view of the white box in the image above (scale bars: 2 µm). (B) ROS production in TA sections of Foxk2^fl/fl and Foxk2^fl/fl; Myod1-Cre littermates at 8 weeks by DHE staining (n = 4 areas). The image below (scale bars: 100 µm) is an enlarged view of the white box in the image above (scale bars: 200 µm). P value: **P = 0.0014. (C) ATPase activity detection in TA sections of Foxk2^fl/fl and Foxk2^fl/fl; Myod1-Cre littermates at 8 weeks. The image below (scale bars: 100 µm) is an enlarged view of the white box in the image above (scale bars: 200 µm). (D) DRP1 and OPA1 IF staining of MuSCs from Foxk2^fl/fl and Foxk2^fl/fl; Myod1-Cre littermates with fluorescence quantitative statistics shown on the right (n = 3 areas). Scale bars: 100 µm. P value: DRP1, *P = 0.0119; OPA1, ***P = 0.0009. (E) Representative images of MuSCs from Foxk2^fl/fl and Foxk2^fl/fl; Myod1-Cre littermates by JC-1 staining. The zoomed-in area (scale bars: 100 µm) is an enlarged view of the boxed area in the image (scale bars: 200 µm). The red/green ratio fluorescence quantitative statistics to measure mitochondrial membrane potential are shown on the right (n = 3 areas). P value: *P = 0.0193. (F) Representative images of MuSCs (scale bars: 200 µm) and quantitative of fluorescence intensity to measure ROS levels in MuSCs from Foxk2^fl/fl and Foxk2^fl/fl; Myod1-Cre littermates by DCFH-DA staining (n = 5 areas). P value: ****P < 0.0001. (G) Quantification of the ATP content in MuSCs from Foxk2^fl/fl and Foxk2^fl/fl; Myod1-Cre littermates (n = 3 repeats). The values were normalized to the total cellular protein level. P value: *P = 0.0105. (H) Quantitative of mitochondrial DNA copy number in MuSCs from Foxk2^fl/fl and Foxk2^fl/fl; Myod1-Cre littermates by qPCR (n = 3 repeats). P value: ns-P = 0.6130. (I) Quantitative of relative mitochondrial genes expression levels in MuSCs from Foxk2^fl/fl and Foxk2^fl/fl; Myod1-Cre littermates by qPCR (n = 3 repeats). P value: Tfb2m, *P = 0.0125; Cox4l, **P = 0.0039; Cycs, **P = 0.0011; Pgc1a, ns-P = 0.8299; Tfb1m, ns-P = 0.0608. Data were analyzed by Student’s t test. All error bars indicate mean ± standard deviation. TEM transmission electron microscopy, ROS reactive oxygen species, TA tibialis anterior, DHE dihydroethidium, IF immunofluorescence, MuSCs muscle stem cells, DCFH-DA dichlorodihydrofluorescein diacetate, qPCR quantitative real-time polymerase chain reaction. Source data are available online for this figure.

Impaired myogenic differentiation by FOXK2 deficiency and ineffective rescue of R215W mutant in C2C12 cells.

(A, B) (A)Western blot and (B) qPCR analysis (n = 3 repeats) of MyHC, MYOG, MYOD, MYF5, PAX7 and FOXK2 expression levels in C2C12 cells. Cells were induced into myogenic differentiation and whole cell lysates were isolated at the indicated times for analysis. The black arrow points to the specific protein band. (C) Western blot analysis of FOXK2 expression level in C2C12 cells at undifferentiated stage (day 0). The black arrow points to the specific protein band. (D) Quantitative analysis of Myh4 and Myog expression levels by qPCR after 6-day myogenic differentiation (n = 3 repeats). P value for blank vs control, control vs FOXK2-KO, FOXK2-KO vs KO + EV, KO + EV vs KO + WT, KO + WT vs KO + R215W and KO + EV vs KO + R215W: Myh4: ns-P = 0.2985; ****P < 0.0001; ns-P = 0.7651; ***P = 0.0009; ***P = 0.0009; ns-P = 0.2905. Myog: ns-P = 0.3348; **P = 0.0040; ns-P = 0.3182; ****P < 0.0001; **P = 0.0015; ns-P = 0.8951. (E) Representative images of MyHC (red) and DAPI (blue) IF staining in C2C12 cells after 6-day myogenic differentiation (scale bars: 200 µm area). The percentage of MyHC-positive nuclei (n = 3 areas) and the diameter of myofibers (n = 23 myofibers) are shown below. P value for blank vs control, control vs FOXK2-KO, FOXK2-KO vs KO + EV, KO + EV vs KO + WT, KO + WT vs KO + R215W and KO + EV vs KO + R215W: Left: ns-P = 0.6887; ***P = 0.0010; ns-P = 0.8348; ****P < 0.0001; **P = 0.0010; **P = 0.0018. Right: ns-P = 0.7009; ****P < 0.0001; ns-P = 0.8594; **P = 0.0070; **P = 0.0081; ns-P = 0.8277. Data were analyzed by Student’s t test. All error bars indicate mean ± standard deviation. Blank untreated wild-type C2C12 cells, control C2C12 cells transfected with a blank vector as a control for the FOXK2-KO group, FOXK2-KO FOXK2 knockout C2C12 cells, KO + EV FOXK2-KO cells transfected with a blank vector as a control for KO + WT and KO + R215W groups, KO + WT FOXK2-KO cells with overexpression of the human wild-type FOXK2 vector, KO + R215W FOXK2-KO cells with overexpression of the human FOXK2 R215W mutation vector, qPCR quantitative real-time polymerase chain reaction, IF immunofluorescence. Source data are available online for this figure.

Mitochondrial dyshomeostasis by FOXK2 deficiency and low-effective rescue of R215W mutant in C2C12 cells.

(A) Representative TEM images of C2C12 cells. Blank arrows point to the incomplete mitochondrial membrane, and red asterisks indicate indistinct cristae. The zoomed-in area (scale bars: 1 µm) is an enlarged image of the boxed area in the left image (scale bars: 2 µm). (B, C) (B) DRP1 and (C) OPA1 IF staining of C2C12 cells and fluorescence quantitative statistics shown below (n = 4 areas). Scale bars: 50 µm. P value for blank vs control, control vs FOXK2-KO, FOXK2-KO vs KO + EV, KO + EV vs KO + WT, KO + WT vs KO + R215W and KO + EV vs KO + R215W: DRP1: ns-P = 0.2406; ****P < 0.0001; ns-P = 0.9454; ***P = 0.0003; **P = 0.0015; *P = 0.0429. OPA1: ns-P = 0.8264; *P = 0.0456; ns-P = 0.2234; ****P < 0.0001; *P = 0.0433; **P = 0.0092. (D) Representative images of C2C12 cells by JC-1 staining. Scale bars: 100 µm. (E) Quantitative statistics of mean red/green ratio fluorescence intensity of (D) to measure mitochondrial membrane potential (n = 3 areas). P value: **P = 0.0097; *P = 0.0390; blank vs control, ns-P = 0.9385; FOXK2-KO vs KO + EV, ns-P = 0.5861; KO + EV vs KO + R215W, ns-P = 0.1023; KO + WT and KO + R215W, ns-P = 0.3174. (F) Quantification of ATP contents in C2C12 cells (n = 3 repeats). The values were normalized to the total cellular protein level. P value: ****P < 0.0001; **P = 0.0013; blank vs control, ns-P = 0.0961; control vs FOXK2-KO, ***P = 0.0004; FOXK2-KO vs KO + EV, ns-P = 0.1541; KO + EV vs KO + R215W, ***P = 0.0003. (G) Quantitative analysis of fluorescence intensity to measure ROS levels in C2C12 cells by DCFH-DA staining (n = 4 repeats). P value: ****P < 0.0001; ***P = 0.0007; **P = 0.0050; *P = 0.0337; blank vs control, ns-P = 0.7027; FOXK2-KO vs KO + EV, ns-P = 0.2653. (H) Quantitative of mtDNA copy number in C2C12 cells measured by qPCR (n = 3 repeats). P value: blank vs control, ns-P = 0.5916; control vs FOXK2-KO, *P = 0.0214; FOXK2-KO vs KO + EV, ns-P = 0.4779; KO + EV vs KO + WT, *P = 0.0234; KO + WT vs KO + R215W, ns-P = 0.2577; KO + EV vs KO + R215W, ns-P = 0.3716. (I) Quantitative of relative mitochondrial relative gene expression levels in C2C12 cells by qPCR (n = 3 repeats). P value for blank vs control, control vs FOXK2-KO, FOXK2-KO vs KO + EV, KO + EV vs KO + WT, KO + WT vs KO + R215W and KO + EV vs KO + R215W: Pgc1a: ns-P = 0.6530; ***P = 0.0005; ns-P = 0.5826; ***P = 0.0002; ns-P = 0.1210; **P = 0.0011. Tfb1m: ns-P = 0.1746; **P = 0.0032; ns-P = 0.1248; *P = 0.0302; ns-P = 0.0663; ns-P = 0.1662. Tfb2m: ns-P = 0.1434; **P = 0.0064; ns-P = 0.4715; **P = 0.0047; ns-P = 0.0556; ns-P = 0.4704. Data were analyzed by Student’s t test. All error bars indicate mean ± standard deviation. Blank untreated wild-type C2C12 cells, control C2C12 cells transfected with a blank vector as a control for the FOXK2-KO group, FOXK2-KO FOXK2 knockout C2C12 cells, KO + EV FOXK2-KO cells transfected with a blank vector as a control for KO + WT and KO + R215W groups, KO + WT FOXK2-KO cells with overexpression of the human wild-type FOXK2 vector, KO + R215W FOXK2-KO cells with overexpression of the human FOXK2 R215W mutation vector, ROS reactive oxygen species, TEM transmission electron microscopy, IF immunofluorescence, DCFH-DA Dichlorodihydrofluorescein diacetate, mtDNA mitochondrial DNA, qPCR quantitative real-time polymerase chain reaction. Source data are available online for this figure.

Coenzyme Q10 improves mitochondrial dysfunction and skeletal muscle development disorders of mice with Foxk2 deficiency in muscle stem cells.

(A) Illustration of the study design of CoQ10 treatment in mice. (B) Concentration of CoQ10 in TA tissues of mice at 2 weeks (n = 3 mice). P value: **P = 0.0013; ns-P = 0.4524; Foxk2^fl/fl+Vehicle vs Foxk2^fl/fl + CoQ10, *P = 0.0118; Foxk2^fl/fl + CoQ10 vs Foxk2^fl/fl;Myod1-Cre+CoQ10, *P = 0.0435. (C) Quantification of ATP content in TA tissues from Foxk2^fl/fl and Foxk2^fl/fl; Myod1-Cre littermates at 2 weeks, with or without CoQ10 treatment (n = 3 mice). Values were normalized to total protein levels. P value: **P = 0.0012; *P = 0.0119; Foxk2^fl/fl+Vehicle vs Foxk2^fl/fl + CoQ10, ns-P = 0.9377; Foxk2^fl/fl + CoQ10 vs Foxk2^fl/fl;Myod1-Cre+CoQ10, ns-P = 0.2180. (D) Representative TEM images of TA sections from Foxk2^fl/fl and Foxk2^fl/fl; Myod1-Cre littermates at 2 weeks. Black arrows indicate damaged mitochondria. The percentage of damaged mitochondria was quantified using ImageJ software (n = 8 areas). Scale bars: 1 µm. P value: Foxk2^fl/fl+Vehicle vs Foxk2^fl/fl;Myod1-Cre+Vehicle **P = 0.0017; Foxk2^fl/fl+Vehicle vs Foxk2^fl/fl + CoQ10, **P = 0.0066; Foxk2^fl/fl + CoQ10 vs Foxk2^fl/fl;Myod1-Cre+CoQ10 **P = 0.0022; Foxk2^fl/fl;Myod1-Cre+Vehicle vs Foxk2^fl/fl;Myod1-Cre+ CoQ10, **P = 0.0057. (E) ROS production in TA sections of Foxk2^fl/fl and Foxk2^fl/fl; Myod1-Cre littermates at 2 weeks, with or without CoQ10 treatment, measured by DHE staining (n = 4 areas). Quantifications were performed using ImageJ software. Scare bar: 200 μm. P value: Foxk2^fl/fl+Vehicle vs Foxk2^fl/fl;Myod1-Cre+Vehicle *P = 0.0325; Foxk2^fl/fl+Vehicle vs Foxk2^fl/fl + CoQ10, ns-P = 0.1736; Foxk2^fl/fl + CoQ10 vs Foxk2^fl/fl;Myod1-Cre+CoQ10 ns-P = 0.2200; Foxk2^fl/fl;Myod1-Cre+Vehicle vs Foxk2^fl/fl;Myod1-Cre+ CoQ10, *P = 0.0426. (F) H&E staining illustrating histological aspects and the percentage of central nuclei myofibers in skeletal muscles (Gas, TA, IM, TM, Dp, and EM) of Foxk2^fl/fl and Foxk2^fl/fl; Myod1-Cre littermates at 2 weeks, with or without CoQ10 treatment (n = 4 areas). Black arrows indicate central nuclei. The inset in the upper right corner (scale bars: 25 µm) is an enlarged view of the boxed area in the main image, highlighting central nuclei myofibers (scale bars: 100 µm). P value for Foxk2^fl/fl+Vehicle vs Foxk2^fl/fl + CoQ10, Foxk2^fl/fl;Myod1-Cre+Vehicle vs Foxk2^fl/fl;Myod1-Cre+CoQ10, Foxk2^fl/fl+Vehicle vs Foxk2^fl/fl;Myod1-Cre+Vehicle and Foxk2^fl/fl + CoQ10 vs Foxk2^fl/fl;Myod1-Cre+CoQ10: Gas: ns-P = 0.6529; *P = 0.0225; ***P = 0.0003; ***P = 0.0007. TA: ns-P = 0.4640; ns-P = 0.2498; ***P = 0.0008; **P = 0.0011. IM: ns-P = 0.5588; **P = 0.0067; ***P = 0.0004; *P = 0.0337. TM: ns-P = 0.1669; *P = 0.0110; **P = 0.0044; ns-P = 0.1219. Dp: ns-P = 0.2431; *P = 0.0322; **P = 0.0077; ns-P = 0.6149. EM: ns-P = 0.6625; **P = 0.0024; ***P = 0.0001; ns-P = 0.4264. (G) IF staining of Laminin in the TA of Foxk2^fl/fl and Foxk2^fl/fl; Myod1-Cre littermates at 2 weeks with or without CoQ10 treatment. The distribution of CSA of myofibers was calculated (n = 3 areas). Scale bars: 100 µm. P value: 500-899, **P = 0.0063, *P = 0.0198; 900-1299, **P = 0.0010; 1300-1699, *P = 0.03871; 1700-2099, *P = 0.0251; 2100-2499, **P = 0.0022; >2500, **P = 0.0060. Data were analyzed by Student’s t test. All error bars indicate mean ± standard deviation. CoQ10 Coenzyme Q10, Gas gastrocnemius, TA tibialis anterior, IM intercostal muscle, TM tongue muscle, Dp diaphragm, EM eyelid muscle, TEM transmission electron microscopy, ROS reactive oxygen species, DHE dihydroethidium, H&E hematoxylin-eosin staining, IF immunofluorescence, CSA cross-sectional area.

FOXK2 transcriptional regulation in C2C12 Cells.

(A) Genome-wide distribution of FOXK2 binding sites identified by ChIP-seq in C2C12 cells. (B) Enriched motif at FOXK2 binding sites identified by ChIP-seq. (C) GO analysis (BP) of genes with promoter regions bound by FOXK2, with a focus on mitochondrion-related terms. (D) Genome-wide distribution of FOXK2-binding peaks with altered chromatin accessibility in FOXK2-KO C2C12 cells, as determined by ATAC-seq. (E) Heatmap showing changes in chromatin accessibility changes at FOXK2 binding sites in FOXK2-KO C2C12 cells. (F) GO analysis (BP) of genes with upregulated chromatin accessibility in FOXK2-KO C2C12 cells, highlighting mitochondrion-related terms. (G) Venn diagrams show the overlapping (19 genes) among genes FOXK2-bound promoters (ChIP-seq), upregulated chromatin accessibility (ATAC-seq), and mitochondrion-related genes from the MitoCarta 3.0 database. (H) Volcano plot of DEGs (Down: 910 genes; Up: 1107 genes) from RNA-seq, with genes showing |log2FC| > 0.58 and p-value < 0.05. (I) Normalized RNA-seq, ChIP-seq and ATAC-seq profiles at the Pdk4 and Cry1 loci in FOXK2-KO vs. control cells. (J) Quantitative of relative Pdk4 and Cry1 mRNA expression levels in C2C12 cells and MuSCs from Foxk2^fl/fl and Foxk2^fl/fl; Myod1-Cre mice, assessed by qPCR (n = 3 repeats). P value: C2C12: Pdk4, *P = 0.0154; Cry1, **P = 0.0027. MuSCs: Pdk4, ***P = 0.0010; Cry1, *P = 0.0176. Data were analyzed by Student’s t test. All error bars represent mean ± standard deviation. MuSCs muscle stem cells, ChIP-seq chromatin immunoprecipitation sequencing, GO Gene Ontology, BP biological process, control C2C12 cells transfected with a blank vector as a control for the FOXK2-KO group, FOXK2-KO FOXK2 knockout C2C12 cells, ATAC-seq assay of transposase accessible chromatin sequencing, RNA-seq RNA sequencing, DEGs differentially expressed genes, FC fold change, qPCR quantitative real-time polymerase chain reaction. Source data are available online for this figure.