The FGF13-Caveolin-1 Axis: A Key Player in the Pathogenesis of Doxorubicin- and D-Galactose-Induced Premature Cardiac Aging

Abstract

Delaying senescence of cardiomyocytes has garnered widespread attention as a potential target for preventing cardiovascular diseases (CVDs). FGF13 (Fibroblast growth factor 13) has been implicated in various pathophysiological processes. However, its role in premature myocardial aging and cardiomyocyte senescence remains unknown. Adeno-associated virus 9 (AAV9) vectors expressing FGF13 and cardiac-specific Fgf13 knockout (Fgf13KO) mice are utilized to reveal that FGF13 overexpression and deficiency exacerbated and alleviated Doxorubicin/D-galactose-induced myocardial aging characteristics and functional impairment, respectively. Transcriptomics are employed to identify an association between FGF13 and Caveolin-1 (Cav1). Mechanistic studies indicated that FGF13 regulated the Cav1 promoter activity and expression through the p38/MAPK pathway and nuclear translocation of p65, as well as the binding level of PTRF to Cav1 to mediate cardiomyocyte senescence. Furthermore, Cav1 overexpression in murine hearts reversed the alleviatory effects of FGF13 deficiency on the Doxorubicin/D-galactose-induced myocardial aging phenotype and dysfunction. This study has demonstrated that FGF13 regulated the Cav1-p53-p21 axis to augment cardiomyocyte senescence and thereby exacerbated cardiac premature aging and suggests that FGF13 knockdown may be a promising approach to combat CVDs in response to aging and chemotoxicity.

Keywords: cardiac premature aging; cardiomyocyte senescence; caveolin‐1; fibroblast growth factor 13; p53 signaling.

Figure 1

The abundance of FGF13 is downregulated in premature cardiac aging models. A–C) The results were from the analysis of the GSE12480 dataset (A) and GSE56348 dataset (B). A) The corresponding FGFs of the 2 genes correlated with the myocardial aging were visualized by the heatmap in the hearts from 8 young (4‐6 month) and 8 old (25–28 month) mice samples. B) The corresponding FGFs of the 4 genes correlated with heart failure were visualized by the heatmap in the hearts from 5 mice with heart failure (HF) and 5 sham‐operated control samples (Sham‐HF). The darker shade of red or blue represents the higher correlation level. C) Venn diagram visualizing the myocardial aging and failure‐related differentially expressed genes between two datasets. D) The relative mRNA expression level of FGF13 in 5 human young hearts and 12 human aging hearts (from GSE141910). E,F) 8‐Week mice were administered with D‐galactose (200 mg kg⁻¹ per day for 0, 2, 4, 6, 8 weeks). Myocardial tissues were collected for indicated analyses. E) Representative images of hearts stained with p21, FGF13, or Sirius red for the same groups (n = 4 per group) and the quantitative analysis of p21‐positive areas (n = 4 per group), FGF13‐positive areas (n = 4 per group), and interstitial fibrosis (Sirius red, n = 4 per group). F) Representative western blotting for FGF13 in heart tissue and the quantitative analysis (n = 4 per group). G,H) Representative western blotting for FGF13 in G) primary cardiomyocytes and H) fibroblasts after D‐galactose (20 g L⁻¹) stimulation for different times (n = 3 per group) and the quantitative analysis (n = 3 per group). I) Representative image of SA‐β‐gal in the heart tissue, treated with D‐galactose (200 mg kg⁻¹ per day for 8 weeks) and PBS (the upper part), treated with Doxorubicin (2.5 mg/kg/time × 2 per week for 4 consecutive weeks) and PBS (the bottom part), and the quantitative analysis of the percentage of SA‐β‐gal‐positive areas (n = 4 per group). J,K) The results were from the analysis of the GSE230638 dataset. Comparative gene expression profiling analysis of RNA‐seq data for hiPSC‐cardiomyocyte cells treated with Doxorubicin versus DMSO. J) The KEGG pathways. The x‐axis is the rich factor. The y‐axis refers to pathway terms. Volcano plot of differentially expressed genes. Red dots represent upregulated genes, blue dots represent downregulated genes, and gray dots represent genes that were not differentially expressed. We use log2 of the fold change as the source of data for the X axis and‐log10 of the P as the source of data for the y axis. Fold change >1.5× and adjusted p‐value <0.05 indicate statistically significant differences. The differentially expressed genes of Doxorubicin‐induced cardiomyocyte senescence pathway have been labeled. An online platform (https://www.bioinformatics.com.cn) was utilized. K) The relative mRNA expression level of FGF13 was analyzed (from GSE230638) (n = 5 per group) and the relative mRNA expression level of Fgf13 in 5 Doxorubicin or PBS treated murine hearts (from GSE81448). L) 8‐Week mice were administered with Doxorubicin (2.5 mg/kg/time × 2 per week for 4 consecutive weeks) and PBS. Then the myocardial tissues were collected. Representative western blotting for FGF13 in heart tissue and the quantitative analysis (n = 10 per group). M,N) Representative western blotting for FGF13 in M) cardiomyocytes and N) fibroblasts after Doxorubicin (0.1 µm) stimulation for different times and the quantitative analysis (n = 3 per group). The protein level was standardized by GAPDH. Data are means ± SEM.  The P value was determined using two‐tailed unpaired Student's t test or ANOVA with Tukey's multiple comparisons test.

Figure 2

Cardiac‐specific knockout of Fgf13 in cardiomyocytes alleviates D‐galactose‐induced cardiomyocyte senescence and cardiac injury. A–G) For 7 weeks old wild‐type (WT, Fgf13^f/Y ) mice and Fgf13KO (Fgf13^f/Y crossed with αMHC‐MerCreMer) mice, tamoxifen was administered at the dose of 75 mg/kg/day for 5 consecutive days. One week after the injection, mice were subjected to Sham or D‐galactose treatment. H–N) FGF13 overexpression vector (AAV9‐cTnT‐FGF13) and control vector (AAV9‐LacZ) were injected intravenously into tail veins of 6 weeks old male C57BL/ 6J mice, respectively, 2 weeks after the injection, these mice were subjected to Sham or D‐galactose treatment. A,H) The experimental flowchart. B,I) Representative M‐mode echocardiographic images from each group in mice and G,N) representative echocardiographic data for LVEF and LVFS are shown (n = 6 per group). C,J) Sirius Red staining (the upper part) (scale bar, 50 µm) and Masson staining (the lower part) (scale bar, 50 µm) and G,N) quantification (n = 6 per group). D,K) Representative image of SA‐β‐gal in the heart tissue (Arrows represent positive marks) (scale bar, 50 µm) and G,N) quantification (n = 6 per group). E,L) WGA (wheat germ agglutinin; scale bar, 50 µm) (n = 6 per group) and G,N) quantification (n = 6 per group). F,M) Representative whole heart images (scale bar, 3 mm) and G,N) representative data of HW in the indicated groups (n = 6 per group). Data are means ± SEM.  The P value was determined using ANOVA with Tukey's multiple comparisons test.

Figure 3

Cardiac‐specific knockout of Fgf13 in cardiomyocytes alleviates Doxorubicin‐induced cardiomyocyte senescence and cardiac injury. A–G) For 7 weeks old wild‐type (WT, Fgf13^f/Y ) mice and Fgf13KO (Fgf13^f/Y crossed with αMHC‐MerCreMer) mice, tamoxifen was administered at the dose of 75 mg/kg/day for 5 consecutive days. One week after the injection, these mice were subjected to Sham or Doxorubicin treatment. H–N) FGF13 overexpression vector (AAV9‐cTnT‐FGF13) and control vector (AAV9‐LacZ) were injected intravenously into tail veins of 6 weeks old male C57BL/ 6J mice, respectively, 2 weeks after the injection, these mice were subjected to Sham or Doxorubicin treatment. A,H) The experimental flowchart. B,I) Representative M‐mode echocardiographic images from each group in mice and G,N) representative echocardiographic data for LVEF and LVFS are shown (n = 6 per group). C,J) Sirius Red staining (the upper part) (scale bar, 50 µm) and Masson staining (the lower part) (scale bar, 50 µm) and G,N) quantification (n = 6 per group). D,K) Representative image of SA‐β‐gal in the heart tissue (Arrows represent positive marks) (scale bar, 50 µm) and G,N) quantification (n = 6 per group). E,L) WGA (wheat germ agglutinin; scale bar, 50 µm) (n = 6 per group) and G,N) quantification (n = 6 per group). F,M) Representative whole heart images (scale bar, 3 mm) and G,N) representative data of HW in the indicated groups (n = 6 per group). Data are means ± SEM.  The P value was determined using ANOVA with Tukey's multiple comparisons test.

Figure 4

FGF13 knockdown alleviates premature cardiomyocyte senescence. Transfected Neonatal Rat Cardiomyocytes (NRCMs) were either untreated or treated with Doxorubicin (0.1 µm) or D‐galactose (20 g L⁻¹) for 72 h and related quantification. A–D) TRITC Phalloidin staining (the upper part) and E) related quantification (n = 3 per group). A–D) β‐galactosidase staining (the lower part) and F) related quantification (n = 3 per group). A,C) NRCMs were transfected with si‐FGF13 or si‐NC. B,D) NRCMs were transfected with FGF13^oe or empty vector (EV). G,H) Representative western blotting for p53 and p21 in NRCMs and related quantification (n = 3 per group). The protein level was standardized by GAPDH. Data are means ± SEM. The P value was determined using ANOVA with Tukey's multiple comparisons test.

Figure 5

FGF13 regulates the expression of Cav1 in vivo and in vitro in senescent cardiomyocytes. A–C) RNA transcriptome sequencing was performed on D‐galactose‐induced cardiac tissue samples treated with AAV9‐FGF13 (n = 3) and AAV9‐LacZ (n = 3), respectively. A) Volcano plot of differentially expressed genes. We used log2 of the fold change as the source of data for the X axis and‐log10 of the P as the source of data for the y axis. Fold change >1.5× and padj < 0.05 indicate statistically significant differences. Red and blue points represent the upregulated genes and the downregulated genes compared with control group. B) KEGG pathways. The x‐axis is the rich factor. The y‐axis refers to pathway terms. C) The corresponding Cav1‐3 were visualized by the heatmap. The darker shade of red or blue represents the higher correlation level (the upper part). The variation and padj of corresponding Cav1‐3 and Cavin1–4 were shown in the table (the lower part). D) Real‐time qPCR analysis of the Cav1‐3 mRNA expression in D‐galactose‐induced NRCMs treated with FGF13^oe or EV and related quantification (the left part) and in Doxorubicin‐induced NRCMs treated with si‐FGF13 or si‐NC and related quantification (the right part). E) NRCMs were transfected with si‐FGF13 (or si‐NC) and FGF13^oe (or EV). Transfected NRCMs were either untreated or treated with D‐galactose or Doxorubicin for 72 h. Representative western blotting results (the upper part) and related quantification of Cav1 in NRCMs (the lower part) (n = 3 per group). F) Representative western blotting results in D‐galactose or Doxorubicin induced murine heart after myocardial‐specific FGF13 overexpression (the upper part) and related quantification of Cav1 (the lower part) (n = 6 per group). G,H) Representative images of immunofluorescence staining of Cav1 (red) and cTNT (green) and DAPI (blue) in D‐galactose and Doxorubicin induced mouse hearts after myocardial‐specific Fgf13 knockout. Data are means ± SEM. The P value was determined using ANOVA with Tukey's multiple comparisons test.

Figure 6

Myocardial‐specific overexpression of Cav1 reverses the protective effect of FGF13 knockout against cardiomyocyte senescence and cardiac injury. For 6 weeks old wild‐type (WT, Fgf13^{f/Y )} mice and Fgf13KO (Fgf13^f/Y crossed with αMHC‐MerCreMer) mice, tamoxifen was administered at the dose of 75 mg/kg/day for 5 consecutive days. On the 7th week, AAV9‐Cav1 (or AAV9‐LacZ) was injected into the tail vein for 2 weeks. Then these mice were subjected to D‐galactose or Doxorubicin treatment. After successful modeling, cardiac function tests were performed, and tissue samples were collected. A,H) The experimental flowchart. B,I) Representative M‐mode echocardiographic images from each group in mice and G,N) representative echocardiographic data for LVEF and LVFS are shown (n = 6 per group). C,J) Sirius Red staining (the upper part) (scale bar, 50 µm) and Masson staining (the lower part) (scale bar, 50 µm) and G,N) quantification (n = 6 per group). D,K) Representative image of SA‐β‐gal in the heart tissue (Arrows represent positive marks) (scale bar, 50 µm) and G,N) quantification (n = 6 per group). E,L) WGA (wheat germ agglutinin; scale bar, 50 µm) (n = 6 per group) and G,N) quantification (n = 6 per group). F,M) Representative whole heart images (scale bar, 3 mm) and G,N) representative data of HW in the indicated groups (n = 6 per group). Data are means ± SEM. The P value was determined using ANOVA with Tukey's multiple comparisons test.

Figure 7

The regulation of premature senescence in cardiomyocytes by FGF13 in vitro depends on the expression of Cav1. NRCMs were either untreated or treated with Doxorubicin (0.1 µm) for 72 h. A–E) NRCMs were transfected with si‐FGF13 (si‐NC) and Cav1^oe (or EV). F–J) NRCMs were transfected with FGF13^oe (or EV) and si‐Cav1 (or si‐NC). B,G) Representative western blotting results and D,I) related quantification of p53, p21 (n = 5 per group). A,F) NRCMs were transfected with Cav1^oe (or EV) and si‐Cav1 (or si‐NC). Representative Western blotting results for Cav1 and D,I) related quantification (n = 3 per group). The protein level was standardized by GAPDH. C,H) TRITC Phalloidin staining (the upper part) and β‐galactosidase staining (the lower part) and E,J) related quantification (n = 3 per group). Data are means ± SEM. The P value was determined using two‐tailed unpaired Student's t test or ANOVA with Tukey's multiple comparisons test.

Figure 8

FGF13 regulates Cav1 activity and its upstream pathway to participate in the cardiomyocyte senescent process. NRCMs transfected with si‐FGF13 (or si‐NC) or FGF13^oe (or EV), in the presence or absence of JSH‐23 (20 µm) or SB203580 (20 µm) for 3 h, and then were either untreated or treated with Doxorubicin (0.1 µm) or D‐galactose (20 g L⁻¹) for F,I) 24 h, A,C) 48 h, D,E) 72 h. HEK293T cells were transfected with B) si‐FGF13 (0, 25, 50, 75 nm) or FGF13^oe (0.25, 0.5 µg) for 48 h (n = 4 per group). A–C) Luciferase reporter activities of Cav1. D) Representative western blotting results and related quantification of H) Cav1, and the ratio of p‐p38 to p38 (n = 4 per group). The protein level was standardized by GAPDH. E) β‐galactosidase staining (n = 3 per group) and G) related quantification. F) Representative images of immunofluorescence staining of Cav1 (green) and PTRF (red) and DAPI (blue) (scale bar, 50 µm) (n = 3 per group). I) Representative images of immunofluorescence staining of p65 (red) and DAPI (blue) (scale bar, 50 µm) (n = 3 per group). J) The graphic abstract. Created in BioRender. shen, m. (2025) https://BioRender.com/0ul3s50. Data are means ± SEM. The P value was determined using ANOVA with Tukey's multiple comparisons test.