FBLN7 KO attenuates age-related cardiac fibrosis by promoting TGFBR3/ALK1/Smad1 signaling and inhibiting the profibrotic phenotypes of cardiac fibroblasts

Abstract

Rationale: Aging induces structural and functional changes in the heart, including left ventricular (LV) hypertrophy, a decline in diastolic function, and even heart failure. Fibulin 7 (FBLN7) is a key mediator of extracellular matrix (ECM) remodeling under pathological conditions. In our study, we aim to explore whether FBLN7 is also involved in the development of age-related cardiac fibrosis and its underlying mechanisms. Methods: We generated naturally aged FBLN7 knockout and wild-type mice (18 months old). Western blot and immunofluorescence assays were employed to investigate the biological function of FBLN7 in senescent cardiac fibroblasts. The interaction between FBLN7 and cell membrane receptors was explored through molecular docking and co-immunoprecipitation techniques. The interaction between FBLN7 and natural products was explored through virtual screening, molecular dynamics simulations and surface plasmon resonance (SPR). Results: Our results demonstrated that the cardioprotective effects observed in aged FBLN7 knockout (KO) mice are mediated by the inhibition of profibrotic phenotypes in senescent cardiac fibroblasts (CFs), which reduces age-related myocardial fibrosis and ultimately improves cardiac diastolic function. The observation that overexpressing FBLN7 in fibroblast-specific protein 1 positive (FSP1⁺) cells of aged mice exacerbates age-related myocardial fibrosis further supports this finding. Mechanistically, we identified that FBLN7 promotes the proliferation, migration, actin remodeling, and collagen production of senescent CFs at least partially by binding to TGFBR3 and reducing its protein levels, thereby inhibiting the activation of the ALK1-Smad1/5/9 pathway. Additionally, we identified a natural product, Ginsenoside Ro, that physically interacts with FBLN7 and validated its antifibrotic activity both in vitro and in vivo. Conclusions: These findings reveal FBLN7 reverses the impaired profibrotic phenotypes of senescent CFs, thereby aggravating age-related cardiac fibrosis. Given that age-related fibrosis is a significant pathological factor in heart failure with preserved ejection fraction (HFpEF), downregulating FBLN7 and/or interfering with its function may represent an effective therapeutic strategy for HFpEF.

Keywords: FBLN7; TGFBR3; cardiac fibrosis; fibroblasts senescence; natural products.

Figure 1

Loss of FBLN7 exacerbates age-related cardiac hypertrophy and diastolic dysfunction. A) Representative Western blot images and quantification of Collagen I, Collagen III, FBLN7, and P21 protein levels in the hearts of wild-type (WT) mice at 2 months and 18 months of age. Western blot analyses show the expression of these proteins, with 8 mice in each group. B) Representative images of immunohistochemical staining for FBLN7 in heart sections from WT mice at 2 months and 18 months, with quantification displayed on the right. C) ELISA detection of serum FBLN7 levels. D) Gross appearance of WT mice and FBLN7 knockout (KO) mice at 2 months and 18 months of age. E) Gross appearance of the hearts of WT mice and FBLN7-KO mice at 2 months and 18 months of age. F) Heart weight normalized to body weight (HW/BW) of the indicated groups of mice. G) Heart weight normalized to tibia length (HW/TL) of the indicated groups of mice. H) Representative echocardiography images of the four groups of mice. Upper: B-mode; middle: PW Doppler mode; lower: tissue Doppler mode. Echocardiographic analysis of interventricular septal thickness in diastole (IVSd), interventricular septal thickness in systole (IVSs), left ventricular posterior wall thickness in diastole (LVPWd), left ventricular posterior wall thickness in systole (LVPWs), fractional shortening (FS%), LV mass, the ratio of early diastolic filling velocity to left atrial pressure (E/e') and the ratio of early to late diastolic filling velocities (E/A) are presented on the right and bottom. Error bars represent the standard error of the mean (SEM). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Figure 2

FBLN7 KO protects mice against age-related cardiac fibrosis. Cardiac fibrosis was assessed in wild-type (WT) and FBLN7 knockout (KO) mice at 2 months and 18 months of age. A) Representative images of hematoxylin and eosin (H&E) staining of hearts from the four groups of mice. The lower panel shows the enlargement of the black box. B) Representative images of wheat germ agglutinin (WGA) staining for the four groups of mice, with quantifications displayed around. The lower panel shows the enlargement of the white box. C) Representative micrographs of Masson's trichrome staining in heart sections from the four groups of mice, showing three different views: upper (perivascular), middle (global), and lower (interstitial). Quantifications of fibrotic areas are presented at the bottom. D) Representative images of heart sections stained with Sirius red. E) Representative micrographs of immunohistochemical (IHC) staining for Collagen I and III, along with corresponding statistical plots shown below. Both perivascular and interstitial views are included. F) Representative western blot images and quantifications of collagen I, collagen III and FBLN7 protein levels are presented. Error bars are SEM. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Figure 3

FBLN7 reversed the impaired profibrotic phenotypes of senescent CFs in vitro. A) Representative Immunofluorescent (IF) micrographs illustrating the co-localization of FBLN7 with cTnI, α-SMA, Vimentin, and CD31 in wild-type aging hearts (at 18 months of age) and young hearts (at 2 months of age) are presented. White arrowheads indicate co-localization (yellow). B) Ontology (GO) term analysis (biological processes) of differentially expressed proteins in the left ventricular tissues of FBLN7 gene knockout mice and wild-type mice. C) Representative Western blot images depict the expression of Collagen I, Collagen III, FBLN7, α-SMA, and P21 proteins in primary cardiac fibroblasts (CFs) treated with various concentrations of recombinant FBLN7 protein. Quantifications are shown on the right. D-I. The reversal effects of FBLN7 overexpression (adFBLN7) on the impaired profibrotic phenotypes of senescent CFs induced by D-galactose (D-gal) are illustrated. D) Representative Western blot images show the expression of Collagen I and III, α-SMA, P21, and FBLN7 proteins in FBLN7-overexpressing CFs after treatment with D-gal or PBS. Quantifications are displayed on the right. E) Representative images depict P21 expression (red) in CFs labeled with Vimentin (green). F) Representative IF images of CFs stained with α-SMA (green) and Vimentin (red), with nuclei stained using DAPI (blue), are shown. G) Representative images of CFs co-stained with EdU (red) and DAPI (blue) are presented, with the quantification of EdU-positive cells indicated on the right. H) Representative photomicrographs from Transwell assays are included, with the quantification of migrating CFs shown on the right. I) Representative images of senescence-associated β-galactosidase staining are provided, along with the quantification of β-galactosidase-positive cells (blue-green) displayed above the image. Error bars represent the standard error of the mean (SEM). Statistical significance is indicated as follows: *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Figure 4

FBLN7 modulates the impaired profibrotic abilities of senescent cardiac fibroblasts through the ALK1-Smad1/5/9 pathway. A-B) Representative western blot images displaying the levels of p-Smad2, Smad2, p-Smad1/5/9, and Smad1/5/9 proteins in cardiac fibroblasts with FBLN7 silencing (siFBLN7) (A), FBLN7 overexpression (adFBLN7) (B), and corresponding control cardiac fibroblasts (siNC or adNC) following treatment with D-gal or PBS. Quantifications of band intensity are presented on the right. C-D) Representative images illustrating the nuclear translocation of p-Smad1/5/9 (red) (C) and p-Smad2 (red) (D) in siFBLN7 and siNC cardiac fibroblasts after treatment with D-gal or PBS, labeled with Vimentin (green). E) Representative western blot images showing the protein levels of p-Smad2, Smad2, p-Smad1/5/9 and Smad1/5/9 in D-gal treated cardiac fibroblasts transfected with siNC, siFBLN7, or siFBLN7 plus ML347 treatment. Quantifications are displayed on the right. F) Representative western blot images showing the protein levels of Collagen I and III, ALK1, α-SMA and P21 in D-gal treated cardiac fibroblasts transfected with siNC, siFBLN7, or siFBLN7 plus ML347 treatment. Quantifications are displayed on the right. Error bars represent the standard error of the mean (SEM). Statistical significance is indicated as follows: *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Figure 5

TGFBR3 is essential for the differential transduction of TGF-β signals mediated by FBLN7. A) Endogenous co-immunoprecipitation (Co-IP) of FBLN7 and TGFBR3. Immunoprecipitation (IP) was performed with TGFBR3, followed by immunoblotting (IB) with FBLN7 in senescent cardiac fibroblasts (CFs). B) Representative confocal images demonstrating the co-localization of FBLN7 and TGFBR3 in senescent CFs. C) Reciprocal Co-IP of FBLN7 and TGFBR3. IP was conducted with GFP-tag and IB with Flag-TGFBR3 in 293T cells (left panel). Conversely, IP was performed with Flag-tag and IB with GFP-FBLN7 in 293T cells (right panel). D) Representative confocal images illustrating the co-localization of Myc-FBLN7 and Flag-TGFBR3 in 293T cells. E) Results of molecular docking between FBLN7 and TGFBR3. a) The optimal docking conformation of FBLN7 and TGFBR3 reveals a stable complex, with FBLN7 highlighted in blue and TGFBR3 in yellow. b) The white semi-transparent structure represents the entire molecular framework, providing a global view of the docking region. The black box in the center indicates the binding site, emphasizing critical areas on the binding interface. c) The Docking Zoom Map displays key amino acid residues at the binding interface and their interactions. F-G) Representative western blot images showing TGFBR3 and ALK1 protein levels in CFs with FBLN7 silencing (F) and overexpression (G) after treatment with D-gal or PBS. Quantifications of band intensity are presented on the right. H) Representative western blot images illustrating protein levels of FBLN7 and TGFBR3 in D-gal-treated CFs infected with either Null, adenovirus encoding FBLN7 (adFBLN7), or adFBLN7 plus Flag-TGFBR3. Quantifications are displayed on the right. I) Representative western blot images illustrating protein levels of p-Smad1/5/9, Smad1/5/9, p-Smad2 and Smad2 in D-gal-treated CFs infected with either Null, adFBLN7, or adFBLN7 plus Flag-TGFBR3. Quantifications are displayed on the bottom. J) Representative western blot images illustrating protein levels of Collagen I and III, ALK1, α-SMA, and P21 in D-gal-treated CFs infected with either Null, adFBLN7, or adFBLN7 plus Flag-TGFBR3. Quantifications are displayed around the imge. Error bars represent the standard error of the mean (SEM). Statistical significance is indicated as follows: *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Figure 6

Overexpression of FBLN7 in FSP1⁺ cells promotes age-related cardiac fibrosis. A) Schedule for FBLN7 overexpression. B) Gross appearance of mice and hearts in each group. C) Representative echocardiographic images of the 2 groups of mice. Upper: B-mode; middle: PW Doppler mode; lower: tissue Doppler mode. Echocardiographic analysis of interventricular septal thickness in diastole (IVSd) and systole (IVSs), left ventricular posterior wall thickness in diastole (LVPWd) and systole (LVPWs), left ventricular mass, the ratio of early to late diastolic filling velocities (E/A), and the ratio of early diastolic filling velocity to left atrial pressure (E/e'), fractional shortening (FS%) and left ventricular ejection fraction (LVEF) are presented on the right and bottom. D) Representative images of H&E staining of hearts from aging mice (18 months old) injected with AAV-FBLN7 or AAV-Null. E) Representative WGA staining images of the two groups of mice, with quantification shown on the right. F) Representative micrographs of Masson staining in heart sections from AAV-FBLN7 and AAV-Null mice, displaying three different views: upper: perivascular; middle: global; lower: interstitial. Quantifications of fibrotic areas are shown on the right. G) Representative western blot images showing Collagen I and III protein levels in heart tissues from the AAV-Null and AAV-FBLN7 groups. Quantifications are presented on the bottom. H) Results of molecular docking between FBLN7 and Ginsenoside Ro (GRO). a) Structure of GRO. b) Docking zoom map illustrating key amino acid residues at the binding interface and their interactions. c) Predicted structure model of the FBLN7-GRO complex from the GRAMM program, with FBLN7 represented in cartoon form and GRO depicted as sticks. I) Molecular dynamics simulation of the FBLN7-GRO complex over 100 ns. a) Complex root mean square deviation (RMSD) analysis. b) Ligand RMSD analysis. c) Radius of gyration (Rg) analysis. d) Solvent-accessible surface area (SASA) analysis. e) Root mean square fluctuation (RMSF) analysis. f) Hydrogen bond analysis. J) Surface plasmon resonance (SPR) analysis of FBLN7 binding to GRO. Error bars represent the standard error of the mean (SEM). Statistical significance is indicated as follows: *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Figure 7

GRO rescues the impaired profibrotic phenotypes of senescent cardiac fibroblasts and prevents age-related cardiac fibrosis and diastolic dysfunction. A) Representative Western blot images displaying protein levels of Collagen I and III, TGFBR3, ALK1, α-SMA, and P21 in senescent CFs infected with adenovirus encoding FBLN7 (adFBLN7) and treated with DMSO or various concentrations of GRO (50, 100, and 200 μM). Quantifications are presented on the right. *: adNC+D-gal+DMSO vs. adNC+PBS+DMSO, &: adFBLN7+D-gal+DMSO vs. adNC+D-gal+DMSO, $: adFBLN7+D-gal+GRO50 vs. adFBLN7+D-gal+DMSO, #: adFBLN7+D-gal+GRO100 vs. adFBLN7+D-gal+DMSO, %: adFBLN7+D-gal+GRO200 vs. adFBLN7+D-gal+DMSO. B) Schedule for FBLN7 overexpression and GRO supplementation in mice. C) Gross appearances of mice and their hearts in each group. D) Representative echocardiography images of the two groups of mice. Upper: B-mode; middle: PW Doppler mode; lower: tissue Doppler mode. Echocardiographic analysis of interventricular septal thickness in diastole (IVSd) and systole (IVSs), left ventricular posterior wall thickness in diastole (LVPWd) and systole (LVPWs), left ventricular mass, the ratio of early to late diastolic filling velocities (E/A), and the ratio of early diastolic filling velocity to left atrial pressure (E/e'), left ventricular ejection fraction (LVEF) and fractional shortening (FS%) are shown at the bottom. E) Representative images of H&E staining of hearts from aging mice (18 months old) injected with AAV-FBLN7 and administered 40 mg/kg/d of GRO or solvent control via gavage. F) Representative WGA staining images of the two groups of mice, with quantification shown on the right. G) Representative micrographs of Masson staining in heart sections from the AAV-FBLN7-18M-GRO and AAV-FBLN7-18M-CTL mice, displaying three different views: upper: perivascular; middle: global; lower: interstitial. Quantifications of fibrotic areas are presented on the right. H) FBLN7 exerts its reversing effect on the impaired profibrotic phenotypes of senescent CFs by downregulating the TGFBR3/ALK1/Smad1 pathway, which may contribute to the pathogenesis of age-related cardiac fibrosis. GRO exerts antifibrotic effects possibly through interaction with FBLN7. Error bars represent the standard error of the mean (SEM). Statistical significance is indicated as follows: *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.