Fig 1.
Fibromodulin is upregulated in hearts of patients with heart failure.
(A) FMOD mRNA expression in left ventricular biopsies from explanted hearts of end-stage dilated heart failure (HF) patients compared to controls (Cntr), n = 5–17. Ribosomal protein L32 (RPL32) was used as reference gene. (B) Representative immunoblots and (C) quantitative levels of extracellular FMOD (FMODext, ~60kDa) protein in HF patients and Cntr, n = 5–10. FMODext from cell culture medium of FMOD-transfected HEK293 cells was used as positive control (Pos cntr), see S1 Fig for details. Vinculin was used for loading control. HF patient characteristics are found in S3 Table. Data are shown as single data points and mean±SEM. Statistical differences were tested using an unpaired t-test, HF vs. Cntr.
Fig 2.
Fibromodulin is upregulated in hearts of mice in response to pressure overload.
(A) FMOD mRNA in the left ventricle (LV) of wild-type (WT) mice subjected to aortic banding (AB) for 24 hours (h) -18 weeks (w), compared to sham-operated controls, n sham = 3–10, n AB = 7–10. Ribosomal protein L32 (RPL32) was used as reference gene. (B) Representative immunoblots and (C) quantitative levels of extracellular FMOD (FMODext, ~60kDa) protein in mouse LVs 3w and 16w post-AB or sham operation, n = 5–7. Culture medium of FMOD-expressing HEK293 cells was used as positive control (Pos cntr), see S1 Fig for details. Coomassie staining was used for loading control. (D and E) Pearson regression analysis of FMOD mRNA and (D) left ventricular weight (LVW)/tibia length, or (E) lung weight (LW)/tibia length in sham- and AB-operated mice, n = 87. (F) FMOD, lumican (LUM), biglycan (BGN) and decorin (DCN) mRNA normalized to RPL32 in LVs of mice 24h-18w post-AB, relative to the average of sham-operated controls, n sham = 3–10, n AB = 7–10. Mouse characteristics are found in S4 Table. Data are shown as mean±SEM. Statistical differences were tested using an unpaired t-test vs. respective sham-operated controls (A and C), Pearson regression analysis (D and E) and one-way ANOVA with Bonferroni post-hoc test vs. FMOD mRNA (F). *p≤0.05; ** p≤0.01; ***p≤0.005.
Fig 3.
Fibromodulin is expressed in cardiomyocytes and cardiac fibroblasts and regulated by inflammatory pathways.
(A) FMOD mRNA in cultured cardiomyocytes (CM) and cardiac fibroblasts (CFB) from neonatal rats, n = 10–11. Ribosomal protein L4 (RPL4) was used as reference gene. (B) Quantitative levels of extracellular FMOD (FMODext) protein in culture medium from CM and CFB, n = 6–10, with (C) representative immunoblots. (D) CM and (E) CFB cultures were treated with the innate immunity mediator lipopolysaccharide (LPS) or the nuclear factor (NF)κB-inhibitor SM7368, alone or in combination. CM, n = 9–18, and CFB, n = 10–19. Data are shown as mean±SEM. Statistical differences were tested using an unpaired t-test, CM vs. CFB (A-B), *p≤0.05; ***p≤0.005, and (D-E) one-way ANOVA with Holm-Sidak post-hoc test vs. Cntr or co-treatment with SM7368 vs. LPS alone, *** p≤0.005.
Fig 4.
FMOD-KO mice show mildly exacerbated hypertrophic remodeling compared to wild-type, associated with increased ERK signaling upon aortic banding.
Cardiac phenotype assessed by serial echocardiography at 2–12 weeks (w) (A-B and D-E, n sham = 5–10, n AB = 5–24) and organ weights at 2w and 12w (C and F, n sham = 5–14, n AB = 9–14) post aortic banding (AB). (A) Interventricular septum thickness in diastole (IVSd), (B) left ventricular posterior wall thickness in diastole (LVPWd), (C) heart weight-to-body weight (HW/BW) ratio, (D) LV interior diameter in diastole (LVIDd), (E) fractional shortening (FS) and (F) lung weight-to-body weight (LW/BW) ratio in fibromodulin knock-out (FMOD-KO) and wild-type (WT) control mice subjected to AB. (G) Representative histology image of a whole heart section at low magnification(scale bar 1000μm) and insets showing representative histology images at high magnification (scale bar 50μm) of the different groups at 12w post operation, green (WGA staining) shows outline of cardiomyocyte cross sectional area (CSA) and blue (DAPI staining) shows nucleus. (H) Quantitative changes in cardiomyocyte (CM) cross-sectional area (CSA) respective to average sham controls, measured on midventricular sections of FMOD-KO and WT mice 2, 4, and 12w post-AB, n = 2191–14485 CMs from n sham = 1–4 mice, and n AB = 2–4 mice. (I) Representative immunoblots and (J) quantitative phosphorylated (p) and total levels of the extracellular signal–regulated kinase 1 and 2 (ERK1 and ERK2, respectively) in LVs of FMOD-KO and WT mice 4w post-AB (n sham = 2, n AB = 3–5). Coomassie staining was used as loading control. Data are shown as mean±SEM. Statistical differences were tested using an unpaired t-test vs. WT AB, #p≤0.05; ###p≤0.005 (A, B, D and E 4-10w, H and J), one-way ANOVA with Dunn's post-hoc test vs. WT sham, *p≤0.05; **p≤0.01; ***p≤0.005, or vs. WT AB (A-F 2w and 12w).
Fig 5.
FMOD-KO and wild-type mice have comparable levels of fibrillar collagens and collagen cross-linking in vivo, but fibromodulin decreases the migration and alters the expression of fibrosis-associated genes in cultured cardiac fibroblasts in vitro.
(A) COL1A2 and COL3A1 mRNA in the left ventricle (LV) of fibromodulin knock-out (FMOD-KO) and wild-type (WT) mice at baseline, 2, 4, or 12 weeks (w) post-aortic banding (AB), relative to average of sham-operated controls (n sham = 2–10, n AB = 3–8). Ribosomal protein L32 (RPL32) was used as reference gene. (B) Collagen protein levels assessed by HPLC from whole LV, in mice 2w and 12w post-AB or sham-operation, n sham = 5, n AB = 5–6. (C) Amount of collagen cross-linking quantified from short axis midventricular histological sections of hearts of FMOD-KO and WT mice 2, 4 and 12w after sham or AB operation, n sham = 5–7, n AB = 2–4. (D) mRNA expression of collagens (COL1A2, COL3A1), cell signature proliferation marker PCNA, myofibroblast differentiation signature marker ACTA2, and fibrosis-associated transcripts in cardiac fibroblast (CFB) cultures from neonatal rats transduced with adenovirus encoding FMOD (AdFMOD), or adenovirus-vehicle (AdVeh) as control, n = 6–7. Ribosomal protein L4 (RPL4) was used as reference gene. (E) Representative images of CFB with AdVeh or AdFMOD 24h post-scratching, and (F) quantification of CFB migration 6-24h post-scratching, n = 8–10. Scale bar 250μm. Data are shown as mean±SEM. Statistical differences were tested using one-way ANOVA with Dunn's post-hoc test vs. WT sham, *p≤0.05; ***p≤0.005, or vs. WTAB (A, B and C), unpaired t-test vs. AdVeh, *p≤0.05; **p≤0.01 (D), or two-way ANOVA (F), #p≤0.05; ##p≤0.01; ###p≤0.005.
Fig 6.
Fibromodulin affects the cardiac immune response after pressure overload.
mRNA expression of immune cell adhesion molecules (ICAM1 and VCAM1) and signature molecules of macrophages (ADGRE1/F4.80), T-lymphocytes (CD3), or leukocytes (CD11a and CD45) in the left ventricle of FMOD knock-out (FMOD-KO) and wild-type (WT) mice 12 weeks (w) post-aortic banding (AB), relative to respective sham-operated controls, n sham = 10, n AB = 4–7. Ribosomal protein L32 (RPL32) was used as reference gene. Data are shown as mean±SEM. Statistical differences were tested using one-way ANOVA with Dunn's post-hoc test vs. WT sham, **p≤0.01, or vs. WT AB, #p≤0.05.
Fig 7.
Schematic illustration of the proposed role of fibromodulin in the failing heart.
The schematic illustrates the main findings of our study. (A) Fibromodulin (FMOD) expression in the heart is increased during cardiac remodeling and heart failure progression in patients and mice. (B) FMOD is an extracellular matrix (ECM) proteoglycan produced by both cardiomyocytes and cardiac fibroblasts, the two major cell types in the heart. (C) FMOD knock-out (FMOD-KO) mice show mildly exacerbated hypertrophic remodeling of the heart with attenuated infiltration of leukocytes in response to experimental pressure overload. (D) FMOD overexpression (FMOD-OE) in cultured cardiac fibroblasts decreases their migratory capacity and reduces the expression of fibrosis-related ECM molecules such as periostin (POSTN), transglutaminase 2 (TGM2), and the collagen-cross linking enzyme lysyl oxidase (LOX). Thus, we propose that FMOD has anti-hypertrophic, anti-fibrotic and pro-inflammatory effects in the pressure-overloaded heart.