Matricellular protein CCN1 promotes collagen alignment and scar integrity after myocardial infarction

Abstract

Background: Members of the cellular communication network family (CCN) of matricellular proteins, like CCN1, have long been implicated in the regulation of cellular processes underlying wound healing, tissue fibrogenesis, and collagen dynamics. While many studies suggest antifibrotic actions for CCN1 in the adult heart through the promotion of myofibroblast senescence, they largely relied on exogenous supplementation strategies in in vivo models of cardiac injury where its expression is already induced-which may confound interpretation of its function in this process. The objective of this study was to interrogate the role of the endogenous protein on fibroblast function, collagen structural dynamics, and its associated impact on cardiac fibrosis after myocardial infarction (MI).

Methods/results: Here, we employed CCN1 loss-of-function methodologies, including both in vitro siRNA-mediated depletion and in vivo fibroblast-specific knockout mice to assess the role of the endogenous protein on cardiac fibroblast fibrotic signaling, and its involvement in acute scar formation after MI. In vitro depletion of CCN1 reduced cardiac fibroblast senescence and proliferation. Although depletion of CCN1 decreased the expression of collagen processing and stabilization enzymes (i.e., P4HA1, PLOD1, and PLOD2), it did not inhibit myofibroblast induction or type I collagen synthesis. Alone, fibroblast-specific removal of CCN1 did not negatively impact ventricular performance or myocardial collagen content but did contribute to disorganization of collagen fibrils and increased matrix compliance. Similarly, Ccn1 ablated animals subjected to MI showed no discernible alterations in cardiac structure or function one week after permanent coronary artery ligation, but exhibited marked increases in incidence of mortality and cardiac rupture. Consistent with our findings that CCN1 depletion does not assuage myofibroblast conversion or type I collagen synthesis in vitro, Ccn1 knockout animals revealed no measurable differences in collagen scar width or mass compared to controls; however, detailed structural analyses via SHG and TEM of scar regions revealed marked alterations in their scar collagen topography-exhibiting changes in numerous macro- and micro-level collagen architectural attributes. Specifically, Ccn1 knockout mice displayed heightened ECM structural complexity in post-MI scar regions, including diminished local alignment and heightened tortuosity of collagen fibers, as well as reduced organizational coherency, packing, and size of collagen fibrils. Associated with these changes in ECM topography with the loss of CCN1 were reductions in fibroblast-matrix interactions, as evidenced by reduced fibroblast nuclear and cellular deformation in vivo and reduced focal-adhesion formation in vitro; findings that ultimately suggest CCN1's ability to influence fibroblast-led collagen alignment may in part be credited to its capacity to augment fibroblast-matrix interactions.

Conclusions: These findings underscore the pivotal role of endogenous CCN1 in the scar formation process occurring after MI, directing the appropriate arrangement of the extracellular matrix's collagenous components in the maturing scar-shaping the mechanical properties that support its structural stability. While this suggests an adaptive role for CCN1 in regulating collagen structural attributes crucial for supporting scar integrity post MI, the long-term protracted expression of CCN1 holds maladaptive implications, potentially diminishing collagen structural complexity and compliance in non-infarct regions.

Keywords: Acute myocardial infarction; Cardiac fibroblast; Cardiac fibrosis; Cellular communication network; Collagen alignment; Extracellular matrix; Fibroblast-matrix interactions.

Figure 1:. Elevated CCN1 expression in acutely injured and chronically failing hearts.

A. Whole heart Ccn1 mRNA expression (n=3, RNA sequencing from Kim et al. [Accession: GSE114695]) and B–C region-specific cardiac CCN1 protein levels (n=5–6 each, 2–3 males each) in post-myocardial infarction mice. D. UMAP and E. dot plot of single-cell RNA sequencing analysis of Ccn1 expression in dilated cardiomyopathy patients (n=18) and healthy donors (n=27), using data from Koenig et al. [Accession: GSE183852]. Graphs depict arithmetic mean ± SEM. Statistical procedures: normality was assessed in A–C via Shapiro-Wilk test. A two-way ANOVA and B–C one-way ANOVA with a Holm-Sidak’s post hoc multiple comparisons test. D Wilcoxon rank-sum method.

Figure 2:. Reductions in fibroblast proliferation and senescence following CCN1 knockdown.

A. CCN1 immunoblot performed on murine cardiac fibroblasts transfected with siRNA targeting CCN1 (siCCN1) or non-target control (siNT) (n=5 each; 3 males, 2 females). GAPDH served as a loading control B. Graph illustrating PrestoBlue viability assays conducted on cardiac fibroblasts transfected with either siCCN1 or siNT (n=4 each; 2 males, 2 females) over a 72-hour period (24 h intervals). The vertical axis reports relative fluorescence units (RFU). C. Representative fluorescent images and accompanying quantification of nuclear incorporated EdU in siCCN1- or siNT-transfected cardiac fibroblasts (n=6 each; 5 males, 1 female). Cells were counterstained with DAPI to label nuclei. D. Representative brightfield images of senescence-associated beta-galactosidase (SA-β-gal)-stained siCCN1- or siNT-transfected cardiac fibroblasts (n=6 each; 5 males, 1 female). All graphs depict arithmetic mean ± SEM. Statistical procedures: normality was assessed via Shapiro-Wilk test. A Mann-Whitney U test, B repeated measures two-way ANOVA followed by Holm-Sidak’s multiple comparisons test, and C–D two-tailed unpaired Student’s t-test.

Figure 3:. CCN1 knockdown does not impede fibroblast activation or type I collagen synthesis.

A. Representative immunoblots and quantification of B. CCN1, C. COL1A1, and D. ⍺-SMA in murine cardiac fibroblasts stimulated with increasing concentrations of TGFβ1 for 48 h (n=5 each; 2 males, 3 females). E. Representative immunoblots and F. quantification of CCN1 knockdown, fibroblast activation markers, and collagen synthetic/stabilization enzymes in TGFβ1 stimulated (10 ng/ml for 48 h) murine cardiac fibroblasts transfected with siCCN1 or siNT control (n=5 each; 3 males, 2 females). GAPDH served as a loading control. All graphs depict arithmetic mean ± SEM. Statistical procedures: normality was assessed via Shapiro-Wilk test. B and D one-way ANOVA with Holm-Sidak’s multiple comparisons test, C non-parametric Kruskal-Wallis test with Dunn’s multiple comparisons test, and F two-tailed unpaired Student’s t-test.

Figure 4:. Fibroblast-specific CCN1 knockout mice do not show impediments in cardiac function, but exhibit alterations in ECM structure and biophysics.

A. Confirmation of CCN1 ablation in isolated cardiac fibroblasts from Ccn1^fl/fl and Ccn1^−/− mice two weeks post TAM injection. B. Echocardiographic assessment of LVEF in Ccn1^−/−, Ccn1^fl/fl, and transgene only (Col1a2-Cre) control mice prior to TAM (BSL), and one (1 mo)- and two-months (2 mo) post TAM injection [Ccn1^−/−: n=4 (4 males), Ccn1^fl/fl: n=5 (4 males, 1 female), Col1a2-Cre: n=4 (3 males, 1 female)]. C. Decellularized cardiac ECM segments from Ccn1^−/− (n=3; 3 males) and Ccn1^fl/fl (n=4; 3 males, 1 female) mice, two months post TAM, were subjected to incremental stretching of 0.5 mm on a myograph instrument, and their internal tension forces were recorded and plotted. The interaction (p=0.0013), stretch distance (p<0.0001), and genotype (p<0.0008) were analyzed. Inset Westerns show corresponding CCN1 levels in fibroblasts from mice two months post TAM. D. Myocardial hydroxyproline levels in Ccn1^−/− (n=3; 2 males, 1 female) and Ccn1^fl/fl (n=3; 1 male, 2 females) mice two months post TAM. E. Representative TEM of decellularized myocardial ECM from Ccn1^−/− (n=3; 2 males, 1 female) and Ccn1^fl/fl (n=3; 2 males, 1 female) mice two months post TAM. All graphs depict arithmetic mean ± SEM. Statistical procedures: data distribution was evaluated via a Shapiro-Wilk normality test. B Repeated measures two-way ANOVA with Holm-Sidak’s multiple comparisons test, C two-way ANOVA with Bonferroni multiple comparisons test, and D two-tailed unpaired Student’s t-test.

Figure 5:. Fibroblast-specific removal of CCN1 prior to MI leads to higher mortality and incidence of cardiac rupture.

A. Kaplan-Meier survival curves for Ccn1^−/− (n=30; 19 males, 11 females) and Ccn1^fl/fl (n=21; 13 males, 8 females) with TAM injection three weeks prior to MI. Summary table reports number of post-MI deaths and autopsy results. B. Representative image of cardiac rupture (arrow denotes site of rupture). C. Echocardiographic-based measures of cardiac structure [end-diastolic volume (EDV) and end-systolic volume (ESV)] and function [ejection fraction] in Ccn1^fl/fl (n=19; 11 males, 8 females) and Ccn1^−/− (n=14; 9 males, 5 females) mice one week after MI. D. Gravimetric analyses of post-MI hearts normalized to tibia lengths in Ccn1^fl/fl (n=15; 8 males, 7 females) and Ccn1^−/− (n=9; 4 males, 5 females) mice one week after MI. E–F. Representative images and quantification of cardiomyocyte cross-sectional areas in myocardial tissue sections from both Ccn1^fl/fl (n=15; 8 males, 7 females) and Ccn1^−/− (n=9; 4 males and 5 females) mice, obtained one week after MI. Sections were stained with wheat germ agglutinin (WGA; cell borders) and anti-α-sarcomeric actin antibody (α-SA; cardiomyocytes), and counter-stained with DAPI (cell nuclei). G. Picrosirius red staining of transverse myocardial sections (spanning apex to base) used in the determination of scar H. mass and I. width in Ccn1^fl/fl (n=15; 8 males, 7 females) and Ccn1^−/− (n=9; 4 males and 5 females) mice one week post MI. All graphs depict arithmetic mean ± SEM. Statistical procedures: data distribution was evaluated via a Shapiro-Wilk normality test. A log-rank test. C [EF], D, and F two-tailed unpaired Student’s t-test. C [EDV], C [ESV], H, and, I Mann-Whitney U test.

Figure 6:. CCN1 deletion reduces macrostructural collagen alignment and fiber linearity in post-infarct cardiac scars.

A. Representative second-harmonic generation (SHG) microscopy images depicting infarct scar regions in Ccn1^fl/fl and Ccn1^−/− mice. Images are of high resolution (4506 × 4506 pixels) and were taken one week after permanent ligation. B. Random regions of interest (ROIs: 450 × 450 pixels) within the SHG images, focusing on the longitudinal aspect of collagen fibers, were selected. These ROIs underwent computational analysis through CurveAlign/CT-FIRE, enabling the quantitative evaluation of macrostructural attributes of the collagen network. C. CurveAlign-generated images revealing the spatial distribution (indicated by red dots) and directional orientation (depicted by green lines) of individual fiber segments within the collagen matrix. D. Corresponding heatmaps depicting the relative alignment of collagen fibers across the region. Intensely aligned areas are highlighted in red, emphasizing regions of structural coherence. E. CT-FIRE computational reconstructed image of analyzed ROIs, showcasing the arrangement of extracted collagen fibers. F. CT-FIRE program extracted collagen fibers overlaid onto the cropped ROI images, providing a holistic view of overall fiber architecture. Fibers are differentiated by varying colored lines, enhancing visual contrast. Graphs illustrating mean G. fibrillar collagen alignment, H. straightness, and I. width ± SEM in Ccn1^fl/fl and Ccn1^−/− animals. Data are representative of analyses performed on six distinct ROIs from n=3 animals per genotype. Statistical procedures: Data normality was assessed using the Shapiro–Wilk test. Data in G were analyzed using the non-parametric Mann–Whitney U test. Data in H and I were analyzed by an unpaired, two-tailed Student’s T-test.

Figure 7:. CCN1 ablation perturbs collagen microstructural organization and reduces fibril diameter in cardiac scars post-MI.

A. Representative global view and B. magnified transmission electron micrographs (TEM) of decellularized scars extracted from Ccn1^fl/fl and Ccn1^−/− mice, one week after MI. Asterisk (*) denotes suspected collagen degradation products. C. Corresponding histograms illustrating the distribution of collagen fibril diameters (number of fibrils vs. diameter [bin center; nm]) in Ccn1^fl/fl and Ccn1^−/− mice subjects. Bin width set at 10 nm. D. Box and whisker plot reporting the median fibril diameter (interquartile range) derived from both experimental groups. Inset legend below the graph provides details on mean fibril diameter ± standard deviation, along with the total fibril count per group. Analysis was conducted using data obtained from pooled biological samples, with n=3 distinct animals per genotype (all males). Statistical procedures: data distribution was evaluated using the D’Agostino & Pearson normality test, and significance was determined using the Mann-Whitney U test.

Figure 8:. Reduced fibroblast-matrix interactions with CCN1 deficiency.

A. Representative immunohistochemical images of α-SMA⁺ myofibroblasts located within scar regions of Ccn1^fl/fl and Ccn1^−/− mice one week after MI. Transverse myocardial sections were stained with antibodies against α-smooth muscle actin (α-SMA; myofibroblasts) and ⍺-sarcomeric actin (⍺-SA; cardiomyocytes), and counterstained with DAPI (nuclei). Images were used to measure and enumerate myofibroblast morphological features, including B. cell circularity, C. nuclear circularity, D. cell area, and E. cell perimeter, of Ccn1^fl/fl (n=5; 2 males, 3 females) and Ccn1^−/− mice (n=5; 3 males, 2 females). F. Representative brightfield (top) and immunocytochemical (ICC) images depicting focal adhesion formation in cultured murine cardiac fibroblasts transfected with siRNA targeting CCN1 (siCCN1) or non-target (siNT) control. Cardiac fibroblasts were stained with anti-paxillin antibody (focal adhesions), phalloidin (F-actin, cytoskeleton), and counterstained with DAPI (nuclei). G. Quantification of total number of focal adhesions in siCCN1- and siNT-transfected murine cardiac fibroblasts (n=7 each, all males). H.–I. Accompanying Westerns and quantification of phospho-FAK (pFAK) in siRNA-transfected cardiac fibroblasts (n=5 each; 3 males, 2 females). Total FAK served as loading control. J. Graphical interpretation of research findings (described in text). All graphs depict arithmetic mean ± SEM. Statistical procedures: data distribution was evaluated via a Shapiro-Wilk normality test. B‒F analyzed by an unpaired, two-tailed Student’s t-test and G by a Mann-Whitney U test.