Myocardial contraction and hyaluronic acid mechanotransduction in epithelial-to-mesenchymal transformation of endocardial cells

Abstract

Epithelial-to-mesenchymal transition (EMT) of endocardial cells is a critical initial step in the formation of heart valves. The collagen gel in vitro model has provided significant information on the role of growth factors regulating EMT but has not permitted investigation of mechanical factors. Therefore we sought to develop a system to probe the effects of mechanical inputs on endocardial EMT by incorporating hyaluronic acid (HA), the primary component of endocardial cushions in developing heart valves, into the gel assay. This was achieved using a combination collagen and crosslinkable methacrylated HA hydrogel (Coll-MeHA). Avian atrioventricular canal explants on Coll-MeHA gels showed increased numbers of transformed cells. Analysis of the mechanical properties of Coll-MeHA gels shows that stiffness does not directly affect EMT. Hydrogel deformation from the beating myocardium of explants directly led to higher levels of regional gel deformation and larger average strain magnitudes associated with invaded cells on Coll-MeHA gels. Inhibition of this contraction reduced EMT on all gel types, although to a lesser extent on Coll-MeHA gels. Using the system we have developed, which permits the manipulation of mechanical factors, we have demonstrated that active mechanical forces play a role in the regulation of endocardial EMT.

Keywords: Collagen; Epithelial-to-mesenchymal transition (EMT); Heart valve; Hyaluronic acid; Mechanical properties.

Fig. 1

EMT behaviors as a function of collagen and MeHA presentation. (A) Endocardial (EC) sheet sizes. (B) Quantification of EMT via counts of invaded cells. (C) Maximum invasion depth of invaded cells. (D) Viability of explants seeded on different gel compositions. (E) Proliferation of cells on collagen and Coll-MeHA gels as measured by BrdU. All data presented as average ± SEM and represents 3 technical replicates with n ≥ 10 biological replicates. *p < 0.05 vs. 0.12wt% collagen at same time point.

Fig. 2

Gel characterization. (A–F) SEM images of 0.12wt% collagen (A,B), 0.2wt% Coll + 0.5wt% MeHA (C,D), and 0.4wt% Coll + 0.5wt% MeHA (E,F). Scale bar = 300μm (A,C,E); Scale Bar = 15μm (B,D,F). (G) Modulus data of gels obtained via AFM. All data presented as average ± SEM with n ≥ 8 scans per composition on N ≥ 2 different gels. (H–I) Fluorescent images of avian AVC scanned by AFM. Scale bar = 100μm (H) and 25μm (I). Inset in (I) highlights scanned area. (J) 3d topographical map with modulus value overlaid. (K) Representative distributions of modulus vs. scan area. Inset numbers represent median values. N = 3 embryos. * p<0.05 vs. 0.12wt% collagen; @ p<0.01 vs. 0.2wt% Coll + 0.5wt% MeHA.

Fig. 3

Mechanical regulation of EMT. (A) Regional gel deformation for different presentations of collagen and MeHA. 7d data unavailable for some samples due to slow myocardium beating. All data presented as average ± SEM and represents 3 technical replicates with ≥ 4 biological replicates. (B) Strain field maps induced by beating myocardial overlaid with position of invaded cells (*) Scale bar = 200μm. (C) Histograms showing number of invaded cells as a function of local strain magnitudes. Inset in 1d is re-scaled to show detail. Numbers correspond to weighted average strain ± SEM with n ≥ 3 replicates (see A for color legend). *p < 0.05 vs. 0.12wt% collagen at same time point.

Fig. 4

Pharmacological inhibition of myocardial contraction. (A–B) Live(green)-Dead(red) staining on explants and cells seeded on 0.12wt% collagen with either no treatment (A) or 1.5mM tricaine (B). Inset in (A) shows negative control, 4% PFA treatment for 5min. (C) Quantification of myocardial contractions with tricaine treatment. (D) Regional gel deformation induced by explants treated with tricaine. (E) EC sheet areas for tricaine treated explants. (F) EMT quantification via number of invaded cells in the presence of tricaine. (G) Maximum invasion depth of transformed cells. All data presented as average ± SEM and represents 3 technical replicates with n ≥ 6 biological replicates. *p < 0.05 vs. 0.12wt% collagen at same timepoint. ^p < 0.05 vs. same gel type at same time point.

Fig. 5

Physical inhibition of myocardial contractions. (A) EC sheet sizes as a function of myocardium presence. Note: These EC sheet values include area covered by explant. (B) EMT quantification via invaded cell counts with or without myocardium. (C) Maximum invasion depth of transformed cells with or without myocardium. All data presented as average ± SEM and represents 3 technical replicates with n ≥ 10 biological replicates. *p < 0.05 vs. 0.12wt% collagen at same timepoint. ^p < 0.05 vs. same gel type at same time point. +p < 0.05 vs. “−Myo” on same gel type at same time point.