Remodeling of cardiac fibroblasts following myocardial infarction results in increased gap junction intercellular communication

Abstract

Background: We have recently shown that native murine ventricular fibroblasts express both connexin43 (Cx43) and Cx45, and that the level of Cx43 expression influences intercellular coupling and cell proliferation. Relatively little is known, however, about how myocardial infarction (MI) influences expression of Cx43, or how altered Cx43 expression may affect fibroblast function post-MI. Fibroblasts are critical for infarct healing and post-infarct ventricular remodeling. They can couple electrically with cardiac myocytes and influence myocardial activation patterns. Thus, Cx43 remodeling and the level of intercellular communication in fibroblasts expressed in the infarcted heart were the subject of the present investigation.

Methods: Fibroblasts were isolated from both infarct scar and remote, noninfarcted regions of murine hearts 6 d after coronary ligation. Expression levels of Cx43, α-smooth muscle actin and N-cadherin were quantified by immunoblotting. Gap junctional intercellular communication was quantified by Lucifer yellow dye transfer.

Results and conclusions: Fibroblasts isolated from infarcted hearts exhibited marked up-regulation of Cx43 protein expression and enhanced intercellular coupling. Exogenous administration of transforming growth factor-β (TGF-β) to fibroblast cultures from normal, non-operated hearts produced comparable up-regulation of Cx43, suggesting that increased intercellular communication between fibroblasts in infarct and peri-infarct regions may be secondary to activation of a TGF-β pathway. Unlike cardiac myocytes that down-regulate Cx43, presumably to limit intercellular transmission of biochemical mediators of ischemic injury, fibroblasts may up-regulate Cx43 to maintain electrical and metabolic coupling at a time when intercellular communication is compromised.

Figure 1

Top: α-SMA immunoreactive signal demonstrating increased myofibroblasts in a culture of fibroblasts isolated from the infarct region of a heart subjected to coronary ligation (right) compared to cells isolated from a sham-operated heart (left). Bar, 100 μm. Bottom: Representative immunoblots of α-SMA and actin protein expression in fibroblasts isolated from sham-operated hearts and from noninfarcted and infarct regions of hearts subjected to MI showing increased α-SMA protein expression in the fibroblasts isolated from infarcted hearts.

Figure 2

Representative immunoblots of CD-31, smoothelin, vimentin and actin protein expression in cultured fibroblasts isolated from sham-operated hearts (lanes 1 and 2), noninfarcted (lanes 3 and 4) and infarct (lanes 5 and 6) regions of hearts subjected to MI, and tissue homogenates of mouse aorta (ao, lane 7), intestine (int, lane 8) and lung (lane 9) demonstrating that our fibroblast cultures (lanes 1–6) are devoid of measurable endothelial and smooth muscle cell contamination. The endothelial cell marker CD-31, or PECAM-1, is expressed in lung tissue; the smooth muscle cell marker smoothelin is present in aorta and intestine. The nonspecific marker, vimentin, is seen in all lanes.

Figure 3

Top: Representative immunoblots of Cx43, Cx45, N-cadherin (probed with an anti-pan cadherin antibody), and actin expression in fibroblasts isolated from sham-operated hearts and from noninfarcted and infarcted regions of hearts subjected to coronary ligation. Equal protein (30 μg) was loaded in each lane. Cx43, Cx45 and N-cadherin band densities were normalized to actin band densities to control for protein loading. Bottom: Histograms of summarized immunoblot data showing that Cx43 expression was increased significantly in fibroblasts from both noninfarcted and infarct regions compared to those isolated from sham-operated hearts. *, p < 0.05.

Figure 4

Top: Fluorescence images showing Lucifer yellow dye transfer after scrape-loading in fibroblasts isolated from a sham-operated heart (left) and from the infarct region of a heart subjected to MI (right). Outlines beside each fluorescence image represent the areas of dye transfer on one side of the scrape line obtained using ImageJ software. The total area of dye transfer was divided by the length of the microscopic field analyzed (top to bottom) to calculate the mean distance of dye transfer in μm. Bar, 100 μm. Bottom: Histograms of summarized data showing that the distance of dye transfer in fibroblasts from both noninfarcted and infarct regions was significantly greater than that in fibroblasts isolated from sham-operated hearts. *, p < 0.05.

Figure 5

Top: Representative immunoblots showing increased expression of Cx43 and α-SMA in cultured fibroblasts isolated from normal (non-operated) hearts and exposed to exogenous TGF-β (10 ng/ml). Equal protein (30 μg) was loaded in each lane. Cx43 and α-SMA band densities were normalized to actin band densities to control for protein loading. Bottom: Histograms of summarized immunoblot data showing that both Cx43 and α-SMA protein expression were increased significantly in fibroblasts treated with TGF-β, whereas N-cadherin (N-cad) was unchanged. *, p < 0.05.