Effects of emodin, a plant-derived anthraquinone, on TGF-β1-induced cardiac fibroblast activation and function

Abstract

Cardiac fibrosis accompanies a number of pathological conditions and results in altered myocardial structure, biomechanical properties and function. The signaling networks leading to fibrosis are complex, contributing to the general lack of progress in identifying effective therapeutic approaches to prevent or reverse this condition. Several studies have shown protective effects of emodin, a plant-derived anthraquinone, in animal models of fibrosis. A number of questions remain regarding the mechanisms whereby emodin impacts fibrosis. Transforming growth factor beta 1 (TGF-β1) is a potent stimulus of fibrosis and fibroblast activation. In the present study, experiments were performed to evaluate the effects of emodin on activation and function of cardiac fibroblasts following treatment with TGF-β1. We demonstrate that emodin attenuates TGF-β1-induced fibroblast activation and collagen accumulation in vitro. Emodin also inhibits activation of several canonical (SMAD2/3) and noncanonical (Erk1/2) TGF-β signaling pathways, while activating the p38 pathway. These results suggest that emodin may provide an effective therapeutic agent for fibrosis that functions via specific TGF-β signaling pathways.

Keywords: TGF-β; emodin; fibroblast; fibrosis.

Fig. 1.

This illustrates relative quantification of the effects of TGF-β1 and emodin on collagen type I and collagen type III protein accumulation in conditioned medium of cardiac fibroblasts as determined by immunoblot analyses. Figure 1A illustrates representative immunoblots stained with fast green to verify even protein loading (upper panel) and blots probed with type I collagen (middle panel) and type III collagen (lower panel) antisera. Graphic analyses of collagen type I (Figure 1B) and collagen type III (Figure 1C) are shown. (* p<0.05 relative to untreated controls, # p <0.05 relative to TGF-β1 treatment alone as determined by ANOVA, n = 4).

Fig. 2.

This figure demonstrates the effects of TGF-β1 and emodin on 3-dimensional collagen hydrogel contraction. Data are presented as the size of the hydrogel perimeter relative to the starting size (shorter bars indicate greater contraction). The inset shows representative images of collagen hydrogels. (* p<0.05 relative to untreated controls, # p <0.05 relative to TGF-β1 treatment alone as determined by ANOVA, n = 4).

Fig. 3.

This demonstrates the effects of TGF-β1 and emodin on fibroblast migration (Figure 3A) and proliferation (Figure 3B). Data are presented as fold of the untreated controls. (* p <0.05 relative to untreated controls, # p <0.05 relative to TGF-β1 treatment alone as determined by ANOVA, n = 4).

Fig. 4.

This figure illustrates the quantitative effects of TGF-β1 and emodin on the expression of α-smooth muscle actin (Figure 4B) and fibroblast activation protein (Figure 4C) levels as determined by immunoblots of cellular lysates. Data are presented as fold of the untreated controls. (* p <0.05 relative to untreated controls, # p <0.05 relative to TGF-β1 treatment alone as determined by ANOVA, n = 4). Images of representative immnoblots are shown in Figure 4A (α-SMA is α-smooth muscle actin, FAP is fibroblast activation protein and GAPDH is glyceraldehyde phosphate dehydrogenase).

Fig. 5.

This figure demonstrates the effects of TGF-β1 and emodin on the levels of α1 integrin (Figure 5B), α2 integrin (Figure 5C) and β1 integrin (Figure 5D) protein levels as determined by immunoblots of cellular lysates. Data are presented as fold of the untreated controls. (* p <0.05 relative to untreated controls, # p <0.05 relative to TGF-β1 treatment alone as determined by ANOVA, n = 4). Images of representative immunoblots are shown in Figure 5A.

Fig. 6.

Graphic representation of the effects of TGF-β1 and emodin on the activation of TGF-β signaling components including SMAD 2 (Figure 6A), SMAD 3 (Figure 6B), Erk 1/ 2 (Figure 6C) and p38 (Figure 6D). The insets illustrate representative immunoblots with the respective signaling component antisera. Data are presented as the signal with antisera to the phosphorylated protein relative to antisera to total protein. (* p <0.05 relative to untreated controls, # p <0.05 relative to TGF-β1 treatment alone as determined by ANOVA, n = 3).