Curcumin attenuates angiogenesis in liver fibrosis and inhibits angiogenic properties of hepatic stellate cells

Abstract

Hepatic fibrosis is concomitant with sinusoidal pathological angiogenesis, which has been highlighted as novel therapeutic targets for the treatment of chronic liver disease. Our prior studies have demonstrated that curcumin has potent antifibrotic activity, but the mechanisms remain to be elucidated. The current work demonstrated that curcumin ameliorated fibrotic injury and sinusoidal angiogenesis in rat liver with fibrosis caused by carbon tetrachloride. Curcumin reduced the expression of a number of angiogenic markers in fibrotic liver. Experiments in vitro showed that the viability and vascularization of rat liver sinusoidal endothelial cells and rat aortic ring angiogenesis were not impaired by curcumin. These results indicated that hepatic stellate cells (HSCs) that are characterized as liver-specific pericytes could be potential target cells for curcumin. Further investigations showed that curcumin inhibited VEGF expression in HSCs associated with disrupting platelet-derived growth factor-β receptor (PDGF-βR)/ERK and mTOR pathways. HSC motility and vascularization were also suppressed by curcumin associated with blocking PDGF-βR/focal adhesion kinase/RhoA cascade. Gain- or loss-of-function analyses revealed that activation of peroxisome proliferator-activated receptor-γ (PPAR-γ) was required for curcumin to inhibit angiogenic properties of HSCs. We concluded that curcumin attenuated sinusoidal angiogenesis in liver fibrosis possibly by targeting HSCs via a PPAR-γ activation-dependent mechanism. PPAR-γ could be a target molecule for reducing pathological angiogenesis during liver fibrosis.

Keywords: VEGF; angiogenesis; curcumin; hepatic stellate cell; liver fibrosis; peroxisome proliferator-activated receptor-γ.

Fig. 1

Curcumin attenuates CCl₄-caused liver fibrosis in rats. Rats were grouped: group 1, vehicle control (no CCl₄, no treatment); group 2, model group (with CCl₄, no treatment); group 3, curcumin-treated group (100 mg/kg + CCl₄); group 4, curcumin-treated group (200 mg/kg + CCl₄); group 5, curcumin-treated group (400 mg/kg + CCl₄). (A) Liver sections were stained with haematoxylin and eosin, masson reagents and sirius red. (B) Measurement of hydroxyproline levels in liver and blood. (C) Real-time PCR analyses of α-SMA, α(I)procollagen, fibronectin and peroxisome proliferator-activated receptor-γ in liver tissues. (D) Western blot analyses of liver proteins with densitometry. For the statistics of each panel in this figure, ^#P < 0.05 versus group 1, ^##P < 0.01 versus group 1, *P < 0.05 versus group 2, **P < 0.01 versus group 2, n = 6.

Fig. 2

Curcumin alleviates sinusoidal angiogenesis in rats with CCl₄-caused fibrosis. Rats were grouped: group 1, vehicle control (no CCl₄, no treatment); group 2, model group (with CCl₄, no treatment); group 3, curcumin-treated group (100 mg/kg + CCl₄); group 4, curcumin-treated group (200 mg/kg + CCl₄); group 5, curcumin-treated group (400 mg/kg + CCl₄). (A) Liver sections were stained with immunofluorescence by using antibodies against CD31, vWF, CD34, VEGF-R2, PDGF-βR, and VEGF. (B) ELISA measurement of VEGF levels in liver and blood. (C) Real-time PCR analyses of HIF-1α, VEGF-R2, PDGF-βR, and VEGF in liver tissues. (D) Western blot analyses of liver proteins with densitometry. For the statistics of each panel in this figure, ^#P < 0.05 versus group 1, ^##P < 0.01 versus group 1, *P < 0.05 versus group 2, **P < 0.01 versus group 2, n = 6.

Fig. 3

Curcumin does not affect endothelial cells and aortic ring vascularization in vitro. (A and B) LSECs were treated with DMSO (0.02%, w/v), imatinib and curcumin for 24 or 48 hrs. Cell viability was evaluated by MTS assay (A). Lactate dehydrogenase activity in supernatant was assessed (B). (C) LSECs were incubated with DMSO (0.02%, w/v), imatinib and curcumin on matrigel for 24 hrs. Tubulogenesis was visualized and quantified. (D) Explanted aortic rings were placed on matrigel and subjected to DMSO (0.02%, w/v), imatinib and curcumin for 7 days (n = 6). The histogram shows quantification of aortic sprouts specifically from the endothelial cell lumen. Values are represented as arbitrary units. For the statistics of each panel in this figure, *P < 0.05 versus DMSO, **P < 0.01 versus DMSO.

Fig. 4

Curcumin interrupts PDGF-βR/ERK and PI3K/AKT/mTOR pathways linking to reduced VEGF expression in HSCs. (A–D) HSCs were treated with DMSO (0.02%, w/v), PDGF (20 ng/ml) and curcumin for 24 hrs. Real-time PCR analyses of VEGF mRNA (A). Western blot analyses of VEGF protein expression with densitometry (B). ELISA measurement of VEGF level in supernatant (C). Immunofluorescence by using antibody against VEGF (D). (E and F) HSCs were treated with DMSO (0.02%, w/v) and curcumin for 24 hrs prior to PDGF stimulation for an additional 3 hrs. Western blot analyses of PDGF-βR/ERK signals (E) and PI3K/AKT/mTOR signals (F). (G and H) HSCs were treated with DMSO (0.02%, w/v), imatinib (10 μM), U0126 (10 μM), rapamycin (10 nM) and curcumin (20 μM) for 24 hrs prior to PDGF stimulation for an additional 3 hrs. Real-time PCR analyses of VEGF mRNA (G). Western blot analyses of VEGF protein expression with densitometry (H). (I) LSECs were incubated with conditioned media from HSCs treated with DMSO (0.02%, w/v), PDGF (20 ng/ml), imatinib (10 μM), U0126 (10 μM), rapamycin (10 nM) and curcumin (20 μM) for 24 hrs. Tubulogenesis was visualized and quantified. For the statistics of each panel in this figure, ^#P < 0.05 versus DMSO, ^##P < 0.01 versus DMSO, *P < 0.05 versus DMSO + PDGF, **P < 0.01 versus DMSO + PDGF.

Fig. 5

Curcumin interrupts PDGF-βR/FAK/RhoA pathway linking to inhibited HSC invasion and vascularization. (A) HSC invasion for 24 hrs was evaluated by Boyden chamber assay with DMSO (0.02%, w/v) and curcumin in the upper well and PDGF in the lower well. (B–D) HSCs were treated with DMSO (0.02%, w/v), imatinib (B), Y15 (C), or curcumin (D) for 24 hrs prior to PDGF stimulation for an additional 3 hrs. Western blot analyses of FAK/RhoA signals. (E) HSC invasion for 24 hrs was evaluated by Boyden chamber assay with DMSO (0.02%, w/v), imatinib (10 μM), Y15 (10 μM), fasudil (10 μM) and curcumin (20 μM) in the upper well and PDGF in the lower well. (F) HSCs were incubated with DMSO (0.02%, w/v), PDGF (20 ng/ml), imatinib (10 μM), Y15 (10 μM), fasudil (10 μM) and curcumin (20 μM) on matrigel for 24 hrs. Tubulogenesis was visualized and quantified. For the statistics of each panel in this figure, ^#P < 0.05 versus DMSO, *P < 0.05 versus DMSO + PDGF.

Fig. 6

Peroxisome proliferator-activated receptor-γ activation is required for curcumin inhibition of angiogenic properties of HSCs. (A and B) HSCs were treated with DMSO (0.02%, w/v), curcumin, 15d-PGJ2 (A), or PD68235 or transfected with siRNA (B) for 24 hrs prior to PDGF stimulation for an additional 3 hrs. Western blot analyses of PDGF-βR/ERK and mTOR signals. (C and D) HSCs were treated with DMSO (0.02%, w/v), PDGF, curcumin, 15d-PGJ2 (10 μM), PD68235 (10 μM) or transfected with siRNA for 24 hrs. Real-time PCR analyses of VEGF mRNA (C). Western blot analyses of VEGF protein expression with densitometry (D). (E) LSECs were incubated with conditioned media from HSCs treated with DMSO (0.02%, w/v), PDGF, curcumin, 15d-PGJ2 (10 μM), PD68235 (10 μM), or transfected with siRNA for 24 hrs. Tubulogenesis was visualized and quantified. (F and G) HSCs were treated with DMSO (0.02%, w/v), curcumin, 15d-PGJ2 (F), or PD68235 or transfected with siRNA (G) for 24 hrs prior to PDGF stimulation for an additional 3 hrs. Western blot analyses of FAK/RhoA signals. (H) HSC invasion for 24 hrs was evaluated by Boyden chamber assay with DMSO (0.02%, w/v), curcumin, 15d-PGJ2 (10 μM), PD68235 (10 μM), or transfection with siRNA in the upper well and PDGF in the lower well. (I) HSCs were treated with DMSO (0.02%, w/v), PDGF, curcumin, 15d-PGJ2 (10 μM), PD68235 (10 μM), or transfected with siRNA for 24 hrs. Tubulogenesis was visualized and quantified. For the statistics of each panel in this figure, ^#P < 0.05 versus DMSO, *P < 0.05 versus DMSO + PDGF, ^&P < 0.05 versus curcumin + PDGF.

Fig. 7

Schema of the underlying mechanism of curcumin inhibition of angiogenic properties of HSCs. Curcumin activates peroxisome proliferator-activated receptor-γ possibly leading to transrepression of PDGF-βR, which disrupts ERK and PI3K/AKT/mTOR pathways and thereby inhibits VEGF mRNA and protein expression in activated HSCs. Inhibition of PDGF-βR also blocks FAK/RhoA cascade resulting in reduced HSC motility. These actions in concert attenuate the angiogenic effects of HSCs. The identified mechanism possibly accounts for curcumin attenuation of HSC-based pathological angiogenesis in liver fibrosis.