Hepatocyte-specific expression of constitutively active Alk5 exacerbates thioacetamide-induced liver injury in mice

Abstract

While Transforming growth factor-βs (Tgf-βs) have been known to play an important role in liver fibrosis through an activation of Hepatic Stellate Cells (HSC), their fibrotic role on hepatocytes in liver damage has not been addressed thoroughly. To shed more light on the hepatocyte-specific role of Tgf-β signaling during liver fibrosis, we generated transgenic mice expressing constitutively active Tgf-β type I receptor Alk5 under the control of albumin promoter. Uninjured mice with increased Tgf-β/Alk5 signaling in hepatocytes (caAlk5/Alb-Cre mice) did not show characteristics related to hepatocyte death, fibrosis and inflammation. When subjected to thioacetamide (TAA) treatment, caAlk5/Alb-Cre mice exhibited more severe liver injury, when compared to control littermates. After TAA administration for 12 weeks, an increase in pathological changes was evident in caAlk5/Alb-Cre livers, with higher number of infiltrating cells in the portal and periportal area. Immunohistochemistry for F4/80, myeloperoxidase and CD3 showed that there was an increased accumulation of macrophages, neutrophils and T-lymphocytes, respectively, in caAlk5/Alb-Cre livers. Coincidently, we observed an exacerbated liver damage as seen by increases in serum aminotransferase level and number of apoptotic hepatocytes in caAlk5/Alb-Cre mice. Sirius staining of collagen demonstrated that the fibrotic response was worsened in caAlk5/Alb-Cre mice. The enhanced fibrosis in mutant livers was associated with marked production of α-SMA-positive myofibroblast. Hepatic expression of genes indicative of HSC activation was greater in caAlk5/Alb-Cre mice. In conclusion, our data indicated that elevation of Tgf-β signaling via Alk5 in hepatocytes is not sufficient to induce liver pathology but plays an important role in amplifying TAA-induced liver damage.

Keywords: Pathology; Toxicology.

Fig. 1

Analysis of caAlk5 transgenes in liver of control and caAlk5/Alb-Cre mice. (A) PCR strategy to assess the DNA recombination of caAlk5 transgene. (B) Genomic DNA extracted from control and caAlk5/Alb-Cre livers were used for PCR analysis with pCAG and pTβRI primers. (C) Detection of caAlk5 mRNA by quantitative RT-PCR performed on total RNA from livers of control and caAlk5/Alb-Cre mice.

Fig. 2

Representative images of immunohistochemical staining for phospho-Smad2 and 3 (pSmad2/3). While pSmad2/3 immunoreactivity was barely detected in control liver (A), nucleus of hepatocytes were clearly stained with pSmad2/3 antibody in caAlk5/Alb-Cre livers (B). Immunochemistry with NovaRed substrate, counterstained with hematoxylin.

Fig. 3

Histochemical analysis of caAlk5/Alb-Cre mice. Hematoxylin and Eosin-stained liver sections obtained from control (A) and caAlk5/Alb-Cre mice (B). Sirius Red/Fast green-stained liver sections obtained from control (C) and caAlk5/Alb-Cre mice (D).

Fig. 4

Histological changes of TAA-induced liver injury in control and caAlk5/Alb-Cre mice. Following TAA exposure for 12 weeks, control liver developed fibrotic lesions that connected among centrilobular- centrilobular, centrilobular-portal and portal-portal area (I-L).The fibrotic lesions were populated by a number of small hematoxylin-positive stromal cells. While fibrous bridgings were also seen in liver of TAA-treated caAlk5/Alb-Cre mice, hematoxylin-positive cells were more marked along fibrotic lesions (M-P).

Fig. 5

Immunohistochemistry for macrophages: F4/80 (A and B), for neutrophils: myeoperoxidase (D and E) and for T lymphocytes: CD3 (G and H) in livers of control (A, D, G) and caAlk5/Alb-Cre mice (B, E, H) treated with TAA for 12 weeks. Quantification of F4/80-(C), myeloperoxidase-(F), and CD3-(I) positive cells in livers from saline-treated and TAA-treated mice. Data represent means ± SD, n = 5 mice per experimental group, ten 20x fields per mouse. There is a significant difference after TAA treatment when compared with genotype-specific saline control (*P < 0.05). There is a significant difference when compared with genotype-specific saline control and between genotypes after TAA treatment (**P < 0.05). Immunochemistry with NovaRed substrate, counterstained with hematoxylin. CV denotes the central vein. PT denotes the portal tract.

Fig. 6

Serum level of aspartate transaminase (AST) and alanine transaminase (ALT) in control and caAlk5/Alb-Cre mice treated with saline or TAA for 12 weeks. AST (A) and ALT (B) level was higher in caAlk5/Alb-Cre mice compared to those of control mice. Data are means ± SD of n = 5 per experimental group. There is significant difference compared with saline control group (*P < 0.05). There is a significant difference when compared with genotype-specific saline control and between genotypes after TAA treatment (**P < 0.05).

Fig. 7

Immunohistochemistry for cleaved caspase 3 (CC3) in liver sections from TAA-treated control and caAlk5/Alb-Cre mice. While CC3 positivity was infrequently seen in liver parenchyma of control mice (A and B), increased number of CC3-positive cells were observed in centrilobular area of caAlk5/Alb-Cre liver (C-D). (E) Quantification of CC#-positive hepatocytes at 12 weeks of saline or TAA treatment. There is a significant difference after TAA treatment when compared with genotype-specific saline control (*P < 0.05). There is a significant difference when compared with genotype-specific saline control and between genotypes after TAA treatment (**P < 0.05). CV denotes the central vein.

Fig. 8

Hepatic fibrosis after TAA exposure. After 6-week (A-D) and 12-week (E-H) treatment, mice were sacrificed and their livers were analyzed with Sirius red staining. Images are representative of n = 5 mice per experimental group and 10 fields per mouse. (I) Morphometric analysis of Sirius red density from mice exposed to saline or TAA. Data are means ± SD of n = 5 per experimental group. *P < 0.05, significantly different compared with saline control for the same genotype. There is a significant difference after TAA treatment when compared with genotype-specific saline control and between genotypes at the same time point (**P < 0.05).

Fig. 9

Immunohistochemistry for myofibroblast: α-smooth muscle actin (α-SMA) in liver sections from TAA-treated control and caAlk5/Alb-Cre mice. While α-SMA-positive cells were seen in the interface of portal area and liver parenchyma (A and B), more immunoreactivity for α-SMA were found in the fibrous septa of caAlk5/Alb-Cre livers (C and D). (E) Quantification of α-SMA-positive cells in liver sections from TAA-treated control and caAlk5/Alb-Cre mice. Data are means ± SD of n = 5 per experimental group. *P < 0.05, significantly different compared with saline control for the same genotype. There is a significant difference when compared with genotype-specific saline control and between genotypes after TAA treatment (**P < 0.05). CV denotes the central vein.

Fig. 10

Hepatic expression of fibrogenic factors after chronic TAA exposure. Relative mRNA expression of PDGF-A (A), PDGF-B (B), Tgf-β (C) and CTGF (D) was determined quantitative RT-PCR. Expression level was normalized to β-actin. Five mice from each experimental group were used for the analysis. There is a significant change in gene expression when compared with genotype-specific saline control (*P < 0.05). There is a significant difference when compared with genotype-specific saline control and between genotypes after TAA treatment (**P < 0.05).