Cholangiocyte endothelin 1 and transforming growth factor beta1 production in rat experimental hepatopulmonary syndrome

Abstract

Background & aims: Hepatic production and release of endothelin 1 plays a central role in experimental hepatopulmonary syndrome after common bile duct ligation by stimulating pulmonary endothelial nitric oxide production. In thioacetamide-induced nonbiliary cirrhosis, hepatic endothelin 1 production and release do not occur, and hepatopulmonary syndrome does not develop. However, the source and regulation of hepatic endothelin 1 after common bile duct ligation are not fully characterized. We evaluated the sources of hepatic endothelin 1 production after common bile duct ligation in relation to thioacetamide cirrhosis and assessed whether transforming growth factor beta1 regulates endothelin 1 production.

Methods: Hepatopulmonary syndrome and hepatic and plasma endothelin 1 levels were evaluated after common bile duct ligation or thioacetamide administration. Cellular sources of endothelin 1 were assessed by immunohistochemistry and laser capture microdissection of cholangiocytes. Transforming growth factor beta1 expression and signaling were assessed by using immunohistochemistry and Western blotting and by evaluating normal rat cholangiocytes.

Results: Hepatic and plasma endothelin 1 levels increased and hepatopulmonary syndrome developed only after common bile duct ligation. Hepatic endothelin 1 and transforming growth factor beta1 levels increased over a similar time frame, and cholangiocytes were a major source of each peptide. Transforming growth factor beta1 signaling in cholangiocytes in vivo was evident by increased phosphorylation and nuclear localization of Smad2, and hepatic endothelin 1 levels correlated directly with liver transforming growth factor beta1 and phosphorylated Smad2 levels. Transforming growth factor beta1 also stimulated endothelin 1 promoter activity, expression, and production in normal rat cholangiocytes.

Conclusions: Cholangiocytes are a major source of hepatic endothelin 1 production during the development of hepatopulmonary syndrome after common bile duct ligation, but not in thioacetamide-induced cirrhosis. Transforming growth factor beta1 stimulates cholangiocyte endothelin 1 expression and production. Cholangiocyte-derived endothelin 1 may be an important endocrine mediator of experimental hepatopulmonary syndrome.

Figure 1

Liver histology in common bile duct ligation (CBDL)-treated and thioacetamide (TAA)-treated animals. Compared with control animals (A), 1-week CBDL animals developed bile duct proliferation and periductal fibrosis (B). Two-week CBDL animals had progressive duct proliferation with expansion of portal tracts and bridging fibrosis (C). Three-week CBDL animals developed biliary cirrhosis (D). TAA-treated animals had progression from focal hepatocyte necrosis and bridging fibrosis with minimal inflammation at 2 weeks (E) to micronodular cirrhosis at 8 weeks (F) (Masson’s trichrome stain; original magnification, 40×).

Figure 2

Immunohistochemical localization of endothelin-1 (ET)-1, cytokeratin 19 (CK19), and α-smooth muscle actin (α-SMA) in liver sections from CBDL animals. Representative images are shown of ET-1 (top), CK19 (middle), and α-SMA (bottom) in normal (A, E, and I), 1-week (B, F, and J), 2-week (C, G, and K), and 3-week (D, H, and L) CBDL liver. In normal liver (A), ET-1 staining was most prominent in a pattern consistent with sinusoidal endothelial staining. After CBDL (B–D), staining increased markedly in the cytoplasm of biliary epithelial cells and focally in periductal regions. CK19 stained the biliary epithelium in normal (E) and CBDL (F–H) animals in a pattern similar to CBDL ET-1 staining. In normal liver (I), α-SMA staining was most prominent in vascular smooth muscle. After CBDL (J–L), α-SMA staining increased in the periductal regions (but not in biliary epithelium) in a pattern consistent with stellate cells or portal myofibroblast staining (original magnification, 40×).

Figure 3

Immunohistochemical localization of ET-1 and α-SMA in liver sections from 8-week TAA-treated animals. ET-1 staining (A) was found in lobular regions in a sinusoidal distribution similar to that of normal controls and additionally in cells adjacent to fibrous bands. α-SMA staining (B) was seen in stellate cells adjacent to fibrous bands in a pattern similar to perifibrous ET-1 staining (original magnification, 40×).

Figure 4

Liver CK19 and α-SMA protein levels in CBDL- and TAA-treated animals and correlation of liver CK19 levels with liver and plasma ET-1 levels in CBDL animals. (A) The top panel shows representative liver CK19 and α-SMA immunoblots, and the bottom graph summarizes Western blot results. Single bands of approximately 40 kilodaltons for CK19 or 42 kilodaltons for α-SMA were seen. Values are expressed as means ± SE (n = 5–7 animals for each group). *P < .05 compared with normal controls. (B) The relation between liver CK19 content and liver or plasma ET-1 levels in control and CBDL animals. The linear correlation coefficient (r) was significant for each comparison. **P < .0001.

Figure 5

Laser capture microdissection (LCM) of cholangiocytes and hepatocytes and ET-1 mRNA analysis from CBDL liver. Representative H&E-stained images from CBDL liver show a bile ductule cross section before (A) and after (B) LCM and after capture to a cap (C). The graph summarizes ET-1 mRNA levels in total RNA amplified after isolation from cholangiocytes and hepatocytes from 1-, 2-, and 3-week CBDL liver (D). Data are expressed relative to 18S levels as means ± SE and are from 4 to 5 real-time quantitative RT-PCR assays. *P < .05 compared with normal controls (n = 4–5 animals for each group).

Figure 6

Liver transforming growth factor (TGF)-β1 and phospho-Smad2 (p-Smad2) protein levels and correlation of liver ET-1 content with liver TGF-β1 or p-Smad2 levels in CBDL animals. Single bands of approximately 50 kilodaltons consistent with TGF-β1 (A) and 58 kilodaltons consistent with p-Smad2 or Smad2/3 (B) are seen. Graphs summarize TGF-β1 and p-Smad2 protein levels in liver. Values are expressed as means ± SE (n = 5–7 animals for each group). *P < .05 compared with normal controls. The relation is shown between ET-1 levels and TGF-β1or p-Smad2 levels in liver from control and CBDL animals (C). The linear correlation coefficient (r) was significant for each comparison. **P < .0001. A graphical summary is shown of TGF-β1 mRNA levels in cholangiocytes from control and 1-, 2-, and 3-week CBDL liver (D). Data are expressed relative to 18S levels as means ± SE and are from 4 to 5 real-time quantitative RT-PCR assays. P < .05 compared with normal control (n = 4–5 animals for each group).

Figure 7

Immunohistochemical localization of TGF-β1 and p-Smad2 in liver sections from 1-, 2-, and 3-week CBDL animals. Representative images of TGF-β1 (top) and p-Smad2 (bottom) are shown in normal (A and E), 1-week (B and F), 2-week (C and G), and 3-week (D and H) CBDL liver. In normal animals (A), TGF-β1 staining was minimal and was not observed in ductular epithelium. After CBDL (B–D), TGF-β1 staining increased most notably in the cytoplasm of cholangiocytes beginning within 1 week. There was also a focal increase in staining in periductal cells and in a distribution consistent with sinusoidal staining. Smad2 phosphorylation and nuclear localization were also minimal in normal animals (E). After CBDL (F–H), p-Smad2 staining increased and was found most prominently in cholangiocyte nuclei (arrows) and in the nuclei of some periductal cells and periportal hepatocytes.

Figure 8

Evaluation of TGF-β1 modulation of ET-1 expression in NRCs. (A) RT-PCR analysis shows the expression of the TGF-β type I and II receptors in NRCs. (B) Dose and time course RT-PCR analysis of ET-1 mRNA levels relative to α-actin mRNA levels after TGF-β1 stimulation of NRCs. (C) NRC ET-1 Northern blot analysis. Top panel insert shows representative ET-1 and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) Northern blots in control NRCs and in NRCs incubated with TGF-β1 alone or TGF-β1 after pretreatment with a specific anti–TGF-β antibody (TGF-β Ab) or a nonspecific rabbit immunoglobulin G (10 μg/mL). Bottom graph summarizes effects of TGF-β1 stimulation on ET-1 mRNA levels in NRCs normalized to GAPDH mRNA levels. ET-1 mRNA levels are expressed relative to control as means ± SE (n = 4–9 for each group). *P < .05 compared with control. (D) NRC ET-1 radioimmunoassay. NRCs were treated with TGF-β1 in the presence or absence of a specific TGF-β Ab. The graph summarizes effects of TGF-β1 on ET-1 production in NRCs. ET-1 levels are expressed as means ± SE (n = 4–9 for each group). *P < .05 compared with control. (E) Analysis of TGF-β1 effects on ET-1 promoter activity in NRCs by using 2 promoter constructs (pET-184 and pET-649) determined by a luciferase reporter gene assay normalized for transfection efficiency. Data are expressed in relative light units (RLU) per gram of protein as mean ± SE (n = 6 for each group). *P < .05 compared with the control that was transfected without any treatment.