Protective and damaging effects of platelets in acute cholestatic liver injury revealed by depletion and inhibition strategies

Abstract

Alpha-naphthylisothiocyanate (ANIT) causes cholestatic hepatitis characterized by intrahepatic bile duct epithelial cell injury and periportal hepatocellular necrosis. The progression of ANIT-induced hepatocyte injury is reported to involve extrahepatic cells including platelets. We showed recently that the procoagulant protein tissue factor (TF) is essential for ANIT-induced coagulation and contributes to ANIT-induced liver necrosis. Platelets have been shown to express TF and can contribute to coagulation cascade activation. To this end, we tested the hypothesis that platelet-dependent coagulation contributes to ANIT-induced liver injury. In ANIT (60 mg/kg)-treated mice, activation of the coagulation cascade occurred prior to a decrease of platelets in the blood. Immunostaining for glycoprotein IIb (CD41) revealed platelet accumulation along the borders of necrotic foci in livers of ANIT-treated mice. Antibody-mediated platelet depletion did not affect coagulation but markedly affected liver histopathology in ANIT-treated mice. Platelet depletion induced marked pooling of blood within necrotic lesions consistent with parenchymal-type peliosis as early as 24 h after ANIT treatment. In contrast, treatment with the P2Y(12) inhibitor clopidogrel significantly reduced ANIT-induced hepatocyte necrosis and serum alanine aminotransferase activity but did not exaggerate bleeding into necrotic foci. Clopidogrel also reduced hepatic neutrophil accumulation but did not affect induction of Intercellular adhesion molecule-1 or chemokine CxC motif ligand-1 messenger RNA expression in liver. The data indicate that ANIT-induced coagulation is platelet independent and that platelets contribute to ANIT-induced hepatocyte necrosis by promoting neutrophil accumulation. In contrast, severe thrombocytopenia induces parenchymal-type peliosis in the livers of ANIT-treated mice, a rare hepatic lesion associated with pooling of blood in the liver.

FIG. 1.

Time course of ANIT-induced liver injury, coagulation, and blood platelets in mice. Mice were given ANIT (60 mg/kg, po) or vehicle (corn oil). (A) Serum ALT activity, (B) serum ALP activity, (C) plasma TAT levels, and (D) platelets in EDTA-anticoagulated whole blood were determined 24 and 48 h after ANIT administration. Levels of all biomarkers in mice given vehicle were not different between 24 and 48 h. Forty-eight-hour data are shown. Data are expressed as mean ± SEM; n = 5 mice per group. *Significantly different from Vehicle-treated mice. p < 0.05.

FIG. 2.

Platelet accumulation and fibrin deposition in livers of ANIT-treated mice. Mice were given ANIT (60 mg/kg, po) or vehicle (corn oil) and livers removed 48 h later. (A and B) Representative photomicrographs showing (A) CD41 (platelet, green) staining and (B) fibrin (red) staining in a liver section from a vehicle-treated mouse. The digital merge of these images is shown in panel (C). Asterisk indicates fibrin staining (red) outlining the border of a central vein. (D and E) Representative photomicrographs showing marked (D) CD41 (platelet, green) staining and (E) fibrin (red) staining in areas of necrosis in a liver section from an ANIT-treated mouse. The digital merge of these images is shown in panel (F). Bar = 20 μm. Arrow indicates an area of hepatic necrosis containing fibrin and bordered by marked platelet accumulation.

FIG. 3.

Effect of platelet depletion on ANIT-induced coagulation and liver injury. Mice were given 1 mg/kg of anti-CD41 antibody (MWReg30) or isotype control antibody 16 h prior to treatment with ANIT (60 mg/kg, po). (A) Platelets in EDTA-anticoagulated whole blood, (B) plasma TAT levels, (C) serum ALP activity, and (D) serum ALT activity were determined 24 and 48 h later. Data are expressed as mean ± SEM; n = 5 mice per group. *Significantly different from the same group of mice at 24 h; #significantly different from ANIT-treated mice pretreated with isotype control antibody.

FIG. 4.

Effect of platelet depletion on ANIT-induced liver histopathology. Mice were given 1 mg/kg of anti-CD41 antibody or isotype control antibody 16 h prior to treatment with ANIT (60 mg/kg, po). Representative photomicrographs showing hematoxylin and eosin–stained liver sections from isotype control antibody–pretreated mice (A) 24 and (D) 48 h after ANIT treatment and in anti-CD41 antibody–pretreated mice 24 h after ANIT treatment (B and C) and 48 h after ANIT treatment (E and F). Bar = 20 μm (A and B), 10 μm (C), and 50 μm (D-F). Arrows indicate areas of periportal hepatocellular necrosis. Asterisks are near areas of parenchymal-type peliosis.

FIG. 5.

Effect of clopidogrel treatment on ANIT-induced liver injury. Mice were treated with ANIT (60 mg/kg, po) and then subsequently with clopidogrel (30 mg/kg, ip) or sterile PBS 8, 20, and 32 h later. Serum (A) ALT activity and (B) ALP activity were determined 48 h later. Representative photomicrographs showing hematoxylin and eosin–stained (C and D, Bar = 50 μm) and anti-CD41 (platelet, red)–stained (E and F, Bar = 20 μm) liver sections from ANIT-treated mice given vehicle (C and E) or clopidogrel (D and F). (G) Quantification of CD41-positive staining (platelets) in liver sections from each group of mice. Data are expressed as mean ± SEM; n = 3–6 mice per group. *Significantly different from ANIT-treated mice given sterile PBS; p < 0.05. Arrows indicate areas of periportal hepatocellular necrosis.

FIG. 6.

Effect of clopidogrel treatment on indicators of neutrophil accumulation. Mice were treated with ANIT (60 mg/kg, po) and then subsequently with clopidogrel (30 mg/kg, ip) or sterile PBS 8, 20, and 32 h later. Hepatic (A) KC and (B) ICAM-1 mRNA levels were determined 48 h later. (C) Liver homogenate MPO activity was measured 48 h after ANIT treatment. Data are expressed as mean ± SEM; n = 3–6 mice per group. *Significantly different from ANIT-treated mice given sterile PBS; p < 0.05.