Role of peroxisome proliferator-activated receptor-alpha in hepatobiliary injury induced by ammonium perfluorooctanoate in mouse liver

Abstract

Peroxisome proliferator-activated receptor-alpha (PPARalpha) has been suggested to protect against chemically induced hepatobiliary injuries in rodents. This function could mask the potential toxicities of perfluorooctanoic acid (PFOA) that is an emerging environmental contaminant and a weak ligand of PPARalpha. However its function has not been clarified. In this study, PFOA was found to elicit hepatocyte and bile duct injuries in Pparalpha-null mice after 4 wk treatment with PFOA ammonium salt (0, 12.5, 25, 50 micromol/kg/d, gavage). In wild-type mice, PFOA caused major hepatocellular damage dose-dependently and minor cholangiopathy observed only at 25 and 50 micromol/kg. In treated Pparalpha-null mice, PFOA produced marked fat accumulation, severe cholangiopathy, hepatocellular damage and apoptotic cells especially in bile ducts. Oxidative stress was also increased 4-fold at 50 micromol/kg and TNF-alpha mRNA was upregulated more than 3-fold at 25 micromol/kg in Pparalpha-null mice. Biliary bile acid/phospholipid ratios were higher in Pparalpha-null mice than in wild-type mice. Results from these studies suggest that PPARalpha is protective against PFOA and have a critical role in drug induced hepatobiliary injury.

Fig. 1.. Effects of PFOA on the mouse liver by oral gavage for 4 wk.

Hematoxylin-eosin stained sections of liver from control wild-type mice (A), wild-type mice treated with PFOA at 12.5 μmol/kg (B), 25 μmol/kg (C), 50 μmol/kg (D), control Pparα-null mice (E), Pparα-null mice treated with PFOA at 12.5 μmol/kg (F), 25 μmol/kg (G), 50 μmol/kg (H, I). Original magnification, × 200 (A–H), × 40 (I). Wild-type mice treated with PFOA (B–D) have diffuse hepatocyte hypertrophy with numerous eosinophilic cytoplasmic granules. Control Pparα-null mice (E) has scattered small fat vacuoles. Centrilobular fat accumulations were increased dose-independently in Pparα-null mice treated with PFOA at 12.5 μmol/kg (F), 25 μmol/kg (G), 50 μmol/kg (H, I). Focal necrosises are scattered with fat accumulation and proliferation of bile ductules is prominent in the portal tracts in Pparα-null mice treated with PFOA at 50 μmol/kg (I). Diffuse hepatocyte hypertrophy was observed in both mouse lines treated (B–D, F–H). Bile duct epithelial thickness (arrow) was observed in both mouse lines treated at 25 μmol/kg (C, G) and 50 μmol/kg (D, H). Diffusely distributed, fine, fatty droplets and ground-glass appearance is showed at 12.5 μmol/kg (F) and 25 μmol/kg (G) in Pparα-null mice. Note hyperplastic changes in the biliary duct epithelium with bile plaque (arrow head) and fibrosis (open circle) as evidenced by proliferation of bile ductules (arrow) in Pparα-null mice treated with PFOA at 50 μmol/kg (H). cv, central vein; pv, portal vein; ha, hepatic artery; bd, bile duct; f, fat droplet; ne, necrosis.

Fig. 2.. Distribution of apoptotic cells in liver PFOA treated by oral gavage for 4 wk by immunohistochemistry for TUNEL.

Wild-type mice treated with PFOA at 25 μmol/kg (A, E) and 50 μmol/kg (B, F), Pparα-null mice treated with PFOA at 25 μmol/kg (C, G) and 50 μmol/kg (D, H). Original magnification, × 100 (A–D), × 400 (E–H) the extended a part surrounded with a square in A–D, respectively. Wild-type mice treated with PFOA at 25 μmol/kg (A, E) and 50 μmol/kg (B, F) show diffuse positive stains in hepatocyte, vessel wall, and bile duct epithelium (arrow). Pparα-null mice treated with PFOA at 25 μmol/kg (C, G) and 50 μmol/kg (D, H) show positive stains mainly in bile duct epithelium (arrow head). cv, central vein; pv, portal vein; ha, hepatic artery; bd, bile duct.

Fig. 3.. Ultrastructure of hepatocyte and bile duct epithelium cells in control and after treatments of wild-type mice and Pparα-null mice with PFOA by oral gavage for 4 wk.

Hepatocytes from control wild-type mice (A), wild-type mice treated with PFOA at 12.5 μmol/kg (B), 25 μmol/kg (C), 50 μmol/kg (D, I), Control Pparα-null mice (E), Pparα-null mice treated with PFOA at 12.5 μmol/kg (F), 25 μmol/kg (G), 50 μmol/kg (H, J), Bile duct epithelial cell (BEC) of Pparα-null mice treated with PFOA at 50 μmol/kg (K). Numerous glycogen granules (circle) are observed in control wild-type mice (A). The increased number and size of dark staining peroxisomes were shown in treated wild-type mice (B–D, I). Hepatocytes from control Pparα-null mice (E) are similar to control wild-type mice with fewer fat droplets (f) in cytoplasm. In contrast to controls, treated Pparα-null mice (F–H, J) also display hepatocyte hypertrophy, decreased glycogen granules, degranulation and disruption of the rough endoplasmic reticulum, and increased mitochondria in dose-dependently. The marked different points contrasts to wild-type mice treated with PFOA are increased fat droplets in cytoplasm, a few peroxisomes, and a variable size and shape of mitochondria (F–H, J). Note that peroxisomes are markedly increased and slightly enlarged in size in wild-type mice treated with PFOA at 50 μmol/kg (I), and mitochondria are pleomorphic, enlarged (*), and disorganization of cristae (arrowhead) in Pparα-null mice treated with PFOA at 50 μmol/kg (J). BECs (K) showed degradation of cytoplasmic structure, vacuolization, disintegration of nuclei and organelles, and were surrounded with fibroblasts and collagen. p, peroxisome; f, fat droplet; v, vacuole. (A–H) Bar=4 μm, (I, J) Bar=1 μm, (K) Bar=10 μm.

Fig. 4.. Effects of PFOA on biomarkers associated with liver injury.

(A) Effects of PFOA on 8-hydroxydeoxyguanosine from unfractionated livers of wild-type and Pparα-null mice. This figure reveals that the levels of 8-OHdG tend to increase dose-dependently in Pparα-null mice (Jonckheere’s test, p<0.05), in which the levels are increased significantly at 50 μmol/kg (p<0.05). (B) The expressions of TNF-α mRNA are significantly increased in Pparα-null mice treated with PFOA at 25 (p<0.01) and 50 μmol/kg (p<0.05). (C) The expressions of Mdr2 mRNA are significantly up-regulated in wild-type mice treated with PFOA at all doses (at 12.5 μmol/kg, 25 μmol/kg and 50 μmol/kg, p<0.05, p<0.01, respectively). In Pparα-null mice treated with PFOA, the expressions of Mdr2 mRNA are not induced at 12.5 μmol/kg, however induced at 25 μmol/kg (p<0.05) and 50 μmol/kg (p<0.01) significantly. (D) Effects of PFOA on biliary total bile acid/phospholipid (BA/PL) ratio. Biliary BA/PL ratios show significant decrease in wild-type mice treated with PFOA dose-dependently (p<0.05). However, no such significant adaptation is observed in Pparα-null mice treated with PFOA. Data are presented as mean ± SD. Trend test is Jonckheere’s test. *p<0.05, **p<0.01 versus control controls in each group. Log-transformation was performed for expressions of Mdr2 mRNA levels due to heteroscedusticity.

Fig. 5.. Effects of PFOA on Hepatic BSEP and MRP2 protein levels.

Each panel represents an individual experiment. There is a significant decrease in BSEP protein level in wild-type mice treated with PFOA at 50 μmol/kg (p<0.01). In Pparα-null mice treated with PFOA, the levels are increased significantly at 12.5 μmol/kg (p<0.01), however decreased significantly at 50 μmol/kg (p<0.05). There is a significant decrease in MRP2 protein levels in both wild-type and Pparα-null mice treated with PFOA at 50 μmol/kg (p<0.05). Control wild-type mice, w0; wild-type mice PFOA treated with 12.5 μmol/kg, w12.5; 25 μmol/kg, w25, 50 μmol/kg, w50; control Pparα-null mice, n0; Pparα-null mice treated with PFOA at 12.5 μmol/kg, n12.5; 25 μmol/kg, n25, 50 μmol/kg, n50. Black bars, wild-type mice; white bars, Pparα-null mice. Densitometric values are presented as mean ± SD of 3 animals in each group. *p<0.05, **p<0.01 versus control in each group. Trend test is Jonckheere’s test.