Sinusoidal endothelial dysfunction precedes inflammation and fibrosis in a model of NAFLD

Abstract

Non-alcoholic fatty liver disease (NAFLD) is the hepatic manifestation of the metabolic syndrome. Most morbidity associated with the metabolic syndrome is related to vascular complications, in which endothelial dysfunction is a major pathogenic factor. However, whether NAFLD is associated with endothelial dysfunction within the hepatic vasculature is unknown. The aims of this study were to explore, in a model of diet-induced overweight that expresses most features of the metabolic syndrome, whether early NAFLD is associated with liver endothelial dysfunction. Wistar Kyoto rats were fed a cafeteria diet (CafD; 65% of fat, mostly saturated) or a control diet (CD) for 1 month. CafD rats developed features of the metabolic syndrome (overweight, arterial hypertension, hypertryglyceridemia, hyperglucemia and insulin resistance) and liver steatosis without inflammation or fibrosis. CafD rats had a significantly higher in vivo hepatic vascular resistance than CD. In liver perfusion livers from CafD rats had an increased portal perfusion pressure and decreased endothelium-dependent vasodilation. This was associated with a decreased Akt-dependent eNOS phosphorylation and NOS activity. In summary, we demonstrate in a rat model of the metabolic syndrome that shows features of NAFLD, that liver endothelial dysfunction occurs before the development of fibrosis or inflammation.

Figure 1. Hematoxylin/eosin (H/E), oil-red, Mason’s Thricrome images and steatosis area from livers from rats fed a control diet (CD) or a high fat diet (CafD).

(original magnification, x100 or x200 as displayed in the figures).

Figure 2. Liver inflammation and hepatic stellate cells phenotype.

A) Immunohistochemistry showing CD43 (a pan-leukocyte marker) immunostaining in livers from high fat feeding (CafD) and control rats (CD). The administration for 1 month of a high fat diet did not induce liver inflammation (original magnification: x200). B) Cafeteria diet did not induce activation of HSC as assessed by immunohistochemistry for α-SMA.

Figure 3. Ex-vivo assessment of liver circulation.

Portal Perfusion Pressure (PPP) in CafD (n = 12) and CD rats (n = 12) in the absence (A) or the presence (B) of the NO donor sodium nitroprusside (SNP). Livers from rats fed cafeteria diet showed an increased PPP. No differences were observed in the presence of SNP (CD: control diet; CafD: cafeteria diet).

Figure 4. CafD rats show liver endothelial dysfunction.

A) Response to ACh in livers from control diet rats (CD; black squares; n = 6) and high fat diet rats (CafD; white circles; n = 6). B) NOS activity in liver homogenates from control rats (CD; n = 4) and cafeteria fed-rats (CafD; n = 4). C) Representative blots and densitometry readings of liver P-Akt (at Ser⁴⁷³) to Akt ratio and D) P-eNOS (at Ser¹¹⁷⁶) to eNOS ratio (western blotting). AU: Arbitrary Units.

Figure 5. Representative blots and densitometry readings of liver P-eNOS (at Ser¹¹⁷⁶) to eNOS ratio (western blotting), five minutes after a portal injection of vehicle (white bars and -) or insulin (black bars and +).

Insulin increased eNOS phosphorylation in CD rats (A) but not in CafD rats (B).

Figure 6. Endothelial cells phenotype.

Cafeteria diet did not induce changes in A) SE-1 expression nor B) CD31, which are closely correlated with fenestrations and capillarization at the sinusoidal endothelial cells. C) Cafeteria diet did not induce de novo expression of CD34, a marker of loss of LSEC phenotype.