Selective Hepatic Cbs Knockout Aggravates Liver Damage, Endothelial Dysfunction and ROS Stress in Mice Fed a Western Diet

Abstract

Cystathionine-β-synthase (CBS) is highly expressed in the liver, and deficiencies in Cbs lead to hyperhomocysteinemia (HHCy) and disturbed production of antioxidants such as hydrogen sulfide. We therefore hypothesized that liver-specific Cbs deficient (LiCKO) mice would be particularly susceptible to the development of non-alcoholic fatty liver disease (NAFLD). NAFLD was induced by a high-fat high-cholesterol (HFC) diet; LiCKO and controls were split into eight groups based on genotype (con, LiCKO), diet (normal diet, HFC), and diet duration (12 weeks, 20 weeks). LiCKO mice displayed intermediate to severe HHCy. Plasma H₂O₂ was increased by HFC, and further aggravated in LiCKO. LiCKO mice fed an HFC diet had heavier livers, increased lipid peroxidation, elevated ALAT, aggravated hepatic steatosis, and inflammation. LiCKO mice showed decreased L-carnitine in the liver, but this did not result in impaired fatty acid oxidation. Moreover, HFC-fed LiCKO mice demonstrated vascular and renal endothelial dysfunction. Liver and endothelial damage correlated significantly with systemic ROS status. In conclusion, this study demonstrates an important role for CBS in the liver in the development of NAFLD, which is most probably mediated through impaired defense against oxidative stress.

Keywords: NAFLD; cystathionine-β-synthase; high-fat diet; hydrogen sulfide; hyperhomocysteinemia; knockout mice; liver damage; reactive oxygen species.

Figure 1

CBS is selectively deleted in liver. (a) CBS protein expression in LiCKO is significantly downregulated in liver; (b) CBS protein expression in LiCKO is unaffected in kidneys; (c) Down-regulation of CBS in liver is associated with impaired hydrogen sulfide production. * p < 0.05.

Figure 2

Body weight, liver weight, WBC, HCT, and PLT. (a,b) At 12 and 20 weeks of diet, body weight was higher in ND-fed LiCKO mice than controls on ND. Body weight of all groups on HFC diet were significantly increased compared with mice on ND, without differences between genotypes; (c) Liver weight of LiCKO mice on HFC diet was significantly increased over LiCKO on ND; (d) Liver weight of LiCKO mice on HFC diet was significantly increased over controls on HFC diet and over LiCKO on ND; (e) At 12 weeks of diet, no differences were observed in WBC; (f) At 20 weeks of diet, WBC in LiCKO mice on HFC diet was significantly increased over controls on HFC diet and over LiCKO on ND; (g) At 12 weeks of diet, no differences were observed in HCT; (h) At 20 weeks of diet, HCT in LiCKO mice on HFC diet was significantly increased over controls on HFC diet; (i) At 12 weeks of diet, no differences were observed in PLT; (j) At 20 weeks of diet, PLT in LiCKO mice on HFC diet was significantly decreased over controls on HFC diet and over LiCKO on ND; WBC; white blood cell count, HCT; hematocrit; PLT; platelet count; * p < 0.05 difference between con and LiCKO on same diet; † p < 0.05 difference between ND and HFC for con or LiCKO; ns = non-significant.

Figure 3

Plasma homocysteine, serum hydrogen peroxide, liver MDA staining, and plasma ALAT. (a) At 12 weeks of diet, homocysteine was elevated in plasma of ND-fed LiCKO mice and further significantly increased in LiCKO HFC mice; (b) At 20 weeks of diet, plasma homocysteine was similarly elevated in both LiCKO ND and LiCKO HFC; (c,d) At 12 and 20 weeks of diet, two-way ANOVA indicated that genotype (LiCKO) and diet (HFC) were both associated with further increased serum H₂O₂ concentrations; (e) No differences in liver MDA staining were observed between groups at 12 weeks of diet; (f) At 20 weeks of diet, liver MDA staining was significantly increased only in LiCKO HFC; (g,h) At 12 and 20 weeks, plasma ALAT activity was increased in LiCKO HFC over con HFC. * p < 0.05 difference between con and LiCKO on same diet; † p < 0.05 difference between ND and HFC for con or LiCKO; ns = non-significant.

Figure 4

Liver histology. (a,b,e,f) No abnormal histological patterns were detected in mice on ND at 12 and 20 weeks of diet; (c) Control mice on HFC demonstrate mild steatosis (white arrow); (d) Steatosis is more aggravated in LiCKO at 12 weeks of HFC; (g) At 20 weeks of HFC, mild steatosis is present in control mice; (h) Further increased steatosis and inflammatory cells (black arrow) can be seen at 20 weeks of HFC in LiCKO; (i) NAFLD activity scores (NAS) at 12 weeks of diet; (j) NAFLD activity scores (NAS) were significantly increased in LiCKO mice at 20 weeks of HFC diet compared to controls on HFC diet. * p < 0.05 difference between con and LiCKO on same diet; † p < 0.05 difference between ND and HFC for con or LiCKO; ns = non-significant.

Figure 5

Liver triglyceride and cholesterol levels at weeks 12 and 20 of diet. (a,b) Liver triglyceride levels; (c,d) free cholesterol levels in liver; (e,f) total cholesterol and (g,h) cholesteryl levels. * p < 0.05 difference between con and LiCKO on same diet; † p < 0.05 difference between ND and HFC for con or LiCKO; ns = non-significant.

Figure 6

Acylcarnitines in liver. (a) Acylcarnitines at 12 weeks of diet; (b) Acylcarnitines at 20 weeks of diet; * p < 0.05 difference between con and LiCKO on same diet; † p < 0.05 difference between ND and HFC for con or LiCKO; ns = non-significant.

Figure 7

Effect of diet and genotype on vascular and endothelial function. (a,c) No effects of HFC diet on EDR were observed in control mice at 12 or 20 weeks of diet. Inserts show AUC for EDR; (b,d) Impaired EDR in LiCKO mice at both 12 and 20 weeks of HFC diet. Inserts show AUC for EDR; Quantification of the AUC demonstrated decreased EDR in LiCKO mice at both 12 and 20 weeks of HFC diet; (e) Increased ACR in urine of LiCKO mice at 20 weeks of HFC indicated renal endothelial damage. * p < 0.05 difference between con and LiCKO on same diet; † p < 0.05 difference between ND and HFC for con or LiCKO; ns = non-significant.

Figure 8

Correlations between liver and endothelial damage with oxidative state. (a,c) plasma ALAT and ACR activity correlated positively with serum H₂O₂ at week 20; (b) EDR correlated negatively with serum H₂O₂ at week 20.