AKR1D1 knockout mice develop a sex-dependent metabolic phenotype

Abstract

Steroid 5β-reductase (AKR1D1) plays important role in hepatic bile acid synthesis and glucocorticoid clearance. Bile acids and glucocorticoids are potent metabolic regulators, but whether AKR1D1 controls metabolic phenotype in vivo is unknown. Akr1d1-/- mice were generated on a C57BL/6 background. Liquid chromatography/mass spectrometry, metabolomic and transcriptomic approaches were used to determine effects on glucocorticoid and bile acid homeostasis. Metabolic phenotypes including body weight and composition, lipid homeostasis, glucose tolerance and insulin tolerance were evaluated. Molecular changes were assessed by RNA-Seq and Western blotting. Male Akr1d1-/- mice were challenged with a high fat diet (60% kcal from fat) for 20 weeks. Akr1d1-/- mice had a sex-specific metabolic phenotype. At 30 weeks of age, male, but not female, Akr1d1-/- mice were more insulin tolerant and had reduced lipid accumulation in the liver and adipose tissue yet had hypertriglyceridemia and increased intramuscular triacylglycerol. This phenotype was associated with sexually dimorphic changes in bile acid metabolism and composition but without overt effects on circulating glucocorticoid levels or glucocorticoid-regulated gene expression in the liver. Male Akr1d1-/- mice were not protected against diet-induced obesity and insulin resistance. In conclusion, this study shows that AKR1D1 controls bile acid homeostasis in vivo and that altering its activity can affect insulin tolerance and lipid homeostasis in a sex-dependent manner.

Keywords: bile acid; chenodeoxycholic acid; cholic acid; cortisol; metabolic syndrome; non-alcoholic fatty liver disease; steroid.

Figure 1

Hepatic and serum bile acids are lower in Akr1d1^–/– mice with sexually dimorphic changes to bile acid metabolism and composition. Mature (30 weeks) male and female Akr1d1^–/– mice have decreased total hepatic and serum bile acids (A) and altered bile acid composition with reduced 12α-hydroxylated/non-12α-hydroxylated bile ratio in the liver and serum (B and C) (n  = 12–16 mice). Principal component analysis shows greater divergence from WT in male Akr1d1^–/– mice (D). Akr1d1 deletion has a sexually dimorphic effect on mRNA expression of hepatic bile acid-metabolising genes and levels of bile acid intermediates. Cyp8b1 expression is increased in Akr1d1^–/– females but not in males (E) (n  = 10 mice) as are the AKR1D1 substrates 7α,12α-dihydroxy-4-chol-3-one (F) and 7α-hydroxy-4-chol-3-one (G) (n  = 10 mice). The oxysterol 27-hydroxycholesterol (27-OHC) is decreased in Akr1d1^–/– females (H) (n  = 9 mice). Female Akr1d1^–/– mice also have increased expression of the bile acid-detoxifying genes Cyp3a11, Cyp2c55, Cyp4a12a and Sult2a7 (I) (n  = 10 mice) and serum levels of LCA sulphate (J) (n  = 9 mice). Data are presented as mean ± s.d., log2(FC), ratio or mean relative abundance. *P  < 0.05, **P  < 0.01, ***P  < 0.005, ^∅ P  < 0.001, ^∅∅ P  < 0.0005, ^∅∅∅ P  < 0.0001 compared to WT. P values for bile acid composition compare WT and Akr1d1^–/– within sex (male/female). (WT = WT C57BL/6; –/– = Akr1d1^–/–). A full colour version of this figure is available at https://doi.org/10.1530/JOE-21-0280.

Figure 2

Schematic overview of the steps involved in the classic and alternative bile acid synthesis pathways in humans and mice. Figure produced in BioRender. A full colour version of this figure is available at https://doi.org/10.1530/JOE-21-0280.

Figure 3

At maturity, male Akr1d1^–/– mice have reduced fat mass and improved insulin tolerance. Young (10-week) Akr1d1^–/– mice have normal body weight and composition (lean mass empty bar; fat mass filled bar) (A) and energy expenditure (B) is unchanged (n  = 14–16). Male, but not female, Akr1d1^–/– mice have increased dark phase food intake (C) and a preference for carbohydrate over lipid as an energy source, measured as an increase in respiratory exchange ratio (D) (n  = 14–16). Male (blue lines), but not female (red lines), Akr1d1^–/– mice gain less weight than their WT (black line) littermates (E) and at 30 weeks have lower fat mass without change in lean mass (F) (lean mass empty bar; fat mass filled bar) (n  = 14–16). Akr1d1^–/– males have exaggerated glucose clearance after i.p. insulin injection (G), normal 30 min kITT (H), reduced blood glucose after a 4-h fast (I) and reduced quadricep (J) (n  = 11–15), but normal liver, glycogen (K). Although i.p. glucose tolerance (L) and fed serum glucose (M) are comparable to WT animals (n  = 11–15), fed insulin concentration is decreased (N) (n  = 7–8). Data are presented as mean ± s.d. *P  < 0.05, **P  < 0.01 compared to WT of the same sex. (WT = WT C57BL/6; –/– = Akr1d1^–/–). A full colour version of this figure is available at https://doi.org/10.1530/JOE-21-0280.

Figure 4

Male Akr1d1^–/– mice have reduced adipose and hepatic lipid accumulation and hypertriglyceridemia. Mature (30 week) male Akr1d1^–/– mice (grey points) have smaller gonadal, subcutaneous and peri-renal adipose tissue weights compared to WT littermates (black points) (A) (n  = 14–15 mice) with smaller adipocytes in the gonadal and subcutaneous depots (B and C) (n  = 13–14 mice) and reduced hepatic triacylglycerol (D) (n  = 14–15 mice). Serum triacylglycerol (E) (n  = 14–16), monoacylglycerols and diacylglycerols (F) and non-esterified fatty acids (G) are increased, but total and HDL cholesterol (H) are normal (n  = 10 mice). Intra-muscular triacylglycerol is increased in Akr1d1^–/– quadricep muscle (I) (n  = 9 mice). The mRNA expression of lipid metabolism genes in the gonadal fat is unchanged (J), but in the liver, the expression of fatty acid esterification (Gpat3) and lipid droplet (Cidec & Plin4) genes as well as the transcription factor Pparg are reduced (K) (n  = 10 mice). *P  < 0.05, **P  < 0.01, ***P  < 0.005, ^∅ P  < 0.001, ^∅∅ P  < 0.0005 compared to wildtype (WT = WT C57BL/6; –/– = Akr1d1^–/–).

Figure 5

Akr1d1 deletion does not protect male C57BL/6 mice against diet-induced obesity, hypercholesterolemia or insulin resistance but improves hypertriglyceridemia. On a high fat diet (HFD), male Akr1d1^–/– mice (grey line) gain weight at the same rate as their WT littermates (black line) (A) and after 20 weeks body composition is not different between genotypes (lean mass lower bar; fat mass upper bar) (B) (n  = 14–15 mice). Adipose weight (C) (n  = 14–15), hepatic triacylglycerol (D) (n = 12–15), HDL and total cholesterol (E) (n  = 14–15) were unchanged between HFD fed WT and Akr1d1^–/– males. Serum triacylglycerol was reduced in HFD fed Akr1d1^–/– males but not to levels seen in the control diet (F) (n  = 10 mice). Akr1d1^–/– males were not protected against diet-induced reduction in ipGTT (G) or ipITT (H) (n  = 14–15 mice). Data are presented as mean ± s.d.*P  < 0.05, **P  < 0.01, ***P  < 0.005 compared to wildtype (WT = WT C57BL/6; –/– = Akr1d1^–/–).

Figure 6

Bile acid profile in male Akr1d1^–/– mice on a high fat diet (HFD). After 20 weeks on a HFD, male Akr1d1^–/– mice have reduced total liver (A) and serum (B) bile acids and altered bile acid composition (liver: C; serum D) compared to high fat-fed WT littermates. Serum 12α-hydroxylated/non-12α-hydroxylated bile ratio in the liver (E) and serum (F) is reduced compared to control and HFD fed WT littermates. Data are presented as mean ± s.d. in n  =10–15 mice. *P  < 0.05, **P  < 0.01, ***P  < 0.005, ^∅ P  < 0.001, ^∅∅ P  < 0.0005, ^∅∅∅ P  < 0.0001. WT control diet data are presented for comparison, P value for bile acid composition compares WT and Akr1d1^–/– on HFD (WT = WT C57BL/6; –/– = Akr1d1^–/–). A full colour version of this figure is available at https://doi.org/10.1530/JOE-21-0280.