Constitutive androstane receptor mediates the induction of drug metabolism in mouse models of type 1 diabetes

Abstract

Untreated type 1 diabetes increases hepatic drug metabolism in both human patients and rodent models. We used knockout mice to test the role of the nuclear xenobiotic receptors constitutive androstane receptor (CAR) and pregnane and xenobiotic receptor (PXR) in this process. Streptozotocin-induced diabetes resulted in increased expression of drug metabolizing cytochrome P450s and also increased the clearance of the cytochrome P450 substrate zoxazolamine. This induction was completely absent in Car(-/-) mice, but was not affected by the loss of PXR. Among the many effects of diabetes on the liver, we identified bile acid elevation and activated adenosine monophosphate-activated protein kinase as potential CAR-activating stimuli. Expression of the CAR coactivator peroxisome proliferator-activated receptor gamma coactivator (PGC)-1alpha was also increased in mouse models of type 1 diabetes.

Conclusion: The CAR-dependent induction of drug metabolism in newly diagnosed or poorly managed type 1 diabetes has the potential for significant impact on the efficacy or toxicity of therapeutic agents.

Figure 1. CAR is activated in STZ-induced type 1 diabetic mice and drug clearance is enhanced

(A) Wild type and CAR^-/- mice (n=5) were treated with or without STZ for 7 days and subgroups of diabetic mice were treated with insulin for 7 days. Blood glucose levels were measured on indicated days. (B) Liver total RNA was isolated from mice of different treatments and equal amounts of RNA were pooled from individual mice. Northern blot was performed with indicated probes. (C) Zoxazolamine was dosed to non-diabetic mice, diabetic mice and diabetic mice with insulin treatment. Paralysis time was compared in each group. (**p < 0.01).

Figure 2. CAR activation maintains in diabetic hCAR mice, NOD mice and PXR^-/- mice

Total liver RNA of each different group was isolated. Northern blot or qPCR was performed with indicated probes. (A) hCAR mice (n=4) were treated with or without STZ for 2 weeks. Glucose and gene expression were studied. (B) 22-week male NOD non diabetic and diabetic mice (n=3-4) were checked for glucose and CAR target gene expression. (C) Wild type and PXR^-/- mice (n=3-4) were treated with or without STZ for 2 weeks. Gene expression of CAR targets were studied for CAR activation. (**p<0.01)

Figure 3. CAR activation is not observed in mice with either acute exposure to STZ or β-hydroxybutyrate

(A) Wild type and CAR^-/- mice (n=5) were treated with or without STZ for 2 days and on the third day, mice were sacrificed and their liver samples were harvested. Total RNA were isolated from each group and northern blots were performed with indicated probes. (B) Wild type and CAR^-/- mice (n=5) were treated with β-hydroxybutyrate and mice were sacrificed at indicated hours. Liver samples were harvested and total RNA were isolated and prepared. Northern analysis was performed with indicated probes.

Figure 4. Elevation of bile acid levels in STZ-induced diabetic mice

Wild type and CAR^-/- mice (n=7) were treated with or without STZ for 2 weeks and serum and hepatic bile acid levels were measured. (*p<0.05; **p<0.01)

Figure 5. AMPK is activated in STZ-induced diabetic mice

Wild type and CAR^-/- mice (n=4) were treated with or without STZ for 2 weeks and liver samples were harvested. Liver tissue from different groups was homogenized and every two samples from same groups were pooled together and 30 μg of protein was separated on SDS gel. Phospho-AMPK and total AMPK level were determined by Western blot using antibodies as indicated. The graphs demonstrate the quantification of phosphorylation of AMPK. (**p<0.01)

Figure 6. PGC-1α expression is induced in mouse models of type 1 diabetes

Liver RNA from WT mice (n=4) treated with or without STZ for 2 weeks and NOD nondiabetic and diabetic mice (n=3-4) was extracted and analyzed for PGC-1α by quantitative RT-PCR. (**p<0.01)