Liver-specific reactivation of the inactivated Hnf-1alpha gene: elimination of liver dysfunction to establish a mouse MODY3 model

Mice deficient in hepatocyte nuclear factor 1 alpha (HNF-1alpha) develop dwarfism, liver dysfunction, and type 2 diabetes mellitus. Liver dysfunction in HNF-1alpha-null mice includes severe hepatic glycogen accumulation and dyslipidemia. The liver dysfunction may appear as soon as 2 weeks after birth. Since the HNF-1alpha-null mice become diabetic 2 weeks after birth, the early onset of the liver dysfunction is unlikely to be due to the diabetic status of the mice. More likely, it is due directly to the deficiency of HNF-1alpha in liver. Although the HNF-1alpha-null mice have an average life span of 1 year, the severe liver phenotype has thwarted attempts to study the pathogenesis of maturity-onset diabetes of the young type 3 (MODY3) and to examine therapeutic strategies for diabetes prevention and treatment in these mice. To circumvent this problem, we have generated a new Hnf-1alpha mutant mouse line, Hnf-1alpha(kin/kin), using gene targeting to inactivate the Hnf-1alpha gene and at the same time, to incorporate the Cre-loxP DNA recombination system into the locus for later revival of the Hnf-1alpha gene in tissues by tissue-specifically expressed Cre recombinase. The Hnf-1alpha(kin/kin) mice in which the expression of HNF-1alpha was inactivated in germ line cells were indistinguishable from the HNF-1alpha-null mice with regard to both the diabetes and liver phenotypes. Intriguingly, when the inactivated Hnf-1alpha gene was revived in liver (hepatic Hnf-1alpha revived) by the Cre recombinase driven by an albumin promoter, the Hnf-1alpha(kin/kin) mice, although severely diabetic, grew normally and did not develop any of the liver dysfunctions. In addition, we showed that the expression of numerous genes in pancreas, including a marker gene for pancreas injury, was affected by liver dysfunction but not by the deficiency of HNF-1alpha in pancreas. Thus, our hepatic-Hnf-1alpha-revived mice may serve as a useful mouse model to study the human MODY3 disorder.     

Mutant HNF-1alpha and mutant HNF-1beta identified in MODY3 and MODY5 downregulate DPP-IV gene expression in Caco-2 cells

Dipeptidylpeptidase IV (DPP-IV) is a well-documented drug target for the treatment of type 2 diabetes. Hepatocyte nuclear factors (HNF)-1alpha and HNF-1beta, known as the causal genes of MODY3 and MODY5, respectively, have been reported to be involved in regulation of DPP-IV gene expression. But, it is not completely clear (i) that they play roles in regulation of DPP-IV gene expression, and (ii) whether DPP-IV gene activity is changed by mutant HNF-1alpha and mutant HNF-1beta in MODY3 and MODY5. To explore these questions, we investigated transactivation effects of wild HNF-1alpha and 13 mutant HNF-1alpha, as well as wild HNF-1beta and 2 mutant HNF-1beta, on DPP-IV promoter luciferase gene in Caco-2 cells by means of a transient experiment. Both wild HNF-1alpha and wild HNF-1beta significantly transactivated DPP-IV promoter, but mutant HNF-1alpha and mutant HNF-1beta exhibited low transactivation activity. Moreover, to study whether mutant HNF-1alpha and mutant HNF-1beta change endogenous DPP-IV enzyme activity, we produced four stable cell lines from Caco-2 cells, in which wild HNF-1alpha or wild HNF-1beta, or else respective dominant-negative mutant HNF-1alphaT539fsdelC or dominant-negative mutant HNF-1betaR177X, was stably expressed. We found that DPP-IV gene expression and enzyme activity were significantly increased in wild HNF-1alpha cells and wild HNF-1beta cells, whereas they decreased in HNF-1alphaT539fsdelC cells and HNF-1betaR177X cells, compared with DPP-IV gene expression and enzyme activity in Caco-2 cells. These results suggest that both wild HNF-1alpha and wild HNF-1beta have a stimulatory effect on DPP-IV gene expression, but that mutant HNF-1alpha and mutant HNF-1beta attenuate the stimulatory effect.