The MODY1 gene HNF-4alpha regulates selected genes involved in insulin secretion

Abstract

Mutations in the gene encoding hepatocyte nuclear factor-4alpha (HNF-4alpha) result in maturity-onset diabetes of the young (MODY). To determine the contribution of HNF-4alpha to the maintenance of glucose homeostasis by the beta cell in vivo, we derived a conditional knockout of HNF-4alpha using the Cre-loxP system. Surprisingly, deletion of HNF-4alpha in beta cells resulted in hyperinsulinemia in fasted and fed mice but paradoxically also in impaired glucose tolerance. Islet perifusion and calcium-imaging studies showed abnormal responses of the mutant beta cells to stimulation by glucose and sulfonylureas. These phenotypes can be explained in part by a 60% reduction in expression of the potassium channel subunit Kir6.2. We demonstrate using cotransfection assays that the Kir6.2 gene is a transcriptional target of HNF-4alpha. Our data provide genetic evidence that HNF-4alpha is required in the pancreatic beta cell for regulation of the pathway of insulin secretion dependent on the ATP-dependent potassium channel.

Figure 1

Derivation of β cell–specific HNF-4α knockout mice. (A) HNF-4α^loxP/loxP mice in which exon 2 was flanked by loxP sites were bred to Ins.Cre-transgenic mice expressing Cre recombinase under control of the rat insulin promoter. The resulting HNF-4α^loxP/+; Ins.Cre offspring were then mated with HNF 4α^loxP/loxP homozygotes to obtain HNF-4α^loxP/loxP; Ins.Cre mutants and their littermate controls: HNF-4α^loxP/+, HNF-4α^loxP/loxP, and HNF-4α^loxP/+; Ins.Cre. (B) Primers 1, 2, and 3 (red, blue, and green in A) were used for PCR genotyping of isolated islets from HNF-4α^loxP/loxP; Ins.Cre and HNF-4α^loxP/loxP mice. In the absence of Cre, amplification by primers 1 and 2 results in a 620-bp product. Cre-mediated excision of exon 2 results in a 450-bp product amplified by primers 1 and 3. Quantification of the bands shows that deletion occurs in approximately 70% of all islet cells (note that non–β cells make up 20–30% of the islet cell numbers). (C) Concordant with the results in B, mRNA levels of HNF-4α were reduced by 63% in mutant islets, as determined by quantitative PCR using primers specific to exon 2. *P < 0.05; n = 3 per group. (D and E) Immunostaining of pancreatic sections from adult control (D) and mutant (E) mice using an antibody against HNF-4α indicates that the number of β cells expressing HNF-4α protein is reduced by approximately 85–90% (arrow) in the mutant mouse. Non–β cells in the islet mantle still express HNF-4α protein (arrowhead) in the mutant mouse. Thus, HNF-4α is deleted efficiently and specifically in pancreatic β cells. Magnification, ×200.

Figure 2

Deletion of HNF-4α in β cells results in hyperinsulinemia and impaired glucose tolerance in vivo. (A) In the fed state (Fed) and after an overnight (16-hour) fast (Fast), blood glucose concentrations are decreased in HNF-4α^loxP/loxP; Ins.Cre mice compared with littermate controls. (B) Plasma insulin levels are elevated in HNF-4α mutants in both the fed and overnight-fasted states. (C) Fasting plasma glucagon levels in the mutants were indistinguishable from controls. (D) The ratio of plasma insulin to plasma glucagon is elevated 70% in HNF-4α^loxP/loxP; Ins.Cre mice. (E) Glucose tolerance test. After an overnight fast, 3- to 5-month-old HNF-4α^loxP/loxP; Ins.Cre mice and littermate controls were challenged with 2 grams of glucose per kilogram of body weight. The blood glucose elevation is significantly higher in HNF-4α^loxP/loxP; Ins.Cre mice than in controls, indicating impaired glucose tolerance in the HNF-4α mutants. (F) Following glucose injection (3 g/kg body weight), HNF-4α mutants exhibit a diminished first-phase insulin secretory response in comparison to controls. (G) Insulin tolerance test. Mutant and control mice that had fasted for 4 hours were injected with 0.75 units of insulin per kilogram of body weight. The insulin sensitivity of HNF-4α mutants is indistinguishable from that of controls. *P < 0.05 by Student’s t test or ANOVA; n = 8–13 animals per group for each experiment.

Figure 3

HNF-4α is not required for the maintenance of islet architecture or β cell mass. (A–D) Immunofluorescence detection of the pancreatic hormones insulin, glucagon, and somatostatin, which label β cells, α cells, and δ cells, respectively. Similar to controls (A and C), 3- to 5-month-old HNF-4α^loxP/loxP; Ins.Cre mice contain glucagon-positive α cells (B) and somatostatin-positive δ cells (D). In addition, insulin-positive β cells are centrally located in the islets of both controls and mutants, while less-frequent α cells and δ cells are found along the periphery (A–D), indicating normal islet architecture in the HNF-4α^loxP/loxP; Ins.Cre mice. Magnification, ×200. (E) Point-counting morphometry of 4-month-old mice reveals no significant difference in β cell mass between controls and mutants (control, 0.93 ± 0.17 mg, n = 6; mutant = 1.04 ± 0.22 mg, n = 5; P = NS).

Figure 4

Glucose-stimulated insulin secretion is dysregulated in isolated islets of HNF-4α^loxP/loxP; Ins.Cre mice. (A) Isolated islets from HNF-4α mutants (open circles) lack a robust first-phase insulin secretory response to glucose perifusion compared with that of controls (filled squares), and fail to rapidly terminate insulin secretion upon switching to 0 mM glucose (n = 3). (B) ATP levels in isolated islets from HNF-4α^loxP/loxP; Ins.Cre mice (white bars) stimulated with 2, 5, or 10 mM glucose for 60 minutes are virtually indistinguishable from those of control mice (black bars) (n = 2 per group), indicating that glucose metabolism is not adversely affected in HNF-4α–deficient β cells. (C) The intracellular calcium concentration ([Ca²⁺]_i) increases rapidly in response to 16.7 mM glucose (2.9 nM/s), 1 μM glyburide (10.0 nM/s), and 30 mM KCl in control islets. (D) In contrast, the intracellular calcium concentration increases at a slower rate in response to glucose (0.7 nM/s) and glyburide (2.0 nM/s) in HNF-4α^loxP/loxP; Ins.Cre mice. For all calcium-imaging experiments, n = 4 per group. These are representative plots.

Figure 5

Gene expression analysis in isolated islets of HNF-4α^loxP/loxP; Ins.Cre mice. (A) Levels of mRNA of MODY genes, as determined by real-time PCR. (B) Levels of mRNA of pancreatic enriched transcription factors, as determined by real-time PCR. (C) Levels of mRNA of genes involved in glucose and lipid metabolism, as determined by real-time PCR. (D) Levels of mRNA of genes involved in stimulus-secretion coupling, as determined by real-time PCR. *P < 0.05; n = 3–5 for all PCR experiments. HPRT, hypoxanthine guanine phosphoribosyl transferase. (E) Western blot analysis of HNF-1α in isolated islets normalized to α-tubulin protein levels. C, control; M, mutant. (F) Western blot analysis of Kir6.2 in isolated islets normalized to α-tubulin protein levels. For all protein quantification, n = 3, controls, and n = 2, mutants.

Figure 6

HNF-4α directly activates the Kir6.2 gene. (A) The consensus binding site for HNF-4α is 13 bp long and was derived from 71 known HNF-4α binding sequences from the literature (55) using the program Weblogo (http://weblogo.berkeley.edu/). The size of the letters reflects the frequency at which the nucleotide appears at that position in the binding site. (B) Putative HNF-4α binding site in the Kir6.2 promoter identified using NUBIScan, which uses a transcription factor–binding site–identification algorithm to identify nuclear receptor binding sites. Note that the site located at position –2,300 matches all determinant nucleotides in the HNF-4α consensus site shown in A. (C) EMSA demonstrates that HNF-4α binds to the identified binding site in the Kir6.2 gene as well as the HNF-4α consensus site. In supershift experiments using 2 different antibodies raised against HNF-4α, the identity of the bound protein is confirmed to be HNF-4α. (D) Cotransfection of BHK cells with HNF-4α and pGL3-Kir6.2, expressing luciferase under the control of the 237-bp region of Kir6.2 containing the binding site, results in a dose-dependent increase in luciferase activity, indicating that this element serves as an HNF-4α–dependent enhancer. Mutation of this binding site abolishes the transcriptional activation. Statistical analysis was performed by ANOVA; n = 3 for each transfection condition.