Beta-Cell hyperplasia induced by hepatic insulin resistance: role of a liver-pancreas endocrine axis through insulin receptor A isoform

Abstract

Objective: Type 2 diabetes results from a combination of insulin resistance and impaired insulin secretion. To directly address the effects of hepatic insulin resistance in adult animals, we developed an inducible liver-specific insulin receptor knockout mouse (iLIRKO).

Research design and methods: Using this approach, we were able to induce variable insulin receptor (IR) deficiency in a tissue-specific manner (liver mosaicism).

Results: iLIRKO mice presented progressive hepatic and extrahepatic insulin resistance without liver dysfunction. Initially, iLIRKO mice displayed hyperinsulinemia and increased beta-cell mass, the extent of which was proportional to the deletion of hepatic IR. Our studies of iLIRKO suggest a cause-and-effect relationship between progressive insulin resistance and the fold increase of plasma insulin levels and beta-cell mass. Ultimately, the beta-cells failed to secrete sufficient insulin, leading to uncontrolled diabetes. We observed that hepatic IGF-1 expression was enhanced in iLIRKO mice, resulting in an increase of circulating IGF-1. Concurrently, the IR-A isoform was upregulated in hyperplastic beta-cells of iLIRKO mice and IGF-1-induced proliferation was higher than in the controls. In mouse beta-cell lines, IR-A, but not IR-B, conferred a proliferative capacity in response to insulin or IGF-1, providing a potential explanation for the beta-cell hyperplasia induced by liver insulin resistance in iLIRKO mice.

Conclusions: Our studies of iLIRKO mice suggest a liver-pancreas endocrine axis in which IGF-1 functions as a liver-derived growth factor to promote compensatory pancreatic islet hyperplasia through IR-A.

FIG. 1.

Liver histology and metabolic gene expression in inducible liver-specific insulin receptor knockout (iLIRKO) mice. A: Periodic acid Schiff, hematoxylin-eosin (H/E), and Masson staining of liver sections from fed 6- and 12-month-old male control (upper panel) and iLIRKO (lower panel) mice. The images are representative of four independent experiments. Magnifications 20×. B: Protein extracts of liver from 6-month-old Wt (wild type), IR^(loxP/loxP), and iLIRKO mice were analyzed by Western blot (WB) with anti-IR β-chain, PEPCK, FAS (fatty acid synthase), GK, and β-actin antibodies as indicated in each panel. The blots are representative of five independent experiments. The corresponding autoradiograms were quantitated by scanning densitometry and are expressed as mean ± SEM. (A high-quality digital representation of this figure is available in the online issue.)

FIG. 2.

Inducible liver-specific insulin receptor knockout (iLIRKO) shows insulin resistance in the liver and extrahepatic tissues. In vivo insulin signaling studies were performed in 6-month-old control and iLIRKO mice. Mice of both experimental groups were injected with 1 unit/kg body wt of human insulin (Novo Nordisk) into the peritoneal cavity. After 10 min of treatment, liver, muscle, BAT (brown adipose tissue), and brain were removed. Protein extracts of these tissues were analyzed by Western blot (WB) with anti-IR β-chain, phospho-Akt, phospho-p70, phospho-ERK1/2, and β-actin antibodies. The results are representative of four independent experiments. Histograms summarize the densitometric analysis of the corresponding Western blots.

FIG. 3.

Progressive insulin resistance and glucose intolerance in inducible liver-specific insulin receptor knockout (iLIRKO) mice. A: Insulin tolerance tests were performed on 2-, 4-, 6-, and 12-month-old male control (○) and iLIRKO (●) mice. Fed animals were injected intraperitoneally with 1 unit/kg body wt of human regular insulin. Blood glucose was measured immediately before injection and 15, 30, and 60 min after the injection. Results expressed as percentage of initial blood glucose concentration are means ± SEM (n = 10–20). B: Glucose tolerance tests were performed on 2-, 4-, 6-, and 12-month old control (○) and iLIRKO (●) mice that had been fasted for 16 h. Animals were injected intraperitoneally with 2 g/kg body wt of glucose. Blood glucose was measured immediately before injection and 30, 60, 90, and 120 min after the injection. Results are expressed as mean blood glucose concentration ± SEM; n = 10–20 of each genotype. **P < 0.005; *P < 0.05; iLIRKO versus control.

FIG. 4.

Progressive IGF-1 liver expression, β-cell hyperplasia, and defective insulin secretion with aging in inducible liver-specific insulin receptor knockout (iLIRKO). A: Liver extracts from 6-month-old control and iLIRKO mice with graded IR deletion were analyzed by Western blot with anti-IR β-chain, IGF-1, IGFBP1, IGFBP3, and β-actin antibodies. The Western blot autoradiograms of IGF-1 were quantified by scanning densitometry from four independent experiments. B, upper panel: β-cell mass was evaluated by point-counting morphometry in 6-month-old control (□) and iLIRKO mice (mice with 50% IR, ; mice with 25% IR, ▤; mice with 0% IR, ■), results shown as fold increase of β-cell mass in control animals, n = 7–15 of each group. Lower left panel: plasma insulin content was measured in 6-month-old male control (□) and iLIRKO mice (50% IR, ; 25% IR, ▤; 0% IR, ■) by RIA (Linco); values are expressed as means ± SEM from four animals per genotype. Lower right panel: plasma IGF-1 content was measured in 6-month-old male control (□) and iLIRKO mice (50% IR, ; 25% IR, ▤; 0% IR, ■) by RIA (Diagnostic Systems Laboratories). Values expressed as mean ± SEM from four animals per genotype. C: Liver extracts from 1-year-old control and iLIRKO mice were analyzed by Western blot with anti-IR β-chain, IGF-1, IGFBP1, IGFBP3, and β-actin antibodies. A representative experiment out of four is shown. D, upper panel: β-cell mass was evaluated by point-counting morphometry in 1-year-old control and iLIRKO mice; results are presented as fold increase of the control β-cell mass. Lower left panel: plasma insulin content was measured in 1-year-old male control and iLIRKO mice by RIA (Linco); values are expressed as means ± SEM from four animals per genotype. Lower right panel: plasma IGF-1 content was measured in 1-year-old male control and iLIRKO mice by RIA (Linco); values are expressed as means ± SEM from four animals per genotype.*P < 0.05, **P < 0.005, and ***P < 0.001 iLIRKO versus control.

FIG. 5.

Increase of IR-A isoform in pancreatic islets: β-cell lines expressing IR-A (Rec A), but not IR-B (Rec B), induce proliferation in response to insulin or IGF-1. A: mRNA levels of insulin receptor and the insulin receptor isoforms distribution were analyzed by real-time quantitative PCR in 6-month-old control and iLIRKO mice as described in research design and methods. Values are expressed as mean ± SEM; n = 4 of each genotype, (*P < 0.05; **P < 0.005; ***P < 0.001 iLIRKO vs. control). B: Pancreatic islet proliferation was assessed by BrdU incorporation in 6-month-old control and inducible liver-specific insulin receptor knockout (iLIRKO) mice. The islets were isolated as described and seeded in 96-well plates for 16 h in RPMI medium. After that, the medium was removed and fresh medium with the stimuli (10 nmol/l insulin, ; 10 nmol/l IGF-1, ▤) and BrdU was added for 24 h. Finally, the BrdU incorporation was measured as indicated by the manufacturer. Values are expressed as mean ± SEM; n = 4 of each genotype. Data were subjected to ANOVA with Bonferroni post-test (*P < 0.05; IGF-1 vs. basal in iLIRKO mice). C: IR expression was analyzed by Western blot and RT-PCR in IRLoxP, IRKO, Rec A, and Rec B β-cells. Arrow at the RT-PCR panels indicates the IR-A and IR-B. A representative experiment out of four is shown. D: Functional assessment of insulin receptor reconstitution was carried out by immunoprecipitation of insulin– or IGF-1–stimulated β-cells with antibodies against IR or IGF-1R and subsequent Western blot against phospho-tyrosine residues. E: IRLoxP, IRKO, Rec A, and Rec B β-cells were cultured to 80% confluence and then serum and glucose starved for 4–6 h. Glucose uptake induced by insulin () or IGF-1 (▤) was measured as described in research design and methods. Data were subjected to ANOVA with Bonferroni post-test (*P < 0.05, IRKO and Rec A vs. IRLoxP in basal conditions; #P < 0.05, IGF-1 vs. basal of each cell line). F: IRLoxP, IRKO, Rec A, and Rec B β-cells were cultured to 50% confluence in 10% FBS-DMEM overnight. After that, 10 nmol/l insulin (), 10 nmol/l IGF-1 (▤), or both (■) were added to the wells in serum-starved 5 mmol/l glucose DMEM. After 24 h, the medium was withdrawn and the four cell lines were stained with violet crystal as described in research design and methods. Statistical significance was carried out by Student's t test by comparison of basal conditions with insulin-stimulated conditions of each cell line (*P < 0.05) or basal conditions with IGF-1–stimulated conditions of each cell line (#P < 0.05). G: IRLoxP, IRKO, Rec A, and Rec B β-cells were cultured to 50% confluence in 10% FBS-DMEM overnight. After that, 10 nmol/l insulin (), 10 nmol/l IGF-1 (▤), or both (■) were added to the wells in serum-starved 5 mmol/l glucose DMEM for 24 h. Thymidine incorporation was measured in these conditions as described in research design and methods. Statistical significance was carried out by Student's t test by comparison of basal conditions with insulin-stimulated conditions of each cell line (*P < 0.05) or basal conditions with IGF-1–stimulated conditions of each cell line (#P < 0.05).