Pancreatic and duodenal homeobox-1 (PDX1) contributes to β-cell mass expansion and proliferation induced by Akt/PKB pathway

ABSTRACT

Maintenance of pancreatic β-cell mass and function is fundamental to glucose homeostasis and to prevent diabetes. The PI3 K-Akt-mTORC1 pathway is critical for β-cells mass and function, while PDX1 has been implicated in β-cell development, maturation, and function. Here we tested whether Akt signaling requires PDX1 expression to regulate β-cell mass, proliferation, and glucose homeostasis. In order to address that, we crossed a mouse model overexpressing constitutively active Akt mutant in β-cells (β-caAkt) with mice lacking one allele of PDX1gene (β-caAkt/pdx1^+/-). While the β-caAkt mice exhibit higher plasma insulin levels, greater β-cell mass and improved glucose tolerance compared to control mice, the β-caAkt/pdx1^+/- mice are hyperglycemic and intolerant to glucose. The changes in glucose homeostasis in β-caAkt/pdx1^+/- were associated with a 60% reduction in β-cell mass compared to β-caAkt mice. The impaired β-cell mass in the β-caAkt/pdx1^+/- mice can be explained by a lesser β-cell proliferation measured by the number of Ki67 positive β-cells. We did not observe any differences in apoptosis between β-caAkt/pdx1^+/- and β-caAkt mice. In conclusion, PDX1 contributes to β-cell mass expansion and glucose metabolism induced by activation of Akt signaling.

KEYWORDS:

• PDX1
• β-cell proliferation
• AKT/PKB signaling
• diabetes
• β-cell expansion

Introduction

Type 2 diabetes (T2D) is characterized by defective β-cell expansion and adaptation in conditions of increased insulin demand. In addition to alteration in mass, defects in insulin secretion also contribute to the development of hyperglycemia during the pathogenesis of T2D.^1,2 The capacity of β-cells to increase its mass in response to insulin resistance is critical to maintain glucose homeostasis resulting from a dynamic balance of neogenesis, proliferation, cell size and apoptosis.³ The contribution of these different components in regulation of β-cell mass under normal or regenerative conditions varies but replication of β-cells is a major determinant to maintain beta cell mass under basal conditions.^4,5 How β-cells adapts to insulin resistance is not completely understood but published evidence suggest that Akt is involved.¹

Akt is a converging pathway for growth factors, insulin, incretins, and glucose in β-cells.^6–11 Several lines of evidence suggest that activation of the IRS2/phosphoinositide 3-kinase (PI3 K)/Akt pathway plays a critical role in the regulation of β-cell mass and function.^8,12-20 Transgenic mice with gain and loss of function of Akt have provided further evidence for the role of this kinase in β-cells.^18,20,21 Overexpressing a constitutively active form of Akt1 specifically in islet β-cells (β-caAkt mice) induces striking increases in β-cell mass, proliferation, neogenesis and cell size as well as improved glucose tolerance and insulin secretion.^20,21 The increase in β-cell proliferation was associated with increased cyclin D1, D2, p21 levels, and Cdk4.⁵ On the other hand, overexpression of a kinase-dead Akt1 in β-cells provokes hypoinsulinemia and mild hyperglycemia accompanied by reduced glucose-stimulated insulin secretion and impaired glucose tolerance, increasing the susceptibility to diabetes.¹⁸

Pancreatic duodenal homeobox-1 (PDX1), also known as insulin promoter factor 1 (IPF1) in humans, is a homeobox gene that plays critical roles in the developing pancreas.²² PDX1 is required for the development of both pancreatic exocrine and endocrine cells since inactivating mutation in exon 1 of IPF1 in humans and targeted disruption of exon 2 in mice lead to pancreas agenesis.^22,23 PDX1 heterozygous mice develop glucose intolerance later on in life due to alterations in β-cell survival.²⁴ In addition, disruption of PDX1 in developing β-cells using insulin-driven Cre recombinase results in hyperglycemia, compromised insulin-expressing cells, and increased glucagon-expressing cells.^25,26 Interestingly, PDX1 expression becomes restricted to mature β-cells, suggesting that PDX1 plays a role in β-cells function. Indeed, conditional deletion of PDX1 in 80% of mature β-cells (after the first four weeks of age in mice) in a tamoxifen-inducible conditional knockout model induces diabetes by switching the identity of β-cells to a α-cell-like phenotype.²⁷ Importantly, PDX1 has a pro-survival role in β-cell²⁸ and it regulates the expression of a number of β-cell genes, e.g., insulin, islet amyloid, glucokinase, and the glucose transporter 2.^29–34 Therefore, PDX1 is critical for pancreas development, β-cell maturation, and function. The published evidence underscores the importance of PDX1 in development, insulin secretion and β-cell survival, yet the role of PDX1 in proliferation is less known. For instance, it has been shown that PDX1 deficiency during β-cell adaptation to insulin resistance impairs β-cell proliferation and negatively impact β-cell expansion.³⁵

There are several lines of evidence linking insulin and/or Akt signaling pathway to PDX1: i) Foxo1 acts as a repressor of Foxa2 which activates PDX1 promoter; ii) Diabetes in the insulin receptor substrate knockout mice (IRS2KO) is rescued by genetically haploinsufficiency of Foxo1;³⁶ iii) Insulin protects islets from apoptosis via PDX1, partially mediated by Akt²⁸ iv) GSK3β phosphorylates PDX1 and tags for proteasome degradation³⁷ and v) expression of a kinase dead mTOR decreases PDX1 protein content in islets.³⁸ However, the role played by PDX1 in Akt-induced β-cell expansion has not been addressed.

In the present study, we assessed the role of PDX1 in β-cell proliferation by crossing mice overexpressing a constitutively active Akt mutant in β-cell with the PDX1 haploinsufficient mice. Remarkably, deficiency in PDX1 blunted all the improvements in glucose tolerance, β-cell proliferation, and β-cell mass induced by increased Akt signaling. These studies show that β-cell expansion induced by Akt is mediated, at least in part, by PDX1.

Methods

Mice

All procedures were approved by University of Miami IACUC (#15-196). Mice overexpressing constitutively active Akt1 in β-cells (β-caAkt) have a C57BL/6 J background and have been described previously.^5,20,39 PDX1 haploinsufficient mice have been also described previously.^22,24 We generated a new mouse model by crossing the β-caAkt mouse with the pdx1^+/- mouse to study the role of PDX1 in the context of Akt-induced β-cell proliferation and expansion. Male mice were used in the experiments.

Metabolic studies

Fasting blood samples were obtained from the tail vein after fasting the mice overnight. Glucose was measured on whole blood using Accu-Chek II glucometer (Roche Diagnostics, Indianapolis, IN). Plasma insulin levels were determined using Rat insulin ELISA kit (Linco, Saint Louis, MO). Glucose tolerance tests were performed in 18-h–fasted animals by injecting glucose (2 g/kg) intraperitoneally as described previously.⁴⁰

Islet morphometry and proliferation analysis

Pancreata were fixed overnight in formalin solution, and embedded in paraffin using standard techniques as described previously.²⁰ Sections (5 μm) were placed on glass slides, deparaffinized with xylene, and stained for insulin by using guinea pig anti-human insulin AB (1:300; BioGenex Laboratories, San Ramon, California, USA) and the Histomouse-SP kit (Zymed Laboratories Inc., South San Francisco, California, USA) using 3-amino-9-ethyl carbazole (AEC) as chromogen according to the manufacturers’ instructions. Endogenous peroxidase was quenched with 0.6% H202/methanol for 30 minutes. The β-cell mass was calculated by point counting morphometry from five insulin-stained sections (5 µm) separated by 200 µm as previously described.⁴⁰ Proliferation was assessed in insulin- and rabbit anti-human Ki67 (1:200; # Ki67p/VPK451, Leicabiosystems)-stained sections. At least 2000 insulin-stained cells were counted for each animal.

Caspase 3 immunostaining

Sections were mounted on poly-L-lysine-coated slides, deparaffinized and microwaved. Guinea pig anti-insulin (DAKO, Carpinteria, CA) and rabbit anti-cleaved caspase 3 (1:50; Cell Signaling, Beverly, MA) were applied overnight at 4°C and washed in PBS. After washes, the slides were incubated simultaneously with anti-guinea pig-FITC and Anti Rabbit-Cy3 for 4 h at room temperature. Slides were washed and covered it with DAPI.

Imaging

Leica DM5500 immunofluorescence microscope with motorized stage and Image-Pro analysis software was used for imaging acquisition.

Statistical analysis

Quantitative data are presented as the mean ± s.e.m. One-way ANOVA (Analysis of Variance) followed by tukey posthoc test was used to compare all four groups. All statistical analysis was performed using Prism 7 and p < .05 was considered statistically significant.

Results

PDX1 is required for the β-caAkt mouse metabolic phenotype

It is well documented that Akt regulates β-cell mass, survival, proliferation, and function.^5,18,20 Mice expressing a constitutively active Akt specifically in β-cells (β-caAkt) exhibit a remarkable increase in β-cell mass due to high proliferative rates, and greater insulin levels in the blood.^20,21 Since PDX1 is critical for a proper function of mature β-cells by regulating important genes and insulin/Akt signaling modulates PDX1 expression, we tested the hypothesis that PDX1 mediates some of the biological functions induced by Akt. To address that, we generated β-caAkt mice deficient in one allele of pdx1 gene (pdx1^+/-). The pdx1^+/- and β-caAkt mice are normoglycemic in fed conditions (Figure 1A). However, glucose levels in the β-caAkt/pdx1^+/- mice are higher compared to WT mice (Figure 1A). Fasting blood glucose levels are also increased in the β-caAkt/pdx1^+/- mice compared to β-caAkt mice (Figure 1B). As expected, β-caAkt mice exhibit high plasma insulin levels compared to WT and pdx1^+/- mice (Figure 1C). Interestingly, insulin levels in β-caAkt/pdx1^+/- were restored to the levels observed in WT mice (Figure 1C). Examination of glucose tolerance in 8-week-old mice showed that while pdx1^+/- mice have normal tolerance to glucose, β-caAkt mice exhibited improved glucose tolerance (Figure 1D,E). However, deletion of one allele of pdx1 induces glucose intolerance when compared to WT mice and pdx1^+/- (Figure 1D,E).

Figure 1. Assessment of glucose homeostasis and glucose tolerance in β-caAkt and pdx1^+/- intercross. (A) Fed glucose levels in blood; (B) Fasting glucose levels; (C) Fed plasma insulin levels; (D) Blood glucose levels after intraperitoneal glucose injection (2 g/kg) and (E) Area under the curve of values obtained in (C). * p < .05, ** p < .01 and *** p < .001 compared to WT. # p < .05, ## p < .01 and ### p < .001 compared to β-caAkt. N = 4–13.

Overexpression of caAkt induces increase in β-cell mass at least in part by a PDX1-dependent mechanism

The data presented above suggest that the improved glucose tolerance and hyperinsulinemia found in β-caAkt mice were significantly affected by decreased PDX1 levels, suggesting that changes in β-cell mass could be responsible for these observations. β-cell mass in the β-caAkt/pdx1^+/- intercross mice was then assessed by islet morphometry (Figure 2A-D). In adult pancreas from 12-14-week-old male mice, pdx1^+/- mice exhibited comparable β-cell mass to WT mice (Figure 2E). Overexpression of the caAkt mutant was associated with six-fold induction in β-cell mass (Figure 2E). This striking increase in the β-caAkt mice was decreased by 75% in β-caAkt mice lacking one PDX1 allele (β-caAkt/pdx1^+/- mice; Figure 2E).

Figure 2. PDX1 deficiency limits β-cell mass expansion induced by Akt. (A-D) Representative pancreatic morphology in WT, pdx1^+/-, β-caAkt and β-caAkt/pdx1^+/- mice stained for insulin (red) and counterstained with hematoxylin (blue). (E) β-cell mass quantification performed at 10 weeks of age. * p < .05 and *** p < .001 compared to WT. ### p < .001 compared to β-caAkt. N = 4.

Normal levels of PDX1 are required for Akt-induced β-cell proliferation

In order to understand the mechanism behind the blunted β-cell mass expansion in β-caAkt/pdx1^+/- mice, we assessed proliferation and apoptosis in the four groups (Figures 3 and 4, respectively). Analysis of proliferation by immunostaining for Ki67 in β-cell nuclei demonstrates that compared to WT mice, pdx1^+/- mice exhibit a trend to decreased proliferation but this was not statistically significant (Figure 3B). Proliferation is remarkably increased in β-caAkt mice compared to WT mice. Interestingly, β-cell proliferation in the β-caAkt/pdx1^+/- mice is not statistically different to WT mice (Figure 4AB). The β-caAkt mice exhibited increased apoptosis suggesting that these mice have high β-cell turnover due to concomitant elevation of proliferation and apoptosis (Figures 3B and 4B). Apoptosis in β-caAkt/pdx1^+/- mice was elevated to levels similar to those observed in β-caAkt mice (Figure 4A,B).

Figure 3. PDX1 expression is important for Akt-induced β-cell proliferation. (A) representative images of WT, pdx1^+/-, β-caAkt and β-caAkt/pdx1^+/- mice after staining paraffin pancreas slices for Dapi (blue), Insulin (green) and Ki67 (red). Arrows denote Ins⁺/Ki67⁺ cells. Images were taken with 40x objective (B) Quantification of Ins⁺/Ki67⁺ cells among all β-cells. *** p < .001 compared to WT. ## p < .01 compared to β-caAkt. N = 3–4.

Figure 4. Quantification of β-cell death. (A) representative images of WT, pdx1^+/-, β-caAkt, and β-caAkt/pdx1^+/- mice after staining paraffin pancreas slices for Dapi (blue), Insulin (green) and Caspase 3 (red). Arrows denote Ins⁺/Caspase3⁺ cells. Images were taken with 40x objective (B) Quantification of Ins⁺/Caspase3⁺ cells among all β-cells. * p < .05 compared to WT. N = 3–4.

Discussion

The current studies were designed to assess the role of PDX1 in the phenotype observed in mice with activation of Akt signaling in pancreatic β-cells. We demonstrate that PDX1 haploinsufficiency is sufficient to limit β-cell expansion induced by activation of Akt signaling in vivo (Figure 2) and that PDX1 levels are essential for β-cell proliferative signals induced by activation of Akt (Figure 3). Collectively, the current data place PDX1 as a central modulator of Akt-induced β-cell expansion.

Challenged by insulin resistance, β-cells are required to increase insulin secretion to maintain glucose at normal levels. Acutely, there is an increment in insulin secretion stimulated by increasing levels of glucose imposed by insulin resistance. The current evidence suggest that hyperinsulinemia is an important component to stimulate β-cell expansion in these conditions.^41,42 Downstream of insulin receptor signaling, Akt/PKB pathway is critical for this response and its overactivation induces a remarkable increase in β-cell mass due to higher rates of β-cell proliferation (Figures 3 and 4, and²⁰). Previous studies suggested that IRS2 and PDX1 are important for β-cells compensation to insulin resistance.^35,43 Using two independent mouse models of insulin resistance – insulin receptor (IR)/IR substrate–1 (IRS-1) double heterozygous mice and liver-specific IR KO (LIRKO) mice – Kulkarni and coworkers showed that the compensatory islet-growth response to insulin resistance is impaired by PDX1 heterozygosity.³⁵ Our studies extend these observations by placing PDX1 downstream of Akt signaling and suggest that β-cell mass expansion in conditions of insulin resistance could be mediated at least in part by an IR/IRS2/Akt/PDX1 axis.

We confirmed in the present studies that β-caAkt and WT mice have similar blood glucose despite high plasma insulin levels as previously reported.²⁰ These results suggest peripheral insulin resistance. In contrast, Tuttle et al. have shown a 20-25% decrease in glucose levels concomitantly with an 4-5-fold increase in plasma insulin.²¹ Although they did not show any difference in the insulin tolerance test between β-caAkt and WT mice, the fact that a 4-5-fold increase in plasma insulin only slightly impact glucose levels also suggest a degree of peripheral insulin resistance and/or insulin clearance.

The fed hyperglycemia and glucose intolerance in βcaAkt/pdx1^+/- mice was associated with increased β-cell mass compared to WT and pdx1^+/- mice, highlighting the important role of PDX1 in β-cell function and not only in β-cell mass. These findings suggest that PDX1 levels are essential in coordinating appropriate function in proliferative conditions induced by Akt and underline the role of this transcription factor in maintaining a differentiated phenotype in conditions of increase proliferation. In contrast to previous studies, our results showed that pdx1^+/- mice exhibited normal glucose tolerance. This is most likely explained by the use of young male mice (2-month-old) in our studies.

Evidence that Akt-PDX1 axis plays a major role in β-cell proliferation is based on the following findings: i) PDX1 haploinsufficiency partially blocks the proliferation and β-cell mass observed in β-caAkt mice (Figure 3); ii) Studies performed in islets from β-caAkt mice and caAkt MIN6 cells showed that Akt regulates protein levels of cyclin-dependent kinase 4 and cyclin Ds;^5,39 The mechanisms whereby PDX1 regulates β-cell proliferation are unclear but it is possible that PDX1 could regulate transcription of critical cell cycle regulators by binding directly to cis regulatory elements or indirectly by regulating transcription of other molecules that will alter transcription of components of the cell cycle machinery;^5,44 iii) decreased expression of β-catenin in β cells from PDX1 heterozygous islet provides additional potential mechanisms in coordinating β cell expansion by the Akt/PDX1 axis.³⁵ Future studies can be designed to assess additional mechanisms by performing gene expression in islets from β-caAkt/pdx1^+/- mice.

PDX1 also regulates β-cell mass by regulating cell survival.^24,28 The current studies showed a trend to increased apoptotic rates in PDX1 heterozygous mice (Figure 3) by staining β-cell (insulin⁺ cells) for caspase3, confirming previous findings.²⁴ Interestingly, higher apoptotic rates were also observed in β-caAkt mice (Figure 4). However, we did not observe any difference between the β-caAkt and β-caAkt/pdx1^+/- mice indicating that Akt-induced apoptosis in β-cells islets in the β-caAkt mice is not mediated by PDX1. It also indicates that the differences in β-cell mass between the β-caAkt and β-caAkt/pdx1^+/- mice are not explained by changes in apoptosis. The unexpected increased apoptosis in β-caAkt mice was observed previously by Tuttle et al.²¹ Given that Akt has been described as an oncogene, it is not unexpected to observe a concomitant increase in proliferation and apoptosis.⁴⁵ Indeed, high tissue turnover (increased proliferation and apoptosis) could provide an advantage to tissues with high growth rate after activation of oncogenic pathways.⁴⁵

In summary, the current work uncovers that Akt/PDX1 axis regulates β-cell proliferation and survival.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.

Additional information

Funding

This work was supported by National Institutes of Health Grant RO1-DK-073716 and RO1-DK084236 and Juvenile Diabetes Research Foundation (JDRF) Grant 17-2013-416.