Altered β-Cell Distribution of pdx-1 and GLUT-2 After a Short-Term Challenge With a High-Fat Diet in C57BL/6J Mice

All animal experiments were approved by the local ethics committee and followed the guidelines for experiments in animals (European Economic Community Council Directive 86/609/EEC). Four-week-old female C57BL/6J mice obtained from the Bomholtgaard Breeding and Research Center, Denmark, were used for this study. Animals were divided into two groups, receiving either standard rodent food or a high-fat diet (diet nos. D12310 and D12309, respectively; Research Diets, New Brunswick, NJ). On a caloric base, the high-fat diet consisted of 16.4% protein, 25.6% carbohydrates, and 58.0% fat, which corresponds to an energy content of 23.6 kJ/g. The normal diet consisted of 25.8% protein, 62.8% carbohydrates, and 11.4% fat (12.6 kJ/g). The animals were kept under a standard 12 h:12 h light-dark cycle and given food and water ad libitum. After 1, 4, and 8 weeks, blood was drawn in awake mice from the intra-orbital bulbar plexus for the measurement of basal insulin and glucose levels. After centrifugation, plasma was stored at −20°C until further analysis. After 1, 4, and 8 weeks of challenge with high-fat or control diet, the mice were also anesthetized with an intraperitoneal injection of midazolam (Dormicum, 0.14 mg/mouse; Hoffman-La-Roche, Basel) and a combination of fluanison (0.9 mg/mouse) and fentanyl (0.02 mg/mouse) (Hypnorm; Janssen, Beerse, Belgium). Thereafter, a blood sample was taken from the retrobulbar intra-orbital capillary plexus in a heparinized pipette, after which d-glucose (1 g/kg; British Drug Houses, Poole, U.K.) was injected rapidly intravenously. The volume load was 10 μl/g body wt. New blood samples were taken after 1, 5, 20, 50, and 75 min. The samples were taken in heparinized tubes. After immediate centrifugation at 4°C, plasma was separated and stored at −20°C until analysis. At these time points after the start of the respective diets, islets were also isolated for in vitro studies. In brief, after a midline laparotomy, the common bile duct was cannulated and ligated at the papilla of Vater. The pancreas was filled with 3 ml ice-cold Hank’s balanced salts solution (HBSS) supplemented with 0.4 mg/ml collagenase P (Boehringer Mannheim, Mannheim, Germany) before removal and incubation at 37°C for 19 min. After washing three times in HBSS, the islets were handpicked under a stereomicroscope. The isolated islets were then cultured for 48 h in an RPMI medium supplemented with 2.06 mmol/l l-glutamine (Life Technologies, Täby, Sweden), 10% fetal bovine serum, 100 units/ml penicillin, and 0.5 mg/ml streptomycin (all from Kebo Laboratory, Spånga, Sweden) at 37°C in 5% CO₂–95% air. After the overnight incubation, islets were washed in a HEPES buffer containing 125 mmol/l NaCl, 5.9 mmol/l KCl, 1.28 mmol/l CaCl₂, 1.2 mmol/l MgCl₂, 25 mmol/l HEPES, 3.3 mmol/l glucose, and 0.1% human serum albumin (pH 7.36) and preincubated in the same medium for 60 min. Thereafter, three islets were incubated in 100 μl of the medium for 60 min at 37°C in the presence of glucose at varying concentrations. After incubation, aliquots of 25 μl in duplicate were collected and stored at −20°C until analysis of insulin content. Finally, at 1, 4, and 8 weeks, pancreases were also dissected and fixed in Stefanini’s fixative at 4°C overnight. After rinsing in phosphate-buffered saline (PBS), pancreases were cryoprotected by immersion in 20% sucrose in PBS overnight at 4°C, sectioned on a cryostat at 10 μm, and used for immunohistochemical analysis.

Insulin was determined by double antibody radioimmunoassay using guinea pig anti–rat insulin, mono-¹²⁵I-human insulin, and, as standard, rat insulin (Linco Research, St. Charles, MO). Glucose was determined by the glucose oxidase method.

Pancreatic sections were separately incubated with primary antibodies specific for pdx-1 and GLUT2 (Chemicon International, Temecula, CA). In double-immunofluorescence experiments, antisera against insulin and glucagon were used simultaneously. For information on antisera and dilutions, see Table 1. Primary antibodies were diluted in blocking buffer (10 mmol/l phosphate buffer, pH 7.2; PBS containing 0.25% bovine serum albumin and 0.25% triton-X), incubated at 4°C overnight, and rinsed three times for 5 min in PBS. Sections were incubated with fluorochrome-conjugated secondary antibodies (goat anti–guinea pig Ig-Alexa 546 and goat anti–rabbit Ig-Alexa 488, both 1:300 in blocking buffer) for 45 min at room temperature. After rinsing in PBS as before, sections were mounted in PBS:glycerol (1:1) and viewed in an Olympus BX 51 microscope (LRI, Lund, Sweden). Images of the specimen were taken with a digital camera.

From the intravenous glucose tolerance test (IVGTT), the total area under the curve (AUC) for both glucose and insulin concentrations was calculated individually with the trapezoid rule for each animal. Furthermore, the acute insulin response, which is the arithmetic mean of the suprabasal 1- and 5-min insulin value, was used for assessment of insulin secretion. The results were compared group-wise. Statistical differences were computed with the software StatView from the SAS Institute (Cary, NC). Comparisons for differences between high-fat–fed and normal diet–fed mice were performed by the Student’s two-tailed unpaired t test with Bonferroni’s post hoc analysis. For all comparisons, a difference of P < 0.05 was considered significant.

The high-fat diet induced a larger increase in body weight than the normal control diet, with the total weight change after 8 weeks being 48% in high-fat–fed mice compared with 17% in control mice (P < 0.001; Table 2). The high-fat group also showed a higher variation in body weight (coefficient of variation, 11 vs. 4.5%). The high-fat diet induced significant hyperglycemia after 1 week of treatment, which remained throughout the 8-week study period, whereas hyperinsulinemia evolved after 8 weeks of the high-fat diet (Table 2).

Glucose elimination after the intravenous glucose challenge was delayed in high-fat–fed mice compared with mice fed a normal diet after 1, 4, and 8 weeks, as judged by the high AUC_glucose (Table 3). In contrast, the insulin response to the intravenous glucose was not significantly different between the groups at any time point.

Insulin secretion over a wide range of glucose concentrations was similar in islets from controls and high-fat–fed mice after 1 and 4 weeks of treatment, with the exception of enhanced insulin secretion in islets from high-fat mice at 22 mmol/l glucose (786 ± 49 vs. 287 ± 51 pmol/l/three islets at 1 week [P < 0.001] and 777 ± 99 vs. 473 ± 86 pmol/l/three islets at 4 weeks [P = 0.003]). In contrast, after 8 weeks, insulin secretion was enhanced at all glucose concentrations in islets from high-fat–fed mice compared with that in islets from control mice (Fig. 1).

Double immunofluorescence showed that the relative distribution of insulin- and glucagon- containing cells was not changed in animals receiving a high-fat diet (Figs. 2A and B). Similarly, the ObRb immunoreactivity was not altered (Fig. 2C and D). In contrast, the immunocytochemical appearance of islet pdx-1 and GLUT2 differed between the two groups. Thus, in control mice, staining with an antiserum specific for pdx-1 showed immunoreactivity in both the cytoplasm and the nucleus of β-cells, and in most islets, nuclear staining was more prominent than cytoplasmic staining. In contrast, pancreases from high-fat–fed mice displayed pdx-1 immunoreactivity mostly in the cytoplasm, and only weak nuclear staining was observed (Fig. 3A and B). In mice fed a normal diet, GLUT2 was localized to the majority of islet cells, and the immunoreactivity was confined to the cell membrane. The staining intensity was decreased in sections from animals that had received a high-fat diet, and in these islets, accumulation of immunoreactive material was located over the cytoplasm (Fig. 3C and D, arrows). This difference in pdx-1 and GLUT2 immunoreactivity between the groups occurred after 4 and 8 weeks of the high-fat diet, but not after 1 week.

This study monitored short-term islet effects of a high-fat diet in C57BL/6J mice to increase our knowledge on early islet adaptation to high-fat feeding–induced insulin resistance. This mouse strain has been shown to be susceptible to this diet and develops glucose intolerance more readily than other strains (3,16). The long-term effect of a high-fat intake on glucose metabolism in this mouse strain has been established in several studies (2–7,16), but so far, no early and sequential data are available that show after what time period these effects begin to become manifest. In this study, we related the islet morphology to changes in glucose tolerance and insulin secretion during the first 8 weeks of high-fat feeding in mice. We found that impairment of glucose elimination occurred after 1 week on the high-fat diet. The impaired glucose elimination was associated with insulin responses that were not potentiated, i.e., the insulin resistance was not accompanied by a compensatory increase in insulin secretion, suggesting that the high-fat diet had compromised islet adaptation. Interestingly, adaptation was evident in vitro after 8 weeks when glucose-stimulated insulin secretion was exaggerated. This shows that a considerable time (2 months) is required for a detectable islet compensation. However, in spite of this increased glucose-stimulated insulin secretion in vitro, no such augmentation was evident in vivo, suggesting that integrative actions in vivo opposed primary islet adaptation.

In association with the development of insulin resistance and glucose intolerance, we found compromised β-cell translocation of the homeobox transcription factor pdx-1 after the high-fat diet. pdx-1 is known to be associated with normal proliferation of pancreatic cells during development, and it has been shown that targeted disruption of the pdx-1 gene results in pancreatic agenesis, whereas heterozygous pdx-1–deleted mice develop glucose intolerance in adulthood (18). This would suggest an important role of this transcription factor with regard to β-cell function. We found that the inability of the islets to adapt to the insulin resistance after a high-fat diet was associated with compromised translocation of pdx-1 to the nucleus in that pdx-1 was localized mainly in the cytoplasm in high-fat animals, whereas in control mice, both a nuclear and cytoplasmic localization was observed. pdx-1 is normally translocated into the nucleus and binds to the insulin gene, which leads to the transcription of insulin mRNA; in fact, together with E47 and β2/neuroD, pdx-1 is recognized as a main effector of insulin promoter transactivation (19,20). The mechanism of the compromised nuclear translocation of pdx-1 in high-fat–fed mice is not known. Activation of pdx-1 seems to be induced by nutrient stimulation (21) and, as shown by Macfarlane et al. (22), high glucose induces nuclear translocation of pdx-1 in rat islets by a mechanism governed by SAPK2. Furthermore, pdx-1 expression has been shown to be impaired in a patient with persistent hyperinsulinemic hypoglycemia of infancy (23).

We also found that the translocation of GLUT2 to the plasma membrane was compromised in mice fed the high-fat diet, which would perturb the islets’ capacity to adequately respond to glucose. This result is analogous to reduced expression of GLUT2 in several animal models of diabetes (11–14). Interestingly, GLUT2 expression in islets is known to be regulated by pdx-1, and therefore, the impaired translocation of these two factors might have a common mechanism (24). Finally, we also studied the immunolocalization of ObRb in islets from diet-treated mice. Islet β-cells express this functional form of the ObRb (15), and circulating leptin is increased in the mice fed a high-fat diet (16,17). The hyperleptinemia may be a consequence of the obesity due to increased expression in triglyceride-storing adipocytes but may also be a sign of leptin resistance. Leptin has been demonstrated to inhibit insulin secretion in vitro (15,25,26). However, in vivo studies in diabetic models have shown that leptin treatment improves insulin secretion in ob/ob mice (27), and overexpression of the leptin receptor in a model homozygous for a mutation of the ObRb, the Zucker diabetic rat, restores GLUT2 expression and insulin secretion (28). By using an ObRb-specific antibody, we observed immunoreactivity for this receptor in the β-cells of both normal and high-fat islets with no apparent difference. This finding would suggest unaltered islet ObRb expression in this model. It would therefore be of interest to also examine the effect of leptin in this animal model of glucose intolerance.

In conclusion, an 8-week high-fat diet challenge to glucose-intolerant prone C57BL/6J mice results in impaired glucose tolerance and insufficient compensation in insulin secretion. This is accompanied by compromised translocation of pdx-1 to the nucleus and GLUT-2 to the plasma membrane but unaltered expression of the ObRb. Our results therefore indicate that high-fat feeding impairs β-cell trafficking of factors of importance both for β-cell glucose recognition and insulin gene expression.

This study was supported by the Swedish Medical Research Council (grant number 6834), Novo Nordisk, Albert Påhlsson, Crafoord, the Swedish Diabetes Association, and the Faculty of Medicine, Lund University.

The expert assistance of Lilian Bengtsson, Kerstin Knutsson, and Lena Kvist is gratefully acknowledged. The authors also thank Dr. P. Serup, Novo Nordisk, for the kind gift of the pdx-1 antiserum.