Pancreatic beta cell senescence contributes to the pathogenesis of type 2 diabetes in high-fat diet-induced diabetic mice
- First Online:
- Cite this article as:
- Sone, H. & Kagawa, Y. Diabetologia (2005) 48: 58. doi:10.1007/s00125-004-1605-2
- 1.7k Downloads
During the pathogenesis of type 2 diabetes insulin resistance causes compensatory proliferation of beta cells. As beta cells have a limited replication potential, this compensatory proliferation might accelerate cellular senescence and lead to diabetes. We examined the cellular senescence of beta cells after proliferation during lipoglucotoxicity.
Senescence-associated markers in beta cells were examined in nutrient-induced diabetic C57BL/6J mice that were fed a high-fat diet. After 4 and 12 months of the high-fat diet, intraperitoneal glucose tolerance tests (IPGTTs) and histochemical analyses of Ki-67, p38, senescence-associated beta-galactosidase, and beta cell mass were performed.
At 4 months, the AUC for plasma insulin levels during the IPGTT (AUCinsulin) was higher, beta cell mass was 3.1-fold greater, and the proliferation of beta cells was 2.2-fold higher than in the control group. However, at 12 months, AUCinsulin declined, the frequency of Ki-67-positive beta cells decreased to one-third that of the control group, and the senescence-associated, beta-galactosidase-positive area increased to 4.7-fold that of the control group. Moreover, small amounts of p38, which is induced by oxidative stress and mediates cellular senescence, were found in beta cells from the high-fat diet group, but not in beta cells from the control group. Furthermore, the senescence-associated, beta-galactosidase-positive area in the high-fat diet group had a highly significant negative correlation with AUCinsulin (r=−0.852, p<0.01).
Beta cell senescence occurred in diet-induced type 2 diabetes and led to insufficient insulin release. These findings suggest that cellular senescence contributes to the pathogenesis of diet-induced diabetes.
KeywordsCellular proliferation Cellular senescence High-fat diet Pancreatic beta cell Type 2 diabetes
AUC for plasma glucose levels during IPGTT
AUC for plasma insulin levels during IPGTT
extracellular signal-regulated kinase
intraperitoneal glucose tolerance test
c-Jun N-terminal kinase
mitogen-activated protein kinase
reactive oxygen species
- SA beta-gal
terminal deoxynucleotidyl transferase-mediated dUTP nick-end labelling
Defective insulin secretion in type 2 diabetes mellitus is caused by beta cell dysfunction and reduced beta cell mass. Beta cell dysfunction is caused by chronic hyperglycaemia and/or hyperlipidaemia respectively referred to as glucotoxicity and lipotoxicity, which have been previously reviewed in detail [1, 2, 3].
Long-term regulation of beta cell mass is determined by the balance of beta cell growth, including replication and neogenesis, and beta cell loss through apoptosis, necrosis and senescence [4, 5, 6, 7]. Decreased beta cell mass occurs in humans with type 2 diabetes and animal models of type 2 diabetes [6, 7, 8, 9, 10, 11, 12]. Chronic hyperglycaemia leads to progressive loss of beta cell mass with a prolonged increase in the rate of beta cell apoptosis without a compensatory increase in beta cell growth. In addition to necrosis, markedly increased beta cell apoptosis is regarded as one of the major factors in the pathogenesis of diabetes. The cellular mechanisms of increased beta cell apoptosis, which include hyperglycaemia-induced oxidative stress [8, 12], are being intensively studied [13, 14, 15, 16, 17, 18].
Beta cell senescence also affects beta cell mass, but has not been studied in diabetes. Mammalian cells, with the exceptions of germ line cells and stem cells, have a limited replication potential in vivo, defined as replicative senescence, which amounts to irreversible growth arrest after a limited number of cell divisions . Cellular senescence is also associated with reactive oxygen species (ROS), the incessant damaging products generated from aerobic metabolism such as oxidative stress [20, 21, 22, 23]. During the pathogenesis of type 2 diabetes, beta cell proliferation increases to compensate for the increased insulin demand caused by insulin resistance, and the generation of ROS is induced by hyperglycaemia [6, 7, 8, 12, 24]. It would seem that beta cell proliferation and/or ROS generation accelerate cellular senescence and lead to diabetes. Type 2 diabetes in C57BL/6J mice is provoked by a high-fat diet and is associated with obesity, hyperglycaemia and hyperinsulinaemia [25, 26], which induce beta cell proliferation and ROS production in beta cells. To test the hypothesis that cellular senescence occurs in beta cells and contributes to the pathogenesis of type 2 diabetes, we fed C57BL/6J mice a high-fat diet for 12 months and examined cellular proliferation and ageing markers in the beta cells.
Materials and methods
Six-week-old C57BL/6J male mice were purchased from Clea Japan (Tokyo, Japan). The control group was fed a normal diet (AIN-93; 14% protein, 72% carbohydrates, 4% fat) , with free access to water. The high-fat diet group was given free access to a high-fat diet, in which beef tallow replaced 36% of the carbohydrates in the normal diet (high-fat diet: 14% protein, 36% carbohydrates, 40% fat). The mice were housed in a temperature-controlled room with a 12 h light–dark cycle. All experiments were approved by the Animal Ethics Committee of Kagawa Nutrition University and complied with the “Principles of laboratory animal care” (NIH publication no. 85-23, revised 1985).
Blood analysis and glucose tolerance tests
For the intraperitoneal glucose tolerance test (IPGTT), mice fasted for 12 h and blood was drawn from the tail vein at 0, 30, 60 and 120 min after i.p. injection of glucose (2 g/kg of body weight). To measure plasma lipids, blood was collected from the tail vein before the i.p. injection of glucose. Plasma glucose, NEFA, triglycerides and total cholesterol were measured by commercial kits (Wako Pure Chemical Industries, Osaka, Japan). Plasma insulin was measured using a rat insulin enzyme immunoassay system (Amersham Biosciences, Piscataway, NJ, USA). At each time point, blood samples (approximately 70 μl) were taken in one capillary tube, with the exception of the blood samples taken at the 0 time point (approximately 100 μl in 1.5 capillary tubes). Because of the small body weight of the mice, the blood volume needed for microanalyses was only 2 μl, with the exception of the insulin analyses (20 μl). The trapezoid rule was used to determine the AUC for glucose and insulin concentrations in each animal. The insulinogenic index was calculated by dividing the AUC for plasma insulin levels during IPGTT (AUCinsulin) by the AUC for plasma glucose levels during IPGTT (AUCglucose) (0–30 min).
Isolation of islets and islet study
Islets were isolated from the pancreas by collagenase digestion and collected under a microscope after 4 and 12 months of the high-fat diet. Islets were cultured for 12 h in RPMI 1640 medium containing 11.1 mmol/l glucose, 10% heat-inactivated fetal bovine serum, 100 IU/ml penicillin and 100 μg/ml streptomycin, as previously described [28, 29]. Then, size-matched islets were pre-incubated for 60 min at 37 °C with KRB containing 0.2% BSA and 2.8 mmol/l glucose. After pre-incubation, 10 islets were stimulated with KRB at the indicated concentration of glucose for 1 h at 37 °C. Immediately thereafter, the supernatant was collected and stored at −80 °C until the insulin assay. Then, islets were washed with PBS three times and homogenised with acid ethanol (0.18 mol/l HCl in 95% ethanol) to determine the insulin content.
For double-immunohistochemical staining of insulin and glucagon, Ki-67, or phospho-p38 (Thr180/Tyr182), pancreases from mice after 4 and 12 months of the experimental period were weighed, fixed overnight in 4% paraformaldehyde at 4 °C, embedded in paraffin, and sectioned. Three sections (3 μm) separated by 200 μm were incubated with rabbit anti-human glucagon polyclonal antibody (DakoCytomation, Carpinteria, CA, USA) diluted to 1:100, rat anti-mouse Ki-67 monoclonal antibody (TEC-3; DakoCytomation) diluted to 1:50, or mouse anti-human phospho-p38 monoclonal antibody (28B10; New England BioLabs, Beverly, MA, USA) diluted to 1:50. As secondary antibodies, swine anti-rabbit IgG (DakoCytomation) diluted to 1:1,000, rabbit anti-rat IgG antibody (DakoCytomation) diluted to 1:200, or rabbit anti-mouse IgG antibody (Rockland, Gilbertsville, PA, USA) diluted to 1:1,000 were used. We used 3,3′-diaminobenzidine tetrahydrochloride (DAB; Dojin, Tokyo, Japan) as substrate chromogen. Next sections were washed with 0.1 mol/l glycine–HCl buffer (pH 2.2) following incubation with guinea pig anti-porcine insulin antibody (DakoCytomation) diluted to 1:200. As a secondary antibody, rabbit anti-guinea pig IgG coupled to peroxidase (Rockland) diluted to 1:1,000 was used. Sections were stained with 3-amino-9-ethylcarbazole (Sigma-Aldrich, St. Louis, MO, USA) as substrate chromogen and counterstained in haematoxylin.
Pancreatic islet morphology
For morphometric analysis of beta cell mass, three sections of each pancreas were covered systematically by capturing images from non-overlapping fields with a digital camera (Nikon 950; Nikon, Tokyo, Japan). The islet area and the section area were analysed using NIH Image 1.63 software, as previously described . The percentage of beta cell area in the islet area was calculated by dividing the area of all insulin-positive cells by the islet area and by multiplying by 100. The islet mass was calculated by multiplying the pancreas weight by the percentage of islets per pancreas. The beta cell mass was calculated by multiplying islet mass by the percentage of beta cells per islet. To assess the relative rate of beta cell neogenesis from exocrine duct cells, we quantified the percentage of exocrine duct cells that were immunoreactive for insulin, as previously described . The resulting data were expressed as the percentage of duct cells that were positive for insulin in each pancreas, and the mean was calculated for each group.
Beta cell proliferation
Beta cell replication per islet was calculated by comparing the insulin and Ki-67 stains and averaging the number of Ki-67-positive beta cells from approximately 80 islets per mouse.
Detection of cellular senescence
Cellular senescence was evaluated by immunohistochemistry with cellular senescence marker phospho-p38 and by senescence-associated beta-galactosidase (SA beta-gal) staining. The frequency of beta cell senescence determined by immunohistochemistry was calculated using the method described for cell replication. The detection of SA beta-gal was performed as previously described  with minor modifications. After 12 months of the experimental period, pancreases were fixed for 3 h in 4% paraformaldehyde at 4 °C, placed overnight in PBS (pH 7.2) containing 30% sucrose at 4 °C, embedded in Tissue-Tek OCT compound (Miles, Elkhart, IN, USA), frozen, and sectioned into 8 μm thick slices by a refrigerated microtome. SA beta-gal was detected by a senescence detection kit (BioVision, Mountain View, CA, USA) and counterstained with Kernechtrot. For each pancreas, the images of islets in two sections separated by 200 μm were captured with a digital camera (Nikon). The SA beta-gal-positive area in the islets and the islet area were analysed using NIH Image 1.63 software, as previously described . The percentage of SA beta-gal-positive area was calculated by dividing the SA beta-gal-positive area in the islets by the islet area and multiplying by 100.
Detection of apoptosis
To detect apoptosis, three sections (3 μm) separated by 200 μm were stained by the terminal deoxynucleotidyl transferase-mediated dUTP nick-end labelling (TUNEL) method using an in situ apoptosis detection kit (Wako Pure Chemical Industries, Japan). DAB was used as substrate chromogen. Frequency of pancreatic islet cell apoptosis, which was expressed as events per islet, was calculated by averaging the number of TUNEL-positive beta cells in approximately 60 islets from each pancreas.
All data analyses were conducted using StatView 5.0 (SAS Institute, Cary, NC, USA). Data are shown as the mean ± SEM. Correlation analysis was performed between cellular senescence and cellular proliferation, AUCglucose, AUCinsulin, insulinogenic index or p38 mitogen-activated protein kinase (MAPK). The significance of correlation coefficients was analysed by Fisher’s z-transformation. Statistical analysis of the other data was performed with the Mann–Whitney U-test. We considered p values of less than 0.05 to be statistically significant.
Body weight, blood analyses, and glucose tolerance tests
Body weight, plasma triglycerides, NEFA, total cholesterol, glucose, and insulin levels of C57BL/6J mice before and after 4 and 12 months of a high-fat diet
Body weight (g)
Total cholesterol (mmol/l)
AUCglucose, AUCinsulin, and insulinogenic index during IPGTT in C57BL/6J mice after 4 and 12 months of a high-fat diet
AUCglucose(mmol l−12 h−1)
AUCinsulin(ng ml−12 h−1)
Insulinogenic index (ng/mmol)
Beta cell mass and proliferation
Morphometric analysis of pancreas of C57BL/6J mice after 4 and 12 months of a high-fat diet
High fat diet
% Beta cell/islet
% Islet area/pancreas area
Pancreas weight (g)
Islet mass (mg)
Beta cell mass (mg)
% Beta cell/islet
% Islet area/pancreas area
Pancreas weight (g)
Islet mass (mg)
Beta cell mass (mg)
Beta cell senescence and apoptosis
Correlation between cellular senescence and proliferation, AUCglucose, AUCinsulin, insulinogenic index and p38 MAPK
The markers of cellular senescence in beta cells were increased in mice fed a high-fat diet. Preceding senescence of the beta cells, increased proliferation was observed under hyperglycaemic conditions. There was a highly negative correlation between cellular proliferation and senescence. These findings indicate that cellular senescence occurs after rapid proliferation of beta cells and causes decreased cellular proliferation and beta cell mass.
Reduced pancreatic beta cell mass is a major factor in the pathogenesis of type 2 diabetes, as well as beta cell dysfunction. The long-term regulation of beta cell mass depends on the balance of beta cell supply from replication and neogenesis and beta cell death mainly from apoptosis [4, 5, 6, 7]. To adapt to hyperglycaemia during the pathogenesis of type 2 diabetes, augmented beta cell proliferation causes a compensatory increase in beta cell mass. The disruption of an adequate increase in beta cell mass leads to type 2 diabetes [6, 7, 8, 9, 10, 11, 12]. After a long-term, high-fat diet, the beta cell proliferation markedly decreased and islet cellular senescence was observed. We were unable to perform beta-galactosidase staining on the islets from 4-month-old mice. However, if senescence stops proliferation, then the necrosis and apoptosis that follow will reduce islet mass. Cell division and cell death are independent events as in the case of neurons, but beta cells are vulnerable to lipoglucotoxicity, and islet mass may be decreased by apoptosis. Reduced hypertrophy may also contribute to the reduction of islet mass. To judge from these findings, beta cell senescence causes decreased beta cell mass and may lead to type 2 diabetes.
Cellular senescence and decreased cellular proliferation occur in chronic diseases with high cellular turnover. For example, liver cirrhosis, in which chronic hepatocyte death and renewal is the major predisposing factor, is caused by hepatocyte senescence with concomitant telomere shortening that limits the number of cell divisions in humans . Recent in vitro studies  have found that senescence of human primary pancreatic beta cells is responsible for limiting the number of cell divisions. It seems likely that repeated replication during differentiation, such as high cellular turnover during the pathogenesis of liver cirrhosis, could cause islet cells to become senescent. Many studies indicate that the high cellular turnover of beta cells that is caused by beta cell replication and apoptosis is observed in the pathogenesis of type 2 diabetes. However, no previous reports discuss whether beta cell senescence occurs in type 2 diabetes. The current theory is that insufficient beta cell proliferation occurs at an early stage of type 2 diabetes pathogenesis and that the markedly increased beta cell apoptosis is the main factor of reduced beta cell mass in the pathogenesis of type 2 diabetes. Actually, we observed markedly increased cellular apoptosis after a long-term, high-fat diet. However, in our study (for 4 months) and another  longitudinal study in rodents, proliferation of beta cells was found even after a long-term prediabetic stage, suggesting that compensatory beta cell proliferation is maintained until the loss of the remaining beta cell divisions.
In addition to high cellular turnover, there may be another pathway of beta cell senescence in diabetes. Senescence of human primary pancreatic beta cells is caused by a telomere-independent pathway because senescence is not prevented by the expression of telomerase, which extends telomere length . Additionally, telomere length in mice is much longer than in humans, suggesting that telomeres are not related to cellular senescence in mice [35, 36]. Many studies demonstrate that cellular senescence is caused by ROS and is mediated by enhancing the activity of MAPK, which consists of at least three enzymes, including extracellular signal-regulated kinase (ERK), c-Jun N-terminal kinase (JNK), and p38 MAPK. For example, in rat kidney cells the ageing process parallels increases in ROS that strongly enhance the activities of ERK, JNK, and p38 MAPK . Isawa et al. reported that p38 MAPK is an important causative molecule in both telomere-dependent and telomere-independent cellular senescence, including ROS-induced senescence in both human and mouse fibroblasts . Oxidative stress occurs in the diabetic state [8, 38], and ROS are produced in islets through the non-enzymatic glycosylation reaction and the electron transport chain in mitochondria [12, 39]. It was recently reported that p38 MAPK in isolated rat islets is activated by oxidative stress . In our study, the active form of phospho-p38 was observed in the islets of diet-induced diabetic mice at the late stage of the high-fat diet, but the frequency of phospho-p38 was small compared with the percentage of SA beta-gal staining in the islets. Moreover, there was a significant positive correlation between the frequency of phospho-p38-positive beta cells and the percentage of SA beta-gal-positive area in islets of the high-fat diet group. The cyclin-dependent kinase inhibitors p16 and p21 are increased in p38-activated senescence . Furthermore, oxidative stress induces p21 (WAF1/CIP1/Sdi1) expression in isolated rat islet cells by treatment with H2O2 and pancreatic rat islet cells of hyperglycaemic Zucker diabetic fatty rats . Our results and the previous findings suggest that beta cell senescence is caused by oxidative stress and may be mediated by the p38 MAPK pathway, resulting in the elevation of p21.
In conclusion, decreased beta cell proliferation and cellular senescence are found in diet-induced type 2 diabetic mice. Cellular senescence may be caused by the compensatory rapid proliferation of beta cells needed to maintain adequate beta cell mass and/or the accumulation of ROS in beta cells. Beta cell senescence leads to insufficient insulin release as a consequence of decreased beta cell mass and contributes to the pathogenesis of diet-induced diabetes.
This study was supported by the High-Technology Research Centre Project of Kagawa Nutrition University from the Ministry of Education, Japan. We thank Dr Yasunori Kanazawa (Jichi Medical School Tochigi, Japan) for helpful discussions.