A total of 90 male C57BL/6J mice (age, 12 months, 26±2 g) were purchased from the Affiliated Animal Institute of Sichuan Academy of Medical Science & Sichuan Provincial People's Hospital (Chengdu, China). Mice were maintained in a 12-h light/dark cycle in an atmosphere of 0.03% CO₂ with free access to water and food. All mice were raised under specific pathogen-free conditions with a controlled temperature (23±2°C). Mice were randomly divided into three groups (n=30 in each group): i) A sham group, in which mice received a sham operation; ii) a PPx control group, in which mice were administered normal saline after PPx surgery; and iii) a mangiferin group, in which mice were intraperitoneally (i.p.) administered 90 mg/kg mangiferin for 28 consecutive days after PPx surgery (17,22). Mangiferin (purity, >98%) was purchased from Desite Co. (Chengdu, China). The animal procedures were approved by the Ethics Committee of Sichuan Academy of Medical Science & Sichuan Provincial People's Hospital (Chengdu, China).

All animals were fasted for 12 h prior to and 5 h after surgery, after which the animals were given free access to a standard diet and water. Mice were anesthetized by i.p. injection of 50 mg/kg pentobarbital (Beijing Propbs Biotechnology, Beijing, China), and the spleen and entire splenic portion of the pancreas were then surgically removed, while the mesenteric pancreas between the portal vein and duodenum was left intact. For the sham operation, the spleen was removed, while the pancreas was left intact. All mice were labeled with 0.8 mg/ml 5-bromo-2-deoxyuridine (BrdU; Sigma-Aldrich; Merck KGaA, Darmstadt, Germany), continuously provided in drinking water after the PPx surgery.

After the animals were fasted for 8 h, their blood glucose levels were measured with a SureStep Blood Glucose meter (LifeScan, Milpitas, CA, USA). An intravenous glucose tolerance test (IVGTT) was performed by tail vein injection of 1 g/kg D-glucose (Sigma-Aldrich; Merck KGaA) on days 14 and 28. Plasma insulin levels were determined by using an ultrasensitive mouse insulin ELISA kit (cat. no. 80-INSMS-E01; ALPCO, Salem, NH, USA) and plasma glucagon levels were determined by using a glucagon ELISA kit (cat. no. 81518; Crystal Chem Inc., Downers Grove, IL, USA).

After sacrificing the mice with 150 mg/kg pentobarbital, the remaining pancreas was perfused with collagenase P (Roche, Indianapolis, IN, USA), and the islets were then isolated as previously described (23). The islet tissues were lysed and total protein was quantified as previously described (24), followed by loading onto a NuPage Novex 10% Bis-Tris gel (Thermo Fisher Scientific, Inc., Waltham, MA, USA) for electrophoresis. After electrophoresis, the proteins were transferred onto polyvinylidene fluoride membranes (Pall Corp., Port Washington, NY, USA). The membranes were blocked and subsequently incubated with primary antibodies at 4°C overnight, followed by incubation with horseradish peroxidase (HRP)-conjugated secondary antibodies at room temperature for 2 h. Chemiluminescence detection was performed with Immobilon Western Chemiluminescent HRP Substrate (EMD Millipore, Billerica, MA, USA) and measured directly with a BioSpectrum Imaging System (UVP, Upland, CA, USA). The following primary antibodies were used: Rabbit polyclonal anti-caspase-3 (cat. no. sc-7148; 1:1,000 dilution), rabbit poly-clonal anti-B-cell lymphoma 2-associated X protein (Bax; cat. no. sc-493; 1:500 dilution), mouse monoclonal anti-BH3 domain interacting death agonist (Bid; cat. no. sc-135847; 1:1,000 dilution), rabbit polyclonal anti-poly(ADP ribose) polymerase (PARP; cat. no. sc-7150; 1:1,000 dilution), mouse monoclonal anti-cyclin D1 (cat. no. sc-450; 1:1,000 dilution), rabbit polyclonal anti-cyclin D2 (cat. no. sc-450; 1:1,000 dilution), rabbit polyclonal anti-cyclin D3 (cat. no. sc-182; 1:1,000 dilution), rabbit polyclonal anti-cdk4 (cat. no. sc-260; 1:1,000 dilution), mouse monoclonal anti-phospho-signal transducer and activator of transcription 3 (p-STAT3; cat. no. sc-8059; 1:500 dilution), mouse monoclonal anti-STAT3 (cat. no. sc-8019; 1:1,000 dilution), rabbit poly-clonal anti-retinoblastoma (Rb; cat. no. sc-7905; 1:1,000 dilution), rabbit polyclonal anti-insulin (cat. no. sc-7953; 1:1,000 dilution), rabbit polyclonal anti-glucokinase (GCK; cat. no. sc-7908; 1:1,000 dilution), rabbit polyclonal anti-insulin promoter factor 1 (PDX-1; cat. no. sc-25403; 1:1,000 dilution) and mouse monoclonal anti-β-actin (cat. no. sc-47778; 1:5,000 dilution) were purchased from Santa Cruz Biotechnology, Inc. (Dallas, TX, USA); rabbit monoclonal anti-p16^INK4a (cat. no. ab108349; 1:500 dilution), rabbit monoclonal anti-p27^Kip1 (cat. no. ab32034; 1:1,000 dilution), rabbit poly-clonal anti-phospho-Rb (anti-phospho-S780, cat. no. ab47763; 1:1,000 dilution) and rabbit polyclonal anti-glucose transporter 2 (GLUT-2; cat. no. ab54460; 1:1,000 dilution) antibodies were from Abcam (Cambridge, MA, USA). Rabbit polyclonal anti-cleaved caspase-3 (cat. no. 9661; 1:1,000 dilution) was purchased from Cell Signaling Technology, Inc. (Beverly, MA, USA). The HRP-conjugated goat anti-mouse polyclonal immunoglobulin (Ig)G (cat. no. 115-035-003) and HRP-conjugated goat anti-rabbit polyclonal IgG (cat. no. 111-035-003) were purchased from Jackson ImmunoResearch Laboratories (West Grove, PA, USA) and diluted at 1:5,000. β-actin served as the loading control and its levels were used for normalization.

The Cdk4 kinase assay was performed based on a previously described protocol (25). The amount of ³²P-labeled glutathione S-transferase-Rb (cat. no. SRP5124, Sigma-Aldrich; Merck KGaA) was evaluated by autoradiography and quantified by analysis with a PhosphorImager and an ImageQuant (Molecular Dynamics, Sunnyvale, CA, USA).

For analysis of neurogenin (Ngn)3 mRNA abundance, RNA was extracted from the remnant pancreas, and for analysis of other genes, RNA was extracted from the islets. Total RNA from the islets and the remnant pancreas was isolated using an RNeasy micro kit (Qiagen, Valencia, CA, USA) and reverse transcribed as previously described (26). The primer sequences and probes were as previously described (21,23). The RNA and primer were heated to 72°C and slowly cooled prior to reverse transcription at 42°C for 1 h. When cooled to room temperature, the reaction was diluted to 100 µl with RNase-free water. qPCR was performed with SYBR-Green PCR Master mix in a total reaction volume of 20 µl using the following amplification steps: Initial denaturation at 95°C for 10 min; followed by 40 cycles of denaturation at 95°C for 15 sec; and then elongation at 55°C for 30 sec. The expression levels were normalized to those of the internal standard GAPDH. Real-time qPCR was performed on an ABI 7900 system using an Applied Biosystems Power SYBR Green PCR Master Mix kit (Thermo Fisher Scientific, Inc.). The copy number was calculated using the 2^−ΔΔCq method. The Cq value for GAPDH was used to normalize the samples (27).

Pancreas tissues were fixed in 10% formalin (cat. no. HT-5011; Sigma-Aldrich, Merck KGaA) for 24 h at room temperature, dehydrated, embedded in paraffin (cat. no. A601888; Sangon Biological, Shanghai, China) and cut into 5 µm sections. For immunohistochemical staining, rat monoclonal BrdU antibody (1:100 dilution, cat. no. ab6326; Abcam), rabbit polyclonal Ki67 antibody (1:100 dilution, cat. no. PA1-21520; Thermo Fisher Scientific, Inc.), rabbit polyclonal proliferating cell nuclear antigen (PCNA) antibody (1:100 dilution, cat. no. ab18197; Abcam), and guinea pig polyclonal insulin antibody (1:100 dilution, cat. no. A0564; Dako, Glostrup, Denmark) were used according to a previously described protocol (25). The primary antibodies were detected with corresponding secondary antibodies (1:5,000 dilution; Jackson Immunoresearch Laboratories Inc.), including horseradish peroxidase (HRP)-conjugated goat anti-rat IgG polyclonal antibody (cat. no. 112-035-003), HRP-conjugated goat anti-rabbit polyclonal IgG (cat. no. 111-035-003), Alkaline Phosphatase (AP)-conjugated goat anti-guinea pig IgG polyclonal antibody (cat. no. 106-055-003), FITC-conjugated goat anti-guinea pig IgG polyclonal antibody (cat. no. 106-095-003), Alex Fluor 647-conjugated goat anti-guinea pig IgG polyclonal antibody (cat. no. 106-005-603), Cy5-conjugated goat anti-rabbit IgG polyclonal antibody (cat. no. 111-175-144). The secondary antibodies were incubated with the section for 2 h at room temperature. For terminal deoxynucleotidyl transferase deoxyuridinetriphosphate nick end labeling (TUNEL) staining, a Promega Fluorometric DeadEnd kit (Promega Corp., Madison, WI, USA) was used following the manufacturer's protocols.

Images of the islets were captured with a Nikon 80i microscope (Nikon Corp., Tokyo, Japan). All of the microscopy imaging and quantification procedures were performed by two technicians who were blinded to the experimental conditions of each sample. The numbers of β-cells, BrdU(+) β-cells, Ki67(+) β-cells, PCNA(+) β-cells and TUNEL(+) β-cells were manually counted and checked with Image-Pro Plus 6.3 software (Media Cybernetics, Silver Spring, MD, USA). At least 10 islets containing at least 1,000 β-cells were counted per mouse, and at least 10 consecutive sections from eight mice per group were stained. The 1st and 10th section were selected at a 50-µm distance. Islet diameters were determined with an intraocular calibrated grid. Islet size was also measured in images of insulin-stained islets converted to gray scale at a magnification of ×400.

Analysis of relative β-cell volume was performed via point counting morphometry using a 56-point grid. An average of 10,000 points/mouse were counted. The islet β-cell mass was calculated by multiplying the relative β-cell volume by the total weight of the remnant pancreas.

Male adult (age, 3 months, 22–24 g, n=20) and aged (age, 12 months, 26–28 g, n=90) mice were anesthetized and sacrificed by i.p. injection of 150 mg/kg pentobarbital. The pancreas was subsequently perfused with 1 ml collagenase P (1.0 mg/ml) and digested at 37°C for 15 min. Isolated cells were then centrifuged at 250 × g for 2 min at 4°C. The supernatant was discarded and the cells were re-suspended in 10 ml Hank's balanced salt solution. The intact re-suspended islets were handpicked using a pipette under a dissection scope, and primary β-cell cultures were prepared from the handpicked islets. Subsequently, 400-500 islets were incubated in 1 ml 1X accutase solution (cat. no. A6964; Sigma-Aldrich; Merck KGaA) for 15 min at 37°C. The digestion was stopped by adding cell culture medium, and the cells were collected by centrifugation at 300 × g for 5 min at 4°C. Subsequently, the islet cells were cultured in RPMI-1640 medium (Gibco, Thermo Fisher Scientific, Inc.) supplemented with 10% fetal bovine serum (Gibco, Thermo Fisher Scientific, Inc.), 50 µM β-mercaptoethanol (Sigma-Aldrich; Merck KGaA), 1 mM sodium pyruvate (Sigma-Aldrich; Merck KGaA), 2 mM L-glutamine (Sigma-Aldrich; Merck KGaA) and 10 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (Sigma-Aldrich; Merck KGaA) at 37°C in a humidified atmosphere containing 5% CO₂.

The PCR product of the p16^INK4a gene was cloned into pcDNA3 plasmids (Addgene, Cambridge, MA, USA) under the control of a cytomegalovirus promoter, and the plasmids were transfected into the primary cultured islet cells in the presence of Lipofectamine LTX (Thermo Fisher Scientific, Inc.). Small interfering RNA (siRNA) specific for p16^INK4a pool (cat. no. 12578) and non-targeting siRNA controls were obtained from GE Dharmacon (Lafayette, CO, USA). Transfection with the siRNAs was performed using Dharmafect (GE Dharmacon).

Islet cells at the logarithmic growth phase were seeded in a 24-well plate (2×10⁵ cells per well) and incubated at 37°C for 24 h. Mangiferin was then added and the cells were incubated for another 24 h. Subsequently, 10 µl of a 5 mg/ml solution of MTT was added to each well, followed by incubation at 37°C for 4 h. Subsequently, the medium was removed and the plates were thoroughly agitated for 1 h. Finally, termination buffer was added to each well, and the absorbance at 570 nm was measured with a spectrophotometer (Model 3550 Microplate Reader; Bio-Rad Laboratories, Inc., Hercules, CA, USA). The proliferation rates were calculated from the optical density (OD) according to the following formula: Cell proliferation (%) = [OD 570 nm (drug)/OD 570 nm (control)] × 100%.

Values are expressed as the mean ± standard error of the mean from at least three independent experiments. Statistical significance between multiple groups was determined by analysis of variance followed by a Bonferroni post hoc test, and between 2 groups using Student's t-test using GraphPad Prism (version 5.0 for windows) statistical software package (GraphPad Software, Inc., La Jolla, CA, USA). P<0.05 was considered to indicate a statistically significant difference between groups.

According to a previous study by our group, treatment with mangiferin (90 mg/kg) maintained glucose homeostasis in adult mice by islet regeneration induced by mangiferin (21). Therefore, the present study investigated the regulatory role of mangiferin in aged mice. As presented in Fig. 1A, compared with the mice in the sham group, the PPx control mice exhibited significantly increased blood glucose levels. In turn, mangiferin treatment markedly reduced the blood glucose levels in PPx mice. Although the mice that received mangiferin treatment did not maintain normal blood glucose to the same extent as the sham-operated mice, a time-dependent decrease in blood glucose levels was observed from day 3 to day 10. Subsequently, the body weight of the mice in each group was assessed, revealing that the body weight of the mangiferin-treated mice were higher when compared with that of the PPx control mice (Fig. 1B). Further analysis by IVGTT clearly substantiated the hypothesis that islet regeneration was induced by mangiferin. According to the IVGTT data on days 14 (Fig. 1C) and 28 (Fig. 1D), mangiferin partially recovered the impaired glucose tolerance of the PPx mice on day 14 and maintained the partially recovered glucose tolerance on day 28. Furthermore, the serum insulin levels (Fig. 1E) and glucagon levels (Fig. 1F) were measured, and it was observed that mangiferin-treated aged mice had elevated levels of insulin secretion and had no effect on glucagon secretion compared with those in the PPx group. These experiments illustrated that mangiferin treatment maintained homeostasis in aged mice after PPx.

To further explore the mechanisms via which mangiferin promotes homeostasis, the remaining pancreas tissues were stained with specific antibodies for proliferating cells, and the proliferation rates were subsequently quantified. Representative immunohistochemical staining images of the islet β-cells and proliferating cells are provided in Fig. 2A. It was evident that the islets of mangiferin-treated mice had more proliferating cells (indicated by arrows) on day 14 than those of the PPx mice. Of note, with prolonged treatment with mangiferin for 28 days, an increased quantity of BrdU-labeled proliferating cells was present, which provided evidence that mangiferin induces islet regeneration in a time-dependent manner. To further validate this hypothesis, cell proliferation was assessed with two endogenous markers, PCNA and Ki67. According to the levels of proliferating cells indicated by BrdU-labeling (Fig. 2B), Ki67 labeling (Fig. 2C) and PCNA labeling (Fig. 2D), mangiferin induced robust cell proliferation in the islets of aged mice. Although it has been reported that pancreatic duct epithelial cells may be considered as progenitor cells that contribute to neogenesis, no proliferation of the duct cells was observed (data not shown). Therefore, it appears that mangiferin induced islet regeneration in aged mice mainly via inducing the proliferation of β-cells. However, whether those proliferated cells resulted in islet hyperplasia remains elusive. To investigate this, the remaining pancreas was subsequently weighed (Fig. 2E) to calculate the relative β-cell volume (Fig. 2F) and β-cell mass (Fig. 2G). It was revealed that the relative β-cell volume and β-cell mass of mangiferin-treated mice were significantly increased relative to those in PPx control mice, which is consistent with the results on β-cell proliferation.

Previous clinical studies on T2DM patients clearly demonstrated that their β-cell mass was decreased due to increased β-cell apoptosis (3,28). Therefore, therapeutic approaches comprising the inhibition of β-cell apoptosis may not only palliate hyperglycemia, but reverse and prevent the progression of the disease, and thus, the present study further focused on the effect of mangiferin on β-cell apoptosis. A previous study by our group illustrated that mangiferin inhibited β-cell apoptosis in young mice (21). In the present study, the remnant pancreas was stained using an in situ TUNEL apoptosis detection kit. Of note, the islets of PPx mice lacking mangiferin treatment exhibited more apoptotic cells (green cells indicated by arrows; Fig. 3A) than those of mice treated with mangiferin. Consistent with these qualitative observations, calculation of the percentages of apoptotic cells (TUNEL-positive islet β-cell percentage) revealed that mangiferin treatment markedly inhibited apoptosis in the aged mice (Fig. 3B). To further validate that mangiferin inhibited cell apoptosis in the islets of the aged mice, key proteins in the caspase pathway were analyzed (Fig. 3C). Activation of caspases has a central role in apoptosis, and caspase-3 serves as the convergence point of different apoptotic signaling pathways (29). As presented in Fig. 3C, caspase-3 was de-activated upon mangiferin treatment with increased pro-caspase-3. Bid induces a conformational change of Bax (30), and Bax induces the release of cytochrome c from the mitochondria during apoptosis (31). The present results suggested that mangiferin-treated cells had lower expression levels of Bax and Bid compared with those in the control, suggesting that more apoptotic events were inhibited by mangiferin. Furthermore, decreased cleavage of PARP was identified in the mangiferin-treated group. Therefore, changes in the apoptotic tendency of β-cells, along with enhanced proliferation capability, likely contributed to the increased levels of insulin secretion in vivo.

A study by Krishnamurthy et al (9) indicated that p16^INK4a restricted proliferation and regeneration in the islets of aged mice, while mice lacking p16^INK4a exhibited enhanced islet proliferation rates. To assess whether mangiferin treatment prevents β-cell senescence, in situ fluorescence staining of p16^INK4a was performed. Representative images of p16^INK4a immunofluorescent staining are presented in Fig. 4A. Quantification of the staining clearly indicated that >50% of β-cells were labeled with p16^INK4a after PPx and also in the sham operated group, indicating that PPx did not induce marked senescence compared with the sham group. In contrast to the PPx controls, the islets of mangiferin-treated mice exhibited positive staining for active p16^INK4a at a rate of ~35%, indicating that mangiferin markedly inhibits β-cell senescence (Fig. 4B). It was therefore concluded that mangiferin treatment inhibits β-cell senescence in aged mice.

A plethora of literature suggests that cyclin D/Cdk4 complexes have a critical role in the cell cycle regeneration of β-cells (32–34), and a previous study by our group indicated that mangiferin induced β-cell regeneration via the regulation of cell cycle complexes (21). Therefore, to investigate whether cyclin D-Cdk4 complexes may be regulated by mangiferin in aged mice, the expression of cyclin D1, -D2 and -D3, as well as Cdk4, p16^INK4a and p27^Kip1 was assessed at the protein (Fig. 5A) and mRNA (Fig. 5B) level. As expected, mangiferin greatly increased the transcription and translation of cyclin D1 and -D2, as well as Cdk4. Furthermore, the expression levels of p16^INK4a and p27^Kip1 were significantly reduced by mangiferin in the aged mice. Quantification of the western blot results indicated that the expression and phosphorylation levels of STAT3 were markedly elevated in the mangiferin-treated mice relative to those in the untreated PPx controls, and the total STAT3 protein levels of PPx control was also significantly higher than those of the sham group. Notably, PPx could induce the increased expression of STAT3, and mangiferin treatment could induce the phosphorylation and activation of STAT3 (Fig. 5C and D).

To confirm the effects of mangiferin on β-cell proliferation via the regulation of p16^INK4a, islet cells of adult and aged mice were isolated and cultured in the presence of mangiferin or vehicle (dimethyl sulfoxide) (Fig. 6A); furthermore, the p16^INK4a gene was knocked down in the isolated islet cells of the aged mice (Fig. 6B) and p16^INK4a was overexpressed in the isolated islet cells of the adult mice (Fig. 6C). Of note, mangiferin induced the proliferation of the islet cells from the adult and the aged mice (Fig. 6D). Furthermore, p16^INK4a silencing in islet cells from aged mice led to elevated proliferation rates, regardless of whether the islets were treated with or without mangiferin (Fig. 6E). In addition, mangiferin treatment of the isolated islet cells of the adult mice with over-expression of p16^INK4a significantly increased the proliferation rate (Fig. 6F).

Furthermore, to address whether the activity of Cdk4 may be regulated by mangiferin, a Cdk4 kinase assay was performed. Previous studies demonstrated that Rb is an important substrate of the cyclin D1/Cdk4 complex, and that its activation is closely associated with senescence (35). In the Cdk4 assay, increased radioactive labeling of the substrate is considered to indicate greater enzymatic activity of Cdk4. As presented in Fig. 7A and B, the in vitro kinase assay indicated lower levels of labeled phosphorylated Rb in the mangiferin-treated group when compared with that in the control group. In addition, in vivo analysis of Rb phosphorylation on serine 780 revealed that mangiferin treatment greatly reduced the phosphorylation of Rb, suggesting that the enhanced Cdk4 activity resulted in decreased Rb activity (Fig. 7C). Collectively, these results indicate that mangiferin may induce β-cell proliferation through the regulation of cell cycle proteins.

To elucidate the mechanism underlying mangiferin-induced insulin secretion in aged mice, it was assessed whether mangiferin regulates β-cell-specific genes, including insulin, GCK, GLUT-2 and PDX-1, as critical genes for β-cell function. As expected, a significant increase in the protein expression levels of insulin, GLUT-2, GCK and PDX-1 was detected in the islets of the mangiferin-treated mice (Fig. 8A). Accordingly, the mRNA levels of these genes were also elevated (Fig. 8B). Notably, mangiferin treatment led to upregulation of the expression of PDX-1, the upstream gene of insulin. However, no changes in the transcription or translation levels of Ngn3 were observed in mangiferin-treated mice when compared with those in the controls (Fig. 8), suggesting that no neogenesis occurred in the aged mice. Therefore, it was concluded that mangiferin treatment markedly contributes to insulin secretion through the regulation of insulin-associated genes.