A total of 40 NOD/Lt mice were purchased from the Institute of Experimental Animals of the Chinese Academy of Medical Sciences. There were 20 male and 20 female mice, with a weight of 17–25 g. Exendin-4 reagent (Kangtai Biotechnology, Beijing, China); hematoxylin and eosin (H&E) dye liquor (Beijing Solarbio Science and Technology, Co., Ltd., Beijing, China); IL-2, IL-10 and IFN-γ reagent kits (Shanghai Biotechnology, Shanghai, China); monoclonal CD4 rabbit antibody (Abcam, Cambridge, MA, USA; catalog no.: ab133616; dilution, 1:100) and monoclonal CD8 rabbit antibody (Abcam; catalog no.: ab22378; dilution, 1:100); a flow cytometer (Beijing Kexue Yingye Science and Technology Development, Beijing, China); microplate reader (Jinan Guangyao Medical Equipment, Jinan, China); centrifugal machine (Shanghai Puyuan Instrument, Shanghai, China); and an optical microscope (Shenzhen Final Technology, Shenzhen, China) were used in the study.

The present study was approved by the ethics committee of Fujian Medical University (Fujian, China).

A random number table was used to divide 40 clean and healthy NOD mice into 4 groups (n=10/group) as follows: One blank control group D with normal saline intervention, and the remaining three groups with exendin-4 intervention. According to the different exendin-4 doses of 2, 4 and 8 µg/kg/day, they were called the low-dose group A, medium-dose group B and high-dose group C, respectively. The four groups of mice were intervened for 8 weeks, after which they were sacrificed by cervical dislocation. An inner canthus method was used to take ~1 ml of whole blood, which was then centrifuged at 5,000 × g for 10 min at 4°C. Serum was kept at −20°C to test serum IL-2, IL-10 and IFN-γ. Pancreatic tissues were selected to conduct T-cell subset determination, while other subsets continued to be observed.

Pancreatic samples were embedded with paraffin, and then turned into tissue sections with H&E dyeing to be observed under a light microscope (Olympus, Tokyo, Japan). Pathology personnel used a double-blind method to determine the infiltration degree of pancreatitis. The grade of pancreatitis infiltration degree was divided as follows: i) Grade 0: Pancreas islet was complete without lymphocyte infiltration; ii) grade 1: lymphocytes infiltrated around the pancreas islet or <25% pancreas islet area was affected; iii) grade 2: 25–50% pancreas islet area was affected; and iv) grade 3: >50% pancreas islet area was affected. According to the above grading standard, the pancreatic sections of the four groups were observed under a ×5-x400 microscope.

Immunohistochemistry was used to observe the expression of the IL-10 gene in local pancreatic samples. Brownish granules in cytoplasm were considered positive cells. Positive cell percentage and dyeing intensity were calculated. The positive cell percentage was calculated as: i) <1% was 0; ii) 1–10% was 1; iii) 11–50% was 2; iv) 51–80% was 3; and v) >80% was 4. The dyeing intensity of granules was calculated as: i) No dyeing was 0; ii) faint yellow was 1; iii) yellow was 2; and iv) brown was 3. The product of positive cell percentage and dyeing intensity determined the immunohistochemistry score as follows: 0 score (−); 1–4 scores (+); 5–8 scores (++); and 9–12 scores (+++).

After the mice were sacrificed, they were sterilized with alcohol. Pancreatic tissues were separated through an aseptic technique. Nearby connective tissues and pancreatic tissues were removed and rinsed twice in Hanks' liquid. Ophthalmic scissors were used to cut the tissues into pasty sections. A tip sucker was used to blow and disperse pancreatic cells, while a 300-mesh net was used to implement infiltration. A lymphocyte-separating medium was used to extract lymphocytes at rotating speeds of 400 × g. They were then centrifuged for 20 min. The suspended layer of lymphocytes was extracted after a 2 ml pH 7.4 phosphate buffer saline (PBS) working solution was added. The lymphocytes were centrifuged for 5 min at 800 × g. The supernatant was discarded, and the cell concentration was adjusted to 2×10⁹ cells/l. Under a microscope, 0.4% trypan blue was used to count living cells. The cell survival rate was determined to be >95%. A 100 µl pancreatic single-cell suspension at a concentration of 2×10⁹/ml was measured at 25°C and kept in the dark. The suspension was then incubated with 10 µl CD8-FITC and 10 µl CD4-PE fluorescent antibody for 15 min. A 2-m erythrocyte lysate buffer was added to the suspension to incubate it for 10 min. A 2-ml PBS solution was then added to conduct centrifugation at 800 × g for 5 min. PBS washing was then repeated twice (the method used before), the supernatant was removed, and 200 µl PBS solution was added to each tube to conduct FCM detection. Macintosh (Apple, Inc., Cupertino, CA, USA) was the data processing system used. Cells were analyzed using CellQuest software (BD Biosciences, San Jose, CA, USA). Grating was conducted on lymphocytes on forward and lateral angular scattering diagrams. A logarithmic method was used to acquire fluorescence signals before 1×10⁴ cells were taken and analyzed under the same software. Serum of IL-2, IL-10 and IFN-γ were determined using ELISA. Operating steps were performed with ELISA reagent kits of IL-2, IL-10 and IFN-γ. The samples were determined in the same batch.

SPSS 21.0 statistical software (IBM SPSS, Armonk, NY, USA) was used to conduct statistical analysis. Experimental data were presented as mean ± standard deviation, t-test or t'-test. A χ² test was conducted on experimental results. P<0.05 was considered to indicate a statistically significant difference.

Pancreatitis of mice in control group D was mainly at grade 2 and 3. Under a light microscope, it was observed that pancreatic cell morphology was in disorder, and the size and quantity of the pancreas was small. Mouse pancreatitis in exendin-4 low-dose group A, medium-dose group B and high-dose group C was mainly at grade 0 and 1. Under the light microscope, it was observed that pancreatic cell morphology improved, infiltration degree of lymphocyte improved and pancreatic islet size was restored somewhat.

Under the light microscope, numerous brown granules with deep color within pancreatic sample cells in exendin-4 low-dose group A, medium-dose group B and high-dose group C (Figs. 1–3) were observed. There were a few brown granules within pancreatic sample cells in control group D (Fig. 4). IL-10 immunohistochemistry scores in low-dose group A, medium-dose group B and high-dose group C were 3.82±0.72, 4.34±0.86 and 4.81±0.94, respectively, all of which were higher than the immunohistochemistry score of 2.25±0.63 in group D.

The CD4+CD25+T-cell proportions in pancreatic tissues in exendin-4 low-dose group A, medium-dose group B and high-dose group C were 5.31, 5.53 and 5.74%, respectively, all of which were higher than the CD4+CD25+T-cell proportions of 1.62% in group D, as shown in Figs. 5–8 (right upper quadrant of each figure was a CD4+CD25+T cell subset). The proportions of CD4+CD25^high T-cells in CD4+T cells increased in groups A, B and C.

Compared with control group D, serum IL-10 levels in exendin-4 low-dose group A, medium-dose group B and high-dose group C increased (P<0.05), while IL-2 and IFN-γ levels decreased (P<0.05). The difference of serum IL-10, IL-2 and IFN-γ levels existing in IFN-γ was of statistical significance (P<0.05) (Table I).