All animal procedures were conducted according to the Shanghai Laboratory Animal Ordinance and approved by the Ethics Committee of Shanghai Tenth People’s Hospital (Tongji University, Shanghai, China). Sprague-Dawley (SD) rats were purchased from Shanghai SLAC Laboratory Animal Co., Ltd. (Shanghai, China). These animals were housed under standardized conditions in a 12 h dark/light cycle with free access to water and food. Rats were fasted overnight preoperatively. Following being anesthetized by intraperitoneal injection of 3% pentobarbital, a 2 cm midline laparotomy was performed and the first loop of the duodenum and the pancreas were separated. A blunt fine catheter was then introduced into the bile-pancreatic duct and 3% sodium taurocholate (1 ml/kg; Sigma-Aldrich, St. Louis, MO, USA) was retrogradely injected into the common bile duct within 60 sec. Leakage of sodium taurocholate was prevented by temporary ligation of the distal bile duct with a 9-0 prolene suture, while the proximal bile duct was temporarily occluded with a microvessel clip. Following infusion, the blunt fine catheter, the microvessel clip and the suture were removed to allow the physiological flow of bile. The puncture site on the duodenum was closed with a 6-0 prolene suture and the wound with a 4-0 prolene suture.

Four week-old SD rats were sacrificed and the primary BMSCs were isolated from the femurs and tibias under a sterile condition. Following flushing the bone marrow cavity with DMEM-LG medium (Invitrogen Life Technologies, Grand Island, NY, USA) and removing large tissues with a 200-mesh nylon filter, the isolated bone marrow cells were cultured in DMEM-LG medium supplemented with 10% fetal bovine serum (FBS; Invitrogen Life Technologies), 1% penicillin and streptomycin (Invitrogen Life Technologies) at 37°C in an environment with 5% CO₂. The medium was refreshed every 3 days and cells were harvested by multiple digestions and passaged when cell confluence reached >80%. Cells of the fifth passage were collected and subjected to flow cytometry analysis. These cells were maintained in adipogenic, osteogenic and chondrogenic differentiation medium (Cyagen Biosciences, Sunnyvale, CA, USA) followed by Oil Red O staining, Alizarin Red S staining and Toluidine Blue staining, respectively. The surface markers were detected using a BD FACSC™ II flow cytometry system (BD Biosciences, San Jose, CA, USA). The antibodies utilized in this analysis included anti-CD44-PE, anti-CD73-APC, anti-CD90-FITC, anti-CD105-PerCP-Cy5.5, anti-CD11b-PE, anti-CD19-PE, anti-CD34-PE and anti-CD45-PE (BD Biosciences). Isotype-matched antibodies were used as controls. Cells were harvested between 3–5 passages.

Following AP induction, pancreatic tissues were obtained on days 1, 3, 5, 7 and 10, and normal pancreatic tissues were collected as controls. Immunocytochemistry and western blot analysis were employed to detect the SDF-1 expression. Five SD rats were used at each time point.

After preparation of the paraffin sections, Histostain-Plus kit (DAB, Rabbit; Invitrogen Life Technologies) was used to perform immunohistochemistry. Following conventional dehydration, antigen retrieval and antigen blocking, sections were incubated with rabbit anti-SDF-1 antibody (1:50; Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA) overnight at 4°C in a wet box. Then, the biotinylated secondary antibody and streptavidin conjugated horseradish peroxidase (HRP) were independently incubated with these sections for 10 min at 37°C. Following staining with DAB, the sections were counterstained with hematoxylin and then dehydrated and mounted with neutral resin.

Pancreatic tissues were added to liquid nitrogen and crushed. The total protein was extracted with the conventional method and was followed by determination of protein concentration with bicinchoninic acid method. Subsequently, 30 mg of proteins were subjected to 12% polyacrylamide-SDS gel electrophoresis and electroblotted onto nitrocellulose membranes, which were then incubated with rabbit anti-SDF-1 antibody (dilution, 1:200; Santa Cruz Biotechnology, Inc.) overnight at 4°C following blocking with 5% skimmed milk for 1 h. Incubation with the secondary antibody at room temperature for 2 h was conducted and then visualization was performed, and the protein bands were scanned (Gene Company Limited, Hong Kong, China) and analyzed using Odyssey 3.0 analysis software (Li-Cor Bioscience, Lincoln, NE, USA).

The CXCR4 expression profile on BMSCs was assessed by immunofluorescence staining. Cells were collected and fixed in 4% paraformaldehyde for 20 min. Following treatment with 0.5% Triton X-100 and 1% bovine serum albumin (BSA), cells were incubated with the rabbit anti-CXCR4 antibody (dilution, 1:50; Santa Cruz Biotechnology, Inc.) overnight at 4°C. After washing with PBS, cells were incubated with the fluorescein isothiocyanate (FITC)-conjugated secondary antibody for 30 min at 37°C followed by DAPI (Sigma-Aldrich) staining for 2 min, for nuclear counterstaining. The sections were mounted and then observed under a Zeiss LSM 710 confocal microscope (Carl Zeiss, Oberkochen, Germany).

The migration assay was designed using transwell plates (Corning Costar, Cambridge, MA, USA) that were 6.5 mm in diameter with 8 μm pore filters. In the chemotaxis assay, the upper chambers were loaded with 5×10⁴ BMSCs in 200 μl of DMEM containing 0.1% BSA, and the lower chambers with 500 μl of DMEM containing 10% FBS and SDF-1 at different concentrations (Peprotech, Inc., Rocky Hill, NJ, USA). SDF-1 at 0, 10, 50, and 100 ng/ml was used according to the manufacturer’s instructions. In the chemotaxis inhibition assay, the upper chambers were inoculated with 5×10⁴ BMSCs that were incubated with rabbit anti-CXCR4 antibody (10 μg/ml; Santa Cruz Biotechnology, Inc.) or AMD3100 (10 μg/ml; Sigma-Aldrich), an antagonist of CXCR4, for 2 h at 37°C. The lower chambers were added with 500 μl of DMEM containing 10% FBS and SDF-1 at optimal concentration. Following incubation for 15 h, cells in the upper chamber were removed and the membranes were fixed in 4% paraformaldehyde for 20 min. The cells that migrated to the lower side of the filter were stained with 0.1% crystal violet for 10 min and then observed under a light microscope. Crystal violet was dissolved in 300 μl of 33% acetic acid and the absorbance at 573 nm was measured with an enzyme-labeling measuring instrument (Gene Company Limited).

To trace BMSCs in vivo, cells were labeled with chloromethylbenzamido-1,1′-dioctadecyl-3,3,3′,3′-tetramethylindocarbocyanine perchlorate (CM-Dil; 1 μg/ml; Invitrogen Life Technologies) at 37°C for 30 min. Several studies have demonstrated that labeling cells with CM-Dil does not affect cell viability, proliferation or differentiation. The labeling efficiency was determined by fluorescence microscopy (16,17).

Rats were randomly divided into four groups: the control group (n=15), the AP group (n=15), the BMSCs group (n=15) and the anti-CXCR4 group (n=15). Rats were sacrificed at days 1, 4 and 7 (n = 5/time point). Rats in the anti-CXCR4 group received an injection of CM-Dil-labeled BMSCs (1×10⁷ cells/ml/kg) and were pretreated with anti-CXCR4 antibody via the tail vein. Rats in the BMSC group received an injection of CM-Dil-labeled BMSCs (1×10⁷ cells/ml/kg). In the control and AP group, rats were treated with normal saline of equal volume. Following sacrificing, the pancreas and blood were collected. Each pancreatic tissue was divided into two sections: one was fixed in 4% paraformaldehyde for immunohistochemistry and HE staining, and the other was prepared for the frozen sections, in order to detect the migration of BMSCs towards the pancreatic tissues by confocal microscopy (Zeiss). The images were captured at five randomly selected fields at a high magnification and the migrated CM-Dil-labeled BMSCs (red cells) were calculated. Blood was centrifuged and supernatant collected for serum amylase analysis.

Paraffin sections were prepared for histopathological examination with conventional HE staining. In order to quantify the pancreatic injury, 20 microscopic fields were randomly selected as previously described (18). In brief, the edema, inflammatory cell infiltration and acinar necrosis were divided into four grades (edema: 0 = absent, 1 = focally lobular edema, 2 = diffusely lobular edema and 3 = acini edema and separated; inflammatory cell infiltration: 0 = absent, 1 = infiltration around ductal margins, 2 = infiltration into <50% of lobules and 3 = infiltration of >50% of lobules; acinar necrosis: 0 = absent, 1 = necrosis in <5% of lobules, 2 = necrosis in 5–20% of lobules and 3 = necrosis in 20–50% of lobules).

The amylase activity assay kit (Biovision, Mountain View, CA, USA) was used to determinate the serum amylase activity. The nitrophenol standard curve was delineated. Then, the optical density (OD) immediately prior to reaction (ODT₀) and at 10 min following reaction (ODT₁) was measured at 405 nm. The difference of OD was calculated as follows: ΔOD = ODT₁ − ODT₀, and the concentration of nitrophenol was obtained according to the standard curve and then converted into the amylase activity.

Statistical analysis was performed using SPSS version 14.0 for Windows (SPSS, Inc., Chicago, IL, USA). Data were presented as the mean ± standard deviation (SD) and were analyzed with a Student’s t-test. A value of P<0.05 was considered to indicate a statistically significant result.

A large number of nucleated cells were attached to the bottom of the plates under a light microscope. These cells demonstrated fibroblast-like morphology on days 5–6 and cell colonies were found on days 7–10 (Fig. 1A). These cells gradually aged after passaging 6–7 times, as identified by an increased cell size, intracellular fine particulate matter and a decrease in cell proliferation. These cells demonstrated excellent multi-lineage plasticity and successfully differentiated into adipocytes, osteoblasts and chondrocytes following in vitro adipogenic, osteogenic and chondrogenic differentiation induction for 21, 14 and 28 days, respectively (Fig. 1B). These cells were identified by BMSC markers. Flow cytometry identified that these cells were markedly positive for CD44, CD73, CD90 and CD105 (99.93, 99.97, 99.98, 99.78%, respectively) but negative for CD11b, CD19, CD34, and CD45 (0.65, 0.85, 0.70 and 1.20%, respectively; Fig. 1C).

To investigate the effect of the SDF-1/CXCR4 axis on the migration of BMSCs, the SDF-1 expression in damaged pancreas was measured and observed at different time points following AP induction. As summarized in Fig. 2D, in the control group, pancreatic tissues exhibited a low SDF-1 expression on the cell membrane and in the cytoplasm. The SDF-1 expression was gradually increased following AP induction, peaking on days 5–7, however declining from day 10. The SDF-1 protein expression as determined by western blot analysis was markedly increased in the inflammatory area as compared with the control group, which was consistent with findings in the immunohistochemistry assay. The SDF-1/β-actin ratio on day 7 was the highest, being 8.6-fold greater than that in the control group. The SDF-1/β-actin ratio in AP group on days 3, 5, 7, 10 was significantly different from that in control group (P<0.001; Fig. 2A and B).

BMSCs were treated with anti-CXCR4 antibody for 2 h and then the CXCR4 expression on BMSCs was detected using confocal microscopy. Five fields were randomly selected for analysis, and the results demonstrated that >90% of BMSCs expressed CXCR4, with certain cells having a particularly high expression (Fig. 2C).

Our results revealed that SDF-1 induced BMSC migration in a dose-dependent manner, and that maximum migration was observed following treatment with SDF-1 at 100 ng/ml, which indicates that SDF-1 is important in BMSC migration (Fig. 3A and B). The absorption at 405 nm following treatment with SDF-1 at 100 ng/ml was significantly increased as compared with the other groups (P<0.001; Fig. 3C). In addition, SDF-1-induced migration was markedly inhibited when BMSCs were pretreated with anti-CXCR4 antibody or AMD3100 (Fig. 3D and E). The absorption at 405 nm in the anti-CXCR4 group and AMD3100 group was significantly decreased as compared with the SDF-1 group (P<0.001). However, there was no difference between the anti-CXCR4 group and AMD3100 group (P>0.05; Fig. 3F), which indicate that the inhibitory effect of anti-CXCR4 antibody is similar to that of AMD3100.

In order to trace the cells in vivo, CM-Dil was utilized to label migrating BMSCs. Results demonstrated >95% of BMSCs were successfully labeled by CM-Dil (Fig. 4A). To investigate the effect of CM-Dil-labeled BMSCs in damaged pancreas, frozen sections were prepared as previously described. The results demonstrated CM-Dil positive BMSCs in the BMSCs and anti-CXCR4 groups, however not in the AP and control groups. Concurrently, at each time point, the number of CM-Dil-labeled BMSCs in the anti-CXCR4 group was significantly lowered as compared with the BMSCs group (Fig. 4B). Cells were counted in five randomly selected fields at a high magnification and the total number of cells per section was determined. The results revealed that blocking CXCR4 expression on BMSCs significantly inhibited the migration of BMSCs towards the damaged pancreas (Fig. 4C).

Serum amylase analysis revealed that the amylase activity in the AP group was significantly higher compared with the control group on day 1 (P<0.001). Amylase activity appeared to decrease after BMSCs were transplanted and was significantly decreased as compared with the AP group (P<0.001). Concurrently, serum amylase activity in the anti-CXCR4 group was markedly higher than that in the BMSC group (P<0.001; Fig. 5B). The results of HE staining were consistent with the change in serum amylase activity, and revealed that the BMSCs could reduce pancreatic edema, hemorrhage and necrosis. This effect was compromised following pre-treatment with the anti-CXCR4 antibody. Furthermore, a large number of tubular complexes were observed during pancreatic repair in the BMSC group (Fig. 5A).