The Seoul National University (SNU) designated pathogen free (DPF) pigs aged at least 6 months were used as xenogenic islet donors [22]. Male alpha 1, 3-galactosyltransferase knock-out mice (GalT KO) aged 7 to 12 weeks (kindly provide from Prof. Chung-Gyu Park, SNU) were used as recipients. Type 1 diabetes was confirmed with a portable glucometer (GLUCOCARD X-METER, Arkray, Kyoto, Japan); if random blood glucose levels from the tail vein exceeded 300 mg/dL for three consecutive days after intraperitoneal one-shot injection of 200 mg/kg dose of streptozotocin (STZ, Sigma-Aldrich, St. Louis, MO, USA; freshly dissolved in 0.01 M citric acid buffer, pH 4.5), the diagnosis of type 1 diabetes was made. The GalT KO mice were maintained at a constant temperature in a cycle of 12 hours of light followed by 12 hours of dark and provided with laboratory chow and water. All of surgical interventions and presurgical and postsurgical animal care were provided in accordance with the Laboratory Animals Welfare Act, the Guide for the Care and Use of Laboratory Animals and the Guidelines and Policies for Rodent Survival Surgery provided by the IACUC (Institutional Animal Care and Use Committee) in College of Medicine, The Catholic University of Korea (approval number: CUMS-2014-0105-02).

Pronova ultrapure low viscosity guluronic acid (UP-LVG) sodium alginate (20 to 200 kDa) was purchased from Pronova (Novamatrix, Sandvika, Norway). Povidone was purchased from Sungkwangpham (Cheonan, Korea). Acridine orange (AO), magnesium sulfate (MgSO₄), magnesium chloride (MgCl₂), sodium phosphate monobasic (NaH₂PO₄), sodium chloride (NaCl), chitosan (from crab shells, minimum 85% deacetylated), barium chloride (BaCl₂), nicotinamide, 4-(2-hydroxyethyl)piperazine-1-ethanesulfonic acid (HEPES), propidium iodide (PI), STZ, luria/miller agarose (LB agarose), paraformaldehyde (PFA), fluorescein isothiocyanate (FITC)-labeled dextran, sodium hydroxide (NaOH), H₂O₂, potassium chloride (KCl), calcium chloride (CaCl₂), gluconic acid, potassium hydroxide (KOH), L-histidine, D-mannitol, sodium pyruvate, monobasic potassium phosphate (KH₂PO₄), citric acid, 4′-6-diamidino-2-phenylindole (DAPI), hydrochloride (HCl), sodium bicarbonate (NaHCO₃), 3,3′-diaminobenzidine (DAB), harris hematoxylin, dithizone (DTZ), mayer’s hematoxylin, biebrich scarlet aqueous, acid fuchsin, glacial acetic acid, phosphotungstic acid, light green, and OptiPrep™ were purchased from Sigma Aldrich. Dulbecco’s phosphate-buffered saline (DPBS), RPMI 1640, Media 199 (M199), fetal bovine serum (FBS), antibiotic/antimycotic solution, gentamycin, and porcine serum (PS) were obtained from GIBCO (Carlsbad, CA, USA). Dexa was purchased from Jeil pharm (Seoul, Korea). Cefazolin was purchased from ChongKunDang (Seoul, Korea), ketamine HCl from Yuhan Corporation (Seoul, Korea), CIzyme collagenase MA and CIzyme BP protease from VitaCyte (Indianapolis, IN, USA) and Phenol red free 1×Hank’s balanced salt solution (1×HBSS; without Ca²⁺, Mg²⁺) was purchased from WelGENE Inc. (Gyeongsan, Korea).

Porcine islets were isolated from SNU DPF adult pigs according to established protocols [22,23]. After cannulation of the pancreatic duct using cold histidine-tryptophan-ketoglutarate (HTK) solution (Custodiol™, Koehler Chemi, Bensheim, Germany), the pancreas was delivered in cold HTK solution. The delivered pancreas was washed sequentially with 1% cold povidone, a cold solution of antibiotics (0.2% cefazolin, 0.016% gentamycin dissolved in DPBS), and cold DPBS. Washed pancreas was trimmed to remove non-pancreatic tissues in cold HTK solution, then CIzyme collagenase MA (0.12 CDA/pancreas gram)/BP (0.03 million neutral protease activity units per pancreas gram) protease in 200 mL of phase I solution (10 mM NaOH, 5.8 mM KOH, 1 mM CaCl₂, 1.6 mM MgSO₄, 45 mM D-mannitol, 28.8 mM NaH₂PO₄, 105 mM NaCl) was injected into the pancreatic duct. Injected pancreas was digested in a Ricordi chamber (Biorep technologies, Miami Lakes, FL, USA) and 500 mL phase I solution addition, where it was incubated without shaking for 20 to 25 minutes at 35°C to 37°C until the pancreatic tissue was loosened, followed by gentle shaking and serial sampling. When free islet cells were observed in the sample, digestion was stopped by adding 20% PS in cold phase II solution (90 mM gluconic acid, 30 mM L-histidine, 20 mM KOH, 0.3 mM CaCl₂, 10.2 mM MgSO₄, 50 mM D-mannitol, 5 mM sodium pyruvate, 20 mM KH₂PO₄). Digested pancreatic tissues were collected and washed in 10% PS in cold phase II solution and centrifuged for 2 minutes at 1,000 rpm (Hanil science industrial, Gimpo, Korea). Islets were purified using a continuous OptiPrep™ density gradient (1.100, 1.085, and 1.060 g/cm³ density) and a Cobe 2991 cell separator (Gambro BCT Inc., Lakewood, CO, USA). The purified islets were washed with cold phase II solution and cultured in M199 supplemented with 10% PS, 16 mM nicotinamide, 24 mM NaHCO₃ and 1% penicillin/streptomycin at 37°C in humid conditions with 5% CO₂.

Alginate solution for microencapsulation was prepared that 2% (w/v) UP-LVG alginate was dissolved in Ca²⁺-free Krebs–Ringer–HEPES buffer (KRH buffer: 135 mM NaCl, 4.7 mM KCl, 25 mM HEPES, 1.2 mM KH₂PO₄, 1.2 mM MgSO₄). After 1 to 2 days in cultured, the isolated SNU DPF porcine islets were suspended in a solution of 2% UP-LVG alginate and at a concentration of 10,000 IEq/mL for SNU DPF porcine islets. Microcapsules were made with a coaxial bead generator (Nisco Engineering Inc., Zürich, Switzerland) that expelled islet-solution mixture using a syringe pump (Harvard apparatus, Holliston, MA, USA) into a 150 mm petri-dish (SPL, Pocheon, Korea) of 10 mM BaCl₂ solution (10 mM BaCl₂, 2 mM KCl, 135 mM NaCl, 10 mM HEPES, pH 7.4). The encapsulated islets were washed with 1×HBSS and 1 mg/mL Dexa contained pH 4.5 1% filtrated chitosan solution (w/v in deionized distilled water) for 10 minutes with stirring for Dexa-chitosan coating. Chitosan, high affinity positive charged residue [24], was used as an intermediary to link alginate and Dexa because both alginate and Dexa have negative charges.

All these procedures are made at room temperature. After chitosan coating, microcapsules were washed 3 to 4 times with 1×HBSS and cultured in M199 supplemented with 10% PS, 16 mM nicotinamide, 24 mM NaHCO₃ and 1% penicillin/streptomycin at 37°C in humid conditions with 5% CO₂.

Dexa-chitosan-coated alginate surface was observed using transmission electron microscopy (TEM; JSM-1010, JEOL Ltd., Tokyo, Japan). Microcapsule pore size was determined by the diffusion FITC-labeled dextran conjugates of various sizes [25]. Alginate and Dexa-chitosan-coated alginate microcapsules were incubated for 3 days in 1×HBSS with FITC-labeled dextran conjugates (4, 10, 20, 40, 70, 150, 250 kDa). After incubating, each microcapsule was imaged under a confocal fluorescence microscope (AxioVert 200TM, Carl Zeiss, Oberkochen, Germany).

The standard curve was produced by dissolving Dexa in 1×HBSS; a range of 0 to 500 μg/mL was obtained. The standard Dexa concentration was measured using a UV/Vis spectrometer (Nanodrop, Thermo Fisher Scientific, Waltham, MA, USA) at an absorption wavelength of 240 nm. The 1,000 ea Dexa-chitosan-coated alginate microcapsules were incubated in a 100 mm petri-dish with 10 mL 1×HBSS solution, at 37°C under humid conditions with 5% CO₂, for the Dexa release test. After 35 days, 10 μL supernatant samples were obtained from the incubated petri-dish and measured using the same UV/Vis wavelengths as the standard preparation spectrometer [26]. A regression analysis of the in vitro Dexa release was used to determine the following equation for the Dexa standard curve: Y=[0.412×X (R²=1.000)], where Y is the concentration of Dexa (μg/mL), X is the value of the absorbency of Dexa, and R² is the regression coefficient.

The standard AO/PI method was used to evaluate islet viability 0, 1, 3, 5, 7, and 14 days after microencapsulation. Naked, alginate microencapsulated, and Dexa-chitosan-coated alginate microencapsulated islets were incubated in 1×HBSS with 1% FBS with AO/PI for 10 minutes. Viability of these islets was evaluated using AO/PI double staining to observe viable (AO; green) and dead cells (PI; red) with an inverted fluorescence microscope (Observer.Z1, Carl Zeiss).

Tests were performed on islet samples of 100 IEq (150 to 250 μm diameter islet selection) for each group. Islet samples were transferred to a 60 mm petri-dish (SPL) and incubated in a 5 mL 1% FBS contained glucose-free Krebs-Ringer-Bicarbonate buffer (KRB buffer: 24 mM NaHCO₃, 5 mM KCl, 115 mM NaCl, 2.5 mM CaCl₂, 1 mM MgCl₂, 25 mM HEPES) for 1 hour at 37°C in humid conditions with 5% CO₂ before performing the glucose-stimulated insulin secretion (GSIS) test. Then, each group of islets was incubated in a 5 mL 1% FBS contained low-glucose (2.8 mM) KRB buffer for 1 hour 37°C in humid conditions with 5% CO₂. After incubation, each group of islets was washed with an 1% FBS contained glucose-free KRB buffer and incubated in a 5 mL 1% FBS contained high-glucose (16.8 mM) KRB buffer for 1 hour at 37°C in humid conditions with 5% CO₂. The amount of insulin was determined using the human insulin radioimmunoassay kit (RIA kit, Millipore, Burlington, MA, USA).

Male GalT KO mice weight 20 to 30 g were used as recipients for each type microencapsulated islet transplantation. Their presence of diabetes was considered if two or more non-fasting blood glucose levels from the tail vein exceed 300 mg/dL after intraperitoneal injection of STZ. Two groups of transplanted animals were studied: (1) alginate microencapsulated islets were transplanted (group 1; n=5); (2) Dexa-chitosan-coated alginate microencapsulated islets were transplanted (group 2; n=6). Microencapsulated islets (10,000 IEq) were slowly infused into the peritoneal cavity via a flank incision. After transplantation, blood glucose levels were monitored every day in the first week and twice a week afterwards. Graft failure was determined if three consecutive non-fasting blood glucose levels exceeds 300 mg/dL.

Two hundred and thirty-one days after transplantation, microcapsules were harvested under ketamine HCl anesthesia by peritoneal lavage using warm 1×HBSS. The harvested microcapsules were stained with DTZ and assessed under an inverted microscope to establish the degree of fibrotic cell infiltration (CKX41, Olympus, Tokyo, Japan). Images were captured by an E330 digital camera (Olympus) connected to the inverted microscope. Fibrosis infiltration score was defined on the harvested microcapsule surface (Fig. 1) [27]. Fibrosis infiltration score =(0×0%/total)+(3.3×<50%/total)+(6.6×>50%/total)+(10×100%/total).

Harvested microcapsules were fixed 15 to 30 minutes in 4% PFA at 4°C. Fixed capsules were embedded with 5% LB agarose at room temperature for 10 minutes. And, it was sectioned at 4 to 7 μm. Immunohistochemistry staining was conducted using specific antibodies such as insulin, macrophage, CD3. Specimens were incubated overnight at 4°C with diluted primary antibodies (insulin, CD3, macrophage; Abcam, Cambridge, MA, USA), and then incubated for 1 hour with fluorescence-conjugated secondary antibodies (FITC, Rhodamine, Vector Lab, Burlingame, CA, USA) at room temperature. DAPI (Sigma-Aldrich) was employed for nuclear staining. Immunofluorescence was detected under a fluorescence microscope (Observer.Z1). For infiltrated fibrosis on harvested microcapsules stain, deparaffinized specimens were incubated with weigert’s iron hematoxylin at room temperature for 10 minutes. Next, Biebrich scarlet-acid fuchin solution at 15 minutes and 5% aqueous phosphotungstic acid at 3 minutes and 1% glacial acetic acid at 3 minutes. Finally, it was incubated with light green solution at 15 minutes.

Results were presented as the mean±standard error of mean values. Survival rate was analyzed using log-rank test survival curves, differences between two groups were analyzed using the two-tailed unpaired t test using GraphPad Prism version 5.0 (GraphPad software Inc., La Jolla, CA, USA). P value of less than 0.05 was assumed statistical significance.

The layer of Dexa-chitosan coating on alginate microcapsules was observed by TEM (Fig. 2A). The permeability of each microcapsule was estimated using a diffusion assay of FITC-labeled dextran molecules of various sizes (Fig. 2B). Dexa-chitosan-coated alginate microcapsules exhibited similar diffusion of FITC-dextran at a molecular weight of 4 to 40 kDa compared with alginate microcapsules. This finding suggests that the Dexa-chitosan coating did not affect the permeability of gas exchange and nutrients through the microcapsule inner core. However, diffusion of FITC-labeled dextran in Dexa-chitosan-coated alginate microcapsules was blocked between 40 and 70 kDa. This finding suggests that the molecular cutoff weight of the capsule membrane is at least 70 kDa, which can prevent immune cells and immunoglobulin G (IgG; 150 kDa) from penetrating into the microcapsules. Dexa release from Dexa-chitosan-coated alginate microcapsules was measured using a UV/Vis spectrometer (Fig. 2C). Dexa release was relatively constant over 35 days after microencapsulation.

The viability of encapsulated islets was confirmed by AO/PI assay on days 0, 3, and 7 after microencapsulation and compared with naked islet viability (Fig. 3A). The survival rate of microencapsulated islets stabilized after 3 days. Fourteen days after microencapsulation, the percentage of PI-positive cells in the Dexa-chitosan-coated alginate group (6.05%±1.73%) was similar to that of the alginate group (5.99%±0.80%) and did not differ from that of the naked islets (4.13%±0.91%) (Fig. 3B). Islet function assessed by GSIS was also similar among naked, alginate, and Dexa-chitosan-coated alginate groups 14 days after microencapsulation. These results suggest that islet viability and function were well maintained after microencapsulation (Fig. 3C).

Before transplantation, the nonfasting blood glucose levels of all diabetic mice were greater than 300 mg/dL. Blood glucose levels significantly decreased after transplantation and remained within the normal range until 231 days after transplantation in both the Dexa-chitosan-coated alginate and alginate groups (Fig. 4A). The average nonfasting blood glucose level for 231 days after transplantation was significantly reduced in the Dexa-chitosan-coated alginate group compared with the alginate group (176.0 mg/dL vs. 202.7 mg/dL, P<0.05). The graft survival rate was higher in the Dexa-chitosan alginate group compared with the alginate group, but the difference was not significant (log rank test, P=0.228) (Fig. 4B).

Pericapsular fibrosis was observed from harvested microcapsules 231 days after transplantation. As shown in Fig. 5A (upper panels), the fibrosis of Dexa-chitosan-coated alginate microencapsulated islets was conspicuously decreased compared with that of the alginate group. The fibrosis infiltration score was significantly reduced in the Dexa-chitosan-coated alginate group compared with the alginate group (Fig. 5B). Similarly, Masson’s trichrome staining revealed significant collagen deposition around alginate microcapsules but not around Dexa-chitosan-coated alginate microcapsules (Fig. 6A). In alginate microcapsules, most infiltrated cells were positive for macrophage staining, and a few cells were positive for CD4 staining (Fig. 6B). In contrast, inflammatory cell infiltration was not observed in Dexa-chitosan-coated alginate microcapsules (Fig. 6B). Taken together, these results suggest that Dexa-chitosan-coated alginate microcapsules can significantly suppress pericapsular fibrosis and inflammatory cell infiltration in xenogeneic microencapsulated islet transplantation compared with the alginate group.

The graft function of harvested microencapsulated islets was assessed by DTZ and insulin staining. In the Dexa-chitosan-coated alginate group, most harvested cells were positive for DTZ and insulin (Fig. 5A, lower panel and Fig. 6C). Although alginate microcapsules exhibited significant pericapsular fibrosis, DTZ and insulin staining were positive in most harvested cells (Fig. 5A, lower panel and Fig. 6C).