Lewis rats were purchased from Japan SLC Inc (Shizuoka, Japan). For all experiments, scaffold recipients were 14–16 weeks old, scaffold and pancreatic islet donors were 10–12 weeks old, and BM and ADSC donors were 8 weeks old. Except for the BM donors, all of the rats used in this study were males. All animals had free access to a standard diet and water.

The experiments were approved by the local ethics committee (approved protocol 2019MdA-104) and performed in accordance with national and institutional regulations. Animals were maintained in a specific-pathogen-free environment. All surgeries were performed under anesthesia, and all efforts were made to minimize suffering.

Livers were obtained from male Lewis rats after cannulation of the portal vein (PV) and hepatic artery (HA) with a 22-G venflon cannula, and a tube (TOP X2-100 extension tube, TOP Co., Tokyo, Japan) was placed in the infra-hepatic inferior vena cava (IVC-tube).

The liver was decellularized following the protocol described by Vishwakarma et al. (2019). In brief, after washing out the blood, the liver was sequentially perfused with 0.1% w/v sodium dedocylsulphate (SDS; Wako, Osaka, Japan) for 2 h, SDS 0.5% w/v for 18 h, distilled water for 10 min, 1% Triton X-100 v/v (Wako) for 30 min, and finally distilled water (DW) for 30 min to wash out the detergents. Perfusion was performed through the HA cannula at a flow rate of 1 ml/min using a peristaltic pump (Perista AC-2110; ATTO Corporation, Tokyo, Japan). The decellularized liver (DLM) was kept in 100 ml of DW containing 1% (1,000 U) heparin sodium (Mochida, Tokyo, Japan) and 1% penicillin/streptomycin (Thermo Fisher Scientific, Waltham, MA, United States) in a sterile container at 4°C until the time of implantation (Figure 2).

All procedures were performed using sterile techniques and sterile DW. Detergent perfusion was performed in a laminar flow cabinet. After completion of decellularization protocol, 0.5 × 0.5-cm samples were excised and fixed in 4% paraformaldehyde for the histological assessment of the decellularization in selected samples.

ADSCs were isolated from male Lewis rats as described previously (Ohmura et al., 2010), with a few modifications. In brief, adipose tissue was obtained from the inguinal fat pad and cut into fine pieces before being placed in ice-cold phosphate-buffered saline (PBS; Sigma-Aldrich, St. Louis, MO, United States) and then in Hank’s balanced salt solution (HBSS; Sigma-Aldrich) containing 1 mg/ml collagenase II (Sigma-Aldrich) and 1% penicillin/streptomycin, at 37°C for 30 min with gentle agitation. The digested tissue was filtered through a sterile 70-μm cell-strainer, centrifuged at 400 g for 5 min at room temperature, and resuspended. This process was repeated twice. ADSCs were seeded into T-75 cell culture flasks (Thermo Fisher Scientific) in ADSC-1 medium (Kohjin Bio Co., Ltd., Sakado, Japan) containing 10% fetal bovine serum (Thermo Fisher Scientific) and 1% penicillin/streptomycin (Figure 3). Passage 2 cells were cryopreserved after resuspension in CELL BANKER^® 1plus (Takara Bio Inc., Kusatsu, Japan).

Fourteen-to 16-week-old male Lewis rats were used as recipients (315–370 g). Induction of anesthesia was achieved using 3–4% isoflurane inhalation, and 1.5% isoflurane was used for maintenance during surgery. After hair clipping, the skin of the lower abdomen and left thigh region was prepared with 70% alcohol, and a sterile surgical drape was sutured around the location. A curved skin incision was performed, and the saphenous vessels were dissected using microsurgical techniques and electrocautery. Heparin sodium (100 U) was injected intravenously, followed by ligation of the arteriovenous (AV) bundle at the knee level (before bifurcation). The liver was placed inside the subcutaneous pouch, and the AV bundle was threaded through the scaffold’s IVC to enter from the supra-hepatic IVC and exit through the infra-hepatic IVC by passing the AV bundle ligation thread through the IVC tube. The IVC tube was then removed, and the AV bundle was fixed to the muscle using 5/0 nylon sutures. The scaffold was oriented, and a drop of Viti-bond glue (3M, St. Paul, MN, United States) was used to keep the scaffold in place (Figures 4A–D). The cannulas to the PV and HA were cut and kept inside the subcutaneous pocket for future use. Hemostasis was checked, and the wound was closed in 2 layers using 4/0 Nylon sutures. Saline (5 ml) containing 100 µL of 30% glucose was injected subcutaneously for rehydration. Rats were kept in a warm cage until complete recovery.

Before wound closure, 2.5 ml of HBSS (Thermo Fisher Scientific) was injected through the PV cannula using a butterfly 23-G needle containing no cells (HBSS group; n = 4), freshly obtained BM preparation (BM group; n = 3), or ADSCs (ADSC group; n = 3). Injection was manually performed at a rate of 1 ml/min (Figures 4E–G).

BM was collected from the femur of female Lewis rats after decontamination with 1% Isodine solution (Mundipharma, Tokyo, Japan). Using sterile instruments, the femur was cut near its end, and BM was extruded onto a cell culture plate using HBSS. The process was repeated for the other femur (5 ml total). The collected crude BM was then gently aspirated using a 25-G butterfly cannula for mechanical dispersion to produce a single-cell suspension and collected into a sterile 5-ml Eppendorf tube (Figure 3). A 2.5-ml aliquot of the single-cell suspension was injected into the decellularized scaffold through the PV. Cell counting was then performed using a hemocytometer.

The cryopreserved ADSCs were thawed at 37°C, resuspended in 10 ml of ADSC-1 medium, and centrifuged at 400 g for 5 min at 4°C. The supernatant was discarded, and the pellet was resuspended in medium. The cells were cultured in T-75 flasks. The medium was changed every two to 3 days. Before the time of transplantation, the ADSCs from passage 3 or 4 were collected after detachment with 0.05% trypsin/EDTA (Thermo Fisher Scientific) and washed. A total of 4×10⁶ cells were then suspended in 2.5 ml of HBSS, and the 2.5-ml cell suspension was injected into the decellularized scaffold through the PV. The ADSCs used in this study belong to the same lot (from the same isolation).

After surgery, the rats were followed up by body weight measurements until day 3 and then weekly for 4 weeks. Four weeks after surgery, streptozotocin (STZ; Sigma-Aldrich) was injected intravenously through the penile vein (65 mg/kg). The blood glucose was monitored using a portable glucometer (Freestyle; ABBOTT, Tokyo, Japan). Rats were considered diabetic if 2 consecutive measures of blood glucose were >400 mg/dl.

Pancreatic islets were isolated from male Lewis donors as described previously (Saitoh et al., 2021). In brief, rats were anesthetized by isoflurane inhalation (Abbott Japan Co., Ltd., Tokyo, Japan). The bile duct was identified and clamped at the papilla of Vater. Cold HBSS (10 ml) containing 1 mg/ml collagenase (Sigma type V; Sigma Chemicals, St. Louis, MO, United States) was injected into the common bile duct leading to the pancreas. The pancreas was removed and incubated in a water bath at 37°C for 12 min before being digested, and the cell suspension was washed 3 times in cold HBSS and centrifuged for 1 min. Density-gradient centrifugation was performed for 10 min using a Histopaque-1119 (Sigma Diagnostics, St. Louis, MO, United States) and LymphoprepTM (Axis-Shiled, Oslo, Norway) to isolate pancreatic islets. The islets were cultured in Roswell Park Memorial Institute Medium 1640 (RPMI-1640) (Thermo Fisher Scientific) containing 5.5 mmol/L glucose, 10% fetal bovine serum (Thermo Fisher Scientific), and 1% penicillin/streptomycin (Thermo Fisher Scientific) at 37°C in 5% CO₂ and humidified air (Figure 3).

One day after isolation, the islets were collected for transplantation. Anesthesia and surgical site preparation were performed as described earlier. The previous incision was re-opened, and the scaffold was exposed and carefully dissected from the overlying fibrofatty tissue. The cannula to the PV was exposed, and 300 µL of normal saline was injected using a 25-/23-G butterfly needle connected to a 1-ml syringe to test the patency. A total of 2,600 islet equivalents (IEQs; approximately 8 IEQs/g of rat weight) were then transferred to 100 µL of saline and aspirated using the butterfly needle. The islets were slowly injected into the scaffold through the PV in a total volume of 0.5–1 ml of saline and at a rate of 1 ml/min (manually). Occasional slight leakage of clear fluid out of the scaffold was considered acceptable.

The wound was then closed, and the rat was allowed to recover as described earlier. In one extra rat, both islets and hepatocytes (freshly isolated from male Lewis rats as described previously (Saitoh et al., 2021)) were injected, and the scaffold was retrieved immediately after injection to check the distribution of cells at day 0.

The rats were followed up for 30 days (±2 days) after transplantation for their general condition and correction of hyperglycemia. Rats were considered cured if 2 consecutive glucose measures were below 200 mg/dl. In cured animals and those with partial correction of hyperglycemia, an intravenous glucose tolerance test (IVGTT) was performed on day 28, and the scaffold was removed to confirm recurrence of hyperglycemia. The scaffold was fixed in 4% paraformaldehyde overnight and used for immunohistochemical analyses. 5-bromo-2′-deoxyuridine (BrdU; Abcam, Cambridge, United Kingdom) was injected intraperitoneally (i.p.) on days -4, -3, and -2 relative to the scaffold removal day (dose = 500 mg/kg). In one rat per group, micro-computed tomography (micro-CT) was performed at the study end-point to evaluate the AV bundle, as described later.

The IVGTT was performed as described previously (Jimbo et al., 2014; Inagaki et al., 2021). In brief, after fasting for 14 h with free access to water, the BW and BGL were measured, and 1 g/kg glucose was injected intravenously. The blood glucose was measured at 5, 10, 20, 30, 60, 90, and 120 mins, and the blood glucose curve was generated. The area under the curve (AUC) was then used for comparisons.

Μicro-CT was performed in one rat per group as described previously (Eweida et al., 2018a), with a few modifications. The abdominal aorta was cannulated with a 22-G venflon cannula, and 100 ml of normal saline containing 100 U of heparin sodium per mL was perfused using gravity to wash out the blood. A total of 20 ml of Microfil^® (MV-122; Flow Tech Inc., Carver, MA, United States) with 0.6 ml of curing agent was then injected using a 30-ml syringe through the cannula. After 90 min, the implanted scaffold was dissected and carefully removed. The scaffold was then imaged using a LaTheta™ LCT-200 (Hitachi Aloka Medical Ltd., Tokyo, Japan). The VGSTUDIO MAX™ software program (version 3.0) (Volume Graphics, Heidelberg, Germany) was used for the 3D reconstruction and video acquisition.

Removed scaffolds were fixed in 4% paraformaldehyde, dehydrated, and embedded in paraffin blocks. In brief, 4-μm sections were incubated with rabbit anti-von-Willebrand factor (anti-vWF;Sigma Aldrich, ab7356), rabbit anti-insulin (Abcam, ab181547), and rat anti-BrdU antibody-HRP (Abcam, ab220507). Rabbit Envision and goat anti-rabbit-HRP (4003; DAKO, Glostrup, Denmark) were used as secondary antibodies for anti-vWF and anti-insulin staining, respectively. Either methylene green or Hematoxylin was used for nuclear staining. Hematoxylin-Eosin (HE) staining was also performed. Six to 8 sections with equal intervals (250 µm) were evaluated per sample and scored by a pathologist in a blinded manner. The positive cells for vWF and BrdU were counted, and the total islet area was measured in each section. Then, we calculated the number of positive cells for vWF and BrdU per µm² of islet area (number/µm²) in each section. The resulting scores were used for statistical comparison.

Data for histological scoring are expressed as the mean ± standard error of mean (SEM). All statistical analyses were performed using the JMP pro 15 software program (SAS institute Inc., Carry, NC, United States). For histological scoring, a one-way analysis of variance (ANOVA) was used to compare the means, and the Tukey-Kramer test was used for the post hoc analysis. A student t-test was used to compare the means of AUC from the IVGTT results. p values of <0.05 were considered to indicate statistical significance.

At the end of the decellularization protocol, the DLM demonstrated a whitish, transparent appearance (Figure 2B). With adequate dissection of the subcutaneous plane up to the inguinal ligaments, placement of the whole-liver decellularized scaffold was possible. At the time of islet transplantation, no significant adhesion or fibrous capsule was observed, and the scaffold was easily visualized. A considerable yet expected decrease in size and change in shape was noted at this point. Injection of pancreatic islets into the scaffold was feasible with minimal leakage.

At the time of scaffold retrieval, the size of the scaffold was comparable to the size at islet transplantation. Vascularization of the scaffold surface was visible in most specimens. The AV bundle above and below the scaffold was clearly seen, and pulsation was frequently observed through the scaffold (Figure 5 and Supplementary Video 1). Cut sections through the scaffold showed an intact AV bundle, and the intra-graft vasculature was observed in silicon-injected samples.

Two rats in the BM group and 1 in the ADSC group showed normalization of blood glucose, (blood glucose <200 mg/dl on 2 consecutive measurements). One rat in each group showed partial correction of glucose level. No rats were cured in the HBSS group. After graft removal, hyperglycemia recurred, confirming the function of the graft. In one rat in the BM group, hyperglycemia did not completely recur after scaffold removal. Figure 6 shows the blood glucose levels of the rats during the study period.

The IVGTT was performed in rats with improved blood glucose levels (n = 6), and the results are shown in Figure 7. The AUC in rats with functioning grafts were significantly lower than values in diabetic rats (n = 3), p = 0.002, while significantly higher compared to normal Lewis rats (n = 6), p < 0.001.

In all three groups, there was evidence of vascular sprouting from the AV bundle inside the scaffold. Figure 8 shows representative micro-CT 3D reconstruction images. The micro-CT examination also confirmed the patency of the AV bundle in samples from the three groups (HBSS, BM, and ADSC). A representative video through the axial scans can be found in the supplementary data (Supplementary Video 2).

Histological examination showed insulin-positive cells in all three study groups (Figure 9). However, intact islets were more frequently encountered in the BM and ADSC groups than in the acellular scaffolds. The number of vWF-positive cells was higher within the islet region in the ADSC and BM groups than in the HBSS group (p = 0.08 and 0.10, respectively). BrdU + cells were significantly less frequent in the BM group than in both the ADSC and acellular groups (p = 0.003 for both), with no significant difference noted between the HBSS and ADSC groups.

The AV bundle was intact and patent in all analyzed samples with evident sprouting into the scaffold (Figure 10). No evidence of fibrosis or capsule formation was noted. A honeycomb appearance of the empty DLM was noted in some regions, while evidence of resorption could also be seen at other locations.