Isolation of human liver cells
Adult liver cells were isolated and cultured as previously described [11, 14]. This study was carried out according to the guidelines and with the approval of the Institutional Review Board of Asan Medical Center (IRB number: 2014-1182 Seoul, Republic of Korea). Liver tissue was obtained from donors undergoing abdominal surgery; informed consent was obtained from the patients for this procedure. In the case of non-diabetic patients and patients with type 2 diabetes, liver tissues were obtained with consent from patients who underwent pancreatic surgery due to benign pancreatic disease. In the case of patients with type 1 diabetes, liver tissue was obtained with consent from patients who underwent abdominal surgery for pancreatic transplantation. The selection criteria for liver tissue biopsy participants included patients undergoing pancreatic resection due to pancreatitis, pancreatic transplantation, or hepatectomy. Pancreatic tumors and pancreatic cancer were excluded from analyses. Exclusion criteria included patients with cholagiohepatitis prior to surgery, bilirubin >2 mg/dl or more, class B, C on liver function tests, and patients with uncorrected coagulation disorder. Patients whose liver function was expected to be abnormal were excluded. In addition, the liver tissues were confirmed to be normal liver tissues through histology analysis. Donors had an average age of50.2±16.1;14 men and 24 women acted as donors. Liver tissues were digested by 0.16% collagenase type I (Worthington Biochemical, NJ, USA) and plated on fibronectin-coated plates (3 μg/cm2, Sigma-Aldrich, St. Louis, MO, USA). The cells were cultured in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal bovine serum (FBS), 1% antibiotic-antimycotic and Glutamax (Life Technologies, Carlsbad, CA, USA). The medium was changed daily during the first three days to remove non-adherent cells and from the fourth day onwards, the medium was changed every 2 or 3 days. Liver cells were passaged when approaching confluence using trypsin-EDTA.After propagation, liver cells from passages 5 to 7 were used for the experiments. To evaluate the growth characteristics of the isolated liver cells, the time required by the cells to proliferate (doubling time) was measured. At every passage, we seeded 2.5x105 cells in 100-mm culture dishes. At confluence,the cells were harvested and countedusing trypan blue staining. Population doubling time was determined using a previously reported formula [20], and was confirmed using anonline doubling time calculator application.
Characterization of liver cells
We confirmed surface marker expression and differentiation capacity of liver cells. Liver cells at early (1-2), mid (6-7), and late (12-14) passages were blocked with 1% bovine serum albumin and incubated for 1 h at 4°C with the following anti-human antibodies conjugated with phycoerythrin (PE): Mouse IgG1 isotype control, CD31, CD45, CD90, CD73, and CD105 (BD Biosciences, San Jose, CA, USA). Additionally, HLA-DR was analyzed to determine the immune properties. Cells were analyzed using a FACSCalibur device (BD Biosciences). For analyzing albumin expression and ectopic gene expression of PDX1, NEUROD1, and MAFA, liver cells were fixed and permeabilized using eBioscienceTM Foxp3/Transcription factor staining buffer set (Thermo Fisher Scientific, Waltham, MA, USA) and were stained using indirect method. Furthermore, the cells were stained using rabbit anti-PDX1, mouse anti-NEUROD1, and rabbit anti-MAFA (1:200; Abcam). For secondary fluorescence labeling, the cells were incubated with anti-rabbit IgG Alexa Fluor 488 and anti-mouse Alexa 555 (1:200; Thermo Fisher Scientific). To confirm the introduction of PDX1, NeuroD1, and MAFA genes in cells using flow cytometry, the liver cells in which the transcription factor genes (PDX1, NEUROD1, and MAFA) was not introduced were stained in the same way as gene transduced cells. For albumin control, primary mouse IgG1 isotype control was used as primary antibody instead of mouse anti-albumin.
For adipogenic and osteogenic differentiation, we used the adipogenic or osteogenic differentiation media (Lonza, Basel, Switzerland). After 14 days in culture, the adipogenic culture formed vacuoles. The plates were fixed and stained with Oil Red O (Sigma-Aldrich). The osteogenic differentiation cultures were incubated for 28 days and fixed and stained with 1% Alizarin Red solution pH 4.1 (Sigma-Aldrich). We also confirmed the expression of albumin by isolated liver cells at early (1-2), mid (6-7) and late (12-14) passages. Albumin expression were analyzed by flowcytometry and immunofluorescence staining. For immunofluorescence staining, cells were fixed with 4% paraformaldehyde (Merck, Darmstadt, Germany) and permeabilized with 0.1% Triton X-100. After antibody blocking, primary antibodies were incubated overnight at 4°C. We used mouse anti-albumin (1:100, Santa Cruz Biotechnology, Dallas, TX, USA) and anti-mouse Alexa 488 (1:200; Thermo Fisher Scientific, Waltham, MA, USA). ProLong Gold antifade reagent with 4′,6-diamidino-2-phenylindole (DAPI, Thermo Fisher Scientific) was used to stain the nuclei and for mounting. The slides were examined using the EVOS® FL auto cell imaging system (Thermo Fisher Scientific).
Preparation of IPCs and IPC sheets
To induce trans-differentiation, ß cell-related transcription factors were transduced into liver cells. PDX1, NEUROD1 and MAFA were selected and transduced with adenovirus expressing vectors. Ad-CMV-hPDX1, Ad-CMV-hNEUROD1, Ad-CMV-MAFA and Ad-CMV-GFP were obtained from Vector Biolabs (Burlingame, CA, USA). Ad-RIP-luciferase was generously provided by Orgenesis, Inc. (Germantown, MD, USA). The expression of green fluorescent protein (GFP) after transduction of GFP for 2 days was assessed by flow cytometry. Liver cells were transduced with PDX1 and NEUROD1 for 2 days, followed by replacement of the medium with medium containing MAFA for 3 days. Liver cells were cultured in 10 mM nicotinamide (Sigma-Aldrich), 20 ng/mL epidermal growth factor (PeproTech, Rocky Hill, NJ, USA), and 5 nM exendin-4 (Sigma-Aldrich) included in the culture medium. The optimal MOI was determined according to GFP transduction efficiency, cell survival, insulin promotor activity, and insulin gene expression under various conditions. MOIs of the viruses were 500, 250, 50, 200, and 200 for PDX1, NEUROD1, MAFA, GFP and luciferase, respectively.
The scheme of this study is summarized in Figure 1a. IPCs and the IPC sheets grown on culture dishes were transduced simultaneously with PDX1 and NEUROD1 for 2 days. Next, cells were harvested and re-seeded in commercially available PIPAAm dishes (UpCell; CellSeed, Inc., Tokyo, Japan)—having temperature-responsive properties—to prepare the IPC sheets. For the differentiation of IPCs, cells were re-seeded in culture dishes at a concentration of 106/100mm. MAFAwas transduced into IPCs and IPC sheets when cells were re-seeded. On the next day (day 3 after initial virus exposure), the differentiation media was replenished in IPC-seeded culture plates. In the case of IPC sheets, the cell sheets cultured on the UpCell dish were harvested by reducing the temperature from 37°C to 20°C (Fig. 1b) on day 3. For the in vitro assay, the IPC sheet was attached to the insert well with an 8-μm pore size (SPL, Gyeonggi-do, Korea). Five days after the initial virus exposure, the IPC sheets and IPCswere harvested to analyze mRNA levels and insulin production; these values were compared to those obtained on culturing the IPCs on the monolayer culture dishes.
Luciferase assay
To optimize the virus transduction conditions, luciferase activity controlled by the insulin promoter was assessed. At day 2, cells were co-infected by adenovirus containing the firefly luciferase gene controlled by the rat insulin promoter (Ad-RIP-luciferase). At day 5 after initial virus exposure, the cells were harvested and lysed in passive lysis buffer (Promega, Madison, WI, USA). Luciferase activity was measured using the Luciferase Assay System (Promega) and normalized to protein levels. Luminescence was detected by a VICTOR2TM 2030 multilabel plate reader (Perkin Elmer, Waltham, MA, USA).
Real-time quantitative PCR (qPCR)
Total RNA was extracted from liver cells and IPCs on predetermined days using TRIzol reagent (Thermo Fisher Scientific). cDNA was synthesized from 1 μg RNA template by an oligo-dT primer using a SuperScript III First-Strand Synthesis System (Thermo Fisher Scientific). Real-time PCR was performed using LightCycler 480 SYBR Green I Master mix (Roche Applied Science, Mannheim, Germany) in a LightCycler® 480 II real-time thermal cycler (Roche Applied Science). The pancreatic endocrine gene-specific primer sets used are listed in Table 1. Gene expression was normalised to the glyceraldehyde 3-phosphate dehydrogenase housekeeping gene, and relative quantification was performed.
Immunofluorescence staining
Cells were fixed with 4% paraformaldehyde and permeabilized with 0.1% Triton X-100. After antibody blocking, primary antibodies were incubated overnight at 4°C. We used rabbit anti-insulin and mouse anti-glucagon (1:1000; Abcam, Cambridge, UK), rabbit anti-PDX1, goat anti-PDX1, mouse anti-NEUROD1, rabbit anti-MAFA (1:200; Abcam) and mouse anti-albumin (1:100, Santa Cruz Biotechnology). For secondary fluorescence labelling, the cells were incubated with anti-rabbit IgG Alexa Fluor 488, anti-rabbit IgG Alex Fluor 555, anti-goat IgG Alex Fluor 488, anti-mouse Alexa 555 and anti-mouse Alexa 488 (1:200; Thermo Fisher Scientific). ProLong Gold antifade reagent containing DAPI (Thermo Fisher Scientific) was used to stain the nuclei and for mounting. The slides were visualized using the EVOS® FL auto cell imaging system (Thermo Fisher Scientific). Under the microscope, the cells with red fluoresce were insulin or glucagon positive and blue fluorescence highlighted the nucleus and counted in 10 randomly selected fields per IPC sheet or IPC cells with total of 3 donors in each group. The results were presented as insulin or glucagon positive cells per 100 cells.
Insulin and C-peptide analysis
To measure the insulin and C-peptide contents in IPCs, cells were washed and lysed with RIPA lysis buffer[21]. Insulin and C-peptide levels were measured using a commercial ultrasensitive insulin ELISA kit (80-INSHUU-E01.1 ALPCO, Salem, NH, USA) and ultrasensitive C-peptide ELISA kit (10-1141-01, Mercodia, Uppsala, Sweden), respectively. Also, insulin secretion was measured in the culture medium of IPCs and IPC sheets by static incubation for 2 days in differentiation culture medium containing 5.5 mM glucose. Insulin and C-peptide contents were ascertained in cells from four different donors. Prior to glucose stimulation of the cells, any residual insulin released from the islets wasremoved by incubation in serum- and glucose-free RPMI 1640 medium. The tubes were kept at 37 ºC for 1 h with shaking and then the medium was completely removed by centrifugation and replaced with serum-free RPMI 1640 medium supplemented with 2.8 mM glucose for 1 h. Following collection of the medium, the cells were then incubated with serum-free RPMI 1640 containing 28 mM glucose for an additional 1 h. The supernatant from each sample was collected. The assays were performed in triplicate. The stimulation index was calculated as a ratio between the insulin secreted at high and low glucose media[22].
Transmission electron microscopy (TEM)
To analyze the granular ultrastructure, the cells were fixed with 1% glutaraldehyde and 1% PFA in 0.1 M sodium cacodylate buffer (pH 7.2) at 4°C. Specimens were then fixed in 2% osmium tetroxide for 60 min at 4°C. Dehydration of the fixed samples was performed, and the samples were transferred to Lowicryl resin (Polyscience, Niles, IL, USA). Samples were then sectioned (60 nm) with an ultramicrotome (UltracutUCT, Leica, Wetzlar, Germany) and collected on nickel grids. Post-embedding immunogold labelling was performed for insulin and glucagon labelling using the rabbit anti-insulin (Abcam), mouse anti-glucagon (Abcam), 5-nm colloidal gold conjugated to goat anti-rabbit IgG (Sigma-Aldrich) and 9–11-nm colloidal gold conjugated to goat anti-mouse IgG (Sigma-Aldrich). Following immunogold labelling, the sections were double-stained with 2% uranyl acetate for 20 min and lead citrate for 10 min. The sections were then viewed using TEM.
Transplantation of IPCs and IPC sheets to diabetic mice
This study was reviewed and approved by the Institutional Animal Care and Use Committee (IACUC No. 2015-12-133) Asan Institute for Life Sciences. The committee abides by the Institute of Laboratory Animal Resources (ILAR) guide. All experiments related to animals were performed in accordance with the relevant guidelines and regulations. To develop insulin-dependent diabetic mouse model, male 8-week-old BALB/c nude mice were treated with 180 mg/kg STZ (Santa Cruz Biotechnology) dissolved in 0.1 M citrate buffer (pH 4.5). After inhalational anesthesia with isoflurane, the abdomen of each mouse was swabbed with Betadine and an approximately 1.5 cm incision was made. Once the portal vein was exposed, it was flat enough to allow a 30-gauge needle to be inserted without restriction. Slight traction was kept on the pancreas near the vein bed with a sterile Q-tip to create tissue tension for injection through the needle. A syringe was attached to the 30-gauge needle and a suspension of IPCs (1 × 106 cells in 100 μL of phosphate buffered saline) was released into the portal vein for 1 min. We did not routinely inject more than 106 cells because the injection of greater numbers of cells often was lethal to the mice. Using Q-tips, the injection site was pressed gently for at least 5 min to stop any bleeding. The abdomen was closed and treated with Betadine.
Before cell sheet transplantation, liver capsules at the intended site were removed by swabbing the liver surface. An IPC sheet with CellShifterTM was placed on the liver surface, and the CellShifte TMr was removed after 5 min. One IPC sheet contained approximately 106 cells. We transplanted one cell sheet for the biodistribution study to match the cell number with the portal vein injection group. To increase the efficacy of IPC sheets, we transplanted one IPC sheet or three IPC sheets onto the liver of the diabetic mice for efficacy evaluation. Blood glucose was monitored after transplantation using a Codefree blood glucose monitoring system (SD Biosensor, Suwon, Korea). To confirm human insulin secretion after transplantation, we collected blood serum on the predetermined date and evaluated human serum insulin using an ultrasensitive insulin ELISA kit (80-INSHUU-E01.1 ALPCO, Salem, NH, USA).
Cell labelling and imaging ex vivo and in vivo
IPCs were labelled with Qdot 800 (Qtracker 800; Molecular Probes, Inc., Eugene, OR, USA). In vivo optical imaging was conducted using an IVIS Spectrum imaging system (Caliper Life Science Inc., Waltham, MA, USA). Images were acquired using an excitation and emission wavelength of 430 and 800 nm, respectively. Fluorescence was quantified as the sum of all detected photon counts per second within a constant region of interest for each transplant site (in vivo). At 7 days after transplantation, the mice were sacrificed and the livers, kidney, spleen, pancreas and lung were harvested. The fluorescence in each organ was detected (ex vivo image).
Liver toxicity
Blood samples were collected from the retro-orbital sinus. Hepatic toxicity measurements included ALT and aspartate aminotransferase AST in the blood serum at 1, 2 and 5 days after transplantation. Serum ALT and AST levels were analyzed using a model 7180 automatic biochemistry clinical analyzer (Hitachi, Tokyo, Japan).
Histological analysis
After sacrificing the mice on days 7 and 14, livers were removed and fixed with 4% formalin. A paraffin block was prepared and cut into 4-μmsections. The tissue slides were deparaffinized and dehydrated. Samples subjected to antigen retrieval were blocked with 3% bovine serum albumin. Immunohistochemistry was performed using primary antibodies for rabbit anti-PDX1 (dilution 1:200, Abcam) and rabbit anti-insulin (dilution 1:1000, Abcam). To confirm, neovascularization in transplanted site, we also stained liver tissue using CD31 (dilution 1:1000, Abcam). A DAB Detection Kit (REAL EnVision detection system, DAKO, Glostrup, Denmark) was used for immunohistochemistry. To perform hematoxylin and eosin staining, samples were deparaffinized and dehydrated, followed by staining with hematoxylin (Sigma-Aldrich) and eosin (Sigma-Aldrich).
Statistical analyses
The data are presented as the mean ± standard deviation, with the number of samples indicated in the figure legends. The statistical significance of the differences between multi-groups was analyzed with one-way ANOVA test and by determining the differences between the means with Tukey’s post-hoc test. The statistical significance of the difference between the two groups was analyzed with Student’s t-test. P < 0.05 indicated a statistically significant difference.