Long-term reversal of diabetes by subcutaneous transplantation of pancreatic islet cells and adipose-derived stem cell sheet using surface-immobilized heparin and engineered collagen scaffold

Animals

Male Sprague-Dawley rats and BALB/c-nu mice purchased from Orient Bio (Seongnam, Korea) were used as donors and recipients, respectively. All animal experiments were approved by the Institutional Animal Care and Use Committee of Asan Medical center (IACUC-2014-13-217).

Preparation of heparinized esterified collagen

Esterified collagen (EC) were fabricated by 1% (w/v) EC solution (Dalim Tissen, Korea) in 0.1 M acetic acid (Junsei Chemical, Tokyo, Japan). Lyophilized EC were crosslinked with 20 mM 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDC; Tokyo Chemical, Tokyo, Japan) in ethanol (Merck Millipore, Burlington, Massachusetts, USA). For heparin immobilization, freeze-dried EC were soaked in 50 mM 2-(N-morpholino)ethanesulfonic acid hydrate buffer (Sigma-Aldrich, St. Louis, MO, USA) supplemented with 1% (w/v) heparin (Wako Pure Chemical, Osaka, Japan), 20 mM EDC and 7.8 mM N-hydroxysuccinimide (Sigma-Aldrich).

Characterization of heparinized EC and growth factor release assay

The presence of heparin in collagen sponges was verified by toluidine blue staining.17 To quantify heparin crosslinking, sponge was immersed to 1 mL of 0.005% toluidine blue. After shaking, 2 mL of hexane was added. The absorbance of the aqueous phase was measured at 620 nm.18

A Hitachi S800 scanning electron microscope (Hitachi, Tokyo, Japan) was used to examine the morphology of the scaffolds. For in vitro enzymatic degradation of scaffold, EC and EC-heparin (EC-Hep) were incubated in 2 units/mL of collagenase and the residual mass was determined over time. The tensile strength of the scaffold was measured using universal testing machine (ST-1001, SALT, Korea) in accordance with ASTM D-882–83. The tensile strength was determined to be 3.62±1.83 n with a speed of 2 mm/min.

In vitro growth factor release profiles from the EC and EC-Hep were determined. Briefly, 200 μL of basic fibroblast growth factor (bFGF) or hepatocyte growth factor (HGF) solution (2 and 20 ng/mL, respectively; Koma Biotech, Seoul, Korea) was loaded on EC or EC-Hep. The sponges were soaked in 1 mL phosphate buffered saline (PBS) and incubated at 37°C. The entire incubation medium was collected after 4, 8, 24, 48 and 72 hours. After 72 hours of incubation, the remaining growth factors in EC or EC-Hep were obtained with 1 mL of PBS containing 1.5 M NaCl by vigorous shaking and squeezing of the sponges. The amount of bFGF or HGF was determined using ELISA (Koma Biotech).

Isolation of rat islets and hADSCs

Rat islets were isolated by digestion with collagenase type XI (Sigma-Aldrich) and cultured in Roswell Park Memorial Institute (RPMI) 1640 medium (Invitrogen, Carlsbad, California, USA) as previously described.19 Human adipose tissue was obtained from donors who were to receive abdominal surgery. hADSCs isolated from adipose tissue were cultured in Dulbecco’s modified Eagle’s medium (Invitrogen) as previously described by Zuk et al.20 The cells were classified as hADSCs on the basis of their adherence to plastic and expression of the surface markers CD45, CD73, CD90 and CD105 (BD Pharmingen, San Diego, California, USA). All experiments were performed using hADSCs after three to nine passages (online supplementary figure S1a, b).

Isolation of human islets

Pancreatic tissue collected with appropriate consent (online supplementary table S1) and ethical approval (IRB: 2019–0442) was used for islet isolation. Islets were isolated by the Ricordi method as previously described.21 Following isolation, islets were cultured in CMRL1066 medium (Invitrogen).

Islet and hADSC co-culture

To facilitate the coating of hADSCs to the islet surface, we used the previously reported method.22 Briefly, the hADSCs and islets in RPMI medium were co-incubated at 37°C for 1 hour in culture tubes and subsequently seeded. To compare the number of islets with different diameters and volumes, individual islets were mathematically converted to standard islet equivalents (IEQs) with a diameter of 150 µm. Islets (100 IEQs per well) were cultured with or without hADSCs (1×10⁴ cells) on various dishes for 7 days. Adhesion and shape of the hADSCs on the islet surface were analyzed using inverted microscopy, and viability was determined by calculating the percentage of viable (fluorescein diacetate (FDA)-positive, green, 0.5 µM) versus non-viable (propidium iodide (PI)-positive, red, 75 µM) cells within each islet.

Measurement of insulin release

Islets were stimulated with 1.67 or 16.7 mM glucose concentration in Krebs-Ringer buffer for 1 hour. The concentration of released insulin was analyzed using ELISA (Mercodia, Uppsala, Sweden).

Human growth factor array in the co-culture media

After incubation for 7 days, the conditioned medium was collected. RPMI medium served as the control. We measured 40 different growth factors using the Quantibody Human Growth Factor Array 1 (RayBiotech, Norcross, Georgia, USA). The data were analyzed using the Q-Analyzer (RayBiotech). The amount of each growth factor in the original extracts was measured and subtracted from that in the medium to calculate the amount of additional growth factor produced by the cells.

Transplantation of heparin-esterified collagen-hADSC-islet sheet

Diabetes was induced in BALB/c-nu mice by intraperitoneal injection of streptozotocin (Sigma-Aldrich, 250 mg/kg). Mice were considered to be diabetic when blood glucose (BG) levels were >300 mg/dL for at least two consecutive days. Mice were randomly divided into five experimental groups (n=5 in each group). Rat islets (4000 IEQ) alone and rat islets with ADSCs (1×10², 10³, 10⁴, 10⁵) loaded onto EC-Hep were transplanted in the mice’s dorsal subcutaneous region. For human islet transplantation, mice were divided into two experimental groups: 1) human islets (8000 IEQ) or 2) human islets with ADSCs (1×10⁶). Body weight and BG levels were monitored, and blood samples were obtained before transplantation and once a week afterward for measurements of insulin levels.

Intraperitoneal glucose tolerance test

After 6 hours of fasting, the mice received a glucose solution intraperitoneally (2 g/kg). BG levels were determined at 0, 5, 15, 30, 60 and 120 min after injection. Normal and mice with diabetes were used as controls.

Histological study

Retrieved grafts were fixed in 4% paraformaldehyde (Sigma). Samples were serially sectioned and sections were stained for H&E, CD31 (1:100; Abcam, Cambridge, Massachusetts, USA), alpha smooth muscle actin (αSMA, 1:200, Abcam) and insulin (1:200, Abcam). Imaging was acquired using inverted fluorescence microscopy and analyzed using ImageJ software.

For determination of β-cell mass, the total area and insulin positive area of each section was measured using ImageJ software.

Quantification of inflammatory and angiogenic factors using real-time PCR

To evaluate the expression of specific genes in the graft, mRNA levels were evaluated by real-time PCR. Following RNA extraction, as per the manufacturer’s protocol, cDNA was reverse transcribed using an Omniscript RT kit (Qiagen, Mississauga, Ontario, Canada). Real-time PCR was performed using SYBR Green Master Mix kit (Applied Biosystems, Foster City, California, USA). The sequences of used primers are shown in online supplementary table S2. Data were calculated using beta-actin as an internal reference.

Statistical analysis

Data are presented as means±SEM. A paired two-tailed t-test was applied for comparison between two groups or analysis of variance (ANOVA) with Tukey’s post hoc test (GraphPad Prism V.8, GraphPad Software, La Jolla, California, USA) was used when comparisons were made across more than two groups. A p value <0.05 indicated a statistically significant difference.

Fabrication and characterization of heparin immobilized collagen scaffold

The hypothetical mechanism of scaffold fabrication is schematically illustrated in figure 1A. hADSCs co-cultured with islets (figure 1B) can increase the expression of growth factors. These growth factors can be attached to esterified collagen crosslinked with heparin (EC-Hep; figure 1C), anticipating that this could promote angiogenesis by the local, sustainable release of the growth factors.

Request permission

To address the feasibility of this functional scaffold, the scaffolds were fabricated in a disc-form construct (figure 1D) and stained with toluidine blue to confirm the presence of heparin (figure 1E). After the native collagen (NC) and EC were crosslinked with heparin, the amount of heparin bound to the scaffolds was 4.85 and 9.43 mg heparin/g scaffold, for NC-Hep and EC-Hep, respectively (figure 1F). These results showed that a greater amount of heparin was bound to the EC than NC. The relative water solubility of EC produce a relatively larger density of the EC matrix compared with the NC matrix, and thus is attributed to the larger area that binds heparin.

Uniform pores with sizes between 60 and 300 µm in both scaffolds were clearly observed under the SEM (figure 1G). The physical properties of EC were also assessed to evaluate suitability. EC showed a dramatic increase in viscosity in the neutral buffer solution; the wettability of EC was 71±2.61 at the beginning before the slipping of water drops, and 46.17±4.41 after 2 min. In vitro degradation was assessed by measuring the weight loss; at 72 hours, the testing samples had degraded by 100% (figure 1H, left). The biocompatibility and biodegradability of the scaffolds were evaluated in vivo; the sponges had degraded 4 weeks later, and the mechanical strength was similar between the two scaffolds (figure 1H, right).

In vitro release curves of bFGF or HGF from EC and EC-Hep are depicted in figure 1I. The EC displayed a relatively rapid release profile for both bFGF and HGF. In contrast, a slower release of bFGF from EC-Hep was observed. The release profile of HGF was quite similar compared with that of bFGF. We checked the remaining amount of bFGF and HGF in the scaffold at 72 hours. The remaining amount of bFGF from EC-Hep and EC was 176.51±1.01 pg and 65.31±1.21 pg, respectively, and the remaining amount of HGF in EC and EC-Hep was 0.80±0.03 and 1.77±0.06 ng, respectively (figure 1I; online supplementary table S3). These results indicate that the interaction affinity between growth factors and EC-Hep could restrict the release of bFGF and HGF from EC-Hep.

Effect of heparin-esterified collagen-hADSC-islet sheet on islet viability and function

Rat islets were co-cultured with hADSCs, which adhered to the islet complex to form hADSC-rat islets (hA-rat islets). Morphological characteristics were clearly distinguishable. The hA-rat islets in EC-Hep featured more well-defined islet clusters than the rat islets alone (figure 2A–D). The co-culture of rat islets with hADSCs promotes the improvement of islets stability. Both rat islets and hA-rat islets were then loaded into EC or EC-Hep to evaluate long-term stability and morphological changes. Growth factors attached to EC-Hep provide an enriching microenvironment, and the paracrine effect helps the uptake of the islet-stabilizing growth factors. Thus, the function and viability of the complex were assumed to be enhanced. In this regard, cell viability, as determined by FDA/PI staining, was higher in hA-rat islets incubated in EC-Hep than in non-treated rat islets (figure 2E, left). Typically, in vitro culture of rat islets resulted in a tremendous loss of IEQs over time. However, the number of IEQs was remarkably preserved by hA-rat islet formation and heparin functionalization on the EC after 7 days of incubation (figure 2E, right).

Request permission

To further examine the effects of hADSC coating and heparin immobilization on the islet functionality, we evaluated the level of insulin secretion at 1.67 and 16.7 mM glucose. As shown in figure 2F, the amount of insulin released was higher in hA-rat islets than in the rat islets cultured alone. In addition, the heparin functionalization on the EC significantly increased glucose-stimulated insulin secretion. The findings indicated that co-culture with hADSCs improved islet secretory function in vitro. These beneficial effects on islet function could be partly attributed to heparin functionalization on the EC scaffold.

Effect of heparin-esterified collagen-hADSC-islet sheet on secretion of growth factors

To determine the growth factors enhanced by co-culture in EC-Hep, we used a commercially available antibody array. As shown in figure 2G,H, compared with culture with islet alone, hA-rat islet led to a marked increase in growth differentiation factor 15 (GDF-15), placenta growth factor (PIGF), and vascular endothelial growth factor (VEGF) levels. Of the growth factors assayed, the signal intensities of HGF, insulin-like growth factor binding protein-2 (IGFBP-2), osteoprotegerin (OPG), IGFBP-6, insulin, GDF-15 and VEGF were significantly greater in the heparin-esterified collagen-hADSC (HCA)-rat islet sheet than in other groups. Among them, increases in HGF, IGFBP-2, IGFBP-6 and OPG were markedly greater in the HCA-rat islet sheet than hA-rat islet or hA-rat islet-EC.

Glucose monitoring after transplantation of the heparin-esterified collagen-hADSC-islet sheet

To examine the impact of the HCA-rat islet sheet, we transplanted HCA-rat islet sheets into the subcutaneous sites (figure 3). The mice were transplanted with rat islets or rat islets containing hADSCs (1×10² to 1×10⁵ cells). EC only and 4000 IEQ rat islets only were used as controls and did not correct the hyperglycemia. Although rat islets containing 1×10² hADSCs displayed reduced BG levels compared with rat islets alone, the levels were still maintained in the range of 200–300 mg/dL. When the rat islets were transplanted with a higher number of hADSCs (1×10³ or 1×10⁴ cells), hyperglycemia was successfully reversed a week after transplantation. In addition, an increased number of hADSCs (1×10⁵ cells) resulted in a stable BG level at the normal range in the first day of transplantation, and the mice immediately recovered to normoglycemia (figure 3A).

Request permission

The observed dose-dependent improvement of the BG level control suggested a protective role of hADSCs. The immediate recurrence of hyperglycemia after removing the islet grafts further supported the crucial role of the transplanted rat islets (figure 3B). To evaluate the efficacy of rat islet engraftment, the intraperitoneal glucose tolerance test (IPGTT) was performed on euglycemic recipients. The BG level was significantly lower on glucose injection in the recipients of the HCA-rat islet sheet than in mice with diabetes (figure 3C). The area under the curve of IPGTT, representing the overall response to high glucose challenge, showed remarkable improvement in recipients of HCA-rat islet sheet (figure 3D). To assess the efficiency of our experiment in human islets, we transplanted HCA-human islet sheets. Isolated islets had typical features of the human islet, and functioned well on glucose stimulation (online supplementary figure S2a-d) as well as in vivo (online supplementary figure S3e-f). Based on the rodent islet results, a high number of hADSCs was chosen for human islet experimental group. Human islets and hADSCs loaded within the EC-Hep were transplanted. Graft function was evaluated by BG and body weight measurements (figure 3E-F). BG levels in mice transplanted with HCA-human islet sheet were significantly lower compared with BG levels in mice transplanted with human islet alone sheet. The HCA-human islet sheet transplant group displayed elevated serum human insulin levels compared with the human islet alone sheet (figure 3G). Results of glucose tolerance tests performed on day 40 were significantly better in recipients of hA-human islet than recipients of only human islets (figure 3H).

Histological finding after transplantation of heparin-esterified collagen-hADSC-islet sheet

Finally, we assessed whether the improved islet function was attributable to increased angiogenesis. To determine angiogenesis in the early post-transplantation phase, histological assessment was performed on day 7 after transplantation. To confirm the newly formed microvasculature, retrieved tissues were stained for vascular marker. The numbers of detectable vessels in the HCA-rat islet sheet were significantly higher than in the other groups (figure 4). Thus, addition of hADSCs induced effective early vascularization. The secretion of growth factors from embedded hADSCs was assumed to help induce angiogenesis, and was attributed to the rapid development of islet vasculature in the post-transplantation period. The vessel density was significantly higher in the hA-rat islets than the rat islets in EC-Hep, indicating that one of the beneficial effects provided by hADSCs is increased neovascularization of the graft.

Request permission

Furthermore, we surgically retrieved the implants and performed histological examinations to investigate the morphologies of the islet grafts. Immunohistochemical analysis confirmed the presence of insulin-stained islet grafts with well-preserved structure in the rat islet-hADSC (figure 5A) and human islet-hADSC (figure 5C) transplant group. The result indicated the benefit of hADSCs and EC-Hep in composing the islet sheet for the stable engraftment and glycemic control in the recipients.

Request permission

To assess the mechanisms underlying early graft loss, graft inflammation and vascularization were characterized by quantitative RT-PCR on day 7 after transplantation. F4/80 is a macrophage-specific marker and CD3 is expressed by T cells. Real-time PCR analysis for F4/80 and CD3 mRNA expression in the graft showed that F4/80 mRNA expression was slightly decreased in the cured group, compared with the non-cured group. However, CD3 mRNA expression was not expressed in both groups (figure 5B). In cured mice, interleukin (IL)-1β expression was lower, but the VEGF and PIGF mRNA expressions in the graft were significantly higher than in non-cured mice. Thus, the HCA-rat islet sheet may be beneficial for islet engraftment by preventing IL-1β-induced islet injury and promoting angiogenic ability.