Improved islet recovery and efficacy through co-culture and co-transplantation of islets with human adipose-derived mesenchymal stem cells

Islet transplantation is an established clinical procedure for select patients with type 1 diabetes and severe hypoglycemia to stabilize glycemic control. Post-transplant, substantial beta cell mass is lost, necessitating multiple donors to maintain euglycemia. A potential strategy to augment islet engraftment is the co-transplantation of islets with multipotent mesenchymal stem cells to capitalize upon their pro-angiogenic and anti-inflammatory properties. Herein, we examine the in vitro and in vivo effect of co-culturing murine islets with human adipose-derived mesenchymal stem cells (Ad-MSCs). Islets co-cultured with Ad-MSCs for 48 hours had decreased cell death, superior viability as measured by membrane integrity, improved glucose stimulated insulin secretion and reduced apoptosis compared to control islets. These observations were recapitulated with human islets, albeit tested in a limited capacity. Recipients of marginal mouse islet mass grafts, co-transplanted with Ad-MSCs without a co-culture period, did not reverse to normoglycemia as efficiently as islets alone. However, utilizing a 48-hour co-culture period, marginal mouse islets grafts with Ad-MSCs achieved a superior percent euglycemia rate when compared to islets cultured and transplanted alone. A co-culture period of human islets with human Ad-MSCs may have a clinical benefit improving engraftment outcomes.


Introduction
Islet transplantation is a therapeutic procedure that can restore endogenous insulin production and maintain euglycemia for a sustained period in patients with difficult to control type 1 diabetes mellitus (T1DM). The recent Clinical Islet Transplant Consortium's, National Institute of Health (NIH) sponsored phase 3 trial demonstrated islet transplantation's ability to stabilize glycemic control in select patients with T1DM presenting hypoglycemia unawareness where the primary end-point revealed 88% and 71% of recipients maintained euglycemia for 1-year and 2-years post-islet transplant respectively [1]. This Federal Drug Administration (FDA)-Biologics License Application enabling study may allow product licensure for islet transplant, facilitating reimbursement through insurance in the USA. Despite its apparent success, this procedure is not without limitations. A major challenge is overcoming suboptimal acute engraftment, where up to 60% of the initial transplanted islet mass is potentially lost due to innate instant blood-mediated inflammatory reaction (IBMIR), delayed re-vascularization, or hypoxic stress [2][3][4][5]. Multiple islet infusions are therefore often required to maintain periods of insulin independence. Co-transplantation of islets with multipotent stem cells (MSCs) is a potential strategy to mitigate early islet cell loss in culture and after transplantation [6,7].
MSCs are ubiquitous throughout cell types, and their capacity for self-renewal and differentiation into cells of mesoderm lineage includes adipocytes, chondrocytes, osteoblast and myocytes [8]. Previous studies have demonstrated the ability of MSCs to augment islet function, in part due to MSC's immunomodulatory and trophic properties, and their ability to secrete several paracrine factors [9][10][11]. Notably, MSCs modulate angiogenesis through gene expression of cytokines, including vascular endothelial growth factor (VEGF), fibroblast growth factors (FGFs), transforming growth factor-βs (TGF-βs) and Annexin-1 (ANXA1) [12][13][14]. MSCs can modulate the secretion of cytokines and promote the concentration of growth factors at the islet engraftment site and may aid neovascularization [15,16].
In the clinical islet allograft setting, islets undergo an obligate culture period of up to 72 hours before transplantation. The culture period facilitates recipient conditioning, and may allow for transplantation of a more immunologically quiescent graft [17,18]. Conversely, culture may be detrimental to islet survival due to limited nutrients, and exposure to oxidative, hypoxic, and inflammatory stressors [19]. These stressors lead to impaired islet viability and decreased cell mass. In addition, islet endothelial cells are compromised during the islet isolation and culture process that diminishes islet recovery and function prior to transplantation [20,21]. During isolation, islets are stripped of their native vascularization and rely on diffusion of nutrients and oxygen to survive. Moreover, the intra-islet endothelial cells rapidly decline to 5% by 4 days post culture [20]. The disrupted vascular supply hinders the post-transplant revascularization process. Over this period, insufficient vascularization causes increased cell death and graft failure due to inadequate nutrient delivery and prolonged ischemia [21]. In consequence, multiple donors and infusions are often required to maintain recipient insulin independence [19,22,23]. To decrease donor demand, it has been shown co-culturing islets with human bone marrow-derived MSCs or mouse adipose-derived MSCs (Ad-MSCs), for an extended period, can augment islet function and improve islet engraftment post-transplant [24][25][26].
In the present study, we explored the use of human Ad-MSCs co-cultured with murine and human islets for 48 hours at islet to Ad-MSC ratios of 1:300 [24] and 1:2000 [25]. We hypothesize co-culturing islets with human Ad-MSCs will improve islet recovery, islet function and augment engraftment outcomes.

Murine pancreatectomy and islet solation
Pancreatic islets were isolated from 8-to 12-week-old BALB/c mice (Jackson Laboratories, CA) and housed under conventional conditions in accordance with the Canadian Council on Animal Care. All experimental procedures were approved by the University of Alberta Research Ethics and Animal Use Committee (Study ID: AUP00000331). Prior to pancreatectomy, the common bile duct was cannulated and the pancreas was distended with 0.125 mg/ mL cold Liberase TL Research Grade enzyme (Roche Diagnostics, Laval, QC, CA) in Hanks balanced salt solution (Sigma, St. Louis, MO, USA). Pancreas digestion was continued in a 37˚C water bath for 14 minutes with light agitation. After the pancreatic digestion phase, islets were purified using histopaque-density gradient centrifugation (1.108, 1.083 and 1.069 g/mL, Sigma, St. Louis, MO, USA). Upon purification, islets were placed in Connaught Medical Research Laboratories (CMRL-1066) (Corning-cellgro, Manassas, VA, USA) supplemented with 10% fetal bovine serum, 1% L-glutamine (200 mmol/L, Sigma, St. Louis, MO, USA), 1% sodium pyruvate (100 mmol/L, Sigma, St. Louis, MO, USA), 1% non-essential amino acid 100x (Sigma, St. Louis, MO, USA) and 100 U/mL penicillin-G and100 μg/mL streptomycin (Sigma Aldrich Canada Co., Oakville, ON, CA) at 37˚C/5%CO 2 at pH 7.4.

Human islet isolation, purification and culture
Human islet preparations were isolated from deceased consenting multi-organ deceased donors, as described [19], with intent for clinical transplantation. Islets were only used for research when yields failed to meet that required for a listed recipient and with research consent from donor families. Permission for these studies was granted by the University of Alberta Health Research Ethics Board (REB), Edmonton, Alberta, Canada. The REB ensures that individual research projects involving human participants, identifiable data and/or human biological material meet the requirements of the current Tri-Council Policy Statement: Ethical Conduct of Research Involving Humans and University policy as well as provincial, federal and other legislation and regulations, as applicable.

Adipose derived mesenchymal stem cell expansion and culture conditions
To prepare the human adipose-derived MSCs, infrapatellar fat pad were removed from 2 donors (18 and 25 year old males) undergoing orthopedic knee surgery at the University Hospital of Alberta, Edmonton, Canada, and processed as described before [27]. The Ad-MSCs meet the International Society for Cellular Therapy guidelines [28] and were verified through chondrogenic differentiation after stimulation with TGFbeta and dexamethasone and CD cell surface markers [27]. Ethics committee waived the need for written informed consent of patients, as specimens used in this study were intended for discard in the normal course of the surgical procedure. All research involving human participants were reviewed and approved by the University of Alberta Health REB (Study ID: PRO 00001416 and PRO 000018778).

Percent islet recovery
Post isolation and subsequently post 48-hour culture, islets were harvested and counted to determine islet yield. Aliquots from respective samples were stained with dithizone (Sigma Aldrich Canada Co., Oakville, ON, CA) and counted in triplet. The percentage of islet recovery was determined by ratio of total islets harvested 48 hours' post-culture relative to the number of islets harvested immediately post-isolation.

Insulin secretory activity
Following isolation and post 48-hour culture, triplicates of 50 mouse and human islet equivalents were collected for respective groups and a static glucose-stimulated insulin secretion (s-GSIS) assay and dynamic insulin perfusion was performed. Islets were washed of residual glucose in glucose free medium, followed by incubation in RPMI-1640 (Sigma Aldrich Canada Co., Oakville, ON, CA) containing low (2.8 mmol/L) glucose, followed by high (16.7 mmol/L) glucose, each, for one hour at 37ºC. The supernatant was harvested and stored at 20ºC. Insulin levels were measured by an enzyme-linked immunosorbent assay (ELISA) (Mercodia, Uppsala, Sweden).

Apoptosis TUNEL staining
Apoptosis of islets was assessed quantitatively using Tdt-mediated dUTP nick-end labeling (TUNEL) staining. Prior and post-cultured islets were fixed in 4% paraformaldehyde, embedded in agar, processed and embedded in paraffin. Tissue sections were co-stained with antiinsulin antibody at 1:200 concentration (Agilent Technologies, Mississauga, ON, CA) and labeled with Rhodamine (TRITC) conjugated anti-guinea pig IgG (1:200, Jackson ImmunoResearch, West Grove, PA, USA). To identify the apoptosis, fluorescein isothiocyanate-dUTP with TdT enzyme (Promega, Madison, WI, USA) was added and counterstained with the nuclear stain, DAPI (ProLong Gold DAPI, Invitrogen, Calrsbadm CA, USA). Apoptosis was determined by analyzing the number of positive TUNEL-stained cells as a percentage of both insulin and nuclei positive cells utilizing FIJI ImageJ Software (National Institute of Health, USA).

Mesenchymal stem/stromal cell tracking
Following isolation, MSCs were stained with a vybrant DiO cell labeling dye (ThermoFisher Scientific, Waltham, MA) dosed at 5μl per 10 6 cells. Stained islets were distributed amongst respective groups of MSCs alone, 1:300 and 1:2000 islet to MSC ratios. Islets cultured alone did not contain MSCs. Microscopic evaluation was performed prior to culture (0 hours) and post culture (48 hours). MSCs microscopic evaluation and quantification was performed by FIJI ImageJ Software (National Institute of Health, USA).
The mice were transplanted with post-isolation or post-culture mouse islets with or without the presence of Ad-MSCs (islet to Ad-MSC ratio: 1:2000) or islets alone. Islets were aspirated into polyethylene (PE-50) tubing and centrifuged. A left lateral paralumbar subcostal incision was made and islets were delivered under the left kidney capsule. Islet engraftment was assessed through non-fasting blood glucose measurements, three times per week for 60 days. Blood glucose monitoring was conducted using a portable glucometer (OneTouch Ultra 2, LifeScan, CA, USA). Two consecutive readings maintained at <11.1 mmol/L confirmed graft function and reversal of diabetes. At day 60, a recovery nephrectomy of the grafted kidney confirmed graft dependent efficacy, when animals returned to hyperglycemia (�18 mmol/L).

Intraperitoneal glucose tolerance test (IPGTT)
In vivo glucose tolerance and islet function was assessed by an intraperitoneal glucose tolerance test (IPGTT) in euglycemic mice 6 weeks' post-transplant. The mice were fasted overnight and administered 25% dextrose intraperitoneally at a dose of 3 g/kg. Naïve normoglycemic mice served as controls. Blood glucose was monitored at baseline (t = 0), 15, 30, 60, 90 and 120 minutes' post injection. Significance was measured as area under the curve amongst groups.

Islet graft assessment
Sixty days' post islet transplant, kidney bearing grafts were removed and placed in -80˚C. Kidneys were homogenized and sonicated with 2mM of acetic acid in 0.5% of bovine serum albumin (Thermo Fisher Scientific, Waltham, MA, USA). Supernatant was collected and insulin levels were measured by an enzyme-linked immunosorbent assay (ELISA) (Mercodia, Uppsala, Sweden).
Immunohistochemistry was performed on a subset of kidneys bearing the islet grafts post 60 days' post islet transplant and which were fixed in 10% formalin. The tissue was sectioned and upon heat retrieval, were blocked with 20% goat serum (Sigma-Aldrich, St. Louis, MI, USA), and subsequently incubated overnight at 4˚C primary antibodies guinea pig anti-insulin

Statistical analysis
Data were analyzed using GraphPad Prism (GraphPad Software, La Jolla, CA, USA). Student t test or analysis of variance (ANOVA) was used to compare islet yield, membrane integrity, TUNEL, sGSIS and AUC for IPGTT's and represented as scatter plots. Kaplan-Meier graft survival function curves were compared using the log-rank (Mantel-Cox) statistical method. Statistical significance was considered when p-values < 0.05. Graphical representation of data is as mean ± standard deviation (s.d) of small in vitro data sets, and standard error of the mean (s.e.m) for in vivo analysis where � p < 0.05, �� p < 0.01, ��� p < 0.001, ���� p < 0.0001.

Pro inflammatory cytokine expression after islet transplant
Post 48-hour culture media was collected and analyzed for murine and human pro-inflammatory cytokines (S1 Fig). Murine interleukin (IL)-2p70 expression was reduced by the presence of Ad-MSCs during culture, whereas islets cultured alone revealed increased expression (p<0.05, S1B Fig). Human and mouse IL-1β, IL-6, IL-10, and tumor necrosis factor (TNF)-α were detected but their expression amongst respective groups did not provide statistical significance (islets vs. islets +Ad-MSCs) (S1A, S1C and S1E Fig).

Co-transplantation of islets and Ad-MSCs without prior co-culture is detrimental to engraftment
To determine if a period of islet-MSC co-culture was needed, we co-transplanted islets + Ad-MSCs without prior culture. One hundred and fifty mouse islets (marginal islet mass) were isolated and co-transplanted under diabetic murine kidney capsule with or without 1:2000 Ad-MSCs combined at the time of transplant. Ten of 13 (76.9%) control islet recipients without MSCs became euglycemic at 21.0 ± 6.4 days after transplant, whereas 7 of 13 (53.8%, p<0.001) islets + Ad-MSCs became euglycemic by 38.6 ± 6.1 days (Fig 3A and 3B). All euglycemic animals became hyperglycemic after graft recovery nephrectomy 60 days after transplant, confirming graft dependent euglycemia. Seven weeks' post-transplant, an intraperitoneal glucose tolerance test (IPGTT) was performed on all euglycemic mice (Fig 3C) and demonstrated no significance amongst the mean area under the curve (Fig 3D).

Co-culture for 48 hours of islets and Ad-MSCs followed by cotransplantation improves engraftment
Post-isolation, islet aliquots were placed into non-adherent petri dishes and co-cultured with or without the presence of Ad-MSCs for 48 hours (1:2000 islet to Ad-MSC ratio, 48hrs co-culture period) and subsequently transplanted into diabetic mice under the kidney capsule (200 islets per recipient). Mice co-transplanted with Ad-MSCs (1:2000 islet to Ad-MSC ratio) reversed to euglycemia at a faster rate (22.3 ± 4.7 days) than the islets alone group (38.5 ± 7.6 days) (Fig 4A and 4B). As well, percent engraftment was improved in those mice transplanted with islets co-cultured with 1:2000 Ad-MSCs (islets: 9 out of 19 (47%) vs. 1:2000: 18 out of 21 (86%)) (p<0.05) (Fig 4A and 4B). Sixty days post-transplant, all euglycemic mice returned to hyperglycemia post recovery nephrectomy. IPGTTs were performed on all euglycemic recipients 7 weeks post-transplant (Fig 4C) and there was no significant difference for area under the curve amongst respective groups (Fig 4D).

Discussion
In this study, we show that co-culturing mouse islets for 48 hours with human Ad-MSCs improves islet function and efficacy relative to islets cultured alone. Furthermore, we demonstrated at a limited capacity human islets co-cultured with Ad-MSCs may improve in vitro islet recovery and function. Collectively, these findings provide a fundamental basis for applications within a clinical setting. Clinical islet transplantation is restricted currently to patients with brittle T1DM that cannot be stabilized by alternative measures. Multiple intraportal islet infusions are often required to sustain euglycemia. Human islet loss during culture and intraportal transplantation is substantial, related in part to IBMIR, hypoxia, apoptosis, and other inflammatory or immune activating events [30][31][32][33]. Supplementation of islets with additives during culture may support islet function, morphology, and vitality [34][35][36]. It has been shown previously that human and mouse islet function can be maintained for up to 30 days after co-culture with MSCs [37,38]. During the culture period, high doses of mouse derived MSCs and islets adhere to the edges of islets and can penetrate the islet core to form MSC-islet composites [26]. Consequentially, most studies evaluate the utilization of bone marrow derived MSCs and limited research evaluates the use of Ad-MSCs [39,40]. Studies utilizing Ad-MSCs have shown improved islet function and reduced islet mass to reverse diabetes in rodent models [41][42][43].
Herein, this study examined the supplementation of mouse or human islets with humanderived Ad-MSCs during a 48-hour co-culture period to evaluate their potential clinical relevance and therapeutic benefits for human islet survival and engraftment. Our in vitro observations demonstrated the ability to increase mouse islet yield, maintain vitality, cell survival and insulin secretion with the supplementation of Ad-MSCs during coculture relative to islets cultured alone. Our observations are potentially applicable for future translation to the clinic as we demonstrated the capacity of human Ad-MSCs to also improve human islet in vitro potency and it can further investigate in vivo through the Rag -/model. Our findings support previous literature where human islets with human-derived Ad-MSCs improved islet function in vitro after 72-hour culture on an engineered cell sheet [44] and improved islet efficacy in mice by pre-48-hour-hypoxic cultured human-derived Ad-MSCs intraperitoneally injected into mice [45]. Likewise, rat islets following a 72-hour post-culture with human-derived Ad-MSCs had improved islet engraftment upon intraportal co-transplantation in rats and [46]. Direct static co-culture with MSCs has the potential to release local cytokines, including VEGF and ANAX1, to promote cell survival [13,14,47]. We did not measure such trophic factors but in an effort to comprehend and analyze the activity of Ad-MSCs Islets co-cultured with mesenchymal stem cells we collected media after 48h co-culture to quantify proinflammatory expression [48][49][50]. Mouse and human cytokine profiles was neither elevated nor down regulated, except for murine IL2p70. IL2p70 is a lymphocyte membrane glycoprotein (p70, IL-2RP) which aids IL-2 recruitment of lymphocytes and natural killer activity to the islet engraftment [51,52]. The down regulation of IL2p70 may be potentially advantageous in dampening early immune responses. From these data, we interpret that MSCs are not harmful for in vitro islet survival and function, and the ability to downregulate pro-inflammatory cytokines could be dependent on in vivo function [3,53].
Based on our in vitro observations, the 1:2000 islet to Ad-MSC ratio was used for in vivo transplantations due to their superior in vitro islet function. Without a culture period, we utilized a marginal mass dose of 150 islets, and attempted to transplant 150 islets post 48-hour culture, but due to the islet loss and subsequent decreased function, all mice transplanted maintained hyperglycemia and had to be euthanized prior to the end-point. We increased the dosage post culture to 200 islets per recipient to circumvent hyperglycemia to evaluate the islet engraftment for up to 60 days. Although islet mass transplanted was different, we have shown improved graft efficacy when islets were co-cultured with Ad-MSCs relative to islets co-transplanted with Ad-MSCs immediately after isolation without prior co-culture. In this study, as a result of the islet loss in the control group compared to the islet and Ad-MSC group, the number of donor pancreata in the control group had to be increased to ensure that all groups were transplanted with a similar islet mass after 48-hour culture. We speculate improved graft function were due to Ad-MSCs formation and trophic effects during a prolonged direct cell contact of islets and Ad-MSC. The state of the Ad-MSCs prior to culture are single cells, and when transplanted MSCs could migrate away from the transplant site and therefore may not be as beneficial for engraftment. During a co-culture period, Ad-MSCs adhere to islets and one another and are more likely to remain at the engraftment site thereby increasing the probability for their trophic factors to help graft [54]. Extended culture times can reduce exocrine tissue and endotoxin carry-over [53] which may account for our observed marked improvement in islet potency after co-culture. We and others have demonstrated that MSCs can decrease islet loss, maintain islet morphology and function during co-culture with islets compared to culture of islets alone [24]. These results should be viewed with caution because early time points during in vivo assessment were not conducted and it is unclear whether MSCs benefits are derived only within the first days after engraftment, or have a prolong durable effect.
These findings are potentially readily translatable to the clinical setting, where human islets routinely undergo up to 72 hours of culture. Addition of human Ad-MSCs manufactured under good manufacturing conditions (GMP) during this period could be highly protective, and may decrease the need for multiple donors and risk of Human Leukocyte Antigen sensitization. Autologous MSCs offer potential promise in mitigating allo-rejection, perhaps allowing transplants to proceed with less immunosuppression and therefore less potential risk. To further elucidate the cytoprotective capacity of human Ad-MSCs our future research efforts will further explore the impact that co-culture on human islet function and engraftment.
Herein, we highlight a potential translatable strategy to protect islets in culture and early after transplantation that can increase the frequency of donor engraftment rates with the supplementation of Ad-MSC during culture.
Supporting information S1 Fig. Murine and human proinflammatory cytokine analysis. Murine and human proinflammatory profiles from the medium after 48-hour culture (n = 3). Mouse cytokines are displayed on the left y-axis in blue and human cytokines on the right y-axis in red. Upon analysis, (b) murine IL2p70 expression was significantly downregulated for islets co-cultured with Ad-MSCs (1:2000) compared to control islets ( � p<0.05, ANOVA) (mean ±s.e.m). IL-1β, IL-6, IL-10, and tumor necrosis factor (TNF)-α expression was detected, but no significance difference was demonstrated amongst groups (p>0.05, ANOVA). Data represented (mean ± s.d).