Enhancing Human Islet Transplantation by Localized Release of Trophic Factors From PLG Scaffolds

Abstract

Islet transplantation represents a potential cure for type 1 diabetes, yet the clinical approach of intrahepatic delivery is limited by the microenvironment. Microporous scaffolds enable extrahepatic transplantation, and the microenvironment can be designed to enhance islet engraftment and function. We investigated localized trophic factor delivery in a xenogeneic human islet to mouse model of islet transplantation. Double emulsion microspheres containing exendin-4 (Ex4) or insulin-like growth factor-1 (IGF-1) were incorporated into a layered scaffold design consisting of porous outer layers for islet transplantation and a center layer for sustained factor release. Protein encapsulation and release were dependent on both the polymer concentration and the identity of the protein. Proteins retained bioactivity upon release from scaffolds in vitro. A minimal human islet mass transplanted on Ex4-releasing scaffolds demonstrated significant improvement and prolongation of graft function relative to blank scaffolds carrying no protein, and the release profile significantly impacted the duration over which the graft functioned. Ex4-releasing scaffolds enabled better glycemic control in animals subjected to an intraperitoneal glucose tolerance test. Scaffolds releasing IGF-1 lowered blood glucose levels, yet the reduction was insufficient to achieve euglycemia. Ex4-delivering scaffolds provide an extrahepatic transplantation site for modulating the islet microenvironment to enhance islet function posttransplant.

Introduction

Human islet transplantation has the potential to be an effective treatment for type 1 diabetes (1,2). Despite advances in glucose monitoring and therapeutics, chronic micro- and macro-vascular complications and treatment-induced hypoglycemia are major sources of morbidity and mortality in diabetic patients, and these complications are frequently delayed rather than prevented by insulin therapy (3). Human islet transplantation as a beta cell replacement strategy reached a major milestone in 2000 when seven recipients achieved insulin independence following intra-hepatic portal infusion of islets from more than one donor pancreas with corticosteroid-free immunosuppression (1). Unfortunately, this clinical procedure has resulted in only 44% of patients achieving insulin independence after 3 years (4). Factors that have been implicated in islet failure include the instant blood-mediated inflammatory reaction, foreign hepatic environment, chronic exposure to low oxygenation levels and close proximity to metabolites (5–7). Challenges with hepatic delivery have motivated the development of alternative strategies for controlling the islet microenvironment to promote enhanced function and decrease the number of islets required (8,9). The transplantation site must have several basic properties, such as appropriate oxygen tension, a vascular supply, ability for glycemic detection and insulin delivery, a minimally invasive procedure, and compatibility with immunosuppressive or tolerance strategies (8). Further, treatment of islets with growth factors pretransplantation or via systemic delivery has been proposed as another means to enhance islet function posttransplant (10,11).

Herein, we investigate the localized delivery of exendin-4 (Ex4) or insulin-like growth factor-1 (IGF-1) at the implantation site as a means to enhance islet engraftment and function. Ex4 is a potent and long-acting glucagon-like peptide-1 (GLP-1) receptor agonist that stimulates glucose-dependent insulin secretion and islet cell replication, protects against apoptosis and improves outcomes in islet transplant models (12–17). IGF-1 is a growth factor that decreases apoptosis (18,19), stimulates cellular growth and proliferation, and is involved in regulating islet beta cell mass, proliferation and regeneration (19–21). Biomaterial scaffolds are employed for transplanting islets, functioning as a (i) support for islets that distributes islets throughout the implantation site and allows islet revascularization, and (ii) vehicle for the localized delivery of trophic factors to facilitate engraftment and reduce the number of transplanted islets needed to restore euglycemia. Localized delivery to the islet environment can maintain therapeutic concentrations of these factors while maintaining low levels systemically, which may limit unintended side effects. Microporous scaffolds were fabricated using a layered approach consisting of porous outer layers designed for islet transplantation and a central layer for local protein delivery (22). The scaffolds were fabricated from poly(lactide-co-glycolide) (PLG), a biodegradable material used for numerous applications including localized delivery lasting from days to weeks (22–26), and we characterized the loading, release and bioactivity of these proteins. We have previously reported that human islets were able to restore glycemic control following transplantation on PLG scaffolds (27), and this study investigates the ability of trophic factor delivery to enhance islet survival and function.

Materials and Methods

Protein encapsulation in PLG microspheres

Ex4 (Anaspec, Fremont, CA) or IGF-1 (Shenandoah Biotechnology, Warwick, PA) was encapsulated into PLG microspheres using a double emulsion technique previously described (28). Two concentrations of PLG (3% or 6% w/w) and ratios of the lactide:glycolide polymers (50:50, 75:25 or a blend) were tested. Encapsulation efficiency was determined with ¹²⁵I-labeled proteins.

Layered scaffold fabrication and protein release kinetics

Scaffolds (5mm diameter, 3mm height) were made with a protein-loaded center layer sandwiched between two porous outer layers. The outer layers were fabricated from nonprotein-loaded microspheres as previously reported with single-layer scaffolds (29–31). To determine release kinetics, layered scaffolds were incubated at 37°C in phosphate buffered saline (PBS). At multiple time points, scaffolds were transferred to fresh PBS and ¹²⁵I-labeled protein in the release buffer was determined.

Scanning electron microscopy

Structural characteristics of scaffolds were imaged with a scanning electron microscope (Hitachi SU8030 FE SEM; Hitachi High-Technologies Corp., Tokyo, Japan). A 15-nm osmium coating was applied and the microscope was operated at 5 kV.

Ex4 bioactivity assay

The bioactivity of released Ex4 was determined using rat insulinoma cells (RINm5f) exposed to conditioned or supplemented media; insulin release was quantified by ELISA (Millipore, Billerica, MA). RINm5f cells (ATCC#CRL-11605; Manassas, VA) were plated at 4.0 × 10⁵ cells/well in a 48-well plate in RPMI 1640, 10% fetal bovine serum (heat inactivated), and penicillin/streptomycin. Cells were treated with conditioned or control media for 1 day. Scaffold-treated conditioned medium was prepared by immersing Ex4 or blank scaffolds in media for 8 days at 37°C. These conditions were compared to media supplemented with 100nM Ex4 or media containing no protein. After 1 day of culture with conditioned media, the media was collected and assessed for insulin concentration with a rat insulin ELISA.

IGF-1 bioactivity assay

The bioactivity of IGF-1 released from scaffolds was determined using INS-1 cells exposed to conditioned or supplemented media and assayed for phospho-ERK. INS-1 cells (ATCC #30-2001) were plated at 1.7 × 10⁶ cells/60-mm dish in RPMI 1640,10% fetal bovine serum (heat inactivated), 2mM glutamine, penicillin/streptomycin and 50 µM β-mercaptoethanol. Cells were treated with scaffold-treated conditioned medium, IGF-1 supplemented medium or control low serum medium. Scaffold-treated conditioned medium was prepared by immersing IGF-1 scaffolds in medium for 8 days at 37°C. This conditioned medium and conditioned medium diluted 1:1 with control medium were compared with medium supplemented with IGF-1 (50 ng/mL). Cells were recovered in RIPA buffer with protease and phosphatase inhibitors. Western blots were prepared by loading 20 µg of total protein per lane, transferring to Immobilon (Millipore) nylon membranes and hybridizing to anti-phospho-ERK and anti-ERK antibodies (#9101 and #9102; Cell Signaling, Danvers, MA) following manufacturer’s instructions. The blot was stained with anti-phospho-ERK antibodies, stripped and reprobed with anti-ERK antibodies.

Human islet acquisition, culture and assays

Human islets were obtained from eight shipments through the Integrated Islet Distribution Program (IIDP). The donor age was 49.4 years with a standard deviation of 12.1 years, with six male and two female donors. After overnight shipment, human islets were washed in CMRL 1066 media (MM1 type) (Mediatech, Manassas, VA) supplemented with 2.5% human AB serum (Mediatech). Islets were cultured at 27°C for up to 6 days prior to transplantation in media supplemented with 10% serum changed every 48 h. On the day of transplant, islets were assayed for viability as measured by ATP/DNA (moles/g) by performing plate-based assays on triplicate aliquots for ATP with CellTiter-Glo (Promega, Madison, WI) kit and DNA with Quant-iT picogreen (Invitrogen, Grand Island, NY) kit (32). Islet equivalents (IEQs) were calculated based on 10.5 ng DNA/IEQ. Total IEQ per shipment was adjusted by purity determined by hand counting of dithizone (Sigma–Aldrich, St. Louis, MO) stained aliquots.

Islet transplantation

Male NOD-scid IL2Rgamma^null (NSG; Jackson Laboratories, Bar Harbor, ME) mice (12–16 weeks) were used as transplant recipients. Mice were rendered diabetic 5–6 days pretransplantation by intraperitoneal injection of 130 mg/kg streptozotocin (Sigma–Aldrich) in citric acid and sodium citrate buffer. Nonfasting blood glucose was measured using tail blood and a glucometer (Accucheck Aviva; Roche Diagnostics, Indianapolis, IN). Diabetic recipients exhibited blood glucose levels greater than 300 mg/dL on 2 consecutive days pretransplant. Mice were anesthetized by intraperitoneal ketamine (Ketaset^®; Fort Dodge Animal Health, Fort Dodge, IA) and xylazine (Anased^®; Lloyd Inc., Shenandoah, IA) injection. Layered scaffolds were seeded with islets using a custom-made micropipetting device and immediately wrapped in each epididymal fat pad. Based on the surface area of the scaffold, the seeding density ranges from 3700 to 5000 IEQ/cm², though the islets are distributed throughout the outer 500 µm of the scaffold (27,33). Recipients were followed by nonfasting blood glucose every 1–2 days for at least 30 days posttransplant. An intraperitoneal glucose tolerance test (IPGTT) was performed to assess graft function. A dextrose solution (50%, 2 g/kg body weight; Hospira, Lake Forest, IL) was administered intraperitoneally to fasting mice. Blood glucose was measured at time points after injection. Murine studies were approved by the Northwestern University Animal Care and Use Committees.

Graft histology, quantification and analysis

Grafts from mice transplanted with 3% Ex4 and blank scaffolds were explanted at Day 30, fixed and paraffin-embedded. Sections (5 µm) were stained with hematoxylin and eosin (H&E) to identify islets, after which adjacent sections were stained for insulin/Ki67 with a Hoechst nuclear counterstain as previously described (33). Stained slides were imaged using Leica DMIRB inverted fluorescence microscope (Buffalo Grove, IL) and analyzed with NIH ImageJ (http://rsb.info.nih.gov/ij/). For islet area, the pixel areas of insulin-positive tissue in nonadjacent sections of each graft were quantified. Ki67+ cells were counted and compared to nuclei number within insulin-positive areas.

Statistics

Statistical analyses were done using the statistical package Graphpad Prism (Graphpad, La Jolla, CA). Results are presented as mean ± standard error of the mean (SEM) in all figures. Student’s t-test, paired t-test or one-way analysis of variance were used to determine statistical significance of groups. The log-rank statistic was used for comparison of graft function in Kaplan–Meier analysis between groups. A probability (p) less than 0.05 was considered statistically significant.

Results

Protein loading and release

Encapsulation and release of Ex4 and IGF-1 were initially investigated as a function of the polymer (either a 50:50 or 75:25 ratio of lactide:glycolide, or a blend of the two polymers) and polymer concentration (3%, 6%). The encapsulation efficiencies measured for Ex4 and IGF-1 ranged from 58% to 95% and 11% to 48%, respectively (Table 1). The protein-loaded microspheres were used to form the central layer of the scaffolds (Figure 1A – C).

Table 1.

PLG encapsulates proteins in double emulsion polymer microspheres

PLG designation	3% PLG	6% PLG	3% PLG blend	6% PLG blend
Encapsulation efficiency (%)
Ex4	95.3 ± 2.91	79.7 ±10.0	58.0 ±9.8	63.9 ±1.9
IGF-1	11.4±0.92	48.1 ±3.1	26.7±3.2	NP

Ex4, exendin-4; IGF-1, insulin-like growth factor-1; NP, not performed; PLG, poly(lactide-co-glycolide).

Three PLG formulations with different lactide to glycolide ratios and PLG concentrations in the emulsion phase were used to encapsulate ¹²⁵I-labeled Ex4 and IGF-1.

¹A 50:50 lactide to glycolide polymer ratio.

²A 75:25 lactide to glycolide ratio.

Figure 1. Layered PLG scaffold.

(A) Cross-sectional view of a layered scaffold (5mm diameter × 3 mm height) demonstrating the outer layers molded around a smaller solid inner layer that does not span the width of the scaffold. (B) Scanning electron micrograph of a layered scaffold cross-sectional view. Outer layers were formed with single emulsion microspheres and NaCl particles to create 250–425 µm pores. Protein-delivering double emulsion microspheres form the center layer. (C) Magnified view showing the porous outer layer for islet seeding and center layer containing protein-releasing microspheres. PLG, poly(lactide-co-glycolide).

The quantity of protein released from the central layer of scaffolds was dependent upon the polymer formulation. Ex4 exhibited a bimodal release, with an initial release within the first 2 weeks, and a second, slower release occurring through 10 weeks (Figure 2A). IGF-1 exhibited greater initial release, where the majority of protein release occurred in the first 12 days (Figure 2B). Ex4 and IGF-1 were encapsulated in microspheres composed of 6% 50:50 PLG, as these conditions demonstrated a sustained release profile, a large amount of total protein released and high encapsulation efficiencies for each protein (80% and 48% for Ex4 and IGF-1, respectively).

Figure 2. Protein release from PLG scaffolds.

Layered scaffolds were fabricated with center layers of microspheres loaded with either (A) Ex4 or (B) IGF-1. The release rate was monitored using radiolabeled proteins (n = 3 per condition). Ex4, exendin-4; IGF-1, insulin-like growth factor-1; PLG, poly(lactide-co-glycolide).

Bioactivity of encapsulated proteins

Bioactivity of the released proteins was assessed using bioassays based on cellular responses. Ex4 released from layered scaffolds significantly increased insulin secretion from RINm5f cells similar to treatment with Ex4 peptide, which had not been previously encapsulated (p < 0.05) (Figure 3A). IGF-1 bioactivity was determined by its ability to stimulate ERK phosphorylation as determined by Western blot analysis for phospho-ERK. INS-1 cells treated with medium previously incubated with an IGF-1-containing scaffold increased P-ERK levels similar to native IGF-1 (Figure 3B). These results demonstrate that protein released from layered scaffolds retains its bioactivity compared with unencapsulated control protein.

Figure 3. Ex4 and IGF-1 bioactivity following release.

(A) Rat insulinoma (RINm5f) β cells were treated for 1 day with conditioned medium previously incubated with Ex4 loaded scaffolds for 8 days. Insulin release from the cells was quantified by ELISA. Controls included medium incubated with scaffolds without Ex4 or medium alone. Error bars represent SEM and *p<0.05. Differences between the Ex4 scaffold and Ex4 conditions or the blank scaffold and control conditions were not significant. (B) INS-1 cells were treated for 1 day with conditioned medium previously incubated with IGF-1 loaded scaffolds for 8 days. P-ERK was evaluated by Western blot analysis. Control conditions included cells prior to treatment (time Oh) or cells treated with control medium from incubation with scaffolds without encapsulated IGF-1, which are denoted as controls A and B, respectively. Ex4, exendin-4; IGF-1, insulin-like growth factor-1.

IGF-1 delivery with islet transplantation

Subsequent studies investigated the ability of protein delivery to enhance graft function following transplantation of human islets on layered scaffolds. Human islets received from two IIDP shipments were transplanted into NSG mice on IGF-1 6% 50:50 scaffolds (herein referred to as 6% IGF-1). Human islet viability, evaluated through the ATP to DNA ratio, ranged from 5.24 to 8.26 nM/ng with a purity range of 37–92%, consistent with previous reports (27). Mice transplanted with 2000 IEQ maintained normal glucose levels for more than 140 days, and graft removal caused hyperglycemia, confirming that euglycemia was due to human islet transplantation (27). In this study, a mass of 1500–2000 IEQ was transplanted, which was not sufficient to restore euglycemia without an alternative intervention. Blood glucose levels in mice transplanted with IGF-1 loaded scaffolds were significantly lower compared to blood glucose levels in mice transplanted with blank scaffolds (n = 8 each). However, glucose levels were not sufficiently decreased to restore euglycemia (Figure 4A and B). Mice transplanted with IGF-1 scaffolds without islets did not demonstrate any decline in blood glucose levels, suggesting that systemic effects of IGF-1 release could not account for the observed reduction in mice transplanted with islets on IGF-1-releasing scaffolds (Figure S1).

Figure 4. Graft function for human islets transplanted on 6% IGF-1 loaded scaffolds.

Blood glucose was measured over time from mice transplanted with 1500–2000 islet equivalents on (A) 6% 50:50 scaffolds containing IGF-1 (n = 8) and blank scaffolds (n = 8), *p<0.05. (B) A Kaplan–Meier plot represents animals with functioning grafts, where NS indicates “not significant.” Graft failure is defined as three consecutive glucose measurements greater than 250 mg/dL. IGF-1, insulin-like growth factor-1.

Ex4 delivery with islet transplantation

Human islets from four IIDP shipments were transplanted into NSG mice on Ex46% 50:50 scaffolds (6% Ex4). Human islet viability was assessed through the ATP to DNA ratio, with values ranging from 3.17 to 6.37 nM/ng and the purity of the shipment ranging from 62% to 75%. Mice receiving 6% Ex4 scaffolds displayed decreased blood glucose levels compared to mice receiving blank scaffolds (Figure 5A). For control mice, the percentage of animals that achieved euglycemia was approximately 35% by Day 10 and remained at that level through 30 days, where graft failure was defined as three consecutive glucose measurements greater than 250 mg/dL. However, in the presence of Ex4, the percentage of animals that achieved euglycemia was significantly higher at 75% compared to blank scaffolds at Day 10 (p < 0.05, Figure 5B), with a gradual decline to 50% by Day 30.

Figure 5. Graft function for human islets transplanted on 6% Ex4 loaded scaffolds.

Blood glucose was measured over time from mice transplanted with 1500–2000 islet equivalents on (A) 6% 50:50 scaffolds containing Ex4 (n = 19) and blank scaffolds (n = 17). (B) A Kaplan–Meier plot demonstrating graft function for the 6% 50:50 Ex4 scaffolds and blank scaffolds. Graft failure is defined as three consecutive glucose measurements greaterthan 250 mg/dL, *p<0.05. Ex4, exendin-4.

We subsequently investigated whether the release profile of Ex4 from the scaffold could extend the duration during which islet function was enhanced. Scaffolds formed with 3% 50:50 PLG Ex4-loaded microspheres in the center layer (3% Ex4) were selected, as the encapsulation efficiency (95%) was increased relative to 6% Ex4, and a greater amount of protein was released in the first 40 days compared to 6% Ex4 scaffolds (Figure 2A). Human islets from two IIDP shipments were transplanted on 3% Ex4 scaffolds, with ATP to DNA ratios ranging from 4.15 to 6.06 nM/ng and purities from 70% to 80%. The impact of Ex4 in reducing blood glucose levels was more pronounced when islets were transplanted on the 3% scaffolds (Figure 6A). For the 3% Ex4 scaffolds, all mice were euglycemic at Day 10, which contrasted with 20% of mice transplanted with control scaffolds. On Day 30 posttransplant, ~60% of mice transplanted with 3% Ex4 scaffolds were euglycemic compared with ~20% of control mice (p<0.01, Figure 6B). Removal of 3% Ex4 grafts at Day 15 posttransplant reversed euglycemia, indicating the transplanted islets maintained euglycemia (Figure S2). Mice transplanted with Ex4 scaffolds without islets had a transient decline in blood glucose levels, yet none of these mice achieved euglycemia, indicating that systemic effects of Ex4 did not account for the euglycemia observed with islets transplanted on protein-releasing scaffolds (Figure S1).

Figure 6. Graft function for human islets transplanted on 3% Ex4 loaded scaffolds.

Blood glucose was measured over time from mice transplanted with 1500–1800 islet equivalents on (A) 3% 50:50 scaffolds containing Ex4(n = 10) and blank scaffolds (n = 10). (B) A Kaplan-Meier plot demonstrating graft function for the 3% 50:50 Ex4 scaffolds and blank scaffolds. Graft failure is defined as three consecutive glucose measurements greaterthan 250 mg/dL, **p<0.01. Ex4, exendin-4.

An IPGTT was performed on these mice to further investigate the function of islets transplanted on 3% Ex4 scaffolds compared with blank scaffolds or control mice. Mice receiving a sub-minimal islet mass were challenged 11 days after transplant. Mice in the 3% Ex4 scaffold group demonstrated lower blood glucose levels (Figure 7A) and a smaller area under the curve (AUC) for glucose (Figure 7B) compared with the blank scaffold group. The AUC was similar to that in control mice.

Figure 7. Intraperitoneal glucose tolerance test for human islets transplanted on 3% Ex4 scaffolds.

Blood glucose measurements (A) and area under the curve (AUC) (***p< 0.001) (B) for naïve mice (n = 6) or mice receiving Ex4 scaffolds (n = 4) or blank scaffolds (n = 5) following challenge with an intraperitoneal dextrose injection 11 days posttransplant. Ex4, exendin-4.

Histology

Islets on scaffolds releasing Ex4 were characterized using histological techniques. H&E staining demonstrated integration of the islets with the host tissue, with no signs of fibrosis that may impact islet function (Figure S3). Islet area was significantly greater in mice receiving 3% Ex4 scaffolds compared with blank scaffolds, indicating that Ex4 acts to enhance islet survival posttransplant (Figure 8A – C). As determined by co-localization of Ki67 and nuclear staining, mice transplanted with 3% Ex4 scaffolds had a higher ratio of proliferating islet cells (Ki67+) compared with blank scaffolds (Figure 8D – F).

Figure 8. Human islets transplanted on 3% Ex4 scaffolds have a greater islet area and demonstrate more proliferation.

Quantification of islet area in 3% Ex4 scaffolds (A) shows a greater area than in blank scaffolds (B), **p<0.01 (n>40 islets) (C). Islets transplanted on 3% Ex4 scaffolds (D) show more proliferating nuclei (white arrows) within islets compared to blank scaffolds (E). Within insulin-positive areas, the ratio of proliferating cells (Ki67+) to nuclei is greater in 3% Ex4 scaffolds than in blank scaffolds (F), ***p < 0.001 (n >35 islets).

Discussion

Microporous biodegradable PLG scaffolds have been highly effective in syngeneic, allogeneic and xenogeneic islet transplantation models (29–31). We have now adapted these scaffolds for localized protein delivery. Most reports of biomaterials for islet transplantation have focused on immunoisolation to protect islets from an allo- or auto-immune response. Our work has focused on the use of microporous scaffolds (27,30,31) as a platform for islet transplantation. These scaffolds allow for the in-growth of host tissue, so they must be combined with immunosuppression or tolerogenic strategies to prevent immune-mediated rejection (31); however, they offer the opportunity to modify the microenvironment of transplanted islets. Alternate locations aside from the liver have been proposed as transplant sites, including kidney capsule, spleen and subcutaneous sites, but they have failed to show clinical effectiveness or significant advantage over the hepatic site (8,9). We have demonstrated that microporous scaffolds are able to serve as a platform for extrahepatic islet transplantation in multiple animal models, allowing for rapid cell infiltration and islet revascularization while successfully reversing diabetes with a minimal mass of islets (29–31,34). We have also demonstrated that application of specific extracellular matrix proteins to the scaffold significantly enhanced islet engraftment and graft outcomes (29,34).

Local and sustained protein release holds the potential to significantly improve the efficacy of islet transplantation. The microenvironment of transplanted islets can impact islet engraftment and function given that the isolation procedure, among other factors, disrupts the basement membrane and vascular network. Further, the current clinical transplantation procedure, intraportal infusion, induces an inflammatory response, and the local environment of the liver is not ideally suited to islet survival and function (5). Hence, islet function may be compromised posttransplantation, and the islets are vulnerable to damage and apoptosis. Islets have been cultured with trophic factors prior to transplantation, or trophic factors have been delivered systemically as a means to enhance islet function; however, systemic delivery comes with the possibility of adverse effects and may not be possible with some factors. An alternative to these systemic delivery approaches has involved engineering islets to secrete factors for localized protein production (35,36). This approach provides localized production of factors that can create elevated concentrations at the islet with low concentrations systemically, but it requires engineering of the transplanted islets. In our study, localized delivery of inductive factors from a bioactive scaffold was used to modify the microenvironment to expose islets to trophic peptides in the early transplant period, obviating the need to alter islets or treat the islets pretransplant. This was accomplished by engineering a layered scaffold with porous outer layers for islet seeding and cell infiltration to integrate islets with the host vasculature, and a central layer of polymer microspheres capable of releasing bioactive protein in a sustained manner into the islet microenvironment posttransplant.

Localized delivery of IGF-1 from scaffolds lowered glucose levels but did not restore euglycemia in mice transplanted with human islets. Previous studies have demonstrated that IGF-1 decreases islet cell apoptosis and stimulates islet cell proliferation (18–21). We observed a small effect of IGF-1 on islet function posttransplantation, which suggested that IGF-1 may only be effective prior to transplantation, or islet apoptosis and replication are not limiting factors dictating graft function in our model. Of note, IGF-1 exhibited a lower encapsulation efficiency than Ex4, but the in vitro studies demonstrated that sufficient bioactive IGF-1 was released to activate IGF-1 signaling pathways, as reflected by ERK phosphorylation. IGF-1 release demonstrated an initial burst with the majority of protein released by Day 7. As islets are most vulnerable to apoptosis in the early transplant period prior to revascularization, these release kinetics match the anticipated need for preventing apoptosis. Finally, studies demonstrating the action of IGF-1 in stimulating murine beta cell replication in vitro may not directly translate to this study, as humans have a lower capacity for beta cell replication than rodents (37).

We next investigated Ex4 delivery as a means to enhance islet function posttransplant. Exenatide, a GLP-1 receptor agonist, is a clinically approved synthetic version of Ex4 used to treat type 2 diabetes. As a GLP-1 receptor agonist, Ex4 may stimulate insulin secretion and β cell replication and reduce apoptosis after transplant (12–16). Initial approaches using Ex4 to enhance islet transplantation involved treating islets prior to transplantation (10). Pre-culture of islets with Ex4 improved transplant outcomes compared to preculture without Ex4 (38); however, fresh islets demonstrated a shorter time to euglycemia than either precultured islet group, as relatively long culture times prior to transplant result in islet hypoxia and necrosis (39). Ex4 has also been delivered systemically by subcutaneous injection to transplant recipients (40). However, systemic administration of Ex4 to islet transplant recipients has been associated with side effects including nausea, vomiting and reduced gastric emptying (11,40,41). Encapsulating Ex4 in PLG microspheres that are injected subcutaneously to achieve sustained release (42,43) has attenuated some adverse effects of systemic Ex4 injection (43). Finally, pancreatic islets have shown enhanced insulin secretion in vitro when encapsulated in GLP-1 functionalized PEG hydrogels and coatings (44,45), but these biomaterial systems have not been tested in vivo.

Localized Ex4 delivery significantly improved graft function when transplanting a minimal mass of human islets. We previously established the minimal mass of human islets to reverse diabetes following transplantation on a scaffold (without interventions such as protein delivery) as 2000 IEQ (within the same viability and purity ranges) (27). We have now demonstrated that significantly more mice receiving 1500–2000 IEQ transplanted on 3% Ex4 compared with control scaffolds maintained graft function 30 days post-transplant (Figure 6B). Furthermore, the IPGTT demonstrated that mice transplanted with islets on 3% Ex4-releasing scaffolds had lower glucose levels at each time point and a significant reduction in the AUC (Figure 7). Subsequent analysis demonstrated a greater islet area in 3% Ex4 scaffolds at Day 30 with an increase in proliferating cells, suggesting that Ex4 scaffolds allow a minimal islet mass to survive posttransplant and undergo modest proliferation in comparison to islets transplanted on blank scaffolds (Figure 8). These results demonstrate a role for local release of Ex4 in human islet transplantation at a site with potential for clinical translation. Although islet function was improved, not all grafts induced euglycemia for long times. Multiple studies have reported that transplantation of a sub-minimal mass may attenuate blood glucose levels posttransplant, yet over time the islets become stressed and the grafts ultimately fail (46,47). Thus, the present study is the first to demonstrate that localized, sustained delivery of Ex4 from a platform used for islet transplantation can enhance islet function posttransplant and suggests an approach for enhancing the early survival of transplanted islets, maximizing transplant graft mass, and thus minimizing stress on the transplanted islets in the weeks and months posttransplantation.

Conclusions

We investigated localized, controlled delivery of inductive factors from the central layer of a polymer scaffold for enhancing human islet transplantation. Trophic factors were encapsulated with high efficiencies with sustained release for several weeks, and bioactivity was preserved after incorporation. IGF-1 lowered blood glucose in vivo, yet this effect was insufficient to restore euglycemia. Scaffold-mediated delivery of Ex4, which stimulates insulin secretion, significantly improved graft function and enhanced glycemic control upon glucose challenge. The release profile of Ex4 significantly impacted the percentage of animals that achieved euglycemia. Human islets are distinct from other species in their cytoarchitecture, physiology and glucose sensitivity (48–50); this xenogeneic transplant model using local protein delivery within an extrahepatic site may identify factors that enhance human islet engraftment and function following transplantation and provides a versatile tool for modulating the posttransplantation islet microenvironment.

Abbreviations

AUC

area under the curve

Ex4

exendin-4

GLP-1

glucagon-like peptide-1

H&E

hematoxylin and eosin

IEQ

islet equivalent

IGF-1

insulin-like growth factor-1

IIDP

Integrated Islet Distribution Program

IPGTT

intraperitoneal glucose tolerance test

PLG

poly(lactide-co-glycolide)