A hydrogel platform for co-delivery of immunomodulatory proteins for pancreatic islet allografts

Abstract

Type 1 diabetes (T1D), an autoimmune disorder in which the insulin-producing β-cells in the islets of Langerhans in the pancreas are destroyed, afflicts over 1.6 million Americans. Although pancreatic islet transplantation has shown promise in treating T1D, continuous use of required immunosuppression regimens limits clinical islet transplantation as it poses significant adverse effects on graft recipients and does not achieve consistent long-term graft survival with 50–70% of recipients maintaining insulin independence at 5 years. T cells play a key role in graft rejection, and rebalancing pathogenic T effector and protective T regulatory cells can regulate autoimmune disorders and transplant rejection. The synergy of the interleukin-2 (IL-2) and Fas immunomodulatory pathways presents an avenue for eliminating the need for systemic immune suppression by exploiting IL-2’s role in expanding regulatory T cells and leveraging Fas ligand (FasL) activity on antigen-induced cell death of effector T cells. Herein, we developed a hydrogel platform for co-delivering an analog of IL-2, IL-2D, and FasL-presenting microgels to achieve localized immunotolerance to pancreatic islets by targeting the upregulation of regulatory T cells and effector T cells simultaneously. Although this hydrogel provided for sustained, local delivery of active immunomodulatory proteins, indefinite allograft survival was not achieved. Immune profiling analysis revealed upregulation of target regulatory T cells but also increases in Granzyme B-expressing CD8⁺ T cells at the graft site. We attribute the failed establishment of allograft survival to these Granzyme B-expressing T cells. This study underscores the delicate balance of immunomodulatory components important for allograft survival – whose outcome can be dependent on timing, duration, modality of delivery, and disease model.

Introduction

Type 1 diabetes (T1D), an autoimmune disorder in which the insulin-producing β-cells in the islets of Langerhans in the pancreas are destroyed, is growing rapidly with 64,000 yearly diagnoses and $15 billion annual healthcare costs in the U.S.^1–3. Daily glucose monitoring, treatment with exogenous insulin, and fear of sudden hypoglycemic episodes and long-term complications make T1D a physiologically and psychologically exhausting disease to live with. Although long-acting insulins, continuous glucose monitoring devices, and insulin pumps have improved quality of life, there is still no cure for T1D and long-term complications can reduce life expectancy by 18 years^{1, 4}. Advances in cadaveric islet transplantation under stringent immunosuppression regimens have allowed for 50–70% of patients to achieve insulin independence at 5 years^{5, 6}. However, human donor shortages and chronic immunosuppression limit the widespread adoption of islet transplantation as immunosuppressive protocols affect the graft and make patients prone to infections and malignancy, while also increasing their risk for cardiovascular and metabolic disease^7–10. In fact, there have only been ~1,500 islet transplantations since 2000, as risks associated with chronic immunosuppression must be balanced against the threats of chronic diabetes^{5, 11}.

Uncontrolled expansion of β cell-reactive T effector (Teff) cells in pre-diabetes is postulated to reach a point where it can no longer be contained by protective T regulatory (Treg) cells¹². This imbalance increases the likelihood for the survival of autoreactive Teff cells that infiltrate the pancreas and cause massive β-cell death. Onset T1D clinical trials have shown that short-term restoration of the Treg/Teff balance can temporarily slow down disease progression¹³. Prior work from our group demonstrated that controlling the Treg/Teff balance at the local transplantation site using FasL-presenting microgels under a 2-week course of rapamycin treatment achieved long-term allogeneic islet graft survival in chemically diabetic mice and nonhuman primates^{14, 15}. Rapamycin was used in conjunction with FasL because of its synergistic effects in inducing apoptosis of alloreactive T cells and generation of Tregs. However, rapamycin increases the number of both Tregs and Teffs at the graft site¹⁴, suggesting that a more Treg-selective strategy could lead to Treg predominance. Furthermore, the regimen of rapamycin will likely be extended in large mammals, and this prolonged systemic immunosuppression may increase the risk of cardiovascular and metabolic diseases and adversely impact allograft function⁸. Thus, the T cell proliferative action of IL-2, of which Tregs are more susceptible towards, and its synergistic effects with FasL, make it an attractive candidate for eliminating rapamycin to achieve allograft tolerance. Binding of IL-2 to IL-2R activates the Stat5 pathway that is key to Treg cell maintenance and proliferation¹⁶. Activation of the Fas/FasL pathway triggers activation induced cell death (AICD), or apoptosis stimulated by repeated activation of lymphocytes¹⁷. Importantly, IL-2 further sensitizes Fas-mediated AICD of mature Teffs, but naïve T cells and Tregs remain unaffected by Fas-induced cell death¹⁸. Therefore, the Fas and IL-2R signaling pathways educate the immune system in such a way to prevent excessive reaction and autoreactivity.

In this study, we engineered a 4-arm poly(ethylene glycol) maleimide (PEG-4MAL)-based hydrogel platform for local delivery of IL-2D, an analog of IL-2 with an RGD amino acid motif to promote Treg activation, along with a chimeric form of FasL fused to the streptavidin binding domain (SA-FasL) presented on microgels for the desired outcome of immune protection against allogeneic pancreatic islets to reverse diabetes in chemically-induced diabetic mice without the use of rapamycin.

Materials and Methods

Production of recombinant proteins

Recombinant SA-FasL was constructed and purified as previously described¹⁸. IL-2D is the human version of the protein modified to include a 6xHis tag followed by an arginine-glycine-aspartate (RGD) motif, an integrin-recognition motif, to enhance apoptosis via conformational changes that promote pro-caspase-3 activation¹⁹. The recombinant protein was produced in the Shirwan and Yolcu labs at the University of Missouri using a Drosophila expression system (Invitrogen) as described previously^{20, 21}. The protein was purified using an immobilized metal ion affinity chromatography. IL-2D activity was quantified using the CTLL-2 cell proliferation assay using commercial IL-2 as a standard²². The protein was assessed for endotoxin contamination (> 0.003 EU/μg) using Pierce^™ Chromogenic Endotoxin Quant Kit (Thermo Fisher, Waltham, MA), aliquoted, and frozen in −80°C until use.

PEG hydrogel synthesis and IL-2D PEGylation

Four-arm maleimide-end functionalized PEG macromer (PEG-4MAL 20 kDa MW, Laysan Bio, >95% purity) was reacted to form covalent bonds with free thiols on IL-2D (3 total) for 30 minutes at room temperature in 20 mM HEPES in PBS (pH=7.4). To confirm PEGylation, the IL-2D-PEG-4MAL product was run on an SDS-PAGE gel followed by protein visualization with Sypro Red (Life Technologies) staining. To synthesize IL-2D-presenting hydrogels, PEGylated IL-2D was mixed with 10 μg/mL VEGF (for enhanced survival, engraftment, and function of islets) and cross-linked into a 5.0% (w/v) PEG-4MAL network using the dithiol protease-degradable peptide RDVPMSMRGGDRCG (VPM, GenScript). Because the molar ratio of macromer:IL-2D and macromer:VEGF were ≥155, the number of unreacted maleimides was essentially equivalent to the total number of initial maleimides. Because of this, crosslinker concentration was adjusted to match the number of total maleimide groups in PEG-4MAL macromer.

Bioactivity of PEGylated IL-2D

IL-2-dependent CTLL-2 cells (ATCC) were grown in complete DMEM media (Life Technologies). Cells were incubated with either 1) human recombinant IL-2 (positive control), 2) unPEGylated IL-2D, or 3) PEGylated IL-2D for 20 hours. For in vitro titration experiments, IL-2D (both PEGylated and unPEGylated) doses were varied for a constant number of CTLL-2 cells (5.0×10⁵), a cell line that is does not proliferate in the absence of IL-2. Since IL-2D was not purified after PEGylation, activity units were quantified by respective dilution factors. EdU proliferation kits (Invitrogen) were used to assess the proliferation of CTLL-2 cells via flow cytometry.

In vitro release kinetics of PEGylated IL-2D

To assess IL-2D release kinetics, IL-2D was labelled with AlexaFluor-488 NHS Ester (Thermo Scientific), purified, PEGylated, and incorporated in hydrogels as described. Hydrogels were incubated in either PBS or 50 μg/mL type I collagenase (Worthington Biomedical). At specified timepoints, supernatant was collected and analyzed for fluorescence. The amount of IL-2D directly incorporated into hydrogels was calculated by comparing the fluorescence signal of the supernatant from hydrogels incubated in PBS after 24 hours to the supernatant from completely degraded hydrogels.

In vivo release kinetics of PEGylated IL-2D

To assess IL-2D release kinetics in vivo, IL-2D was labelled with DyLight-680 NHS Ester (Thermo Scientific), purified, and reacted with PEG-4MAL macromers. The labelled PEGylated IL-2D was incorporated into hydrogels polymerized in situ in the epidydimal fat pads (EFPs) of male BALB/c mice. A second group of mice were implanted with soluble labelled IL-2D. At specific timepoints, mice were imaged using a Perkin-Elmer IVIS Spectrum CT imaging system.

Islet viability

Pancreatic islets were isolated from Lewis rat male donors using Liberase TL (Roche Life Science) and separated from acinar tissue using Ficoll density gradients as described²³. Rat islets were selected for in vitro studies because of the higher islet yield per animal. To verify that IL-2D-encapsulating PEG-4MAL hydrogels were not detrimental towards pancreatic islets, rat islets were incubated with or without IL-2D hydrogels for 48 hours. Islet health was assessed by quantifying secretion of danger-associated molecular patterns (DAMPS), 70 kDa heat shock protein 70 (Hsp 70) and uric acid via ELISA (MyBioSource and Abcam), and by quantifying metabolic activity via alamarBlue staining (Invitrogen).

SA-FasL microgel synthesis and bioactivity

Biotinylated microgels were synthesized as described previously²⁴. Briefly, 5.0% PEG-4MAL functionalized with 1.0 mM biotin-PEG-thiol (1 kDa, Nanocs) was dispersed into droplets using a microfluidic chip. PEG droplets were then crosslinked with an emulsion of dithiothreitol (Sigma) and mineral oil (Sigma) with 2% SPAN80 (Sigma). Excess emulsion was washed off by centrifugation in 1% bovine serum albumin (Sigma). Biotinylated microgels were incubated with SA-FasL for 1 hour and then washed 5 times by centrifugation in PBS. To assess bioactivity, SA-FasL-presenting microgels were co-incubated with 500,000 Fas apoptosis-sensitive A20 cells (ATCC) with or without IL-2D-encapsulating PEG-4MAL hydrogels in 1.0 mL media. After 18 h, the cells were stained with markers of early and late apoptosis (APC annexin V and propidium iodide, BD Biosciences). Samples were analyzed by flow cytometry.

Islet transplantation

All animal procedures were performed under protocols approved by the Georgia Institute of Technology IACUC and in accordance with NIH guidelines. Pancreatic islets from C57BL/6 or BALB/c mice were isolated using Liberase Tflex (Roche Life Science) and purified with Ficoll density gradients (Corning). Streptozotocin-diabetic C57BL/6 (180 mg/kg i.p.; diabetes defined by blood glucose readings >300 mg/dl on two consecutive days) were transplanted with 600 islet equivalents (IEQ), if syngeneic, or 1200 IEQ, if allogeneic, and 2400 biotinylated hydrogel microparticles (microgels) into EFPs of male recipients, sealed with a PEG hydrogel as previously described¹⁴. Depending on the group, microgels were delivered into each EFP presenting either 600 ng SA- or 1.2 μg SA-FasL on their surface and a 15 μL PEG hydrogel delivering VEGF with or without IL-2D, the former with different IL-2D doses (250, 500, 5000, 10,000, 20,000 units) depending on the study. The EFPs are closed around the graft and sealed with an empty 15 μL PEG hydrogel to further secure the islets within the EFPs. Blood glucose and body weight of recipients were subsequently monitored by blinded scientists. For all transplanted mice, blood glucose levels were initially restored to the normoglycemic range, indicating primary graft function. Blood glucose levels >250mg/dL for two consecutive measurements were considered as an indication of islet graft rejection.

Immune cell profiling and IL-2D titration

To optimize IL-2D dosing, a full dose of BALB/c pancreatic islets (1200 islet equivalents) was co-transplanted with 2400 SA-FasL-presenting microgels (1 μg SA-FasL per 1000 microgels) and PEG-4MAL hydrogels encapsulating VEGF with either 1) 5000 units, 2) 10,000 units, or 3) 20,000 units of IL-2D into the EFPs of immunocompetent, streptozotocin-diabetic C57BL/6 mice. At 3 and 6 days after transplantation, EFPs were explanted and digested into isolated single cells by collagenase digestion. Cells were then stained for T cell subpopulations using antibodies to cell surface markers CD4, CD8, CD45, CD3, CD25, CD44, CD62L, and intracellular Granzyme B and FoxP3 (BD Biosciences). Data were then collected via flow cytometry and analyzed on GraphPad.

Statistics

Statistical analyses were performed using GraphPad Prism version 9.0.0 for Windows (GraphPad Software, San Diego, California USA). Data is presented as mean with error bars representing standard error (SE). Comparisons between two groups were conducted through a t-test and those conducted among several groups was performed by one-way analysis of variance (ANOVA). Where appropriate, a two-way ANOVA was conducted. A p value of <0.05 was considered significant.

Results

IL-2D-releasing PEG-4MAL hydrogels

A PEG-4MAL macromer was chosen as the backbone of the hydrogel platform as its maleimide end groups react with free thiols to form covalent bonds via a Michael-type reaction. This reaction allows for direct incorporation of bioactive ligands within the hydrogel, and by crosslinking the macromers using protease-degradable peptides, release of these ligands is obtained via proteolytic degradation (Fig. 1A). PEGylation of IL-2D was confirmed by running an SDS-PAGE gel and observing an increase in molecular weight of ~20 kDa for IL-2D reacted with PEG-4MAL macromer, indicating that the majority of IL-2D is monoPEGylated (Fig. 1B). PEGylated IL-2D was then cross-linked into a hydrogel using the protease-degradable peptide VPM. When subjected to collagenase I, hydrogels released all the encapsulated IL-2D, demonstrating that cleavage of VPM yields controlled release of IL-2D-functionalized PEG-4MAL macromers, which was not observed in hydrogels not subject to collagenase incubation (Fig. 1C). An initial burst release of ~25% of the fluorescently-labeled IL-2D, even in hydrogels not subject to collagenase I, indicates that roughly 75% of total IL-2D becomes directly bound to the hydrogel network; the remaining 25%, which is not covalently bound to the hydrogel but entrapped within the network, is passively released – an observation that was consistent across three different IL-2D doses (Fig. 1D).

Figure 1:

Protease-degradable PEG-4MAL hydrogels for the controlled release of covalently-tethered IL-2D. (A) Schematic detailing the synthesis of IL-2D-encapsulating PEG-4MAL hydrogels and protease-dependent degradation. (B) SDS PAGE gel of IL-2D and PEG-4MAL + IL-2D showing an increase in MW, demonstrating successful covalent tethering of IL-2D to PEG-4MAL macromers. (C) IL-2D in vitro release profile from PEG-4MAL hydrogels treated in PBS or collagenase as measured by fluorescence (mean ± SE, N=4). (D) IL-2D loading efficiency at 3 different doses as measured by fluorescence on hydrogels encapsulating tagged IL-2D (mean ± SE N=4).

To confirm that PEGylation does not affect the bioactivity of IL-2D, proliferation studies were conducted using CTLL-2 cells, which depend on IL-2 for their proliferation²⁵. We found dose-dependent increases in CTLL-2 proliferation for both unmodified and PEGylated IL-2D, of which there are no differences in all tested doses and are comparable to that of human recombinant IL-2 (hIL-2) at 50 U/mL (Fig. 2A). Furthermore, unmodified and PEGylated IL-2D retained the same level of bioactivity when stored at 4°C for 30 days, an important consideration for translation into the clinic (Fig. 2B).

Figure 2:

Bioactivity of IL-2D is not negatively affected after being PEGylated with PEG-4MAL. (A) Dose-dependent proliferation of IL-2-sensitive CTLL-2 cells by unPEGylated and PEGylated IL-2D (mean ± SD, N=3). (B) Bioactivity of stored IL-2D and PEGylated IL-2D over time as measured by CTLL-2 proliferation via EdU proliferation kits and flow cytometry (mean ± SD, N=3).

Hydrogel prolongs presence of IL-2D at graft site

To determine the kinetics and localized nature of IL-2D release from PEG-4MAL hydrogels in vivo, IL-2D was fluorescently labelled with a near infrared fluorophore, PEGylated, and directly incorporated into PEG-4MAL hydrogels implanted into the EFPs of BALB/c mice. Control BALB/c mice received the same concentration of labelled IL-2D without incorporation into a hydrogel (free IL-2D). The EFPs of all recipients were folded and sealed with a PEG hydrogel. Delivery of hydrogel-encapsulated IL-2D was locally sustained in the EFPs over time, as opposed to that of free IL-2D, which was rapidly cleared (Fig. 3A). Normalized signal was significantly higher in those that received IL-2D hydrogels than those that received free IL-2D for the first 12 days (Fig. 3B). Curve fitting of the data with a single-phase decay model showed a 237% increase in half-life (5.15 days vs. 2.17 days) for IL-2D for hydrogel delivery compared to free protein¹⁴. Signal from explanted EFPs in the encapsulated IL-2D group was also significantly higher, highlighting an extended residence time (Fig. 3C).

Figure 3:

Sustained and localized in vivo release of IL-2D from degradable PEG-4MAL hydrogels. (A) Signal from IL-2D encapsulated in PEG-4MAL hydrogels is sustained, in contrast with rapid clearance of free IL-2D. (B) Normalized signal is presented for both conditions; ****P<0.0001; ***P<0.001; **P<0.01, multiple t tests (N=7 mice per group). (C) IL-2D signal from explanted (EFPs) at D19 (n=14); ***P<0.001, unpaired t test.

IL-2D hydrogels do not negatively affect islet health and graft function

Since IL-2D hydrogels were synthesized for pancreatic islet allografts, we needed to ensure that pancreatic islet health was not negatively impacted when cultured with IL-2D hydrogels. We first evaluated in vitro cytocompatibility of IL-2D hydrogels by co-culturing rat islets in the absence or presence of casted IL-2D hydrogels for 48 hours and measuring secretion of danger-associated molecular patterns (DAMPs) via ELISA. We selected HSP70 and uric acid secretion as metrics for pancreatic islet health because they have been associated with apoptosis, autophagy, and necroptosis²⁶. We observed no differences in DAMP secretion between islets cultured in the presence or absence of IL-2D hydrogels (Fig. 4A). We also measured metabolic activity of islets incubated with or without IL-2D hydrogels for 48 hours as it is a strong indicator for islet cell viability²⁷, and we found no differences between islets incubated with or without IL-2D hydrogels (Fig. 4B). Together, these findings indicate that IL-2D hydrogels do not negatively affect the health of islets.

Figure 4:

Islet viability and SA-FasL microgel bioactivity are unaffected by IL-2D-delivering PEG-4MAL hydrogels. (A) Islets co-cultured with or without IL-2D-delivering hydrogels for 48 h and there were no differences in secretion of danger associated molecular patterns (DAMPs) HSP70 and uric acid (mean ± SE, N≥3; unpaired t-test). (B) Metabolic activity of islets was unaffected by presence of IL-2D-delivering PEG-4MAL hydrogels (mean ± SE, N=3; unpaired t-test). (C) No differences were detected in the apoptotic activity of SA-FasL microgels in the presence of IL-2D PEG-4MAL hydrogels (mean ± SE, N=3; one-way ANOVA). (D) The combination therapy does not have detrimental effects on graft function in a syngeneic transplant model (shadowed area represents SE for each group, N=3).

To confirm that IL-2D hydrogels do not negatively impact the bioactivity of SA-FasL microgels, FasL apoptosis-sensitive A20 cells were co-cultured in the presence of casted IL-2D hydrogels and SA-FasL microgels. The addition of IL-2D hydrogels did not interfere with SA-FasL microgels’ ability to induce apoptosis of A20 cells relative to that observed in A20 cells incubated with SA-FasL microgels in the absence of IL-2D hydrogels (Fig. 4C). To ensure that the combination therapy (IL-2D hydrogels with SA-FasL microgels) does not impair graft function in the absence of alloimmunity, we conducted syngeneic transplants in which 600 C57BL/6 IEQ were transplanted into STZ-diabetic C57BL/6 recipients along with 1) SA-microgels + empty hydrogels or 2) SA-FasL microgels + IL-2D hydrogels. We found that IL-2D hydrogels do not have detrimental effects on graft function in this syngeneic transplant model, as indicated by normoglycemic blood glucose levels recorded post-transplantation (Fig. 4D). This suggests that IL-2D hydrogels do not negatively affect the health of pancreatic islets.

IL-2D hydrogels upregulate T cell subpopulations at the graft site.

To examine the immunological effects induced by the combination therapy, we conducted immune profiling studies on STZ-diabetic C57BL/6 mice transplanted with fully immune mismatched BALB/c islets (1200 IEQ) and either (i) VEGF PEG-4MAL hydrogel + SA-presenting microgels, (ii) VEGF PEG-4MAL hydrogel + SA-FasL-presenting microgels, or (iii) VEGF+IL-2D-delivering hydrogels (20,000 U) + SA-FasL-presenting microgels. The initial dose of 20,000 U was chosen as it was the maximum amount that could be delivered in a 5.0% (w/v) hydrogel. At selected time points (3- and 6-days post-transplantation), EFPs containing islet grafts were explanted, digested, and cells were stained for T cell subpopulations. The IL-2D/SA-FasL combination therapy induced a significant upregulation of Tregs at day 3 which was further heightened at day 6 (Fig. 5A) despite no differences in CD4⁺ Teff counts at day 3 and day 6 between all groups (Fig. 5B). This lack of dependence on the combination therapy was paralleled by similar counts of CD8⁺ Teffs in combination therapy-treated mice as that of control and monotherapy-treated mice (Fig. 5C). However, an upregulation of both CD4⁺ and CD8⁺ Teffs was observed at D6 across all groups. Furthermore, the subset of CD8⁺ T cells expressing granzyme B was significantly upregulated by the combination therapy with levels sustained through day 6 (Fig. 5D). Taken together, these observations show a broad effect on T cell subpopulations created by the presence of these hydrogels – particularly, we observe both an upregulation of Tregs and granzyme B-expressing CD8⁺ T cells in mice that received the combination therapy. Since IL-2 can induce either anti- or pro-inflammatory responses (depending on the dose), these broad results indicate a need to tip the balance more towards an anti-inflammatory response to decrease the presence of granzyme B-expressing CD8⁺ T cells²⁸.

Figure 5:

The combination therapy of SA-FasL microgels with an IL-2D-delivering PEG-4MAL hydrogel induces significant changes to T cell subpopulations at the graft site. (A) The combination therapy group exhibited a significant upregulation of Tregs (CD4⁺CD25⁺FoxP3⁺) at D3 that was further heightened at D6. (B) There were statistically significant differences in CD4⁺Teffs (CD4⁺CD44^hiCD62L^lo) across groups, stemming from the decrease in CD4⁺ Teffs at D6 exhibited by the combination therapy group. (C) This decrease in CD4⁺ Teffs is paralleled by a lower presence of CD8⁺ Teffs (CD8⁺CD44^hiCD62L^lo) at D3 and D6 in the combination therapy group. (D) The combination therapy group also exhibited a significant upregulation of GranzymeB⁺CD8⁺ T cells at D3 that was sustained at D6 (mean ± SE, N≥4; two-way ANOVA).

To find a dose with more favorable immune-protective effects, we conducted further immune profiling studies on STZ-diabetic C57BL/6 mice transplanted with fully immune mismatched BALB/c islets (1200 IEQ), SA-FasL microgels, and hydrogels delivering 1) 20,000, 2) 10,000, or 3) 5,000 U of IL-2D at days 3 and 6 post-transplantation. A significant IL-2D-dose-dependent increase in Tregs was observed at both day 3 and day 6 despite no significant increases in CD4⁺ Teffs nor CD8⁺ Teffs (Fig. 6A–C). Like the aforementioned immune profiling results, the presence of CD4+ Teffs (Fig. 6B) and CD8+ Teffs (Fig. 6C) both increased from day 3 to day 6 for all three doses, suggesting a broad activation of Teffs that was not dose dependent. Furthermore, we observed a pronounced dose-dependent effect in the presence of granzyme B-expressing CD8⁺ T cells at both time points. It is worth noting that this upregulation of CD8⁺GranzymeB⁺ T cells is more pronounced at day 6 for all concentrations (Fig. 6D). Despite this last observation, the IL-2D dose-dependent upregulation of Tregs combined with the IL-2D dose-independence of Teff counts led us to proceed with conducting allogeneic islet transplantations to test the combination therapy’s ability to induce allograft tolerance.

Figure 6:

(A) The combination therapy induces an IL-2D dose-dependent upregulation of Tregs that is further heightened at D6. (B, C) This dose dependence is not observed in the presence of CD4+ and CD8+ Teffs. (D, E) A time-dependent decrease in CD4+ Teffs was observed, an indication that SA-FasL-induced Teff apoptosis may be occurring. However, (F) a dose-dependent increase of Granzyme B-expressing CD8+ T cells was observed, hinting at a broader activation of T cells caused by IL-2D (mean ± SE, N≥4; two-way ANOVA).

Combination therapy in the context of T1D

To test the combination therapy’s ability to induce allograft tolerance and reverse diabetes in the absence of any immunosuppression, we conducted allogeneic transplants in which 1200 BALB/c IEQ were transplanted into the EFPs of STZ-diabetic C57BL/6 recipients along with either (i) SA-microgels + VEGF hydrogels with no IL-2D or (ii) SA-FasL microgels + VEGF hydrogels with IL-2D. We first transplanted mice with IL-2D hydrogels (20,000 U) as this dose induced the most pronounced Treg upregulation with minimal Teff upregulation. However, we subsequently lowered the IL-2D dose to 5000, 1000, and 500 U as graft rejection, defined by two consecutive readings >250 mg/dL, was observed at an earlier time for mice treated with 20,000 U compared to control groups (Fig. 7A). We found that there was an inverse correlation between length of graft function and IL-2D dose (Fig. 7B).

Figure 7:

(A) Metabolic allograft function, based on non-fasting blood glucose levels (shadowed area represents SE; N≥4). (B) Graft function and IL-2D dose are inversely correlated is prolonged with a lowered dose of IL-2D (simple linear regression, R² =0.3074, P<0.05).

Discussion

β-cell specific T cells are speculated to be activated by autoreactive B cells that recognize β-cell-like autoantigen-specific T cells or autoreactive naïve T cells that avoid negative selection in the thymus that then go undetected by dysfunctional peripheral tolerance mechanisms^29–32. This can occur due to defects in Tregs leading to decreased numbers in inflamed tissues, lowered IL-2 receptor signaling, and/or inability to suppress reactive cells¹⁶. Efforts have been made to correct Teff autoreactivity in T1D patients; of such have been in vitro-expanded Treg auto transfers aimed at achieving durable immune tolerance³³. These have had limited success due to lack of peripheral space for engraftment, competition from host Tregs, insufficient IL-2 levels, and antigen unavailability for the selective expansion of disease-relevant Tregs^{33, 34}. Immune checkpoint inhibitors, widely leveraged for cancer immunotherapies, have also been used within the context of T1D. One example is that of the PD-1/PD-L1 pathway, which has proven to be vital for successful engraftment of fully mismatched cardiac allografts³⁵. Presentation of a chimeric streptavidin/PD-L1 (SA-PD-L1) on the surface of PEG-4MAL microgels alongside fully mismatched BALB/c islets in the EFPs of diabetic C57BL/6 mice led to local induction of allograft acceptance in combination with a brief rapamycin treatment³⁶. Within a similar context, delivery of fully mismatched BALB/c islets and SA-FasL on PEG-4MAL microgels, with a short course of rapamycin, yielded allograft acceptance and function over 200 days¹⁴. This same technology was extended to chemically diabetic non-human primates, where it induced the acceptance of islet allografts¹⁵. Despite these compelling results, translation into the clinic may require a scaled-up course of rapamycin, hence not fully eliminating the need for systemic immunosuppression – albeit a short course and not for the duration of the graft. Herein, we sought to leverage the synergistic signaling pathways of Fas and IL-2R, as they help to maintain a stable Treg/Teff balance via multiple routes, by engineering a hydrogel platform for the delivery of IL-2D in conjunction with SA-FasL microgels to eliminate immunosuppressive regimens from allogeneic islet transplantation. The protease-dependent degradation of the hydrogel platform allowed for the release of IL-2D in a localized and sustained manner within the EFPs of recipient mice. Furthermore, delivery of IL-2D, in conjunction with SA-FasL-presenting microgels, induced a dose-dependent upregulation of Tregs without significantly upregulating Teffs. However, a subpopulation of CD8⁺GranzymeB⁺ T cells was heightened in an IL-2D dose-dependent manner. We also found that length of graft function was dependent on IL-2D dosage, an indicator that direct hydrogel delivery of IL-2D (even at lower doses), was creating a pro-inflammatory environment. Immune profiling studies found that Treg presence is not as pronounced at lower doses; and at 500 U, Treg upregulation may have not been sufficient to induce allograft tolerance as indefinite graft survival was not achieved. This is despite studies showing that low-dose IL-2 results in more immune protective effects; e.g., clinical trials where patients with chronic graft-versus-host disease treated with low-dose IL-2 showed the selective expansion of functional Tregs and clinical improvement^37–39. Furthermore, a phase 1/2 clinical trial in which T1D patients were administered with low doses of IL-2 found a dose-dependent increase in Tregs without a significant increase in Teffs and NK cells⁴⁰.

Because indefinite graft function was not achieved, the tested lower doses may have not been sufficient to induce allograft tolerance like a short course of rapamycin did with SA-FasL microgels¹⁴. Although IL-2 sensitizes activated T cells to AICD, it may have not been to the same extent that rapamycin induces apoptosis of alloreactive T cells⁴¹. It may also be that the true synergy between FasL and IL-2 is not captured by our approach due to several reasons. Of such, the affinity and/or differential signaling in Tregs vs Teffs as a response to IL-2D in hydrogels may have caused an earlier infiltration of CD8⁺ Teffs; this is within the realm of possibilities for PEGylated IL-2D as others have shown that IL-2-antibody complexes can induce very specific and varied immune responses, including massive expansions of CD8⁺ T cells in vivo⁴². Furthermore, this platform may not have allowed for the synchronized action of SA-FasL (half-life: 3 ± 0.8 days) and IL-2 (half-life: 5.15 days); the prolonged presentation of IL-2D at the graft site may have further affected its pharmacokinetics which, along with significant amounts of alloantigen, shifted towards IL-2’s inflammatory properties. It is also possible that the suppressive function of Tregs has been altered or that the therapy makes CD8⁺ Teffs more resistant to FasL-mediated apoptosis. Furthermore, delivery of our biomaterial platform prior to islet delivery may have induced a more immune-hospitable site for the islet allograft; however, this would come at the cost of adding a second surgical procedure and limited apoptotic activity of SA-FasL by the initial lack of islet antigen. We note that drainage of IL-2D to lymph nodes was not considered a priority during the design of the biomaterial vehicle. Targeted delivery to lymph nodes presents an attractive route for establishing robust immune tolerance and is worth considering for future studies in other contexts. Given the complexities of identifying a “sweet spot” dose for IL-2, others have constructed IL-2 muteins to induce preferential Treg cell enrichment – this includes a study wherein an IL-2 mutein extended resolution of NOD diabetes^{39, 43}. Others have directly and selectively modified IL-2 via site-specific PEGylations for enhancing the immunosuppressive properties of IL-2 by modifying the pharmacokinetics and half-life of IL-2⁴⁴. We posit that a systematic approach for identifying key protein residues for specific modifications, in conjunction with a delivery vehicle like the one described here, would provide an avenue for inducing a more immune protective response and prolonging graft function.

Conclusion

We describe a PEG-4MAL hydrogel platform for the co-delivery of IL-2D and SA-FasL-microgels within the context of allogeneic islet transplantation. This hydrogel platform did not negatively affect bioactivity of either immunomodulatory protein, was compatible with pancreatic islets for cell replacement therapies, and induced significant changes in T cell subpopulations in a dose-dependent manner. Although long-term allograft function was not achieved, this study showcases a robust biomaterial-mediated approach for delivery of bioactive immunomodulatory proteins that results in enhanced residence time in a localized manner.