Abstract
Activation of adenosine A2A receptors inhibits inflammation in ischemia/reperfusion injury, and protects against cell damage at the injury site. Following transplantation 50% of islets die due to inflammation and apoptosis. This study investigated the effects of adenosine A2A receptor agonists (ATL146e and ATL313) on glucose-stimulated insulin secretion (GSIS) in vitro and transplanted murine syngeneic islet function in vivo. Compared to vehicle controls, ATL146e (100 nM) decreased insulin stimulation index [SI, (insulin)high glucose/(insulin)low glucose] (2.36 ± 0.22 vs. 3.75 ± 0.45; n = 9; p < 0.05). Coculture of islets with syngeneic leukocytes reduced SI (1.41 ± 0.17; p < 0.05), and this was restored by ATL treatment (2.57 ± 0.18; NS). Addition of a selective A2AAR antagonist abrogated ATL's protective effect, reducing SI (1.11 ± 0.42). ATL treatment of A2AAR+/+ islet/A2AAR-/- leukocyte cocultures failed to protect islet function (SI), implicating leukocytes as likely targets of A2AAR agonists. Diabetic recipient C57BL/6 mice (streptozotocin; 250 mg/kg, IP) received islet transplants to either the renal subcapsular or hepatic-intraportal site. Recipient mice receiving ATL therapy (ATL 146e or ATL313, 60 ng/kg/min, IP) achieved normoglycemia more rapidly than untreated recipients. Histological examination of grafts suggested reduced cellular necrosis, fibrosis, and lymphocyte infiltration in agonist-treated animals. Administration of adenosine A2A receptor agonists (ATL146e or ATL313) improves in vitro GSIS by an effect on leukocytes, and improves survival and functional engraftment of transplanted islets by inhibiting inflammatory islet damage in the peritransplant period, suggesting a potentially significant new strategy for reducing inflammatory islet loss in clinical transplantation.
Keywords
Introduction
Strategies to restore pancreatic islet β-cells lost to autoimmune destruction in type 1 diabetes mellitus (T1DM) have long been a focus in diabetes research. Successful islet cell transplantation using steroid-free immunosuppression (the “Edmonton protocol”) marked a major breakthrough in islet transplantation and rapidly became the standard of therapeutic immunosuppression in this field (74). Nevertheless, recent clinical studies have indicated that nonspecific inflammation in the peritransplant period mediates the early loss of approximately half of the transplanted islet mass (10, 72). However, the early loss of a majority of grafted islets soon after transplantation cannot be entirely attributed to allogeneic immune responsiveness, nor to autoimmune reactivity (recurrence autoimmune destruction of β-cells) inherent to T1DM, as these losses occur even following syngeneic transplantation (24, 50, 56).
Past reports suggest that innate host inflammatory reactions in the microenvironment of the graft contribute to early cellular loss within the transplanted islet mass (52, 54, 57, 70, 80). The detrimental effect of a rapid inflammatory reaction reasonably explains the peritransplant loss of a significant proportion of transplanted islet mass. Additionally, innate inflammatory responses likely promote alloantigen presentation, thereby strengthening and accelerating subsequent cell-mediated allogeneic immune responsiveness. To counter these inflammatory effects, administration of anti-inflammatory agents has been investigated (31, 32, 49, 51, 82) and adopted in the standard induction immunosuppressive therapy in several clinical trials (2, 16, 27). This study investigates the effect of reducing islet loss in the peritransplant period by inflammation-mediated processes through the inhibitory action of adenosine and adenosine analogs. There are four subtypes of G protein-coupled adenosine receptors; A1, A2A, A2B, and A3 (34, 44). Activation of A2A receptors powerfully inhibits inflammatory responsiveness and ischemia/reperfusion injury (IRI) (15, 25, 42, 46–48, 62, 75). A2A adenosine receptors are expressed on most hematogenous cells including neutrophils, macrophages, and T lymphocytes (14, 21, 36), and their activation inhibits oxidative burst in neutrophils (20, 79), cytokine release by monocytes/macrophages (26, 41, 45), lymphocyte activation (40), and neutrophil-endothelial cell adherence (13, 63). Thus, inhibition of inflammatory activity, and inflammation-mediated cell damage at the graft site, by an adenosine receptor agonist may prove to be an important adjunctive therapy to advance successful islet engraftment and transplant outcomes.
ATL146e and ATL 313 are potent and selective agonists of the adenosine A2A receptor subtype, and are powerful inhibitors of inflammatory cells including neutrophils, macrophages, monocytes, platelets, mast cells, eosinophils, and T cells (43, 48). As these agonists protect tissues from damage induced by inflammation and IRI at concentrations well below the threshold of hemodynamic responsiveness (4, 17, 22, 61, 71), it was reasoned that they could reduce islet loss and increase engrafted islet mass in the immediate posttransplant period, by inhibiting nonspecific inflammatory responses in the microenvironment of the grafted islets. By increasing the number of surviving islets, transplanted islet mass requisite to reverse hyperglycemia and achieve insulin independence would be reduced, allowing more efficient use of available donor islets. This study investigated the efficacy of ATL compounds in protecting grafted islets from early loss, and improving islet enraftment and survival, in a syngeneic murine islet transplant model. These studies also assessed ATL's in vitro effects on glucose responsiveness and insulin secretory capacity in isolated mouse islets, cultured alone or in the presence of neutrophil-enriched mixed leukocytes.
Materials and Methods
Animals
Female retired breeder C57BL/6 (Charles River Laboratories, Wilmington, MA) were used as islet donors for in vitro and in vivo transplant studies. A2A adenosine receptor knock-out (A2AAR-/-) mice (C57BL/6 background) were also used as islet donors. Female C57BL/6 (wild-type) and A2AAR-/- (C57BL/6 background) mice, 8–12 weeks old, served as transplant recipients. The A2AAR-/- (C57BL/6 background) mice were generated by breeding A2AKO mice (B6;129P-adora2atm1Jfc) to a C57BL/6 background by a speed congenic method. In brief, mice with disrupted adora2a were by interbred through several generations (129-Adora2atm1ifc x C57BL/6). A new line was initiated using a founder with a maximal number of informative loci (B6/129 or 129/129). By generation 5, the only detectable 129 contamination surrounded the Adora2a locus (on chromosome 10), and at present, the 129 genome can be detected only in a region between D10Mit31 and D10Mit42 (8cM). All animals were housed in institutional vivaria, and all procedures approved by the Animal Care and Use Committee.
Islet Isolation
Islets were isolated by collagenase digestion and density gradient centrifugation. Under anesthesia, the pancreas was exposed by midline laparotomy, the common bile duct identified and cannulated, and the pancreas distended [collagenase P, 2.5 ml, 2 mg/ml in Hank's balanced salt solution (HBSS); Roche Molecular Biochemicals, Indianapolis, IN]. The distended pancreas was excised, digested (13 min, 37°C), and shaken vigorously to mechanically disrupt the tissue (30 s). Digestion was quenched [excess cold HBSS with 10% fetal calf serum (FCS); Invitrogen, Grand Island, NY], and the digested tissue filtered through a nylon mesh (1,000 μm, Nitex; Fisher Scientific) and washed (3x, HBSS, 0°C). Islets were purified by density gradient centrifugation using Histopaque 1077 (Sigma-Aldrich, St. Louis, MO). Islets were washed and cultured (~18 h, 37°C, 5% CO2) in DMEM (Invitrogen) supplemented with 10% FCS, 2 mM l-glutamine, 1% penicillin/streptomycin, 0.1 mM MEM nonessential amino acids, and 25 mM HEPES buffer (all supplements from Invitrogen; 20 ml per plate), then hand-picked to high purity.
Mixed Leukocytes Isolation
Peripheral whole blood (2 ml) was collected in heparinized, siliconized tubes, and diluted with HBSS (2 ml). Diluted blood was layered onto Ficoll-Paque PLUS (3 ml, 15 ml tube), centrifuged (400 × g, 30 min, 20°C), supernatant discarded, and the leukocyte-containing pellet was treated with cold NaCl (0.2%, 20 ml, 30 s) to lyse RBCs (repeated as necessary). Leukocyte preparations were washed, resuspended in cold PBS, and counted by hemocytometer.
Glucose-Stimulated Insulin Secretion
Islets were hand-picked for purity and consistency in size, distributed to six-well plates (50 islets/well), and cultured for ~18 h in low-glucose (2.8 mM) media, with or without adenosine A2A receptor agonist (100 nM; ATL146e or ATL313). For coincubation experiments, islets were cocultured with neutrophil-enriched mixed leukocytes (1 × 105 leukocytes/well) for 2 h in the presence or absence of ATL (100 nM) and/or SCH58261 (A2AAR antagonist; 25 nM). Islets were washed (2x), suspended in Krebs-Ringer bicarbonate (KRB) buffer supplemented with either low (2.8 mM) or high (28 mM) glucose, incubated for 2 h (all incubation conditions performed in triplicate), and insulin content of collected samples assessed by insulin ELISA (mouse insulin ELISA kit, Mercodia Inc, Winston Salem, NC).
Islet Transplantation
For all transplants, islets were isolated, cultured overnight, aliquoted by hand-picking, and syngeneically transplanted. Islets from each islet batch were transplanted to both control and experimental groups to minimize potential differences in outcomes due to islet batch variation. Groups and treatments are outlined in Table 1. For histological studies, islets were transplanted to the renal subcapsule in aliquots of 300 islets per recipient for histological evaluation and assessment (Table 1). Islet-bearing kidneys were excised at 1, 3, 5, 7, or 9 days posttransplant. For functional studies, mice were rendered diabetic by a single IP injection of streptozotocin (STZ, 250 mg/kg; Sigma). For the renal subcapsular site, diabetic recipient mice were anesthetized, and the left kidney exposed through the left flank. Islets were placed into the subcapsular space (in aliquots of 150, 100, or 50 islets) through a small incision in the renal capsule. The kidney was returned to anatomical position, and the skin and muscle closed separately. For the hepatic-intraportal site, diabetic recipient mice were anesthetized, and the hepatic pedicle exposed by a midline laparotomy. Islets were suspended in HBSS in aliquots of 75, 150, or 300 islets, and gently injected into the exposed portal vein. Hemostatic gauze (Surgicel, Eth-icon Inc., Somerville, NJ) was used to minimize bleeding, and the skin and muscle closed separately.
Donor/Recipient Combinations and Therapeutic Regimens Used in This Study
The table displays the combinations of donor and recipient animals used in this study, and the therapeutic regimen, including the number of islets transplanted, and the dose, day of initiation, and duration of ATL146e therapy, administered to each experimental group.
Blood Glucose Monitoring
Recipient blood glucose levels (nonfasting) were monitored daily (AccuChek blood glucose meter, Roche), and diabetes was defined as blood glucose (BG) >300 mg/dl on 2 consecutive days, with cure defined as a return to normoglycemia (BG <200 mg/dl for 2 consecutive days). Reversion to hyperglycemia following left nephrectomy (30 days posttransplant) confirmed islet graft function.
Administration of Adenosine Agonists
ATL compounds (gifts of PGxHealth, LLC, Adenosine Therapeutics Group, Charlottesville, VA) were administered (60 ng/kg/min, 7 days, IP, beginning on day 0, -2, or -3 as indicated) by osmotic minipump (Alzet, Cupertino, CA). Pumps were primed in sterile saline (4 h) prior to implantation to insure immediate drug release. Dosages of ATL compounds used in these studies were empirically determined in transplant studies employing multiple dosing protocols. Initial doses were estimated from concentrations required to observe in vitro effects of ATL compounds. Studies commenced using ATL146e (Stedivaze™, apadenoson; scheduled to enter Phase III clinical trials in 2009 as a pharmacologic stress agent for myocardial perfusion imaging). However, licensing issues and exclusivity for clinical trials precluded its use in later studies. ATL313, a unique variant of ATL146e with improved solubility and a longer in vivo half-life, was generously provided for the continuation of these studies.
Histology
Islet-bearing kidneys were excised, fixed (10% neutral buffered formalin), prepared for routine paraffin histology, and sections stained with hematoxylin and eosin (H&E). Adjacent sections evaluated for apoptosis (TUNEL; Upstate, Lake Placid, NY). Images (40x) were captured using MRGrab Software (Zeiss, München-Hallbergmoos, Germany), and visualized (Photoshop v7.0; Adobe, San Jose, CA). Infiltrating cell density, graft necrosis, fibrosis, and lymphoid follicle size were observed.
Data Presentation and Analysis
GSIS data (secreted insulin levels or stimulation index) are presented as means ± SE. Statistical analyses were performed using a one-way ANOVA with a Tu-key's Multiple Comparison Test performed as a posttest. Outcomes of transplant studies are presented as median (and mean ± SE) time to cure. Statistical analyses were performed using the Wilcoxon Signed Rank Test and the Mantel-Cox Log-rank Test.
Results
In Vitro Studies
The direct effect of adenosine A2A receptor agonists on GSIS by isolated mouse islets (50 hand-picked islets/well) was examined in vitro using ATL146e. The SI [calculated as (insulin)high glucose/(insulin)low glucose] for islets alone was 3.75 ± 0.45 (n = 9). Islets cultured in media supplemented with ATL146e exhibited a stimulation index of 2.36 ± 0.22, which represented a statistically significant decrease. Notably, this decrease in SI resulted from increased basal insulin secretion by islets incubated in low-glucose ATL146e-supplemented KRB buffer (125.2 ± 44.3 ng/ml) versus low-glucose KRB buffer alone (61.7 ± 13.2 ng/ml, p < 0.05), and not from a decrease in stimulated insulin secretion, which remained relatively constant with or without ATL (242.3 ± 52.6 ng/ml vs. 203.4 ± 33.1 ng/ml, NS) (Fig. 1).
Levels of insulin secreted in response to static 2-h incubation of 50 isolated, hand-picked murine islets in either low (2.8 mM) or high (28.0 mM) glucose, with or without ATL146e (100 nM). *Compared to mean insulin secretion of untreated islets in low glucose; ‡compared to mean insulin secretion of ATL-treated islets in low glucose.
To assess the effect of inflammatory cells on insulin secretion, islets cocultured with syngeneic leukocytes (50 islets plus 1 × 105 leukocytes/well) were challenged by GSIS. Insulin assay demonstrated that SI was inhibited by leukocytes (2.28 ± 0.17 for ATL-treated islets vs. 1.34 ± 0.10 for islets plus leukocytes; p < 0.05). Addition of ATL146e to culture media preserved islet insulin secretory capacity in islet/leukocyte cocultures, restoring SI to control values of islets treated with ATL146e (2.28 ± 0.17 for islets plus ATL146e vs. 2.58 ± 0.14 for islets plus leukocytes plus ATL146e; NS) (Fig. 2). Levels of insulin secretion in low-glucose conditions were not statistically different, while insulin level in islet/leukocyte cocultures exposed to high-glucose conditions showed a significant decrease compared to all other high-glucose conditions.
Insulin stimulation index (SI) calculated from static incubations of murine islets in low (2.8 mM) or high (28.0 mM) glucose, cultured with or without ATL146e (100 nM), in the presence or absence of mixed leukocytes (1 × 105 per well). *Compared to mean SI of ATL-treated islets; †compared to mean SI of untreated islets cocultured with leukocytes; ‡compared to mean SI of islet/leukocyte cocultures treated with ATL146e.
To evaluate specificity of ATL146e action, the A2AAR-selective antagonist SCH58261 was added to islet/leukocyte cocultures. Treatment of islet/leukocyte cocultures with SCH58261 (25 nM final concentration) abrogated the protective effect of ATL146e, causing a drop in SI to the level of islets cocultured with mixed leukocytes (2.58 ± 0.14 for islets, leukocytes, and ATL146e vs. 1.11 ± 0.42 for islets, leukocytes, ATL146e, and SCH58261; p < 0.05) (Fig. 2). This result supports the hypothesis that islet secretory function is protected by the specific activation of A2AAR by ATL146e.
To determine if ATL146e was acting through A2A receptors on islets or effector mixed leukocytes, islets were cocultured with mixed leukocytes derived from A2AAR-/- (A2AAR-KO) mice. Islets (wild-type C57BL/6, A2AAR+/+) and leukocytes (A2AAR-KO) were cocultured and glucose challenged, as above. Stimulation index calculated from insulin levels in wells containing A2AAR+/+ islets and A2AAR-/- leukocytes was significantly reduced (SI 1.24 ± 0.40) compared to control C57BL/6 islets alone (3.32 ± 0.40; p < 0.05). The addition of ATL146e to cocultures of islets and A2AAR-/- leukocytes did not alter insulin levels nor insulin SI compared to identically challenged wells in the absence of ATL (1.24 ± 0.40 for islet/A2AAR-KO leukocyte cocultures vs. 0.90 ± 0.10 for islet/A2AAR-KO leukocyte cocultures treated with ATL; NS) (Fig. 3).
Insulin SI in response to static 2-h incubation of C57BL/6 murine islets in low (2.8 mM) or high (28.0 mM) glucose, cultured with or without ATL146e (100nM), in the presence or absence of A2AAR-KO (C57BL/6 background) mixed leukocytes (1 × 105 per well). *Compared to mean SI of untreated islets.
Transplant Studies
In histological islet transplant studies, islet graft-bearing kidneys were harvested at specific time points and examined using H&E staining (for general graft morphology and cell identify infiltrating), and TUNEL staining (to identify apoptotic cells). Infiltrating neutrophils were observed in islet graft biopsies 24 h following transplantation (posttransplant day 1), with neutrophil cell density noticeably greater than that observed on posttransplant day 3. By posttransplant day 5, lymphocytes (and lymph nodules) were also noted within islet grafts (H&E) (Fig. 4). Increased apoptosis was observed within grafts beginning on day 3, with reduced apoptosis observed by day 9 (TUNEL) (Fig. 4). In contrast, recipients dosed with A2AAR agonists (60 ng/kg/min, IP) demonstrated marked reductions in both infiltrating neutrophils and lymphocytes (H&E) (Fig. 4). The area of tissue necrosis and degree of apoptosis (TUNEL) also appeared markedly reduced (n = 3) (Fig. 4). Although A2AAR binding has been reported to stimulate collagen production in fibroblasts (69), no histological evidence of increased fibrosis at the graft site of animals treated with ATL146e was observed in these studies.
Grafts of 300 syngeneic islets, transplanted to untreated nondiabetic recipients (A) or to nondiabetic recipients treated with ATL146e (B), were harvested 1, 3, 5, 7, or 9 days posttransplant. In the absence of ATL146e therapy, grafts harvested on day 1 and day 3 exhibited extensive necrosis within the graft, with the development of graft fibrosis and appearance of lymphatic nodules by day 5 (H&E). Extensive apoptosis was revealed within the region of the islet graft beginning on day 3, with resolution and reduction of apoptosis observed by day 9 (TUNEL). In contrast, grafts from recipients receiving ATL therapy (60 ng/kg/min, IP) reduced necrosis and fibrosis, minimal lymphocyte infiltration, and little to no lymphatic nodule development were observed (H&E). Additionally, apoptosis, as assessed by TUNEL staining, was markedly absent, suggesting that systemic ATL146e therapy in the peritransplant period reduced inflammatory cell infiltration, and reduced the incidence of apoptosis within the graft.
Transplant studies to reverse STZ-induced diabetes were initiated using ATL146e (10 ng/kg/min) for 7 days, beginning on posttransplant day 0. All recipients of 150 syngeneic islets alone achieved normoglycemia within 24 days (n = 6), while transplantation of 100 islets to diabetic mice, in the absence of ATL146e, failed to cure diabetes. All recipients of 150 or 100 islets, treated with ATL146e (at the above dose), achieved normoglycemia within 14 days. Transplantation of 50 islets, either alone or with ATL therapy (10 ng/kg/min, 7 days, beginning posttransplant day 0), failed to reverse diabetes in any recipient (data not shown). In these studies, untransplanted STZ-diabetic mice receiving ATL146e (n = 8) or ATL 313 (n = 5) were never observed to revert to normoglycemia.
As the half-life of ATL146e in systemic circulation is extremely short, it was hypothesized that increasing ATL146e dose (60 ng/kg/min; 7 days), and initiating therapy earlier (beginning transplant day -1), might improve islet transplant success. In fact, normoglycemia was achieved within 1 day of transplantation of either 150 or 100 syngeneic islets to diabetic mice, in recipients receiving ATL146e therapy (60 ng/kg/min, 7 days, beginning posttransplant day -1). However, transplantation of 50 islets to diabetic recipients under this ATL regimen resulted in only a transient improvement in glycemic status, with normoglycemia being temporarily achieved in only one recipient (return to hyperglycemia within 24 h) (Fig. 5A). Interestingly, transplantation of 50 islets to diabetic mice dosed with ATL146e (60 ng/kg/min, 7 days) beginning posttransplant day -3 resulted in normoglycemia in all mice within 5 days of transplantation (Fig. 5B), suggesting that up to 72 h of ATL146e therapy was required to achieve full anti-inflammatory efficacy, perhaps by achieving optimal circulating ATL blood levels, or perhaps by inhibiting nominal inflammatory signaling by reducing local cytokine levels. Although bioassay of islet recipients' serum (ATL-treated and untreated recipients) for a panel of inflammatory cytokines yielded circulating levels below detectable values, cytokine expression in the microenvironment of the graft is likely substantially different.
Graph displays transplant outcomes for transplantation of 50 islets to syngeneic diabetic (STZ-induced) recipients receiving ATL therapy (60 ng/kg/min, for 7 days, beginning on transplant day -1, dashed lines), and in recipients not receiving ATL therapy (A, solid lines). ATL146e was administered by osmotic minipump, and control (untreated sham) recipients received minipumps loaded with saline. In grafts of 50 islets, graft outcomes were improved when ATL therapy was initiated on transplant day -3 (B, 3 days prior to transplantation). The administration of ATL146e resulted in improved islet graft outcomes, allowing accelerated rate of cure (a more rapid reversal of STZ-induced hyperglycemia), and successful islet transplantation with a small number of donor islets.
To assess the role of A2AAR in protecting and improving islet graft function, islet transplants to the renal subcapsular site were performed using A2AAR-/- recipients. As previously, STZ-diabetic recipients received grafts of 50 C57BL/6 islets (syngeneic to the recipients' genetic background). ATL therapy significantly improved results of syngeneic islet transplantation between C57BL/6 donors and recipients (Table 2, Fig. 6A); however, no significant improvement in graft outcome (reduced time to normoglycemia) was observed between C57BL/6 recipients of syngeneic islets and A2AAR-/- recipients of congenic C57BL/6 islets. Additionally, no difference was observed between A2AAR-/- recipients of congenic C57BL/6 islets with or without administration of ATL146e therapy (60 ng/kg/min, 7 days, posttransplant day -3) (Table 2, Fig. 6A, B).
Kaplan-Meier curves comparing transplant outcomes for transplantation of islets to untreated diabetic recipients, and to diabetic recipients receiving ATL therapy. (A) Comparison of syngeneic transplantation to the renal subcapsular site in C57BL/6 mice (50 islets; ATL therapy at 60 ng/kg/min for 7days, beginning on transplant day -2). (B) Comparison of transplantation of C57BL/6 islets to the renal subcapsule of congenic A2AAR-KO recipients (50 islets; ATL therapy at 60 ng/kg/min for 7 days, beginning on transplant day -2). (C) Compararison of transplantation of C57BL/6 islets to the hepatic intraportal site in A2AAR-KO congenic mice (300 islets; ATL therapy at 60 ng/kg/min for 7 days, beginning on transplant day -3), and demonstrating a significant improvement in outcome with ATL therapy. (D) Comparison of transplantation of C57BL/6 islets to the hepatic intraportal site in A2AAR-KO congenic mice (150 islets; ATL therapy at 60 ng/kg/min for 7 days, beginning on transplant day -3).
Syngeneic Transplantation of Wild-Type C57BL/6 Islets to Wild-Type and to BL/6-A2AAR-KO Congenic Recipients, With and Without ATL146e Therapy
The table shows the outcomes of transplantation of 50 C57BL/6 wild-type islets to syngeneic or congenic BL/6-A2AAR-KO (A2A adenosine receptor knockout mice with a C57BL/6 genetic background) recipient mice. Recipients received either no ATL146e therapy, or IP infusion of ATL146e (60 ng/kg/min, 7 days beginning posttransplant day -3) by osmotic minipump. Animals' blood glucose values were monitored daily from the time of induction of diabetes by streptozotocin injection (250 mg/kg body weight, IP). ATL therapy exerted a significant effect on islet graft success in wild-type C57BL/6 recipients of syngeneic C57BL/6 islets, as assessed by the interval of time from transplantation to restoration of recipient normoglycemia (time to cure). In contrast, ATL therapy had no effect on islet graft success, as assessed by time to cure, in BL/6-A2AAR-KO recipients of syngeneic C57BL/6 islets.
Compared to untreated syngeneic controls.
Compared to untreated knock-out congeneic controls.
To assess the effect of A2AAR agonist (ATL313) on islet transplantation to a clinically relevant site, isolated islets (hand-picked to high purity) were transplanted to the hepatic-intraportal site by embolization into the portal vein (IPO). As previously, a group of untransplanted subject animals was monitored following STZ injection, and spontaneous reversion to normoglycemia was not observed in either untreated mice or in ATL313-treated mice (data not shown). All recipients of 300 syngeneic islets transplanted to the hepatic-intraportal site (n = 6) achieved normoglycemia within 19 days, with a mean time to cure of 15.7 ± 1.6 days, while administration of ATL313 (60 ng/kg/min, 7 days, posttransplant day -2; n = 6) restored normoglycemia in all recipients within 12 days posttransplantation (mean time to NG 8.9 ± 1.4 days, p < 0.05) (Table 1, Fig. 6C). In IPO recipients of 150 syngeneic islets, normoglycemia was achieved within 24 days, with a mean time to cure of 19.5 ± 1.3 days. Recipients of 150 islets to the IPO site, treated with ATL313 (n = 6) achieved normoglycemia within 21 days, with a mean time to cure of 17.2 ± 1.4 (p = 0.26, not significant compared to untreated recipients of 150 islets). Transplantation of 75 syngeneic islets failed to reverse hyperglycemia in any IPO recipient. However, blood glucose levels remained within a measurable range (<600 mg/dl) for longer periods in recipients treated with ATL313 compared to untreated recipients (data not shown).
Discussion
Significant advances have been made in the field of clinical islet transplantation in the past decade (12, 38, 58, 74). However, recipients continue to require infusions of islets derived from multiple pancreata to achieve insulin independence, even after transplantation of a large number of islets from a single pancreas [>5,000 islet equivalents (IEQ); islet volume standardized to a diameter of 150 μm/kg recipient body weight]. In considering the factors that constrain successful clinical islet transplantation, the emergence of this procedure as a standard of care in diabetes mellitus requires, among other protocol enhancements, improved functional engraftment with minimal islet loss following transplantation.
Inflammatory activity in the microenvironment of the islet graft represents a key obstacle to successful islet transplantation. The majority of grafted islets are lost in the peritransplant period by a mechanism that cannot be accounted for by allogeneic immunity, or recurrent autoimmune activity (18, 35, 72), as early loss of grafted islets occurs in experimental models of both syngeneic and allogeneic islet transplantation (8, 10, 24, 35, 52, 54, 57, 72, 73). Moreover, this early and rapid graft loss is rarely observed in whole pancreas transplantation (76), suggesting islet loss in islet transplantation cannot be attributed to either allo- or autoimmune processes. Thus, strategies to effectively inhibit inflammatory cascades in the peritransplant period will greatly impact clinical islet transplant outcomes.
Isolated human islets express many genes/products that play a role in inflammatory responses, and that would likely promote antigen presentation, accelerating and intensifying a subsequent cell-mediated immune response (29, 68). Expression of cytokines that stimulate inflammation at the graft site, such as IL-1β, IL-8, MIP-2, TNF-α, MCP-1, and NO, are increased significantly following islet isolation, and islets are highly sensitive to these (pro)inflammatory mediators in the microenvironment of the graft (3, 67, 72).
Intraportal embolization of isolated human islets has been shown to stimulate an instantaneous blood-mediated inflammatory reaction (IBMIR), described as an immediate clotting event, potentially resulting from the production, by islet cells, of tissue factor (a strong trigger of coagulation) (6, 7, 53). IBMIR induces a rapid infiltration of neutrophils and monocytes (via increased endothelial adherence and migration) into underlying tissues in response to chemoattractants released by damaged tissues (a hallmark of acute inflammation). Suppression of IBMIR in preclinical islet transplant models has been achieved using anticoagulants (e.g., heparin, melagatran) and/or ex vivo pretransplant treatment with antioxidants (23, 55, 65). Notably, A2AAR agonists also strongly inhibit adherence of stimulated (but not unstimulated) neutrophils (13, 19), suppress secretion of pro-inflammatory mediators such as TNF-α (30, 37, 64, 77), inhibit reactive oxidative species, and inhibit cAMP-mediated release of preformed nonoxidative products from neutrophil lysosomes (1, 5, 19, 33, 78, 81).
Infiltration of neutrophils and monocytes contributes to tissue injury by disrupting islet morphology and triggering islet cell death by direct cytotoxicity or apoptotic induction. Lymphocytes (particularly NKT cells) are also activated, triggering an inflammatory cascade enhancing neutrophil recruitment. Findings suggest that the rapid inflammatory damage in ischemia reperfusion injury in the liver, and possibly in the kidney, is initiated through the activation of NKT cells. Further, the activation of these cells is inhibited by A2AAR activation (39, 83). Conversely, preactivation of NKT cells has been shown to protect the liver from ischemia reperfusion injury through an A2AAR-mediated pathway (11).
Transplantation of C57BL/6 Islets to the Hepatic-Intraportal Site of Syngeneic Recipients, With and Without ATL313 Therapy
Outcomes of intrahepatic portal vein (IPO) transplantation of C57BL/6 islets to syngeneic recipient mice. Recipient animals received either no ATL313 therapy, or IP infusion of ATL313 (60 ng/kg/min, 7 days beginning posttransplant day -2) by osmotic minipump. Animals' blood glucose values were monitored daily from the time of diabetes induction (streptozotocin injection; 250 mg/kg body weight, IP). ATL therapy exerted a significant effect on islet graft success in IPO recipients of 300 syngeneic C57BL/6 islets, as assessed by the interval of time from transplant to restoration of recipient normoglycemia (time to cure). In contrast, ATL therapy had no significant effect on islet graft success, as assessed by time to cure, in IPO recipients of 75 or 150 syngeneic C57BL/6 islets. While restoration of normoglycemia was not achieved in any IPO recipient of 75 syngeneic islets, retention of hyperglycemia within a detectable range (<600 mg/dl) was prolonged in recipients treated with ATL313 compared to untreated recipients.
Compared to untreated syngeneic IPO recipients of 300 islets.
Compared to untreated syngeneic IPO recipients of 150 islets.
In these studies, in vitro assessments of ATL effects on islets demonstrated an increased basal insulin secretion in low glucose, but no effect on glucose-stimulated insulin secretion. ATL's effects were blocked by the A2AAR selective antagonist SCH58261, supporting the hypothesis that these A2AAR agonists act through ATL/A2AAR receptor binding. Additionally, the hypothesis that ATL/A2AAR binding on inflammatory cells would inhibit inflammation-mediated islet cell damage and dysfunction was supported by the observed failure of ATL146e to protect in vitro insulin secretion by islets cocultured with A2AAR-/- leukocytes, suggesting limited endogenous protection in this model (i.e., neither ATL nor A2AAR are constitutively active without their appropriate ligand). Thus, ATL compounds exert their protective effects by selectively activating A2A receptors, and not other adenosine receptor subtypes.
These studies suggest that hematogenous cells express A2A receptors and that A2AAR-mediated signaling inhibits cell migration, activation, and proliferation. In syngeneic islet transplantation, administration of ATL146e appeared to reduce immune cell infiltration cells, and the size and number of lymphoid nodules. The ATL compounds used in these studies are synthetic agonists and display greater potency and a longer half-life than endogenous adenosine. An extensive literature describes the numerous physiologic roles of adenosine, which are usually manifest under conditions of hypoxia, ischemia, or tissue injury. These conditions provoke the release of adenosine or adenine nucleotides, which are very rapidly metabolized by adenosine deaminase and adenosine kinase. These results correlate with reports that even low concentrations of adenosine inhibit T-cell receptor-triggered proliferation and upregulation of CD25 (IL-2 receptor chain) in lymphocytes (28), and that these effects are selectively mediated by A2A receptors (9, 40). Conversely, the suppression of neutrophil activity inhibits lymphocyte activity through neutrophil-mediated modulation of lymphocyte responsiveness (66). Although adenosine receptor agonist administration to prediabetic mice has been reported to prevent diabetes onset, these effects were not mediated through A2AAR receptors (59), and A2AAR-selective agonists (ATL146e and ATL313) failed in these studies, to prevent/ameliorate STZ-induced diabetes. Finally, recent studies following the same line of inquiry as outlined in this report have demonstrated that the administration of adenosine to murine recipients of intrahepatic islet transplants, at the time of transplantation, suppresses NKT cell-mediated IFN-γ production by neutrophils, and prevents early islet loss of transplanted islets (60).
The Edmonton protocol introduced a corticosteroid-free, and less diabetogenic, immunosuppressive regimen (sirolimus, low-dose tacrolimus, induction with anti-IL-2 receptor antibody) for pancreatic islet transplantation, which produced prolonged islet graft function and improved rates of insulin independence. However, early islet loss mediated by innate immune activity continues to present a major obstacle to long-term functional islet graft success, despite modifying the induction regimen to include the anti-inflammatory agent etanercept (9–12).
The studies reported here demonstrate that A2A receptor agonist infusion in the peritransplant period reduces inflammation-mediated early cell loss in pancreatic islet transplantation. Therapeutic administration of ATL compounds reduced the islet dose required to achieve sustained normoglycemia in syngeneic murine islet transplants, including at the clinically relevant intrahepatic site. In islet transplant protocols that minimize systemic immunosuppression, including a short course of an A2AAR agonist as part of induction therapy, could minimize early peritransplant islet loss, thereby improving islet graft outcomes. Additionally, by reducing the number of islets required for successful clinical islet transplantation, a greater efficiency in the use of human islets for clinical transplantation could lead to the availability of this procedure to a greater number of potential recipients. Ongoing studies, using a murine islet allotransplant model, will continue to evaluate the benefit of anti-inflammatory A2A receptor agonists as adjuncts to clinical immune suppressive regimen, in minimizing excessive peritransplant islet loss.
Footnotes
Acknowledgments
Funding for this project was provided by the generosity of Paul and Diane Manning and the Focus to Cure Diabetes Foundation, and by grants from the National Institutes of Health (R01 HL37942), and the Juvenile Diabetes Research Foundation (5-2006-997, 1-2008-79). The authors acknowledge the kind gift of the ATL146e and ATL 313 compounds used in these experiments, which were provided by Adenosine Therapeutics Group (Charlottesville, VA), of PGxHealth LLC, a Division of Clinical Data, Inc. The authors would also like to acknowledge the assistance of Dr. George Stukenborg, Associate Professor of Public Health at the University of Virginia, for his advice on the statistical analyses presented in this work. Dr. Joel Linden conducts research supported by, and involving technology owned by, the Adenosine Therapeutics Group (Charlottesville, VA), of PGxHealth LLC, a Division of Clinical Data, Inc., and also receives income from the Adenosine Therapeutics Group. Additionally, Dr. Kenneth Brayman has stock and/or stock options, in Adenosine Therapeutics Group (PGxHealth LLC, a Division of Clinical Data, Inc.).