Resolving the Conundrum of Islet Transplantation by Linking Metabolic Dysregulation, Inflammation, and Immune Regulation

Abstract

Although type 1 diabetes cannot be prevented or reversed, replacement of insulin production by transplantation of the pancreas or pancreatic islets represents a definitive solution. At present, transplantation can restore euglycemia, but this restoration is short-lived, requires islets from multiple donors, and necessitates lifelong immunosuppression. An emerging paradigm in transplantation and autoimmunity indicates that systemic inflammation contributes to tissue injury while disrupting immune tolerance. We identify multiple barriers to successful islet transplantation, each of which either contributes to the inflammatory state or is augmented by it. To optimize islet transplantation for diabetes reversal, we suggest that targeting these interacting barriers and the accompanying inflammation may represent an improved approach to achieve successful clinical islet transplantation by enhancing islet survival, regeneration or neogenesis potential, and tolerance induction. Overall, we consider the proinflammatory effects of important technical, immunological, and metabolic barriers including: 1) islet isolation and transplantation, including selection of implantation site; 2) recurrent autoimmunity, alloimmune rejection, and unique features of the autoimmune-prone immune system; and 3) the deranged metabolism of the islet transplant recipient. Consideration of these themes reveals that each is interrelated to and exacerbated by the other and that this connection is mediated by a systemic inflammatory state. This inflammatory state may form the central barrier to successful islet transplantation. Overall, there remains substantial promise in islet transplantation with several avenues of ongoing promising research. This review focuses on interactions between the technical, immunological, and metabolic barriers that must be overcome to optimize the success of this important therapeutic approach.

• I. Introduction

• II. Barriers to Successful Clinical Islet Transplantation

• III. The Role of the Innate Immune Response in Clinical Islet Transplantation

  • A. Overview
  • B. Initiation of the innate response
  • C. Improving islet survival immediately after transplant
• IV. New Strategies to Achieve Tolerance in the Presence of Autoimmunity

  • A. Overview
  • B. Development and function of regulatory T cells
  • C. Limitations to the clinical application of tregs and the role of innate immunity
• V. Inflammation and the Emerging Role of Dysmetabolism

  • A. Dysmetabolism drives inflammation
  • B. Management and assessment of the metabolism of the islet transplant recipient

• VI. Summary

I. Introduction

TYPE 1 DIABETES MELLITUS (T1D) results from autoimmune destruction of the insulin-secreting pancreatic β-cells by the coordinated interaction of CD4⁺ and CD8⁺ T cells, B lymphocytes, dendritic cells, and macrophages, all of which infiltrate the islets (1,2,3,4,5). The current standard of therapy requires lifelong exogenous insulin administration by either insulin pump or multiple daily injections (2). However, this therapeutic approach fails to achieve physiological control of blood glucose levels, and this failure results in microvascular complications in a high percentage of patients. Islet transplantation can improve metabolic control of blood glucose to an extent that has not been possible using injectable insulin, even when the insulin is delivered on a continuous basis. Based on the many reports of successful clinical islet auto- and allotransplantation trials since 1990, including adoption of the Edmonton protocol, which markedly increased the rate of insulin independence (6,7,8), islet transplantation remains a definitive treatment option for T1D. Despite the initial success and early enthusiasm for this procedure, recent reports demonstrate that long-term insulin independence in islet transplant recipients is frequently lost by 5 yr of follow-up (9,10). Nevertheless, the positive results reported in these studies have stimulated efforts to improve long-term islet transplantation outcomes.

Both short- and long-term loss of graft function and immune rejection are barriers to the success of islet transplantation (2,11,12). These barriers may be accentuated by characteristics unique to islet transplantation. First, islet transplantation represents a graft of nonvascularized cells/cell clusters. Second, transplanted islets are placed in a heterotopic location, a site other than the natural location of the tissue (12). Furthermore, the grafted islets are not only subject to allogeneic immune rejection after transplantation but are also exposed to the autoimmune process that initiated the original disease. Pathogenic T cells and autoantibodies preexist in T1D patients; thus islets, when they are transplanted to a recipient with T1D, represent a cellular transplant to a presensitized host (9). Finally, when compared with solid organ transplantation, the isolated islets may be subjected to longer ischemic times, additional damage resulting from enzymatic activity (both autologous pancreatic enzymes and collagenase used in the preparative process), and to other process-related deleterious effects (13,14). In addition to the unique technological and immunological features that accompany islet transplantation, recipient characteristics such as labile blood sugars and chronic hyperglycemia may drive a systemic inflammatory response that creates an environment that is hostile to islet engraftment (15,16,17,18).

Despite the challenging features of this process, islet transplantation offers a remarkable cellular transplant paradigm for the development of new tolerogenic strategies that can be rapidly moved to the clinical setting. The nature of the islet isolation process allows for the ex vivo culture of islets, providing the opportunity to modify islet immunogenicity by a variety of techniques, such as antigen-presenting cell depletion (19,20) or the introduction of specific immunomodulating cytokines or chemokines by gene transfer techniques (21,22). In the same way that islets can be prepared before transplantation, the deranged metabolism of the islet recipient, which may be directly toxic to islet β-cells, can also be improved before and after transplantation. To extend and improve the success of this procedure, a new understanding and approach to the patient with diabetes is likely to be required based on the unique immunological and metabolic challenges presented by the patient with T1D.

This review will focus on the hypothesis that each of the identified barriers to islet transplantation (technical, immunological, and metabolic) results in an accentuated state of inflammation that is detrimental to islet health (Fig. 1). This inflammatory state further strengthens each of the identified barriers, engendering a vicious cycle that creates extreme resistance to islet engraftment and long-term survival. This review will discuss this challenge in terms of its components and identify biological, immunological, and metabolic approaches to improve the success of clinical islet transplantation.

Figure 1.

Overview of the interaction between the barriers to islet transplantation. Significant barriers to islet transplantation into the individual with diabetes include the underlying immune dysregulation, the presence of alloreactive lymphocytes that mediate graft rejection, ongoing islet cell destruction of both native islets and graft, and chronic hyperglycemia. Each of these factors either contributes to or is exacerbated by an underlying state of chronic inflammation. In addition, the traumatic nature of the transplant procedure may directly cause inflammation that will further feed the disruptive cycle.

II. Barriers to Successful Clinical Islet Transplantation

Clinical islet transplantation at established centers has resulted in approximately 80% of recipients achieving insulin independence at 1 yr after transplant. However, insulin independence falls to approximately 10% after 5 yr, although 80% of recipients continue to produce C-peptide as recently reported (23,24). Despite efficient immunosuppressive regimens that use a combination of daclizumab, sirolimus, and tacrolimus (25), a significant number of transplanted islets are lost in the first 10–14 d after transplantation (26,27). In experimental models of islet transplantation, loss of 60% of the transplanted syngeneic islet mass in the peritransplant period has been reported, indicating the sensitivity of islet-cell transplants to early damage (11,28).

Nevertheless, several reports of long-term clinical islet allograft survival and function have demonstrated the promise of islet transplantation to normalize blood glucose, return physiological glycemic control, and reduce the risk of severe episodes of hypoglycemia (7,23,24). These reports suggest that a large islet mass (numbers in excess of 10,000 islet equivalents (IEQ)/kg recipient body weight) derived from multiple (as many as two to four) donors is necessary to achieve insulin independence (24). This requirement suggests that islet transplantation under the Edmonton protocol may not be clinically practical given the limited number of donor pancreata. To make islet transplantation feasible, it is critical to maximize the number of islets recovered from every islet donor and also enhance islet engraftment and posttransplant survival.

Current challenges to highly efficient islet cell purification and successful transplantation include: 1) the isolation and characterization of islets for transplantation, under conditions that preserve islet function; 2) the loss of functional islet mass resulting from inflammatory responses, autoimmune-mediated islet destruction, and alloimmune rejection; 3) the failure of the newly transplanted islets to successfully revascularize, resulting in primary non-function-related to inefficient engraftment of islet cells; 4) transplantation of an insufficient islet mass coupled with high metabolic demand on islets grafted to a diabetic environment; and 5) the limited supply of pancreata for use in producing adequate numbers of islets (see Fig. 2 for overview) (6,29,30,31).

Figure 2.

Obstacles in the process of islet transplantation. An overview of the technical and immunological steps in islet transplantation is presented. For each process, a brief list of the known obstacles is given.

In general, two major issues contribute heavily to islet loss after intraportal transplantation. First, the innate immune response creates an environment that is not conducive for islet cell survival. Pancreatic islets are exposed to substantial oxidative stress during organ procurement (32), organ processing (islet isolation and purification), islet culture, and islet transplantation (11,33). Multiple signaling pathways can be triggered during these procedures, which can lead to cytokine-mediated β-cell injury and death. Many of the cytokines, which can be produced by the inflammatory cells that migrate to the site of the islet graft, are capable of triggering islet loss by both cell necrosis and apoptosis (34,35). Increasing attention is required to define the role of innate immunity and the inflammatory response, a factor that now appears to significantly affect the functional survival of transplanted islets (36).

Second, the adaptive immune response, and specifically allograft rejection, has been the traditional focus of immunosuppressive therapies in transplantation. Current immunosuppressive protocols (25) are able to effectively control acute rejection in whole organ transplantation. Even in islet-cell transplantation, alloimmunity seems to be controlled by conventional immunosuppression as evidenced by continued C-peptide production in approximately 85% of islet recipients 5 yr after islet transplantation. Reduced islet graft rejection in comparison to transplantation of the whole pancreas might be expected, due to a decrease in the mass/volume of tissue transplanted, which would represent an overall reduction in graft antigen load particularly in major histocompatibility complex (MHC) class II antigen (37). It is unclear whether loss of islet function after intraportal infusion results predominately from chronic alloimmune rejection, from toxicity of the immunosuppressive agent(s), from recurrence of the autoimmune reactivity that initially destroyed the β-cell, or from metabolic exhaustion. A clearer understanding of the pathways involved in the chronic degradation of islet graft function would allow targeting of specific steps, which could inhibit chronic loss of islet function. The control of the adaptive immune response, which mediates autoimmunity and allograft rejection, is discussed further in Section IV.A.

III. The Role of the Innate Immune Response in Clinical Islet Transplantation

A. Overview

Although prevention of islet allograft rejection has been a focus of many investigative teams since islet transplantation was first described in 1972–73 (38,39), primary nonfunction of a newly transplanted islet graft appears to result from events that are not related to classic cell-mediated immune responses commonly associated with acute rejection of whole organ transplants. Factors that influence primary nonfunction include: islet quality, innate immunity—which induces apoptosis and/or necrosis, coagulation, and complement fixation—and hypoxic injury due to lack of islet vascularity (40,41). Additionally, islets are subjected to high metabolic stress after transplantation due to the toxic (including diabetogenic) effects of the majority of immunosuppressive agents and increased physiological demand as a consequence of a suboptimal islet mass transplanted in a diabetogenic milieu (hyperglycemia, inflammation, oxidant stress, and insulin deficiency).

Nonspecific innate immune (inflammatory) responses, thought to be part of a primitive response to infection or invasion, also occur as a consequence of damage to tissues. The function of the innate immune response, its interaction with adaptive immunity, and the resultant local and systemic inflammatory response represent a central theme within this review. The detailed cellular and molecular mechanisms and their physiological consequences will be discussed throughout the text. In brief, the inflammatory response results in the production and release of a number of proinflammatory cytokines that act as chemoattractants and activate inflammatory cells (9,40,42,43). These cytokines include TNF-α, IL-1β, IFN-γ, and others all of which have been implicated in the pathogenesis of autoimmune diabetes and are known to exert deleterious effects on islet β-cell function by inducing apoptosis and cell death. In fact, cytokine production may be the end-stage mechanism by which islet β-cells are eliminated during spontaneous diabetes development (44). Nonetheless, attempts to target these inflammatory cytokines with individual neutralizing therapies have been routinely unsuccessful. This failure may be in part due to the systemic nature of inflammation. Although high levels of islet-toxic cytokines are obvious features of the diabetogenic state, the activation of the inflammatory cascade involves the coordinated induction of an entire transcriptional program encompassing many hundred gene products (45). Thus, targeting the induction of the inflammatory program may be more efficacious than targeting its downstream effectors and may serve to alleviate many of the factors that complicate islet transplantation including early islet loss (Section III.C), activation of adaptive immunity (Section IV.A), resistance to tolerance induction (Section IV.C.3), and chronic insulin resistance with hyperglycemia (see Section V.A).

The pancreatic islet may be uniquely susceptible to inflammatory injury because each islet contains a significant number of native inflammatory cells (macrophages, dendritic cells, nature killer T cells, etc.). These inflammatory cells may be especially numerous because islets are extremely well perfused in vivo, receiving 5 to 15% of the total blood flow to the pancreas although they constitute less than 1% of the pancreas by weight. This design also makes the islet more sensitive to injury from nutrient and oxygen deprivation and from ischemia/reperfusion (46,47,48), and this injury may provoke the innate immune response. During organ procurement and cold storage, the rich vascular network is also disrupted, compromising viability of the whole organ and significantly limiting survival of cells at the core of the islet (32), a site that may be rich in insulin-producing β-cells.

B. Initiation of the innate response

1. Instant blood-mediated inflammatory reaction (IBMIR).

After isolation, purified islets are most commonly transplanted to the liver parenchyma by embolization into the hepatic portal vein. Although intraportal transplantation has been the most commonly used approach in clinical islet transplant studies, certain features of intraportal transplantation result in significant damage to the transplanted islets. The potential benefit of exposure to the nutrient-rich blood of the portal vein is tempered by inflammatory reactions initiated when islets come into direct contact with blood elements. Previous studies have demonstrated a negative effect on islet cell viability upon exposure to blood or blood elements such as monocytes/macrophages, as well as platelets and complements proteins. This thrombocytic/inflammatory reaction has been more recently termed the IBMIR (49). IBMIR is characterized by the rapid activation and binding of platelets, activation of the coagulation and complement systems, and by the rapid infiltration of leukocytes (mainly CD11+) into the islets. This response occurs within 5 min of islet transplantation (36) and has been observed in vitro in both allogeneic and syngeneic donor/recipient combinations in rodent models (50).

More recent evidence suggests that the IBMIR response is triggered by inflammatory mediators such as tissue factor (TF) and monocyte chemoattractant protein 1 (MCP-1) (51). These molecules function in coagulation and are normally expressed at the cell surface of islet β-cells and α-cells (52). TF functions to activate the extrinsic pathway, and elevated levels of TF are directly correlated to poorer clinical outcome in islet transplantation. Expression of these mediators on pancreatic islets is increased due to the physiological stresses associated with the events related to islet isolation, including brain death, donor organ procurement, and the isolation process (53).

Neither of these molecules, when expressed on islets in situ, comes into direct contact with blood under normal conditions. However, embolization of islets into the portal vein at the time of transplantation immediately places these donor-derived, cell surface inflammatory mediators into direct contact with recipient blood within the hepatic-portal system, where they become potent stimulators of innate immunity in the microenvironment of the newly transplanted islets. This mechanism may explain the reduced number of islets required to cure diabetes in rodent models, when islets are placed at the renal subcapsular site instead of the intraportal site. Islets transplanted to the renal subcapsule are not placed directly into a vascular channel, and therefore are less able to trigger an IBMIR response (54).

Resident intraislet macrophages may also express and secrete the critical factors involved in initiation of IBMIR and may serve as another source of exposure after islet transplantation. This exposure could also trigger the immediate inflammatory response in the microenvironment of the grafted islets. This “cytokine storm” induces thrombotic and apoptotic events, resulting in islet cell death and dysfunction both in allogeneic and syngeneic islet transplantation (55). However, in an allograft setting, this IBMIR response creates an environment that enhances allogeneic responsiveness by the adaptive immune system (a response absent in syngeneic or autologous islet transplantation) (26). The integration of inflammatory mediators and adaptive immunity is discussed in detail in Section IV.C.1.

2. Response at the transplant site.

The dysfunction and nonspecific, inflammation-driven activation of host cells within the liver also initiate events that damage intraportally transplanted islets. Endothelial cells of the hepatic sinusoids up-regulate expression of cell adhesion molecules such as intercellular adhesion molecule-1 and P-selectin in response to their new association with the transplanted islets (56,57). This interaction increases nitric oxide production and stimulates the secretion of proinflammatory cytokines (e.g., TNF-α, IL-1, and IFN-γ). This increase in proinflammatory cytokine production results in the activation of donor macrophages (resident within the islet), host macrophages (Kupffer cells resident within the liver), and hepatic endothelial cells. These cells enhance local inflammatory activity resulting in the loss of transplanted islets (58) and may also stimulate adaptive immunity as they emigrate from the graft to the draining lymph nodes. Additionally, the intraportal infusion procedure uses a transcutaneous and transhepatic approach to catheterize the portal vein, resulting in intrahepatic trauma with some bleeding. Furthermore, embolization of islets induces some level of portal hypertension (59,60). These factors also increase the proinflammatory events contributing to inflammation-mediated islet loss. Thus, innate inflammatory activity is initiated by the transplant procedure and by the exposure of isolated islets to blood at the time of intraportal infusion, and negatively impacts islet engraftment and viability in the crucial period immediately after transplant. Whether or not these inflammatory events or intrahepatic insulin production lead to areas of steatosis, which are also known to complicate this strategy, may merit further consideration as the transplant process evolves (61). Of note, a recent report also questions whether the local deposition of fat within the liver parenchyma may form the framework for later islet injury. This publication demonstrated that inhibition of steatosis with leptin therapy and caloric restriction led to less long-term islet loss by apoptosis (62). As with many factors in islet transplantation, this phenomenon may be the direct consequence of the islet transplantation procedure as well as a later cause of islet loss.

C. Improving islet survival immediately after transplant

Prevention of islet injury in the initial hours and days after intraportal transplantation has become a major focus of investigation in the islet transplant community. No uniformly effective approach to preventing damage to grafted islets has been developed. These investigations have centered on two basic strategies: pretreatment of the donor pancreas or islets and treatment of the recipient at or immediately before transplantation.

1. Modifications of culture methods for islet isolation.

Donor antigen-presenting cells (APCs), which reside within isolated islet tissue, are able to evoke a robust immune response by emigrating from the tissue to the local lymph nodes where they mediate lymphocyte activation via direct antigen presentation. They may further exacerbate systemic inflammation by activation of IBMIR as discussed in Section III.B. Eliminating donor APCs from the graft before transplantation can interrupt these pathways. Faustman et al. (29,63) demonstrated the efficacy of this strategy by pretreating purified islets with anti-MHC class II antibody or antidendritic cell antibody and complement to remove donor passenger leukocytes. However, because islets themselves can up-regulate MHC II and may even be injured by complement activation, other more gentle methods have been sought to deplete passenger APCs (64) and have included: low temperature (24 C) culture (65,66,67), high oxygen culture (68), and exposure to UVB irradiation (69,70,71). Rat islets devoid of MHC class II cells (putative APCs), generated in vitro from perinatal donor pancreata in the absence of a hematogenous cell source, also demonstrated reduced immunogenicity and were freely transplantable across MHC barriers (30).

In addition to the ability of specific culture conditions to alter islet antigenicity, newer strategies may also directly enhance islet cell survival by adjusting the environment to meet the specific needs of islet cells. By promoting β-cell survival, these methods may minimize islet tissue injury and thereby decrease the stimuli for innate immune activation. New methods of pancreas preservation, designed to minimize tissue and cellular damage and enhance cell survival have been explored. One such approach has been to store and transport the pancreas suspended at the interface between traditional preservation solution and an oxygen-rich perfluorocarbon solution, perfluorodecalin. This procedure, known as the two-layer method, has been successfully applied to both animal and human pancreas preservation (72,73,74,75). The two-layer method of preservation improves oxygenation of, and maintains ATP levels within, the pancreas during organ storage and transportation. Islets isolated from two-layer method-preserved pancreata demonstrated superior integrity and viability compared with islets isolated from pancreas preserved in University of Wisconsin solution (76,77,78). However, other studies have questioned whether this preservation strategy truly leads to better islet survival after transplant into human recipients (79); thus, identification of other islet-preserving methods remains a vital problem.

In other studies, intraductal glutamine administration significantly improved islet yield and viability (80). Additionally, endogenous glutathione levels were augmented in islets isolated from these pancreata, procured after a clinically relevant period of ischemia. Supplementation of collagenase solution with glutamine before delivery of the digestive enzyme to the pancreas via the pancreatic duct may be an effective method to improve islet viability and minimize islet injury during the isolation process. Similarly, researchers at the University of Miami reported improved islet yield and islet viability by supplementation of solutions used in the islet isolation process with nicotinamide (NA) (31). A significant increase in islet yield was observed in isolations using NA-supplemented solutions (4343 ± 348 IEQ/g, n = 19) when compared with isolations without NA supplementation (2789 ± 348 IEQ/g, n = 36; P = 0.005). Furthermore, an additive effect on islet yield was observed when NA supplementation was used in the processing of pancreata preserved using the two-layer method (5538 ± 413 with NA, n = 36; P ≤ 0.05). NA supplementation of isolation and/or culture media also reduced the levels of TF within the islet and secreted MCP-1 (31,81). These changes may have a continued benefit after transplant by reducing the islet-mediated inflammation that results in chemoattraction of local monocytes, macrophages, and other immune cells. Quelling the instigators of the posttransplant inflammatory response represents an attractive opportunity for intervention presented by the ex vivo phase of transplant preparation.

2. Improving islet revascularization.

Although improving islet isolation may enhance islet cell survival, the inability of islets to rapidly revascularize threatens their survival immediately after transplant. In the normal setting in situ, the islet is a richly vascularized microorgan that carries five to 10 times the circulating blood of the surrounding acinar tissue in a normal pancreas (82,83). This structural design enables islets to effectively receive oxygen and nutrients and then deliver secreted hormones to the portal and systemic circulation. Isolated islets, however, lose their endogenous vasculature before transplantation. Revascularization of the islet graft is estimated to require 7–14 d (83,84,85,86). Vascular endothelial cells from both donor (87) and host (82,83) stimulate angiogenesis to form intragraft blood vessels in 3–5 d, and full blood circulation is reestablished within approximately 1 wk. Decreased vascular density and low revascularization have been reported in transplanted mouse (88) and human pancreatic islets (89), although studies have also shown near normal vascularity (86). To optimize graft function, islets must be revascularized and reinnervated to approximate normal endogenous blood flow so that afferent nerve inputs and hormonal inputs, as well as drainage and secretion, are best suited to regulating metabolism. Differences between human and rodent islet microvasculature and cytoarchitecture (90,91) and conflicting kinetic models of islet blood flow (86) complicate our present understanding of the secretion and behavior of the various islets cell types.

Several factors may improve graft angiogenesis and vascularity. Vascular endothelial growth factor-A, a well-known angiogenic factor that is secreted by islets in response to hypoxia (92), appears to play a significant role in angiogenesis and improvement of graft function (93,94). Additional growth factors, such as platelet-derived growth factor, epidermal growth factor, and fibroblast growth factor are also thought to promote graft revascularization (83,95,96). Inhibition of angiotensin II in syngeneic mouse islet grafts also appears to improve blood perfusion, oxygen tension, and glucose-stimulated insulin secretion (97). However, angiotensin II blockade, as in Ref. 97 with losartan, may only be effective after establishment of the vascular network. Early use after transplant had negative effects on outcome, and thus, care must be used in the clinical setting where islet transplant candidates may have been on these therapies before transplantation. Other factors that may improve revascularization include transplanting smaller islets (98), which are less prone to hypoxia in culture and revascularize more quickly and efficiently than larger islets (see also Refs. 83 and 99, suggesting advantages of larger islets), reducing acinar contamination (100), and avoiding grafts from older donors with lower vascular density (83).

In addition to the important role played by the vascular endothelium in supporting islet cell survival, the endothelium may also play a direct role in the immune response. Indeed, endothelial cells are now appreciated for their ability to present antigen via both MHC class I and class II as well as to up-regulate costimulatory molecules in response to local tissue injury (101,102,103). In models of cardiac allograft rejection, the vascular endothelium can directly activate alloreactive CD8⁺ T cells even in the absence of a contribution from professional APCs (101). Whether the islet endothelium has similar capability is not known at present. However, in the progression of spontaneous diabetes, the endothelium has been shown to be capable of presenting autoantigens such as glutamic acid decarboxylase (104). Interestingly, the islet vascular endothelium is also thought to be responsible for homing of CD8⁺ T cells to the islet during spontaneous diabetes development (105). Thus, in rebuilding the vascular endothelium to support islet survival, care will be needed to ensure that the newly formed vasculature does not serve to augment the local immune response and attract islet-destructive lymphocytes to the transplantation site.

3. Modification of transplant site: omentum and pancreas.

In addition to altering the isolation and culture conditions to optimize islet yield and orchestrating the reconstruction of the islet microvasculature, an equally simple and attractive approach to increasing islet survival is to consider whether sites and techniques other than intraportal infusion into the liver are better adapted for islet implantation. Although the liver appeared as a promising initial site due to the ease of access to the portal vein and the nutrient-rich environment that hepatic embolization was imagined to provide, this site appears to fall short for the reasons mentioned above, including the intrahepatic inflammatory response, the development of portal hypertension, and the evolution of hepatic steatosis (Section III.B). Indeed, we now realize that liver embolization places the fragile islet cell into an environment of low oxygen tension (5–10% mm Hg) leading to further islet loss because islets do not immediately revascularize (106). Additionally, the islet tissue is exposed to the highest systemic glucose concentration in addition to the highest concentrations of islet-toxic immunosuppressive drugs (107,108). Finally, when we consider the potential need for long-term physical monitoring of the grafted islets by biopsy, the liver represents a highly unfavorable location. Nonetheless, the question remains as to whether any other site would be more attractive. Of sites that have been well studied, the omentum has been a strong candidate. In fact, it was first offered as a location during the pioneering experiments that established the techniques of islet transplantation (109,110). This site preserves the advantage of insulin and glucagon delivery to the portal system (111) while relaxing some of the stringent requirements for islet purity needed to prevent thrombosis after intrahepatic embolization. It may also limit direct blood contact and thus reduce both IBMIR and activation of the innate response mechanisms.

Finally, delivery into the pancreas is a technically challenging but theoretically attractive consideration. At least one comparison study suggested that pancreatic implantation leads to different gene expression and function of the transplanted tissue compared with the intraportal route (112). Overall, these investigations represent ongoing efforts to preserve islet function without any significant change to standard methods and therapies currently in place.

4. Modification of transplant site: thymus.

In addition to choosing alternative sites that are more favorable for islet survival, it is also possible to place islets in a position to exert a positive modulatory effect on the immune system. The thymus represents the central location of T cell development and education. Within this environment, T cells are exposed to self-antigens, and self-reactive T cells are modulated so that they are either unable to harm the host or acquire beneficial regulatory properties (reviewed in detail in Section IV.A). T cell development in the thymus is determined strongly by the concentration and quality of available antigens, a feature that is highlighted by autoimmune-prone mice and humans that are deficient in the autoimmune regulator, a transcription factor that controls what antigens are available for thymic presentation (113,114,115,116,117,118). Absence of appropriate thymic antigen expression in these individuals leads to peripheral autoimmunity. Conversely, it has been attractive to instill novel antigens into the thymus so that they will become included in the definition of self and be afforded the aegis provided by thymic selection. Because islets are a cellular transplant, they have been an outstanding candidate for placement into this location.

In fact, intrathymic islet transplantation has been highly successful in mouse, rat, and canine models (22,119,120,121,122,123,124). The first reports in the early 1990s demonstrated the ability of intrathymic transplantation of islet tissue to prevent alloimmune rejection in addition to preventing recurrence of disease (119,120). These studies also demonstrated that introduction of islet tissue into the thymus permitted survival of a matched second islet graft transplanted intraportally or under the kidney capsule as would be expected after tolerance induction. Because it may be difficult to transplant adequate islet mass into the thymus to promote permanent euglycemia in larger animals or humans (125), the extension of tolerance from the central site to the periphery was a valuable aspect of this approach. This same approach has also been extended to prevent spontaneous autoimmunity in diabetes models as well as in other autoimmune disorders, indicating its efficacy for long-term immunomodulation (122,124,126,127). Interestingly, the effect of intrathymic inoculation appears to rely, in part, on the presence of intraislet passenger APCs (128). These cells may participate directly in the education of the T cell repertoire, and hence, the intrathymic transplant may provide more than just antigen. The presence of these cells and the appropriate antigen leads mechanistically to the deletion of CD4 and CD8 T cells in addition to the selection of regulatory T cells and their emigration to the periphery (129,130,131,132). As a caveat to this appealing procedure, it is important to note that the injected antigens are interacting directly with potential islet-reactive lymphocytes. In at least one study, injection of the thymus with certain candidate β-cell antigens accelerated disease progression (133), a finding that necessitates care in this approach especially with islets that may have suffered prolonged ischemia or significant autodigestion.

Despite the attractiveness of the thymus as a transplantation site with immunomodulatory function, there are also some limitations to this approach. In particular, because the average transplant recipient is likely to be postpubertal, thymic involution may represent a substantial barrier to the success of this strategy. The thymic epithelium decreases in area by at least 90% by the time a person reaches 50 yr of age. The rate of decrease is thought to reach its peak during or shortly after puberty. Gonadectomy or chemical inactivation of sex steroid production seems to rejuvenate thymic function, suggesting a role of pubertal hormone development in thymic decline (134). In addition to endocrine control of thymic function, intrinsic hematopoietic decline also contributes to decreased thymopoiesis (135,136,137). This distinction is important if investigators wish to augment thymic production for therapeutic purposes in aged individuals. On the one hand, adjustment of adrenocortical hormone levels and sex steroids may provide some effect, but replacement of thymic progenitors with a younger pool as is accomplished with bone marrow transfer may be necessary for reversal of thymic decline. The utility of bone marrow transfer for diabetes and transplantation is discussed further below (Section IV.A).

Although the decline in overall thymic output is well documented, the ability of the thymus to regulate the immune system also depends on exactly which cell types continue to be produced as patients age. One study has assessed the frequency of TRECs [T cell receptor (TCR) excision circles], a marker of recent generation in the thymus, in T cell subgroups in children and adults. They reported the continued production of CD45RA+CD4RO- cells into nearly the seventh decade. However, the same study demonstrated a greater diversity of TREC+ subgroups in children consisting of CD45RA+CD45RO−, CD45RA+CD62L+, and CD45RO−CD27+ fractions (138). Because not all groups of T cell progenitors will have the same propensity to produce regulatory T cells, it is important to understand which subsets are responsible for development of T cells with regulatory function. At present, the precise cell subtype that gives rise to regulatory cells is not known. In addition, another study using similar technology demonstrated a decrease in TCR diversity with age (139). Diversity may be a necessary factor to promote regulatory cell function, and loss of diversity may prevent the generation of T cells with appropriate specificity for recruitment into the islet-protective regulatory T cell pool. Indeed, TCR specificity may be necessary for the efficient protection of transplanted material by graft-specific regulatory T cells (129,140).

In addition to the theoretical appeal of direct delivery of therapy to the thymus, there is also evidence that compound transplants of donor organs in combination with donor thymus may enhance tolerance induction in part by providing a selecting milieu replete with donor antigen (141,142,143,144). In fact, cotransplantation of whole thymus or vascularized thymic lobes together with either kidney or heart allotransplants leads to long-term graft acceptance in a miniature swine model (145). The transplantation of a thymic lobe previously prepared with islets has also been successfully tested in a murine model (146). Because some recipients with diabetes could require a kidney transplant in addition to islet restoration, whether or not there is a role for compound islet-thymus-kidney grafts may warrant future consideration. Overall, transplantation into the thymus may permit avoidance of peripheral immune surveillance and early immune-mediated islet loss while also exerting a beneficial immunomodulatory effect.

5. Modification of islets by gene therapy.

Although modifying the culture conditions in which islets are preserved or the transplantation site into which they are delivered is a relatively straightforward way to alter their survival and immunogenicity, many investigators are taking advantage of the ex vivo period of islet culture to consider approaches such as gene therapy to permanently modify islets and promote islet resistance to inflammatory injury. These efforts have primarily been directed at modifying islet sensitivity to the various stresses that they encounter during the isolation and transplantation procedure.

Within the apoptosis-provoking cytokine pathway, secretion of the inflammatory mediators IFN-γ, IL-1, and TNF-α is most likely to result in diabetes progression (147,148,149,150,151). Islet cells can directly receive signals through these mediators, and contact with these ligands involves activation of multiple islet transcription factors including signal transducer and activator of transcription-1 (STAT1), activator protein-1 (AP-1), and nuclear factor κB (NF-κB) (152,153). Signal transduction through these cytokine pathways leads to transcription of proapoptotic genes such as Bax and, ultimately, to islet cell death (154). These pathways have been amenable to intervention through several gene therapy approaches. Interference in the IL-1 mediated apoptotic pathway has been obtained by transduction of islet cells with a construct expressing IRAP, the IL-1 receptor antagonist protein. Tellez et al. (155) demonstrated improved islet cell survival and replication after adenoviral transduction with this construct, an extension of the prior findings of Giannoukakis et al. (155,156). Protection from cytokine-mediated apoptosis has also been achieved after islet cell transduction with TNF-α receptor immunoglobulin fusion protein, baculovirus p35 protein, and IGF-I (157,158,159). In addition to preventing the initial activation of the apoptotic pathway, strategies to overexpress antiapoptotic proteins such as the TNF-α inducible transcription factor A20 or bcl-2 can also extend islet survival by protecting islets from deleterious proinflammatory signaling (160,161,162).

In addition to protecting islet cells from apoptosis, gene therapy approaches may also induce islet cell proliferation or augment directly the function of the transplanted tissue to promote disease cure with fewer donor islets. The inability to transplant sufficient islet mass leads to continued hyperglycemia. As discussed in detail below (Section V.A), this metabolic derangement may be directly islet toxic and may also indirectly injure islets by provoking proinflammatory cytokine production. Modification of donor islets to promote immediate euglycemia may overcome these barriers. Interestingly, intensive early control of new-onset T1D has also been suggested to preserve islet mass, and further studies of this approach are anticipated soon (163).

Islet cell proliferation can be induced by methods that allow for pharmacological activation of promitotic pathways. This approach has been pursued using lentiviral-mediated cellular transduction with a chimeric construct in which the erythropoietin receptor signaling apparatus is fused to FK506-binding protein domains; cross-linking of this construct by exposure to a chemical inducer of dimerization can then drive islet cell proliferation (164). In this study, there was successful proliferation demonstrated both in vitro and in vivo, retention of appropriate glucose responsiveness, and stability of the differentiated phenotype of the transduced cells. The induction of islet cell proliferation with decreased minimal islet mass for cure has also been found after islet transduction with hepatocyte growth factor (165,166). The physiological role of this pathway has been supported by demonstrating decreased islet function in the absence of the hepatocyte growth factor receptor (167).

In addition to the induction of islet proliferation, the function of individual pancreatic β-cells can be augmented such that one islet may subserve the function of many, a “super-islet” approach. Because secretion of adequate levels of insulin to maintain physiological euglycemia in response to glucose and other nutritional stimulation is the primary goal of islet transplantation and may also prevent hyperglycemia-mediated islet injury, transduction of islet β-cells with the insulin gene has been investigated to increase the ability of individual β-cells to produce insulin in response to hyperglycemic challenge (168). This procedure reduces the required islet mass for cure of diabetes in mice such that only 25–50% of the previously needed islet mass is rendered curative.

A combination of these approaches may drastically reduce the number of islets needed for curative transplant and may provide the groundwork for future exploration of living-related islet donation. These gene therapy approaches to augment islet function may also permit adequate islet transplant mass to achieve immediate euglycemia. Overall, these strategies to modify the surgical approach to islet transplantation and the islet tissue’s resiliency may greatly expand our ability to perform islet transplantation by maximizing the effective yield from every donor pancreas. However, there remain significant barriers within the recipient, most notably the response of the recipient immune system to the islet transplant procedure. Because islet transplantation is elective, there is also opportunity to condition the recipient either before or at the time of transplantation. These particular recipient characteristics and candidate approaches are the focus of the following sections.

IV. New Strategies to Achieve Tolerance in the Presence of Autoimmunity

A. Overview

The ultimate goal of islet transplantation is to cure diabetes by transplanting insulin-producing cells derived from a plentiful source without the need for lifelong recipient immunosuppression (37,169). Although we have discussed the significant barriers to obtaining numerous, high purity, highly functional islets, there are even more significant barriers inherent to the interaction of the potential transplant with the islet-reactive immune system. For islet transplantation to succeed, the concept of immunological tolerance must be understood in the context of T1D. In its simplest definition, tolerance is the absence of immune responsiveness to a particular immune target. Most individuals possess an immune system that becomes “educated” during development and acquires the ability to discern self from nonself. In T1D, the loss of the ability to identify insulin-producing islet β-cells as self leads to autoimmune responsiveness, resulting in the destruction of β-cells within the native pancreas (1,170). This autoimmune response is mediated primarily by T cells, which are the main effector cells in the process of β-cell destruction.

Although we have discussed the possibility of modulating immune function by targeting T cell development in the thymus (Section III.C.4), the overall derangements in the autoimmune-prone immune system likely begin in the bone marrow—the generative site of all lymphoid cells. In fact, it has long been known that bone marrow from diabetes-prone nonobese diabetic (NOD) mice can transfer diabetes to normal mice after lethal irradiation and destruction of the recipient’s bone marrow (171,172). However, it is also well established that the presence of bone marrow from a diabetes-resistant strain can suppress the ability of NOD bone marrow to cause disease (173). This process depends on the presence of protective professional (non-B lymphocyte) APCs (174), a cell class that is known to be defectively generated in NOD bone marrow (175,176,177,178). The ability of normal lymphoid cells to prevent diabetes progression had been appreciated even before the study of bone marrow chimeras. In pioneering early work in the BioBreeding rat, Naji et al. (179) observed that infusion of neonatal rats with lymphocytes from diabetes-resistant donors led to lifelong diabetes prevention. This finding established the concept that lymphoid cells with normal function could rescue the phenotype of the diabetes-prone immune system.

This concept has been extended into a therapeutic approach by considering the use of bone marrow transplantation for diabetes prevention and reversal. Preliminary studies in rat and murine models have established proof-of-principle for this approach. Although initial studies focused on elimination of the entire diabetes-provoking bone marrow and its complete replacement by allogeneic cells, this approach has not been ideal for humans where the risk and complications of graft-vs.-host disease are high. Instead, induction of a state of mixed chimerism in which cells from the diabetes-prone immune system coexist with, and are controlled by, donor cells has become the ideal. This approach has been successful in both preventing and reversing autoimmunity and permitting the transplantation of islet tissue that is MHC-matched to the protective marrow (180,181,182,183,184). Interestingly, the presence of as few as 1% donor cells in the presence of 99% diabetes-prone NOD cells can suppress diabetes progression (180).

Despite the exciting success of bone marrow transplantation in rodent models of diabetes, there has not yet been a human clinical trial of this approach. The tentative approach to translating this finding relates in part to the risks that have been associated with bone marrow engraftment, which include graft-vs.-host disease and the stringent immunosuppression required to support bone marrow engraftment. However, there have been exciting recent advances in the transplantation literature, with new reports of simplified conditioning regimens. In one recent study, five patients with end-stage renal disease received matched bone marrow and renal grafts without myeloablation and with only transient immunosuppression (185). Four of five transplants had long-term graft function, and assessment of the recipient immune system showed expansion of regulatory cells. Whether similar success can be achieved in patients with autoimmune disease remains to be seen. Interestingly, the presence of immune chimerism has been associated as a risk factor for the development of autoimmunity in women who are at risk for multiple sclerosis (186,187). Whether a chimeric state that follows pregnancy is a trigger of autoimmune progression is under continued scrutiny. There may be important characteristics of the donor chimeric cells that may either promote or inhibit progression to autoimmunity in predisposed individuals. These interactions will require careful consideration as this technique moves forward.

The ability of procedures such as bone marrow transplant to induce immune tolerance to specific antigenic targets could prevent both the autoimmune processes leading to T1D and the anti-islet allogeneic immune reactivity that leads to islet graft rejection. The development of a therapeutic protocol that could successfully induce tolerance to both auto- and alloimmune responsiveness would dramatically alter treatment strategies in diabetic and prediabetic patients (188,189).

Although replacement of the immune system by bone marrow transplant represents one method to reeducate the immune repertoire, there are also numerous opportunities to intervene in immune function by pharmacological means. Although numerous therapies for enhanced immunosuppression or tolerance induction are emerging, all strategies are similar in that they attempt to reproduce endogenously generated immune tolerance. Antigen-reactive T cells (in the case of autoimmunity, antigen-reactive cells are also self-reactive) can be subject to one of several processes to ensure that their target antigen is not destroyed. Removal of specific reactivity can be achieved through immunological ignorance, anergy, deletion, and/or immune regulation (see Fig. 3 for overview of active mechanisms) (reviewed in Ref. 190). Sequestration of target tissue from immune surveillance so that tissue-reactive lymphocytes do not encounter the target antigen (i.e., ignorance) is the theoretical underpinning for approaches to transplant islets into immunologically privileged sites such as the testes (reviewed in Ref. 191). However, this mechanism is passive and does not alter the underlying propensity for the immune system to recognize and destroy islet tissue. For both anergy and deletion, exposure of the reactive T cell to the target antigen results either in physical elimination (deletion) or functional inactivation (anergy) of that specific antigen reactive clone (reviewed in Ref. 192). Newly emerging self-reactive T cells must continue to be anergized or deleted to prevent the development of autoimmune responsiveness, and these processes must continue over a lifetime as new cells are generated so long as anergy and deletion are the only mechanisms employed.

Figure 3.

Overview of active mechanisms of induced immune tolerance. The mechanisms that support immunological tolerance are illustrated. For each a brief listing of the respective advantages and limitations is provided.

Self-reactive T cells can also be controlled (down-regulated) by regulatory T cells. In fact, T cells bearing self-reactive receptors can be diverted to the regulatory pathway during their thymic development (129,193). Regulatory T cells offer the distinct advantages that they are long-lived, capable of self-renewal, and, importantly, capable of protecting the tissue expressing their cognate antigen from other self-reactive T cells (194,195,196). Given the desirable functions of this regulatory cell class, we will present a further discussion of the features, clinical applications, and limitations of regulatory T cells.

B. Development and function of regulatory T cells

The landscape of regulatory T cells has become very broad in the last decade, and these cells and their mechanisms of action have been reviewed extensively in many other locations (190,197,198). At present, we appreciate that there are at least four potential classes of T lymphocytes with regulatory cell function, and these include Treg cells (CD4⁺CD25⁺foxp3+), Th3 cells (T helper type 3 cells; CD4 cells producing TGFβ), Tr1 cells (T regulatory type 1 cells, CD4 cells producing TGFβ and IL-10) and CD8⁺CD28− cells. Although many of the features of these cells overlap, there are experimental strategies to differentiate the CD4 cell types.

Treg cells are differentiated from the others by their generation within the thymus, a feature that has also led them to be called “natural T regulatory cells” (193). Both Th3 and Tr1 cells appear to be induced in the periphery after antigen contact in the correct cytokine milieu and possibly in the presence of regulation-facilitating dendritic cells (199,200,201,202,203,204,205,206). Much of the characteristics of CD8-lineage regulatory cells are only now emerging, and their role in diabetes is largely unknown although they have shown some suppressive activity in transplantation and other autoimmune models (207,208,209).

Mechanistically, these cell types also demonstrate considerable overlap in function. All classes of regulatory cell have some propensity to regulate naive and effector T cells in a cell-contact-dependent manner. Treg cells additionally utilize CTLA-4, IL-10, and TGFβ to effect regulation. Th3 cells are characterized by production of TGFβ alone, whereas Tr1 cells produce both IL-10 and TGFβ. The mechanism(s) of CD8-mediated regulation is not well understood at present.

Because Tr1 and Th3 cells are thought to be generated through recruitment from the naive pool of peripheral T cells, they have made good candidates for the effectors of tolerance that can be induced by peripheral antigen administration. Indeed, these cells were first described in oral tolerance induced in the setting of experimental autoimmune encephalitis and the clinical setting of multiple sclerosis after oral exposure to known autoantigens (199,210,211). These cell types have been similarly thought to participate in the suppression of diabetes progression that follows oral administration of insulin (212), although in many studies that focus predominantly on TGFβ production as an outcome it is difficult to identify the precise cell subtype involved.

Although many, if not all, classes of regulatory cell may be involved in the prevention of autoimmune disease, the role of naturally occurring CD4⁺CD25⁺ thymically derived Tregs has been the best characterized. These cells can be identified by additional expression of CTLA-4 and CD62L but are best known for the expression of the transcription factor foxp3, which directs their development (213). The role of foxp3 in Treg development is appreciated not only from basic science models but also from clinical experience because it is now known that patients with the autoimmune X-linked polyendocrinopathy (IPEX) have foxp3 deficiency (214,215,216). Because diabetes is part of the IPEX spectrum, foxp3+ Treg cells have a clear role in the prevention of diabetes. The ability of these cells to mediate highly effective suppression of both autoimmunity in disease models and allograft rejection in transplant studies has made them attractive candidates for therapeutic application. Many strategies have attempted to purify and expand them in vivo; however, although we have multiple cellular markers of this class, none are entirely specific, and generally isolated “Treg” samples are contaminated with a significant percentage of activated cells. The emergence of new surface identifiers such as CD127, as recently reported, may allow further progress in this effort (217).

Because Tregs strongly suppress the immune response to transplantation of syngeneic islets and improve survival and function of syngeneic islet grafts (218), several approaches are now emerging to induce/increase host T regulatory cell activity in the transplant setting.

First, given the central role of TGFβ in regulatory cell function, this molecule has been used as a therapeutic agent (219). In fact, systemic TGF-β1 therapy, administered by iv Adv-TGF-β1, inhibits pancreatic islet destruction in diabetes-prone NOD mice. This inhibition appears to result from the development of a population of Foxp3+ regulatory T cells (220). This same therapeutic regimen was also able to prolong survival of islet grafts placed 7 to 14 d after treatment (median survival time of 50 d in treated animals, compared with 17 d in untreated hosts). These grafts demonstrated a peri-islet mononuclear cell infiltrate that stained positively for CD4, CD25, and Foxp3. These studies suggest that TGF- β1 induces a protective effect, which could act on native or transplanted islets, and could reduce both autoimmune and alloimmune reactivity. Further studies have shown that CD4⁺CD25⁺ regulatory T cells derived from peripancreatic lymph nodes inhibit the differentiation of islet-reactive CD8⁺ T cells into cytotoxic effector cells and arrest the diabetes disease process (221). The specific mechanism of this inhibition is unknown; however, a critical role for CD4⁺CD25⁺ Treg cells and TGF-β receptor signaling in the process has been established both in vitro and in vivo. These findings indicate that TGF-β acts on CD4⁺ T cells that also express CD25 and demonstrate that increased level of expression of TGF-β1 correlates with the delayed onset of diabetes.

Second, recent studies indicate that naive CD4 cells can develop into regulatory Tr1 or Th3 cells in response to antigen exposure, and these cells can act therapeutically to regulate immune responsiveness (222). Weber et al. observed that murine effector regulatory T cells, derived from islet-specific CD4⁺ cells by immune stimulation through the TCR in the presence of IL-2 and TGF-β1, were able to inhibit lymphocyte infiltration within the pancreas, and prevented the onset of spontaneous diabetes in the NOD mouse (223). The generation and function of this population of regulatory cells was not dependent on innate CD25⁺ cells, and, in fact, they did not express CD25. The results suggest that adaptive regulatory T cells derived from the innate CD4⁺ cell population may prevent diabetes in the NOD mouse by inhibiting infiltration of autoreactive Th1 cells into the pancreatic islets (170).

C. Limitations to the clinical application of Tregs and the role of innate immunity

1. Immune control of regulatory cell function.

Although regulatory T cells represent an attractive means of affording long-term immunological protection to islet transplants, the ability to induce or apply these cells in the autoimmune individual warrants careful consideration. In fact, regulatory T cells likely failed the diabetic individual in the initial progression to disease and rejuvenating their functions necessitates better understanding of the initial loss of tolerance.

Regulatory T cells themselves are subject to important control mechanisms that may be deranged in the setting of autoimmunity. In the normal individual, these mechanisms exist so that regulatory cells can be deactivated to facilitate a robust immune response when defense against pathogens is required. The control of regulatory T cells is, in fact, one crucial point at which the innate immune system interacts with and influences the function of adaptive immunity. Most important in this control mechanism appears to be the activation state of local APCs. Whether APCs mediate their counter-regulatory effect by directly deactivating regulatory cells or by inducing resistance to regulation in naive cells is still under investigation. Nevertheless, one such mechanism of counter-regulation is the interaction between the glucocorticoid-induced TNF receptor family-related receptor (GITR) and its ligand (GITR-L). GITR is expressed on regulatory CD25⁺ cells as well as nonactivated CD25− T cells. Stephens et al. (224) demonstrated that the presence of GITR on naive T cells and its engagement by GITR-L on activated APCs is the critical event that promotes inhibition of suppressor function by rendering the naive T cell insensitive to the actions of regulatory cells. Whether GITR signaling can also deactivate regulatory T cells or whether there are other ligand-receptor interactions that perform this function is a fertile area for investigation. Building on this idea, the cloning of GITR-L and its human equivalent Activation-inducible TNF receptor suggests that therapies that target its structure may emerge in the near future (225,226). Whether this mechanism is appropriately regulated in patients with autoimmune diabetes or in the rodent models of disease is not known. However, progressive resistance of the naive T cell pool to the action of regulatory cells has been demonstrated in at least one experimental setting in NOD mice (227), the most widely used murine model of human diabetes.

2. Molecular mechanisms of innate immunity.

Understanding the molecular control of inflammation has given substantial insight into how innate immune signals can stimulate adaptive immunity and counter-regulate regulatory T cells. Central to this pathway is the role of toll-like receptors (TLRs). These molecules are the central sensors of the innate immune system, and their activation plays an important role in the initiation of the immune response (228). TLRs function as pattern recognition receptors that coordinate the response to immune-stimulatory ligands including lipopolysaccharide (LPS), viral RNA, viral and bacterial DNA, and molecules characteristic of parasitic infection (reviewed in Ref. 229). Given the important epidemiological role in diabetes progression played by environmental agents, the function of these receptors may be vital to the autoimmune process (230,231). The signaling cascade that follows TLR triggering by environmental agents is shown in Fig. 4. For all TLR molecules, their effects are mediated through coupling to Toll/IL1R receptor domain-containing adaptor proteins. These adaptors include Mal, MyD88, TRIF, TRAM, and SARM. The TLR pathways engage combinatorially most of these adaptors; however, two of them, MyD88 and TRIF, are essential for signaling triggered by dimerization of TLRs. Whereas MyD88 integrates signals from the majority of TLRs including IL1R and IL18R, TRIF is solely responsible for transducing signals from TLR3, which recognizes viral RNA and triggers type 1 interferon production. These proteins initiate the downstream cascade, which ultimately couples to the activation of NF-κB, AP-1, and other stress-responsive transcription factors. Importantly, TLR adaptor proteins, in addition to NF-κB and AP-1, have been implicated in diabetes pathogenesis and β-cell apoptosis (153,232,233,234,235,236,237).

Figure 4.

TLR complexes, their ligands, and their downstream signaling pathways. The TLR complexes involved in activation of the innate immune system and control of allogeneic tolerance (241,243,246) are shown on the cell surface including homodimeric complexes of TLR3, -4, and -9, and the heterodimeric TLR1/2. Natural ligands are shown above in black; ligands that are used in laboratory investigation of signaling are noted in red; ligands that may be produced by secondary necrosis during normal tissue turnover are shown in green. The TLR1/2, TLR 9, and TLR 4 complexes all activate MyD88. MyD88 in turn stimulates TRAF6 via IRAK, which is involved in activation of the NEMO-IKK (IκB kinase) complex. TLR4 and TLR3 also use the TRIF adaptor to activate the same NEMO-IKK molecules. This common step in TLR signaling results in the phosphorylation and degradation of IκB, the constitutive repressor of NF-κB. After release of repressor control, NF-κB translocates to the nucleus via the karyopherin/importin α-1 dimer and then activates the transcription of stress responsive genes. IFN, Interferon.

3. TLR-mediated counter-regulation.

Although TLRs are critical for the activation of the innate immune response, activation of these receptors also leads to robust activation of adaptive immunity—that part of the immune response directed by antigen-specific T and B lymphocytes. As previously reported by Pasare and Medzhitov (238,239,240), APCs, including both dendritic cells and B lymphocytes, when stimulated through TLRs become able to counter-regulate the function of Tregs and thereby permit the development of an immune response that would otherwise be suppressed. They further demonstrated that in mice that cannot activate the TLR signaling cascade due to deficiency in MyD88, a central molecule of TLR signaling, depletion of regulatory T cells is required for immune activation. In addition to the crucial role emerging for these TLR molecules in the physiological activation of adaptive immunity, there is also an awareness of a central role played by these signaling molecules in directing the sensitivity of the immune system to interventions that promote tolerance, a feature highlighted strongly in recent transplantation studies. Thornley et al. (241) have demonstrated that engagement of several TLRs (TLR2, -3, -4, or -9) also deactivates regulatory T cells and overrides tolerance induced by costimulatory blockade in a model of skin transplantation. Subsequent analyses have shown a similarly central role of TLRs in models of cardiac and islet allotransplantation (242,243,244,245,246). The presence of ongoing inflammation mediated by TLR triggering may form an additional barrier to the induction of transplantation tolerance in individuals with autoimmune diabetes.

4. Inflammation and tolerance resistance in the presence of autoimmunity.

The concern that the inflammatory state associated with autoimmunity is a barrier to transplantation tolerance is well-founded in transplantation studies in mice. Although normal murine strains can be modified easily by many strategies that induce donor-specific tolerance (247,248,249,250,251,252,253), the NOD mouse has proven highly resistant to these approaches (254,255,256). Currently, no pharmacological therapy that promotes allogeneic transplantation tolerance in the presence of an intact immune system has reliably translated from normal murine strains into the NOD background, even when the transplant is other than islets and hence not subject to autoimmune recurrence. The difficulty of inducing tolerance in this model suggests that autoimmunity may alter the immune system in ways that are not reversible by current strategies. Moreover, studies in intercrosses between the NOD mouse and other murine strains demonstrate that the resistance to tolerance is a genetically dominant trait (257). These progeny remain resistant to conventional tolerance-inducing therapies although none of these mice develop autoimmunity. Because agonists for several TLRs (TLR2, -3, -4, and -9) prevent the establishment of long-term tolerance to skin and islet allografts in nonautoimmune strain mice (241,246), chronic or dysregulated TLR activation in NOD mice could connect these findings.

Indeed, the function and expression of several TLR molecules is dysregulated in the NOD background. TLR1 has been linked to the genetic locus encoding the Idd6 gene, and its expression is altered in NOD mice although the role of its ligand in diabetes progression remains unknown (258). NOD mice also have a well-characterized hypersensitivity to TLR4 triggering by LPS (259). Human studies are beginning to suggest similar genetic and functional linkages. One weakly powered study correlated a polymorphism in a TLR3 intron to diabetes development in South African Blacks (260). A stronger study indicated altered expression and signaling of both TLR2 and TLR4 in monocytes of patients with established T1D (261). The role of TLR signaling in diabetes progression has been more intensively studied in the murine model, although the results remain confounded there as well. Exposure to the TLR4 ligand LPS is able to trigger diabetes onset in TCR transgenic BDC2.5/NOD mice, suggesting that it may destabilize the balance of regulatory and autoreactive activity, which normally suppresses diabetes in these mice (262). Moreover, triggering of TLR3 and TLR9 is involved in the mechanism of diabetes progression induced by viral infection in a rat model (263,264). Despite these studies that support a role for TLRs in deactivating immune regulation, there are also several studies in which bacterial wall components and specific TLR triggering by CpG or poly-IC prevented disease (265,266). However, in these studies, the stimulus was administered at a very young age before the development of a robust diabetogenic repertoire, and these studies have not all been repeatable (267). Moreover, exposure to the TLR3 ligand poly-IC in the presence of IFN-γ does activate β-cell apoptosis, and this cell death requires TRIF-dependent activation of TLR3 (268). Given the substantial epidemiological evidence of environmental triggers of disease progression, it is vital to know the consequences of TLR triggering and innate immune activation on the sensitivity of NOD mice to tolerance induction and maintenance.

Although the role of environmental agents in TLR triggering has been long appreciated, there is emerging evidence that there are endogenous TLR ligands that may play pivotal roles in the progression of autoimmune disease (reviewed in Refs. 269 and 270). These endogenous ligands (Fig. 4) are most likely to be produced from cells undergoing apoptosis. Although an earlier paradigm suggested that necrotic cells induce inflammation whereas apoptosis serves to maintain immune quiescence, many tissues undergoing apoptosis also have areas of secondary necrosis (271,272,273,274). These areas may be immunogenic and serve to stimulate TLR receptors. Indeed, β-cells in mice experience a wave of physiological apoptosis around 2 wk of age (275,276) and shortly thereafter, TLR-responsive monocytes begin to accumulate in the area. A recent study by Kim et al. (277) suggests that secondary necrosis of islets can activate islet-reactive T cells by stimulating local antigen presenting cells such as macrophages and dendritic cells. Their analysis suggested that this phenomenon may be mediated by TLR2 through contact with proteinaceous debris. Interestingly, TLR2 has been found to be up-regulated in patients with T1D (261) and associated with diabetes susceptibility (278).

In addition to the developing role of TLRs and their ligands in autoimmune diabetes, TLR-sensitive cell types, which include dendritic cells, macrophages, and B lymphocytes, are well known for their central participation in the development of disease (177,259,279,280,281,282). As such, understanding their sensitivity to these potential stimulants could reveal pathways for disease prevention. After TLR stimulation, dendritic cells are induced to mature, migrate, and become able to inhibit the function of regulatory T cells, thereby activating the immune response (239). B Lymphocytes from NOD mice also represent an important target of TLR signaling because they are hyperresponsive to LPS, the TLR4 ligand, and are a requisite coactivator for islet-destructive T cells (174,259,279,283). These alterations in the B cell compartment, which include up-regulation of T cell stimulatory molecules such as MHCII, CD86, and CD40 (174,259), have more recently been linked to hyperactivation of NF-κB, the paradigmatic mediator of TLR signaling (234,235,284). Because B lymphocytes are required for certain tolerance-inducing therapies (285,286,287) and are able to expand regulatory T cells (288), their activation state and sensitivity to these innate immune signals may be paramount in directing the outcome of exposure to tolerance-inducing therapies during transplantation.

5. Mediators of inflammatory injury.

With regard to targets to overcome this TLR-mediated transplantation resistance, one follow-up study linked the tolerance resistance to production of type 1 interferons (246). These interferon molecules may play an important role in tissue injury during autoimmune disease and may be produced after ligation of TLR3. Studies in lupus pathogenesis have most recently identified genetic risk conferred by an interferon regulatory element and other innate immune factors (289,290). Similarly, interferons are emerging as important mediators in the pathogenesis of autoimmune diabetes. Although interferon production has been detected in the serum and pancreata of subjects with diabetes (291) and the role of interferon IFN-γ as a proinflammatory mediator has been long appreciated, new insights are demonstrating a pathogenic role for interferon β as well. IFN-β is the principal product of TLR3-mediated signaling that follows exposure to viral RNA (292,293). Expression of IFN-β under the insulin promoter results in insulitis and low-grade diabetes progression in normal mice (294). Introduction of this transgene into the diabetes-prone NOD background leads to substantially accelerated diabetes progression (295). The local production of IFN-β did not result in islet β-cell dysfunction but instead led to a robust insulitis. Of interest, production of type 1 interferons has also been suggested to promote myeloid dendritic cell dysfunction in NOD mice (296). These studies emphasize how environmental factors, mediated through TLR signaling, may interact with a background propensity to autoimmunity to result in overt β-cell destruction and diabetes.

6. Therapeutic targeting of proinflammatory pathways.

With the emerging role of inflammation in diabetes progression, there are also attempts now to specifically target these pathways. As illustrated in Fig. 4, there are several key steps in the pathway that are amenable to intervention; in particular, MyD88, which mediates signaling from most TLRs, and NF-κB, which is the final functional molecule in the cascade, are important therapeutic targets and proof-of-principle for both in diabetes has been demonstrated. Eldor et al. (297) have transgenically expressed a super-repressor of NF-κB in β-cells and shown resistance to apoptosis as a result. Similarly, expression of a dominant-negative MyD88 was also islet-protective (233). These data suggest that inhibition of most major TLR pathways can be accomplished and that this effect may be beneficial to islet survival. The role of the MyD88-independent TLR3 pathway, which has also been implicated in diabetes progression (264,268), has not been as well studied. In this pathway, TRIF, rather than MyD88, plays a central signaling role, but whether TRIF is involved in autoimmune diabetes or β-cell apoptosis has not been specifically investigated. Importantly, the above studies focused on TLR signaling pathways within the islets themselves. Islets do express these TLR molecules, and there may be important interactions between islet cells and the innate immune system (298,299). This interaction represents an attractive mechanism by which viral infection can induce anti-islet autoimmunity in individuals with the appropriate genetic predisposition.

Beyond these initial investigations of islet-directed antiinflammatory therapy, there are also strategies being developed to inhibit systemic inflammation that may drive immune activation. Lisofylline, originally identified for its ability to inhibit IL-1β signaling and induction of apoptosis in islet cells (300), has shown robust activity to down-regulate the systemic inflammatory response (301). This amelioration of the autoinflammatory state has been associated with interference with STAT4 signaling (302) and has resulted in prevention and amelioration of autoimmune diabetes as well as improved islet transplant survival in murine models (303,304,305) and improved human islet function (58). Koulmanda et al. (306) similarly identified underlying inflammation as a barrier to diabetes reversal and applied a triple therapy approach consisting of rapamycin plus agonist IL-2-related and antagonist-type mutant IL-15-related Ig cytolytic fusion proteins (IL-2.Ig and mutIL-15.Ig). They hypothesized that augmenting the proregulatory effects of IL-2 while inhibiting the proinflammatory mediator IL-15 could favorably alter the balance between regulation and inflammation (307). This therapeutic approach had a striking ability to reverse diabetes and ameliorate inflammation.

7. Horizons in antiinflammatory therapy.

Although there is likely to be success through direct targeting of the signaling pathways involved in proinflammatory cytokine signaling as discussed above, the human body itself also has substantial defense mechanisms designed to constrain the effects of the inflammatory response once it is unleashed. Several emerging therapeutic strategies seek to augment these pathways to down-regulate inflammation and protect tissues from injury. Some of these strategies are even beginning to be applied to therapy in diabetes and in transplant settings.

Nonspecific inflammation within the first 3 d after transplant may result in the early loss of nearly half of the transplanted islet mass (18,218,308,309,310,311,312,313,314). This loss is mediated largely by the actions of the innate immune system, which responds to islet tissue damage caused by pathogens, toxins, ischemia-reperfusion injury, or iatrogenic interventions such as surgery. However, this inflammatory reactivity is counter-productive because it produces collateral damage to the newly placed graft (315). Recent studies show that adenosine, acting through the adenosine receptor (AR) subtype 2A, can powerfully inhibit inflammation and reperfusion injury (316,317). The ARs are G protein-coupled and comprised of four subtypes: A₁, A_2A, A_2B, and A₃ (318,319,320). The A_2AAR subtype is found on most marrow-derived cells (e.g., neutrophils, platelets, macrophages, and T cells) (319,321,322,323) and on platelets, where they are known to inhibit ADP-induced aggregation via a cAMP-mediated process (322,324). Deletion of A_2AAR in transgenic mice revealed the critical nature of this receptor in the inhibition of both tissue-specific and systemic inflammatory responses (315). Despite widespread tissue distribution, the activation of A_2AAR on CD4+ T lymphocytes is primarily responsible for reducing injury and inflammation in the liver, heart, and kidney (325). Responses to A_2AAR activation include the inhibition of the oxidative burst in neutrophils (326,327), reduced secretion of inflammatory cytokines by monocytes and macrophages (328), reduced platelet activation (329), inhibition of lymphocyte activation (325,330,331), and decreased neutrophil-endothelial cell adherence (332). In human monocytes, endotoxin-induced production of proinflammatory cytokines such as IL-6, IL-8, and TNF-α, is inhibited by A₂AR activation (328), whereas adenosine (10–100 μm) inhibits stimulated human monocyte production of IL-12 and enhances production of IL-10 (333). In mouse models, A_2AAR agonists inhibit LPS-induced release of IL-12 and TNF from isolated mouse macrophages (334,335,336). These studies demonstrate that the activation of a single receptor subtype, the A_2AR receptor, can broadly regulate inflammatory responsiveness by affecting production and secretion of a number of proinflammatory cytokines.

A large body of evidence suggests that the major signaling pathway that links A_2AAR activation and reduction of inflammation is the cAMP-protein kinase A pathway (326). A_2AAR agonists increase cAMP in a dose-dependent manner, an effect that is enhanced in the presence of a phosphodiesterase inhibitor (337). Activation of the NF-κB/Rel family of transcription factors is mediated by the phosphorylation and subsequent degradation of inhibitor of κB (IκB), which renders NF-κB free to enter the nucleus (Fig. 4; also reviewed in Ref. 338). Although a considerable amount of effort has focused on examining the protective effect of A_2AARs in animal models of ischemia-reperfusion injury, including lung, heart, and brain, no study has yet addressed the protective anti-inflammatory effect of A_2AAR agonists in islet transplantation. Given the functional consequences of nonspecific activation of ARs, identification of selective agents targeting specific ARs may be critical in developing novel intervention therapies for use in islet transplantation. Toward this end, numerous A_2AAR agonists have been synthesized and await testing. Preliminary studies have shown that A_2AAR agonist therapy in the peritransplant period reduces the islet dose required to achieve euglycemia and shortens the time-to-cure in syngeneic murine islet transplantation (339,340). Presumably, this improvement in islet graft outcome results from reducing nonspecific inflammation occurring in the graft microenvironment in the initial 3 to 5 d after transplant.

In addition to the ability of adenosine to signal through extracellular receptors to attenuate inflammation, there is also a class of cell internal proteins that provide cellular protection against inflammatory signals. These molecules, known as suppressors of cytokine signaling (SOCS), appear to play an important role in down-regulating cytokine signaling pathways (341,342,343,344). Briefly, these proteins possess SH2 domains, which permit interaction with phosphorylated signaling molecules. After engagement of their targets, which include cytokine signaling intermediates such as Janus kinase and various STAT proteins, the SOCS molecules, through a sequence known as a SOCS box, recruit the ubiquitin-transferase system (345,346). This recruitment results in degradation of the activated signaling intermediates and attenuation of the cytokine signaling cascade.

Overexpression of both SOCS-1 and SOCS-3 protects islet β-cells from cytokine-mediated damage in both spontaneous diabetes and transplant settings (347,348,349,350,351,352,353,354). However, a second line of investigation has also suggested that these proteins interact with the insulin signaling system by binding insulin receptor substrate-2 and can thereby promote insulin resistance (355,356,357,358,359). Because inflammatory cytokines up-regulate SOCS proteins, this new finding suggests that a chronic inflammatory state may promote insulin resistance. Therefore, whether transgenic expression of SOCS proteins would be of long-lasting benefit after transplant now holds less promise than initially anticipated. Whether conversion of SOCS proteins to pharmacological agents, which can be given short-term and have been used in sepsis models (360), would overcome this limitation is an open area for future investigation.

8. Connection between inflammation and hyperglycemia.

The concept that inflammation may promote insulin resistance and exacerbate hyperglycemia has also been confirmed in approaches described above (Section IV.C.6) utilizing antiinflammatory therapies for diabetes prevention and reversal. Interestingly, investigations using lisofylline and the “triple therapy” approach revealed that targeting the underlying inflammatory state also led to improved glucose sensitivity (306,361). This sensitivity did not seem to be accounted for by β-cell proliferation or changes in insulin secretion. Koulmanda et al. (306) further concluded that inflammation begets insulin resistance that can be reversed. Interestingly, insulin resistance in humans is strongly associated with prandial hyperglycemia, which in turn is associated with greater oxidant stress and activation of inflammatory prostanoids than fasting hyperglycemia (362). Although insulin resistance is not usually thought of as characteristic of patients with T1D, the metabolic syndrome, which is linked to insulin resistance, happens commonly in a large subset of patients with tight control of T1D as shown in the Diabetes Control and Complications Trial (363). The important interaction between insulin resistance and inflammation and how inflammation attenuates immune regulation remains a fertile area for future investigation. Overall, there appears to be an important linkage between inflammation, metabolic dysregulation, and tolerance induction that can be targeted to promote β-cell function and survival.

V. Inflammation and the Emerging Role of Dysmetabolism

A. Dysmetabolism drives inflammation

Although immune system-mediated inflammation may drive insulin resistance and hyperglycemia, there is also emerging evidence that dysmetabolism may cause inflammation and reinforce this barrier. Indeed, hyperlipidemia and hyperglycemia are potent inducers of proinflammatory cytokines including TNF-α and IL-1β (364). Although many initial studies documenting the connection between hyperglycemia and proinflammatory cytokines have taken place in patients with type 2 diabetes, some of the same metabolic complications, such as insulin resistance, are often present in the pretransplant patient with diabetes resulting from autoimmune disease. These studies have characterized increased levels of circulating TNF-α, IL-6, MCP-1, and vascular cell adhesion molecule as well as modest elevations in C-reactive protein in individuals with metabolic dysregulation (365,366,367,368). Interestingly, there are also reports that hyperglycemia may lead to endothelial cell dysfunction, a clinical feature that may be harmful to successful islet cell engraftment (369,370). In many cases, these elevated cytokines could be correlated to postprandial rises in blood sugar, suggesting that the rise in inflammatory gene expression may be glucose sensitive (17). This hypothesis has been further supported by a recent demonstration that in patients with metabolic syndrome there is a direct link between hyperglycemia and IL-6 and TNF-α expression at the genetic level (367). This dysregulation of the inflammatory axis may even occur in normal subjects because glucose loading promotes activation of NF-κB, a crucial proinflammatory transcription factor (371). In addition to these studies in patients with dysregulated glucose metabolism, one recent transplantation study by Montolio et al. (18) suggested that posttransplant hyperglycemia in a murine model of islet transplantation also augments production of TNF-α and IL-6.

In addition to immune-mediated effects, dysmetabolism can induce islet damage by other means. Hyperglycemia is a well-established factor contributing to the decline of the pancreatic β-cell, as initially shown in studies in which rats infused for 48 h with glucose showed severe reduction in β-cell glucose-stimulated insulin secretion (372). Chronic exposure to high glucose results in β-cell glucotoxicity (373), a term referring to reduced insulin secretion and synthesis and increased β-cell apoptosis resulting in large part from oxidative stress (374,375). Hyperglycemia-induced deficits in islet function include reduced glucose stimulation, insulin content, pdx-1 binding capacity, calcium homeostasis, and glucokinase expression (376,377,378,379,380). Hyperlipidemia can exacerbate islet damage by producing additional lipotoxic effects (381), particularly in combination with hyperglycemia, as recently reviewed (382).

Because many of the candidate patients for islet transplantation will have very labile diabetes, their metabolic state may influence their susceptibility to immune regulation. Moreover, their hypoinsulinemic and hyperglycemic state may foster the production of proinflammatory cytokines that are islet-toxic. Therefore, tight glycemic control and early transplantation of adequate islet mass may be important future avenues for both tolerance induction and the promotion of islet cell survival. Pharmacological therapy may also have a role because recent reports have suggested that treatment with statins may have a beneficial effect on inflammation (383). Additionally, treatment with angiotensin-converting enzyme inhibitors or angiotensin receptor blockers may prevent some of the endothelial cell dysfunction seen with hyperglycemia and may also decrease cytokine expression (369). Overall, there are a number of emerging opportunities to consider metabolic control as an important supporting factor to enhance the success of immunological interventions.

B. Management and assessment of the metabolism of the islet transplant recipient

Although we have highlighted the important interaction between the inflammatory state and the metabolic control of the patient with T1D, there are relatively little data and guidance for how to manage the glycemic excursions of the islet recipient in the peritransplant period. This period contains a number of substantial challenges, including introduction of new immunosuppressive agents that may cause hyperglycemia, alterations in peripheral insulin sensitivity, and successive changes in daily insulin requirement as islet function recovers after transplant. Developing an optimal approach to posttransplant diabetes management will require an integrated understanding of the alterations in physiology that follow transplant. Moreover, as improved glucose control and insulin treatment itself may both ameliorate the inflammatory state, more aggressive approaches including prandial control may be warranted rather than simply continuing standard therapy until β-cell function improves.

A substantial physiological change occurs after islet cell transplantation for the reversal of T1D. Unfortunately, most of our detailed data are from follow-up studies performed 2 to 6 months after transplant. However, by 6 months, there is a substantial reduction in hypoglycemic events and an improvement in the production of counter-regulatory hormones during hypoglycemia, suggesting a partial restoration of normal metabolism (59,384,385,386). At the same time point, there is generally also a reduction in daily insulin requirement and an improvement in glycosylated hemoglobin. More importantly, there is also evidence of increased insulin sensitivity in the same time frame. A study by Rickels et al. (387) demonstrated significant improvement in insulin sensitivity as calculated from a minimal model approach at a mean of 6 months after transplantation. This same study also demonstrated improvement in the dynamic handling of free fatty acids.

That insulin sensitivity improves over time is also a reminder that it is abnormal in the recipient who is receiving the islets. Whether insulin resistance plays a role in β-cell loss after transplantation is not known but merits further investigation. It is well known from in vitro studies of islet function that exposure to hyperglycemia leads to down-regulation of insulin production and decreased sensitivity to glucose (388,389). More long-term exposure to hyperglycemia can also lead to β-cell death. Interestingly, insulin may itself play a role in the islet response to glucose and in glucose-mediated islet toxicity. A recent study by Johnson et al. (390) demonstrated that low levels of insulin within the physiological range exerted an antiapoptotic effect on islet β-cells. This effect was lost at higher concentrations of insulin, leaving an open question as to whether insulin resistance and resulting higher levels of serum insulin either from exogenous administration or produced by the transplanted islets may unexpectedly enhance islet loss in the immediate posttransplant period.

Overall, there is substantial need for intensive monitoring of blood glucose after transplantation and for aggressive responses to the changes that are occurring in the host with respect to insulin needs, blood glucose levels, and insulin sensitivity. Optimizing these features may lead to substantial improvements in transplantation outcomes. There are now several techniques that allow for assessment of insulin sensitivity and action, β-cell function, and glucose variability and its associated risks for hypo- and hyperglycemia. Emerging tools, such as continuous glucose monitoring, and their application to blood glucose control in the posttransplant period may permit significant improvements in the metabolic management of the islet transplant recipient (391). This recipient metabolic conditioning may in turn be rewarded with reduced inflammation and a wider window for successful application of tolerance inducing therapy.

VI. Summary

Islet transplantation remains an exciting avenue to improve the quality of life for many individuals with T1D. If restoration of normal glucose tolerance can be achieved and immunological tolerance to islet tissue can be induced, this therapy would represent the definitive intervention for patients with this disease. However, there are still many barriers that prohibit the widespread application of this approach. First, there is the limited pool of islet donors. Current approaches seek to optimize the yield of islets from every donor pancreas, to preserve and expand their number ex vivo, and to ensure their maximal survival after transplantation. Strategies include a variety of supplementations to the culture conditions to allow islets to thrive as well as advanced techniques for modification of the islet β-cell that may include gene therapy and other approaches. Enhancement of these strategies may eventually allow for the use of living related donors, which would make a significant impact on our ability to perform islet transplantation.

An even more significant and challenging obstacle is the underlying defects in immune tolerance that created the diabetic state. Although diabetes is often described as an organ-specific immune disease, individuals with diabetes are also at high risk for the development of other comorbid autoimmune conditions, a fact that suggests this disease is one of global immune dysfunction (392). This chronically inflamed state may also create resistance to some of the therapies that are now being developed for immune modulation. Whether individuals with autoimmune disease will be resistant to these new therapies will require careful monitoring because the mouse model of T1D has proven highly resistant to therapies that successfully induce antigen-specific tolerance in numerous other normal murine strains (254,256,257). Moreover, to what extent chronic metabolic dysregulation further feeds inflammation requires more investigation but also represents a clinical opportunity. At the same time, understanding how to safely and aggressively manage the metabolic state in the immediate posttransplant period is a fertile area for the acquisition of new data and the introduction of new approaches. One possibility is the combination of islet transplantation with an emerging engineering approach known as the “artificial pancreas,” which combines continuous glucose monitoring with automated insulin delivery driven by closed-loop control algorithms. Although this engineering area is still in infancy, the first reports of closed-loop control experiments are promising (393,394,395). The combination of islet transplantation with artificial pancreas technology has the potential to have a greater effect than each of its components.

This review has highlighted the important role of inflammatory mediators underlying each of the barriers that we must overcome for successful islet transplantation. In this regard, the challenges that we face in achieving islet transplantation are each connected to the other and as such require an integrated approach. The chronic hyperglycemia and dysregulated immune function of the diabetic recipient create a hostile environment for islet transplantation. The transplantation procedure itself, which causes islet and hepatic tissue injury, further feeds this process. Our inability to gain early euglycemia and specific immunoregulation allows the inflammatory state to fester, leading to further islet loss and a self-perpetuating cycle. However, there are opportunities to intervene at each of these steps as well as to target the inflammatory state directly (Fig. 5). A combination approach to these barriers may disrupt this deleterious cycle and enhance the success of islet transplantation. Overall, the expansion of clinical experience and expertise in the isolation and transplantation of islet tissue, the opportunity to study and modify this tissue ex vivo, our growing appreciation of the underlying immunology and metabolism, and our understanding of the interactions between these factors indicate a bright future for the application of islet transplantation to T1D.

Figure 5.

Interventions to break the cycle of inflammation that prevents islet transplantation. Significant barriers to islet transplantation into the individual with diabetes include the underlying immune dysregulation, the presence of alloreactive lymphocytes that mediate graft rejection, ongoing islet cell destruction of both native islets and graft, and chronic hyperglycemia. Each of these factors either contributes to or is exacerbated by an underlying state of chronic inflammation. As discussed, there are interventions that can be directed to each of these factors in the transplant process (shown in bold). Immune regulation can be restored by expansion and replacement of regulatory T cells. This expansion may aid in controlling auto- and alloreactive lymphocytes that can be further targeted by immunosuppressive or immunomodulatory agents. Down-regulation of the immune response may temper inflammation and permit enhanced survival of islet transplants. These transplants can be further protected by transplantation of an adequate mass and by stress-resisting modifications. Achievement of adequate islet mass may correct the hyperglycemic state, which can also be addressed before and after transplant by intensive therapy and monitoring approaches. Promotion of euglycemia may further attenuate inflammation, which may support regulatory cell function. In addition to therapies targeting the individual processes, we have also discussed new interventions to directly target the inflammatory state. Together, these interventions can combine to support and sustain each other and enhance the opportunity for successful islet transplantation.