What's New and What's Hot in Transplantation: Basic Science ATC 2003
*Corresponding author: Peter S. Heeger, firstname.lastname@example.org
It has been approximately 50 years since the initial descriptions of acquired transplant tolerance, and our understanding of the immune response to a transplanted organ has progressed enormously during the ensuing years. Recent studies have shed new light on the molecular and cellular basis of transplant rejection, have better defined the mechanisms of allograft tolerance with particular emphasis on a role for regulatory T cells, have identified important new hurdles to overcome in order to prolong allograft survival, have brought xenotransplantation closer to becoming a clinical reality, and have led to the development of novel techniques that may permit analysis of immune responses to transplanted organs in vivo.
This year marks the 50th anniversary of Peter Medawar's classic description of acquired immune tolerance (1), an observation that helped to usher in the modern era of transplantation science and that earned him a share of the 1960 Nobel Prize in Physiology or Medicine. Our understanding of the pathogenesis of transplant rejection and its prevention has progressed enormously in the subsequent years, with important new insights now developing as rapidly as we can publish them. This report represents a brief summary of selected, relevant findings in transplantation science published from 2002 through mid 2003, including newly reported data presented at the 2003 American Transplant Congress (ATC).
New Findings Relevant to Mechanisms of Transplant Rejection
Priming of alloreactive T cells outside of secondary lymphoid organs
T lymphocytes are central mediators of acute and chronic allograft rejection. Current immunologic dogma dictates that activation of naïve T lymphocytes is tightly controlled so as to prevent priming of pathogenic, autoreactive, effector T cells that could damage the host. To partially achieve this goal, it is thought that activation and differentiation of naïve T cells into effector T cells requires antigen presentation by professional antigen presenting cells (APCs) within the secondary lymphoid organs (lymph nodes or spleen) and in the context of appropriate costimulatory signals (2). Cognate interactions between naïve T cells and nonprofessional APCs, particularly those interactions occurring outside of the secondary lymphoid organs, are thought to be tolerogenic. Reports from two groups have challenged this notion. Rosengard and colleagues provided evidence that donor endothelial cells may be able to prime naïve alloreactive CD8 T cells in vivo (3,4). This group used a cleverly designed chimeric, transplantation model, in which the relevant alloantigens were expressed on graft endothelium but not on graft-derived hematopoietic cells. Both alloantigen-specific, TCR transgenic CD8 T cells and polyclonal CD8 T cells were primed following transplantation of the chimeric heart grafts, and the primed T cells destroyed the organ. In other studies, Allegre and colleagues showed that lymphotoxin-deficient mice and lymphotoxin β-receptor-deficient mice can reject skin and cardiac allografts despite a lack of lymph node and spleen tissue (5). Overall, the data provide suggestive, albeit not fully conclusive, evidence that vascularized transplants have the potential to activate naïve T cells in the absence of secondary lymphoid organs, when no other alloactivation pathways are present. Theoretically, this priming and effector pathway could be operative for the lifetime of the allograft and could thus contribute to the development of chronic graft injury. It remains to be seen whether activation of naïve alloreactive T cells outside of secondary lymphoid organs contributes to the alloimmune repertoire under more standard conditions.
Blurring the allorecognition pathways
Recipient T cells can be directly primed by donor MHC: donor peptide complexes expressed on donor-derived APCs that have migrated out of the graft into the recipient lymphoid organs (6–8). Donor-derived transplant antigens can also be processed and presented by recipient APCs and thereby prime recipient T cells through the indirect pathway (6–14). Multiple studies by numerous investigators in transplant immunology over the last decade have provided a large body of evidence showing CD4+ T cells can be indirectly primed to donor antigens. These primed CD4+ T cells are important mediators of acute and chronic allograft injury, and they can play a dominant role in tolerance induction (6–18). Lechler and colleagues have now added a new wrinkle to the allorecognition story. This group first confirmed a previous report that T cells can acquire intact MHC molecules found on the surface of other cells, express these donor-derived MHC molecules, and present the alloantigen to other T cells (19,20). Unpublished reports from this group presented at the ATC 2003 provide evidence that recipient APCs can acquire donor MHC molecules in an analogous fashion (presented as part of a symposium on dendritic cell biology by R. Lechler, on June 1, 2003). Under such conditions, donor MHC molecules seem to be expressed on the surface of recipient cells, permitting ‘direct’ recognition of donor antigen found on a recipient APC. In theory, this ‘semidirect’ pathway could result in T-cell tolerance if the antigen is presented to naïve T cells in the absence of costimulatory signals, or alternatively could contribute to T-cell activation if expressed in the presence of appropriate costimulation. Whether this antigen presentation pathway plays a significant role in priming or tolerizing donor-reactive T cells to transplant antigens in vivo remains to be determined, but the possibility remains an intriguing one.
Alloimmunity and autoimmunity are interrelated
In another twist, emerging data from several laboratories have provided convincing evidence that autoreactive T cells can contribute to destruction of a transplanted organ. Allograft transplantation primes pathogenic, recipient-MHC-restricted T cells specific for peptides derived from cardiac myosin (heart grafts), collagen V (lung transplants), heat-shock proteins (skin grafts and heart grafts) and some unknown autoantigens (12,21–30). The primed autoreactive T cells are not simply innocent bystanders, because (1) they can be isolated from allografts undergoing rejection, (2) immunization with these autoantigens before transplantation accelerates allograft rejection, and (3) induction of a pathogenic immune response to these autoantigens through experimental immunization can precipitate rejection of an isograft. Interestingly, the primed, autoreactive T cells capable of rejecting a transplanted isograft do not seem to cause injury to the native organs of the recipient. A number of questions remain unanswered, including whether autoimmunity only develops in the context of uncontrolled alloimmune responses, whether the autoreactive T cells are susceptible to conventional immunosuppression, and whether local inflammation to a native organ (for example from an infection) during the rejection process can precipitate autoimmune-mediated injury of the native organ.
The detection of autoreactive T cells following transplantation might be anticipated. While many indirectly presented peptides derive from donor MHC molecules, any antigen found in donor cells could theoretically be processed and presented by recipient APCs. The proinflammatory state of the transplanted organ owing, in part, to surgical trauma and ischemia reperfusion injury, along with the enormous antiallograft T-cell immune response focused towards donor MHC molecules, could permit priming of autoreactive T cells. Chemoattractants produced by the inflamed organ could preferentially facilitate migration of the activated autoreactive T cells to the graft. Notably, the link between alloreactivity and autoreactivity is not limited to mouse models; one fascinating report from the ATC 2003 showed that lung allograft recipients with bronchiolitis obliterans preferentially expressed autoreactive, collagen-V-specific T cells and auto-antibodies in their peripheral blood when compared with control patients (31).
Perhaps even more intriguing are the observations that induction of tolerance to an organ-specific autoantigen before transplantation can delay or even prevent rejection of a subsequently placed allograft (23,27). Overall, these experimental data suggest that one can harness a naturally developing, autoreactive regulatory T-cell repertoire, and that expansion of such T cells can result in prevention of a pathogenic alloimmune response. As mechanisms become better elucidated, this concept of manipulating autoimmunity as a means of controlling alloreactivity could have important implications for human transplantation.
Complexities of costimulation
In addition to T-cell receptor (TCR) interaction with peptide/MHC complexes, additional costimulatory signals delivered to the responding, alloreactive T cells regulate the activation, expansion and ultimately the collapse of the alloreactive immune repertoire. It is now well established that T-cell-expressed CD28 and CD154 interactions with APC-expressed CD80/86 and CD40, respectively, are important signals involved in initiating differentiation of naïve T cells into effector cells (32). CTLA4 expressed on T cells has also been shown to interact with CD80/86, but ligation of this molecule results in down-regulation of the T-cell response. A number of additional costimulatory pathways have recently been identified, and over the past year it has become increasingly apparent that many of these costimulatory molecules can participate in the activation or inhibition of alloreactive T cells (32). Blockade of CD134/CD134L interactions in vivo, for example, synergies with CTLA4-Ig therapy to prolong heart graft survival in mice (33). Perhaps more surprising was the fact that this approach was effective in prolonging allograft survival even in the context of presensitized donor-reactive T cells (33). The PD–1/PD-1 L pathway has additionally been identified as an important inhibitory signal of the alloimmune repertoire (34). Multiple abstracts presented at the ATC 2003 confirmed and extended our understanding of this novel inhibitory pathway (35,36). Other costimulatory signals of potential relevance include CD70/CD27, LIGHT, BAFF, and ICOS (32,37,38). How these molecules interact with more standard costimulatory signals to control the alloimmune response will be an exciting and rapidly evolving story to follow over the next year.
Allorecognition at the graft site
Alloreactive effector T cells down-regulate surface expression of the lymph node homing receptor CD62L, up-regulate expression of the activation/adhesion molecule CD44, leave the secondary lymphoid organs, and circulate in the periphery where they can re-encounter their specific ligands directly on graft cells within the transplanted organ (39). T cells only capable of recognizing and directly interacting with donor MHC/peptide complexes on donor cells within the graft can indeed mediate acute graft rejection (40,41). Alternatively, indirectly primed T cells entering an allograft can re-encounter their specific ligands on graft-infiltrating recipient APCs that have processed and presented donor antigen in the graft itself (6,14,18). This interaction can theoretically result in bystander killing and/or delayed type hypersensitivity (DTH) reactions that lead to local tissue damage. Work from our laboratory recently provided evidence for an additional unique effector pathway of T-cell-mediated transplant rejection (42). Using T-cell-receptor transgenic T cells of defined specificity, we showed that endothelial cells have the ability to process antigenic proteins derived from exogenous sources, and to present the relevant peptides to recipient CD8 T cells in the context of recipient MHC class I. We further showed that T cells capable of recognizing cross-presented antigen on endothelial cells could mediate skin graft rejection (42). Whether this antigen presentation pathway will have important implications for vascularized transplants in normal mice or humans remains to be defined. In addition to the potential relevance to transplantation, cross-presentation of exogenous antigens by endothelial cells may be an important target of autoimmune responses. Consistent with this contention, Savinov and colleagues provided strong evidence that endothelial cells can cross-present insulin (43) and that the endothelial expression of insulin-derived determinants was essential for the development of murine autoimmune diabetes.
Innate immunity and transplantation
There is increasing awareness by the transplant immunology community that there is an important interrelationship between the innate and adaptive immune systems that contributes to antiallograft immunity capable of mediating rejection. Studies published by Goldstein and colleagues, for example, demonstrated that mice deficient in Toll-receptor-signaling pathways (MyD88 knockout mice) were unable to reject minor antigen disparate skin grafts (44). An additional intriguing and unanticipated result was the observation that wild-type recipient mice did not reject allogeneic renal transplants deficient in complement component C3 (45). These latter studies suggest that complement deficiency can affect priming and/or effector function of alloreactive T cells, further highlighting the interaction between adaptive and innate immunity. Better comprehension of how innate immunity influences alloreactive T-cell function could theoretically provide new therapeutic targets capable of prolonging graft survival.
Advances in Tolerance
Immune tolerance is the ultimate goal of the transplant scientist and transplant clinician. Interest in inducing, understanding and monitoring immunologic tolerance remains high, particularly in light of the funding of the cooperative Immune Tolerance Network, a consortium of studies and institutes dedicated to bringing immune tolerance to the clinic. Studies over the last year have provided a new basic insight into mechanisms of tolerance, have identified new hurdles to tolerogenesis, and have in some selected cases provided solutions to overcome these problems.
Regulatory T cells are here to stay
Accumulating evidence suggests that regulatory T cells may be essential mediators of peripheral tolerance (46) to autoantigens and transplant antigens. Distinct subsets of regulatory T cells seem to emerge directly from the thymus as mature lymphocytes with a predetermined and defined function. In addition, naïve peripheral, αβ TCR-expressing T cells can differentiate into activated T cells with regulatory capacity under certain conditions. Regulatory T cells therefore represent a diverse set of lymphocytes that function through multiple, distinct effector mechanisms, and presently can only be categorized based on certain cell-surface markers and or cytokine-secreting properties.
Five to 10% of peripheral CD4+ T cells coexpress the IL-2Rα chain, CD25 (CD4+ CD25+), and are thought to be direct thymic emigrants with regulatory function (47,48). Additional markers found on this form of regulatory T cell include CTLA4 and the glucocorticoid-induced tumor-necrosis factor receptor family related gene (49,50). CD4+ CD25+ regulatory T cells apparently suppress other T cells through direct cell:cell contact, although some reports suggest that suppressive cytokines (IL-10 and or TGFβ) may be involved, and detailed mechanisms remain to be elucidated (50–52). The antigen specificity of CD4+CD25+ regulatory T cells is not known, and it is furthermore not fully understood whether antigen specificity plays any role in mediating their regulatory effects. Despite a paucity of information regarding exact mechanisms of regulation, recently published studies by three individual laboratories revealed that a newly described transcription factor, Foxp3, can act as the ‘master switch’ for inducing regulatory cell function in this cell population (53–55). In support of this contention, Foxp3 is preferentially expressed in CD4+ CD25+ regulatory T cells, transduction of the Foxp3 gene into naïve CD4+ CD25− T cells results in a conversion into T cells with regulatory function, Foxp3 KO mice exhibit autoimmunity, and humans with Foxp3 mutations develop autoimmune disease. In addition to the novel insight provided by this work, the studies provide scientists with an identifiable molecular marker of regulatory T cells that could be invaluable for future studies.
Other forms of regulatory T cells have also been described and include NK T cells, TGFβ-secreting Th3 CD4 cells, IL-10-secreting Tr1 CD4 cells, CD8+ regulatory T cells and TCR negative regulatory cells. The interested reader is referred to several thorough reviews for a detailed discussion (46,52,56).
Although a detailed understanding of regulatory T-cell ontogeny, differentiation and mechanisms of action remain only partially understood, the phenomenon of suppression/regulation is indisputable and is particularly relevant to transplantation science. A multitude of publications on regulatory T cells in transplantation have appeared in press over the last 12–24 months. One particularly intriguing report demonstrated that regulatory T cells can be isolated from the tolerant organ, thereby suggesting that these regulatory cells function locally to inhibit pathogenic antiallograft immune responses (57). Understanding how to harness regulatory T cells and defining how they function in the context of transplantation will continue to be an extremely ‘hot’ area of research over the next several years.
Newly recognized problems in transplantation tolerance
While our basic understanding of mechanisms of transplantation tolerance is slowly improving, it has become apparent that tolerance-inducing strategies will need to be adjusted in order to overcome newly recognized hurdles.
One such hurdle is the memory T cell. Memory T cells exhibit specific properties that permit rapid and effective control of infectious agents previously encountered by the host. These features, which include stable, differentiated cytokine phenotypes, relatively low activation thresholds and costimulatory requirements (when compared with naïve T cells), the ability to rapidly engage effector functions, and the ability to migrate to peripheral tissues, are integral to host defense (39). In contrast, these same features are likely to be deleterious in the context of attempting to induce transplantation tolerance. In support of this contention, several groups, including Valujskikh and colleagues, recently published studies showing that memory donor-reactive T cells prevent tolerance induction in mouse models (58–61). As noted earlier, targeting newly identified costimulatory pathways may be one approach to tolerize these problematic T cells. It is anticipated that studies defining the mechanisms of tolerance resistance and elucidating novel approaches to specifically controlling donor-reactive memory T cells will remain an important focus of transplantation research over the ensuing several years.
In addition, long-term survival of allografts following attempts at tolerance induction may be limited by the development of donor-reactive alloantibodies (62). Such antibodies have been shown to contribute to transplant vasculopathy and late organ failure. Several investigators are now attempting novel strategies to specifically tolerize B cells in the context of transplantation. One approach that seems to be highly effective in mice is to administer donor bone marrow in its natural environment using fragments of intact bone as a tolerizing stimulus at the time of the transplant (63).
Progress in Understanding Mechanisms of Pharmacological Immunosuppression
FTY720 is an immunosuppressant medication in phase 2 human trials with a mechanism of action that differs from all other agents in general use. Administration of this drug results in a rapid and dramatic decrease in the number of peripheral blood T cells, shunting the T cells into secondary lymphoid organs. Studies published by two groups have now identified the molecular basis of FTY720s effects by demonstrating that it binds to a sphingosine-1-phosphate receptor, and thereby prevents naïve and activated T cells from leaving lymph node tissues (64–66).
As noted in the earlier discussion, T-cell activation/differentiation into effector lymphocytes requires costimulatory signals, one of the most important of which is T-cell-expressed CD28 interaction with CD80/86 on the APC. Studies in animal models have shown that CTLA4-Ig, a fusion protein that binds with high affinity to CD80/86, can prolong graft survival. Based on previous in vivo and in vitro data, the effect of CTLA4-Ig has largely been attributed to induction of anergy because the alloreactive T cells interact with their specific ligand in the absence of CD28 signals. An eye-opening paper published in Nature Immunology in late 2002 demonstrated that CTLA4-Ig may in fact be functioning through an entirely unanticipated mechanism of action (67). This work showed that CTLA4-Ig interaction with CD80/86 signals through the APC and results in up-regulation and secretion of an immunoregulatory enzyme called indole amine 2,3 dioxygenase (IDO). IDO catabolizes tryptophan, and the absence of this amino acid (and/or the presence of a tryptophan catabolite) actively inhibits T-cell function. This latter mechanism of action, rather than induction of anergy as a result of a lack of costimulatory signals through the T-cell itself, seems to explain the long-term islet graft survival induced by CTLA4-Ig in some experimental animal models.
Xenotransplantation Is Getting Hotter
Perhaps one of the most exciting advances in transplantation science in several years is the production of 1,3 galactosyl-transferase-deficient pigs (68,69). The α1,3 galactose residue expressed on cells of nonprimate mammals is the dominant target antigen of preformed xenoreactive antibodies that mediate hyperacute rejection of xenografts (70). The newly developed 1,3 galactosyl transferase knockout pigs lack the enzyme that produces this residue and the α1,3 galactose antigen is not found on the donor organs (68,69). Preliminary unpublished studies from the Massachusetts General Hospital and Immerge BioTherapeutics, Inc. (Sachs, Yamada Cooper et al., personal communication) suggest that there is, in fact, no hyperacute rejection of 1,3 galactosyl transferase knockout pig kidneys or hearts transplanted into nonhuman primate recipients. If confirmed, the development and large-scale production and breeding of this knockout pig significantly enhances the possibility that xenotransplantation could be a real solution for the organ shortage problem. Of course, a number of additional barriers will still need to be overcome, including understanding how to control the xenoreactive T-cell repertoire and addressing the potential of porcine retroviruses to infect humans, before xenotransplantation from pigs to humans becomes a clinical reality.
In addition to the important scientific advances, newly developed research techniques are now being applied to transplantation models, thereby providing novel insights into mechanisms of T-cell biology and transplant rejection in vivo. Intravital microscopy and two-photon microscopy, for example, permit real-time visualization of specific lymphocyte interactions and trafficking patterns in live anesthetized animals (71,72). These types of direct visualization analyses have begun to yield a number of unanticipated results that could not be inferred from traditional experimental designs. In addition, one innovative report published in early 2003 used positron emission tomography scanning to repeatedly track labeled, antigen-specific T cells as they accumulated at the site of a tumor in live animals over a 2-week period (73). One can easily envision how this type of approach might be modified to study transplant-reactive T cells responding to an allograft and/or to follow the fate of labeled islet transplants in vivo. It is likely that the further development of technology for in vivo imaging of immune responses will result in better insights into the mechanisms of transplant rejection and tolerance that could ultimately be applicable to human allograft recipients.
Funding source: Dr Heeger is supported in part by grant AI43578-01 from the National Institutes of Health.