What's New–What's Hot in Basic Science: American Transplant Congress 2004

Introduction

Basic transplantation science is inherently linked to the practice of organ transplantation, and is thus appropriately viewed in the light of translational intent. What is termed basic to transplantation may indeed be superfluous to physiological human immunity, since transplantation is a most artificial (although medically relevant) condition. Furthermore, most of what is reported as new in our field is usually novel in terms of its relevance to our discipline. That being said, one significant event of this past year has been the public identification by the Director of National Institutes of Health (NIH), Elias Zerhouni, that translational science is of critical importance and requires increased attention to achieve an optimal balance between academic progress and tangible improvements in healthcare delivery (1). As evidenced by the scientific program at the 2004 American Transplant Congress (ATC), transplantation stands poised to deliver exceptional gains by linking basic human immunology with therapeutic benefit. While practicality is at times viewed as the lesser sibling of basic science, in transplantation the two are melding nicely.

The Variable and Complex Nature of Alloimmunity

Alloimmune responses have been understood largely through the results of experiments measuring isolated in vitro responses against defined antigens at fixed moments in time. While helpful in understanding the mechanics of immunity, this approach often fails to predict in vivo allograft rejection, and even in vivo models infrequently explain adult human allograft behavior. At this year's ATC, several investigators have begun to place fundamental immune mechanisms into the context of an interconnected maturing immune system.

Heterogeny, the concept that immune specificity is not exclusive to a single antigen but exists in a cross-reactive spectrum of specificities with structurally related antigens, now seems well entrenched in transplant biology [reviewed in (2)]. This concept, more than any other, is helping to explain the gaps between the test tube, mouse and clinic. Experimental studies by Adams and the Emory group appeared this year demonstrating that prior viral infections alter subsequent alloimmune responses and in particular adversely affect tolerance induction protocols (3). Furthermore, while the influence of an active anti-viral infection could be predicted as “stirring up trouble” for an allograft recipient, the group demonstrated at the ATC that even latent viral infections impair tolerance (4).* This was hypothesized to be due to cells with allospecific cross-reactivity persisting in a heightened activation state that precludes dependence on accessory signals like co-stimulation. The immune repertoire is thus being viewed not only as a product of thymic selection, but also as a mosaic of past immune experience and ongoing suppressive responses to latent virus. As if the polymorphism of the major histocompatability complex (MHC) were not enough, we now recognize that each person's uniqueness is a product of polymorphism plus environmental response. It is no wonder that one well-attended ATC symposium was on complexity theories.

Closely related to heterogeny is the recognition that immune memory is an important aspect of the human alloimmune response [recently reviewed in (5)]. Through heterogeny, allospecific memory need not be a result of alloantigen exposure. This year it was demonstrated in non-human primates (NHPs) (6) and humans (7) by Pearl et al. from the NIH that depletional induction antibodies possess less efficacy against memory cells. Memory phenotype cells were shown to predominate following alemtuzumab and rabbit anti-thymocyte globulin-mediated T-cell depletion. In related experiments, calcineurin inhibitors (CNIs) were shown in vitro to be particularly efficient in preventing memory cell activation. As CNIs target the T-cell receptor (TCR) more than accessory or co-stimulation molecules, this fits well with our understanding of how antigen-experienced cells are activated. This observation has important implications given the considerable enthusiasm for depletional induction evident at this year's meeting. Allospecific memory cell formation was also characterized relative to viral-specific memory by the Lakkis laboratory from Yale. In rodent studies, they showed that both CD4 and CD8 memory could be generated in the absence of secondary lymphoid tissues (8). The same group also published work this year showing that allosensitized T-cells mediate chronic rejection when encountering a healed allograft and acute rejection when presented with a freshly reperfused allograft indicating that the phenotype of memory-driven rejection may be determined as much by the state of the allograft as by the effector arm of the immune system (9).

With memory being interjected into the transplant vernacular, it is interesting to note that some in the basic sciences are questioning whether memory exists at all (10,11). A strong teleological and experimental argument has been put forth by Zinkernagel suggesting that T-cells only maintain a memory-like posture in the face of persistent antigen, and that viral memory only indicates some subclinical source of viral antigen, or of antigen that induces heterogeny. Given that our intent is to prevent antigen (graft) elimination, this may bode poorly for peripheral tolerance mechanisms.

Another relevant concept clearly developed this year is that of homeostatic activation. Homeostatic proliferation is apparently spontaneous activation (or pseudo-activation) driven by TCR interactions with self-peptide: MHC complexes occurring in a lymphopenic environment. Cells gain several qualities of antigen-experienced cells in that they lose the requirement for many accessory signals for activation and thus become pseudo-memory T-cells. Originally defined in the artificial world of T-cell deficient animals, homeostatic proliferation is now recognized to occur in patients treated with lymphodepleting protocols. This is especially true in adults where reduced thymic output limits the ability to reconstitute a sufficient T-cell repertoire. Studies in mice and monkeys were presented this year by Turka at the University of Pennsylvania that established homeostatic repopulation as another barrier to tolerance and a relevant foe to anti-rejection therapies (12). Aggressive depletion was shown to induce homeostatic proliferation that produces memory-like cells capable of resisting tolerance induction and promoting rejection. Again, given the clear enthusiasm for depletional trials in the past several years and at this year's ATC, the concept that depletion is not solely an immunosuppressive maneuver should be adopted by clinicians.

As alloimmunity becomes increasingly seen as a complex phenomenon, the tools used to characterize it have kept pace by becoming more comprehensive. Both applications of genomic and proteomic techniques were well represented at this year's ATC, particularly in the Monday invited Basic Science Symposium, and a Wednesday morning session by Sarwal reflecting important published advances from the past year (13–15). Cellular rejection was shown to have several subtypes with widely varying prognoses that were characterizable by genomic analysis. This suggests that treatment could be tailored to specific types of rejection based on the specific molecular fingerprints. In addition, the non-invasive urine proteomics approach appears to have matured to the point where a prospective trial investigating its diagnostic ability is warranted. Although the applicability of these techniques is being defined, it is clear that powerful multiplex assays are capable of being applied to transplanted tissues and will certainly yield data with mechanistic and diagnostic significance.

Newly Targeted Pathways for Immunosuppression

Novel pathways and molecular mechanisms rarely make their debut at the ATC; however, transplantation is frequently a field in which appreciation for the therapeutic potential for their manipulation becomes apparent. This year several pathways were newly highlighted as relevant to transplantation. Most prominent among these were the Janus kinase (Jak)-related cascades and associated regulatory molecules (16–24).

The Jaks have been established as critical proximal signal transduction proteins for cytokine signaling, particularly for the common γ-chain cytokines [recently reviewed in (16,17)]. They convey signals to the signal transducers and activators of transcriptions (STATs), DNA binding proteins that translocate to the nucleus to influence gene transcription. A series of resulting transcripts known as suppressors of cytokine signaling (SOCS) proteins feed back onto Jak proteins to curb their response to cytokines. The pattern of Jak/STAT/SOCS activity is known to influence the Th1/Th2 phenotype of activated cells and small molecules capable of exploiting these tuning mechanisms have been hotly sought after.

This year the first of several Jak inhibitors, the Jak3 inhibitor CP690–550, made its debut in transplantation (18). Although the efficacy was not exceptional, it is clear from abstracts presented from Borie's Stanford group that this will add a new option in immune manipulation (19,20). Another member of this class, the Jak3 inhibitor NC1153, was debuted at the ATC by Kahan's group in Houston, and was touted as having more selectivity for Jak3 versus other Jak family members and to have a more tolerable side effect profile (21,22). Other abstracts (23,24) discussed nuances of these proteins and the take home message is that these compounds will provide a very attractive option to apply in immunosuppressive and tolerogenic protocols. The optimal use of these and related compounds will likely be forthcoming.

In a related line of investigation that has gained prominence in the past year, the differential expression of SOCS proteins has recently been elucidated (25,26). SOCS expression appears to be critical in the maturation of dendritic cells (DCs) and in Th1 versus Th2 differentiation. Zhang et al. from Houston presented interesting data demonstrating that SOC3 uniquely identified tolerized cells in mice, a finding with potential for evaluating tolerant patients and interrogating cells prior to adoptive transfer (27).

Vascular endothelial growth factor (VEGF) was also highlighted by several investigators this year particularly with reference to chronic allograft nephropathy. A well-established angiogenesis factor of importance in tumor biology, wound healing and chronic inflammation, VEGF is also being shown to have proinflammatory effects, and most recently has been related to chemokines with known relevance to monocyte chemotaxis. Studies this year have shown it to induce the chemokines MCP-1, IL-8 and IP-10 (28), and to be expressed in association with acute rejection in animals and in humans (29,30). New data were presented by Briscoe's lab at the Brigham and Women's Hospital highlighting the chemotactic effects of VEGF and showing that VEGF-mediated leukocyte trafficking and function is dependent on MCP-1 (31). It has been known that sirolimus inhibits VEGF-induced angiogenesis—a theoretical factor in favor of mTOR inhibition. In data providing nice clinical correlation to the growing in vitro and rodent data regarding VEGF, it was shown by a collaborative study between the University of Wisconsin and the NIH that when patients could be maintained rejection free on sirolimus monotherapy, that their intragraft expression of VEGF (and type I collagen) was significantly reduced compared to rejection-free patients maintained on a CNI-based regimen (32). This was associated with decreased interstitial fibrosis in patients maintained on CNI-free regimens suggesting that when patients could be maintained without rejection with sirolimus the benefits of CNI avoidance could be linked with the VEGF pathway. This was not to say that on the whole, CNI avoidance was best because the risk of acute rejection without CNIs could potentially negate these benefits. Data were also presented challenging the concept that sirolimus was unique in its VEGF inhibition. In surprising data, again from Briscoe's group, mycophenolate mofetil was also shown to inhibit VEGF-inducted angiogenesis in vitro (33).

Innatity

This year provided considerable insight into the function of innate immune cells and pathways. The most prevalent emerging pathway is that of the Toll-like Receptors (TLRs). Originally described as molecules establishing dorsal–ventral polarity, and subsequently fungal immunity, in Drosophila, TLRs have grown into a family of receptors that bind various pathogen-associated molecular patterns and critically effect the activation of antigen presenting cells [recently reviewed in (34)]. Interestingly, this year it was also shown that TLRs provide direct signals to T-cells and thus influence T-cell responses independent of APCs (35). Although generally associated with responses to pathogens, TLRs have recently been evaluated for their relevance to allotransplantation. Mice without the critical signal TLR signal transduction protein myeloid differentiation primary-response protein 88 (MyD88) were shown this year to be unable to reject minor mismatched (H-Y antigen) allografts (36). The lack of MyD88 had minimal effect on fully mismatch allograft survival. Nevertheless, additional evidence presented at the ATC by Goldstein's lab at Yale demonstrated that MyD88 is critical for production of Th1-associated proinflammatory cytokines (interferon-γ and IL-2, but not IL-4) after an allo-DC stimulus or after exposure to DCs activated by LPS (37). These data thus help to characterize this as another relevant pathway connecting innate and acquired and perhaps alloimmunity.

TLRs were also highlighted in studies trying to establish a biological explanation for the apparent pro-tolerant posture of hepatic DCs. It was shown by Thompson's Pittsburgh group that liver DCs have reduced expression of TLR4 compared to splenic DCs, and that this correlates with a reduced ability of LPS-stimulated liver DCs to activate T-cells to express interferon-γ (38). This study provided a nice explanation for hepatic DC behavior, but more importantly provided a potential marker for pro-tolerant DCs or for therapeutic intervention to skew a response toward tolerance.

The relevance of the innate response to transplantation was also demonstrated in several platelet studies this year. Two studies, one in mice (39) and another in humans (40), showed that platelet-derived CD154 was sufficient to induce DC maturation and that platelet activation via thrombin could be a potential bridge between surgically induced trauma and acquired immune system activation. In a study presented at the ATC by Xu from the NIH it was shown that CD154 derived from platelets was sufficient to induce allograft rejection in CD154-deficient mice and that blockade of platelet-derived CD154-induced long-term allograft survival when T-cell CD154 was absent (41).

Thus, innate responses to infectious organisms, trauma, ischemia and other perturbations, long recognized clinically to influence allograft rejection, are now being mechanistically linked to acquired alloimmunity. This additional level of understanding adds complexity to our understanding of allograft rejection, but also makes it easier to reconcile known associations between injury and rejection.

The Expansion of Regulation

T-cells with the ability to attenuate or modify immune responses have long been thought to exist, although the transplant community's acceptance of regulatory or suppressor populations has waxed and waned through the decades. One thing that was apparent from this year's Congress is that the concept of regulatory T-cells has matured considerably. Fewer investigators now insist upon a single definable phenotype as the regulatory population, and more accept a variety of types of governance in different situations.

Strong evidence in support of regulatory T-cell activity in draining lymph nodes (as opposed to locally in the allograft) was presented in the opening session of the ATC by Markmann's group from the University of Pennsylvania (42). Through an elegant series of skin graft studies in mice, it was shown that suppression could occur in the absence of an intact graft. That is, when the allograft was removed, regulatory cells persisted referable to CD4+CD25+ T-cells in the nodes draining the prior allograft site. This speaks against a peripheral surveillance function for regulatory cells in lieu of a more centralized control strategy.

Expanding on work examining the effect of donor-specific transfusion on CD4+CD25+ regulatory T-cell (Treg) formation (43) data were presented by Wood in an invited lecture critically examining the role of CNI- and sirolimus-based immunosuppression on Treg formation. One growing clinical concern is whether immunosuppression negatively influences potentially beneficial regulation. Quoting studies by Bushell in her laboratory, Wood showed that CD4+CD25+ Tregs were inhibited by CNIs given at the time of donor antigen exposure, but not after the formation of the regulatory function. Sirolimus did not alter Treg function regardless of when it was given. This suggests that while CNIs may alter the formation of Tregs, they may not inhibit their function once formed.

The prospect of harnessing Tregs gained momentum with reports from Lechler's London group that Tregs could be generated ex vivo in humans (44). Expanding upon this published work the group showed that this strategy could be used in mice to induce tolerance (45). In an invited Basic Symposium, Lechler also presented the concept of semi-direct presentation, the uptake and representation of intact MHC molecules, to explain the phenomenon of unlinked help and suppression. Data supporting the exchange of molecules between donor and recipient was also presented supporting this concept (46).

While the prominence of CD4+CD25+ regulatory cells has grown in recent years, a surprising resurgence of interest in and evidence for CD8+ Tregs reappeared over the past year (47–51). Minor histocompatibility-specific CD8+ T-cells were isolated from renal transplant recipients with functional tolerance and shown to have antigen-specific suppressor function in a trans-vivo DTH assay (47) and CD8+ Tregs responsive to latent TGF-β were shown to be associated with metastable tolerant renal allograft recipients (48). Interest in the CD8+CD28− phenotype as a source of regulatory T-cells emerged (49–51). One proposed mechanism for CD8+CD28− involves the T-cell-mediated contact dependent induction of immunoglobulin-like transcripts (ILT) 3 and 4 on DCs rendering them co-stimulation poor and pro-tolerant to CD4+ T-cells [reviewed in (50)]. New data regarding the function of this subset was presented by Suciu-Foca's group from Columbia showing that allograft endothelial cells could be induced to express ILT3 and ILT4 as a result of contact with CD8+CD28− T-cells from allograft recipients, perhaps inducing a local graft accommodating effect. As ILT3 and ILT4 are easily detected immunohistochemically, this may provide a method of detecting an accommodated allograft, and suggests that regulation or suppression need not involve direct contact between the suppressor and effector cells.

A similar link between accommodation and regulatory immunity in fact did not involve direct cell contact or T-cells. It has recently been observed that low doses of anti-HLA antibodies have the effect of inducing cell survival genes and other evidence of endothelial cell accommodation (52). The hypothesis discussed during several sessions was that low dose exposure to antibody that was not injurious promoted subtle adaptation of the target cells to subsequent immune injury.

New translational Studies in Islet and Xenotransplantation

Enthusiasm for both islet transplantation and xenotransplantation has fluctuated in the past decade. While significant advances have been highly publicized, the clinical and even pre-clinical outcomes have not proceeded as quickly as some might have predicted. Both fields, however, have been attended by dedicated teams that have methodically worked through problems associated with these new disciplines. Several significant steps forward were reported at the ATC indicating that progress continues.

Diabetogenic immunosuppression has been an impediment to islet transplantation. Exciting data were presented by the Naji lab at the University of Pennsylvania indicating that long-term islet survival off immunosuppression could be achieved in NHPs with a combination of rabbit anti-thymocyte globulin, the CD20-specific monoclonal antibody rituximab and a 200 day course of sirolimus (53). Similarly, a collaborative group, including Herring from Minnesota and Kenyon and Ricordi from Miami, reported this year that a combination of FTY720, RAD and basiliximab allowed for indefinite islet survival in NHPs (54). Both of these regimens appear clinically applicable and are likely to serve as reasonable regimens for new islet trials. In a striking report by Herring, it was shown that islet xenografts could be maintained indefinitely with this regimen when the CD154-specifc antibody ABI793 was added (55). Although it is unlikely that trials with CD154-specific therapies will advance until their thrombotic risk is better understood, this is nevertheless a remarkable achievement in its simplicity and potential impact.

Finally, real progress in the field of vascularized xenotransplantation was reported by the Transplantation Biology Research Center of the Massachusetts General Hospital (56–58). Survival of hearts and kidneys derived from Gal–α-1,3-Gal knockout animals was shown to exceed 80 days using an aggressive immunosuppressive regimen. Although clearly not clinically applicable at this point, the consistency with which hyperacute rejection was avoided was admirable and promises to open the way to quickened progress in the coming years.