Answers are grounded in this paper and cite their sources.
In 1986, Mosmann and Coffman observed that individual CD4+ T helper(Th*) cell clones express phenotypically distinct cytokine profiles, thereby giving rise to a unifying concept “the Th1/Th2 paradigm” (Fig. 1)(1). Central to this paradigm is the role of certain cytokines which act as cross-regulators of Th1/Th2 cell function. Interferon-γ (IFNγ), a Th1 product, inhibits expression of the Th2 program, whereas interleukin (IL) 4 and IL-10, Th2 products, can act to block the Th1 program. Moreover, polarized cytokine programs (i.e., Th1+ and Th2- or Th1- and Th2+) have been detected in vivo, most notably in murineLeishmania major infections (2). In appreciation of these facts, an intensive effort has been launched to reach an understanding of the CD4+ T-cell programs associated with allograft rejection and tolerance. Indeed, new therapeutic strategies based on the concept of manipulating cytokine networks have been espoused as a means to guide the allograft response toward a tolerant state (Fig. 1). These strategies include: (a) inhibition of production/effects of Th1 cytokines, and (2) enhancement of systemic or local levels of Th2 cytokines. Recently it has become apparent that the Th1/Th2 model, while providing a valuable framework for probing various models, has undoubtedly oversimplified the complexity of cytokine responses by individual T cells in vivo(3, 4). Indeed, as suggested by Mosmann and Kelso, these two patterns of cytokine expression likely represent extremes of many possible outcomes. The aim of this Overview is to consider the cytokine programs associated with allograft rejection and tolerance in the light of recent studies that highlight two important phenomena: first, the redundancy of cytokines able to act as T-cell growth factors; second, the pleiotropic nature of individual cytokines. It is with the characteristics of redundancy and pleiotropism in mind that we consider whether the manipulation of specific cytokine pathways provides a robust opportunity for inducing transplantation tolerance in the clinic. PROINFLAMMATORY CYTOKINE PROGRAMS Consistently, intragraft IL-2, IFNγ, and the cytotoxic T lymphocyte(CTL)-specific marker granzyme B, proteins and/or transcripts have been detected in rejecting allografts, whereas IL-4 expression has not been as consistently detected (5-8). Moreover, the sequential analysis of biopsies of either rejecting experimental allografts(5-7) or human renal allografts (M.J. Dallman, Imperial College of Science, Technology and Medicine, London, UK, personal communication, 1996) usually detects IL-2 and IFNγ gene expression preceding graft dysfunction. The precise roles of IFNγ and IL-2 in graft rejection are uncertain. However, several CD4+ T-cell-dependent effector mechanisms (i.e., delayed-type hypersensitivity and CTL activity) participate in acute allograft rejection (Fig. 2a). In this context, IL-2 is believed to stimulate both the proliferation of T cells and activation of cytotoxic function of CTLs, and IFNγ is thought to recruit macrophages into the graft, cause macrophage activation, enhance CTL activation, and up-regulate MHC expression by the graft. Clearly, IL-2 expression is strongly associated with the rejection process(5-7). Is allograft rejection an IL-2-dependent process? IL-2, when administered as an adjunct to various tolerizing regimens, precludes long-term engraftment (9-11 and H. Waldmann, Oxford, UK, personal communication, 1996). Does this mean that the absence of IL-2 gene expression, ipso facto, causes tolerance? Several studies have, indeed, demonstrated that application of tolerance-inducing therapies often skews the allograft response from a Th1 pattern usually noted during unmodified rejection to a Th2 pattern (6, 11-13). The possibility that tolerance might directly result from an allograft response conducted in the absence of IL-2 expression and in the presence of IL-4 expression (11, 13-15) has proven irresistible. To test this hypothesis, we transplanted allogeneic islets (DBA2/J, H2d) into mice whose IL-2 gene had been functionally silenced by targeted gene disruption (IL-2 knockout mice, H2b[16]). Interestingly, all IL-2 knockout mice reject their allografts at a tempo that is only modestly delayed in comparison to mice that have inherited two copies of the intact IL-2 gene. Immunohistologic analysis of the allograft response in IL-2 knockout graft recipients reveals CD4+ and CD8+ T-cell infiltrates in rejecting grafts. The intragraft pattern of gene expression in IL-2 knockout recipient mice demonstrated the presence of transcripts for IL-4, IFNγ, and the CTL-specific marker granzyme B, but not IL-2 (17). Thus, in this model the absence of IL-2 gene expression, even in the presence of IL-4 gene expression, does not inevitably lead to graft tolerance. Insofar as the immune system in the IL-2 knockout mouse matures in the absence of IL-2, IL-2 knockout mice could bear an expanded population of T cells that are reliant on T-cell growth factors other than IL-2, which do not normally participate in rejection. However, Minasi et al. (18), studying a normal murine immune system, which matured in the presence of IL-2, demonstrated that vigorous in vitro T-cell proliferation can occur in the absence of IL-2. Hence, IL-2-independent T-cell proliferation may not be an aberrant phenomenon peculiar to the IL-2 knockout mouse. Which growth factors might support IL-2-independent T-cell proliferation? IL-4, IL-7, IL-9, and IL-15 are potent T-cell growth factors(19, 20). IL-7-triggered signals, not blocked by cyclosporine, are important for lymphocyte differentiation and stimulation of CTL proliferation (21). IL-9, which is produced by T cells, albeit later than IL-2, powerfully supports T-cell stimulation. IL-15 is produced by a wide variety of cell types, with the notable exception of mature T cells (20). Interestingly, IL-15 binds to a trimolecular receptor complex that includes two members of the IL-2 receptor complex (i.e., the IL-2R β and common gamma [γc] chains, as well as the IL-15-selective IL-15Rα chain). IL-15 supports the proliferation of CD4+ and CD8+ T cells; IL-15-driven proliferation is blocked by anti-IL-2Rβ, but not by anti-IL-2Rα mAb. Further, IL-15 supports CTL activity, albeit suboptimally in the absence of IL-2 (reviewed in[20]). In our studies of the allograft response elicited in IL-2 or IL-2+IL-4 double knockout mice and studies of human renal transplant biopsies, amplified IL-7 and IL-15 gene expression is evident in all sampled tissues undergoing rejection. Clearly, allograft rejection elicits the expression of multiple T-cell growth factor programs. In addition to attempts to neutralize or prevent expression of IL-2, one could target proinflammatory cytokines or chemokines that are expressed during allograft rejection (i.e., IFNγ, tumor necrosis factor-α or RANTES[22]). In theory, this approach could be effected through blockade of gene expression with antisense oligonucleotides, neutralization with antibodies, or soluble cytokine receptor fusion proteins (reviewed in[23]). Although appealing, one must ask, as in the case of T-cell growth factors, whether the plethora of proinflammatory cytokines exerting overlapping function will nullify this approach. Initial preclinical studies utilizing this approach highlight inherent difficulties and complexities. IFNγ knockout hosts reject heterotopic heart transplants with a time course and histologic pattern similar to those of the wild-type host, and expression of other proinflammatory cytokines noted during the allograft response in IFNγ knockout mice is unchanged (i.e., tumor necrosis factor-α) (24). Similarly, islet allograft rejection in IFNγ receptor knockout hosts is brisk and T-cell dependent(25). While more studies are required, these data indicate that individual proinflammatory cytokines are not indispensable for effecting allograft rejection. We are pessimistic that targeting any one T-cell cytokine pathway will prevent clinical graft rejection, as indicated through analysis of cytokine gene deficient mice (vida supra). Moreover, given the redundant, overlapping nature of cytokine functions, it may prove difficult to block the production or neutralization of the myriad of cytokines that are apparently involved in supporting alloantigen-specific T-cell activation, proliferation, and effector functions. However, it may prove feasible to inhibit intracellular signaling pathways that are activated by several cytokine receptors. Members of a given cytokine family often utilize common receptor subunits (reviewed in[26]). For example, the IL-2 receptor complex comprises an α, β, γc heterotrimer. IL-15 and IL-2 both bind to the IL-2R β and γc chains and utilize these chains for signal transduction (27). Moreover, other cytokines, which act as T-cell growth factors (i.e., IL-4, IL-7, and IL-9), also utilize theγc chain of the IL-2R, in addition to their own, cytokine-specificα receptor chains, as receptor components(28-30). The discovery that X-linked severe combined immunodeficiency is caused by a defect in the γc chain proves that the lack of signaling through this pathway affects multiple cytokines and leads to far more profound immunodeficiency than noted in IL-2 knockout mice(16, 31). Thus, γc chains are critical for signal transduction for a number of cytokine T-cell growth factors. Hence, the selective destruction of T-cell-activation-induced growth factor receptor, i.e., IL-2Rα, bearing T cells with cytolytic anti-IL-2Rα mAb, or blocking their ability to bind growth factors with anti-γc mAb, or blocking the common pathway of intracellular growth factor signaling with drugs such as rapamycin, may well lead to graft prolongation in the clinic(32, 33). There are rumors of promising early clinical results achieved through the use of humanized or chimeric anti-IL-2Rα mAbs as induction therapy. Although this approach may well prove useful, it is not likely, given results in IL-2 knockout mice, that the therapeutic activity is completely due to IL-2R blockade. These antibodies bear antibody-dependent cellular cytotoxicity activating/complement-activating residues and are able to lyse activated T cells. As a consequence, we will not consider the use of these agents in the context of manipulating cytokine pathways, because they likely destroy activated lymphocytes that are sensitive to multiple cytokines. CYTOKINE PROGRAMS AND IMMUNOREGULATION Active immunoregulatory phenomena are central to the creation and/or maintenance of peripheral allograft tolerance. Many laboratories, perhaps most elegantly in the demonstration of infectious tolerance by Qin et al.(34), have shown the presence of donor-specific suppressor cells through the use of passive transfer systems. One hypothesis suggests that tolerance is due to the ability of antigen-specific, T-cell clones to migrate to the target tissue and release immunoregulatory cytokines into the local microenvironment. Is there evidence to support this contention? Oral administration of antigen can induce a state of immunologic hyporesponsiveness (“oral tolerance”; reviewed in[35]). Weiner's laboratory determined that oral tolerance in the experimental autoimmune encephalomyelitis model is associated with the appearance of central nervous system homing, myelin basic protein-specific CD8+ T cells that express transforming growth factor-β (TGFβ)(36). TGFβ blocks IL-2/IL-4-stimulated T-cell proliferation (37). In the tolerized experimental autoimmune encephalomyelitis model, bulk CD4+ T cells, isolated from the mesenteric lymph node, produce IL-4, IL-10, and TGFβ, whereas the CD8+ T cells express only TGFβ (38). Our group has shown that an anergic CD8+ T cell islet-infiltrating clone, which releases an inhibitory factor (“IL-X”) able to block IL-2- and IL-4-dependent T-cell proliferation in a manner similar to TGFβ, abrogates diabetogenic autoimmunity in nonobese diabetic mice(39). This study tends to link anergy and immunosuppressive T-cell phenomena and, together with the aforementioned studies, would support the existence of cytokine-driven immunoregulatory networks. What evidence exists for a role of immunoregulatory cytokines in either the induction or the maintenance of allograft tolerance? In animal models, tolerizing therapies appear to inhibit the magnitude and/or alter the quality of CD4+ T-cell activation during the induction of tolerance. A switch from the expression of IL-2 and IFNγ, detected in the grafts of acutely rejecting hosts, to the expression of IL-4 in the grafts of hosts being treated with tolerizing therapies is often manifest(6, 11, 12, 34). In an extension of the Th1/Th2 paradigm, it has been suggested that induction of tolerance to an allograft is dependent upon immune deviation of the CD4+ T-cell program to a Th2 phenotype (i.e., IL-4, IL-10) and away from the Th1 (i.e., IL-2, IFNγ) profile associated with rejection (40, 41)(Fig. 2b). Interestingly, there is a suspicion that immune deviation to a Th2-like pattern, while permissive for long-term engraftment, may paradoxically support chronic rejection(42). Moreover, a recent study by Larson's group has raised the possibility that chronic vascular rejection in a heterotopic cardiac allograft model is associated with an ongoing immune response characterized by prolonged cytokine expression. In this model, long-term acceptance and avoidance of chronic rejection requires a silencing of both Th1 and Th2 cytokine programs (43). What role might IL-10 have in abrogating allograft rejection? Recent in vitro studies have determined that rIL-10 exerts inhibitory effects upon: IFNγ release, proliferation of Th1 cells stimulated by macrophages, MHC class II expression on monocytes, IL-12 release from macrophages, and the expression of B7 costimulatory ligands on the surface of macrophages(44-47). Thus, IL-10 is a potent blocker of Th1 responses stimulated by macrophages. Nonetheless, IL-10 may not exact the same inhibitory effects upon all antigen-presenting cells (APCs). In particular, IL-10 does not block the initiation of an alloimmune response by allogeneic dendritic cells and B cells (44). For example, IL-10 fails to block IL-2 production and T-cell proliferation during a primary mixed lymphocyte reaction, a system in which dendritic cells are the principal APC (44, 48, 49). Moreover, the effects of IL-10 are not uniformly immunosuppressive or anti-inflammatory: IL-10 augments IL-2-supported proliferation of CD8+ T cells and enhances IL-2-induced cytotoxicity by these cells (50). Finally, IL-10 stimulates monocyte FcγRI surface expression and antibody-dependent cellular cytotoxicity (51). Thus, depending on the nature of the APC and the local microenvironment, IL-10 may exert anti- or proinflammatory effects. Initial studies suggest that mammalian IL-10 may not readily block allograft rejection. In a transgenic mouse model in which the murine IL-10 gene was constitutively expressed by the pancreatic β cells, islets from these transgenic mice were transplanted into MHC-incompatible recipients. Despite documented IL-10 expression within the microenvironment of the graft, allograft rejection was not delayed (52). In order to determine the effects of systemic levels of IL-10 during an alloimmune response, we treated the recipients of allogeneic islets with a long-acting form of rIL-10. This Ig-linked, long-acting IL-10 has a biological half-life of 5 days, whereas rIL-10 has a half-life of minutes. Yet treatment with IL-10-Ig did not delay the onset of allograft rejection(53). Insofar as the predominant APC in the allograft response may be the dendritic cell and since IL-10 is unable to inhibit dendritic-cell-dependent activation of naive T cells, IL-10 may not block allograft rejection (44). Indeed, in our studies, IL-10 treatment resulted in an augmented CTL response to the graft. Interestingly, Qin et al. (54) have found that retroviral transfection of cardiac grafts with viral IL-10, not murine IL-10, leads to prolonged but not indefinite cardiac allograft survival, which suggests that the structural differences of viral IL-10 and murine IL-10 may determine whether the immunosuppressive versus immunostimulatory effects of IL-10 are exerted during the allograft response. Finally, these studies highlight the inherent difficulty of extrapolating the reductionist approach of in vitro studies(i.e., the immunoregulation by IL-10 on macrophage-based T-cell activation) to the complex dynamics of in vivo biology. In vivo veritas; in vitro maybe! Does IL-4 inhibit graft rejection and directly promote allograft tolerance? High in vitro levels of IL-4 skew Th0 cells toward a Th2 response and away from a Th1 response (Fig. 1). As stated previously, the expression of IL-4 during the period of administration of tolerizing therapies and the concomitant silencing of IL-2 gene expression are frequent markers for tolerance induction (6, 11, 12). While these events may aid tolerance induction, the expression of IL-4 in the absence of IL-2 is not sufficient for the induction of tolerance (vida supra)(17). Furthermore, IL-4 can elicit certain proinflammatory effects, including increases in: MHC class II expression on B cells, expression of B7 costimulatory ligands on B cells, chemoattraction of macrophages, vascular cell adhesion molecule-1 and intercellular adhesion molecule-1 expression on vascular endothelial cells, and CTL responses(55-63). Finally, the administration of a soluble IL-4 receptor to allograft recipients leads to graft prolongation, not hastened rejection (64). Thus, it is entirely unclear whether IL-4 expression during an alloimmune response will favor the acquisition of tolerance. What is the relative role of IL-4 in abrogating rejection? Is IL-4 in and of itself sufficient to block the allograft response? We treated islet allograft recipients with a biologically active, long-lived IL-4 molecule. Allograft rejection occurred, without delay, despite the maintenance of a serum concentration of IL-4 at >100 U/ml throughout the transplant period(X.X. Zheng, unpublished observations). Clearly, this pharmacologic dose of IL-4 was unable to abolish rejection, which may relate to dominant proinflammatory effects achieved at this serum level. A more physiologic approach is to ask whether IL-4-producing cells, able to localize to the same tissues as IL-2- and IFNγ-producing cells, are able to abrogate the proinflammatory response mediated by these “Th1”-like cells. Diane Mathis's group, in a murine adoptive transfer model of autoimmune diabetes, demonstrated that IL-4-producing T cells, even in 10-fold excess of IL-2- and IFNγ-producing T cells, were unable to preclude the development of diabetes (65). These studies suggest that: (a) the beneficial effects of IL-4 may depend upon the level of IL-4 in the microenvironment, (b) IL-4 by itself is insufficient, and/or (c) the importance of the detection of IL-4 during tolerance induction lies not in the expression of IL-4 but rather the absence of IL-2 and IFNγ expression. The corollary of the aforementioned is to ask whether IL-4 is an in rejection? To this we transplanted allogeneic islets into IL-4 gene knockout mice As IL-4 knockout recipients islet in the absence of IL-4 is T-cell as A course of fusion which blocks T-cell costimulatory signals, also engraftment in IL-4 gene knockout Moreover, we and have achieved similar results in IL-4 knockout mice a heart allograft either human or combined and murine as a tolerizing Interestingly, the administration of long-lived IL-4 and in the IL-4 knockout hosts leads to an inhibition rather than an enhancement of the Finally, et al. demonstrated that oral tolerance to could be achieved in IL-4 knockout In the absence of IL-4 does not preclude The aforementioned does not one to that IL-4 role in allograft tolerance. the of donor-specific to a state of tolerance may be IL-4 of in with precludes the of tolerance in mice and in a mouse heart transplant model Oxford, UK, personal communication, 1996). Moreover, the adoptive transfer of by and in a allograft model was shown to be dependent on IL-4 Thus, in IL-4 can aid tolerance IL-4 is not a for the induction of peripheral tolerance (vida supra). There is evidence that IL-4 is for the maintenance of expression of IL-4 may aid in the induction of tolerance through the of to the Th1 program. to of the alloimmune response, with IL-4 and IL-10, are in their and it is likely that we have only the surface in immunosuppressive cytokines. cytokines are the microenvironment in which they are may determine whether the effects of a cytokine are immunoregulatory or in the case of TGFβ there is as to whether it can be (reviewed in Furthermore, cytokines, in physiologic act not Thus, therapeutic that a cytokine may lead to outcomes. gene which the for the of a given cytokine or factor into the graft microenvironment, can and systemic effects. However, it is notable that in to has allograft tolerance been achieved blocking the production/effects of a proinflammatory cytokine IL-2, IFNγ, and/or the production or administration of a cytokine with immunoregulatory IL-4, IL-10, In all of allograft the common is to therapies that prevent T-cell activation during by naive T a that is cytokine Furthermore, has data to suggest that peripheral tolerance exists at multiple immune cytokine in is to be sufficient to induce or the complex of events that are to the and maintenance of CD4+ T-cell activation programs indicate indicate inhibitory (a) of T-cell cytokine programs associated with allograft rejection. (b) of T-cell cytokine programs associated with CTL requires a to into an effector
No takes yet. Share an insight, caveat, or question.
Nickerson et al. (1997) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: