Regulatory T cell
Although CD4+CD25+ regulatory T cells (Treg) represent a well-characterized population of T cells with in vitro and in vivo suppressive capacity, the basic mechanisms of suppression are still not understood. The constitutive expression of the high-affinity receptor for IL-2 has raised the question about the role of IL-2 in Treg function. Here, we review recent data indicating that IL-2 is not only necessary for the homeostasis of Treg but is also critical for the activation of Treg function. Since Treg do not produce IL-2 by themselves, their capacity to utilize IL-2 secreted by other T cells appears to be an essential component of Treg biology. This indicates that Treg suppressive activity is controlled by interaction with activated target cells via the soluble mediator IL-2. In Treg, IL-2 has been identified as a potent inducer of the immunosuppressive cytokine IL-10, an important mediator of Treg suppression in vivo. The efficient capture of IL-2 by Treg may, under conditions of limited IL-2 supply, cause IL-2 deprivation of responder T cells. This competition can explain some of the currently discussed discrepancies between in vivo and in vitro activity of Treg.
CD4+CD25+ regulatory T cells (Treg) play an important role in the maintenance of tolerance to self as well as in the negative control of immune reactions against non-self antigens (reviewed in 1, 2). Although CD4+CD25+ Treg have attracted very considerable interest during the last few years, it is still not clear how Treg actually suppress or regulate immune responses. The suppressive mechanisms acting in vitro and in vivo are controversial. It has been claimed that suppression is contact-dependent, because soluble factors like IL-10 or TGF-β do not seem to play a role in vitro and suppression is abrogated by physical separation of Treg and responder T cells 3–5. Also, Treg isolated from secondary lymphoid organs of normal mice do not produce IL-10 or TGF-β upon re-stimulation. In contrast, there is evidence that IL-10 and TGF-β may be relevant mediators of suppression in vivo6–11 and Treg isolated from sites of active immune reactions are able to produce high amounts of IL-10 upon re-stimulation 7, 11, 12. Recent data indicate that IL-2 is critical for the activation of Treg effector function, such as production of IL-10. Since Treg do not produce IL-2 they seem to depend on their capacity to efficiently compete for IL-2 with their target T cells – activated dendritic cells or conventional T cells 5, 13–17.
This review focuses on the role of IL-2 in the regulation of Treg functional activity and highlights the importance of IL-2-mediated intercellular communication between Treg and their target cells. A model of mutual interaction between Treg and target cells via competition for IL-2 not only explains how Treg activity is controlled in vivo but also helps to clarify current discrepancies between the results of in vitro and in vivo studies of Treg function.
IL-2 is essential for Treg homeostasis
Constitutive expression of CD25, the α-chain of the IL-2 receptor, initially allowed the identification of Treg. For a long time, CD25 was merely regarded as a surrogate Treg marker without functional relevance. Only recently has it become clear that IL-2 plays an essential role for Treg development and homeostasis (reviewed in 18). IL-2 was originally identified as an important T cell growth factor in vitro 19. Paradoxically, strong T cell hyperproliferation and autoimmune diseases are observed in mice deficient for IL-2 or components of its receptor 20–23. This unexpected pathology can be explained by a reduced number of Treg or their functional impairment 24–29, since transfer of wild-type Treg into neonatal IL-2R-deficient mice prevents disease development 24, 26, 28, 30. IL-2 mainly plays a role for the peripheral expansion and maintenance of Treg 1, 18, 31 rather than their thymic generation since Treg develop in IL-2-deficient mice 16, 28.
Treg compete for IL-2 with responder T cells
The IL-2 that Treg need for survival has to be provided by other cells since Treg are not able to produce IL-2 themselves. The constitutive expression of all three chains of the high-affinity IL-2 receptor enables Treg to take up IL-2 without previous activation 15, 26. Using the CD25 expression level, which is positively regulated by IL-2 32, 33, as an indicator for IL-2 uptake, it has been demonstrated in vitro and in vivo that competition for IL-2 indeed occurs 12, 15, 17. During co-culture of Treg with responder T cells, Treg up-regulate CD25, whereas induction of CD25 expression on responder T cells is suppressed. Up-regulation of CD25 on Treg is inhibited by addition of anti-IL-2, whereas addition of IL-2 restores CD25 expression on responder T cells, demonstrating that the differential expression of CD25 is regulated by IL-2 15. Thus, Treg competition for IL-2 is boosted by the positive-feedback loop of IL-2 uptake and CD25 up-regulation. The same inverse regulation of CD25 expression on Treg versus responder T cells has been observed in vivo 12, 17 upon adoptive co-transfer of Treg and responder T cells.
Is IL-2 only required for induction of suppressive activity or is IL-2 consumption a suppressive mechanism itself?
Recent in vitro studies show that IL-2 enhances Treg suppressive activity 5, 13, 15, 16. Furthermore, the blockade of Treg uptake of IL-2 during in vitro co-culture with responder T cells abrogates suppression 14, 15 showing that IL-2 uptake by Treg is essential for suppression. Interestingly, whereas Treg-mediated suppression of IL-2 protein production is abolished by addition of exogenous IL-2, Treg-mediated reduction of IL-2 mRNA is not.
These results suggest that Treg control IL-2 production at the mRNA and at the protein level by different mechanisms. (1) IL-2 protein production by responder T cells depends on a positive IL-2 feedback loop that can be interrupted by Treg via competition for IL-2; since IL-2 is essential for in vitro T cell proliferation, the net effect of IL-2 competition is decreased proliferation, which is the standard read-out of in vitro inhibition assays. (2) Transcription of the IL-2 gene is regulated by Treg via another mechanism that only requires induction by IL-2 5, 14. In an in vitro proliferation assay, the rather late effect of this inducible activity may be difficult to detect. In proliferation assays, the effect of IL-2 competition is dominant because it is evident immediately 15. In vivo, the IL-2-induced suppressive activity may be more relevant since Treg control immune responses that are not dependent on IL-2 and that are therefore insensitive to competition. For the identification of these IL-2-dependent suppressive factor(s) the IL-2 mRNA level instead of proliferation should be used as a read-out of in vitro suppression.
Does IL-2 activate in vivo suppression and what is the suppressive factor?
In a series of papers dating back to 1996, the effect of in vivo administration of an agonistic IL-2–IgG2b fusion protein on adaptive immunity was analyzed 34–38. Interestingly, IL-2–IgG2b administration suppressed cellular and humoral responses 34, 36, 38 as well as an experimental model of colitis 37 and organ transplantation 35. The suppression was not due to depletion of IL-2-binding cells such as activated T cells; rather, the frequency of CD25+ T cells was increased. Furthermore, in one study in-vivo IL-10 levels were also increased and anti-IL-10 antibodies abrogated the protective effect of IL-2–IgG2b administration on experimental colitis 37.
Although in these studies the effect of IL-2 on Treg was not analyzed directly, it is tempting to speculate that Treg were indeed the main targets. In line with these indirect data, de la Rosa et al. 15 and others 17 showed that IL-2 is critical for priming Treg for IL-10 production. Preliminary data from our laboratory also indicate that IL-2 is required in vivo to prime Treg for IL-10 production (A. Scheffold, unpublished observations). However, IL-10 production by Treg is not essential in all disease models. It remains to be clarified whether IL-2 also primes Treg for production of additional suppressive factors apart from IL-10. With respect to the effector mechanism, it might be helpful to clarify which cells are the actual targets of Treg in vivo. Direct suppression of T cells has only been observed in in vitro experiments but has so far not been proven in vivo. In contrast, suppressive molecules like IL-10 mainly act on APC but not on T cells. Therefore one possible scenario of suppression could be via “education” of APC, as has recently been shown for bystander suppression in oral tolerance 39.
Soluble mediators versus cell–cell contact: proximity matters!
The role of soluble factors for Treg function has previously been questioned because Treg suppressive activity has been claimed to be contact-dependent, since suppression is abrogated in transwell cultures 4, 5. Does this observation really contradict the critical role of IL-2-based communication between responder T cells and Treg? In fact, sequestration of IL-2 by Treg will be most effective in close proximity to the responder T cells, since Treg constitutively express high-affinity IL-2 receptors, whereas responder T cells only up-regulate CD25 upon activation and in the presence of IL-2. Thus in a transwell culture, where Treg and responder T cells are separated by a distance of the order of 100-times the diameter of a lymphocyte, interference of Treg with the autocrine IL-2 feedback loop is prevented. It has been predicted that 1–2 cell diameters is an effective distance for competition in a mathematical model that integrates topological features, such as the distance between source and target cells, as well as biophysical parameters, including the mediator's secretion rate, the numbers and turnover rates of receptors, and autoregulatory feedback loops (40 and D. Busse et al., manuscript in preparation).
Similarly, under physiologic conditions most cytokines also act locally: directed secretion towards the target cell, the short serum half-life of cytokines and binding to components of the extracellular matrix 41 suggest the spatial restriction of their activity.
Therefore, considering the obvious discrepancies between in vitro and in vivo systems the available data on suppression mechanisms should be interpreted with care, as discussed above. We propose here that Treg require close proximity to their target cell but so far no direct evidence for a contact-mediated suppression mechanism is available.
Regulation of Treg function by the activity of their target cells
The identification of IL-2 as an essential “on-switch” for Treg suppressive activity has two important implications. (1) TCR triggering alone does not induce full suppressive activity in Treg. This is of fundamental relevance since most Treg probably recognize autoantigens with high affinity 42–45 and are thus constantly activated by antigen in vivo 46, which would result in permanent systemic suppression. (2) Treg activation depends on a second neighbouring cell providing IL-2. Under physiological conditions IL-2 is mainly produced by activated T cells and dendritic cells, which both represent potential Treg targets. Thus, the target cells might control Treg activity.
Via the level of secreted IL-2, Treg can adapt their suppressive mechanism to the state of immune activation (Fig. 1). (1) In non-immunized mice, Treg constantly take up low levels of IL-2, which is required to maintain CD25 expression and survival of Treg 1, 31. This steady IL-2 supply may result from low-affinity TCR stimulation of responder T cells, as required for T cell homeostasis. Whether Treg consumption of IL-2 supports maintenance of peripheral tolerance, e.g. by increasing the T cell activation threshold of autoreactive T cells, is not known. However, this low IL-2 supply is not sufficient to induce the capacity of Treg to activate effector function, e.g. IL-10 production. (2) At sites of immune reaction, high levels of IL-2 are generated locally and enable efficient activation of responder T cells, and at the same time prime colocalized Treg e.g. for IL-10 production. Priming for IL-10 also requires TCR stimulation of Treg which is assured by the continuous presence of autoantigens on all APC. The temporal delay between responder T cell activation and induction of suppressive activity of Treg prevents interference with T cell priming but would allow control of the T cell effector phase within inflamed tissue. Indeed, to acquire the capacity to produce IL-10, Treg had to be activated in vitrovia their TCR in the presence of IL-2 for at least 2–3 days 15. Furthermore, Treg isolated from sites of ongoing immune responses in vivo are able to produce high amounts of IL-10 upon re-stimulation ex vivo7, 11, 12, 47.
Finally, when the immune response is terminated, persistence of Treg expressing IL-10 would result in systemic immune suppression. However, there is evidence that the activated Treg state is restricted only to the short time-window of active immunity. Effector/memory-type Treg, as identified via expression of αEβ748, also depend on IL-2 for IL-10 production, and in vitro IL-2-primed Treg lose the capacity to produce IL-10 when cultured without IL-2 for a few days (A. Scheffold and J. Huehn, unpublished data). Thus, the ability of Treg to suppress seems to be switched on by IL-2 only transiently and remains dependent on external IL-2. This contrasts with cytokine memory in conventional Th2 or Tr1 cells, which requires cytokines like IL-4 or other signals but not IL-2 for priming and eventually becomes independent of the initial differentiation signal 49.
Treg compete for IL-2 with their target cells and suppression by Treg requires their activation by IL-2 signals – it takes (IL-)two to tango. Such a model of mutual interaction between Treg and target cells naturally explains how Treg activity is adapted to the state of the immune reaction.
Considering the differential role of IL-2 for T cell proliferation in vitro and in vivo, the capacity of Treg to compete for IL-2 can also explain some of the current discrepancies between in vivo and in vitro suppression by Treg. This implies that the relevant parameters for determining Treg function in vitro have to be carefully re-evaluated especially in the human system where Treg can only be analyzed in vitro. It also remains to be clarified what additional suppressive mediators are induced by IL-2 and what are the actual target cells of Treg in vivo. However, the pivotal role of IL-2 for the suppressive function of Treg offers a promising strategy for the targeted activation of Treg for immunosuppressive therapeutic intervention.
We thank Andreas Radbruch, Sascha Rutz and Takeshi Takahashi for critical reading of the manuscript as well as Dorothea Busse, Maurus de la Rosa, Susan Brandenburg and Heike Dorninger for provision of unpublished results. The figure was designed by Luise Fehlig. This work was supported by DFG-SFB 618, EC-grant QLK3-CT2002-02026 and the German National Genome Research Network, grant 01 GS 0413.