T regulatory cell suppression of colitis: the role of TGF‐β

Short abstract

Transforming growth factor β (TGF‐β) and interleukin 2 may be involved in IBD peripheral regulatory T cell pathophysiology, raising the possibility of therapeutic application of TGF‐β induced regulatory T cells in IBD patients

CD4+CD25+ regulatory T cells (T_reg) expressing the lineage marker Foxp3 control immune responses to self‐ and foreign antigens and are an intensely studied member of the heterogenous group of regulatory T cells. Although initially described as a population of suppressor T cells required to avoid organ specific autoimmunity, it has subsequently become clear that T_reg control immune responses in a much broader sense, including transplantation tolerance and immune responses to pathogens and tumours.¹,²,³ Relevant to inflammatory bowel diseases (IBD), T_reg prevent⁴ and treat established⁵ colitis in animal models of IBD and are numerically deficient in patients with active IBD.⁶ This underlines the fact that the immune dys‐equilibrium characteristic of chronic inflammatory diseases involves concomitant disturbances in inflammatory and suppressive immune mechanisms and opens up novel approaches for IBD therapy by strengthening T_reg mediated suppression.

Currently, efforts in laboratories worldwide are addressing the central questions of T_reg immunology, resolution of which will help us see more clearly the role of T_reg in disease pathogenesis and therapy. Some of these central questions are: where do T_reg come from; how are they generated, expanded, and maintained; how and where do they function; why do they fail to control immune responses in disease; and how can their therapeutic potential be used most efficiently and safely. As current data indicate that T_reg from IBD patients are functionally normal⁶,⁷,⁸ but numerically deficient during active disease,⁶ it will be interesting to understand the underlying mechanisms and, based on that knowledge, devise the most promising strategy to enlarge their numbers to therapeutic levels.

T_reg originating in the thymus

It was first described that during ontogeny, CD4+CD25+ T_reg originate in the thymus, become detectable in the periphery of normal mice from around day 4 after birth, and increase to adult numbers by week 3.⁹ Production of Foxp3 expressing T_reg by the thymus is considerably delayed relative to non‐regulatory thymocytes,¹⁰ and neonatally thymectomised mice show an early and substantial reduction in peripheral CD4+CD25+ T_reg, associated with the development of autoimmune diseases. It was demonstrated in transforming growth factor β1 (TGFβ1)¹¹ and interleukin 2 (IL‐2)¹² deficient mice, that both cytokines, in contrast with their important function on mature peripheral T_reg, are dispensable for intrathymic T_reg development. Efficient generation of CD4+CD25+ thymocytes results from relatively high affinity interaction of the T cell receptor (TCR) with agonist ligands expressed in thymic epithelial cells.¹³,¹⁴ This process requires CD28 dependent costimulation of developing thymocytes¹⁵ and thymic stromal lymphopoietin expressed by human Hassall's corpuscles to activate thymic CD11c+ dendritic cells to express high levels of CD80 and CD86.¹⁶ Whether a thymic defect contributes to T_reg pathophysiology in IBD is not known. Interestingly, however, thymic T_reg are functionally impaired in patients with autoimmune myasthenia gravis,¹⁷ and it was shown in animal models that colitis induces aberrant thymic development with impaired T_reg cell production.¹⁸

T_reg originating in the periphery

An important next step in our understanding of T_reg biology was when it was shown that Foxp3+ T cells with regulatory function in vivo do not only originate intrathymically but can also be generated from peripheral CD4+CD25− cells. Apostolou and von Boehmer showed that prolonged subcutaneous infusion of low doses of peptide by means of osmotic pumps transforms mature T cells into CD4+25+ suppressor cells that can persist for long periods of time in the absence of antigen and confer specific immunological tolerance after challenge with antigen.¹⁹ Conversion of peripheral CD25− T cells into T_reg in mice requires the presence of TGF‐β and downregulation of the TGF‐β signalling inhibitor Smad7²⁰ in addition to TCR and IL‐2R mediated signals.²¹,²² In contrast, studies by Walker and colleagues²³ suggest that human FOXP3+ T cells with regulatory function in vitro can be generated from peripheral CD4+CD25− T cells via CD3/CD28 signalling in the absence of TGF‐β.

In the present issue of Gut, Fantini and colleagues²⁴ report that T_reg generated in vitro from CD25− T cells under the influence of TGF‐β are functionally active and suppress Th1 mediated colitis induced in SCID mice after transfer of CD4+CD62L+ T cells (see page 671). Furthermore, the authors provide evidence that such TGF‐β induced regulatory T (Ti‐T_reg) cells require exogenous IL‐2 for their in vivo expansion and long term maintenance. Ultimately, these data raise the question of the potential involvement of TGF‐β and IL‐2 in IBD peripheral T_reg pathophysiology and the potential therapeutic application of Ti‐T_reg in IBD patients.

The picture that emerges from the present studies is that TGF‐β and IL‐2 have complementary but partly opposing functions in the biology of peripheral T_reg. Interestingly, in a system in which conversion of naive CD4+CD25− T cells into T_reg was achieved by minute antigen doses with suboptimal dendritic cell activation, both addition of TGF‐β and absence of IL‐2 production enhanced conversion but reduced proliferation.²⁵ A negative effect of IL‐2 mediated expansion on the conversion rate was also supported by the observation that reduced proliferation in the absence of autocrine IL‐2 production was associated with an increased conversion rate of the least divided cells.²⁵ On a similar line, elegant studies using mice containing the Foxp3GFP knock‐in allele that were genetically deficient in either IL‐2 (IL‐2−/−) or CD25 (Ilra−/−), demonstrated that IL‐2 signalling seems to be critically required for promoting the survival and population expansion of T_reg in vivo but is dispensable for induction of Foxp3 expression in developing thymocytes and for peripheral T_reg suppressive function.²⁶ The latter finding notwithstanding, it was shown with T_reg derived from wild‐type mice that IL‐2 uptake by T_reg is required for their suppressive activity in vitro.²⁷,²⁸ In the face of increased IL‐2 levels in active IBD lesions and no known defects in IL‐2 signalling in IBD, it seems unlikely that the numerical T_reg deficit in IBD is due to a lack of IL‐2 mediated T_reg expansion. Furthermore, the findings of Fantini and colleagues²⁴ that Ti‐T_reg were maintained under the influence of IL‐2 when transferred into colitic mice support the fact that natural T_reg expanded for therapeutic purposes and Ti‐T_reg should be maintained after transfer in IBD patients. The reason for the discrepant behaviour of natural T_reg, which expand in lymphopenic hosts, and Ti‐T_reg, which are not maintained when transferred into SCID mice under both syngenic and MHC mismatched conditions, remains to be clarified. It is similarly unknown whether the numerical deficit observed in IBD is a consequence of insufficient conversion of CD4+CD25− T cells into CD4−CD25+ T_reg or whether this is caused by insufficient TGF‐β levels or signalling. The former does not seem very likely as colitis is associated with high TGF‐β levels, whereas the latter might be possible as Smad 7, an inhibitor of TGF‐β signalling, is expressed at high levels in IBD T cells.²⁹

T_reg function

Once activated, peripheral Foxp3+ T_reg are robust suppressors and mediate inhibition of CD4+CD25− responder T cells by an as yet undefined mechanism. Current studies strongly indicate a critical role for TGF‐β in this process. However, conflicting results in the literature make the exact nature of the involvement and cellular source of TGF‐β a matter of current debate.³⁰,³¹,³²,³³,³⁴ The important role of TGF‐β in T_reg suppression was again evident from recent studies which demonstrated that T cells with a dominant negative TGF‐β receptor which cannot respond to TGF‐β escape control by T_reg in vivo.³³,³⁵,³⁶ Thus high levels of Smad7, an inhibitor of TGF‐β signalling, in intestinal IBD T cells²⁹ may not only decrease conversion of CD25− T cells into CD25+ T_reg, as discussed above, but could also be responsible for the fact that CD25− T cells escape TGF‐β mediated control. This however does not seem to be critical as peripheral T_reg from IBD patients⁶,⁸ and intestinal T_reg from active IBD patients were functionally normal in their inhibition of CD25− T cells.⁷

T_reg therapy

Thymic generation of natural T_reg seems constitutive and relatively stable, with low levels of approximately 1–2% of human T_reg in peripheral blood lymphocytes and even lower levels in the intestine.⁶ The potential for their therapeutic manipulation greatly increased when it was shown that human CD4+CD25+ T_reg can be expanded in vitro in the presence of high dose IL‐2 and allogeneic feeder cells without loss of function.³⁷,³⁸ Additional studies showed that their large scale in vitro expansion for clinical purposes can be achieved via repeated stimulation with anti‐CD3, anti‐CD28, and high dose IL‐2.³⁹ Importantly, these expanded T_reg retain their phenotypic characteristics, maintain expression of lymph node homing receptors, and show an even increased suppressive activity compared with freshly isolated cells. Although this indicates that expansion of CD4+CD25+ T_reg has definite potential application in designing future clinical therapy for autoimmune diseases, inflammation, and transplantation, their physiological scarcity and numerical deficiency in IBD and/or their functional deficit in other human diseases support alternative approaches to convert normal naive CD4+CD25− T cells into CD4+CD25+ FOXP3 regulatory T cells. Here, several alternative approaches may be an option and are likely to compete depending on further information and specifics of T_reg pathophysiology in different diseases. These approaches include for example protocols which use TGF‐β together with IL‐2 and TCR mediated signals or nuclear factor κB blocked immature APC⁴⁰ for the ex vivo generation of T_reg, monoclonal antibodies,⁴¹ or copolymer I,⁴² or infusion of low dose antigen¹⁹ for their in vivo generation and further strategies to strengthen T_reg mediated suppression locally.⁴³

In summary, it is quite well established that TGF‐β effectively promotes the generation of Ti‐ T_reg and has an essential role in T_reg function. Current data indicate that a colitis induced defect in the generation of thymic T_reg or a defect in TGF‐β signalling involving high expression of Smad7 may play a role in IBD, but further studies are required to determine the molecular mechanisms of T_reg pathophysiology in IBD, which so far has been characterised as a numerical deficit in active disease. Treatment of IBD by strengthening T_reg mediated suppression has become a feasible option. Further studies will help to choose the best and least cumbersome approach towards achieving T_reg levels that are therapeutic and readjust over time.