Thursday 18 November 2010
Regulatory T cells in health and disease; Putting the pieces together
Promotor: Prof.dr A.B.J. Prakken and prof.dr P.J. Coffer
Defence: 18 November 2010
In this thesis, we investigated human regulatory T cells as targets for therapeutic intervention in autoimmune diseases. We have extended findings from murine to human Treg, from protein modifications to in vitro Treg function, but also from ‘in vitro’ to ‘in vivo’ Treg properties. Our aim has been to establish a clear overview of the human Treg phenotype, mechanism of suppression, and suppressive function in vivo. Also, we evaluated the clinical significance of Treg in peripheral blood and muscle tissue of Juvenile dermatomyositis patients. Moreover, we investigated whether self-antigen stimulation or post-translational modification of FOXP3 increases human Treg frequency and function.
We have clearly demonstrated that that there are important differences between human and murine Treg function. In Chapter 2 we showed that, in contrast to murine Treg, human Treg mediated suppression of effector T cell responses is not mediated by induction of apoptosis due to cytokine consumption. Both peripheral blood Treg, and Treg from synovial fluid of JIA patients suppressed effector T cell proliferation and cytokine production, without inducing efffector T cell apoptosis .
In Chapter 3, we explored the differences between in vitro expanded Treg (nTreg) and induced Treg (iTreg) in both in vitro assays and to evaluate the in vitro results, we also analyzed Treg function in an in vivo model for xenogeneic Graft versus Host Disease (x-GvHD). In vitro assays demonstrated that both populations of Treg exhibited suppressive capacity. However, in vivo, only the expanded nTreg suppressed x-GvHD, while iTreg lost FOXP3 expression and, thereby, suppressive function shortly after being administered in vivo. In conclusion, in vitro suppression assays, which have so far been considered the golden standard for determining Treg suppressive function, are not sufficient to predict Treg suppressive function in vivo. Therefore, human Treg suppressive capacity should be analyzed in vivo in humanized mouse models.
To increase Treg numbers, we have explored the possibility of using HSP60; a self protein recognized by T cells, highly expressed at inflammatory sites (5, 6), and able to induce tolerogenic responses in murine models (7, 8). We showed that HSP60 (both complete protein and selected epitopes) induced Treg from human peripheral blood CD4+CD25- T cells. Furthermore, we observed that CD30 expression on HSP60 induced Treg distinguished functional Treg from non-suppressive cells (Chapters 4 and 7). Thus, HSP60 has the potential to increase Treg frequency by Treg induction, and is therefore a suitable candidate for Treg modulation in vivo.
While we have shown induction of functional Treg with HSP60, it is also important to establish how FOXP3 expression is regulated, since loss of FOXP3 causes a loss of Treg suppressive capacity (Chapter 3). We show in Chapter 5 that FOXP3 expression is regulated by protein acetylation. FOXP3 was acetylated by the HAT p300 and de-acetylated by the HDAC SIRT1. Acetylation of FOXP3 promoted protein stability by preventing proteasomal degradation. Accordingly, treatment of anti-CD3/ anti-CD28 induced human Treg with HDAC inhibitor nicotinamide (NAM), promoting acetylation, caused an increase of FOXP3 expression. Both the frequency of FOXP3+ Treg, as well as expression of FOXP3 per cell was increased. Furthermore, we demonstrated that murine iTreg treated with NAM exhibited increased suppressive function in in vitro assays. Thus, HDAC inhibitors promote stable FOXP3 expression in induced Treg and increase suppressive function.
Before investigating Treg targeting as a treatment for a disease, it is important to establish whether Treg are involved in the pathogenesis of an inflammatory disorder. In Chapter 6 we showed that patients with Juvenile dermatomyositis (JDM) had normal Treg percentages in peripheral blood compared to age-matched controls, independent of disease activity. Furthermore, Treg from JDM patients with remitting disease exhibited suppressive function in vitro, while Treg from patients with active disease were not consistently suppressive. In muscle tissue of JDM patients with active disease many FOXP3+ Treg were present. Thus, Treg were able to migrate towards the inflamed tissue, but somehow were unable to suppress inflammation. In conclusion, Treg may be involved in JDM pathogenesis.