Helios expression was restricted to the Foxp3+ population and was not detectable in CD4+CD25+Foxp3− T cells. We therefore assume that we expanded alloreactive nTreg cells in our aCD4+Rapa- or aCD4+TGF-β+RA-treated cultures, which stably kept their Helios expression. AZD8055 purchase Alternatively, addition of TGF-β may have induced Helios expression
as was shown by Neill et al. . Recently, it has been reported by several groups that Helios− within the Foxp3+ Treg cells are responsible for the release of proinflammatory cytokines such as IL-17 or IFN-γ whereas the Foxp3+Helios+ subset secreted almost no cytokines [60, 61]. This was also seen in our setting where over 70% of the aCD4-mAb+TGF-β+RA and aCD4-mAb+Rapa Treg cells were positive for Foxp3 and Helios (Fig. 3A) but secreted almost no proinflammatory cytokines (Fig. 2A). aCD4+TGF-β+RA Epigenetic Reader Domain inhibitor aTreg cells showed the highest co-expression of Helios, which was associated with reduced IFN-γ and almost no TNF-α expression. Interestingly, addition of Rapa but even more TGF-β+RA to anti-CD4-treated cultures could abrogate downregulation of Neuropilin-1 expression within Foxp3+ cells (Fig. 3B). Thus, altogether especially
addition of TGF-β+RA did stabilise the phenotype of our generated aTreg cells. Furthermore, aCD4+TGF-β+RA aTreg cells displayed the highest regulatory potential in vivo reflecting the relevance of Helios co-expression as a quality property of generated Treg cells. In 2007, Huehn et al. identified the TSDR, a CpG island, which is completely demethylated in stable nTreg cells whereas it is partially or completely methylated in unstable iTreg cells, naïve T cells and effector T cells . When we assessed the demethylation of the TSDR, the purified Foxp3+ cells
from all culture settings showed 100% demethylation unless (Fig. 3E), whereas Foxp3− cells from the same cultures showed no demethylation and iTreg cells showed only partial demethylation of the TSDR. This let us assume that the aTreg cells obtained from the different cultures show the same stability. However, we detected diverse changes in the Foxp3 frequency when we restimulated the cells with alloantigen. Restimulation of aCD4+TGF-β+RA aTreg cells resulted in an increased frequency of Foxp3+ T cells as compared to the primary culture. In contrast, we detected a reduction in the frequency of Foxp3+ cells in CD4+CD25+ T cells obtained from all other cultures. One explanation may be an outgrowth of contaminating CD4+CD25+Foxp3− Teff cells. However, CD4+CD25+ cells from aCD4+Rapa cultures contained also very low numbers of contaminating Teff cells similar to those of aCD4+TGF-β+RA cultures. The addition of TGF-β+RA might have negatively influenced the few contaminating T effector cells in the primary culture so that after restimulation these cells proliferated less or became apoptotic.