Marilen Benner

SUMMARY AND DISCUSSION 223 8 production (28-30). Especially their capacity to secrete IL-10 has gained attention, starting with the observation that the inability of B cell-deficient mice to recover from experimental autoimmune encephalomyelitis stems from the lack of IL-10 producing B cells (31, 32). The idea that B cells are not the cause, but a suppressor of autoimmune disease contrasted all other approaches focusing on auto-antibody production. The intriguing observation that a major part of decidual B cell samples, especially of the second trimester, expressed CD24 and CD38 to a high extent (Chapter 4) , often connected to the ability to produce IL-10, sparked a new question: could B cells of the fetal-maternal interface contribute to local maintenance of tolerance? Recently, few studies have started to acknowledge a possible role for B cells in healthy pregnancy. Rolle et al. initially demonstrated that systemic B cells of pregnant women were able to suppress TNF α production of CD4 + T cells, which could not be mimicked in the same setting by non-pregnant donors (33). An increase in the pool of CD19 + CD24 Hi CD27 + B cells due to pregnancy was not shared by women who had suffered from a spontaneous abortion. Serum of women during pregnancy, but not after pregnancy loss, was able to induce IL-10 production by B cells of peripheral blood (33). This was attributed to pregnancy-induced high levels of human chorionic gonadotropin (hCG) as 95% of CD19 + CD24 Hi CD27 + expressed the hCG receptor and CD19 + cells produced IL-10 upon hCG stimulation (33). HCG is secreted by the blastocyst even before its implantation and before serum levels are measurable (34, 35). In this period, it likely serves a juxtacrine function by modulating levels of e.g. metalloproteinases, insulin-like growth factor-binding protein -1 (IGFBP-1), Leukemia Inhibitory Factor (LIF), M-CSF, vascular endothelial growth factor (VEGF) (36). Its local action in stimulating IL-10 production by B cells was then shown in co-cultures of trophoblast cell lines, and murine placental explants, with B cells. The interaction of these fetal cells and B cells induced IL-10, but this was abrogated when hCG was blocked (37). Supporting the notion that induction of IL-10 production of B cells is a local phenomenon upon trophoblast encounter, we observed that decidual B cells are potent producers of IL-10 upon CpG/CD40L stimulation, an accepted strategy to detect IL-10 secretion by B cells (38-40), and do so to a higher extent than systemic B cells. This fits the results of Fettke et al., who pointed out that co- culture with fetal cells had an additive effect on IL-10 levels compared to CD40L/CpG stimulation alone (37). Likewise, we observed that, in a number of samples, decidual B cells produced IL-10 without the need for CpG/CD40L stimulation supporting the idea that B cells have already been primed in utero , through hCG, trophoblast cells, or other trophoblast related receptor-ligand interactions, to secrete IL-10. Our observation that B cells are located in clusters, together with T cells, points out that even though their numbers are limited, decidual B cells might facilitate IL-10 mediated inhibition of neighboring cells. Of note, lymphocyte clusters could not be found for all decidual samples that we investigated by immunohistochemistry, but in approximately half of all cases. This might be sampling issue, clusters might not be located all throughout the decidua and only a small proportion of tissue is embedded for imaging. Another possible explanation might be time