MAIT cells in placental tissues and their reconstitution following allogeneic hematopoietic stem cell transplantation

Abstract

The placenta is a temporary organ of human reproduction. Both the fetus and placenta are covered in a membrane of maternal origin, the decidua. After the 1st trimester, the fetus is supplied with oxygen and nutrients by maternal arterial blood that penetrates the decidua and fills the intervillous space of the placenta. Fetal blood vessels covered in a thin cell layer, the villi, protrude down into this intervillous blood (IVB), and gases and molecules are transported in and out of the fetal circulation. The maternal immune system recognizes fetal antigen as foreign, yet in the case of a successful pregnancy, it tolerates the fetus while still maintaining immunity against pathogens. Maternal immune cells come into contact with fetal antigen in the decidua and the intervillous space, and these immunological sites are referred to as the feto-maternal interface. Mucosal associated invariant T (MAIT) cells is a large subset of antigen-specific T cells. MAIT cells are activated by metabolites from the synthesis of vitamin B2 by certain species of bacteria and fungi, which are presented on the MHC class I related molecule (MR1). As humans cannot synthesize vitamin B2, MAIT cells can thus discern between self and non-self.

In this thesis, we have studied the immune cell composition and function at the feto-maternal interface with special focus on MAIT cells. We also studied the reconstitution of MAIT cells in the conceptually interesting context of allogeneic hematopoietic cell transplantation (HCT). We found that IVB has a fundamentally different composition of immune cells compared to peripheral blood (PB) from the same donors. The IVB was enriched in MAIT cells, effector memory T cells and B cells. The MAIT cells in IVB had a more potent response when stimulating mixed mononuclear cell cultures with bacteria compared to PB. MR1 was readily expressed by fetal macrophages inside the villi. We compared immune cells isolated from the two different parts of the decidual membrane, the decidua basalis (DB) that covers the placenta from the maternal side, and the decidua parietalis (DP) that is attached to the edge of the placenta, and covers the fetus and amniotic fluid. DP contained more T cells and CD56high NK cells, whereas DB was enriched in MAIT cells, B cells, monocytes and CD56dim NK cells. Immune cells in DP had a higher expression of co-inhibitory markers such as PD-1, TIM-3 and LAG-3. In spite of this expression, MAIT cells, T cells and NK cells from both DP and DB were functional when stimulated with bacteria or PMA/Ionomycin.

Non-pregnant women had a higher frequency of MAIT cells in PB compared to pregnant women at term. Using supernatants from placenta tissue explant cultures, we observed that MAIT cells and CD8+ effector T cells were selectively recruited in migration assays. When analyzing the chemokine and cytokine patterns in these supernatants and plasma samples from IVB, many similarities were seen. In contrast, the levels of chemokines in PB plasma was strikingly different compared to that of IVB plasma. Among these, macrophage migration inhibitory factor (MIF) was 182-fold higher in IVB compared to PB plasma. The frequency of MAIT cells in both IVB and DP correlated with levels of MIF in IVB. When blocking MIF together with CCL20 and CCL25, the migration of MAIT cells towards the placenta tissue explant supernatant was reduced. Using recombinant MIF protein, we could show that MIF attracted MAIT cells, probably by binding the chemokine receptor CXCR4. IVB contained a higher proportion of B cells compared to PB, and the IVB B cells were primarily of the mature/naïve phenotype. This subset of B cells was correlated with levels of CCL20 in IVB plasma. Mature/naïve B cells all expressed the receptor for CCL20, CCR6, and they had a higher median fluorescent intensity of CCR6 compared to immature B cells. Using the same migration assay as previously, we could see that placenta tissue explant supernatant attracted B cells. Lastly, we investigated the reconstitution of MAIT cells following HCT. MAIT cells did not start to increase in total number until two years after the transplantation, and as the non-MAIT T cells proliferated during this time, the relative proportion of MAIT cells remained at the same low levels during the observation period. When stimulating mixed mononuclear cells with bacteria, the MAIT cells from 2-6 months after HCT had an impaired interferon-γ response, whereas the response at 24 months was similar to that seen in healthy controls. Lastly, we showed that proliferating MAIT cells were more sensitive to common immunosuppressive drugs used in patients after HCT.

In conclusion, the immunological constitution and function of IVB seems to be shaped by soluble factors secreted by placental tissue. IVB is a specialized immunological environment enriched in MAIT cells and other effector T cells, as well as mature B cells, and challenging IVB mononuclear cells with bacteria led to a more potent MAIT cell response. This points to that the placenta attracts certain subsets of immune cells in order to uphold immunity at the feto-maternal interface. MAIT cell reconstitution following HCT was impaired, both in terms of cell number, frequency and function. This could partly be explained by an increased sensitivity of dividing MAIT cells to common immunosuppressive drugs used after transplantation. The impaired MAIT cell reconstitution could partly be an explanation of the increased risk of infectious complications following HCT