In accordance with its original
In accordance with its original discovery, EBI2 may play its major role in response to viral infections or in a pathological context such as in autoimmunity. Although we found a significant delay of onset in the Th17 transfer EAE model, active EAE was unchanged in absence of EBI2. At the moment, we can only speculate about this difference. It might be that active EAE induction using CFA/MOG induces a stronger inflammatory array of immune cells, pro-inflammatory factors, cytokines, and chemokines, which might override the need for EBI2-mediated migration. Alternatively, there might be compensatory mechanisms for the lack of this GPCR, so that mice are susceptible to EAE also independently of EBI2 in knockout animals. We showed in transfer experiments with EBI2 mutant T cells that EAE was significantly delayed but after onset, disease proceeded normally. This might be due to the previously described plasticity of Th17 cells, which tend to change their expression profile toward the Th1 lineage together with high levels of GM-CSF (Hirota et al., 2011, Kurschus et al., 2010) (and herein). In absence of EBI2, EAE initiation may more depend on Th1 Rufinamide co-expressing GM-CSF than on Th17 cells. This hypothesis is supported by our data showing a major population of Th1 cells co-expressing GM-CSF in the CNS during disease. We found that Th17 cells, derived either from in vivo immunization or in vitro polarization under pathogenic conditions (in presence of IL-23 and/or IL-1β) expressed high levels of EBI2. In contrast, Th17 cells that were generated in absence of IL-23 or IL-1β lost EBI2 expression during differentiation. Addition of various cytokines revealed that IL-2 was the strongest in promoting down-modulation of EBI2 (data not shown). This result is in line with the inhibitory action of IL-2 on Th17 development (Laurence et al., 2007). Our data are also in line with a previous report showing that Gpr183 is specifically upregulated in Th17 differentiation under pathogenic conditions (using medium containing TGF-β3) (Lee et al., 2012). Hence, EBI2 expression seems to be part of a pathogenic T cell signature and, based on our results with the Th17-mediated EAE transfer model, we suggest that EBI2 confers pathogenicity to encephalitogenic T cells by enhancing migration into the inflamed CNS. In mice, our EBI2EGFP reporter showed that EBI2 is expressed consistently in naive T helper cells. Already single positive CD4+ thymocytes express EBI2 (not shown). In effector and regulatory T cells in naive mice, the percentage of EBI2 expression was lower (∼40%). Like others (Liu et al., 2011), we found that for efficient in vitro migration, polyclonal pre-stimulation of the T cells was needed (not shown). Our findings that 2D2 cells upregulate EBI2 in vivo during priming is in line with enhanced levels of EBI2 expression necessary for directed migration toward 7α,25-OHC. Concomitant with a function of EBI2 for effector cell migration, human CD45RO− memory/Teff cells expressed higher surface EBI2 protein than naive T helper cells. Our data indicate that EBI2 is one of several molecules mediating migration of T cells into the inflamed CNS. Because other important molecules for tissue infiltration, such as CD44, VLA4, and LFA1, are not expressed on naive T cells, those cells may be precluded from immigration to the inflamed CNS despite expression of EBI2. Here, we show that enzymes responsible for synthesis of 7α,25-OHC from cholesterol are highly expressed in the CNS of mice in response to EAE induction. This was paralleled by increased levels of the EBI2 ligands 7α,25-OHC and 7α,27-OHC in SCs of mice with EAE. Therefore, we suggest that during inflammation, expression of CH25H and CYP7B1 sustain lymphocyte accumulation at sites of inflammation/infection. It has been suggested that expression of CH25H supports transmigration of activated CD44+ T helper cells into activated tissues such as the inflamed CNS. Similarly to our data obtained from the EAE transfer model, the lack of CH25H delays onset of EAE with similar end scores and similar infiltration and demyelination at the end of the experiments (Chalmin et al., 2015). Bone marrow chimeric experiments by the same group suggested that CH25H was needed to be expressed in cells of hematopoietic origin (supposedly monocytes/macrophages or moDCs) rather than in cells of the CNS tissue itself to promote EAE development. Contrary to this, we found that Ch25h is expressed by CNS-derived microglia in EAE and not by infiltrating myeloid cell populations. Interestingly, the second enzyme for generating 7α,25-OHC CYP7B1 was expressed by infiltrating lymphocytes and monocytes. This suggests that activated microglia and infiltrating cells have to cooperate in production of the cell-permeable ligand as it was demonstrated previously for different cell lines expressing these enzymes separately (Yi et al., 2012). We did not define the lymphocyte population expressing CYP7B1. Interestingly, NK cells can express this enzyme (Liu et al., 2011) and were also shown to infiltrate the CNS in EAE (Huang et al., 2006). We also did not analyze astrocytes for expression of these enzymes in EAE. It was recently shown that microglia (Butovsky et al., 2014, Olah et al., 2012) as well as murine astrocytes can express the necessary enzymes for production of the EBI2 ligand (Rutkowska et al., 2015, Rutkowska et al., 2016). Alternative sources for EBI2 ligand production in EAE under investigation by us may be endothelial cells. How microglia cells become activated to express Ch25h in EAE is not clear. Because we observe major differences in transfer EAE of Th17 cells, we assume that expression of Ch25h is not a response to CFA used in active EAE induction but rather induced by inflammatory cytokines.