89 TSPO in neurodegeneration In the meta-analysis, there was a non-significant trend towards a reduction in human TSPO expression under pro-inflammatory conditions (Fig. 1b). In the individual datasets, TSPO was unchanged in 33/42 (79%) of the datasets, significantly downregulated in 8/42 (19%) and significantly upregulated in 1/42 (2%). In contrast to the findings in mice, our analysis thus suggests that TSPO expression is not upregulated in human microglia and macrophages after pro-inflammatory stimulation in vitro. To test whether TSPO gene expression changes are regulated at an epigenetic level, we analysed publicly available ChIP-seq datasets for histone modification in mouse and human macrophages before and after treatment with IFNγ23 35 (Fig. 1c-f). Levels of H3K27Ac and H3K4me1 histone marks in the enhancer regions are associated with increased gene expression23,62. While both histone modifications were increased after IFNγ treatment in TSPO promoter regions in macrophages from mouse, they were decreased in humans (Fig. 1c,d). Consistent with this epigenetic regulation, Tspo gene expression was upregulated in mouse macrophages after IFNγ but not in human macrophages in RNAseq data from the same set of samples (Fig. S1a). The PU.1 transcription factor is a master regulator of macrophage proliferation and macrophage differentiation63,64. Because PU.1 increases Tspo gene expression in the immortalised C57/BL6 mouse microglia BV-2 cell line65, we next investigated whether TSPO expression in macrophages is regulated by PU.1 binding in human in publicly available ChIPseq datasets. An increase in PU.1 binding in the mouse Tspo promoter after IFNγ treatment was observed (Fig. 1c). However, PU.1 binding to the human TSPO promoter was decreased after IFNγ treatment (Fig. 1d). To test whether the reduced PU.1 binding at the human TSPO promoter was due to reduced PU.1 expression, we analysed RNAseq data from the same set of samples. Expression of SPI-1, the gene that codes for PU.1, was not altered in human macrophages after IFNγ treatment (Fig. S1b), suggesting that the reduced binding of PU.1 to the human TSPO promoter region was unlikely to be due to reduced PU.1 levels. This suggests that repressive chromatin remodelling in the human cells leads to decreased PU.1 binding, a consequence of which could be the downregulation of TSPO transcript expression. This is consistent with the meta-analysis (Fig. 1a,b); although TSPO expression with inflammatory stimuli did not significantly change in most studies, in 8/9 (89%) of studies where TSPO did significantly change, it was downregulated (Fig. 1b). Together this data shows that in vitro, pro-inflammatory stimulation of mouse myeloid cells increases TSPO expression, histone marks in the enhancer regions and PU.1 binding. These changes are not found following proinflammatory stimulation of human myeloid cells. The presence of the AP1 binding site in the TSPO promoter and LPS inducible TSPO expression is unique to the Muroidea superfamily of rodents To understand why TSPO expression is inducible by pro-inflammatory stimuli in mouse but not human myeloid cells, we performed multiple sequence alignment of the TSPO promoter region of 15 species including primates, rodents, and other mammals (Fig. 2). We found that an AP1 binding site is present uniquely in a subset of species within the Muroidea superfamily of rodents including mouse, rat and chinese hamster (Fig. 2a). These binding sites were not present in other rodents (squirrel, guinea pig), nor in other non-rodent mammals (Fig. 2a). We generated a phylogenetic tree which shows a clear branching in the TSPO promoter of
RkJQdWJsaXNoZXIy MTk4NDMw