91 TSPO in neurodegeneration Muroidea promoter (Fig. 2c). We expanded this motif search and TSPO promoter sequence divergence analysis to a wider range of 24 rodent species from the Muroidea superfamily and other non-Muroidea rodents. Again, we found that the AP1 site is confined only to a subset of the superfamily Muroidea (Fig. S2). Silencing AP1 impairs LPS induced TSPO expression in the immortalized mouse BV2 cell line65. We therefore tested the hypothesis that LPS inducible TSPO expression occurs only in species with the AP1 binding site in the promoter region. In species that lack the AP1 binding site (human, pig, sheep, rabbit), TSPO expression was not induced by LPS (Fig. 2d). However, in the rat, where the AP1 binding site is present, TSPO was increased under these conditions (Fig. 2d). Microglial TSPO expression is unchanged in the AD hippocampus, but is increased in amyloid mouse models Microglia-neuronal interactions, which modulate microglia inflammatory phenotype1, are lost in monocultures in vitro. We therefore examined TSPO expression within inflammatory microglia in situ with quantitative neuropathology using postmortem samples from AD (Table S1). We compared data from human postmortem AD brain to the AppNL-G-F and TAUP301S mouse models. We examined the hippocampal region, one of the most severely affected regions in AD66,67, comparing it to non-neurological disease controls (Fig. 3a-c). No increases were observed in the number of IBA1+ microglia (Fig. 3d), HLA-DR+ microglia (Fig. 3e) or astrocytes (Fig. 3f) and the density of TSPO+ cells in AD did not differ compared to controls (Fig. 3g). Additionally, there was no increase in TSPO+microglia (Fig. 3h,i) and astrocytes (Fig. 3j). We then quantified TSPO+ area (µm2) in microglia and astrocytes as an index of individual cellular expression (see methods). There was no difference in individual cellular TSPO expression in microglia (Fig. 3k) or astrocytes (Fig. 3k) in AD relative to controls. We next conducted multiplexed proteomics with imaging mass cytometry (IMC) for further characterisation of cellular phenotype. As with the IHC, we did not see an increase in microglial density, as defined by the number of IBA1+ cells per mm2, (Fig. S3a) nor in the density of astrocytes (Fig. S3b). Furthermore, again in agreement with the IHC, we did not see an increase in the number of microglia and astrocytes expressing TSPO (Fig. S3c,d). However, IMC did reveal an increase in CD68+ microglia cells (Fig S3e) in AD compared to control, providing evidence, consistent with the literature68,69, that microglia are activated in AD. However, despite microglial activation, we did not find an increase in individual cellular TSPO expression, defined here as mean cellular TSPO signal, in either microglia (Fig. S3f) or astrocytes (Fig. S3g) in AD donors relative to control. Because proximity to amyloid plaques is associated with activation of microglia68, we next tested whether cellular TSPO expression was higher in plaque microglia relative to (more distant) non-plaque microglia in the same tissue sections from the AD brains only. We saw no differences in cellular TSPO expression between the plaque and non-plaque microglia (Fig. S3h). We next compared the human AD data to that from mouse AppNL-G-F (Fig. 4a,b) and TAUP301S (Fig. 4,i,j). The AppNL-G-F model avoids artefacts introduced by APP overexpression by utilising a knock-in strategy to express human APP at wild-type levels and with appropriate cell-type and temporal specificity18. In this model, APP is not overexpressed. Instead, amyloid plaque
RkJQdWJsaXNoZXIy MTk4NDMw