79 TSPO in neurodegeneration Introduction Neuronal-microglial signalling limits microglial inflammatory responses under homeostatic conditions1. The loss of this cross talk in central nervous system (CNS) pathology partly explains why microglia adopt an activated phenotype in many neurodegenerative diseases2,3. Genomic, ex vivo and preclinical data imply that microglial activation also may contribute to neurodegeneration4, for example, by releasing inflammatory molecules in response to infectious or damage-related triggers5. These lead to both neuronal injury and, more directly, pathological phagocytosis of synapses5,6. Development of tools which can reliably detect and quantify microglial activation in the living human brain has been an important goal. By enabling improved stratification and providing early pharmacodynamic readouts, these would accelerate experimental medicine studies probing disease mechanisms and early therapeutics. Detection of 18kDa Translocator Protein (TSPO) with positron emission tomography (PET) has been widely used to quantify microglial activation in vivo7. In the last 5 years alone, there have been ~300 clinical studies using TSPO PET to quantify microglial responses in the human brain, making it the most commonly used research imaging technique for this purpose. The TSPO signal is not specific to microglia, and the contribution from other cell types (particularly astrocytes and endothelial cells) is increasingly acknowledged8. The justification for quantifying TSPO as a marker of microglial activation is based on the assumption that when microglia become activated, they adopt a classical pro-inflammatory phenotype and TSPO expression is substantially increased7,9,10. This has been demonstrated repeatedly inmice, both in vitro and in vivo11-14. We have shown, however, that classical proinflammatory stimulation of human microglia and macrophages in vitro with the TLR4 ligand lipopolysaccharide (LPS) does not induce expression of TSPO15. Furthermore, in multiple sclerosis (MS), TSPO does not appear to be increased in microglia with activated morphology 16. These data appear inconsistent with the assumption that TSPO is a marker of activated microglia in humans. To address this issue, we performed a meta-analysis of publicly available expression array data and found that across a range of pro-inflammatory activation stimuli, TSPO expression is consistently and substantially increased in mouse, but not human macrophages and microglia in vitro. We then performed a comparative analysis of the TSPO promoter region in a range of mammalian species and found that the binding site for AP1 (a transcription factor which regulates macrophage activation in rodents17) is present in and unique to a subset of species within the Muroidea superfamily of rodents. Consistent with the hypothesis that this binding site is required for the increase in TSPO expression that accompanies pro-inflammatory stimulation, we show that TSPO is inducible by LPS in the rat (another Muroidea species with the AP1 binding site in the TSPO core promoter) but not in other mammals. Because neuronal interactions modulate microglial phenotype, we then compared microglial TSPO expression in neurodegenerative diseases affecting the brain and spinal cord (Alzheimer’s Disease (AD) and amyotrophic lateral sclerosis (ALS), respectively) as well as the classical neuroinflammatory brain disease MS which features highly activated microglia. We compared each human disease to its respective commonly used mouse models (amyloid precursor protein (AppNL-G-F)18, tau (TauP301S)19, superoxide dismutase 1 (SOD1G93A)20, and experimental autoimmune encephalomyelitis (EAE) in young and aged animals21. We also studied TSPO expression with EAE in the marmoset in conjunction with frequent MRI scanning that allowed for identification of the acute lesions which contain pro-inflammatory microglia. Consistent
RkJQdWJsaXNoZXIy MTk4NDMw