203 General Discussion investigation of its response to injury and neuroinflammation. Knowledge of the cellular and phenotypic correlates of the TSPO PET signal is important to understand the clinical meaningfulness of TSPO PET as a diagnostic tool. For example, there is heterogeneity in the radioligand uptake of T2-hyperintense lesion in MS before and after initiation of disease modifying therapies24. However, before the relevance of such TSPO PET expression can be explored we need to knowwhere the signal is coming from. Many studies have largely ignored the possible contribution of cell types other than microglia to the TSPO PET signal25-41, either by not investigating expression by cell types than microglia, or not performing quantitative experiments at all. Together, the previously published studies has created the foundation for the misconception that TSPO is almost exclusively present in microglia with almost no association with astrocytes – even though the first study showing TSPO in astrocytes is over two decades old42. Fortunately, emerging studies have reported TSPO in astrocytes in animal models of neurodegenerative diseases5-18 and a few studies show astrocytic TSPO in human disease19-21. TSPO is widely considered as a marker for neuroinflammation, and thus many studies on TSPO PET have been performed in MS, where neuroinflammation, a prominent hallmark of disease is associated with demyelination43, as well as neuronal damage and synaptic loss (Chapter 6). Many studies find increases in TSPO PET signal in active and chronic active lesions in MS40,44-47 and overall the increased TSPO PET signal in the brain predicts disease progression of MS regardless of relapses48. However, as mentioned above this increase in TSPO PET was widely regarded to originate in pathogenic activated microglia2. In Chapter 2 we show that while microglia are the main cells expressing TSPO in MS lesions there is also a significant contribution of astrocytes present in the centre of chronic active and inactive lesions. The finding that in the centre of chronic active and inactive MS lesions the contribution of TSPO is significantly higher in astrocytes, as compared to microglia, is consistent with reports of newer generation ligands for TSPO that bind to reactive astrocytes6,49. Autoradiography with 3H-PBR28 and 3H-PK11195 revealed a strong association between TSPO+ cells and radioligand binding across all cell subtypes. Increased TSPO expression was present not only in white matter lesions but also in leukocortical lesions, lesions that comprise both white and grey matter. The more inflammatory component of the white matter in leukocortical lesions might influence the adjoining grey matter in terms of inflammation. Indeed, purely cortical grey matter lesions show little to no inflammation50-52. These findings indicate that TSPO PET signal must be interpreted in a context-dependent manner. Next, we performed a detailed neuropathological studies to investigate whether TSPO expression was limited to activated microglia – the widely accepted view. In contrast to this view, we showed that TSPO in microglia in MS lesions was present in microglia that expressed anti- and pro-inflammatory markers. Additionally, homeostatic microglia were also found to express TSPO, indicating that while increased TSPO PET signal is a marker of microglia it should not be used as a marker for a change in the activation state of microglia. Additionally, as autoradiography correlated with the number of TSPO+ cells in active MS lesions this indicates that TSPO PET could be a used to identify the microglial density, rather than activation state. A smaller portion of cells did not show colocalisation with GFAP (a marker for astrocytes) or HLA-DR (a marker used to identify activated myeloid cells), but showed strong microglial morphology. In Chapter 3 we thus investigated the possibility of TSPO+ microglia that colocalised with other microglial markers than HLA-DR - a proportion of microglia was found to express either HLA-DR or IBA-1 in the MS brain53. MS lesions were stained for
RkJQdWJsaXNoZXIy MTk4NDMw