Page 19 - Read Online
P. 19

Walker. Neuroimmunol Neuroinflammation 2020;7:194-214  I  http://dx.doi.org/10.20517/2347-8659.2020.09                Page 197

               results have been analyzed comparing AD animals with non-transgenic control animals, or between
                                                      [24]
               AD and controls from aged human tissues . Although the focus of this article is on understanding
               how to define activation states in human brains, findings from rodent models in this context have to
               be considered. Microglia can be directly isolated from the mouse brain by Dounce homogenization to
               break up tissue, filtering through 70 mm mesh, separation by magnetic beads conjugated with anti-myelin
               antibody to remove myelin, and then selected using a fluorescence-activated cell sorter with appropriate
                                                                                      [25]
               labeled antibodies (e.g., CD11b, CD45) to isolate immune cells, including microglia . This basic approach
               will isolate populations of cells that can be diluted to allow the isolation of single cells or analysis in bulk.
               Refinements to these techniques have allowed the sorting and RNA profiling of cellular nuclei from frozen
               human and animal tissue samples [11,26] . Different approaches for microglial profiling are illustrated in Figure 2.

               Using these approaches, it was shown that trans-membrane protein 119 (TMEM119) was a specific marker
                                                   [25]
               for microglia in mouse and human brains . The isolation of microglia from human brains using the same
               methodology is possible but has some limitations. Human brain tissue is not usually amenable to Dounce
               homogenization and requires additional enzymatic digestion to dissociate tissue into single cells, and
                                                                                                [27]
               density gradient centrifugation to separate the myelin content from the cellular components . Studies
               using these approaches aimed to define genes that are unique to microglia and not expressed, or expressed
               at low levels in blood monocytes/macrophages [28-31] .


               Recent important studies relating to inflammatory changes in AD brains identified a type of “disease-
               associated microglia” (DAM) that appear to be associated primarily with preventing inflammatory
               pathology rather than enhancing it [26,32,33] . Another key study involved the meta-analyses of multiple
                                                                     [24]
               different gene profiling studies related to brain inflammation . Amongst other findings, these studies
               confirmed TMEM119, purinergic receptor P2YR12 and fractalkine receptor CX3CR1 as markers with
               highly enriched expression in microglia compared to monocytes/macrophages.

               DAM
               The identification of DAM was primarily carried out using single cell microglia RNA sequencing
               in AD model mice (5xFAD) of different ages followed by validation in human tissue samples. The
               progressive changes in microglia phenotypes was from homeostatic (non-activated) to stage 1 DAM,
               which represents a state of proinflammatory activation, to stage 2 DAM, an altered phenotype that
               restricts neurodegenerative changes. Based on results that included the use of mice that are gene deficient
               (knockout) for the crucial triggering receptor expressed on myeloid cells (TREM2), it was shown that
               the transition from stage 1 to stage 2 DAM was dependent on Trem2 signaling. These data have implied
                                                                                    [32]
               that the activation of Trem2 signaling was protective rather than pathogenic . [Gene identification
               primarily from studies using rodents will use the lower case abbreviation; genes primarily identified
               in human will use the upper-case abbreviation]. This study defined expression of genes Hexb (Beta-
               hexosaminidase subunit beta), Cst3 (Cystatin C), Cx3cr1 (Fractalkine receptor), Ctsd (Cathepsin D),
               Csf1r (Colony stimulating factor-1 receptor), Ctss (Cathepsin S), Sparc (Osteonectin), Tmsb4x (Thymosin
               beta-4), P2ry12 (Purinergic P2Y receptor 12), C1qa (Complement C1q subunit A), C1qb (Complement
               C1q subunit B) as features of homeostatic microglia. Activation and transition to stage 1 DAM involved
               downregulation of identified homeostatic genes Cx3cr1 (Fractalkine), P2ry12 (Purinergic 2Y receptor
               12) and TMEM119 (Transmembrane Protein 119), along with P2ry13 (Purinergic P2Y receptor 13),
               Tgfbr1 (Transforming growth factor receptor beta 1), Txnip (Thioredoxin-interacting protein) and Glu1
               (Glucoamylase 1), and upregulation of Tyrobp (TYRO protein tyrosine kinase binding protein - DAP12),
               Ctsb (Cathepsin B), Cstb (Cytstatin B), Ctsd (Cathepsin D) Apoe (Apolipoprotein E), B2m (Beta-2
               microglobulin), Fth1 (Ferritin heavy chain-1), Timp2 (Tissue inhibitor of metalloprotease-2), H2-D1
               (H2 class 1 histocompatability antigen) and Lyz2 (Lysozyme C-2). The expression of homeostatic genes
               C1qc (Complement C1q subunit C), C1qb, C1qa, Ctss, Hexb, Olfml3 (Olfactomedin-like 3), Csf1r and
   14   15   16   17   18   19   20   21   22   23   24