Page 24 - Read Online

Page 202 Walker. Neuroimmunol Neuroinflammation 2020;7:194-214 I http://dx.doi.org/10.20517/2347-8659.2020.09

proinflammatory cytokines and chemokines such as interleukin (IL)-1α, IL-1β, IL-6, IL-8, tumor necrosis
factor-α (TNF-α), interferon-gamma (IFN-γ) and CCL-2. Demonstration of cytokines IL-1α, IL-1β and
TNF-α in AD brain microglia has been reported but these are not widely-used markers for describing
microglia in tissue [52-55] . There appears to be technical difficulties in localizing these secreted cytokines in
tissue, and it should be noted that these classical cytokines do not prominently feature in the microglial
[24]
disease-associated gene signatures of recent studies [10,11,24,32] . In the paper of Friedman et al. , they provide
supplementary data from gene expression profiles of two studies comparing control and AD samples;
neither of these detected increased expression of these classical cytokines in AD samples (supplementary
[24]
data file in reference ).

DISCREPANCIES BETWEEN GENE PROFILING RESULTS AND TISSUE STUDIES OF
MICROGLIA
Apolipoprotein E and Complement C1q
Apolipoprotein E (APOE/Apoe) and the three complement C1q genes (C1QA/C1qa, C1QB/C1qb, C1QC/
C1qc) have been identified as microglial markers by expression studies, but these proteins have not been
identified in microglia in AD tissue sections. APOE and C1Q proteins can be detected and are associated
with Aβ plaques in AD brains [42,44] , with expression of C1Q protein being detectable in neurons [45,56] . Recent
experimental studies with mouse models showed that the majority of C1QA proteins in mouse brain was
[57]
derived from microglia , and that C1Q overexpression can have a neuroprotective rather than pathogenic
role . As the majority of microglial gene profiling studies used rodent AD models, it is possible that these
[58]
discrepancies are due to species differences in gene expression of human compared to rodent cells.

TREM2
TREM2 has become the most widely studied inflammatory/microglia marker for studies linking
inflammation and AD. This came about due to the identification of heterozygous single nucleotide
polymorphisms (SNP)/mutations associated with increased risk of developing AD [59,60] ; and studies
of Nasu-Hakola disease (NHD), which is associated with homozygous mutations in TREM2 gene or
[61]
TYROBP gene . TYROBP gene encodes DAP12, the essential adaptor protein that mediates TREM2
signaling. Patients with NHD, also known as polycystic lipomembranous osteodysplasia with sclerosing
leukoencephalopathy (PLOSL), develop a type of dementia similar to frontotemporal dementia. This
dementia appears to be directly caused by microglial dysfunction resulting from the loss of function of
TREM2 signaling. The function of TREM2 has been extensively characterized as a receptor for lipids,
[65]
[66]
lipoprotein [62-64] , including Apoe , heat shock protein 60 (HSP60), and Aβ peptide . TREM2 appears
to be a pattern recognition type of receptor rather than being ligand sequence specific. It is still not clear
whether TREM2 is recognizing Aβ in plaques, or one of the many different plaque-associated proteins
[67]
or lipids that accumulate, including lipoproteins such as ApoE and ApoJ (clusterin) . An interaction
of TREM2 and APOE signaling pathways has been implicated in altering microglia to a more damaging
[68]
phenotype . However, there are still inconsistencies about whether TREM2 signaling functions in a pro-
[32]
inflammatory (damaging) or anti-inflammatory (reparative) manner .
If one now considers the theme of this review as to whether TREM2 expression can be used to describe
the phenotypes of microglia in human brain tissue, findings on this have been scarce and divergent.
Studies have consistently shown that TREM2 mRNA is highly expressed in human and animal brains and
in microglia, and many mechanistic studies of TREM2 associated with disease assume that all microglia
express TREM2. However, few studies have successfully localized TREM2 expression to microglia in
human or rodent brain tissue. An earlier study using aged Aβ plaque-developing mice showed TREM2
[69]
expression by microglia associated with plaques , but two studies that used hard-fixed paraffin embedded
human brain tissue with antigen retrieval concluded that TREM2 was not expressed by human brain