Introduction
The neuropathology of AD is characterized primarily by the region-specific accumulation of amyloid beta (Aβ) into senile plaques and hyperphosphorylated tau into neurofibrillary tangles. Both plaques and tangles are widely hypothesized to contribute to the neurodegenerative changes that occur in AD and manifest clinically as dementia [
56]. AD neuropathology also involves several other significant components, including microglial activation, that are associated with disease progression. While it has long been known that activated microglia co-localize with Aβ plaques [
20], as noted by Alzheimer himself [
47], their roles remain incompletely defined. Microglia exhibit a broad range of actions implicated in both normal neural function [
6,
13,
18] and the development of disease [
5,
19]. In the context of AD, activated microglia have been theorized to exert dual effects on disease progression, promoting AD by driving neuroinflammation while also attenuating pathogenesis as a result of phagocytic actions [
29]. Moreover, recent observations suggest that microglia interact with plaques to form a barrier that reduces the outward extension of Aβ fibrils, which may protect nearby neurites from damage [
4,
9,
21,
25,
27,
52,
58,
61,
63,
64].
The regulation of microglial-plaque associations is a topic of high importance. A key molecule in the regulation of these interactions is triggering receptor expressed on myeloid cells 2 (TREM2), a microglial cell surface receptor of the immunoglobulin superfamily that senses damage in the central nervous system [
42,
57]. TREM2 activation is essential for immune function in the brain, promoting proliferation, tropism and survival of microglia [
38,
52]. Heterozygous TREM2 mutations, which yield partial loss of function, confer higher AD risk [
14,
23,
24,
46] and are associated with reduced interactions of microglia with plaques [
63]. Rodent studies confirm that microglial-plaque interactions are TREM2-dependent. Specifically, TREM2 deficiency and haploinsufficiency in mouse models of AD are associated with disrupted clustering of microglia around Aβ plaques [
21,
53,
57,
58] and diminished ability of microglia to form barriers around amyloid deposits, compact plaques, and reduce plaque-associated dystrophic neurites [
63].
Apolipoprotein E (
APOE) genotype, the most widely shared genetic risk factor for late-onset AD [
16,
28], is a strong candidate regulator of TREM2-dependent microglial interactions with amyloid plaques. First,
APOE genotype is an important modulator of microglial activation, with the AD-associated
APOE ε4 allele (
APOE4) linked with increased microgliosis and neuroinflammation [
50]. Second, apoE is a ligand for TREM2 [
2,
3,
61] that directly and/or indirectly activates TREM2-mediated signaling pathways, including those that induce phagocytosis and anti-inflammatory cascades [
22,
53]. Third, cell culture evidence suggests that
APOE4 is associated with greater depletion of TREM2 expression following acute immune challenge [
31], suggesting perhaps that
APOE4 may diminish TREM2-mediated actions. Whether interactions between
APOE and TREM2 extend to microglial plaque interactions is unknown.
Our current study investigates the effects of
APOE genotype on TREM2-dependent microglial interactions with plaques. We utilized the EFAD transgenic mouse model of AD, which includes hemizygous expression of 5xFAD model with knock-in of homozygous human
APOE3 or
APOE4 [
62]. Because sex significantly affects
APOE4 risk for AD in humans [
1,
11,
37] and AD-related pathology in transgenic mice [
7], and because sex regulates microglia phenotype [
55], we also included sex as a modulating variable. Our results indicate that TREM2-dependent microglial interactions with plaques are significantly affected by both
APOE genotype and sex with indices of microglial interactions showing the poorest outcomes with
APOE4 genotype and female sex. These findings identify a new role for
APOE genotype in regulation of microglia and AD pathogenesis and highlight the importance of sex as a modulator of these relationships.
Discussion
In this study, we examined how microglial interactions with amyloid plaques are affected by
APOE genotype and/or sex. Several novel observations document that AD risk factors impact pathogenesis, at least in part, by regulating microglial functions. First, we observed that microglial plaque coverage in EFAD mice is significantly reduced by both
APOE4 and female sex. Consistent with prior findings that microglial plaque coverage is positively associated with plaque compaction [
9], we observed that plaque circularity, an index of plaque compaction [
63], is reduced by both
APOE4 and female sex relative to male E3FAD mice. Previously, microglial interactions with plaques have been shown to be TREM2 dependent [
63]. In support of this we found that TREM2 expression levels in plaque-associated microglia, and specifically within their processes proximal to plaques, were lower in EFAD mice with
APOE4 or female sex relative to male E3FAD mice. Interestingly, the relationship among TREM2 expression,
APOE genotype, and sex was not significant in microglia located away from plaques. This finding, therefore, indicates that some aspects of microglial function differ in the presence versus absence of plaques, which is consistent with recent observations of microglial heterogeneity in relation to plaque proximity [
10,
41]. Importantly, numbers of microglial processes within the plaque environment were the same across
APOE genotypes and sex, suggesting that our observed differences in microglial plaque interactions were associated not with the availability of microglial processes, but rather functional aspects of microglia in
APOE4 and female EFAD mice that may affect their ability to detect and/or interact with plaques. Consistent with this possibility, we observed that the number of processes per microglia in the near plaque environment was significantly lower in
APOE4 and female EFAD relative to male E3FAD mice. In addition, microglial burden and activation were higher in females and E4FAD mice of both sexes than in male E3FAD mice. Further, male E3FAD mice showed the lowest amyloid burden and female E4FAD the highest, a pattern opposite to that observed with microglial plaque coverage. An intriguing possibility is that increased plaque coverage may contribute to lower pathology, for example, as a result of plaque compaction and phagocytosis. Indeed, we observed a significant inverse association between plaque coverage and plaque perimeter specifically in male E3FAD mice. In support of this idea, recent findings have shown that disruption of microglial plaque coverage, resulting from AD-associated TREM2 mutations and TREM2 hemizygosity, are associated with reduced Aβ accumulation [
36].
Our findings add to a growing literature indicating the importance of microglial pathways in
APOE genotype influences on the development of AD. Across both sexes, we observed that
APOE4 was associated with increased overall Iba1 burden and activated microglial phenotype, consistent with prior observations in EFAD mice [
49]. Further, we found that microglia in E4FAD mice exhibit reduced plaque coverage. To our knowledge, this is the first report to define the effects of
APOE genotype on the recently characterized, TREM2-dependent plaque coverage by microglia. Prior studies using less specific analyses have yielded conflicting findings. Yang et al.
, (2013) found that levels of plaque association of bone marrow-derived macrophages transplanted into irradiated APPswe/PS1Δ E9 mice were lower in macrophages from
APOE4 than from
APOE3 mice [
60]. In contrast, Rodriguez et al.
, (2014) reported greater association of microglia with cortical plaques in E4FAD versus E3FAD mice [
44]. Rodriguez and colleagues also reported larger plaque size and a greater proportion of compact plaques in the subiculum of E4FAD mice in comparison to E3FAD mice [
44]. Although we did not quantify plaque size and morphology in the same manner, our findings of higher levels of microglial plaque coverage and plaque circularity in male E3FAD mice may be predicted to yield smaller plaque size and a higher proportion of compact plaque morphology [
63]. Apparent differences between our findings and those of Rodriguez et al.
, may reflect important methodological variables including differences in staining (immunochemistry versus ThioS) and plaque size inclusion criteria. Perhaps most importantly, our data show that differences between E3FAD and E4FAD mice in microglial interactions with plaques are significantly affected by sex, a variable not considered in prior work.
While the mechanisms contributing to the observed regulation of microglial plaque coverage by
APOE are unclear, interactions between
APOE and TREM2 are increasingly recognized as significant contributors to AD-related microglial activity and represent a compelling candidate pathway for the regulation of plaque interactions [
12,
59]. For example,
APOE has been identified as a key regulator of the molecular signature of microglia via interactions with TREM2 [
25,
27,
32,
35,
40]. These investigations reveal that microglia surrounding plaques differ in their morphology and molecular expression profile from microglia that are distal to plaques [
25]. Importantly, an
APOE-TREM2 pathway appears to drive the conversion of homeostatic microglia to a disease-associated phenotype [
25,
27]. Because microglial actions exert both disease-promoting and protective outcomes, regulation of the
APOE-TREM2 pathway is expected to significantly affect AD pathogenesis with the overall effect depending, in part, upon several variables including timing. A recent study found that
APOE knockout in two different AD mouse models was associated with decreased microglial clustering around plaques as well as a reduction in plaque compaction [
54], microglial actions established to be TREM2-dependent. Further, microglia-specific knockout of
APOE in 5xFAD mice attenuated microglial transition to the disease phenotype, partially rescuing neuronal cell death [
27]. How
APOE genotype affects microglial transition to disease phenotypes via the
APOE-TREM2 pathway is not known. Our observations of impaired microglial coverage and plaque compaction in E4FAD mice is qualitatively consistent with findings in
APOE knockout mice [
54], suggesting that apoE4 represents reduced functionality relative to apoE3 in terms of TREM2-dependent microglial actions.
We also identify novel sex differences in microglial interactions with amyloid plaques. The observed sex differences are most apparent in
APOE3 genotype in which male E3FAD mice exhibit increased microglial plaque coverage, plaque compaction, microglial TREM2 expression and reduced plaque burden in comparison to age-matched female E3FAD mice. While the specific mechanisms contributing to these observed sex differences remain to be elucidated, our findings are consistent with abundant recent evidence of wide-ranging sex differences in microglia. Indeed, microglia from male and female rodents exhibit brain-region specific differences in the numbers and activation states during development [
45] that contribute to sexual differentiation of the brain [
30]. Sexually dimorphic features in microglia also exist in the adult brain, including differences in the number of microglia [
34] and expression of several inflammation-related factors [
45]. Further, more extensive transcriptome analyses have revealed significant sex differences in adult microglia that are at least partially independent of adult sex hormone exposure [
55]. As suggested by the data presented here, microglia are increasingly implicated as key regulators of sex differences in several neurological disorders including AD [
15,
43]. There are numerous sex differences in AD risk, pathogenesis, and clinical manifestation that may reflect sexually dimorphic factors in both development and adulthood [
39]. We speculate that these established sex differences, alongside the novel observations presented in this study, include significant contributions of, but are not limited to, microglial actions.
Collectively, these findings suggest that the AD risk factors
APOE genotype and female sex may affect development of AD, in part, by modulating protective microglial functions. An intriguing literature defines both beneficial and deleterious actions of microglia in the context of AD [
8,
17]. Classically, microglia have been viewed as contributors to AD pathogenesis, largely as a consequence of chronic neuroinflammation that is associated with states of microglial activation [
17]. Both activation of microglia and indices of neuroinflammation are regulated, individually and sometimes cooperatively, by
APOE genotype and sex [
51]. In addition to driving disease progression, microglia are also able to combat AD pathogenesis, primarily via plaque interactions that can decrease plaque size and reduce levels of dystrophic neurites presumably by limiting exposure to highly toxic Aβ species [
4,
9,
21,
25,
27,
52,
61,
63,
64]. The present data demonstrate that these protective microglial actions with plaques are attenuated in
APOE4 carriers and females. Thus,
APOE genotype and sex affect both harmful and protective microglial actions. Still unclear is how the balance, or loss thereof, in the heterogeneity in microglial functions and their regulation within plaque environments varies across the disease process.