Background
The role of innate immunity in the pathogenesis of neurodegenerative diseases remains uncertain despite intense, sustained investigation. Recent genetic, genomic and animal model studies have highlighted the importance of several innate immune molecules in the progression of many major neurodegenerative disorders, including Alzheimer’s Disease (AD), the most prevalent neurodegenerative disease worldwide [
1‐
3]. Of particular note, in late 2012, two independent groups identified coding alterations in the
T
riggering
R
eceptor
E
xpressed on
M
yeloid Cells-
2 (
TREM2) gene that dramatically increased the risk of developing late-onset Alzheimer’s Disease (LOAD) [
4,
5].
TREM2 is a single-pass transmembrane glycoprotein that is expressed on myeloid lineage cells, and modulates innate immune function [
6]. In the CNS, TREM2 mRNA is expressed by microglia, although the protein is not readily detected either by immunohistochemistry nor flow cytometry [
7‐
12]. Recessive, loss-of-function mutations in
TREM2 are the genetic basis of Nasu-Hakola disease, characterized by abnormal bone cysts and early onset dementia [
13‐
16]. TREM2 coding variants have also been associated with increased risk for amyotrophic lateral sclerosis (ALS) and frontotemporal-lobar dementia (FTLD) [
17‐
19]. Taken together, these data suggest that TREM2 likely plays a critical role in regulating myeloid cell function within the brain, thereby modulating risk for neurodegenerative diseases. Despite this abundance of genetic validation, the pathways linking TREM2 to neurodegenerative disease have yet to be identified and strategies for therapeutic modulation are not evident.
Recent studies examined the role of TREM2 in regulating disease pathogenesis in mouse models that develop the extracellular deposits of β-amyloid (Aβ) characteristic of AD. Our studies revealed high expression of TREM2 in macrophages associated with Aβ deposits and that TREM2 deficiency virtually abrogated accumulation of these macrophages accompanied by reduced levels of several cytokine mRNAs [
20,
21]. Furthermore, we reported that TREM2 deficiency leads to altered Aβ pathology that varies as a function of disease progression in APP-PS1 mice [
22]. Moreover, the plaque associated TREM2+ cells expressed markers indicating their derivation from peripheral macrophages. The dramatic decrease of plaque-associated macrophages was reported by all groups studying amyloid pathology in TREM2-deficient mice, although there were discordant observations with regard to the origin of TREM2+ cells as well as the effects of TREM2 deficiency on Aβ deposition [
23,
24] which have subsequently been resolved [
22].
Here we address the effect of TREM2 deficiency on another cardinal aspect of neurodegeneration, intracellular neuronal aggregates of hyperphosphorylated microtubule associated protein tau (MAPT; tau), using the hTau mouse model of tauopathy. These studies demonstrate that TREM2 deficiency leads to exacerbated and earlier onset of tau phosphorylation and aggregation coupled with a profound dysregulation of stress related kinase pathways. These results suggest that myeloid cell TREM2 plays a protective role in regulating or suppressing neuronal tau pathologies and potentially implicates TREM2 signaling pathways within discrete myeloid cell populations in linking Aβ and tau pathologies.
Methods
Animal models
hTau
+/−;mTau
−/− were acquired from the Jackson Laboratory which express all six isoforms of the human tau protein under the control of the endogenous human
MAPT promoter, and backcrossed into
Trem2
−/−
mice [
20] to generate hTau
+/−;mTau
−/−:
Trem2
−/−
and hTau
+/−;mTau
−/− control mice. These mice were backcrossed for 4 generations, and maintained on a B6 background. All experiments were preformed and repeated using multiple cohorts of mice, and each individual assay was performed in duplicate. C57BL6/j mice (B6) obtained from the Jackson Laboratories were used side-by-side with
Trem2
−/−
for control experiments.
Mice were housed in the Cleveland Clinic Biological Resources Unit and the Jackson Laboratory, facilities fully accredited by the Association and Accreditation of Laboratory Animal Care. All experimental procedures were approved by the Institutional Animal Care and use Committee at each respective institution.
Western blotting
Mice were deeply anesthetized and perfused with ice-cold PBS, and their brains were removed, snap frozen, and stored at −80°C until use. Cortices and hippocampi were microdissected and homogenized in ice-cold T-per homogenization buffer with added protein phosphatase inhibitor and protease inhibitor cocktails (1:100) for 60 s. Homogenates were sonicated, centrifuged, and total protein concentrations assessed in tissue samples using the Bicinchoninic Acid (BCA) assay, following manufacturer’s instructions (Thermo Fisher, Cat# 23225). Absorbance values were read at 562 nm on a SpectraMax 340 PC plate reader (Molecular Devices, Sunnyvale, CA) using SoftMax Pro 5.2 analytical software. Proteins were denatured at 95°C for 15 min in 35% denaturing buffer containing LDS sample buffer (Life Technologies) and reducing agent (Life Technologies) 30–60 μg total protein were loaded along with 5 μg Magic Mark XP protein ladder (Life Technologies) onto Novex 4–12% Bis-Tris gels (Life Technologies), run at 160 V for 30–60 min, and transferred onto PVDF membranes (EMD Millipore) in 1XTAE buffer at 100 mA overnight at room temperature. After transfer, membranes were blocked with Odyssey Blocking Buffer in PBS (LI-COR Biosciences) for 1 h at room temperature and incubated in the appropriate primary antibodies in blocking buffer with added 0.1% Tween 20 overnight at 4 °C with shaking AT8 (1:5000; Pierce), AT180 (1:2000; Pierce), Tau5 (Thermo Scientific), PHF-1 (1:10,000; a generous gift from Dr. Peter Davies), MC1 (1:1000; a generous gift from Dr. Peter Davies), JNK (1:5000) and pJNK (Thr183/Tyr185; 1:1000) (Cell Signaling Technologies), ERK1/2 (1:5000) and pERK1/2 (Thr202/Tyr204; 1:2000) (Cell Signaling Technologies), GSK3β (1:5000) and pGSK3β (Ser9;1:2000) (Tyr216 1:2000) (Cell Signaling Technologies; ABCAM), p38 MAPK (1:2000) and pP38 MAPK (1:1000) (Cell Signaling Technologies), GAPDH (1:10,000; EMD Millipore) and β-Actin (1:10,000; Thermo Scientific). Membranes were washed with PBST (0.1% Tween 20), incubated for 1 h with the appropriate IR-conjugated secondary antibody, and visualized using an Odyssey IR scanner (LI-COR Biosciences) system. ImageStudio software (LI-COR Biosciences) was used for densitometric analysis and each experimental sample was normalized to total tau (Tau5), GAPDH or β-Actin.
Remaining detergent insoluble pellets prepared for Western Blotting were weighed and homogenized in 10 vol buffer H for 60 s followed by 15 s sonication at 20% amplitude, as performed in (Bhaskar, Korneth et al. 2010). Centrifuged sample supernatants were adjusted to 1% with 10% v/v sarkosyl solution with 1% β-ME. Sarkosyl solutions were incubated for 2 h at 37°C with agitation. Samples were then ultra-centrifuged at 150,000 g for 35 min. Sarkosyl soluble and insoluble fractions were isolated for analysis. Following BCA assay, 60 μg total protein from each fraction were denatured and reduced, run onto gels, transferred, and visualized according to our Western Blotting protocol. Micrographs contain 2 representative samples per genotype, per time-point.
Immunohistochemistry
30 μm free floating tissue sections were prepared for standard immunohistochemistry or immunofluorescence. Standard antigen retrieval via boiling at 95 °C for 10 min in sodium citrate buffer (10 mM) was used before probing with the appropriate primary antibody, AT8, AT180, Iba1, NeuN, CD45, F4/80 diluted at 1:500, or pJNK, pERK, and pGSK3β diluted at 1:100. Secondary antibodies (1:500) conjugated to either biotin (immunohistochemistry; Vector Labs) or Alexa Fluor dyes (for immunofluorescence), or incubated with Avidin: Biotinylated enzyme Complex (ABC; Vector Labs) reagent for 1 h at room temperature. Signals were revealed by developing sections in 3,3′-diaminobenzidine (DAB with/without Ni enhancement; Vector Labs) for immunohistochemistry, or visualization using fluorescent microscopy following mounting with Permount (Fisher Scientific) or DAPI with Hardset (Vector Labs). Quantitation of tau positive cells was performed on a total of 2 medial tissue sections from 4 mice per genotype. Representative sections 200 μm × 2 mm from the apex of the frontal cortex to the white matter were utilized to count Individual p-Tau+ cells in layers II-III as well as IV-VI. % area quantifications were performed using 2mm2 regions of interest across lamina II-III within the medial frontal cortex. Regions were thresholded and quantified using ImageJ particle counter. Data are represented as mean +/− standard error. For microglial morphology analysis, a total of 20–30 microglia were selected randomly from frontal cortex and CA3 region of the hippocampus from each genotype. Individual cells were thresholded and analyzed for soma area and total cell area. Thresholded cells were then skeletonized and analyzed using ImageJ AnalyzeSkeleton plugin which outputs total number of branches along with branch junctions. 2-way ANOVA with Sidek multiple comparison correction was utilized to compare differences between genotypes, and among brain region.
Quantitative RT-PCR (qRT-PCR)
Mice were perfused with PBS, and their brains were removed, snap frozen, and stored at −80 °C until use. Tissue was homogenized in 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS, and 1:100 protease inhibitor cocktail in PBS. RNA was isolated using chloroform extraction and was purified using Purelink (Life Technologies). cDNA was prepared from 1.5 μg using a QuantiTect Reverse Transcription kit (QIAGEN), and real-time PCR was performed for 40 cycles with the StepOne Plus real Time PCR system (Life Technologies). All primers and TaqMan probe were purchased from the Life Technologies database. Relative gene expression was determined using the ∆∆CT method. ANOVA with Bonferroni or Sidak multiple comparison adjustment were used to compare B6 and hTau;Trem2
+/+
mice and 6 and 12-months.
Cytokine multiplexing
Tissue was extracted and processed as described above for western blotting. Cytokine levels were all normalized to total protein concentration following BCA assay. Multiplex assays were conducted according to manufacturer’s instructions using reagents provided with the kit (Invitrogen Mouse 20-plex Cytokine Panel, Cat# LMC0006M). Following sample incubation, plates were washed with a mild detergent solution using a magnetic 96-well separator (Invitrogen Cat# A14179) three times (1 min per wash). After 30-min streptavidin-rpe incubation, plates were again washed three times followed by a final addition of 125 μl wash solution to all wells. Plates were read on a Luminex Magpix unit (Life Technologies) and initial analyses were performed by Xponent software and results exported into Microsoft Excel for further processing. Sample size was set to 50 μl and minimum count was set to 100 events/bead region.
Statistical analysis
Data are presented as mean ± SEM unless otherwise noted. Comparisons between two groups were analyzed using Student’s t test (two-tailed; unpaired) at 95% confidence interval. Multiple group comparison, or multiple comparisons were analyzed using ANOVA or 2-way ANOVA followed by Bonferroni or Sidak post hoc test. Analysis was performed using Prism GraphPad or SPSS software. Significance was determined at P < 0.05 *, P < 0.01 **, and P < 0.001 ***. Each experiment was performed in duplicate, validated, and experimental mice were derived from multiple cohorts. Quantification was performed with the researcher blinded to experimental group and genotype.
Discussion
The current study demonstrates that TREM2 deficiency leads to worsened tau pathology, altered microglial reactivity, and robust signaling abnormalities in the hTau mouse model of tauopathy. These findings have direct implications for the role(s) of TREM2 in the clearance and/or propagation of tau pathology, in regulating Aβ and tau pathologies in opposing manners, in cell or activation specific (macrophage versus microglia) effects on AD pathologies and more broadly for neurodegeneration and AD as outlined below.
Our group and others previously characterized the effects of TREM2 deficiency in mouse models of AD that develop robust Aβ pathology. These studies consistently demonstrated that TREM2 deficiency leads to the loss of the majority of Aβ plaque associated macrophages throughout the diseased parenchyma. In addition to the altered myeloid cell accumulation, these studies demonstrated, an overall reduction in inflammatory markers, reduced reactive astrocytes and a reduction in tau phosphorylation within dystrophic neurites surrounding Aβ deposits. We also demonstrated TREM2 deficiency modestly reduces hippocampal Aβ pathology at 4 months of age. However, analysis of mice 8 months of age revealed an significant enhancement of amyloid burden in the TREM2 deficient mice in both the APPPS1 and 5XFAD mouse models of Aβ pathology [
20,
24,
28]. However, neither the APPPS1 or 5XFAD mouse models of AD develop robust tau phosphorylation and aggregation (besides initial tau phosphorylation within the dystrophic neurites surrounding Aβ deposits). By contrast, the current study demonstrates that TREM2 deficiency results in robustly enhanced intraneuronal tau phosphorylation, aggregation, and kinase dysregulation in the hTau mouse model of tauopathy, suggesting opposing roles in modulating the two cardinal hallmarks of AD pathology.
We have demonstrated that TREM2 deficient microglia exhibit an altered morphology and surface expression profile in response to pathological human tau, which is consistent with other studies using various disease models. Here, we demonstrate a novel mechanism that exacerbates tau pathology independent of the up-regulation of classic pro-inflammatory mediators.
Our studies suggest that TREM2 plays a role in the regulation of JNK, GSK3β and ERK signaling and subsequent downstream tau hyperphosphorylation and aggregation. Taken together with our data demonstrating that reactive microglia localize to regions heavily laden with hyperphosphorylated tau suggests that TREM2 deficient microglia could fail to mount an efficient immune response in reaction to pathological tau accumulation, including potential deficiencies in clearance or phagocytosis of toxic tau, which is consistent with previous studies from the Landreth lab [
12]. Hyperphosphorylation of GSK3β at both the activating Y216 and inactivating S9 epitopes may indicate an attempt by neurons to halt over activation of signaling, as GSK3β(S9) phosphorylation acts to negatively regulate GSK3β activity. This phenomenon could also be due to over-activity of additional upstream kinases that affect multiple signaling molecules at various phosphorylation sites. Additional studies are required to examine the nature of the signal that promotes kinase dysregulation in TREM2 deficient mice, including the timing, cell autonomous and cell non-autonomous mechanisms, and longer-term consequences with aging. Additionally, in-vivo pharmacological blocking agents against JNK, ERK, and GSK will be necessary to further elucidate the exact mechanism by which tau is being phosphorylated, but these studies are beyond the scope of the current report.
Several recent studies have demonstrated the spread of pathological tau into neighboring brain regions throughout the course of pathology via mechanisms which are currently poorly understood. Intriguingly, recent studies by Ikezu and colleagues demonstrated that ablation of microglia or inhibition of exosomal trafficking reduces tau pathology, suggesting that the observed spread of tau may potentially be mediated by microglia [
32]. The observed hippocampal signaling abnormalities detected in 3-month hTau;
Trem2
−/− mice in combination with the localization of pathology in older 6-month mice leads us to posit that pathological tau and/ or signaling dysregulation spreads from the hippocampus into neighboring cortical regions, where the most prominent pathology is localized in hTau;
Trem2
−/−
mice. The failure to effectively clear or process tau species from diseased neurons or the extracellular environment via normal clearance mechanisms could lead to further downstream kinase activation resulting in the eventual seeding and spread of pathological tau species from cell to cell and into different brain regions, giving explanation to our results which demonstrate soluble and insoluble tau pathology is worse in the cortex. These ideas are further supported by recent in-vitro studies which demonstrated that increased neuronal stress kinase signaling leads to increased presence of extracellular tau, although the exact mechanisms remain unclear [
33]. However, future studies will be required to examine whether TREM2 deficiency promotes tau pathology via altered clearance and pathological spread of tau, direct microglial-neuronal contact and/or other possible mechanisms.
Taken together with our previous findings, the current study suggests that TREM2 deficiency has opposing roles on Aβ and tau pathologies, with an amelioration of Aβ-associated pathologies and yet an exacerbation of tau pathology. Several other studies of specific immune pathways have also demonstrated a similar opposing relationship on the development of Aβ and tau pathologies including the neuronal-specific chemokine, fractalkine (CX3CL1) and its cognate microglial receptor (CX3CR1) as well as the cytokine interleukin 1β (IL1β), amongst others [
34,
35]. This accumulating evidence suggests that TREM2 and other innate immune pathways may play detrimental roles in early stages of human AD, when Aβ predominates (estimated to be as many as 10–20 years prior to clinical onset), but beneficial roles at later stages of human AD, when tau pathology develops, that has direct implications for the development of therapeutic strategies targeting these pathways. Furthermore, these results also suggest that innate immune pathways could potentially mechanistically link the two primary pathological hallmarks of AD, albeit in opposing manners. There are several different experimental models via which this might occur, including frustrated phagocytosis, altered clearance/propagation of extracellular/intracellular protein aggregates by myeloid cells and alternative roles of brain resident microglia and macrophages versus peripherally derived macrophages in regulating AD pathologies, which will need to be examined in future studies.
Acknowledgements
We would like to acknowledge the contributions of the Cleveland Clinic Lerner Research Institute Flow Cytometry and Imaging Cores for their assistance in gathering and analyzing data. This work was supported by a BrightFocus Foundation grant (to B.T.L.), a National Research Service Award F31 # 1F31AG050409-01A1 and Stark Neurosciences Research Institute and by the Indiana Clinical and Translational Sciences Institute, funded in part by grant # UL1TR001108 from the National Institutes of Health, National Center for Advancing Translational Sciences. (to S.M.B.) NIH Multi-Center Program Grant (MCPG; to G.E.L., R.M.R., and B.T.L.) and an Investigator Initiated Research Grant (to B.T.L.) from the Alzheimer’s Association, National Institute of Aging grants R01 AG023012 (to B.T.L.) and RF1 AG051495 (to B.T.L. and G.E.L.), National Institute of Neurological Disorders and Stroke grant R01 NS047804 (to B.T.L.),, National Eye Institute grant R01 EY022358 (to S.D.C.), Department of Defense grant W81XWH-13-MRPRA-CSRA (to B.T.L.), the Jane and Lee Seidman Fund and a generous donation from Chet and Jane Scholtz.