Selective removal of astrocytic APOE4 strongly protects against tau-mediated neurodegeneration and decreases synaptic phagocytosis by microglia

In the paper

“Numerous human and rodent model studies have highlighted the important role of astrocytes in AD and tau pathology, with astrocyte activation occurring prior to large scale NFT formation, synaptic dysfunction, and neuronal loss (Bussian et al., 2018; Shi et al., 2017; Yoshiyama et al., 2007).”

What the sources say

• Partly supported

  In the main text (page unavailable), the paper reports senescent astrocytes in PS19 mice and finds that a senescence marker increases before NFT deposition. It supports a role for astrocytes in this mouse model of tau pathology, but does not establish that astrocyte activation precedes synaptic dysfunction or neuronal loss, or provide human-study evidence for that sequence. (LLM, conf=0.86)

  cites Bussian et al., 2018 sourceExperimental — accuracy not yet measured
• Partly supported

  The paper’s astrocyte discussion (page not provided) reports reactive astrocyte markers in 9-month-old TE4 mice and links them to neurodegeneration, supporting an important role in tau pathology. It does not establish that astrocyte activation precedes large-scale neurofibrillary tangle formation, synaptic dysfunction, or neuronal loss; it reports no activation of those astrocytic genes in 3-month-old TE4 mice. (LLM, conf=0.88)

  cites Shi et al., 2017 sourceExperimental — accuracy not yet measured
• Disputed

  The Results section reports weak GFAP staining at 3 months but markedly increased staining at 6 months, when astrogliosis closely corresponded to NFT distribution and density. The paper’s Summary reports synapse loss and impaired synaptic function by 3 months and identifies microglial—not astrocyte—activation as preceding tangles. Thus, this source does not support the claimed sequence in which astrocyte activation precedes NFTs and synaptic dysfunction, although it places substantial neuron loss later. (LLM, conf=0.88)

  Checked by a reasoning model — may be wrong.

  cites Yoshiyama et al., 2007 sourceExperimental — accuracy not yet measured

In the paper

“Specific marker genes for cluster 6 included RNA binding proteins (RBPs) Arpp21 and R3hdm1 (Figure S4, S5A, and Table S3) and several genes (Rorb, Cux1, Cux2, and Mef2c) that are typically repressed in hippocampal pyramidal neurons by Zbtb20 and are normally expressed within cortical neurons (Figure S4, S5B, and Table S3) (Nielsen et al., 2014; Rehfeld et al., 2018).”

What the sources say

• ?Not checked — no full text of the sourcecites Nielsen et al., 2014 source
• ?Could not verify

  The Results section on developmental regulation of the miR-128 host genes discusses Arpp21 and R3hdm1 expression, but does not address repression by Zbtb20. Searches for Rorb, Cux1, Cux2, Mef2c, Zbtb20, and the stated hippocampal-versus-cortical expression pattern found no passage supporting the quoted claim. (LLM, conf=0.78)

  cites Rehfeld et al., 2018 sourceExperimental — accuracy not yet measured

In the paper

“Recently, we reported that microglial engulfment of synapses in P301S mice was reduced in mice that expressed AD-associated TREM2 variants (Gratuze et al., 2020).”

What the sources say

• Partly supported

  The Methods (p. 11) identify PS19 mice as expressing P301S tau, and the Discussion (p. 10) reports reduced engulfment of postsynaptic puncta in mice expressing the AD-associated TREM2 R47H variant. The claim's plural “variants” overstates the mouse finding: the paper tested R47H in mice, while its R62H engulfment result was in human AD tissue (Figure 7, p. 9). (LLM, conf=0.86)

  cites Gratuze et al., 2020 sourceExperimental — accuracy not yet measured

In the paper

“Specifically, the expression of apoE4 resulted in a significant increase in the loss of neurons, brain volume, and glial inflammatory responses compared to other apoE isoforms, while the absence of apoE throughout life alleviated neurodegeneration and inflammation.”

What the sources say

• Supported

  The paper’s account of the P301S tauopathy mouse experiments reports significantly greater brain atrophy and neuronal-layer thinning with ApoE4 than with ApoE2 or ApoE3. It also reports elevated inflammatory glial responses in ApoE4 mice, while mice lacking ApoE were largely protected from neuronal loss, atrophy, and inflammation. Our copy provides section labels but no page numbers. (LLM, conf=0.93)

  cites Shi et al., 2017 source

In the paper

“Although the mechanisms are not fully elucidated, apoE appears to influence AD at least in part by promoting amyloid-β (Aβ) aggregation and reducing Aβ clearance (Castellano et al., 2011; Huynh et al., 2017; Liu et al., 2017).”

What the sources say

• Partly supported

  The Discussion reports that apoE isoforms differentially regulate Aβ accumulation through effects on clearance, supporting the clearance part of the claim. It does not establish that apoE promotes aggregation: the authors say they cannot exclude that possibility, but note conflicting in vitro findings, including reports that apoE inhibits aggregation. The claim presents both mechanisms more definitively than this source supports. (LLM, conf=0.88)

  cites Castellano et al., 2011 sourceExperimental — accuracy not yet measured
• Supported

  The Introduction states that apoE influences AD pathology by promoting Aβ aggregation and impairing its clearance from brain interstitial fluid. The Discussion also describes effects on soluble Aβ clearance and aggregation. Page numbers are not provided in our copy. (LLM, conf=0.94)

  cites Huynh et al., 2017 source
• Partly supported

  The Discussion reports that astrocytic apoE4 impairs Aβ clearance and proposes that apoE4 enhances Aβ aggregation during plaque seeding. The citing sentence says “apoE” generally, while the paper’s evidence and proposed mechanism concern apoE4 specifically, not all apoE isoforms. (LLM, conf=0.84)

  cites Liu et al., 2017 sourceExperimental — accuracy not yet measured

In the paper

“The amount of astrogliosis is also highly correlated with mouse brain volume in a mouse model of tauopathy (Shi et al., 2017).”

What the sources say

• Supported

  The paper reports that GFAP signal, a measure of reactive astrocytes, was elevated in 9-month-old TE4 mice and that the amount of GFAP immunoreactivity was highly correlated with brain volume (passage [1]). This directly supports the claim for the mouse model of tauopathy. (LLM, conf=0.94)

  cites Shi et al., 2017 source

In the paper

“Aquaporin-4 (Aqp4) is upregulated in the rTg4510 model of tauopathy, but exhibits reduced polarization towards astrocytic endfeet in reactive astrocytes in response to tau pathology (Harrison et al., 2020).”

What the sources say

• Partly supported

  The Results section (Fig. 5) reports AQP4 upregulation in rTg4510 mice and reduced polarization and endfoot vessel coverage in the caudal cortex. The Discussion links these findings to reactive astrogliosis associated with tau pathology, but treats that explanation as likely rather than establishing a causal response. Reduced polarization was demonstrated in a specific region, not throughout the model. (LLM, conf=0.83)

  cites Harrison et al., 2020 sourceExperimental — accuracy not yet measured

In the paper

“ApoE also modulates microglial responses to amyloid plaque pathology (Sala Frigerio et al., 2019; Ulrich et al., 2018).”

What the sources say

• Supported

  The Discussion states that Apoe deletion reduces the number of microglia with an ARMs response signature and affects their interaction with Aβ plaques. The Results section, “Deletion of Apoe Suppresses the Microglial Response to β-amyloid,” likewise reports an impaired ARMs response in Apoe-null mice. The paper therefore supports the claim that ApoE modulates microglial responses to amyloid plaque pathology. (LLM, conf=0.94)

  cites Sala Frigerio et al., 2019 source
• Supported

  The Introduction reports reduced plaque-associated microgliosis and activated microglial gene expression in Apoe-deficient mice, and concludes that apoE facilitates plaque-associated microgliosis. The Final comments and conclusion likewise state that apoE affects the innate immune response to amyloid plaques. (LLM, conf=0.94)

  cites Ulrich et al., 2018 source

In the paper

“A similar C4b and H2-D1-expressing oligodendrocyte population was recently described in mouse models of Aβ pathology (Nugent et al., 2020; Zhou et al., 2020).”

What the sources say

• ?Could not verify

  The Results discuss microglial responses to cuprizone-induced demyelination and compare them with microglia in 5XFAD mice. Searches for C4b, H2-D1, and oligodendrocytes found no passage describing the claimed oligodendrocyte population in a mouse model of Aβ pathology. (LLM, conf=0.78)

  cites Nugent et al., 2020 sourceExperimental — accuracy not yet measured
• Supported

  The Results section, “snRNA-seq reveals an Aβ-dependent oligodendrocyte reactive signature,” reports significant upregulation of both C4b and H2-D1 in oligodendrocytes from 5XFAD mice, an Aβ-pathology model. It describes this as a progressive oligodendrocyte response to Aβ accumulation. (LLM, conf=0.94)

  cites Zhou et al., 2020 source

In the paper

“Microglial depletion from tauopathy mouse models strongly inhibits tau-induced neuronal loss and completely blocks tau-mediated brain atrophy (Mancuso et al., 2019; Shi et al., 2019).”

What the sources say

• Partly supported

  The Results (Figure 6; page not provided) report greater preservation of spinal motor neurons in JNJ-527-treated P301S mice, supporting the neuronal-loss part of the claim. The paper describes reducing microglial numbers by inhibiting proliferation, not complete depletion, and our copy does not report that brain atrophy was completely blocked (searched: brain atrophy). (LLM, conf=0.88)

  cites Mancuso et al., 2019 sourceExperimental — accuracy not yet measured
• Supported

  The Results section ('Microglial ablation during a critical time window,' page unavailable) reports that microglial depletion in TE4 tauopathy mice virtually fully blocked neurodegeneration and preserved brain volume at the level of non-tau mice. The Discussion (page unavailable) likewise says depletion fully blocked the development of brain atrophy in this model. (LLM, conf=0.91)

  cites Shi et al., 2019 source

In the paper

“Human tau positron emission tomography (PET) imaging studies found that tau PET burden is highly correlated with cognitive deficits and brain atrophy with APOE-ε4 increasing these associations (Therriault et al., 2020; Weigand et al., 2020).”

What the sources say

• Partly supported

  The Discussion (page unavailable) reports that medial temporal tau pathology is associated with poorer memory function and lower medial temporal gray matter volume. The Results (page unavailable) show greater medial temporal tau-PET uptake in APOE-ε4 carriers, but do not show that APOE-ε4 strengthens tau’s associations with cognitive deficits or atrophy. (LLM, conf=0.86)

  cites Therriault et al., 2020 sourceExperimental — accuracy not yet measured
• Partly supported

  The Results section ('APOE by tau interaction on cognition'; page not provided) reports poorer cognitive performance with higher medial temporal lobe tau PET burden and a stronger tau–memory association in APOE-ε4 carriers. It reports no APOE-ε4 moderation for language or attention/executive performance. The cited passages do not establish an association between tau PET burden and brain atrophy, much less that APOE-ε4 increases it. (LLM, conf=0.88)

  cites Weigand et al., 2020 sourceExperimental — accuracy not yet measured

In the paper

“In a recent study, we found that astrocyte-derived apoE particles are significantly larger in size and are thus contain much more lipid compared to apoE secreted by microglia (Huynh et al., 2019), foreshadowing a differential role and function of apoE derived from these two cell types.”

What the sources say

• Partly supported

  The Results section, “Qualitative assessment of microglia and astrocyte-derived apoE particles,” reports that microglial particles were generally smaller and infers that they were less lipidated than astrocyte-derived particles. The analysis was qualitative: it does not establish statistical significance for the size difference or measure how much more lipid the astrocyte-derived particles contained. (LLM, conf=0.85)

  cites Huynh et al., 2019 sourceExperimental — accuracy not yet measured

In the paper

“Different components of complement, including C1q and C3a receptor also play a role in mediating tau-induced injury (Litvinchuk et al., 2018; Mancuso et al., 2019; Shi et al., 2019; Shi et al., 2017; Wu et al., 2019).”

What the sources say

• Supported

  The Introduction and Results identify C3aR as a mediator of tau pathology: deleting C3ar1 substantially reduced pathological tau in PS19 mice. This supports the C3aR component of the multi-citation sentence; the sentence does not assign its C1q component specifically to Litvinchuk et al. Page numbers are unavailable in our copy. (LLM, conf=0.91)

  cites Litvinchuk et al., 2018 source
• ?Could not verify

  The Discussion describes activated microglia and inflammatory cytokines as possible mediators of tau-related neurodegeneration in P301S mice, but it does not identify complement as the mediator. Searches for C1q, the C3a receptor, and complement activation found no passage addressing their role in tau-induced injury. The indexed passages have section labels but no page numbers. (LLM, conf=0.72)

  cites Mancuso et al., 2019 sourceExperimental — accuracy not yet measured
• Partly supported

  The Introduction (page unavailable) names complement-mediated neuronal damage as a possible mechanism of neurodegeneration, and the Discussion links microglial activation to neurodegeneration in a tauopathy mouse model. The paper does not identify C1q or the C3a receptor as mediators of tau-induced injury; searches for both terms found no relevant passage. (LLM, conf=0.82)

  cites Shi et al., 2019 sourceExperimental — accuracy not yet measured

In the paper

“Accumulating evidence has shown that RBPs can potentially co-localize with tau protein and promote its aggregation (Wolozin and Ivanov, 2019).”

What the sources say

• Supported

  In “Stress granules and tau aggregation” (page not provided), the paper states that tau co-localizes with TIA1 and other stress-granule-associated proteins, and that tau’s association with stress granules stimulates insoluble tau aggregate formation. This supports the citing sentence’s qualified claim. (LLM, conf=0.94)

  cites Wolozin and Ivanov, 2019 source

In the paper

“Previous in vitro and in vivo studies suggested a role for apoE in mediating TREM2-dependent phagocytosis of apoptotic neurons (Atagi et al., 2015; Krasemann et al., 2017).”

What the sources say

• Partly supported

  The experimental-results passage under “ApoA1 and ApoB Also” reports that apoE increased phagocytosis of apoptotic N2a cells by cultured mouse microglia, with reduced uptake by Trem2-KO microglia. This supports the in vitro part of the claim. The Discussion describes an in vivo stroke-model finding concerning TREM2, but does not establish an in vivo role for apoE in phagocytosis of apoptotic neurons. (LLM, conf=0.86)

  cites Atagi et al., 2015 sourceExperimental — accuracy not yet measured
• Supported

  The Discussion (page not provided) states that APOE binds apoptotic neurons and increases TREM2-mediated phagocytosis, citing Atagi et al. The Results also report phagocytosis of injected dying neurons and TREM2-mediated APOE signaling in mouse models. Together, these support the citing sentence’s qualified account of the earlier studies. (LLM, conf=0.90)

  cites Krasemann et al., 2017 source

In the paper

“Dysfunction of oligodendrocytes, the myelinating cells of the CNS, may contribute to AD pathogenesis (Nugent et al., 2020; Zhou et al., 2020).”

What the sources say

• ?Could not verify

  The Introduction links microglial dysfunction—not oligodendrocyte dysfunction—to late-onset Alzheimer’s disease. The paper’s Summary discusses microglial responses to myelin debris, but the passages found when searching for oligodendrocyte dysfunction and Alzheimer’s disease do not address whether oligodendrocyte dysfunction contributes to AD pathogenesis. (LLM, conf=0.76)

  cites Nugent et al., 2020 sourceExperimental — accuracy not yet measured
• Supported

  The Results section reports reduced expression of myelination-related genes in human AD oligodendrocytes. The Discussion also describes a reactive oligodendrocyte signature in the 5XFAD mouse model and cites evidence suggesting a pathogenic role for associated molecules, supporting the claim's cautious 'may contribute' wording. (LLM, conf=0.86)

  cites Zhou et al., 2020 source

In the paper

“Interestingly, a recent study found that Rorb+ excitatory neurons exhibited selective vulnerability in AD (Leng et al., 2021).”

What the sources say

• Supported

  The Results section reports depletion of RORB+ excitatory neurons in the entorhinal cortex across AD stages and concludes that they are highly vulnerable in AD. The Discussion likewise identifies RORB as a marker of selectively vulnerable excitatory neuron subpopulations. (LLM, conf=0.94)

  cites Leng et al., 2021 source

In the paper

“Lower levels of pTau and reduced pTau deposition correlate with less neurodegeneration (Shi et al., 2017).”

What the sources say

• Partly supported

  The mouse-results passage (labeled Abstract in our copy; no page given) reports that p-tau staining patterns correlated with brain atrophy: type 1 accompanied the most preserved brains, and type 4 the greatest atrophy. It also reports greater p-tau-covered area in young TE4 mice, but does not establish that lower overall p-tau levels or reduced deposition correlate with less neurodegeneration; by nine months, differences in overall p-tau staining had decreased. (LLM, conf=0.84)

  cites Shi et al., 2017 sourceExperimental — accuracy not yet measured

In the paper

“In both AD and primary tauopathies, the accumulation of pathological forms of tau is closely linked with brain atrophy, neuronal and synaptic loss, and cognitive and functional decline linked to the location of these changes (Bejanin et al., 2017; La Joie et al., 2020; Petersen et al., 2019).”

What the sources say

• Partly supported

  The Introduction and Discussion (page numbers unavailable) link the regional distribution of tau pathology to cognitive deficits in Alzheimer’s disease and report that grey matter atrophy mediates part of that relationship. The Discussion also describes neuronal loss and synaptic changes, but the study did not establish these associations in primary tauopathies or assess functional decline. Those extensions are not supported by this paper. (LLM, conf=0.87)

  cites Bejanin et al., 2017 sourceExperimental — accuracy not yet measured
• Partly supported

  The Introduction links tau-PET topography with atrophy and the severity of specific clinical deficits in AD, and the Discussion reports that baseline tau-PET predicted where subsequent atrophy occurred. The paper does not establish these relationships in primary tauopathies or directly measure neuronal and synaptic loss. Its Discussion also says clinical decline was not correlated with baseline tau-PET, so it does not support a close link to functional decline in this cohort. (LLM, conf=0.88)

  cites La Joie et al., 2020 sourceExperimental — accuracy not yet measured
• Partly supported

  The Introduction links regional tau burden in AD to patterns of atrophy and domain-specific cognitive performance, and the Discussion links regional NFT accumulation to AD clinical manifestations. The paper does not establish the same relationship in primary tauopathies or link tau accumulation to synaptic loss; its statement about neuronal loss concerns its correlation with cognitive decline, not a demonstrated link to tau accumulation. (LLM, conf=0.86)

  cites Petersen et al., 2019 sourceExperimental — accuracy not yet measured

In the paper

“Microglia play a crucial role in tau pathology and tau-mediated neurodegeneration. Neuroinflammation is one of the hallmarks of tauopathy-linked neurodegeneration, and microglia are a major driving force for this biological process (Shi et al., 2017; Yoshiyama et al., 2007).”

What the sources say

• Partly supported

  In the supplied Abstract passages (page not provided), the paper reports microglial inflammatory activation in tauopathy-model mice and proposes that neuroinflammation may exacerbate neuronal death. It does not establish that microglia are a major driving force of tauopathy-linked neurodegeneration: it describes their harmful role as a possibility and also identifies astrocytes and direct effects of ApoE. (LLM, conf=0.84)

  cites Shi et al., 2017 sourceExperimental — accuracy not yet measured
• Supported

  The paper’s Summary reports prominent microglial activation before tangle formation and links neuroinflammation to early tauopathy progression in P301S mice. The Results describe microglia-mediated neuroinflammation as implicated in tauopathies, supporting the broader background claim. (LLM, conf=0.88)

  cites Yoshiyama et al., 2007 source

In the paper

“How tau and apoE together influence tau-mediated neurodegeneration is not clear, although recent data suggest that microglia are required (Mancuso et al., 2019; Shi et al., 2019), despite the main producers of apoE in the brain being astrocytes (Pitas et al., 1987).”

What the sources say

• Partly supported

  The Discussion (page unavailable) reports that inhibiting microglial proliferation in P301S mice prevented neurodegeneration and concludes that activated microglia play a deleterious role. This supports microglial involvement, but inhibiting proliferation does not establish that microglia are required for tau-mediated neurodegeneration generally. (LLM, conf=0.86)

  cites Mancuso et al., 2019 sourceExperimental — accuracy not yet measured
• Supported

  The Discussion reports that depleting microglia fully blocked brain atrophy in a tauopathy mouse model and concludes that microglial activation drives neurodegeneration in that setting. This supports the citing sentence’s hedged claim that recent data suggest microglia are required. (LLM, conf=0.92)

  cites Shi et al., 2019 source

In the paper

“The C3a receptor is expressed by microglia and C1q-mediated injury is blocked by antibodies to C1q, suggesting neuroinflammation-mediated injury likely involves a multicellular response, an important component of which is elevated microglial-mediated synaptic phagocytosis (Dejanovic et al., 2018).”

What the sources say

• ?Could not verify

  The Results passages describe C1q-dependent microglial synapse engulfment and its reduction by C1q-blocking antibodies, but do not address whether microglia express the C3a receptor. Searches for “C3a receptor” and “C3aR microglia” found no passage establishing that expression. (LLM, conf=0.72)

  cites Dejanovic et al., 2018 sourceExperimental — accuracy not yet measured

In the paper

“Tau-RBP interactions disrupt RNA splicing resulting in gene dysregulation and neurotoxicity (Apicco et al., 2018; Bai et al., 2013; Hsieh et al., 2019).”

What the sources say

• Partly supported

  The Abstract and Discussion report that tau oligomers associate with RBPs and that reducing TIA1 protects against tau-mediated neurodegeneration. The Discussion mentions splicing as a normal RBP function, but does not show that tau–RBP interactions disrupt splicing or thereby cause gene dysregulation and neurotoxicity. (LLM, conf=0.87)

  cites Apicco et al., 2018 sourceExperimental — accuracy not yet measured
• Partly supported

  The Biochemical Confirmation section (page unavailable) reports that U1-70K inclusions are closely associated with tau-positive tangles, but does not establish a tau–RBP interaction that disrupts splicing. The Deep RNA Sequencing and U1 snRNP Deficiency sections report splicing defects and altered APP expression in AD; they do not show that a tau–RBP interaction causes those changes or neurotoxicity. (LLM, conf=0.87)

  cites Bai et al., 2013 sourceExperimental — accuracy not yet measured
• Partly supported

  The Discussion sections report Tau-associated splicing errors, transcriptome disruption, and enhanced Tau toxicity when spliceosome components are reduced. They propose that spliceosome disruption and splicing errors may mediate Tau-induced neurotoxicity, but do not establish the cited sentence’s unqualified causal chain from Tau–RBP interactions through splicing and gene dysregulation to neurotoxicity. Page numbers are unavailable in our copy. (LLM, conf=0.86)

  cites Hsieh et al., 2019 sourceExperimental — accuracy not yet measured

In the paper

“Selective removal of APOE from microglia attenuated facial motor neuron death after axotomy, indicating that microglial apoE expression plays a key role in promoting the DAM/MGnD transcriptional state (Krasemann et al., 2017).”

What the sources say

• Supported

  In the Results section on APOE regulation of MGnD, the paper reports that conditional deletion of Apoe in microglia reduced neuronal loss after facial nerve axotomy and describes microglial Apoe as essential to the MGnD phenotype. The same section reports that Apoe deletion reversed disease-associated transcriptional changes in phagocytic microglia. No page number is available in our copy. (LLM, conf=0.94)

  cites Krasemann et al., 2017 source

In the paper

“Proteomic analysis of AD brain identified a glial metabolism correlated gene network that was significantly correlated with disease severity.”

What the sources say

• Partly supported

  The Results section, “Construction and Validation of a Consensus AD Protein Co-Expression Network,” reports that the M4 astrocyte/microglial metabolism module had the strongest correlations with AD traits, including cognition and pathology. The source identifies it as a protein co-expression module, not a gene network. (LLM, conf=0.86)

  cites Johnson et al., 2020 sourceExperimental — accuracy not yet measured

In the paper

“Microglia secrete apoE in smaller, lipid-poor lipoproteins (Huynh et al., 2019).”

What the sources say

• Supported

  The Results section (‘Qualitative assessment of microglia and astrocyte-derived apoE particles’) reports that particles in microglia-conditioned media were much smaller than astrocyte-derived particles. It concludes that microglia secrete poorly lipidated apoE relative to the larger lipoproteins secreted by astrocytes. (LLM, conf=0.93)

  cites Huynh et al., 2019 source

In the paper

“This subset of microglia was also enriched for several genetic risk factors for AD, including Trem2, and in the case of amyloid may exert a protective role in mitigating amyloid-induced damage (Sala Frigerio et al., 2019).”

What the sources say

• Partly supported

  The Discussion identifies amyloid-responsive microglia (ARMs) as enriched for AD risk genes, including Trem2. But the Results section on ARMs enriched for GWAS AD risk genes says further work is needed to determine whether the changes are protective or harmful. The paper therefore does not establish that this subset mitigates amyloid-induced damage, even as a proposed role. (LLM, conf=0.88)

  cites Sala Frigerio et al., 2019 sourceExperimental — accuracy not yet measured

In the paper

“In the context of CNS damage resulting from Aβ, tau, SOD-1, or other pathologies, apoE is strongly upregulated in microglia coincident with the DAM or MGnD phenotype (Butovsky et al., 2015; Keren-Shaul et al., 2017; Krasemann et al., 2017; Parhizkar et al., 2019; Shi et al., 2017).”

What the sources say

• Partly supported

  The Results and Discussion report that Apoe is strongly upregulated in spinal-cord microglia from SOD1 mice while their normal microglial signature is lost. The paper describes an M1-like phenotype, not a DAM or MGnD phenotype, so it does not establish that specific coincidence; it also does not establish the claim for Aβ or tau pathology. (LLM, conf=0.85)

  cites Butovsky et al., 2015 sourceExperimental — accuracy not yet measured
• Partly supported

  The Introduction and Results report Apoe upregulation during DAM activation in the 5XFAD Alzheimer’s model, and the Results identify a similar DAM-like population in SOD1 ALS mice. The cited passages do not establish apoE upregulation in those ALS microglia or address tau-associated damage or the MGnD phenotype. No page numbers are available in our copy. (LLM, conf=0.86)

  cites Keren-Shaul et al., 2017 sourceExperimental — accuracy not yet measured
• Partly supported

  The Results section (“Reciprocal induction of APOE and suppression of TGFβ signaling”) reports that Apoe was among the most upregulated genes in disease-associated MGnD microglia in APP-PS1 (Aβ), SOD1, and EAE mouse models. The cited paper does not establish the same finding for tau pathology; a search for “tau” found no relevant passage. Page numbers were not available in our copy. (LLM, conf=0.86)

  cites Krasemann et al., 2017 sourceExperimental — accuracy not yet measured

In the paper

“Several of the hub proteins in this glial metabolism network (Cd44, Prdx1, Ldhb, Pkm, Gapdh) were also significantly enriched in Astro2 (Johnson et al., 2020).”

What the sources say

• Partly supported

  The Results section on M4 proteins in cerebrospinal fluid identifies CD44 and PRDX1 as M4 hub proteins and LDHB and PKM as metabolic proteins in that module; the Discussion also identifies GAPDH as an elevated M4 protein. Our copy does not report that these proteins were significantly enriched in Astro2, and it does not identify all five as hub proteins. (LLM, conf=0.88)

  cites Johnson et al., 2020 sourceExperimental — accuracy not yet measured

In the paper

“In the CNS, apoE is only present in HDL-like particles, and the size and form of these particles are dependent on the lipid-transport protein ABCA1 (Hirsch-Reinshagen et al., 2004; Wahrle et al., 2004).”

What the sources say

• ?Not checked — no full text of the sourcecites Hirsch-Reinshagen et al., 2004 source
• Partly supported

  The ‘Plasma and CNS Apolipoprotein Levels’ section says that CSF contains only HDL-like lipoproteins and that apoE is its most abundant apoprotein; it does not establish that apoE throughout the CNS is only present in HDL-like particles. The same section reports that ABCA1 loss changes the size and lipidation of apoE-containing particles, supporting the second clause. Page numbers were not supplied in our copy. (LLM, conf=0.87)

  cites Wahrle et al., 2004 sourceExperimental — accuracy not yet measured

In the paper

“Selective removal of APOE from microglia attenuated facial motor neuron death after axotomy, indicating that microglial apoE expression plays a key role in promoting the DAM/MGnD transcriptional state (Krasemann et al., 2017).”

What the sources say

• Supported

  The Results section (“APOE regulates the transcriptional and post-transcriptional program of the MGnD phenotype and function”) reports that conditional deletion of Apoe in microglia reduced neuronal loss in the facial motor nucleus after axotomy. The same section reports changes in MGnD-associated gene expression after microglial Apoe deletion, supporting the citing sentence’s interpretation of its role in that state. (LLM, conf=0.94)

  cites Krasemann et al., 2017 source

In the paper

“This further supports a potentially dichotomous role for microglial responses, with Apoe and Trem2-dependent functions being protective in response to amyloid, but harmful in the presence of intracellular tau pathology (Gratuze et al., 2020; Shi et al., 2017; Ulrich et al., 2018).”

What the sources say

• Partly supported

  The Introduction (p. 2) describes TREM2 signaling as protective against amyloid-mediated injury, while the Discussion (p. 11) says TREM2 exacerbates neurodegeneration when tau pathology is prominent. These passages support the proposed dichotomy for TREM2, but our copy does not establish the claimed protective role for Apoe in response to amyloid. (LLM, conf=0.85)

  cites Gratuze et al., 2020 sourceExperimental — accuracy not yet measured
• Partly supported

  The discussion in the indexed copy (labeled Abstract; no page provided) says activated microglia may protect against amyloid pathology but could be harmful in tauopathy, and reports ApoE4-associated neuroinflammation in tau-bearing mice. The paper cites TREM2-related work on protective responses to amyloid, but does not establish that TREM2-dependent functions are harmful in tauopathy. (LLM, conf=0.88)

  cites Shi et al., 2017 sourceExperimental — accuracy not yet measured
• Partly supported

  The Results section (“Increased neuritic dystrophy in Apoe -/-;APP PS1 mice”; page unavailable) links Apoe deficiency to reduced plaque-associated microgliosis and increased neuritic damage. The Introduction also describes evidence that Trem2-dependent microglial responses may reduce amyloid-plaque toxicity. This paper does not establish that those responses are harmful in the presence of intracellular tau pathology, so it supports the amyloid half of the cited sentence but not the tau half. (LLM, conf=0.89)

  cites Ulrich et al., 2018 sourceExperimental — accuracy not yet measured

In the paper

“In addition, apoE, particularly in microglia, plays a key role in preventing accumulation of cholesterol ester species both normally and in the setting of demyelination in a TREM2-dependent fashion (Nugent et al., 2020).”

What the sources say

• Partly supported

  The Results (page unavailable) show that Apoe-deficient microglia accumulate cholesteryl esters on both control and demyelinating diets, supporting apoE’s protective role in those settings. The paper identifies Apoe as a TREM2-dependent damage-response gene, but does not establish that apoE’s protection under normal conditions depends on TREM2; it also finds accumulation in astrocytes, not just microglia. (LLM, conf=0.87)

  cites Nugent et al., 2020 sourceExperimental — accuracy not yet measured

In the paper

“The C3a receptor is expressed by microglia and C1q-mediated injury is blocked by antibodies to C1q, suggesting neuroinflammation-mediated injury likely involves a multicellular response, an important component of which is elevated microglial-mediated synaptic phagocytosis (Dejanovic et al., 2018).”

What the sources say

• Partly supported

  The Results report increased microglial engulfment of synapses in Tau-P301S mice and reduced engulfment after C1q-blocking antibody treatment; the Discussion describes this as a microglia–neuron interaction in tauopathy. These findings support the proposed synaptic-phagocytosis component, but the paper does not establish it for neuroinflammation-mediated injury generally. A search for C3a receptor expression did not find support for that additional premise in our copy; page numbers were unavailable. (LLM, conf=0.84)

  cites Dejanovic et al., 2018 sourceExperimental — accuracy not yet measured

In the paper

“In studies of microglial activation in amyloid models, MHC-II and Spp1 were elevated in an advanced activated state (activated response microglia (ARM)) that corresponded with high levels of Apoe expression (Krasemann et al., 2017; Sala Frigerio et al., 2019).”

What the sources say

• Partly supported

  The Results section on plaque-associated MGnD microglia reports increased Spp1 and Apoe in APP-PS1 mice, with Apoe rising across the described microglial subsets (page unavailable in our copy). It does not identify this state as activated response microglia (ARM) or report elevated MHC-II as part of that finding. (LLM, conf=0.88)

  cites Krasemann et al., 2017 sourceExperimental — accuracy not yet measured
• Supported

  In Results, “Microglia Differentiate into Mutually Exclusive Response States” (Figure 2C), MHC class II genes predominate in mid-to-late ARM cells, while Spp1 appears at the farthest end of that branch. Those farthest-end cells also have the highest Apoe levels and are described as an advanced microglial response. (LLM, conf=0.92)

  cites Sala Frigerio et al., 2019 source

In the paper

“In the CNS, apoE is only present in HDL-like particles, and the size and form of these particles are dependent on the lipid-transport protein ABCA1 (Hirsch-Reinshagen et al., 2004; Wahrle et al., 2004).”

What the sources say

• ?Not checked — no full text of the sourcecites Hirsch-Reinshagen et al., 2004 source
• Supported

  The Results section reports a different size distribution for apoE-containing particles in CSF from Abca1-knockout mice and markedly smaller particles from knockout astrocytes. The Discussion states that ABCA1 is required for normal lipidation of astrocyte-secreted apoE, supporting the claim that ABCA1 affects the size and form of CNS apoE particles. (LLM, conf=0.92)

  cites Wahrle et al., 2004 source

In the paper

“C4b+/H2–D1+ oligodendrocytes have also been observed in models of amyloid pathology (Nugent et al., 2020; Zhou et al., 2020).”

What the sources say

• ?Could not verify

  The Results passage discusses microglial responses to cuprizone-induced demyelination and compares them with reports on 5XFAD microglia. Searches for C4b, H2-D1, and oligodendrocytes in amyloid models found no passage in our copy reporting the claimed oligodendrocyte population. (LLM, conf=0.75)

  cites Nugent et al., 2020 sourceExperimental — accuracy not yet measured
• Supported

  The Results section ('snRNA-seq reveals an Aβ-dependent oligodendrocyte reactive signature') reports significant upregulation of both C4b and H2-D1 in oligodendrocytes from 5XFAD mice, a model of amyloid pathology. (LLM, conf=0.90)

  cites Zhou et al., 2020 source

In the paper

“The C3a receptor is expressed by microglia and C1q-mediated injury is blocked by antibodies to C1q, suggesting neuroinflammation-mediated injury likely involves a multicellular response, an important component of which is elevated microglial-mediated synaptic phagocytosis (Dejanovic et al., 2018).”

What the sources say

• Supported

  The Results report that C1q-blocking antibodies prevented microglial synapse loss in culture and reduced synapse engulfment in Tau-P301S mice, with a modest increase in synapse density five days after injection. This supports the claim for C1q-mediated synaptic injury. (LLM, conf=0.90)

  cites Dejanovic et al., 2018 source

In the paper

“Such phagocytosis can be stimulated by opsonins at damaged synapses, such as C1q (Dejanovic et al., 2018; Gratuze et al., 2020; Hong et al., 2016; Stevens et al., 2007).”

What the sources say

• Supported

  The Discussion reports C1q tagging of synapses in tauopathy and increased microglial engulfment of those synapses. The Results and Discussion also report that blocking C1q reduced synapse engulfment, supporting the claim that C1q at affected synapses can promote phagocytic removal. (LLM, conf=0.91)

  cites Dejanovic et al., 2018 source
• Supported

  The Discussion (p. 10) describes C1q-opsonized synaptic markers and says that lower C1q accumulation may compromise the signal for microglia to engulf synapses. The Results (pp. 5–6) report C1q at synapses in the tauopathy model and describe dystrophic synaptic features, supporting the claim in that context. (LLM, conf=0.88)

  cites Gratuze et al., 2020 source
• Supported

  The passage on synapse loss in pre-plaque Alzheimer mouse models (section labeled Abstract; page unavailable) reports that C1q deposits on vulnerable synapses and triggers a complement pathway involving phagocytic microglia. It also reports that inhibiting complement prevents Aβ-induced microglial synaptic engulfment, supporting the claim that synapse-associated C1q can promote phagocytosis. (LLM, conf=0.88)

  cites Hong et al., 2016 source

In the paper

“Genetic deletion of astrocytic APOE4 but not APOE3 was significantly neuroprotective in P301S Tau mice; however, it should be noted that crossing P301S Tau mice to Apoe KO mice or eliminating microglia resulted in no significant evidence of neurodegeneration/brain volume loss by 9 months of age relative to non-Tau transgenic mice (Shi et al., 2019; Shi et al., 2017).”

What the sources say

• Partly supported

  The Results section reports that microglial depletion virtually fully blocked neurodegeneration and left brain volume comparable to non-tau mice. It says Apoe-knockout P301S mice were largely protected, not that they had no neurodegeneration; the Discussion refers to mild neurodegeneration in those mice. Brains were examined at 9.5 months, rather than by 9 months (page numbers unavailable). (LLM, conf=0.86)

  cites Shi et al., 2019 sourceExperimental — accuracy not yet measured
• Partly supported

  The main text reports that ApoE knockout largely attenuated neuronal loss and brain atrophy in nine-month-old P301S mice. It does not establish the claimed lack of a significant difference from non-Tau mice: Methods > Animals says the authors did not statistically compare wild-type mice with the other groups. This paper also did not test elimination of microglia. (LLM, conf=0.88)

  cites Shi et al., 2017 sourceExperimental — accuracy not yet measured

In the paper

“There was a decrease in P2ry12 immunostaining (homeostatic gene) (Butovsky et al., 2014) in oil-treated TAFE3 and TAFE4 mice, which was significantly increased following TAM treatment (Figure 7E and 7F).”

In the paper

“Following removal of poor-quality nuclei, we obtained transcriptional data for 71,129 nuclei (Table S2) that we then subjected to unsupervised clustering by uniform manifold approximation and projection (UMAP) into 16 distinct clusters.”

In the paper

“Micro1 was enriched for the disease-associated microglia (DAM) (Keren-Shaul et al., 2017)/microglial neurodegenerative phenotype (MGnD) (Krasemann et al., 2017) and was dramatically expanded in TAFE3 and TAFE4 mice compared to FE3 and FE4 (Figure 7B).”

In the paper

“We also detected a strong upregulation in MHC-II and Spp1 gene expression in the presence of tau pathology, which was significantly attenuated by astrocytic APOE4 removal (Figure 7G–7I, S8C, and S8D).”

In the paper

“Cluster 10, containing astrocytic nuclei, was rescaled and re-clustered to identify three sub-clusters (Astro1–3) of astrocytes (Figure 6A).”

In the paper

“There was a trend towards increased microglial APOE expression after astrocytic APOE4 deletion in isolated cells (Figure S1A), which was greatly upregulated in P301S Tau mice.”

In the paper

“In TAM-treated female TAFE3 and TAFE4 mice, there was a significant reduction of insoluble pTau in the FA fraction and a trend towards less pTau in TAM-treated male TAFE4 mice.”

In the paper

“Compared to astrocytic APOE3, removal of astrocytic APOE4 had a qualitatively greater effect in altering the proportion of cells in clusters 1 and 3, indicating a stronger protective effect of astrocytic APOE4 removal in these neuronal clusters.”

In the paper

“One major oligodendrocyte cluster (cluster 2) was identified from the original dataset and further rescaled and re-clustered into two sub-clusters (Oligo1 and 2) (Figure 4, S7A and Table S5).”

In the paper

“This reduction was attenuated in TAM-treated animals (Figure 4B and 4C), suggesting a protective effect of astrocytic APOE removal on neurons.”

In the paper

“Single-nucleus RNA sequencing analysis revealed striking gene expression changes in all cell types with astrocytic APOE4 removal decreasing disease-associated gene signatures in neurons, oligodendrocytes, astrocytes, and microglia.”

In the paper

“In P301S mice, there was a strong shift towards oligodendrocytes expressing C4b, and H2–d1, as well as apparent upregulation of more broadly expressed oligodendrocyte genes such as Ptgds, Neat1, and Apod (Oligo2) (Figure S7A–S7C).”

In the paper

“Furthermore, neuronal cluster 6 was drastically expanded in P301S tau transgenic mice (Figure 4B and 4C), an expansion that was attenuated by tamoxifen treatment, suggesting a hippocampal neuronal gene signature associated with tau pathology.”

In the paper

“Deleting astrocyte APOE slightly increased the proportion of homeostatic astrocytes and reduced the proportion of activated astrocytes, consistent with a protective effect of APOE removal (Figure 6B).”

In the paper

“Interestingly, in TAM-treated mice, while only minimal neuronal Arpp21 staining was detected in TAFE4 mice (Figure 5F, second row), no noticeable reduction of neuronal Arpp21 staining was observed in TAFE3 mice (Figure 5F, fourth row).”

In the paper

“TAM treatment reduced AT8 immunoreactivity in female TAFE4, but not in male TAFE4 or in TAFE3 mice. (Figure 3G and 3H, Figure S3D–S3G).”

In the paper

“TAM-treatment significantly attenuated brain atrophy in both male and female TAFE4 mice, but not in TAFE3 mice (Figure 2A–2C).”

In the paper

“However, in TAM-treated mice, immunofluorescent staining of Rorb was co-localized with GFAP-positive astrocytes, but not with NeuN-positive neurons (Figure S5C, bottom row).”

In the paper

“Co-staining of C4b and Olig2 showed that while C4b expression was not detected in non-tau transgenic FE4 mouse brain tissue, oligodendrocyte expression of C4b was clearly observed in the presence of tauopathy.”

In the paper

“Based on the expression of well-known cell-type specific markers, these clusters were categorized into neurons (clusters 1, 3, 4, 6 – 9, 11, 12, 14, and 15), oligodendrocytes (cluster 2), microglia (cluster 5), astrocytes (cluster 10), oligodendrocyte progenitor cells (OPCs) (cluster 13), and choroid plexus epithelial cells (cluster 16) (Figure 4A, 4B, and Table S3).”

In the paper

“Removing astrocytic APOE4 markedly reduced tau-mediated neurodegeneration and decreased p-tau pathology.”

In the paper

“We observed a significant decrease in CD68 coverage in TAM-treated TAFE4 mice and TAM-treated TAFE3 female mice (Figure 8E and 8F).”

In the paper

“We verified upregulation of a subset of Astro2 genes (Gfap, Clu, Vim, Prdx1, and Prdx6) in hippocampal tissue by qPCR and found these genes were upregulated in the presence of tau pathology and significantly decreased after removing astrocytic APOE4 (Figure 6D–6F, S6C, and S6D).”

In the paper

“Reclustering of the microglia cluster (Cluster 5) yielded three sub-clusters (Micro1–3) (Figure 7A and Table S6).”

In the paper

“Following removal of poor-quality nuclei, we obtained transcriptional data for 71,129 nuclei (Table S2) that we then subjected to unsupervised clustering by uniform manifold approximation and projection (UMAP) into 16 distinct clusters.”

In the paper

“In contrast, staining in TAM-treated TAFE mice revealed that apoE co-localized with Iba1+ microglia, but not with GFAP+ astrocytes (Figure 1H and S1B, bottom row.”

In the paper

“Importantly, removal of astrocytic APOE4 decreased tau-induced synaptic loss and microglial phagocytosis of synaptic elements suggesting a key role for astrocytic apoE in synaptic degeneration.”

In the paper

“A third cluster (Micro3) was enriched for markers of peripheral macrophages, and while small, did not appear to vary in abundance with either tau pathology or astrocytic APOE expression (Figure 7B and S8A).”

In the paper

“We observed a significant decrease in CD68 coverage in TAM-treated TAFE4 mice and TAM-treated TAFE3 female mice (Figure 8E and 8F).”

In the paper

“While TAM treatment did not affect Aqp4 coverage in AFE4 mice (Figure S6F), we observed significantly decreased Aqp4 immunoreactivity in TAFE4 hippocampus after knocking down astrocytic APOE4 (Figure 6I and 6J) and more specific co-localization with the vasculature in TAM-treated TAFE4 mice (S6G), suggesting that increased Aqp4 expression and polarization was partially restored after astrocytic AP”

In the paper

“Gene set enrichment analysis (GSEA) for marker genes in cluster 6 found several significantly enriched pathways, including KEGG pathways for spliceosome and ubiquitin-mediated proteolysis and Hallmark pathways for DNA repair (Figure 5A).”

In the paper

“A third, small subcluster, Astro3, was enriched for genes associated with DNA transcriptional activity (Id1, Id3, and Mat2a) as well as Cdkn1a, a marker of senescence (Figure S6A and Table S4).”

In the paper

“Genetic deletion of astrocytic APOE4 but not APOE3 was significantly neuroprotective in P301S Tau mice; however, it should be noted that crossing P301S Tau mice to Apoe KO mice or eliminating microglia resulted in no significant evidence of neurodegeneration/brain volume loss by 9 months of age relative to non-Tau transgenic mice (Shi et al., 2019; Shi et al., 2017).”

In the paper

“We previously showed and again found that in the presence of apoE4 or apoE3, there was marked neurodegeneration and elevations in DAM/MGnD gene expression in P301S Tau transgenic mice which was greater with apoE4 than with apoE3.”

In the paper

“GSEA analysis identified enrichments in pathways pertaining to oxidative phosphorylation, suggestive of an increase in overall metabolic activity within astrocytes.”

In the paper

“We previously showed and again found that in the presence of apoE4 or apoE3, there was marked neurodegeneration and elevations in DAM/MGnD gene expression in P301S Tau transgenic mice which was greater with apoE4 than with apoE3.”

In the paper

“On the contrary, removal of astrocytic APOE, which exerted protective effects in P301S mice, reduced expression of genes associated with ARMs, such as Spp1.”

In the paper

“A similar enrichment in oxidative phosphorylation pathways was observed in the DAM/MGnD-like Micro1 cluster.”

In the paper

“The presence of tau pathology resulted in an expansion of a GFAP+ astrocyte population (Astro2), which appeared to be attenuated by removal of astrocytic APOE.”

In the paper

“Expression of C4b potentially activates the complement pathway, which may induce prolonged inflammation throughout the course of tauopathy.”

In the paper

“How tau and apoE together influence tau-mediated neurodegeneration is not clear, although recent data suggest that microglia are required (Mancuso et al., 2019; Shi et al., 2019), despite the main producers of apoE in the brain being astrocytes (Pitas et al., 1987).”

What the sources say

• ?Not checked — no full text of the sourcecites Pitas et al., 1987 source

In the paper

“Together, these findings suggest that neuronal expression of Rorb in the hippocampus is dramatically increased in the presence of tau pathology, and that removal of astrocytic APOE4 attenuates neuronal Rorb expression, consistent with an association between this distinct neuronal subpopulation and tau pathology.”

In the paper

“Combined, these results indicate that after astrocytic APOE4 removal, astrocyte activation is reduced, which may contribute to the protective effect on tau-mediated neurodegeneration following astrocytic APOE4 removal.”

In the paper

“Together, our results show that our treatment strategy sufficiently and specifically reduces astrocytic apoE mRNA and protein.”

In the paper

“The apolipoprotein E gene (APOE) is the strongest genetic risk factor for late-onset Alzheimer’s disease (AD) with the ε4 allele increasing and the ε2 allele decreasing AD risk relative to the ε3 allele (Holtzman et al., 2012; Strittmatter et al., 1993).”

What the sources say

• Not checked yet

  Checked before — its earlier verdict is held back until a new check runs.

  cites Holtzman et al., 2012 source
• Not checked yetcites Strittmatter et al., 1993 source

In the paper

“Expression of C4b may be involved with other complement pathway genes in inducing prolonged inflammation and increasing the activity of microglial-dependent phagocytosis.”

In the paper

“All P301S Tau transgenic mice involved in the final analysis were obtained from the same generation.”

In the paper

“The modest reduction of these genes in TAM-treated mice suggests another mechanism by which protection may be afforded by astrocytic APOE removal.”

In the paper

“Our data that selective removal of astrocyte-apoE4 reduces microglial-mediated phagocytosis suggests the possibility that astrocyte-secreted apoE4, either alone or together with other molecules such as C1q, could also serve as an opsonin for specific microglia receptors.”

In the paper

“RBPs are a class of proteins responsible for the regulation of mRNA maturation and translation and are increasingly implicated in multiple neurodegenerative diseases, including tauopathies (Wolozin and Ivanov, 2019).”

What the sources say

• Not checked yet

  Checked before — its earlier verdict is held back until a new check runs.

  cites Wolozin and Ivanov, 2019 source

In the paper

“All of the phenotyping and data analysis were accomplished by researchers who were completely blind to the genotype of the mice.”

In the paper

“Littermates of the same sex were randomly assigned to experimental groups.”