β-amyloid accumulation enhances microtubule associated protein tau pathology in an APPNL-G-F/MAPTP301S mouse model of Alzheimer’s disease

May combine two claims.

Human MAPT knock-in models produce tau pathology, but only very late in the murine life span (~15 months).

“Human MAPT knock-in models have also been developed, and these models produce tau pathology, but only very late in the murine life span (~15 months) (Saito et al., 2014; Sakakibara et al., 2018; Sasaguri et al., 2022).”

• cites Saito et al., 2014 sourceNot checked — no full text of the source
• cites Sakakibara et al., 2018 sourceCould not verify

  The Background discusses App knock-in mice and says they do not exhibit tau pathology; it does not discuss human MAPT knock-in models. A search for MAPT knock-in tau pathology at approximately 15 months found no passage addressing that claim. The Results’ reference to 15–18 months concerns reactive gliosis in App knock-in mice, not tau pathology. (LLM, conf=0.75)

  Experimental — accuracy not yet measured
• cites Sasaguri et al., 2022 sourceDisputed

  The section “Humanization of the Entire Murine Mapt Gene to Generate hMAPT Knock-In Mice” (page unavailable) says that mice carrying human MAPT together with an App knock-in lacked evidence of tau pathology even at 24 months, contrary to the claimed onset at about 15 months. It separately reports aging-dependent tau aggregation in mutant MAPT knock-in mice, but gives no 15-month onset for those models. (LLM, conf=0.88)

  Checked by a reasoning model — may be wrong.

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

Other groups have observed that m6A accumulates in models of tauopathy, as well as in other diseases with intracellular aggregates, including ALS and Huntington's disease.

“Other groups have also observed that m6A accumulates in models of tauopathy, as well as in other diseases with intracellular aggregates, including ALS and Huntington’s disease (McMillan et al., 2023; Nguyen et al., 2023; Atrian et al., 2024).”

• cites McMillan et al., 2023 sourceSupported
• cites Nguyen et al., 2023 sourceDisputed

  For the Huntington’s disease portion of the sentence, the m6A-sequencing section reports no significant global change in m6A levels in HD R6/2 mouse cortex. It also reports decreased m6A deposition at specific sites on dysregulated HD genes, rather than accumulation relative to controls. The paper discusses ALS and tau findings, but those do not establish accumulation in its HD model. (LLM, conf=0.88)

  Checked by a reasoning model — may be wrong.

  Experimental — accuracy not yet measured
• cites Atrian et al., 2024 sourcePartly supported

  The Results and Discussion report elevated m6A in tauopathy models, including tau-transgenic Drosophila and MAPT-mutant cerebral organoids. The Discussion mentions ALS and Huntington’s disease in connection with muscleblind, but does not report m6A accumulation in either disease. (LLM, conf=0.88)

  Experimental — accuracy not yet measured

May combine two claims.

These models develop little intracellular MAPT pathology, nor other types of neuronal protein aggregates.

“These models develop little intracellular MAPT pathology, nor other types of neuronal protein aggregates (Oakley et al., 2006).”

• cites Oakley et al., 2006 sourceDisputed

  The 5XFAD mice section reports no neurofibrillary tangles or abnormal phospho-tau staining, supporting the MAPT portion. But the Introduction and the intraneuronal Aβ section explicitly report aggregated intraneuronal Aβ, contrary to the claim that these models do not develop other neuronal protein aggregates. (LLM, conf=0.88)

  Checked by a reasoning model — may be wrong.

  Experimental — accuracy not yet measured

Classified as probably not a substantial claim.

Cognitive loss is only weakly correlated with the accumulation of Aβ.

“Mutations in amyloid precursor protein (APP) cause AD, however cognitive loss is only weakly correlated with the accumulation of Aβ (Scheltens et al., 2021).”

• cites Scheltens et al., 2021 sourceSupported

Probably not a substantial claim: may not pass S3 (load-bearing).

Mutations in MAPT are sufficient to cause dementia (frontotemporal dementia, FTD).

“Cognitive loss is much more robustly correlated with MAPT based NFTs pathology, and mutations in MAPT are sufficient to cause dementia (frontotemporal dementia, FTD) (Scheltens et al., 2021).”

• cites Scheltens et al., 2021 sourcePartly supported

  The Tau-PET section (page not provided) says tau-PET binding topography correlates with cognitive deficits, supporting the sentence’s point about tau pathology and cognition. Our copy does not state that MAPT mutations are sufficient to cause frontotemporal dementia; searches for MAPT mutations and frontotemporal dementia found no such passage. (LLM, conf=0.83)

  Experimental — accuracy not yet measured

Classified as probably not a substantial claim. May combine two claims.

Mutations in MAPT might not cause AD because MAPT pathology does not drive the accumulation of Aβ pathology.

“Mutations in MAPT might not cause AD because MAPT pathology does not drive the accumulation of Aβ pathology (Scheltens et al., 2021).”

• cites Scheltens et al., 2021 sourcePartly supported

  The Pathophysiology section says amyloid β induces the spread of tau pathology, and the Diagnostic criteria section describes amyloid β as a defining feature of Alzheimer’s disease. Neither passage establishes that MAPT mutations do not cause Alzheimer’s disease or that tau pathology cannot drive amyloid β accumulation. A search for MAPT found a reference to the MAPT lifestyle trial, not to mutations in the gene. (LLM, conf=0.76)

  Experimental — accuracy not yet measured

Classified as probably not a substantial claim.

Cognitive loss is much more robustly correlated with MAPT based NFTs pathology.

“Cognitive loss is much more robustly correlated with MAPT based NFTs pathology, and mutations in MAPT are sufficient to cause dementia (frontotemporal dementia, FTD) (Scheltens et al., 2021).”

• cites Scheltens et al., 2021 sourcePartly supported

  The Tau-PET passage (Introduction > Imaging biomarkers; page not provided) says that, unlike amyloid-β deposition, tau-PET binding topography correlates with cognitive deficits. This supports the comparison in the claim, but the passage concerns a tau-imaging measure rather than directly measuring MAPT-based neurofibrillary tangles. The cited sentence also attributes the claim that MAPT mutations suffice to cause frontotemporal dementia to this source; searches for MAPT and frontotemporal dementia did not find support for that claim in our copy. (LLM, conf=0.84)

  Experimental — accuracy not yet measured

Classified as probably not a substantial claim.

Mutations in APP cause AD.

“Mutations in amyloid precursor protein (APP) cause AD, however cognitive loss is only weakly correlated with the accumulation of Aβ (Scheltens et al., 2021).”

• cites Scheltens et al., 2021 sourceCould not verify

  The Genetics section describes an APP mutation associated with prolonged cognitive health, but does not state that APP mutations cause Alzheimer’s disease. A search for causative APP mutations found discussion of clinical mutations in presenilins, not APP; the protective APP variant does not rule out other APP mutations being causative. (LLM, conf=0.72)

  Experimental — accuracy not yet measured

May combine two claims.

Knock-in human APP mice develop robust plaque pathology beginning as early as 3 months of age, however they develop little tau pathology and little neurodegeneration.

“These mice develop robust plaque pathology beginning as early as 3 months of age, however they develop little tau pathology and little neurodegeneration (Saito et al., 2014; Sakakibara et al., 2018).”

• cites Saito et al., 2014 sourceNot checked — no full text of the source
• cites Sakakibara et al., 2018 sourcePartly supported

  The Results report that cortical Aβ amyloidosis in AppNL-G-F/NL-G-F mice began by 2 months, supporting early plaque pathology but not the stated 3-month onset. The Background says none of these App-KI mice exhibit tau pathology or severe neuron loss, supporting the rest of the claim. (LLM, conf=0.86)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

The V-PLEX platform offers analytically validated singleplex and multiplex assay kits, which can provide accurate and reproducible results with consistency from lot to lot.

“The V-PLEX platform offers analytically validated singleplex and multiplex assay kits, which can provide accurate and reproducible results with consistency from lot to lot (van Dyck et al., 2023).”

• cites van Dyck et al., 2023 sourceCould not verify

  The Trial Design and Oversight and End Points sections describe blood testing and the biomarkers assessed, but do not identify the V-PLEX platform or discuss its assay kits. Searches for V-PLEX, singleplex and multiplex kits, assay validation, reproducibility, and lot-to-lot consistency found no supporting passage in our copy. (LLM, conf=0.75)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

The tau pathology does not appear to increase Aβ deposition.

“The tau pathology does not appear to increase Aβ deposition (Saul et al., 2013; Chabrier et al., 2014; Heraud et al., 2014; Stancu et al., 2014; Chen et al., 2016; Kang et al., 2021); indeed, in some cases it accelerates removal of Aβ (Chen et al., 2016).”

• cites Saul et al., 2013 sourceNot checked — no full text of the source
• cites Chabrier et al., 2014 sourcePartly supported

  The Results section, “The presence of human APP increases tau accumulation in hAPPSL/hTau mice” (page unavailable), reports unchanged APP levels in mice with both hAPP and hTau compared with mice with hAPP alone. It does not report a comparison of Aβ deposition between those groups, so unchanged APP levels provide only indirect support for the claim. (LLM, conf=0.76)

  Experimental — accuracy not yet measured
• cites Heraud et al., 2014 sourceNot checked — no full text of the source

May combine two claims.

Human MAPT KI models exhibit delayed tau pathology, as does a cross between the two KI models.

“Human MAPT KI models exhibit delayed tau pathology, as does a cross between the two KI models (Saito et al., 2019; Islam et al., 2023).”

• cites Saito et al., 2019 sourcePartly supported

  The Discussion (page unavailable) says the App/MAPT double-KI mice lacked apparent tauopathy and neurodegeneration despite amyloid pathology, and discusses the timing of subsequent pathological events. This is consistent with pathology developing slowly, but the paper does not demonstrate a delayed onset of tau pathology in either the single MAPT KI or double-KI mice. (LLM, conf=0.82)

  Experimental — accuracy not yet measured
• cites Islam et al., 2023 sourcePartly supported

  The Discussion section on spatially distinct pathology reports that mature tau tangles in the APP/MAPT double-knock-in mice were frequent at 18–22 months but virtually absent at 11–16 months, supporting later development of that form of pathology. It also reports elevated hyperphosphorylated tau in the younger mice. The paper does not establish delayed tau pathology in a human MAPT knock-in model studied on its own. (LLM, conf=0.85)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

Mouse models over-expressing WT or mutant tau exhibit progressive neurodegeneration.

“These models also exhibit progressive neurodegeneration, which is consistent with observations in humans that cognitive loss is more closely associated with tau pathology than Aβ pathology (Lewis et al., 2000; Ramsden et al., 2005; Polydoro et al., 2009).”

• cites Lewis et al., 2000 sourceNot checked — no full text of the source
• cites Ramsden et al., 2005 sourcePartly supported

  The Discussion (pp. 10637–10647; passage page unspecified) describes progressive neurodegeneration and substantial neuron loss in the authors’ P301L tau mice. It also mentions cell death in a non-mutant human-tau model, but cautions that most earlier mutant-tau models did not exhibit or report brain neuron loss. The broad claim about mouse models overexpressing either form therefore goes beyond what this paper establishes. (LLM, conf=0.87)

  Experimental — accuracy not yet measured
• cites Polydoro et al., 2009 sourcePartly supported

  The Introduction reports age-related cortical thinning and fewer detectable nuclei in htau mice carrying nonmutant human tau. The Discussion reports neural loss in a mutant-tau mouse model. These passages support neurodegeneration in both kinds of model, but they do not establish that the mice overexpress tau. (LLM, conf=0.83)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

The patterns of labeling of Aβ and MAPT are consistent with prior studies of APPNL-G-F and P301S MAPT mice.

“These patterns of labeling or Aβ and MAPT are consistent with prior studies of APPNL-G-F and P301S MAPT mice (Yoshiyama et al., 2007; Saito et al., 2014; Sasaguri et al., 2017; Apicco et al., 2018).”

• cites Yoshiyama et al., 2007 sourceNot checked — no full text of the source
• cites Saito et al., 2014 sourceNot checked — no full text of the source
• cites Sasaguri et al., 2017 sourcePartly supported

  The Modeling preclinical AD section shows Aβ labeling in the cortex of an App NL-G-F mouse, and the Second-generation mouse models section describes amyloid pathology in that model. The paper does not describe labeling patterns in P301S MAPT mice; its discussion of tau models concerns other mutations. It therefore supports the Aβ-model portion, but not the attributed MAPT-model comparison. (LLM, conf=0.85)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

The report on the APPNL-G-F/MAPTP290S mouse showed enhancement of MAPT pathology by Aβ but the pathology occurs only in very old mice, and the manuscript did not provide information on other pathologies such as inflammation or m6A.

“The report on the APPNL-G-F/MAPTP290S mouse showed enhancement of MAPT pathology by Aβ but the pathology occurs only in very old mice, and the manuscript did not provide any information on other pathologies, such as inflammation or m6A.”

• cites Huang et al., 2022 sourcePartly supported

  The Discussion (page unavailable) says Aβ significantly increased tau inclusions only from 18 months of age, although some tau inclusions occurred earlier. The Results section on age-dependent neurodegeneration reports nerve-cell loss in the double-knock-in mice from 18 months, so the claim that the paper provided no information on other pathologies is too broad. The passages found do not report inflammation or m6A findings. (LLM, conf=0.88)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

The results are consistent with those recently reported for an APPNL-G-F/MAPTP290S mouse.

“These results are consistent with those recently reported for an APPNL-G-F/MAPTP290S mouse, which is a mouse model in which both human APP and human MAPT are knockins (Huang et al., 2022).”

• cites Huang et al., 2022 sourcePartly supported

  The Abstract and Materials and Methods (Mouse lines; page not provided) identify the APPNL-G-F × MaptP290S mouse and report increased tau pathology, supporting the citation’s identification of the model. But Mouse lines says the P290S mutation was inserted into murine Mapt and describes the APP line as expressing humanized Aβ; it does not support the statement that both human APP and human MAPT were knocked in. The citing paper’s own results are not provided, so their consistency with Huang et al.’s findings cannot be assessed here. (LLM, conf=0.90)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Published work indicates that Aβ pathology strongly stimulates the Disease Associated Microglial phenotype while MAPT pathology elicits microglial responses exhibiting weaker cytokine production.

“These results are also consistent with published work indicating that Aβ pathology strongly stimulates the “Disease Associated Microglial” phenotype while MAPT pathology elicits microglial responses exhibiting weaker cytokine production (Paolicelli et al., 2022).”

• cites Paolicelli et al., 2022 sourcePartly supported

  In “Microglial States in the Diseased CNS” (page unavailable), the paper reports a DAM signature in Alzheimer’s disease models and human samples, as well as a distinct tau-associated microglial cluster. It does not say that Aβ pathology strongly stimulates DAM or that responses to MAPT pathology produce fewer cytokines; it explicitly leaves open whether TREM2 senses amyloid-β pathology. (LLM, conf=0.83)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

TREM2 is one of the strongest risk factors for AD, and evidence suggests that TREM2 acts to recognize Aβ pathology and direct microglial responses toward the pathology.

“For instance, TREM2 is one of the strongest risk factors for AD, and evidence suggests that TREM2 acts to recognize Aβ pathology and direct microglial responses toward the pathology (Guerreiro et al., 2013; Zhou et al., 2020).”

• cites Guerreiro et al., 2013 sourcePartly supported

  The Abstract and Results report that rare TREM2 variants significantly increase Alzheimer’s disease risk. The Discussion reports TREM2 expression in microglia around amyloid plaques and describes it as a possible gateway for microglial responses. It does not establish that TREM2 itself recognizes Aβ pathology or directly guides microglia toward it. (LLM, conf=0.86)

  Experimental — accuracy not yet measured
• cites Zhou et al., 2020 sourcePartly supported

  The introductory passage in the Abstract (page unavailable) says TREM2 variants raise AD risk several-fold and that loss of TREM2 function restricts microglia from surrounding Aβ plaques. The Results section on DAM signatures (page unavailable) also reports that both Aβ and TREM2 are required for the disease-associated microglial response. These passages support a role in directing microglial responses to Aβ pathology, but do not establish that TREM2 itself recognizes that pathology. (LLM, conf=0.83)

  Experimental — accuracy not yet measured

May combine two claims.

Cell body MAPT pathology colocalizes with markers of the translational stress response, such as stress granule proteins HNRNPA2B1, TIA1, EIF3n and PABP, and also neuronal death (cleaved caspase 3 and loss of NeuN positive cells).

“The cell body MAPT pathology is particularly important because this pathology colocalizes with markers of the translational stress response, such as stress granule proteins HNRNPA2B1, TIA1, EIF3n and PABP, and also neuronal death (as shown by cleaved caspase 3 and loss of NeuN positive cells) (Ash et al., 2014; Apicco et al., 2018; Maziuk et al., 2018; Apicco et al., 2019; Jiang et al., 2019, 2021”

• cites Ash et al., 2014 sourcePartly supported

  The 'SGs in Neurodegenerative disease' section reports that TIA-1 colocalizes with pathological tau in affected neurons, supporting that part of the claim. It does not report colocalization with HNRNPA2B1, EIF3n or PABP, or demonstrate neuronal death by cleaved caspase 3 or loss of NeuN-positive cells. The same section cautions that tau pathology in a neuron is not necessarily correlated with that neuron's demise. No page numbers are available in our copy. (LLM, conf=0.88)

  Experimental — accuracy not yet measured
• cites Apicco et al., 2018 sourcePartly supported

  The Results (page not provided) report TIA1 co-labeling of pre-fibrillar tau aggregates, TIA1 puncta co-localizing with PABP, and fewer NeuN-positive neurons in the tauopathy mice. They do not show cell-body tau pathology co-localizing with all four named stress-granule proteins or with markers of neuronal death; the passages found do not report cleaved caspase 3. (LLM, conf=0.86)

  Experimental — accuracy not yet measured
• cites Maziuk et al., 2018 sourcePartly supported

  The Results (Figs. 1 and 4; page not provided) show that TIA1 and PABP colocalize with diffuse phosphorylated tau in neuronal cell bodies, but that colocalization is reduced or absent in mature tangles. The proteins reported in Fig. 4 are HNRNPA0 and eIF2α, not the claim’s HNRNPA2B1 and EIF3n. These passages do not establish neuronal death through cleaved caspase 3 or loss of NeuN-positive cells. (LLM, conf=0.88)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

The classic model of AD is based on the Aβ cascade and proposes that aggregated Aβ stimulates inflammation, and that these two factors produce additive injury to elicit neurodegeneration in AD.

“The classic model of AD is based on the Aβ cascade and proposes that aggregated Aβ stimulates inflammation, and that these two factors produce additive injury to elicit neurodegeneration in AD (Scheltens et al., 2021).”

• cites Scheltens et al., 2021 sourcePartly supported

  The Conclusions identify an amyloid-β cascade hypothesis, and the Pharmacological section says neuroinflammation contributes to neurodegeneration. But the Pathophysiology section places neuroinflammation upstream of or parallel to amyloid-β accumulation; it does not describe aggregated amyloid-β stimulating inflammation or the two producing additive injury. The paper also says the cascade hypothesis has been modified, rather than presenting the cited sequence as its account. (LLM, conf=0.85)

  Experimental — accuracy not yet measured

May combine two claims.

The same studies indicate that the presence of MAPT pathology either does not change the accumulation of Aβ pathology or decreases it.

“These same studies indicate that the presence of MAPT pathology either does not change the accumulation of Aβ pathology or decreases it (Saul et al., 2013; Chabrier et al., 2014; Heraud et al., 2014; Stancu et al., 2014; Chen et al., 2016; Kang et al., 2021), with only one initial study suggesting that MAPT pathology increases Aβ pathology (Ribe et al., 2005).”

• cites Saul et al., 2013 sourceNot checked — no full text of the source
• cites Chabrier et al., 2014 sourcePartly supported

  The Results section on APP and tau accumulation reports unchanged total APP levels in hAPPSL/hTau mice compared with hAPPSL mice, which supports a narrower no-change finding. It does not report a comparison of Aβ accumulation between those groups, so unchanged APP cannot establish that tau pathology leaves Aβ pathology unchanged or decreases it. No page number is available in our copy. (LLM, conf=0.83)

  Experimental — accuracy not yet measured
• cites Heraud et al., 2014 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

This injury is further enhanced by aging, vascular damage, metabolic conditions such as Type II Diabetes and genetic risk factors.

“This injury is further enhanced by aging, vascular damage, metabolic conditions such as Type II Diabetes and genetic risk factors (Scheltens et al., 2021).”

• cites Scheltens et al., 2021 sourcePartly supported

  The Epidemiology section identifies advanced age and APOE ε4 as major Alzheimer’s disease risk factors and associates cardiovascular risks with dementia. The Pathophysiology discussion also places ageing and vascular changes in the disease landscape. But the paper does not establish that type 2 diabetes or these factors enhance the injury described in the claim; it cautions that vascular risks may contribute to dementia without increasing Alzheimer’s pathology. (LLM, conf=0.86)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

The m6A modification occurs in all types of RNA, although it has the lowest abundance in tRNA.

“The m6A modification occurs in all types of RNA, although it has the lowest abundant in tRNA (Linder et al., 2015; Flamand et al., 2023).”

• cites Linder et al., 2015 sourcePartly supported

  The Results sections on crosslinking sites and snoRNAs identify m6A in rRNA and snoRNAs, and the Abstract reports it in mRNA. Our copy does not establish that m6A occurs in all RNA types or that its abundance is lowest in tRNA; a search for tRNA abundance found no supporting passage. Page numbers were not available in the indexed copy. (LLM, conf=0.87)

  Experimental — accuracy not yet measured
• cites Flamand et al., 2023 sourcePartly supported

  The Introduction reports m6A in mRNA, and the section on m6A recognition by the YTH domain identifies m6A sites in 18S and 28S rRNA (pages unavailable). The review does not establish that m6A occurs in all RNA types or that its abundance is lowest in tRNA; its discussion of tRNA in ‘Writers of other mRNA modifications’ concerns other modifications. (LLM, conf=0.82)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

m6A accumulates in the cytoplasm in response to stress.

“m6A is known to accumulate in the cytoplasm in response to stress, but the function of cytoplasmic m6A remains to be clearly delineated (Meier et al., 2016; Briggs et al., 2017; Koren et al., 2019; Ries et al., 2019; Wolozin and Ivanov, 2019; Fu and Zhuang, 2020).”

• cites Meier et al., 2016 sourceCould not verify

  The Results passages describe pathological tau associating with ribosomes and reducing protein synthesis, not m6A localization. Searches for m6A, N6-methyladenosine, and cytoplasmic accumulation under stress found no passage addressing the claim in our copy. (LLM, conf=0.74)

  Experimental — accuracy not yet measured
• cites Briggs et al., 2017 sourceCould not verify

  The Introduction discusses ER stress, unfolded proteins, and stress-related translation of ATF4 mRNA, but not m6A or its location in the cell. Searches for m6A, cytoplasmic accumulation, and N6-methyladenosine found no passage in our copy addressing the claim. (LLM, conf=0.76)

  Experimental — accuracy not yet measured
• cites Koren et al., 2019 sourceCould not verify

  The Results and Discussion passages address tau, ribosomal proteins, and translation in Alzheimer’s disease, not cytoplasmic m6A accumulation. Searches for m6A accumulation under stress and for N6-methyladenosine found no passage in our copy addressing that claim. (LLM, conf=0.76)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

Antibodies against m6A also recognize demethylated adenosine, m6Am.

“Antibodies against m6A also recognize demethylated adenosine, m6Am.”

• cites Linder et al., 2015 sourcePartly supported

  The Abstract reports that anti-m6A antibodies also produce mutational signatures at m6Am, supporting the claim that they recognize it. But it identifies m6Am as N6,2′-O-dimethyladenosine, not demethylated adenosine. (LLM, conf=0.91)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

m6Am non-enzymatically converts to m6A.

“Interestingly, m6Am also non-enzymatically converts to m6A, so the biological impact of m6Am is currently ambiguous (Linder et al., 2015; Flamand et al., 2023).”

• cites Linder et al., 2015 sourceCould not verify

  The Results section, “CITS miCLIP identifies m6Am at the TSS” (page unavailable), says the function of m6Am is poorly understood, but does not say it converts to m6A. Searches for non-enzymatic conversion, demethylation, and loss of the 2′-O-methyl group found no passage addressing the claimed conversion. (LLM, conf=0.72)

  Experimental — accuracy not yet measured
• cites Flamand et al., 2023 sourceCould not verify

  The “Writers of other mRNA modifications” section describes enzymatic formation of m6Am by PCIF1 and uncertainty about m6Am’s biological effects. It does not address non-enzymatic conversion of m6Am to m6A; searches for that conversion and its proposed mechanism found no passage stating it. (LLM, conf=0.74)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

Since aggregated MAPT is thought to mediate Aβ induced degeneration, the absence of MAPT pathology might explain the low levels of neurodegeneration observed in APP mouse models.

“since aggregated MAPT is thought to mediate Aβ induced degeneration, the absence of MAPT pathology also might explain the low levels of neurodegeneration observed in APP mouse models (Oakley et al., 2006).”

• cites Oakley et al., 2006 sourcePartly supported

  The sections “Intraneuronal Aβ, neurodegeneration, and amyloid plaque formation” and “5XFAD mice: rapid amyloid pathology models” report that most APP mouse models lack robust neuron loss and that tangles have not been reported in APP or APP/PS1 models. They do not propose that absent MAPT pathology explains the limited neuron loss; indeed, the paper reports neuron loss in 5XFAD mice despite finding no abnormally phosphorylated tau epitopes. (LLM, conf=0.88)

  Experimental — accuracy not yet measured

• AssumptionRests on an assumption: Aggregated MAPT mediates Aβ induced degeneration.

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

m6Am constitutes 2-10% of total m6A detected in tissues, with brain being on the lower end of the abundance.

“This modification constitutes 2–10% of total m6A detected in tissues, with brain being on the lower end of the abundance (Linder et al., 2015).”

• cites Linder et al., 2015 sourceCould not verify

  The Results section, “CITS miCLIP identifies m6Am at the TSS” (page unavailable), reports mapping m6Am in mRNAs but gives no 2–10% abundance estimate or tissue comparison. Searches for that percentage, relative abundance, and brain found no passage supporting the claim; the Methods describe HEK293 cells and mouse liver, not brain tissue. (LLM, conf=0.75)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

Conditional knockout of METTL3 in microglia attenuated inflammation and Aβ accumulation in a mouse model based on Aβ injection.

“indeed a recent study observed that conditional knockout of METTL3 in microglia attenuated inflammation and Aβ accumulation in a mouse model based on Aβ injection (Yin et al., 2023).”

• cites Yin et al., 2023 sourcePartly supported

  The Results section on Aβ clearance (page not available) reports significantly lower cortical amyloid burden in Aβ-injected mice with METTL3 depletion. But the study attributes this effect to monocyte-derived macrophages, not a microglia-specific knockout. The passages found do not report that inflammation was attenuated. (LLM, conf=0.88)

  Experimental — accuracy not yet measured

May combine two claims.

The increases in m6A are also consistent with the observed increase in the m6A writer METTL3, and decrease in m6A eraser ALKBH5.

“The increases in m6A are also consistent with the observed increase in the m6A writer, METTL3, and decrease in m6A eraser, ALKBH5 (Tong et al., 2018).”

• cites Tong et al., 2018 sourceSupported

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

m6A primarily accumulated in neuronal soma, but also co-localized with a subset of astrocytes and microglia.

“m6A primarily accumulated in neuronal soma, but also co-localized with a subset of astrocytes and microglia.”

May combine two claims.

MAPT pathology neither changed levels of amyloid precursor protein nor potentiated A accumulation.

“MAPT pathology neither changed levels of amyloid precursor protein nor potentiated A accumulation.”

Probably not a substantial claim: may not pass S3 (load-bearing).

The APPNL-G-F/MAPTP301S mouse model showed strong accumulation of N6-methyladenosine (m6A).

“The APPNL-G-F/MAPTP301S mouse model also showed strong accumulation of N6-methyladenosine (m6A), which was recently shown to be elevated in the AD brain.”

• AssumptionRests on an assumption: The m6A changes observed in the APPNL-G-F/MAPTP301S mouse model correspond to those in human AD.

Probably not a substantial claim: may not pass S3 (load-bearing).

Aβ38 levels decreased with age in APPNL-G-F and APPNL-G-F × MAPTP301S mice.

“In contrast, Aβ38 levels decreased with age in APPNL-G-F and APPNL-G-F × MAPTP301S mice (Figure 1D).”

May combine two claims.

In the APPNL-G-F mouse the Aβ42 level peaked at 6 months and remained constant at 9 months.

“Notably, in the APPNL-G-F mouse the Aβ42 level peaked at 6 months and remained constant at 9 months while levels of Aβ42 steadily increased with age (at 3, 6, and 9 months) in APPNL-G-F × MAPTP301S mouse brain (Figure 1C).”

May combine two claims.

Both APPNL-G-F/MAPTP301S and MAPTP301S accumulated hyperphosphorylated tau.

“The result showed that both APPNL-G-F/MAPTP301S and MAPTP301S accumulated hyperphosphorylated tau, but crossing APPNL-G-F to MAPTP301S increased the accumulation of pTau181 and 217 (Figures 2B–D).”

Probably not a substantial claim: may not pass S3 (load-bearing).

BDNF transcript was elevated in each of the mouse models.

“Quantification of BDNF transcript by RT-PCR in the APPNL-G-F/MAPTP301S mouse model showed elevation BDNF each of the mouse models (Figure 3I).”

Probably not a substantial claim: may not pass S3 (load-bearing).

WT brain exhibited very little labeling.

“As expected, WT brain exhibited very little labeling (Supplementary Video S3).”

Probably not a substantial claim: may not pass S3 (load-bearing).

Neuritic plaques were evident throughout the neocortex.

“Neuritic plaques were evident throughout the neocortex, as expected (Supplementary Videos S1–3 Aβ = Green).”

May combine two claims.

Mice carrying MAPTP301S and APPNL-G-F/MAPTP301S genes exhibited levels of TNF-α and IL-1β elevated compared to WT mice but significantly less than APPNL-G-F mice.

“interestingly, mice carrying the MAPTP301S and APPNL-G-F/MAPTP301S genes exhibited levels that were elevated compared to WT mice, but significantly less than that in the APPNL-G-F mice (Figures 3E,H).”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Labeling for phospho-tau was weak due to technical difficulties, but labeling present in APPNL-G-F/MAPTP301S mouse brains occurred in the hippocampus and entorhinal cortex.

“Labeling for phospho-tau encountered technical difficulties, so was weak, but the labeling that was present in the APPNL-G-F/MAPTP301S mouse brains occurred in the hippocampus and entorhinal cortex (Supplementary Videos S2, 3, pS202/5 MAPT, AT8 = Red).”

May combine two claims.

Strong labeling of oligomeric MAPT was selective for the somatic compartment of neurons and exhibited progressive age-dependent accumulation.

“The result showed strong labeling of oligomeric MAPT that was selective for the somatic compartment of the neurons and exhibited progressive age dependent accumulation as the mice (Figure 2F).”

May combine two claims.

Similar results were observed for astrocytosis, along with strong co-localization with Aβ plaques.

“Similar results were observed for astrocytosis (labeled by astrocytic marker GFAP), along with strong co-localization with the Aβ plaques (Figures 3A,D).”

May combine two claims.

Microglia respond to MAPT pathology; Iba-1 reactivity was increased in the P301S MAPT model, but not to the same level as observed in the APPNL-G-F or APPNL-G-F/MAPTP301S model.

“However, it is important to note that microglia do respond to MAPT pathology, Iba-1 reactivity was increased in the P301S MAPT model, but not to the same level as observed in the APPNL-G-F or APPNL-G-F/MAPTP301S model.”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

The amount and distribution of m6A parallels MAPT pathology.

“In addition, the amount and distribution of m6A parallels MAPT pathology.”

May combine two claims.

These microglia were more abundant and showed greater ramifications than in the P301S MAPT model.

“These microglia were more abundant and showed greater ramifications than in the P301S MAPT model.”

May combine two claims.

The APPNL-G-F/MAPTP301S model exhibits strong Aβ accumulation and strong inflammation.

“This model exhibits strong Aβ accumulation and strong inflammation, while avoiding artifacts associated with APP or presenilin over-expression.”

• AssumptionRests on an assumption: The APPNL-G-F/MAPTP301S mouse model is a valid model of Alzheimer's disease relevant to human disease.

May combine two claims.

Some astrocytes and microglia also exhibited strong increases in m6A in the APPNL-G-F/MAPTP301S model, although changes in m6A were most evident in neurons.

“Although changes in m6A were most evident in neurons, the results in Figure 5A show that some astrocytes and microglia also exhibited strong increases in m6A in the APPNL-G-F/MAPTP301S model.”

• AssumptionRests on an assumption: Anti-m6A antibody immunohistochemistry reliably reflects total m6A levels despite cross-reactivity with m6Am.

May combine two claims.

Increases in m6A observed correlate with MAPT pathology and show a corresponding association with increased levels of METTL3 and reduced levels of ALKBH5.

“The increases in m6A observed correlate with MAPT pathology are show a corresponding association with increased levels of METTL3 and reduced levels of ALKBH5, which are the enzymes that, respectively, add and remove m6A from mRNA.”

• AssumptionRests on an assumption: Anti-m6A antibody immunohistochemistry reliably reflects total m6A levels despite cross-reactivity with m6Am.

May combine two claims.

Human APP knockin (KI) models develop robust Aβ pathology, yet do not exhibit artifacts associated with APP over-expression, such as elevated levels of APP cleavage products.

“Human APP knockin (KI) models develop robust Aβ pathology, yet do not exhibit artifacts associated with APP over-expression, such as elevated levels of APP cleavage products (Saito et al., 2014).”

• cites Saito et al., 2014 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

Use of all animals was approved by the Boston University IACUC.

“Use of all animals was approved by the Boston University Institutional and Animal Care and Use Committee (IACUC).”

Classified as probably not a substantial claim.

Mouse models over-expressing WT or mutant tau develop robust tau pathologies.

“These models develop robust tau pathologies, including pathologically phosphorylated, misfolded, oligomeric and/or fibrillar forms of tau pathology.”

• cites Hutton et al., 2001 sourceNot checked — no full text of the source
• cites Sasaguri et al., 2017, 2022Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

May combine two claims.

The APPNL-G-F × MAPTP301S double transgenic mouse model recapitulates the progressive accumulation Aβ plaques and levels similar to that seen in the APPNL-G-F model although total levels of APP and Aβ accumulation are lower in the cross than with the knock-in gene alone.

“These results indicate that the APPNL-G-F × MAPTP301S double transgenic mouse model recapitulates the progressive accumulation Aβ plaques and levels similar to that seen in the APPNL-G-F model although total levels of APP and Aβ accumulation are lower in the cross than with the knock-in gene alone.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Standards with known concentrations were used for Aβ, and all standards and samples were run in duplicate.

“Standards with known concentrations were used for Aβ, and all standards and samples were run in duplicate.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

The quantification of cell numbers was done in a blinded manner, with the investigator analyzing the samples being different than the investigator coding the samples.

“The quantification of cell numbers was done in a blinded manner, with the investigator analyzing the samples being different than the investigator coding the samples.”

May combine two claims.

Neurodegeneration parallels tau pathology, being enhanced in the double transgenic mouse.

“Neurodegeneration parallels tau pathology, being enhanced in the double transgenic mouse.”

Probably not a substantial claim: may not pass S3 (load-bearing).

From an anatomical perspective, the disparate distribution of Aβ and MAPT pathology in the APPNL-G-F/MAPTP301S model might also be accentuated by the strong genetic engineering that drives the MAPT pathology.

“From an anatomical perspective, the disparate distribution of Aβ and MAPT pathology in the APPNL-G-F/MAPTP301S model might also be accentuated by the strong genetic engineering that drives the MAPT pathology, although general toxicity from Aβ accumulation does appear to enhance total MAPT pathology and neurodegeneration in this model.”

May combine two claims.

General toxicity from Aβ accumulation does appear to enhance total MAPT pathology and neurodegeneration in this model.

“From an anatomical perspective, the disparate distribution of Aβ and MAPT pathology in the APPNL-G-F/MAPTP301S model might also be accentuated by the strong genetic engineering that drives the MAPT pathology, although general toxicity from Aβ accumulation does appear to enhance total MAPT pathology and neurodegeneration in this model.”

May combine two claims.

The accumulation of Aβ in the APPNL-G-F/MAPTP301S model clearly does accelerate the appearance and enhance the amount of MAPT pathology moderately.

“Such a strong genetic push reduces the need for a toxin, such as Aβ, to drive the MAPT pathology, although the accumulation of Aβ in the APPNL-G-F/MAPTP301S model clearly does accelerate the appearance and enhance the amount of MAPT pathology moderately (Graphical abstract).”

May combine two claims.

MAPT pathology in the MAPTP301S model accumulates rapidly because of the strong MAPT expression (the human protein is expressed about 5-fold higher than the endogenous mouse protein) and the protein itself has a mutation that enhances the rate of aggregation.

“MAPT pathology in the MAPTP301S model accumulates so rapidly because of the strong MAPT expression (the human protein is expressed about 5-fold higher than the endogenous mouse protein) and the protein itself has a mutation that enhances the rate of aggregation (Yoshiyama et al., 2007).”

• cites Yoshiyama et al., 2007 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

One initial study suggested that MAPT pathology increases Aβ pathology.

“with only one initial study suggesting that MAPT pathology increases Aβ pathology (Ribe et al., 2005).”

• cites Ribe et al., 2005 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

The animal studies were approved by Boston University Institutional Animal Care and Use Committee.

“The animal studies were approved by Boston University Institutional Animal Care and Use Committee.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

MAPT pathology co-localizes with the RNA binding protein HNRNPA2B1, which functions as an indirect m6A reader.

“This finding is consistent with the observation that MAPT pathology co-localizes with the RNA binding protein, HNRNPA2B1, which functions as an indirect m6A reader.”

• cites Jiang L. et al., 2021Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Probably not a substantial claim: may not pass S3 (load-bearing).

The studies were conducted in accordance with the local legislation and institutional requirements.

“The studies were conducted in accordance with the local legislation and institutional requirements.”

Probably not a substantial claim: may not pass S2 (checkable).

The APPNL-G-F/MAPTP301S model provides a useful compromise.

“The current APPNL-G-F/MAPTP301S model thus provides a useful compromise.”

• AssumptionRests on an assumption: The APPNL-G-F/MAPTP301S mouse model is a valid model of Alzheimer's disease relevant to human disease.

Probably not a substantial claim: may not pass S4 (based).

The APPNL-G-F/MAPTP301S model avoids artifacts associated with APP or presenilin over-expression.

“This model exhibits strong Aβ accumulation and strong inflammation, while avoiding artifacts associated with APP or presenilin over-expression.”

• AssumptionRests on an assumption: The APPNL-G-F/MAPTP301S mouse model is a valid model of Alzheimer's disease relevant to human disease.

Probably not a substantial claim: may not pass S3 (load-bearing).

The increases observed for m6A parallel those reported by our group previously.

“The increases observed for m6A parallel those reported by our group previously (Jiang L. et al., 2021).”

• cites Jiang L. et al., 2021Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Probably not a substantial claim: may not pass S3 (load-bearing).

Use of the m6A antibody does not provide insight into methylation of individual RNA molecules.

“Use of the m6A antibody does not provide insight into methylation of individual RNA molecules.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Oligomeric MAPT (which is highly phosphorylated) forms a complex with m6A and the m6A reader HNRNPA2B1.

“Our prior work shows that oligomeric MAPT (which is highly phosphorylated) forms a complex with m6A and the m6A reader HNRNPA2B1 (Jiang L. et al., 2021).”

• cites Jiang L. et al., 2021Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

m6A regulates inflammation, macrophages and also astrocytosis.

“Such results are consistent with emerging studies showing that m6A regulates inflammation, macrophages and also astrocytosis (Cockova et al., 2021; Yi et al., 2021; Sun Z. et al., 2022; Sun J. et al., 2022; Wang et al., 2022).”

• cites Cockova et al., 2021 sourceNot checked — no full text of the source
• cites Yi et al., 2021Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

• cites Sun Z. et al., 2022Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.