CoChat
Sign Up
Cited by 110Alzheimer's disease research and treatmentsNeuroinflammation and Neurodegeneration MechanismsDementia and Cognitive Impairment Research

Genetic variability in response to amyloid beta deposition influences Alzheimer’s disease risk

Open access
Salih, Dervis A, Bayram, Sevinc, Guelfi, Sebastian, Reynolds, Regina H, Shoai, Maryam, Ryten, Mina, Brenton, Jonathan W, Zhang, David +10 more·2019·Brain Communications
DOI 10.1093/braincomms/fcz022PMC PMC7145452
Open this paper in CoChat — free accountOpen PDF

Abstract

Genome-wide association studies of late-onset Alzheimer’s disease risk have previously identified genes primarily expressed in microglia that form a transcriptional network. Using transgenic mouse models of amyloid deposition, we previously showed that many of the mouse orthologues of these risk genes are co-expressed and associated with amyloid pathology. In this new study, we generate an improved RNA-seq-derived network that is expressed in amyloid-responsive mouse microglia and we statistically compare this with gene-level variation in previous human Alzheimer’s disease genome-wide association studies to predict at least four new risk genes for the disease (OAS1, LAPTM5, ITGAM/CD11b and LILRB4). Of the mouse orthologues of these genes Oas1a is likely to respond directly to amyloid at the transcriptional level, similarly to established risk gene Trem2, because the increase in Oas1a and Trem2 transcripts in response to amyloid deposition in transgenic mice is significantly higher than both the increase of the average microglial transcript and the increase in microglial number. In contrast, the mouse orthologues of LAPTM5, ITGAM/CD11b and LILRB4 (Laptm5, Itgam/CD11b and Lilra5) show increased transcripts in the presence of amyloid plaques similar in magnitude to the increase of the average microglial transcript and the increase in microglia number, except that Laptm5 and Lilra5 transcripts increase significantly quicker than the average microglial transcript as the plaque load becomes dense. This work suggests that genetic variability in the microglial response to amyloid deposition is a major determinant for Alzheimer’s disease risk, and identification of these genes may help to predict the risk of developing Alzheimer’s disease. These findings also provide further insights into the mechanisms underlying Alzheimer’s disease for potential drug discovery.

AI snapshot

Written from the paper’s full text.

TL;DR

Using RNA-seq from amyloid-responsive mouse microglia, the authors built a co-expression network enriched for human Alzheimer’s disease GWAS genes and identified four new candidate risk genes: OAS1, LAPTM5, ITGAM/CD11b and LILRB4. They propose that genetic variability in the microglial response to amyloid beta deposition is a major determinant of Alzheimer’s disease risk.

Key findings

  • The RNA-seq microglial module correlated strongly with Aβ deposition (correlation = 0.94; P < 3e−41) and contained mouse orthologues of established GWAS loci including TREM2, ABI3, CD33, INPP5D, MS4A6D, SPI1/PU.1, PLCG2, GAL3ST4, RIN3, HLA and APOE.
  • Human Alzheimer’s disease-associated genes were significantly enriched in the mouse amyloid-responsive network at the P = 0.01 threshold (P = 8.86 × 10−6); enrichment remained after excluding established GWAS loci (P = 1.66 × 10−4 for the 147-gene network; P = 3.68 × 10−4 for the 1584-gene module). The tau-responsive microglial network was not significantly enriched for the top 137 genes (P = 0.78), with only modest enrichment for the entire 2299-gene module (P = 1.74 × 10−2).
  • Four predicted genes were associated with Alzheimer’s disease in updated IGAP data: LAPTM5 (gene P = 6.62E−05), OAS1 (1.58E−03), ITGAM (4.92E−03) and LILRB4 (8.96E−03). Oas1a and Trem2 showed larger increases than the average microglial transcript and microglial number increase: Oas1a 10.0-fold, Trem2 9.2-fold, versus a 3.7-fold microglia increase at 18 months; Laptm5 was 4.1-fold, Lilra5 3.8-fold and Itgam/CD11b 2.3-fold.
  • Three colocalizations were identified between Alzheimer’s disease loci and eQTLs regulating OAS1 expression, all in stimulated monocyte/macrophage states.

Methods

  • Reused RNA samples from Matarin et al. (2015) mice expressing one or two copies of APP (Swedish) and/or PSEN1 (M146V) transgenes plus wild-type controls; RNA-seq was performed by Eurofins Genomics, and weighted gene co-expression network analysis was used to build the microglial network.
  • Human GWAS data were from original IGAP (17,008 Alzheimer’s disease cases and 37,154 controls) and updated IGAP (21,982 cases and 41,944 controls); gene-based P-values were derived using MAGMA/Brown’s method with SNPs within exon boundaries, 1000 Genomes Phase 3 LD reference, and genomic inflation correction.
  • Enrichment analysis converted mouse genes to human orthologues, counted observed versus expected significant genes at alpha = 0.05, 0.01 and 0.001, conservatively counted genes within 0.5 Mb as one, and used a Z-test; colocalization used eQTL datasets from baseline and stimulated human monocytes and iPSC-derived macrophages.

Limitations

  • The paper states that mouse models do not show tau tangles or neuronal loss solely in response to rising Aβ, that bulk RNA-seq was used rather than isolated microglia or single-cell work, that microglial purification may alter transcripts, and that human and mouse microglia differ with some genes expressed more abundantly in human microglia.
  • The paper states that individual gene significance is modest compared with genome-wide levels, that the four genes have not previously been reported as having variants significantly associated with Alzheimer’s disease by traditional GWAS, and that there is no published evidence to date that DNA variation in these genes in humans is linked to Alzheimer’s disease risk.
  • Evident from the text: the four candidate genes are predictions from network overlap and gene-based enrichment rather than genome-wide significant SNP hits, and further functional work is required to understand their mechanisms in disease progression.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1093/braincomms/fcz022

Cites

65 of 65 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

0 of 84 cited sources checked · 0 supported · 0 flagged

  • Findingown data

    The analysis of the genetic network expressed by amyloid-responsive microglia identified five genes within the central portion of mouse microglial network whose human orthologues were associated with Alzheimer's disease from original IGAP data.

    “The analysis of the genetic network expressed by amyloid-responsive microglia identified five genes within the central portion of mouse microglial network whose human orthologues were associated with Alzheimer’s disease from the original IGAP data [described in Salih et al. (2018), using the IGAP data from Lambert et al. (2013)].”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites Salih et al. (2018)

      Not checked in the free run.

    • cites Lambert et al. (2013) source

      Not checked in the free run.

  • Findingown data

    These four genes, OAS1, LAPTM5, ITGAM/CD11b and LILRB4, have not been previously reported as having variants significantly associated with Alzheimer's disease using traditional GWAS approaches.

    “These four genes, OAS1, LAPTM5, ITGAM/CD11b and LILRB4, have not been previously reported as having variants significantly associated with Alzheimer’s disease using traditional GWAS approaches (Table 1, Supplementary Figs 3 and 4).”
    Quote in the paperSubstantial claim
  • Findingown data

    The increase in Oas1a and Trem2 transcripts in response to amyloid deposition in transgenic mice is significantly higher than both increase of average microglial transcript and increase in microglial number.

    “because the increase in Oas1a and Trem2 transcripts in response to amyloid deposition in transgenic mice is significantly higher than both the increase of the average microglial transcript and the increase in microglial number.”
    Quote in the paperSubstantial claim
  • Findingown data

    The enrichment remained significant after established GWAS loci were excluded (P=1.66×10−4 for 147-gene network; P=3.68×10−4 for entire 1584-gene module).

    “The enrichment remained significant even after the established GWAS loci were excluded [P = 1.66 × 10−4 for highly connected network of 147 genes (Fig. 1) and similarly P = 3.68 × 10−4 for the entire module of 1584 genes (Supplementary Table 1)].”
    Quote in the paperSubstantial claim
  • Findingown data

    The new higher-resolution transcriptional network containing co-expressed mRNA most strongly correlated to amyloid deposition again featured primarily microglial genes.

    “The new higher-resolution transcriptional network containing the co-expressed mRNA that most strongly correlated to amyloid deposition again featured primarily microglial genes.”
    Quote in the paperNot a substantial claim
  • May combine two claims. 34
  • The quote only partly states this claim. 1
  • Classified as probably not a substantial claim. 20

  • May combine two claims.

    Findingown data

    Expression patterns for Oas1a, Lilra5 and Itgam/CD11b are similar in both homozygous and hemizygous APP/PSEN1 mice, whereas Laptm5 shows an expression pattern in hemizygous mice more similar to Itgam/CD11b.

    “The expression patterns for Oas1a, Lilra5 and Itgam/CD11b are similar in both the homozygous and hemizygous APP/PSEN1 mice (Supplementary Fig. 6), whereas Laptm5 shows an expression pattern in the hemizygous APP/PSEN1 mice that is more similar to Itgam/CD11b.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Expression levels of putative risk genes relative to age-matched wild-type mice show a range: Oas1a 10.0-fold, Laptm5 4.1-fold, Lilra5 3.8-fold, Itgam/CD11b 2.3-fold, compared to Trem2 9.2-fold and Aif1 3.3-fold at 18 months in homozygous APP/PSEN1 mice.

    “The expression levels of our putative risk genes relative to expression in age-matched wild-type mice shows a range (Oas1a, 10.0-fold increase in homozygous APP/PSEN1 mice relative to wild-type at 18 months of age; Laptm5, 4.1-fold increase; Lilra5, 3.8-fold increase; Itgam/CD11b, 2.3-fold increase; compared to Trem2, 9.2-fold increase, and Aif1, 3.3-fold increase; Supplementary Fig. 6).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    The four novel risk genes expressed by microglia that respond to Aβ deposition show transcript levels rising from 4 months of age in homozygous APP/PSEN1 mice and after 4 months in hemizygous APP/PSEN1 mice, in line with increase in microglial numbers as plaques begin to deposit.

    “Furthermore, in common with the existing seven known GWAS-associated genes in Fig. 1, the four novel risk genes we identify that are expressed by microglia that respond to Aβ deposition show transcript levels rising from 4 months of age in the homozygous APP/PSEN1 mice and after 4 months of age in the hemizygous APP/PSEN1 mice (Supplementary Fig. 5), in line with the increase in microglial numbers”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Medawar et al., 2019 source

      Not checked in the free run.

  • May combine two claims.

    Findingown data

    The mouse orthologues of LAPTM5, ITGAM/CD11b and LILRB4 show increased transcripts in presence of amyloid plaques similar in magnitude to increase of average microglial transcript and microglia number, except Laptm5 and Lilra5 increase significantly quicker as plaque load becomes dense.

    “In contrast, the mouse orthologues of LAPTM5, ITGAM/CD11b and LILRB4 (Laptm5, Itgam/CD11b and Lilra5) show increased transcripts in the presence of amyloid plaques similar in magnitude to the increase of the average microglial transcript and the increase in microglia number, except that Laptm5 and Lilra5 transcripts increase significantly quicker than the average microglial transcript as the”
    Quote in the paperSubstantial claim

  • May combine two claims. The quote only partly states this claim.

    Findingown data

    The RNA-seq analysis revealed many additional genes not detectable with microarray, including more genes previously identified as human Alzheimer's disease risk genes from GWAS.

    “but the mouse RNA-seq analysis revealed many additional genes not detectable with microarray, and included yet more genes previously identified as human risk genes for Alzheimer’s disease from genome-wide association studies (GWAS).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingreasoning

    Itgam/CD11b shows a similar change in relative expression to the average relative transcript level of the entire immune network, and to the increase in microglia number, comparable to relative Spi1/PU.1 expression.

    “Itgam/CD11b shows a similar change in relative expression to the average relative transcript level of the entire immune network, and to the increase in microglia number, comparable to relative Spi1/PU.1 expression.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    The expression levels of the putative risk genes relative to age-matched wild-type mice show a range, with Oas1a showing the greatest relative expression (10.0-fold increase) and Itgam/CD11b showing the lowest (2.3-fold increase).

    “The expression levels of our putative risk genes relative to expression in age-matched wild-type mice shows a range, with Oas1a showing the greatest relative expression (10.0-fold increase relative to wild-type), and Itgam/CD11b showing the lowest relative expression (2.3-fold increase)”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Laptm5 and Lilra5 relative expression are only significantly increased in homozygous APP/PSEN1 mice when the plaque load starts to become heavy (8 months of age), suggesting direct regulation by amyloid only as the plaque load increases.

    “Laptm5 and Lilra5 relative expression are only significantly increased in homozygous APP/PSEN1 mice when the plaque load starts to become heavy (8 months of age), suggesting direct regulation by amyloid only as the plaque load increases”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingcitation

    Oas1a shows increased expression in purified microglia from a number of different mouse models of Alzheimer's disease, using the Myeloid Landscape datasets, suggesting Oas1a is directly up-regulated by amyloid.

    “Oas1a shows increased expression in purified microglia from a number of different mouse models of Alzheimer’s disease, using the Myeloid Landscape datasets suggesting Oas1a is directly up-regulated by amyloid”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Friedman et al., 2018 source

      Not checked in the free run.

  • May combine two claims.

    Findingown data

    In this study there is a positive correlation between Spi1/PU.1 and candidate genes Laptm5 and Itgam/CD11b, as well as established risk genes Trem2 and Abi3, which all have binding sites in their promoters for SPI1/PU.1.

    “in our study we see a positive correlation between Spi1/PU.1 and candidate genes Laptm5 and Itgam/CD11b, as well as established risk genes Trem2 and Abi3, which all have binding sites in their promoters for SPI1/PU.1”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    The authors' data show that microglia respond differently to amyloid deposition versus tauopathy, with around 29% of transcripts in amyloid-responsive microglia showing a stronger correlation to amyloid pathology.

    “Our data also show that microglia respond differently to amyloid deposition versus tauopathy, with around 29% of transcripts in amyloid-responsive microglia showing a stronger correlation to amyloid pathology.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationreasoning

    The individual gene significance is modest compared to genome-wide levels, but the genes are statistically significant and, together with previously identified Alzheimer's disease genes, form the core of a transcriptional gene network.

    “Hence the individual gene significance is modest as compared to the genome-wide levels, but the genes are statistically significant and, together with previously identified Alzheimer’s disease genes, form the core of a transcriptional gene network (Fig. 1 and Table 1).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Premisecitation

    Trem2 expression showed a strong correlation with expression of a network of innate immune genes, suggesting Trem2 is a hub gene and may regulate the entire network.

    “Trem2 expression also showed a strong correlation with the expression of a network of genes in the innate immune system suggesting Trem2 is a ‘hub’ gene, and may regulate the expression of the entire network.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Matarin et al., 2015 source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    The immune module of genes showed a remarkable positive correlation to amyloid deposition and contained orthologues of established Alzheimer's disease risk genes such as Abca7 and Ms4a6d.

    “This immune module of genes showed a remarkable positive correlation to amyloid deposition (Matarin et al., 2015), and contained orthologues of other already established Alzheimer’s disease risk genes such as Abca7 and Ms4a6d (Lambert et al., 2013).”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites Matarin et al., 2015 source

      Not checked in the free run.

    • cites Lambert et al., 2013 source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    ABI3 and PLCG2 were present in this network and were subsequently identified as associated with Alzheimer's disease risk loci.

    “Notably, two genes, ABI3 and PLCG2, that were identified subsequently as being associated with Alzheimer’s disease risk loci (Sims et al., 2017), were also present in this network.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Sims et al., 2017 source

      Not checked in the free run.

  • May combine two claims.

    Interpretationown data

    Trem2 forms a hub gene in the network using either technique, indicating Trem2 expression is highly correlated to many other genes and may drive the response of this network.

    “Trem2 forms a hub gene in our network, using either technique, indicating that Trem2 expression is highly correlated to many other genes in the network, and may drive the response of this network.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim.

    Study design and conductcitation

    Prior to gene-based analyses all individual SNP P-values were corrected for genomic inflation factor (λ) to normalise for unaccounted variation such as population stratification.

    “Prior to the gene-based analyses all individual SNP P-values were corrected for the genomic inflation factor (λ; to normalise for unaccounted variation, due to factors such as population stratification; Devlin and Roeder, 1999).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Devlin and Roeder, 1999 source

      Not checked in the free run.

  • May combine two claims.

    Interpretationown data

    Laptm5 and Lilra5 relative expression are only significantly increased relative to average transcript level of entire network when plaque load starts to become heavy (8 months), but returns to average as disease progresses, suggesting a role in initial response to Aβ.

    “In contrast, Laptm5 and Lilra5 relative expression are only significantly increased relative to average transcript level of the entire network when the plaque load starts to become heavy (8 months of age), but returns to the average relative transcript level of the network as disease progresses, suggesting a role in the initial response to Aβ (Supplementary Fig. 6).”
    Quote in the paperNot a substantial claim

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

    Study design and conductcitation

    Weighted gene co-expression network analyses were performed using recommended parameters from original analysis developers.

    “Weighted gene co-expression network analyses were performed as described in Matarin et al. (2015), using the recommended parameters from the original analysis developers (Zhang and Horvath, 2005; Horvath et al., 2006; Oldham et al., 2006; Langfelder and Horvath, 2008).”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites Zhang and Horvath, 2005 source

      Not checked in the free run.

    • cites Horvath et al., 2006 source

      Not checked in the free run.

    • cites Oldham et al., 2006 source

      Not checked in the free run.

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

    Study design and conductnone given

    Mouse procedures were performed in agreement with UK Animals (Scientific Procedures) Act 1986, with local ethical agreement.

    “The mice procedures used for Matarin et al. (2015), were performed in agreement with the UK Animals (Scientific Procedures) Act, 1986, with local ethical agreement.”
    Quote in the paperNot a substantial claim

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

    Premisecitation

    A decade of GWAS projects for Alzheimer's disease has provided insights into the progression of late-onset Alzheimer's disease, particularly the dependence on the innate immune system, with the identification of genes such as TREM2 and SPI1/PU.1.

    “A decade of GWAS projects for Alzheimer’s disease has provided key and initially surprising insights into the progression of late-onset Alzheimer’s disease, particularly the dependence on the innate immune system, with the identification of genes such as TREM2 and SPI1/PU.1”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites Guerreiro et al., 2013 source

      Not checked in the free run.

    • cites Jonsson et al., 2013 source

      Not checked in the free run.

    • cites Huang et al., 2017 source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Interferon-related genes are expressed in ageing control mice, and their expression is further elevated in mouse models with amyloid pathology.

    “Recent studies have also shown that interferon-related genes are expressed in ageing control mice, and that the expression of interferon-related genes is further elevated in mouse models with amyloid pathology (Friedman et al., 2018; Sala Frigerio et al., 2019)”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites Friedman et al., 2018 source

      Not checked in the free run.

    • cites Sala Frigerio et al., 2019 source

      Not checked in the free run.

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

    Interpretationreasoning

    The discrepancy between Huang et al. and this study may be due to differences in the increase in microglial number between mice and humans.

    “This discrepancy may be due to differences in the increase in microglial number between mice and humans”
    Quote in the paperNot a substantial claim

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

    Premisecitation

    OAS1 can activate ribonuclease L, which degrades viral RNA and inhibits viral replication.

    “OAS1 can additionally activate ribonuclease L, which degrades viral RNA and inhibits viral replication (Donovan et al., 2013).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Donovan et al., 2013 source

      Not checked in the free run.

  • May combine two claims.

    Interpretationown data

    SPI1/PU.1 is a positive regulator of these genes in this mouse model where heavy amyloid load does not lead to tangles and neurodegeneration.

    “suggesting that SPI1/PU.1 is a positive regulator of these genes in this mouse model where heavy amyloid load does not lead to tangles and neurodegeneration.”
    Quote in the paperSubstantial claim

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

    Premisecitation

    A common SNP in the population delays onset of Alzheimer's disease, purportedly via reduced expression of SPI1/PU.1.

    “The study by Huang et al. (2017) shows that a common SNP in the population delays onset of Alzheimer’s disease, purportedly via reduced expression of SPI1/PU.1.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Huang et al., 2017 source

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Study design and conductnone given

    Mouse models for dementia do not show tau tangles or neuronal loss solely in response to rising Aβ.

    “Although mouse models for dementia have clear limitations in that they do not show tau tangles or neuronal loss solely in response to rising Aβ”
    Quote in the paperSubstantial claim

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

    Premisecitation

    Previous analyses including 10 kb flanking sequence did not cause substantial differences to gene-based P-values.

    “Previous analyses including the 10 kb sequence flanking the first and last exons of genes, which may contain potential regulatory SNPs, did not cause substantial differences to the gene-based P-values (Escott-Price et al., 2014).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Escott-Price et al., 2014 source

      Not checked in the free run.

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

    Premisecitation

    Microglial number is increased in homozygous APP/PSEN1 mice by around 3.7-fold in the CA1 region of the hippocampus.

    “We have previously shown that microglial number is increased in these homozygous APP/PSEN1 mice, by around 3.7-fold in the CA1 region of the hippocampus (Medawar et al., 2019)”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Medawar et al., 2019 source

      Not checked in the free run.

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

    Premisecitation

    The authors previously described a human microglial network containing LAPTM5, ITGAM/CD11b and LILRB4.

    “we described a human microglial network containing LAPTM5, ITGAM/CD11b and LILRB4 (Forabosco et al., 2013)”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Forabosco et al., 2013 source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    ITGAM/CD11b was highlighted in recent genetic and functional analyses as likely being important for the progression of Alzheimer's disease, with expression driven by SPI1/PU.1 and related to amyloid deposition in mice and humans.

    “ITGAM/CD11b was highlighted in recent genetic and functional analyses as likely being important for the progression of Alzheimer’s disease, whose expression was driven by SPI1/PU.1, and related to amyloid deposition in mice and humans (Zhang et al., 2013; Hong et al., 2016; Kamphuis et al., 2016; Olmos-Alonso et al., 2016; Huang et al., 2017; Nam et al., 2018).”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites Zhang et al., 2013 source

      Not checked in the free run.

    • cites Hong et al., 2016 source

      Not checked in the free run.

    • cites Huang et al., 2017 source

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Study design and conductcitation

    The gene-wide analysis controlled for linkage disequilibrium and different numbers of SNPs per gene using a statistical approach by Brown (1975), adopted by Moskvina et al. (2011) and de Leeuw et al. (2015).

    “The gene-wide analysis was performed based on the summary P-values while controlling for linkage disequilibrium and different numbers of SNPs per gene using a statistical approach by Brown (1975), and adopted for set-based analysis of genetic data by Moskvina et al. (2011) and de Leeuw et al. (2015).”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites Brown (1975) source

      Not checked in the free run.

    • cites Moskvina et al. (2011) source

      Not checked in the free run.

    • cites de Leeuw et al. (2015) source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    LILRB2 and its mouse orthologue Pirb bind Aβ, and this interaction with Aβ in mice mediates synapse elimination, and deficits in synaptic plasticity and memory.

    “A paralogue of LILRB4, named LILRB2, and its mouse orthologue Pirb have been shown to bind Aβ, and this interaction with Aβ in mice mediates synapse elimination, and deficits in synaptic plasticity and memory (Kim et al., 2013).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Kim et al., 2013 source

      Not checked in the free run.

  • May combine two claims.

    Interpretationreasoning

    This sub-network of genes is regulated together, in part by the SPI1/PU.1 transcription factor, and may function together.

    “and so suggests that this sub-network of genes are regulated together, in part by the SPI1/PU.1 transcription factor, and may function together.”
    Quote in the paperSubstantial claim

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

    Premisecitation

    The latest GWAS studies published during 2019 identified 9 novel risk loci from 71,880 Alzheimer's disease cases and 5 novel risk loci from 35,274 clinically assessed Alzheimer's disease cases in the updated IGAP study.

    “The latest GWAS studies published during 2019 mark the largest of their kind for Alzheimer’s disease featuring 71 880 Alzheimer’s disease cases to identify 9 novel risk loci (Jansen et al., 2019), and 35 274 clinically assessed Alzheimer’s disease cases to identify 5 novel risk loci from the updated IGAP study (Kunkle et al., 2019).”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites Jansen et al., 2019 source

      Not checked in the free run.

    • cites Kunkle et al., 2019 source

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Study design and conductnone given

    To account for linkage disequilibrium, genes within 0.5 Mb of each other were conservatively counted as one.

    “To account for linkage disequilibrium, the genes within 0.5 Mb of each other were conservatively counted as one.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data

    Spi1/PU.1 may be regulating microglial number, and the level of microglial proliferation that can be tolerated by mice and humans may be different.

    “our data suggest that Spi1/PU.1 may be regulating microglial number, and it is possible that the level of microglial proliferation that can be tolerated by mice and humans is different”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim.

    Premisecitation

    OAS1 is induced by interferons.

    “OAS1 is induced by interferons (Donovan et al., 2013)”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Donovan et al., 2013 source

      Not checked in the free run.

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

    Premisecitation

    ITGAM/CD11b is a cell surface receptor involved in activation, migration and phagocytosis of immune cells, and is used as a marker of activated microglia.

    “ITGAM/CD11b (or CR3A), is a cell surface receptor involved in activation, migration and phagocytosis of immune cells, so much so that ITGAM/CD11b is used as a marker of activated microglia (Matsuoka et al., 2001; Heneka et al., 2013; Kamphuis et al., 2016).”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites Matsuoka et al., 2001 source

      Not checked in the free run.

    • cites Heneka et al., 2013 source

      Not checked in the free run.

    • cites Kamphuis et al., 2016 source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    LILRB4 orthologues are increased with amyloid deposition and specifically associated with amyloid plaques.

    “LILRB4 (leukocyte immunoglobulin-like receptor, subfamily B, member 4), orthologues have also been shown to be increased with amyloid deposition and specifically associated with amyloid plaques (Wirz et al., 2013; Kamphuis et al., 2016; Castillo et al., 2017).”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites Wirz et al., 2013 source

      Not checked in the free run.

    • cites Kamphuis et al., 2016 source

      Not checked in the free run.

    • cites Castillo et al., 2017 source

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Premisecitation

    Most GWAS loci are thought to operate by regulating the expression of neighbouring genes.

    “Since most GWAS loci are thought to operate by regulating the expression of neighbouring genes (Bradshaw et al., 2013; Griciuc et al., 2013; Huang et al., 2017)”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites Bradshaw et al., 2013 source

      Not checked in the free run.

    • cites Griciuc et al., 2013 source

      Not checked in the free run.

    • cites Huang et al., 2017 source

      Not checked in the free run.

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

    Premisecitation

    OAS1 is involved in the regulation of cytokine expression.

    “OAS1 (2-prime, 5-prime oligoadenylate synthetase 1) is involved in the regulation of cytokine expression (Lee et al., 2019).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Lee et al., 2019 source

      Not checked in the free run.

  • May combine two claims.

    Interpretationcitation

    Genes showing higher relative transcript levels such as Oas1a and Trem2 are likely to be directly up-regulated by microglia in response to amyloid, and may be promoting a protective response to amyloid.

    “Genes showing higher relative transcript levels such as Oas1a and Trem2 are likely to be directly up-regulated by microglia in response to amyloid, and may be promoting a protective response to amyloid e.g. as described by Lee et al. (2018).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Lee et al., 2018 source

      Not checked in the free run.

  • May combine two claims.

    Interpretationcitation

    These findings are consistent with previous work on Trem2 and CD33, suggesting these risk genes are crucial in controlling the microglial response to amyloid-induced damage.

    “These findings are also consistent with previous work on Trem2 (Wang et al., 2015; Keren-Shaul et al., 2017; Mazaheri et al., 2017; Cheng-Hathaway et al., 2018; Lee et al., 2018), and CD33 (Bradshaw et al., 2013; Griciuc et al., 2013), suggesting these risk genes are crucial in controlling the microglial response to amyloid-induced damage.”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites Wang et al., 2015 source

      Not checked in the free run.

    • cites Bradshaw et al., 2013 source

      Not checked in the free run.

    • cites Griciuc et al., 2013 source

      Not checked in the free run.

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

    Premisecitation

    A number of important studies have shown a significant overlap between the transcriptomes of mouse and human microglia, but also that a number of genes are expressed selectively more abundantly in human microglia.

    “a number of important studies have compared gene expression in microglia from mice and humans, and while they have shown a significant overlap between the transcriptomes of the two species, they have also seen a number of genes are expressed selectively more abundantly in human microglia”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites Miller et al., 2010 source

      Not checked in the free run.

    • cites Galatro et al., 2017 source

      Not checked in the free run.

    • cites Gosselin et al., 2017 source

      Not checked in the free run.

  • May combine two claims.

    Interpretationreasoning

    Not all Alzheimer's disease risk genes have SPI1/PU.1 binding sites; thus, while this core transcription factor plays a substantial role in the progression of disease, there are likely to be auxiliary, environment-dependent transcription factors that modify disease development.

    “Not all Alzheimer’s disease risk genes have SPI1/PU.1 binding sites; thus, while this core transcription factor plays a substantial role in the progression of disease, there are likely to be auxiliary, environment-dependent transcription factors that modify disease development.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationreasoning

    This work is complementary to Huang et al. (2017), showing that microglial SPI1/PU.1-driven transcription is a common feature of many Alzheimer's disease loci.

    “this work is complementary to that of Huang et al. (2017), showing that microglial SPI1/PU.1-driven transcription is a common feature of many Alzheimer’s disease loci.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Huang et al., 2017 source

      Not checked in the free run.

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

    Premisecitation

    A recent study identified a co-expression module within microglia that respond more robustly to amyloid pathology compared to tauopathy.

    “A recent study also presents related data, identifying a co-expression module within microglia that respond more robustly to amyloid pathology compared to tauopathy (Sierksma et al., 2019).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Sierksma et al., 2019

      Not checked in the free run.

48 of 48

  • The paper relies on this without support.

    The mouse co-expression network expressed by amyloid-responsive microglia is sufficiently similar to the human microglial response to Alzheimer's disease to predict human risk genes.

    “verifying the relevance of this gene network to the human condition”

  • The paper relies on this without support.

    The average microglial transcript level is an appropriate baseline for identifying genes directly up-regulated by amyloid beyond changes in microglial number.

    “compared to the average transcript level relative to wild-type mice for the entire innate immune network throughout disease progression”

  • The paper relies on this without support.

    GWAS-discovered risk genes do not account for all of the heritability of late-onset Alzheimer’s disease.

    “Despite all the risk genes that have been discovered by GWAS, they still do not account for all of the heritability of late-onset Alzheimer’s disease.”

  • The paper relies on this without support.

    The remaining risk genes are likely to be of rare mutation frequency or lower effect size.

    “as the remaining risk genes are likely to be of rare mutation frequency or lower effect size”

  • The paper relies on this without support.

    There is no published evidence to date that DNA variation in these putative risk genes in the human population is linked to risk for Alzheimer’s disease.

    “there is no published evidence to date that DNA variation in these genes in the human population is linked to risk for Alzheimer’s disease.”
5 of 5

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • Salih et al. identified a network of microglial genes expressed by amyloid-associated microglia, collated from five transgenic mouse lines, including Spi1.

    “Salih et al. identified a network of microglial genes expressed by amyloid-associated microglia, collated from five transgenic mouse lines, including Spi1, and 74 other genes that were identified as Spi1 dose-sensitive in this study52.”

    From Modest changes in Spi1 dosage reveal the potential for altered microglial function as seen in Alzheimer’s disease (2021) · cites this paper as [52]

    The Results section (‘High-resolution co-expression network using RNA-seq in amyloid-responsive microglia’; page unavailable) reports an amyloid-associated microglial gene network that includes Spi1/PU.1. It describes mice carrying APP and/or PSEN1 transgenes, but our copy does not establish that the network was collated from five transgenic mouse lines. (LLM, conf=0.84)

    Experimental — accuracy not yet measured

  • Using the expression of 11 mouse orthologs of human AD GWAS loci genes, wild-derived mice are transcriptionally distinct from B6 mice regardless of tau genotype.

    “Using the expression of these 11 genes, wild-derived mice are transcriptionally distinct from B6 mice regardless of tau genotype (Fig. S5 B).”

    From Network analysis identifies strain-dependent response to tau and tau seeding-associated genes (2023) · cites this paper as Salih et al., 2019

    The Results section (“High-resolution co-expression network using RNA-seq in amyloid-responsive microglia”; page unavailable) identifies 11 mouse orthologues of human Alzheimer’s disease GWAS loci. It does not report that their expression distinguishes wild-derived mice from B6 mice, with or without a tau genotype; its comparisons concern APP/PSEN1 mice and wild-type controls. (LLM, conf=0.83)

    Experimental — accuracy not yet measured

  • A previous study identified 11 mouse orthologs of genes within human AD GWAS loci that were co-expressed in mouse models of amyloid deposition.

    “First, a previous study identified 11 mouse orthologs of genes within human AD genome-wide association study (GWAS) loci that were co-expressed in mouse models of amyloid deposition (Salih et al., 2019).”

    From Network analysis identifies strain-dependent response to tau and tau seeding-associated genes (2023) · cites this paper as Salih et al., 2019

    The Results section describes a co-expression network in amyloid-responsive mouse microglia containing seven established AD GWAS-associated genes and four newly proposed risk genes—11 genes in total (page unavailable). But it says the four new genes had not previously been reported as significantly associated with AD by traditional GWAS, so it does not support calling all 11 genes within established human AD GWAS loci. (LLM, conf=0.88)

    Experimental — accuracy not yet measured

  • 11 mouse genes previously reported as orthologs of genes within human AD GWAS loci were selected.

    “To characterize whether genes at known AD risk loci were responsible for the effect of genetic background, we selected 11 mouse genes previously reported as orthologs of genes within human AD GWAS loci (Salih et al., 2019).”

    From Network analysis identifies strain-dependent response to tau and tau seeding-associated genes (2023) · cites this paper as Salih et al., 2019

    The Results section (page unavailable) lists seven mouse genes whose human orthologues are established Alzheimer’s GWAS genes. It also describes four additional orthologues associated with Alzheimer’s disease in a gene-based analysis, but explicitly says those four had not previously been reported as significantly associated using traditional GWAS. The source therefore supports the orthologue relationship, not the characterization of all 11 as genes within previously known GWAS loci. (LLM, conf=0.86)

    Experimental — accuracy not yet measured

  • IL-4 treatment significantly up-regulates expression of C1qa, Cd68 and Igf1.

    “Expression of C1qa, Cd68 and Igf1 were all found to be significantly up-regulated with IL-4 treatment, which is similar to what we saw in transgenic mice in correlation with amyloid plaque load (43,44) (www.mouseac.org).”

    From <i>Trem2</i> promotes anti-inflammatory responses in microglia and is suppressed under pro-inflammatory conditions (2020) · cites this paper as [44]

Cites

Flagged first, then by citations

Cited by

Sorted by citations

Related

Authors
Salih, D. A., Bayram, S., Guelfi, S., Reynolds, R. H., Shoai, M., Ryten, M., Brenton, J. W., Zhang, D., Matarin, M., Botia, J. A., Shah, R., Brookes, K. J., Guetta-Baranes, T., Morgan, K., Bellou, E., Cummings, D. M., Escott-Price, V., & Hardy, J.
Affiliations
  • UK Dementia Research Institute (GB)
  • Hitachi Rail Europe Ltd, Holborn, London, UK
  • UCL Queen Square Institute of Neurology (GB)
  • University College London (GB)
  • Department of Neurodegenerative Diseases, Institute of Neurology, UCL, 1 Wakefield Street, London WC1N 1PJ, UK
  • University of Nottingham (GB)
  • Human Genetics, School of Life Sciences, Life Sciences Building, University Park, University of Nottingham, Nottingham NG7 2RD, UK
  • Cardiff University (GB)
  • Dementia Research Institute, MRC Centre for Neuropsychiatric Genetics and Genomics, Cardiff University, UK
Corresponding author
  • Salih, Dervis A
  • Hardy, John
ORCID
Venue
Brain Communications
Publisher
Oxford University Press (OUP)
Published
2019-01-01
Volume
1
Issue
1
Pages
fcz022-fcz022
ISSN
2632-1297 (ISSN-L)
DOI
10.1093/braincomms/fcz022
PMC
PMC7145452
OpenAlex
W2979757494
Semantic Scholar
63c87f8cb1fa9e4867d156fb7806856c83d5b6a4
Cited by
110
References
65
Citations by year
2026: 8 · 2025: 20 · 2024: 16 · 2023: 16 · 2022: 20 · 2021: 16 · 2020: 11 · 2019: 3
FWCI
8.58
Type
journal article
Language
en
Open access
Licence
Locations
  • Brain Communications (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • Repository@Nottingham (University of Nottingham) (publishedVersion, cc-by)
  • ORCA Online Research @Cardiff (Cardiff University) (acceptedVersion)
  • Nottingham Trent University's Institutional Repository (Nottingham Trent Repository) (acceptedVersion)
  • Repository@Nottingham (University of Nottingham) (submittedVersion, cc-by)
  • UCL Discovery (University College London) (submittedVersion)
Funders
  • UK Dementia Research Institute
  • UK Medical Research Council
  • Alzheimer’s Society and Alzheimer’s Research UK
  • Cure Alzheimer’s Fund
  • Dolby Foundation
  • National Institute for Health Research University College London Hospitals Biomedical Research Centre
  • Leonard Wolfson Doctoral Training Fellowship in Neurodegeneration
  • Tenure-track Clinician Scientist Fellowship (MR/N008324/1)
  • University of Nottingham Group is funded by ARUK
  • ARUK Consortium DNA Bank
  • UCL DRI
Topics
  • Health Sciences › Medicine › Physiology › Alzheimer's disease research and treatments
  • Life Sciences › Neuroscience › Neurology › Neuroinflammation and Neurodegeneration Mechanisms
  • Health Sciences › Medicine › Psychiatry and Mental health › Dementia and Cognitive Impairment Research
Keywords
Alzheimer's disease, amyloid-beta deposition, microglia, OAS1, LAPTM5, LILRB4, TREM2, gene co-expression network, transgenic mouse model, amyloid plaques

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-07.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-07.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review