CoChat
Sign Up
Cited by 1,001Alzheimer's disease research and treatmentsDementia and Cognitive Impairment ResearchAdvanced Proteomics Techniques and Applications

Large-scale proteomic analysis of Alzheimer’s disease brain and cerebrospinal fluid reveals early changes in energy metabolism associated with microglia and astrocyte activation

Open access
Johnson, Erik C. B., Dammer, Eric B., Duong, Duc M., Ping, Lingyan, Zhou, Maotian, Yin, Luming, Higginbotham, Lenora A., Guajardo, Andrew +25 more·2020·Nature Medicine
DOI 10.1038/s41591-020-0815-6PMC PMC7405761

Published version of a preprint

Open this paper in CoChat — free accountOpen PDF

Abstract

Our understanding of Alzheimer’s disease (AD) pathophysiology remains incomplete. Here we used quantitative mass spectrometry and coexpression network analysis to conduct the largest proteomic study thus far on AD. A protein network module linked to sugar metabolism emerged as one of the modules most significantly associated with AD pathology and cognitive impairment. This module was enriched in AD genetic risk factors and in microglia and astrocyte protein markers associated with an anti-inflammatory state, suggesting that the biological functions it represents serve a protective role in AD. Proteins from this module were elevated in cerebrospinal fluid in early stages of the disease. In this study of >2,000 brains and nearly 400 cerebrospinal fluid samples by quantitative proteomics, we identify proteins and biological processes in AD brains that may serve as therapeutic targets and fluid biomarkers for the disease. Large-scale, comprehensive proteomic profiling of Alzheimer’s disease brain and cerebrospinal fluid reveals disease-associated protein coexpression modules and highlights the importance of glia and energy metabolism in disease pathogenesis.

Integrity signals

Registry

Crossref · 10.1038/s41591-020-0815-6

Cites

156 of 156 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

0 of 39 cited sources checked · 0 supported · 0 flagged

  • Findingown data

    The M4 module was enriched in microglial and astrocyte protein markers associated with an anti-inflammatory state.

    “This module was enriched in AD genetic risk factors and in microglia and astrocyte protein markers associated with an anti-inflammatory state, suggesting that the biological functions it represents serve a protective role in AD.”
    • Also stated in: § The M4 Astrocyte/Microglial Metabolism Module is Enriched in AD Genetic Risk Factors and Markers of Anti-Inflammatory Disease-Associated Microglia · plain
      “Microglial protein markers that are increased in response to amyloid-β plaques but decreased in response to LPS—or markers of anti-inflammatory disease-associated microglia27—were significantly enriched in the M4 module.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [27] Rangaraju, S. et al. Identification and therapeutic modulation of a pro-inflammatory subset of disease-associated-microglia in Alzheimer’s disease. Mol. Neurodegener. 13, 24 (2018). source

      Not checked in the free run.

  • Findingown data

    A protein co-expression module related to sugar metabolism was among the modules most strongly associated with Alzheimer's disease pathology and cognitive impairment.

    “A protein network module linked to sugar metabolism emerged as one of the most significantly associated modules to AD pathology and cognitive impairment.”
    Quote in the paperSubstantial claim
  • Findingown data

    Proteins from the M4 module were elevated in cerebrospinal fluid in early stages of Alzheimer's disease.

    “Proteins from this module were elevated in cerebrospinal fluid in early stages of the disease.”
    • Also stated in: § M4 Astrocyte/Microglial Metabolism Proteins Are Increased in Cerebrospinal Fluid · plain
      “A number of these proteins were elevated in AsymAD and AD, including M4 hub proteins CD44, PRDX1, and DDAH2.”
    Quote in the paperSubstantial claim
  • Findingown data

    The M4 sugar metabolism module was enriched in Alzheimer's disease genetic risk factors.

    “This module was enriched in AD genetic risk factors and in microglia and astrocyte protein markers associated with an anti-inflammatory state, suggesting that the biological functions it represents serve a protective role in AD.”
    • Also stated in: § The M4 Astrocyte/Microglial Metabolism Module is Enriched in AD Genetic Risk Factors and Markers of Anti-Inflammatory Disease-Associated Microglia · plain
      “We found that the M2 myelin/oligodendrocyte and M4 astrocyte/microglial metabolism modules were significantly enriched in gene products contained within AD risk factor loci (Figure 5A), suggesting that the biological functions or processes reflected by these protein co-expression modules may serve causative roles in AD.”
    Quote in the paperSubstantial claim
  • Findingcitation

    In a prior proteomic study, 7 out of the top 10 plaque-associated proteins most significantly decreased in rapidly-progressive Alzheimer's disease are found in the M4 module, including hubs MSN and PLEC.

    “Notably, when we compare our findings to a prior proteomic study that quantified levels of plaque-associated proteins in normal versus rapidly-progressive AD45, 7 out of the top 10 plaque-associated proteins most significantly decreased in rapidly-progressive AD are found in the M4 module, including M4 hubs MSN and PLEC.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [45] Drummond, E. et al. Proteomic differences in amyloid plaques in rapidly progressive and sporadic Alzheimer’s disease. Acta Neuropathol. 133, 933–954 (2017). source

      Not checked in the free run.

  • May combine two claims. 36
  • Classified as probably not a substantial claim. 32

  • May combine two claims.

    Findingown data

    Increases in expression level of the M4 module are observed with aging but are stronger in Alzheimer's disease, reflecting shared biology between normal aging and Alzheimer's disease.

    “Increases in expression level of the M4 astrocyte/microglial metabolism module are observed with aging, but are stronger in AD, reflecting shared biology between “normal” aging and AD.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    M1 synapse/neuron and M4 astrocyte/microglial metabolism modules showed significant changes in FTLD-TDP and CBD cases similar to Alzheimer's disease, whereas M3 mitochondrial and M10 RNA binding/splicing modules showed more mixed changes across other diseases.

    “We found that the M1 synapse/neuron and M4 astrocyte/microglial metabolism modules showed significant changes in FTLD-TDP and CBD cases, similar to AD, whereas the M3 mitochondrial and M10 RNA binding/splicing modules showed more mixed changes across other diseases.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Other modules affected by aging included M6 cytoskeleton, M7 translation/ribosome, and M9 translation/ribosome.

    “Other modules that appeared to be affected by aging included the M6 cytoskeleton, M7 translation/ribosome, and M9 translation/ribosome modules (Supplementary Figure 9).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    For proteins overlapping with Bai et al.'s validated differentially expressed set after cell type regression, a majority are also differentially expressed in Alzheimer's disease and map to the M4 module.

    “However, for those proteins in this study that overlap with the validated set of differentially expressed proteins in Bai et al. obtained after cell type regression, a majority are also differentially expressed in AD and map to the M4 module.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [35] Bai, B. et al. Deep multilayer brain proteomics identifies molecular networks in Alzheimer’s disease progression. Neuron 105, 975–991.e7 (2020). source

      Not checked in the free run.

  • May combine two claims.

    Findingown data

    Microglial protein markers increased in response to amyloid-β plaques but decreased in response to LPS-markers of anti-inflammatory disease-associated microglia-were significantly enriched in the M4 module.

    “Microglial protein markers that are increased in response to amyloid-β plaques but decreased in response to LPS—or markers of anti-inflammatory disease-associated microglia27—were significantly enriched in the M4 module.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [27] Rangaraju, S. et al. Identification and therapeutic modulation of a pro-inflammatory subset of disease-associated-microglia in Alzheimer’s disease. Mol. Neurodegener. 13, 24 (2018). source

      Not checked in the free run.

  • May combine two claims.

    Findingown data

    All consensus Alzheimer's disease network modules derived from dorsolateral prefrontal cortex were preserved in temporal cortex, and twelve of thirteen were preserved in precuneus.

    “We found that all consensus AD network modules derived from DLPFC were preserved in temporal cortex, and twelve out of the thirteen modules were preserved in precuneus.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingreasoning

    Many M4 proteins are elevated in microglia undergoing plaque phagocytosis, consistent with the strong association of M4 expression with CERAD score.

    “Many M4 proteins are elevated in microglia that are undergoing plaque phagocytosis, which is consistent with the strong association of M4 expression with CERAD score.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    APOE genotype did not have strong effects on module eigenproteins except APOE ε2 on the M2 myelin/oligodendrocyte module, which appeared to attenuate Alzheimer's disease changes.

    “We also assessed the influence of APOE genotype on module eigenproteins, but did not find strong effects except for the APOE ε2 allele on the M2 myelin/oligodendrocyte module, which appeared to attenuate the observed changes in AD (Supplementary Figure 1).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    In cerebrospinal fluid Cohort 2 (96 control, asymptomatic Alzheimer's disease, and Alzheimer's disease), 27 proteins mapped to M4; 17 overlapped with Cohort 1 and showed the same direction of change in Alzheimer's disease cerebrospinal fluid.

    “In Cohort 2, 27 proteins mapped to the M4 astrocyte/microglial metabolism module in brain (Supplementary Figure 14). Of these 27, 17 overlapped with M4 proteins measured in discovery Cohort 1, and showed the same direction of change in AD CSF.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    In normal aging dorsolateral prefrontal cortex, M1 synapse/neuron and M4 astrocyte/microglial metabolism modules decreased and increased with aging, respectively, while M3 mitochondrial and M10 RNA binding/splicing modules were not affected by aging.

    “We found that the M1 synapse/neuron and M4 astrocyte/microglial metabolism modules decreased and increased with aging, respectively, while the M3 mitochondrial and M10 RNA binding/splicing modules were not affected by aging.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    In Cohort 2, many M4 proteins showed significant or trend elevations in asymptomatic Alzheimer's disease, including CD44, LDHB, and PKM, and correlated with cognitive function.

    “In addition, many also showed significant or trend elevations in AsymAD, including CD44, LDHB, and PKM, and correlated with cognitive function (Figure 6C).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Expression of the M4 astrocyte/microglia metabolism module is increased with progression from normal to Alzheimer's disease, and a majority of the most significantly increased proteins in Alzheimer's disease are members of this module.

    “Although expression of the M4 astrocyte/microglia metabolism module is increased with progression from a normal to an AD disease state, and a majority of the most significantly increased proteins in AD are members of this module (Extended Data Figure 7), it is unclear whether these glial responses are deleterious or protective.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    In cerebrospinal fluid Cohort 1 (297 controls and Alzheimer's disease patients), 22 proteins mapped to the M4 module; all showed either increase in Alzheimer's disease or no change, with 10 reaching p<0.05.

    “In Cohort 1, we observed 22 proteins that mapped to the M4 astrocyte/microglial metabolism module in brain (Supplementary Figure 13). All of them showed either an increase in AD or no change, with 10 reaching statistical significance at p < 0.05.”
    Quote in the paperSubstantial claim

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

    Findingown data

    The modules ranged in size from 254 proteins (M1) to 20 proteins (M13).

    “The modules ranged in size from 254 proteins (M1) to 20 proteins (M13).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    Synthetic module eigenprotein values showed similar differences between case groups in temporal cortex and precuneus, with changes in asymptomatic Alzheimer's disease more pronounced in precuneus than dorsolateral prefrontal cortex.

    “Analysis of synthetic module eigenprotein values by case status showed similar differences between and among case groups in temporal cortex (Figure 2D, Supplementary Figure 7) and precuneus (Figure 2E, Supplementary Figure 8) brain regions, with changes in AsymAD more pronounced in precuneus than in DLPFC.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    The M4 sugar metabolism module showed the strongest Alzheimer's disease trait correlations among modules, with cognition r=−0.67, neurofibrillary tangle r=0.49, amyloid-β plaque r=0.46, and functional status r=0.52.

    “The M4 sugar metabolism module showed the strongest AD trait correlations (cognition r=−0.67, p=8.5e−23; neurofibrillary tangle r=0.49, p=4.7e−27; amyloid-β plaque r=0.46, p=1.3e−23; functional status r=0.52, p=2.6e−12).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    The protein co-expression families most strongly correlated to disease reflect synaptic, mitochondrial, RNA binding/splicing, and astrocyte/microglial metabolism biological functions, with astrocyte/microglial metabolism most significantly associated with Alzheimer's disease compared to other biological processes and functions.

    “We find that the protein co-expression families most strongly correlated to disease reflect synaptic, mitochondrial, RNA binding/splicing, and astrocyte/microglial metabolism biological functions, with astrocyte/microglial metabolism most significantly associated with AD compared to other biological processes and functions.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    In ACT, the white matter principal component explained 27% of variance; in the other three cohorts, the white matter marker-correlated principal component explained 7% to 12% of variance.

    “This white matter PC represented 27 percent of variance in the ACT cohort, whereas the white matter marker-correlated PC represented 7 to 12 percent variance in the other three cohorts.”
    Quote in the paperNot a substantial claim

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

    Findingown data

    The APOE protein was not assigned to any network module, so it had no effect on the module enrichment findings.

    “ApoE did not fall within a network module, and therefore did not influence the reported module enrichment results.”
    Quote in the paperNot a substantial claim

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

    Findingown data

    The initial consensus network contained 18 modules with at least 14 proteins.

    “After the initial network construction, 18 modules consisting of 14 or more proteins were detected.”
    Quote in the paperNot a substantial claim

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

    Findingown data

    For the consensus LFQ cohort, 41.3% of quantified proteins were removed when missingness was controlled separately in each cohort before combining for AD network construction.

    “For the consensus LFQ cohort, the dotted line indicates the percent removed (41.3%) when missingness is controlled separately in each cohort prior to combination for construction of the AD network, as was done in this study.”
    Quote in the paperNot a substantial claim

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

    Findingown data

    After merging modules with high kME similarity, the consensus network was reduced to 13 modules.

    “Given high kME similarity between some modules, we used the WGCNA::moduleMergeUsingKME() function to reduce the number of modules to 13”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationown data

    The study identified proteins and biological processes in Alzheimer's disease brain that may serve as therapeutic targets and fluid biomarkers.

    “we identify proteins and biological processes in AD brain that may serve as therapeutic targets and fluid biomarkers for the disease”
    Quote in the paperSubstantial claim

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

    Premisecitation

    Protein co-expression analysis is a powerful tool for understanding biological network, pathway, and cell type changes in human tissue.

    “Protein co-expression analysis is a powerful tool to understand biological network, pathway, and cell type changes in human tissue3,4.”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [3] Miller, J. A., Oldham, M. C. & Geschwind, D. H. A systems level analysis of transcriptional changes in Alzheimer’s disease and normal aging. J. Neurosci. 28, 1410–1420 (2008). source

      Not checked in the free run.

    • cites [4] Oldham, M. C. et al. Functional organization of the transcriptome in human brain. Nat. Neurosci. 11, 1271–1282 (2008). source

      Not checked in the free run.

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

    Premisecitation

    Communities of co-expressed proteins can be linked to disease processes, and hub proteins within co-expression modules are enriched in key drivers of disease pathogenesis.

    “Communities of co-expressed proteins can be linked to disease processes, and the most strongly correlated proteins, or “hubs,” within these co-expression modules are enriched in key drivers of disease pathogenesis5–10.”
    Quote in the paperSubstantial claim0 of 6 sources checked
    • cites [5] Clarke, C. et al. Correlating transcriptional networks to breast cancer survival: a large-scale coexpression analysis. Carcinogenesis 34, 2300–2308 (2013). source

      Not checked in the free run.

    • cites [6] Barabasi, A. L., Gulbahce, N. & Loscalzo, J. Network medicine: a network-based approach to human disease. Nat. Rev. Genet. 12, 56–68 (2011). source

      Not checked in the free run.

    • cites [7] Huan, T. et al. Integrative network analysis reveals molecular mechanisms of blood pressure regulation. Mol. Syst. Biol. 11, 799 (2015). source

      Not checked in the free run.

    • cites [8] Tran, L. M. et al. Inferring causal genomic alterations in breast cancer using gene expression data. BMC Syst. Biol. 5, 121 (2011).

      Not checked in the free run.

    • cites [9] Huan, T. et al. A systems biology framework identifies molecular underpinnings of coronary heart disease. Arterioscler. Thromb. Vasc. Biol. 33, 1427–1434 (2013). source

      Not checked in the free run.

    • cites [10] Cerami, E., Demir, E., Schultz, N., Taylor, B. S. & Sander, C. Automated network analysis identifies core pathways in glioblastoma. PLoS ONE 5, e8918 (2010). source

      Not checked in the free run.

  • May combine two claims.

    Study design and conductown data

    The Alzheimer's disease protein co-expression network was robust and constructed from more than 450 human dorsolateral prefrontal cortex tissues from multiple centers.

    “In summary, we were able to construct a robust AD protein co-expression network from mass spectrometry-based proteomic analysis of greater than 450 human DLPFC brain tissues from multiple centers.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data

    The relationship between aging and Alzheimer's disease at the proteomic level is complex, and some but not all Alzheimer's disease trait-associated modules are influenced by aging.

    “These findings indicate that the relationship between aging and AD at the proteomic level is complex, and that some, but not all, AD trait-associated modules are influenced by the aging process.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Study design and conductown data

    Parallel reaction monitoring-based synthetic eigenprotein changes by disease category were very similar to untargeted label-free quantitation measurements, validating the untargeted findings.

    “We observed very similar AD network module changes across diseases compared to the untargeted measurements, validating the findings from the untargeted LFQ measurements.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationcitation

    The M4 module is enriched in microglial markers upregulated in response to amyloid-β deposition and downregulated in response to LPS, indicating the microglial response reflected in M4 is likely biased toward an anti-inflammatory disease-associated phenotype.

    “The M4 module is also enriched in microglial markers that are upregulated in response to amyloid-β deposition and downregulated in response to LPS, indicating that the microglial response as reflected in M4 module expression is likely biased towards an anti-inflammatory disease-associated phenotype27.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [27] Rangaraju, S. et al. Identification and therapeutic modulation of a pro-inflammatory subset of disease-associated-microglia in Alzheimer’s disease. Mol. Neurodegener. 13, 24 (2018). source

      Not checked in the free run.

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

    Premisecitation

    Targeting hubs within protein co-expression modules most related to disease biology is a promising approach for drug and biomarker development.

    “Therefore, targeting hubs within protein co-expression modules most related to disease biology is a promising approach for drug and biomarker development11–14.”
    Quote in the paperNot a substantial claim0 of 4 sources checked
    • cites [11] Yue, Z. et al. Repositioning drugs by targeting network modules: a Parkinson’s disease case study. BMC Bioinformatics 18, 532 (2017).

      Not checked in the free run.

    • cites [12] Toledo, J. B. et al. Metabolic network failures in Alzheimer’s disease: a biochemical road map. Alzheimers Dement. 13, 965–984 (2017). source

      Not checked in the free run.

    • cites [13] Sun, M., Sun, T., He, Z. & Xiong, B. Identification of two novel biomarkers of rectal carcinoma progression and prognosis via co-expression network analysis. Oncotarget 8, 69594–69609 (2017). source

      Not checked in the free run.

    • cites [14] He, Z. et al. Identifying biomarkers of papillary renal cell carcinoma associated with pathological stage by weighted gene co-expression network analysis. Oncotarget 8, 27904–27914 (2017). source

      Not checked in the free run.

  • May combine two claims.

    Interpretationown data

    M4 astrocyte/microglial module proteins are increased in asymptomatic Alzheimer's disease and Alzheimer's disease cerebrospinal fluid, suggesting that proteins within the M4 module may serve as useful biomarkers for staging Alzheimer's disease progression and for development of novel therapeutic approaches.

    “M4 astrocyte/microglial module proteins are increased in AsymAD and AD CSF, suggesting that proteins within the M4 module may serve as useful biomarkers for staging AD progression and for development of novel therapeutic approaches to the disease.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data

    The M4 module increases in asymptomatic Alzheimer's disease and correlates most strongly with cognitive impairment, suggesting that biological changes reflected by this module occur early and have significant functional consequence on progression to dementia.

    “The M4 astrocyte/microglial metabolism module increases in AsymAD and correlates most strongly with cognitive impairment, suggesting that the biological changes reflected by this module occur early in the disease and have significant functional consequence on progression to dementia.”
    Quote in the paperSubstantial claim

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

    Premisecitation

    LDHB, PKM, and DDAH1 have recently been reported as promising Alzheimer's disease cerebrospinal fluid biomarkers.

    “LDHB, PKM, and DDAH1 have recently been reported as promising AD CSF biomarkers52,53.”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [52] Dayon, L. et al. Alzheimer disease pathology and the cerebrospinal fluid proteome. Alzheimers Res. Ther. 10, 66 (2018). source

      Not checked in the free run.

    • cites [53] Sathe, G. et al. Quantitative proteomic profiling of cerebrospinal fluid to identify candidate biomarkers for Alzheimer’s disease. Proteomics Clin. Appl. 13, e1800105 (2019). source

      Not checked in the free run.

  • May combine two claims.

    Interpretationreasoning

    Elevations in PRDX1, DDAH, and PARK7 were observed, all important anti-oxidant effector proteins likely elevated with increased glycolytic flux.

    “Elevations in PRDX1, DDAH, and protein/nucleic acid deglycase DJ-1 (PARK7) were also observed, all of which are important anti-oxidant effector proteins49–51 and are likely elevated in concert with increased glycolytic flux.”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites [49] Luo, Y. et al. Asymmetric dimethylarginine exacerbates aβ-induced toxicity and oxidative stress in human cell and Caenorhabditis elegans models of Alzheimer disease. Free Radic. Biol. Med. 79, 117–126 (2015). source

      Not checked in the free run.

    • cites [50] Power, J. H. et al. Peroxiredoxin 6 in human brain: molecular forms, cellular distribution and association with Alzheimer’s disease pathology. Acta Neuropathol. 115, 611–622 (2008). source

      Not checked in the free run.

    • cites [51] Clements, C. M., McNally, R. S., Conti, B. J., Mak, T. W. & Ting, J. P. DJ-1, a cancer- and Parkinson’s disease-associated protein, stabilizes the antioxidant transcriptional master regulator Nrf2. Proc. Natl Acad. Sci. USA 103, 15091–15096 (2006). source

      Not checked in the free run.

  • May combine two claims.

    Interpretationown data

    Certain Alzheimer's disease network modules are affected to a greater extent in Alzheimer's disease compared to other neurodegenerative diseases, and FTLD and CBD show many similar changes to Alzheimer's disease, with the caveat that not all neurodegenerative diseases affect the dorsolateral prefrontal cortex region equally at end-stages.

    “These results indicate that certain AD network modules are affected to a greater extent in AD compared to other neurodegenerative diseases, and that FTLD and CBD show many similar changes to AD, with the caveat that not all neurodegenerative diseases affect the DLPFC region equally at end-stages of disease.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationreasoning

    The M4 module is enriched in Alzheimer's disease genetic risk factors, indicating a potential causative role for this module in disease pathogenesis, and appears to serve a protective anti-inflammatory function in model systems, suggesting that genetic risk factor polymorphisms that cluster in this module may induce a loss-of-function phenotype.

    “The M4 module is enriched in AD genetic risk factors, indicating a potential causative role for this protein co-expression module in disease pathogenesis, and appears to serve a protective anti-inflammatory function in model systems, suggesting that genetic risk factor polymorphisms that cluster in this module may induce a loss-of-function phenotype.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Alzheimer's disease genetic risk alleles are more likely to cause loss-of-function changes rather than gain-of-function changes.

  • May combine two claims.

    Premisecitation

    A proteomic relationship between Alzheimer's disease and FTD is supported by TREM2 mutations causing microglial dysfunction and leading to Alzheimer's disease, and PGRN mutations causing microglial dysfunction and leading to FTD.

    “A proteomic relationship between AD and FTD is supported by the fact that mutations in the triggering receptor expressed on myeloid cells 2 (TREM2) protein cause microglial dysfunction and lead to AD39,40, whereas mutations in the progranulin (PGRN) protein also cause microglial dysfunction and lead to FTD41–43.”
    Quote in the paperSubstantial claim0 of 5 sources checked
    • cites [39] Jonsson, T. et al. Variant of TREM2 associated with the risk of Alzheimer’s disease. N. Engl. J. Med. 368, 107–116 (2013). source

      Not checked in the free run.

    • cites [40] Wang, Y. et al. TREM2 lipid sensing sustains the microglial response in an Alzheimer’s disease model. Cell 160, 1061–1071 (2015). source

      Not checked in the free run.

    • cites [41] Baker, M. et al. Mutations in progranulin cause tau-negative frontotemporal dementia linked to chromosome 17. Nature 442, 916–919 (2006). source

      Not checked in the free run.

    • cites [42] Cruts, M. et al. Null mutations in progranulin cause ubiquitin-positive frontotemporal dementia linked to chromosome 17q21. Nature 442, 920–924 (2006). source

      Not checked in the free run.

    • cites [43] Hickman, S., Izzy, S., Sen, P., Morsett, L. & El Khoury, J. Microglia in neurodegeneration. Nat. Neurosci. 21, 1359–1369 (2018). source

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Study design and conductnone given

    A nonparametric bootstrap regression was run separately for each cohort, subtracting the trait of interest (age at death, sex, or postmortem interval) times the median estimated coefficient from 1,000 iterations for each protein.

    “Nonparametric bootstrap regression was performed separately in each cohort by subtracting the trait of interest (age at death, sex, or postmortem interval (PMI)) times the median estimated coefficient from 1000 iterations of fitting for each protein in the cohort-specific log2(abundance) matrix.”
    Quote in the paperSubstantial claim

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

    Premisecitation

    Recent studies have identified additional potential biomarkers from M4 and other brain modules by deep mass spectrometry-based discovery on fractionated cerebrospinal fluid.

    “Indeed, recent studies have identified a number of additional potential biomarkers from M4 and other brain modules by deep mass spectrometry-based discovery on fractionated CSF35,36,54.”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites [35] Bai, B. et al. Deep multilayer brain proteomics identifies molecular networks in Alzheimer’s disease progression. Neuron 105, 975–991.e7 (2020). source

      Not checked in the free run.

    • cites [36] Higginbotham, L., et al. Integrated proteomics reveals brain-based cerebrospinal fluid biomarkers in asymptomatic and symptomatic Alzheimer’s disease. Preprint at bioRxiv https://doi.org/10.1101/806752 source

      Not checked in the free run.

    • cites [54] Dayon, L. et al. Proteomes of paired human cerebrospinal fluid and plasma: relation to blood–brain barrier permeability in older adults. J. Proteome Res. 18, 1162–1174 (2019). source

      Not checked in the free run.

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

    Study design and conductnone given

    A fixed random seed of 1 was used to ensure reproducibility.

    “Random seed was set to 1 for reproducibility”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Study design and conductown data

    The consensus protein co-expression network was built from a matrix of 3,334 log2 protein abundances across 419 case samples after covariate and batch correction.

    “A weighted protein co-expression network for the four-cohort consensus data was generated using the 3,334 log2 protein abundance x 419 case-sample matrix that had undergone covariate and batch correction as described above.”
    Quote in the paperSubstantial claim

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

    Interpretationreasoning

    Monitoring multiple M4 protein levels in biofluid may provide a robust measure of target engagement for Alzheimer's disease therapies.

    “Monitoring multiple M4 protein levels in biofluid may provide a robust measure of target engagement for AD therapies.”
    Quote in the paperNot a substantial claim

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

    Study design and conductown data

    Investigators tested how the missing-value cutoff affected AD network modules by building networks under varying permitted missingness and comparing preservation of the 50% cutoff modules against networks built with stricter cutoffs (10-40% missingness) via Zsummary scores.

    “(B) The effect of missing value threshold on AD network modules. The AD network was constructed using different allowed levels of missing protein measurements. Preservation of AD network modules (50% missingness threshold) in each network generated using a more stringent threshold (10–40% missingness) was assessed by Zsummary score.”
    Quote in the paperNot a substantial claim

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

    Study design and conductown data

    For ACT, a second bootstrap regression round used the white matter principal component as a covariate on the 62 non-outlier case samples, subtracting 28% of the white matter marker-correlated variance to reach a final variance of 12% after recalculating the top five principal components.

    “To adjust for this white matter variance in ACT, we applied a second round of bootstrap regression to the 62 non-outlier ACT case sample log2(abundances), using the white matter PC as a regression covariate, and subtracted 28 percent of the white matter marker correlated variance to achieve a final variance of 12 percent after recalculation of the top 5 PCs.”
    Quote in the paperNot a substantial claim

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

    Study design and conductown data

    In the TMT ROS/MAP database search, Percolator filtered peptide-spectrum matches and peptides to an FDR below 1%, and assembled proteins were further filtered to a final FDR of 1%.

    “Percolator was used to filter peptide spectral matches (PSMs) and peptides to a false discovery rate (FDR) of less than 1%. Following spectral assignment, peptides were assembled into proteins and were further filtered based on the combined probabilities of their constituent peptides to a final FDR of 1%.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationown data

    M4 markers may not be entirely specific for Alzheimer's disease given elevation of the M4 module in FTD and CBD, but may allow assessment of an injury response in Alzheimer's disease with amyloid and tau biomarkers and serve as biomarkers for other neurodegenerative dementias.

    “While M4 markers may not be entirely specific for AD given elevation of the M4 module in FTD and CBD, they may allow for assessment of an injury response in AD in conjunction with amyloid and tau biomarkers, and serve as useful biomarkers for other neurodegenerative dementias in addition to AD.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim.

    Study design and conductnone given

    Case status or diagnosis was explicitly modeled and protected in each regression.

    “Case status/diagnosis was also explicitly modeled (i.e., protected) in each regression.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim.

    Study design and conductcitation

    A median polish algorithm was implemented to remove technical variance, such as tissue collection, cohort, or batch effects, from a two-way abundance-sample data table.

    “We implemented a median polish algorithm for removing technical variance (e.g., due to tissue collection, cohort, or batch effects) from a two-way abundance-sample data table as originally described by Tukey69.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Tukey69

      Not checked in the free run.

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

    Study design and conductnone given

    In the label-free quantification search, the false discovery rate was set to 1% for peptide-spectrum matches, protein identifications, and the site decoy fraction.

    “The false discovery rate (FDR) for peptide spectral matches, proteins, and site decoy fraction were each set to 1 percent.”
    Quote in the paperNot a substantial claim

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

    Study design and conductown data

    For the BLSA precuneus, Mayo temporal cortex, and ROS/MAP cohorts, networks were built with the same pipeline as the consensus network, including batch correction, outlier removal, and regression for age, sex, and PMI.

    “For BLSA precuneus, Mayo temporal cortex, and ROS/MAP cohorts, we confirmed that they built networks using the pipeline described above, including batch correction, outlier removal, and age/sex/PMI regression.”
    Quote in the paperNot a substantial claim

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

    Study design and conductown data

    In the CSF database search, Percolator filtered peptide-spectrum matches and peptides to an FDR below 1%, and assembled proteins were further filtered to a final FDR of 1%.

    “Percolator was used to filter PSMs and peptides to an FDR of less than 1%. Following spectral assignment, peptides were assembled into proteins and were further filtered based on the combined probabilities of their constituent peptides to a final FDR of 1%.”
    Quote in the paperNot a substantial claim

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

    Study design and conductown data

    Outlier detection on the network excluded 31 of 450 cases from the four-cohort consensus network and differential abundance analyses.

    “Finally, network outlier detection was performed as described above, which removed 31/450 cases from consideration in the four-cohort consensus network and differential abundance analyses.”
    Quote in the paperNot a substantial claim

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

    Study design and conductnone given

    The median polish procedure ensures that variance reduction is robust to outliers and preserves biological variance, provided that batches were randomized so that batch is not confounded with diagnosis or other biological traits.

    “The use of median polish ensures that the reduction of variance is robust to outliers while the overall algorithm preserves biological variance, given that batches have been randomized to avoid confounding batch with diagnosis or other biological traits.”
    Quote in the paperSubstantial claim

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

    Study design and conductown data

    Network module preservation across cohorts was assessed using the WGCNA modulePreservation function.

    “We used the WGCNA::modulePreservation() function to assess network module preservation across cohorts.”
    Quote in the paperNot a substantial claim

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

    Study design and conductnone given

    No missing values were imputed in any cohort.

    “No imputation of missing values was performed in any cohort.”
    Quote in the paperNot a substantial claim

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

    Study design and conductnone given

    For the sex-effects analysis on the consensus network, the same cases and pipeline were used, except sex covariance was not regressed out before WGCNA.

    “For analysis of sex effects on the consensus network, the identical cases and analysis pipeline as described above were used to build the consensus network, except that covariance due to sex was not removed by regression prior to WGNCA analysis.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationown data

    The comprehensive study provides a consensus view of the proteomic network landscape of Alzheimer's disease and biological changes associated with asymptomatic and symptomatic stages, and highlights the central role of glial biology in pathogenesis.

    “In summary, our comprehensive study on more than 2000 brains and nearly 400 CSF samples provides a consensus view of the proteomic network landscape of AD and the biological changes associated with asymptomatic and symptomatic stages of the disease, and highlights the central role of glial biology in the pathogenesis of the disease.”
    Quote in the paperSubstantial claim

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

    Study design and conductnone given

    Where indicated, multiple-comparison adjustment used false discovery rate correction.

    “P values were adjusted for multiple comparisons by false discovery rate (FDR) correction where indicated.”
    Quote in the paperNot a substantial claim

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

    Study design and conductown data

    The effect of missing values on the consensus network was assessed using the WGCNA modulePreservation function.

    “We also used this function to assess the effect of missing values on the consensus network.”
    Quote in the paperNot a substantial claim

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

    Study design and conductnone given

    In the ROS/MAP brain tissue TMT labeling, cases were randomized by covariates (age, sex, PMI, diagnosis, and others) into 50 batches of 8 cases each.

    “Prior to TMT labeling, cases were randomized by covariates (age, sex, PMI, diagnosis, etc.), into 50 total batches (8 cases per batch).”
    Quote in the paperNot a substantial claim

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

    Study design and conductown data

    Measurement of additional M4 markers in biofluids is possible because mass spectrometry measurements were performed on unfractionated cerebrospinal fluid not depleted of highly abundant proteins.

    “Measurement of additional M4 markers in biofluids is undoubtedly possible, as our mass spectrometry measurements were performed on unfractionated CSF not depleted of highly abundant proteins.”
    Quote in the paperNot a substantial claim

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

    Study design and conductnone given

    Network module preservation was also assessed using synthetic eigenproteins.

    “We also assessed network module preservation using synthetic eigenproteins.”
    Quote in the paperNot a substantial claim
62 of 62

  • The paper relies on this without support.

    The biochemical and cellular changes in Alzheimer's disease brain beyond amyloid-β and tau deposition remain incompletely understood.

    “the biochemical and cellular changes in the brain that characterize the disease beyond amyloid-β and tau deposition remain incompletely understood”

  • The paper relies on this without support.

    Alzheimer's disease genetic risk alleles are more likely to cause loss-of-function changes rather than gain-of-function changes.

    “AD genetic risk alleles, which are more likely to cause loss-of-function changes rather than gain-of-function changes, are enriched in the M4 module”
2 of 2

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?

  • A recent proteomic network analysis of post-mortem brains revealed activation of a module enriched for glucose metabolism and glial proteins in Alzheimer's disease cases, with the ApoE4 homozygous subgroup most affected.

    “Notably, a recent proteomic network analysis of post-mortal brains revealed the activation of a module enriched for glucose metabolism and glial proteins in AD cases, with the ApoE4 homozygous sub-group most affected (Johnson et al., 2020).”

    From ApoE4 Impairs Neuron-Astrocyte Coupling of Fatty Acid Metabolism (2021) · cites this paper as Johnson et al., 2020

    The Introduction and Results describe an AD-associated M4 module enriched in sugar-metabolism and glial proteins, with increased expression in AD. But the Results report no strong APOE-genotype effects on module levels except an ε2 effect on a different module; they do not support the claim that APOE ε4 homozygotes were most affected. Page numbers were not provided in our copy. (LLM, conf=0.91)

    Experimental — accuracy not yet measured

  • Impaired brain energy metabolism is an early pathological manifestation of neurodegenerative diseases including Alzheimer's disease.

    “Impaired brain energy metabolism serves as an early pathological manifestation of neurodegenerative diseases, including Alzheimer's disease (AD) [1, 2].”

    From Aerobic exercise improves astrocyte mitochondrial quality and transfer to neurons in a mouse model of Alzheimer's disease (2025) · cites this paper as [1]

    The Results (Construction and Validation of a Consensus AD Protein Co-Expression Network; page not provided) report changes in mitochondrial and astrocyte/microglial metabolism protein modules during asymptomatic, preclinical AD. That supports early metabolism-related changes in AD, but not the stronger claim that brain energy metabolism is impaired. The comparison with other neurodegenerative diseases examines network changes in tissue from disease cases, not whether metabolic impairment is an early manifestation of those diseases. (LLM, conf=0.85)

    Experimental — accuracy not yet measured

  • Co-expression networks have been successful for analyzing bulk proteomics datasets in human disease samples.

    “Co-expression networks have been successful for analyzing bulk proteomics datasets in human disease samples,67,68”

    From hdWGCNA identifies co-expression networks in high-dimensional transcriptomics data (2023) · cites this paper as [68]

  • There is important overlap between brain and CSF proteomes.

    “Given the important overlap between brain and CSF proteomes, 80 , 81 , 82 further studies could investigate these proteins as potential sex‐specific biomarkers in AD.”

    From Unraveling sex differences in Alzheimer's disease and related endophenotypes with brain proteomes (2025) · cites this paper as [81]

  • The M4 network module is related to microglial metabolism in human AD brain.

    “Another proteomic analyses of human AD brain identified a network module related to microglial metabolism (termed “M4”) (Johnson et al., 2020).”

    From APOE4 reshapes the lipid droplet proteome and modulates microglial inflammatory responses (2025) · cites this paper as Johnson et al., 2020

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

  • ROS/MAP pooled-standard count does not match the batch design

    Minor

    Fifty batches with two pooled-standard channels per batch require 100 pooled-standard channels, not 64. The labeling methods also report 100 pooled-standard samples.

    Reported 64 · recomputed 100

  • CSF channel-mixing instructions account for only ten channels

    Minor

    The methods say the CSF batches have 11 TMT channels, including a pooled standard. The described mixture combines nine listed channels with the 126 boost channel, leaving one channel unaccounted for.

  • Reported consensus case counts differ across analyses

    Minor

    The sex-analysis counts total 418 cases, whereas the stated post-outlier consensus counts total 419. Their diagnostic-group totals also differ: 226 versus 230 AD, 99 versus 98 AsymAD, and 93 versus 91 controls, despite the methods saying the sex analysis used identical cases.

  • CSF elution steps exceed the stated gradient duration

    Minor

    The four stated CSF elution steps total 183 minutes rather than the reported 180 minutes.

    Reported 180 · recomputed 183

4 of 4

Cites

Flagged first, then by citations

Research works printed but not linked: 1

  1. Tukey, J. W. Exploratory Data Analysis (Addison-Wesley, 1977).

Cited by

Sorted by citations

Related

Authors
Johnson, E. C. B., Dammer, E. B., Duong, D. M., Ping, L., Zhou, M., Yin, L., Higginbotham, L. A., Guajardo, A., White, B., Troncoso, J. C., Thambisetty, M., Montine, T. J., Lee, E. B., Trojanowski, J. Q., Beach, T. G., Reiman, E. M., Haroutunian, V., Wang, M., Schadt, E., Zhang, B., Dickson, D. W., Ertekin-Taner, N., Golde, T. E., Petyuk, V. A., De Jager, P. L., Bennett, D. A., Wingo, T. S., Rangaraju, S., Hajjar, I., Shulman, J. M., Lah, J. J., Levey, A. I., & Seyfried, N. T.
Affiliations
  • Emory University (US)
  • Johns Hopkins University (US)
  • National Institutes of Health (US)
  • Stanford University (US)
  • University of Pennsylvania (US)
  • Banner Sun Health Research Institute (US)
  • University of Arizona (US)
  • James J. Peters VA Medical Center (US)
  • Icahn School of Medicine at Mount Sinai (US)
  • Mayo Clinic in Florida (US)
  • University of Florida (US)
  • Pacific Northwest National Laboratory (US)
  • Rush University Medical Center (US)
  • Baylor College of Medicine (US)
Corresponding author
  • Johnson, Erik C. B.
  • Levey, Allan I.
  • Seyfried, Nicholas T.
ORCID
Venue
Nature Medicine
Publisher
Springer Science and Business Media LLC
Published
2020-04-13
Volume
26
Issue
5
Pages
769-780
ISSN
  • 1078-8956 (ISSN-L)
  • 1546-170X
DOI
10.1038/s41591-020-0815-6
PMC
PMC7405761
OpenAlex
W3015916968
Semantic Scholar
44ee5515609dfe6555c259f551315a322ff2d26b
Cited by
1,001
References
156
Citations by year
2026: 127 · 2025: 165 · 2024: 174 · 2023: 167 · 2022: 149 · 2021: 149 · 2020: 65 · 2013: 1 · 2012: 1
FWCI
83.02
Type
journal article
Language
en
Open access
Licence
Locations
  • Nature Medicine (publishedVersion)
  • bioRxiv (Cold Spring Harbor Laboratory) (acceptedVersion, cc-by-nc-nd)
  • PubMed (publishedVersion)
  • PubMed Central (acceptedVersion)
  • OSTI OAI (U.S. Department of Energy Office of Scientific and Technical Information) (submittedVersion)
  • eScholarship (California Digital Library) (submittedVersion)
Funders
  • U.S. Department of Veterans Affairs (I01BX003853)
  • National Institutes of Health (U24NS07-2026)
  • National Institute on Aging (R01AG017917, P30-AG 10161, P30-AG-19610, R01AG057907, R01AG056533, RF1AG051504, R01‐AG17917, P30AG10124, R01AG050631, R01 AG15819, U01AG046170, U01‐AG61356, RF1AG054014, RF1AG057470, U01AG046152, R01AG057911, P30AG010124, R01AG015819, RF1AG057440, P30AG010161, RF1AG059319, U01AG061357, U01AG061356, P30AG019610, P50AG005146, U01AG046161, RF1AG051633, U01AG046139, RF1AG057471, R01AG061796, P50AG025688, RF1AG058501, R01AG057339, U01AG052411, R01AG053960, R01AG058501, P30AG066507, R01AG072120, R01AG061800, R01AG075827, P30AG066511)
  • National Institute of Neurological Disorders and Stroke (K08NS099474, R01NS114130, U24NS072026)
Topics
  • Health Sciences › Medicine › Physiology › Alzheimer's disease research and treatments
  • Health Sciences › Medicine › Psychiatry and Mental health › Dementia and Cognitive Impairment Research
  • Physical Sciences › Chemistry › Spectroscopy › Advanced Proteomics Techniques and Applications
Keywords
Alzheimer's disease, cerebrospinal fluid proteomics, brain proteomics, energy metabolism, microglial activation, astrocyte activation, sugar metabolism, cognitive impairment, genetic risk factors, therapeutic target, biomarkers, glia
MeSH terms
Alzheimer Disease, Animals, Astrocytes, Brain, Cerebrospinal Fluid, Energy Metabolism, Female, Humans, Male, Nerve Tissue Proteins, Mass Spectrometry, Time Factors, Cohort Studies, Biomarkers, Case-Control Studies, Microglia, Sample Size, Disease Progression, Proteomics, Mice, Gene Regulatory Networks, Metabolic Networks and Pathways, Neurogenesis

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 2026-10-06.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-06.

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