APOE4/4 is linked to damaging lipid droplets in Alzheimer’s disease microglia

In the paper

“This epigenetic signature is also seen in macrophages in atherosclerosis models23 and adipose-associated macrophages24 but this signature is not seen in DAM microglia25.”

What the sources say

• ?Could not verify

  The Results section ('Elevation of ETV6 in aorta macrophages in ApoE−/− mice'; page unavailable) reports increased ETV6 expression in macrophages from an atherosclerosis model. The cited portion of the sentence concerns an epigenetic signature, but neither that section nor the passages retrieved by searching for epigenetic and chromatin terms identifies or measures one. The adipose-macrophage and DAM-microglia portions have separate citations and are not assessed here. (LLM, conf=0.72)

  cites [23] Xiong, X., Yan, Z., Jiang, W. & Jiang, X. ETS variant transcription factor 6 enhances oxidized low-density lipoprotein-induced inflammatory response in atherosclerotic macrophages via activating NF-κB signaling. Int. J. Immunopathol. Pharmacol. https://doi.org/10.1177/20587384221076472 (2022). sourceExperimental — accuracy not yet measured
• Supported

  The citation is attached to the adipose-associated macrophage portion of the sentence. The Results section reports ATAC-seq evidence of distinct chromatin-accessibility patterns in adipose tissue macrophage populations, supporting the attribution that these cells have an epigenetic signature. This source is not cited for the atherosclerosis or DAM microglia comparisons. (LLM, conf=0.84)

  cites [24] Hill, D. A. et al. Distinct macrophage populations direct inflammatory versus physiological changes in adipose tissue. Proc. Natl Acad. Sci. USA 115, E5096–E5105 (2018). source
• ?Could not verify

  The Results section on the microglia trajectory reports depletion of a stage 2 DAM gene signature, but does not identify the epigenetic signature meant by the citing sentence or say that it is absent in DAM microglia. Searches for that epigenetic signature and for a comparison with macrophages did not find a passage establishing the cited claim; page numbers were unavailable in our copy. (LLM, conf=0.72)

  cites [25] Morabito, S. et al. Single-nucleus chromatin accessibility and transcriptomic characterization of Alzheimer’s disease. Nat. Genet. 53, 1143–1155 (2021). sourceExperimental — accuracy not yet measured

In the paper

“A previously published dataset shows that ACSL1 is highly upregulated by the innate immune trigger lipopolysaccharides (LPS) in an independent human iMG chimaeric mouse model22”

What the sources say

• Partly supported

  The Results section, “xMGs respond robustly to a systemic inflammatory challenge” (page not available), reports a dataset comparing human microglia from saline- and LPS-treated chimeric mice and identifies 607 upregulated genes. Our copy does not identify ACSL1 among them or establish that it was highly upregulated; searches for ACSL1 and its full gene name found no supporting passage. (LLM, conf=0.85)

  cites [22] Hasselmann, J. et al. Development of a chimeric model to study and manipulate human microglia in vivo. Neuron 103, 1016–1033 (2019). sourceExperimental — accuracy not yet measured

In the paper

“APOE is one such lipid-related AD risk gene, which is highly upregulated in human microglia in AD3”

What the sources say

• Supported

  The section on systematic differential expression reports that APOE was strongly upregulated in human microglia in AD pathology. The section on cell-type-specific associations also identifies APOE as an AD risk factor whose microglial expression correlates positively with pathology. (LLM, conf=0.92)

  cites [3] Mathys, H. et al. Single-cell transcriptomic analysis of Alzheimer’s disease. Nature 570, 332–337 (2019). source

In the paper

“LDs form in myeloid cells through the upregulation of lipid-synthesis enzymes triggered by the engagement of toll receptors by innate immune triggers, such as bacteria7.”

What the sources say

• Partly supported

  The Discussion reports that TLR agonists increased triglyceride synthesis and the abundance of ACSL1 and DGAT2 in macrophages, supporting the proposed link to lipid droplets. But it describes enzyme upregulation as one part of a multifaceted storage response, alongside increased fatty-acid uptake and decreased lipolysis; it does not establish this as the mechanism across myeloid cells generally. Page numbers are unavailable in our copy. (LLM, conf=0.84)

  cites [7] Huang, Y. et al. Toll-like receptor agonists promote prolonged triglyceride storage in macrophages. J. Biol. Chem. 289, 3001–3012 (2014). sourceExperimental — accuracy not yet measured

In the paper

“Aged mouse microglia accumulate LDs and exhibit a dysfunctional microglial state termed LD-accumulating microglia (LDAM)5”

What the sources say

• Supported

  The Results section describes lipid-droplet-containing aged microglia and names this distinct state “lipid droplet accumulating microglia” (LDAM). The Discussion identifies LDAM in the aging brain and reports severe functional deficits; the Results passages specifically describe deficits in microglia from aged mice. The cited paper therefore supports the sentence. (LLM, conf=0.94)

  cites [5] Marschallinger, J. et al. Lipid-droplet-accumulating microglia represent a dysfunctional and proinflammatory state in the aging brain. Nat. Neurosci. 23, 194–208 (2020). source

In the paper

“Cholesterol-rich lysosomes and LDs in dysfunctional microglia have also been observed in the context of demyelination mouse models and human iPS cell models9–12.”

What the sources say

• Partly supported

  The passage on Pelizaeus–Merzbacher disease mice (page unavailable in our copy) reports demyelination, increased lipofuscin in microglia, and reduced microglial uptake function. The paper also describes lipofuscin as accumulating in lysosomes. It does not identify those lysosomes as cholesterol-rich or report lipid droplets or human iPS-cell models. (LLM, conf=0.88)

  cites [9] Safaiyan, S. et al. Age-related myelin degradation burdens the clearance function of microglia during aging. Nat. Neurosci. 19, 995–998 (2016). sourceExperimental — accuracy not yet measured
• Partly supported

  The main text (p. 1) reports myelin debris in phagocyte lysosomes and lipid-droplet-bearing foam cells in demyelinated old mice; the supplementary Fig. 3 caption (p. 6) identifies microglial cultures treated with myelin debris. Our copy does not report these observations in human iPS cell models, so it supports the mouse-model portion but not that extension. (LLM, conf=0.86)

  cites [10] Cantuti-Castelvetri, L. et al. Defective cholesterol clearance limits remyelination in the aged central nervous system. Science 359, 684–688 (2018). sourceExperimental — accuracy not yet measured
• Partly supported

  The Results (page unavailable) report cholesteryl-ester accumulation in Trem2-deficient microglia after chronic demyelination and increased cholesteryl esters and cholesterol in myelin-treated human iPSC-derived microglia. The paper suggests lipid-droplet storage from experiments in murine macrophages, but it does not report observing cholesterol-rich lysosomes in either microglial model; its evidence for lysosomal dysfunction is indirect. (LLM, conf=0.85)

  cites [11] Nugent, A. A. et al. TREM2 regulates microglial cholesterol metabolism upon chronic phagocytic challenge. Neuron 105, 837–854 (2020). sourceExperimental — accuracy not yet measured
• Partly supported

  The Results (‘Altered lipid metabolism in human APOE4 microglia and astrocytes’; page unavailable) report impaired cholesterol metabolism in human iPSC-derived microglia and predict lysosomal cholesterol accumulation. They do not establish cholesterol-rich lysosomes or lipid droplets in those microglia, or report those observations in demyelination mouse models. The measured increase in lysosomal cholesterol described in ‘Impaired cholesterol trafficking’ was in astrocytes. (LLM, conf=0.84)

  cites [12] Tcw, J. et al. Cholesterol and matrisome pathways dysregulated in astrocytes and microglia. Cell 185, 2213–2233 (2022). sourceExperimental — accuracy not yet measured

In the paper

“a mechanism which is known to regulate LD concentrations27–29”

What the sources say

• Supported

  The Discussion states that autophagy regulates lipid content and that inhibiting it increased lipid droplets in vitro and in vivo. The section “Autophagic pathway components associate with LDs” also reports that activating autophagy significantly decreased lipid-droplet number. (LLM, conf=0.94)

  cites [27] Singh, R. et al. Autophagy regulates lipid metabolism. Nature 458, 1131–1135 (2009). source
• Supported

  The Introduction states that autophagy regulates intracellular lipid stores and describes lipophagy as lipid-droplet breakdown through the autophagy–lysosome pathway. The Results section, “Progressive Atherosclerosis Has Features of Dysfunctional Lipophagy,” likewise describes lipolysis of lipid droplets through that pathway. Page numbers are not provided in our copy. (LLM, conf=0.92)

  cites [28] Qiao, L. et al. Deficient chaperone-mediated autophagy promotes lipid accumulation in macrophage. J. Cardiovasc. Transl. Res. 14, 661–669 (2021). source
• ?Not checked — no full text of the sourcecites [29] Zhou, H. et al. In vivo simultaneous transcriptional activation of multiple genes in the brain using CRISPR-dCas9-activator transgenic mice. Nat. Neurosci. 21, 440–446 (2018). source

In the paper

“cell culture models have demonstrated that hyperactive neurons can release APOE-associated fatty acids which are taken up by astrocytes and buffer lipotoxic stress35.”

What the sources say

• Supported

  The Discussion (p. 11) says hyperactive neurons release toxic fatty acids in lipoprotein particles that astrocytes take up and detoxify, protecting neurons from fatty-acid toxicity. The Discussion (p. 13) says the neuron–astrocyte transfer assays included ApoE-positive particles. (LLM, conf=0.92)

  cites [35] Ioannou, M. S. et al. Neuron–astrocyte metabolic coupling protects against activity-induced fatty acid toxicity. Cell 177, 1522–1535 (2019). source

In the paper

“These results are similar to recent reports describing mouse LDAM that have a dysfunctional and inflammatory microglial state5.”

What the sources say

• Supported

  The Discussion describes LDAM in the aging brain as having severe functional deficits and a pro-inflammatory phenotype. The Results also report increased cytokine release by lipid-droplet-high microglia isolated from aged mice, supporting the claim about mouse LDAM. (LLM, conf=0.95)

  cites [5] Marschallinger, J. et al. Lipid-droplet-accumulating microglia represent a dysfunctional and proinflammatory state in the aging brain. Nat. Neurosci. 23, 194–208 (2020). source

In the paper

“PI3 kinase inhibition is a known modulator of LDs in mouse macrophages26 but this has not been previously shown in human microglia.”

What the sources say

• Supported

  The section on CCR2, ERK, and PI3K activation reports that the PI3K inhibitor LY294002 blocked MCP-1-induced lipid body formation in mouse macrophages. The citation is attached to the mouse-macrophage finding; the statement about human microglia is the citing authors’ own novelty claim. (LLM, conf=0.92)

  cites [26] Pacheco, P. et al. Monocyte chemoattractant protein-1/CC chemokine ligand 2 controls microtubule-driven biogenesis and leukotriene B 4-synthesizing function of macrophage lipid bodies elicited by innate immune response. J. Immunol. 179, 8500–8508 (2007). source

In the paper

“lipid-laden ‘foamy macrophages’ in atherosclerosis which have a senescent phenotype and are drivers of pathology37.”

What the sources say

• Supported

  The paper reports that foamy macrophages bearing senescence markers accumulate in early atherosclerotic lesions and drive pathology through inflammatory and atherogenic signaling. It also reports that removing senescent foamy macrophages from early lesions led to lesion regression. These findings are from an atherosclerosis-prone mouse model. (LLM, conf=0.93)

  cites [37] Childs, B. G. et al. Senescent intimal foam cells are deleterious at all stages of atherosclerosis. Science 354, 472–477 (2016). source

In the paper

“Isogenic APOE4/4 and APOE3/3 iPS cells were generated as previously described16.”

What the sources say

• Supported

  The Results section (‘Gene editing reveals the specificity of apoE4’) states that the authors generated an isogenic APOE3/3 hiPSC line from an APOE4/4 hiPSC line. The Online Methods section describes the gene-editing procedure, supporting the cited description of the cells. (LLM, conf=0.94)

  cites [16] Wang, C. et al. Gain of toxic apolipoprotein E4 effects in human iPSC-derived neurons is ameliorated by a small-molecule structure corrector article. Nat. Med. 24, 647–657 (2018). source

In the paper

“We then used the Python implementation of BBKNN (v.1.5.1), a fast batch correction algorithm suitable for large datasets to integrate data across the samples45. BBKNN calculates a batch-corrected neighbourhood graph from the imputed principal components. We set the individual sample IDs as batch labels to correct for potential sample-wise batch effects.”

What the sources say

• Supported

  The Introduction identifies BBKNN as a fast Python batch-alignment method. The Materials and methods section explains that it constructs a neighbourhood graph balanced across user-defined batches, supporting its use to integrate samples and address sample-wise batch effects. The choice to use individual sample IDs as batch labels describes the citing authors’ procedure, not a result attributed to this source. (LLM, conf=0.92)

  cites [45] Polański, K. et al. BBKNN: fast batch alignment of single cell transcriptomes. Bioinformatics 36, 964–965 (2020). source

In the paper

“alpha-synuclein binding to TLR2 and TLR5 induces microglial NLRP3 inflammasome activation, which is a shared signature seen in LDAM36.”

What the sources say

• Partly supported

  The Discussion (page not available) reports that α-synuclein monomers and oligomers act through TLR2 to prime the microglial NLRP3 inflammasome and through TLR5 to activate it. The paper does not identify this activation as a shared signature of LDAM; searches for LDAM and lipid-droplet-associated microglia found no supporting passage. (LLM, conf=0.86)

  cites [36] Scheiblich, H. et al. Microglial NLRP3 inflammasome activation upon TLR2 and TLR5 ligation by distinct α-synuclein assemblies. J. Immunol. 207, 2143–2154 (2021). sourceExperimental — accuracy not yet measured

In the paper

“This approach takes inspiration from recent work showing the deleterious effects of conditioned media on neurons from astrocytes30 and microglia4.”

What the sources say

• Supported

  The Abstract (page unavailable) states that activated astrocytes secrete factors toxic to neurons, and the Methods (page unavailable) describe collecting reactive astrocyte conditioned media and testing neuronal survival. Citation 30 is attached to the astrocyte portion of the citing sentence; the microglia portion has a separate citation. (LLM, conf=0.91)

  cites [30] Guttenplan, K. A. et al. Neurotoxic reactive astrocytes induce cell death via saturated lipids. Nature 599, 102–107 (2021). source
• Supported

  The citation to this paper is attached to the microglia portion of the sentence, not the astrocyte portion. In Results, under “Imbalance in the net flux of lipids by APOE4 iMGLs,” the authors report that conditioned media from APOE4 microglia-like cells robustly decreased neuronal calcium transients; the finding is specific to APOE4 conditioned media. (LLM, conf=0.92)

  cites [4] Victor, M. B. et al. Lipid accumulation induced by APOE4 impairs microglial surveillance of neuronal-network activity. Cell Stem Cell 29, 1197–1212 (2022). source

In the paper

“Count data were first screened for doublets with the Scrublet (v.0.2.3) Python package43. Once each cell was doublet scored, we applied a separate doublet score threshold per sample to discard doublets from the data.”

What the sources say

• Supported

  The Introduction identifies Scrublet as a method for scoring transcriptomes to detect doublets. The STAR★METHODS General Approach describes removing likely doublets by thresholding those scores, and the dataset methods show thresholds set for individual datasets. The citing authors’ choice of a separate threshold for each of their samples is their own procedure, not a result attributed to this paper. (LLM, conf=0.92)

  cites [43] Wolock, S. L., Lopez, R. & Klein, A. M. Scrublet: computational identification of cell doublets in single-cell transcriptomic data. Cell Syst. 8, 281–291 (2019). source

In the paper

“AD-APOE4/4 brain tissue has the greatest percentage of LDAM, followed by AD-APOE3/3 and the least amount of the LDAM microglia state is found in the aged-matched control brain tissue”

• Also stated in: § Discussion · plain

  “We report that LDAM are more prevalent in AD brains compared to controls and enriched in individuals with the APOE4/4 genotype.”

In the paper

“This screen revealed regulators of triglyceride metabolism as being a top category of genes required for LD accumulation and ACSL1 as one of the most significant genes required for LD formation”

In the paper

“Through single-nucleus RNA sequencing of brain tissue in Alzheimer’s disease, we have identified a microglial state defined by the expression of the lipid droplet-associated enzyme ACSL1 with ACSL1-positive microglia being most abundant in patients with Alzheimer’s disease having the APOE4/4 genotype.”

In the paper

“ACSL1 was upregulated specifically in microglia in AD brain tissue compared to controls and to a greater extent in APOE4/4 compared to AD-APOE3/3 microglia”

In the paper

“The number of lipid bodies is negatively correlated with cognitive performance, as measured by the mini-mental state exam (MMSE) and positively correlated with Aβ plaque amounts and Tau pathology levels”

In the paper

“However, treatment of iMG with fibrillar Aβ (fAβ) led to a strong increase in LD accumulation which was exacerbated by the presence of the APOE4 AD-risk allele.”

• Also stated in: § Discussion · plain

  “We also describe the induction of triglyceride synthesis and LD accumulation following fAβ treatment in iMG and that the LD induction is greater in isogenic APOE4/4 iMG than APOE3/3 iMG.”

In the paper

“These lipid bodies are most prevalent in AD brain tissue with a slight but not significant increase in individuals with AD-APOE4/4 compared to those with AD-APOE3/3”

In the paper

“This yielded ~100,000 single-nucleus transcriptomes with all main cell types of the brain represented”

In the paper

“13C-labelled TAG lipids synthesized in microglia were also observed in mouse neurons treated with conditioned media from BV2s grown in 13C-labelled glucose, however, this was not robustly detected across several replicates perhaps due to the low isotopic enrichment of these lipids”

In the paper

“the LDAM expression signature, such as ACSL1 and NAMPT upregulation, seems to be unique to microglia.”

In the paper

“In accordance with these RNA-seq results, phenotypic measurements of LD-containing APOE4/4 iMGs indicate that they are dysfunctional in phagocytosis, accumulate lysosomes and secrete inflammation-associated chemokines as measured in the cell culture media”

In the paper

“RNA-seq results indicated that LD-high microglia had higher expression of NF-κB associated pro-inflammatory cytokines (for example, TNFA and IL1B) and lower expression of microglial homeostasis markers compared to LD-low microglia”

In the paper

“On GNE-317 treatment, genes in the PI3K/mTOR and autophagy pathways exhibit changes in expression and we also observed an increase in LC3B protein levels”

In the paper

“Likewise, conditioned media from APOE3/3 and APOE4/4 iMGs with a higher concentrations of LDs induced caspase activation in human neurons, whereas conditioned media from APOE-KO iMG had no effect”

In the paper

“Only the LD-high iMG-conditioned media induced high concentrations of pTau in the iPS cell-derived neurons, whereas the LD-low iMG-conditioned media induced similar concentrations of pTau in the iPS cell-derived neurons as the untreated condition”

In the paper

“This effect was similar when conditioned media from APOE3/3 and APOE4/4 iPS cell-derived iMG were used to treat the human neurons but absent when conditioned media from APOE-KO iMGs were used”

In the paper

“Interestingly, the second top hit was S100A1, an AD risk gene downstream of the LPS and TLR4 response in macrophages.”

In the paper

“LDs themselves have antimicrobial properties and are an evolutionarily conserved form of innate immune defence in macrophages8.”

What the sources say

• ?Not checked — no full text of the sourcecites [8] Bosch, M. et al. Mammalian lipid droplets are innate immune hubs integrating cell metabolism and host defense. Science 370, eaay8085 (2020). source

In the paper

“suggesting that the cells containing lipid bodies near plaques could be ACSL1+ microglia.”

In the paper

“ACSL1 is a key enzyme in LD biogenesis and overexpression of ACSL1 is sufficient to induce triglyceride-specific LD formation in several cell types13,14”

What the sources say

• ?Not checked — no full text of the sourcecites [13] Li, L. O. et al. Overexpression of rat long chain acyl-coa synthetase 1 alters fatty acid metabolism in rat primary hepatocytes. J. Biol. Chem. 281, 37246–37255 (2006). source
• ?Not checked — no full text of the sourcecites [14] Zhao, Z. et al. Effects of overexpression of ACSL1 gene on the synthesis of unsaturated fatty acids in adipocytes of bovine. Arch. Biochem. Biophys. 695, 108648 (2020). source

In the paper

“We used the Scanpy (v.1.9.1) package to discard cells with (1) fewer than 500 genes or (2) less than total 1,000 reads or (3) more than 10% mitochondrial reads or (4) more than 10% ribosomal reads.”

In the paper

“To conduct unbiased single-cell level downstream analysis of the microglial cells which is not biased towards any of the patient samples, first we subsampled the microglia cluster with purification50.”

What the sources say

• ?Not matched

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

  cites [50] Palovics, R., Wyss-Coray, T. & Mirzasoleiman, B. Purification of single-cell transcriptomics data with coreset selection. In The 2022 ICML Workshop on Computational Biology Paper 67 (ICML, 2022).

In the paper

“Irreproducible discovery rate (IDR) was used to test for reproducibility between replicates.”

In the paper

“In duplicate culture conditions, cells were treated with 5 μM Aβ fibrils for 24 h, stained with BODIPY 493/503 (1:2,000 from 1 mg ml−1 of stock solution in DMSO; Thermo Fisher) and top 10% highest and 10% lowest BODIPY fluorescent cells were separated by FACS (Sony, MA900).”

In the paper

“Cells were identified by thresholding with the Triangle method.”

In the paper

“ATAC-seq experiments were performed in replicate and peaks were called with parameters -L 0 -C 0 -fdr 0.9 -minDist 200 -size 200.”

In the paper

“A recent report showed that innate immune triggers (for example, Escherichia coli and Salmonella) induce LD formation in peripheral macrophages as part of an evolutionarily conserved antimicrobial defence in which LDs coated with antimicrobial proteins, such as cathelicidin (CAMP), kill bacteria8.”

What the sources say

• ?Not checked — no full text of the sourcecites [8] Bosch, M. et al. Mammalian lipid droplets are innate immune hubs integrating cell metabolism and host defense. Science 370, eaay8085 (2020). source

In the paper

“P values were corrected with Benjamini–Hochberg procedure49 (false discovery rate (FDR) = 0.05) per comparison.”

What the sources say

• ?Not checked — no full text of the sourcecites [49] Benjamini, Y. & Hochberg, Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J. R. Stat. Soc. B 57, 289–300 (1995). source

In the paper

“Screens were performed in duplicate.”

In the paper

“Enhancing autophagy has previously been explored as a strategy to modulate neurodegeneration disease progression but with a focus on autophagic degradation of proteins41, as opposed to the autophagic degradation of intracellular LDs (lipophagy).”

What the sources say

• ?Not checked — no full text of the sourcecites [41] Heras-Sandoval, D., Pérez-Rojas, J. M., Hernández-Damián, J. & Pedraza-Chaverri, J. The role of PI3K/AKT/mTOR pathway in the modulation of autophagy and the clearance of protein aggregates in neurodegeneration. Cell. Signal. 26, 2694–2701 (2014). source

In the paper

“these experiments do not conclusively demonstrate that there is increased autophagic flux of lipids in this context”

In the paper

“It is likely that the APOE genotype contributes to AD pathogenesis through several distinct mechanisms unique to individual cell-type dysfunction or through crosstalk between these cell types.”

In the paper

“All assays were performed under serum-free conditions (MGdM media).”

In the paper

“For all image analyses, a standard threshold value was chosen and automatically applied to each channel of each image before measurement.”

In the paper

“We set gender, expired age, postmortem interval and the percentage of mitochondrial and ribosomal counts as covariates.”

In the paper

“Intracellular lipid features were quantified using the ‘3D Objects Counter’ command.”

In the paper

“which is found on LD surfaces with antimicrobial properties in macrophages exposed to bacteria8.”

What the sources say

• ?Not checked — no full text of the sourcecites [8] Bosch, M. et al. Mammalian lipid droplets are innate immune hubs integrating cell metabolism and host defense. Science 370, eaay8085 (2020). source

In the paper

“The CARS stacks underwent east shadows correction, Gaussian blur filtering (σ = 1) and background subtraction (30 pixel rolling ball radius).”

In the paper

“Lipid particles were identified by thresholding using the FindFoci method, with a search parameter for half peak value of 0.1, peak parameter for minimum peak height relative to background of 0 and a minimum peak size of 200 pixels.”

In the paper

“For each time point, red and green fluorescence was normalized to the phase confluence per well.”

In the paper

“PI3K inhibition, shown to increase autophagy38.”

What the sources say

• ?Not checked — no full text of the sourcecites [38] Ersahin, T., Tuncbag, N. & Cetin-Atalay, R. The PI3K/AKT/mTOR interactive pathway. Mol. Biosyst. 11, 1946–1954 (2015). source