CoChat
Sign Up
Cited by 104Genomics and Chromatin DynamicsGenetic Associations and EpidemiologyEpigenetics and DNA Methylation

Neuronal and glial 3D chromatin architecture informs the cellular etiology of brain disorders

Open access
Hu, Benxia, Won, Hyejung, Mah, Won, Park, Royce B., Kassim, Bibi, Spiess, Keeley, Kozlenkov, Alexey, Crowley, Cheynna A. +149 more·2021·Nature Communications
DOI 10.1038/s41467-021-24243-0PMC PMC8233376
Open this paper in CoChat — free accountOpen PDF

Abstract

Cellular heterogeneity in the human brain obscures the identification of robust cellular regulatory networks, which is necessary to understand the function of non-coding elements and the impact of non-coding genetic variation. Here we integrate genome-wide chromosome conformation data from purified neurons and glia with transcriptomic and enhancer profiles, to characterize the gene regulatory landscape of two major cell classes in the human brain. We then leverage cell-type-specific regulatory landscapes to gain insight into the cellular etiology of several brain disorders. We find that Alzheimer’s disease (AD)-associated epigenetic dysregulation is linked to neurons and oligodendrocytes, whereas genetic risk factors for AD highlighted microglia, suggesting that different cell types may contribute to disease risk, via different mechanisms. Moreover, integration of glutamatergic and GABAergic regulatory maps with genetic risk factors for schizophrenia (SCZ) and bipolar disorder (BD) identifies shared (parvalbumin-expressing interneurons) and distinct cellular etiologies (upper layer neurons for BD, and deeper layer projection neurons for SCZ). Collectively, these findings shed new light on cell-type-specific gene regulatory networks in brain disorders.

AI snapshot

Written from the paper’s full text.

TL;DR

This paper integrates Hi–C chromosome conformation from sorted human neurons (NeuN+) and glia (NeuN−) with transcriptomic and enhancer data to map cell-type-specific 3D regulatory architecture in the brain. It uses these maps to refine the cellular etiology of Alzheimer’s disease, schizophrenia, and bipolar disorder.

Key findings

  • NeuN+ and NeuN− cells show extensive cell-type-specific chromatin architecture: 287 differential FIREs (145 NeuN+, 142 NeuN−), 253 NeuN+ super-FIREs and 157 NeuN− super-FIREs with only 9 shared, and 187,674 NeuN+ versus 167,551 NeuN− promoter-based interactions.
  • AD-associated epigenetic dysregulation maps to different cell classes by direction: AD hyperacetylated H3K27ac peaks were largely active in NeuN− cells and linked to 676 genes, while AD hypoacetylated peaks were largely active in NeuN+ cells and linked to 460 genes.
  • AD genetic risk converged on microglia: AD SNP-based heritability was enriched in NeuN− enhancers (NeuN− p = 8.41 × 10−9; NeuN+ p = 0.205), and NeuN− H-MAGMA identified 181 AD risk genes, enriched in microglial co-expression modules upregulated in AD.
  • SCZ and BD showed neuronal heritability enrichment and both mapped to Glu- and GABA-specific enhancers; risk genes shared parvalbumin-expressing interneuron (In6) enrichment, while BD showed upper-layer Ex1 specificity and SCZ showed Ex2 specificity.

Methods

  • Hi–C libraries were generated from FANS-sorted NeuN+ and NeuN− nuclei from four dorsolateral prefrontal cortex (DLPFC) samples; RNA-seq was also generated from sorted nuclei, and published H3K27ac data from Glu and GABA neurons were integrated.
  • Chromatin architecture was analyzed as A/B compartments, TADs, FIREs/super-FIREs, and promoter-anchored loops; cell-type-specific enhancer–promoter relationships were built by overlapping differential H3K27ac peaks with Hi–C interactions.
  • Disease analyses used LDSC for SNP-based heritability enrichment and H-MAGMA to map GWAS risk variants to genes using NeuN−, Glu, or GABA regulatory maps; GWAS datasets included AD, SCZ, and BD.

Limitations

  • The paper states that neuronal subtype-specific findings need verification with neuronal subtype-specific Hi–C data (e.g., Glu and GABA Hi–C) and that CRISPR/Cas9 enhancer targeting could further test risk-gene enhancer roles.
  • The paper states that AD epigenetic changes may reflect either cellular composition changes or changes in regulatory landscape/cellular function, requiring future investigation.
  • Evident from the text: Hi–C was based on only four DLPFC samples, and NeuN+/NeuN− fractions are broad cell classes rather than pure cell types; Glu/GABA regulatory maps were inferred by integrating subtype H3K27ac peaks with NeuN+ Hi–C rather than subtype Hi–C.

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

Integrity signals

Registry

Crossref · 10.1038/s41467-021-24243-0

Cites

87 of 87 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

0 of 78 cited sources checked · 0 supported · 0 flagged

  • Findingown data

    Genes located in common FIREs are involved in housekeeping functions and pathways that involve both neurons and oligodendrocytes.

    “Genes located in common FIREs were involved in housekeeping functions (e.g., microtubule polymerization, cell junction assembly) as well as pathways that involve both neurons and oligodendrocytes (e.g., main axon, Supplementary Data 1).”
    Quote in the paperNot a substantial claim
  • Findingown data

    1,248 FIREs are shared between NeuN+ and NeuN− cells based on coordinate-level overlap and similarities in FIRE scores.

    “Among them, 1248 FIREs were shared between NeuN+ and NeuN− cells based on the coordinate-level overlap and similarities in FIRE scores (hereafter referred to as common FIREs, “Methods”, Supplementary Data 1).”
    Quote in the paperSubstantial claim
  • Findingown data

    A substantial fraction of chromatin interactions are distal, with ~50% bringing two genomic regions linearly separated by 320 kb or more into close proximity.

    “A substantial fraction of chromatin interactions were distal, as ~50% of chromatin interactions brought two genomic regions linearly separated by 320 kb or more into close proximity (Supplementary Fig. 7c) and captured complex enhancer–promoter interactions.”
    Quote in the paperNot a substantial claim
  • Findingown data

    NeuN− distal enhancers are enriched for motifs for the IRF family, such as IRF4, IRF7, IRF8, and IRF9.

    “NeuN− distal enhancers were also enriched for motifs for the IRF family, such as IRF4, IRF7, IRF8, and IRF9 (Supplementary Fig. 7g), which contains key regulators of neural immune pathways expressed in glial cells31.”
    Quote in the paperNot a substantial claim
  • Findingown data

    Over 75% of promoter-based interactions are detected within TADs.

    “Over 75% of promoter-based interactions were detected within TADs (Supplementary Fig. 7a).”
    Quote in the paperNot a substantial claim
  • May combine two claims. 51
  • Classified as probably not a substantial claim. 7

  • May combine two claims.

    Findingown data

    Over 95% of super-FIREs overlap with differential H3K27ac peaks and all super-FIREs overlap with promoters.

    “Over 95% of super-FIREs overlapped with differential H3K27ac peaks and all super-FIREs overlapped with promoters, indicating that they have a particularly strong cell-type-specific regulatory impact (Supplementary Fig. 5b, c).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Extensive compartment switching occurs between NeuN+ and NeuN− cells: 4,333 regions switch from compartment A to B between NeuN− and NeuN+ cells, and 2,098 regions switch from compartment B to A in NeuN+ to NeuN− cells.

    “We detected extensive compartment switching between NeuN+ and NeuN− cells: 4,333 regions (in 100kb resolution) switched from compartment A to B between NeuN− to NeuN+ cells, while 2098 regions switched from compartment B to A in NeuN+ to NeuN− cells.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Only 9 super-FIREs are detected in both NeuN+ and NeuN− cells, leaving 253 and 157 cell-type-specific super-FIREs in NeuN+ and NeuN− cells, respectively.

    “Only 9 super-FIREs were detected in both NeuN+ and NeuN− cells, leaving 253 and 157 cell-type-specific super-FIREs in NeuN+ and NeuN− cells, respectively (Supplementary Fig. 5a, Supplementary Data 2).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    There are 287 differential FIREs between NeuN+ and NeuN− cells, with 145 NeuN+ FIREs and 142 NeuN− FIREs.

    “detecting 287 differential FIREs between NeuN+ (145) and NeuN− (142) cells (hereby referred to as NeuN+ and NeuN− FIREs, respectively, Fig. 1b, Supplementary Data 2).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    There are 3,966 FIREs in NeuN+ cells and 3,967 FIREs in NeuN− cells.

    “We detected 3966 and 3967 FIREs in NeuN+ and NeuN− cells, respectively (Fig. 1a and Supplementary Data 1)24.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [24] Crowley, C. et al. FIREcaller: detecting frequently interacting regions from Hi-C data. Comput. Struct. Biotechnol. J. 19, 355–362 (2021). source

      Not checked in the free run.

  • May combine two claims.

    Findingown data

    Genes in compartments that switch from A to B between NeuN− and NeuN+ are highly expressed in oligodendrocytes and astrocytes, while genes that switch from B to A in NeuN− to NeuN+ are highly expressed in neurons.

    “Importantly, genes located in compartments that switch from A to B between NeuN− and NeuN+ were highly expressed in oligodendrocytes and astrocytes, while those that switch from B to A in NeuN− to NeuN+ were highly expressed in neurons, suggesting that the difference in chromosome conformation between NeuN+ and NeuN− cells is associated with cell-type-specific gene regulation (Supplementary Fig. 3”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    NeuN+ super-FIREs overlap with genes functioning in synapses and ion gated channels that are highly expressed in neurons, while NeuN− super-FIREs overlap with genes involved in cell adhesion, myelination, and ensheathment of neurons, with high expression in glia.

    “NeuN+ super-FIREs overlapped with genes functioning in synapses and ion gated channels that were highly expressed in neurons, while NeuN− super-FIREs overlapped with genes involved in cell adhesion, myelination, and ensheathment of neurons, with high expression in glia (Supplementary Fig. 5d-f, 6).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    NeuN+ cells show high structural similarity with adult brains, fetal brains, and iPSC-derived neurons.

    “NeuN+ cells showed high similarity with adult brains, fetal brains, and iPSC-derived neurons.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    There are 187,674 promoter-based interactions in NeuN+ cells and 167,551 in NeuN− cells.

    “We detected 187,674 and 167,551 promoter-based interactions3,4 from NeuN+ and NeuN− cells, respectively (Fig. 2a).”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [3] Wang, D. et al. Comprehensive functional genomic resource and integrative model for the human brain. Science 362, eaat8464 (2018).

      Not checked in the free run.

    • cites [4] Won, H. et al. Chromosome conformation elucidates regulatory relationships in developing human brain. Nature 538, 523–527 (2016). source

      Not checked in the free run.

  • May combine two claims.

    Findingown data

    Roughly 37% of chromatin interactions occur between enhancers and promoters and ~23% occur between two different promoters.

    “Roughly 37% of interactions occurred between enhancers and promoters and ~23% of interactions occurred between two different promoters (Supplementary Fig. 7b).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    GRIN2B overlaps with a NeuN+ FIRE, and OLIG1 and OLIG2 overlap with a NeuN− FIRE.

    “Salient examples include GRIN2B, which overlaps with a NeuN+ FIRE, and OLIG1 and OLIG2, which overlap with a NeuN− FIRE (Fig. 1e, f, Supplementary Fig. 4).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    NeuN+ FIREs overlap with neuronal genes enriched for synaptic function, while NeuN− FIREs overlap with genes involved in myelination, glial differentiation, and oligodendrocyte differentiation.

    “Indeed, NeuN+ FIREs overlapped with neuronal genes that were enriched for synaptic function (Fig. 1g), while NeuN− FIREs overlapped with genes involved in myelination, glial differentiation, and oligodendrocyte differentiation (Fig. 1h).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Hypoacetylated peaks in AD are linked to 460 genes using NeuN+ Hi-C data, and hyperacetylated peaks in AD are linked to 676 genes using NeuN− Hi-C data.

    “We were able to link hypoacetylated peaks in AD to 460 genes using NeuN+ Hi–C data and hyperacetylated peaks in AD to 676 genes using NeuN− Hi–C data (hereby referred to as NeuN+ hypo- and NeuN− hyperacetylated genes, respectively, Fig. 3c, Supplementary Data 5).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingcitation

    AD risk genes are highly expressed in postnatal brain samples and microglia.

    “Consistent with previous studies16,42,49, these genes were highly expressed in the postnatal brain samples (Fig. 4d) and microglia (Fig. 4e).”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites [16] Nott, A. et al. Brain cell type–specific enhancer–promoter interactome maps and disease-risk association. Science 366, 1134–1139 (2019). source

      Not checked in the free run.

    • cites [42] Lambert, J. C. et al. Meta-analysis of 74,046 individuals identifies 11 new susceptibility loci for Alzheimer’s disease. Nat. Genet. 45, 1452–1458 (2013). source

      Not checked in the free run.

    • cites [49] Jansen, I. E. et al. Genome-wide meta-analysis identifies new loci and functional pathways influencing Alzheimer’s disease risk. Nat. Genet. 51, 404–413 (2019). source

      Not checked in the free run.

  • May combine two claims.

    Findingown data

    HOMER1 is engaged in NeuN+ specific peaks and loops and its expression is significantly higher in NeuN+ cells compared with NeuN− cells.

    “For example, a gene that encodes a synaptic scaffolding protein, HOMER132, was engaged in NeuN+ specific peaks and loops and its expression was significantly higher in NeuN+ cells compared with NeuN− cells (Fig. 2b).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [32] Tu, J. C. et al. Homer binds a novel proline-rich motif and links group 1 metabotropic glutamate receptors with IP3 receptors. Neuron 21, 717–726 (1998). source

      Not checked in the free run.

  • May combine two claims.

    Findingown data

    45,911 Glu-specific and 32,169 GABA-specific H3K27ac peaks are assigned to 6,234 and 4,342 genes, respectively.

    “We obtained 45,911 Glu-specific and 32,169 GABA-specific H3K27ac peaks and assigned them to 6234 and 4342 genes, respectively (Supplementary Fig. 8a, Supplementary Data 4).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Genes assigned to NeuN+ specific peaks are more highly expressed in neurons, while genes assigned to NeuN− specific peaks are highly expressed in oligodendrocytes and astrocytes.

    “Lastly, genes assigned to NeuN+ specific peaks were more highly expressed in neurons, while those assigned to NeuN− specific peaks were highly expressed in oligodendrocytes and astrocytes (Fig. 2f), demonstrating the tight relationship between cell-type-specific chromatin architecture and expression signature.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    NeuN+ enhancer-promoter interactions are enriched for synaptic and axonal genes, whereas NeuN− enhancer-promoter interactions are associated with actin-based motility.

    “In line with these findings, NeuN+ enhancer–promoter interactions were enriched for synaptic and axonal genes, whereas NeuN− enhancer–promoter interactions were associated with actin-based motility (Supplementary Data 3).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    NeuN+ hypoacetylated genes include synaptic genes, while NeuN− hyperacetylated genes are involved in catalytic activity and glycoprotein binding.

    “GO analysis demonstrated that NeuN+ hypoacetylated genes included synaptic genes while NeuN− hyperacetylated genes were involved in catalytic activity and glycoprotein binding (Supplementary Data 5).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    The BIN1 promoter is connected to multiple enhancers in NeuN− cells, but not in NeuN+ cells.

    “Notably, the BIN1 promoter was connected to multiple enhancers in NeuN−, but not in NeuN+ cells, demonstrating the highly cell-type-specific gene regulatory landscape within the locus.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingcitation

    The T-M8 and T-M14 modules are involved in transcriptional regulation and cellular proliferation and are annotated as astrocyte-specific.

    “These modules were involved in transcriptional regulation and cellular proliferation and were annotated as astrocyte-specific41.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [41] Seyfried, N. T. et al. A multi-network approach identifies protein-specific co-expression in asymptomatic and symptomatic Alzheimer’s disease. Cell Syst. 4, 60–72.e4 (2017). source

      Not checked in the free run.

  • May combine two claims.

    Findingown data

    NeuN+ hypoacetylated genes are highly expressed in neurons, while NeuN− hyperacetylated genes are highly expressed in oligodendrocytes, astrocytes, and endothelial cells.

    “Cellular expression profiles also confirmed these findings, as NeuN+ hypoacetylated genes were highly expressed in neurons, while NeuN− hyperacetylated genes were highly expressed in oligodendrocytes, astrocytes, and endothelial cells (Fig. 3f).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    SOX10 is engaged in NeuN− specific H3K27ac peaks and Hi-C loops and has higher expression in NeuN− cells than in NeuN+ cells.

    “In contrast, a glial gene, SOX1033, was engaged in NeuN− specific H3K27ac peaks and Hi–C loops, and had higher expression in NeuN− cells than in NeuN+ cells (Fig. 2c).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [33] Kuhlbrodt, K., Herbarth, B., Sock, E., Hermans-Borgmeyer, I. & Wegner, M. Sox10, a novel transcriptional modulator in glial cells. J. Neurosci. 18, 237–250 (1998). source

      Not checked in the free run.

  • May combine two claims.

    Findingown data

    AD risk genes are enriched for amyloid−beta pathways, lipoprotein assembly, and immune processes.

    “AD risk genes were enriched for amyloid−beta (Aβ) pathways, lipoprotein assembly, and immune processes (Fig. 4c, Supplementary Data 6).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Genes assigned to Glu peaks are highly expressed in excitatory neurons such as L2/3 pyramidal neurons (Ex1) and L5/6 corticothalamic projection neurons (Ex7), while genes assigned to GABA peaks are highly expressed in inhibitory neurons such as parvalbumin-expressing GABA interneurons (In6).

    “Genes assigned to Glu peaks were highly expressed in excitatory neurons, such as layers (L)2/3 pyramidal neurons (Ex1) and L5/6 corticothalamic projection neurons (Ex7), while genes assigned to GABA peaks were highly expressed in inhibitory neurons, such as parvalbumin-expressing GABA interneurons (In6, Supplementary Fig. 8b, the definition of Ex1–8 and In1–8 neuronal subtypes is available in”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    EHD1 is both hyperacetylated and highly expressed in neurotypical NeuN− cells compared to NeuN+ cells.

    “EHD1 was both hyperacetylated and highly expressed in neurotypical NeuN− cells compared to NeuN+ cells.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    AD-associated hyperacetylated H3K27ac peaks are largely active in NeuN− cells, while AD-associated hypoacetylated peaks are largely active in NeuN+ cells in neurotypical controls.

    “We detected that AD-associated hyperacetylated peaks were largely active in NeuN− cells, while AD-associated hypoacetylated peaks were largely active in NeuN+ cells in neurotypical controls (Fig. 3b).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    CACNG3 promoter forms a loop with a peak that is preferentially active in NeuN+ cells and is hypoacetylated in postmortem AD brain.

    “NeuN+ hypoacetylated genes include CACNG3, whose promoter formed a loop with a peak, which was preferentially active in NeuN+ cells and was hypoacetylated in postmortem AD brain.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    GRIK4 displays complex chromatin interactions with multiple Glu peaks, and GAD1 is engaged in GABA peaks via chromatin interactions.

    “GRIK4, a gene that encodes an ionotropic class of glutamate receptor36, displayed complex chromatin interactions with multiple Glu peaks. In contrast, GAD1, a gene that encodes a well-known cellular marker for inhibitory neurons37, was engaged in GABA peaks via chromatin interactions (Supplementary Fig. 8c).”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [36] Szpirer, C. et al. The genes encoding the glutamate receptor subunits KA1 and KA2 (GRIK4 and GRIK5) are located on separate chromosomes in human, mouse, and rat. Proc. Natl Acad. Sci. USA 91, 11849–11853 (1994). source

      Not checked in the free run.

    • cites [37] Erlander, M. G., Tillakaratne, N. J., Feldblum, S., Patel, N. & Tobin, A. J. Two genes encode distinct glutamate decarboxylases. Neuron 7, 91–100 (1991). source

      Not checked in the free run.

  • May combine two claims.

    Findingown data

    10,167 cell-type-specific H3K27ac peaks are assigned to 7,828 genes via chromatin interactions in NeuN+ cells, and 11,242 peaks are assigned to 8,851 genes in NeuN− cells.

    “We were able to assign 10,167 and 11,242 cell-type-specific H3K27ac peaks to 7828 and 8851 genes via chromatin interactions in NeuN+ and NeuN− cells, respectively (Fig. 2a, Supplementary Data 3).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Glu-BD genes are enriched for neurogenesis, cell adhesion molecules, and synapses, while GABA-BD genes are enriched for transcriptional regulation and NMDA receptor activity.

    “Glu-BD genes were enriched for neurogenesis, cell adhesion molecules, and synapses, while GABA-BD genes were enriched for transcriptional regulation and NMDA receptor activity (Supplementary Data 7).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    AD-associated hyperacetylation was linked to oligodendrocyte genes and astrocyte co-expression modules, while genetic risk factors for AD were mapped onto a different glial cell type, microglia.

    “It is of note that while AD-associated hyperacetylation was linked to oligodendrocyte genes and astrocyte co-expression modules, genetic risk factors for AD were mapped onto a different glial cell type, microglia.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    AD-associated hyperacetylation was enriched for glial enhancers associated with upregulation of astrocytic co-expression modules in AD, while AD-associated hypoacetylation was enriched for neuronal enhancers potentially affecting downregulation of a neuronal co-expression module in AD.

    “Notably, AD-associated hyperacetylation was enriched for glial enhancers that are associated with upregulation of astrocytic co-expression modules in AD, while AD-associated hypoacetylation was enriched for neuronal enhancers that potentially affect downregulation of a neuronal co-expression module in AD.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    753 SCZ risk genes are obtained from Glu-specific regulatory interactions and 624 from GABA-specific regulatory interactions.

    “obtaining 753 SCZ risk genes from Glu (hereby referred to as Glu-SCZ genes) and 624 risk genes from GABA-specific regulatory interactions integrated via H-MAGMA (GABA-SCZ genes) (Supplementary Data 7).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    The genome-wide analysis of chromosome conformation in neurons and glia captures the major known elements of 3D architecture such as compartments, FIREs, and loops, addressing the roles and impact of these different hierarchical units in cell-type-specific gene regulation.

    “The genome-wide analysis of chromosome conformation in these two cell classes captures the major known elements of 3D architecture such as compartments, FIREs, and loops, addressing the roles and impact of these different hierarchical units in cell-type-specific gene regulation.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Glu-SCZ genes display widespread expression among many Glu neuronal subclasses, with relatively higher expression signatures in L3/4 neurons (Ex2) and L5/6 corticothalamic projection neurons (Ex7).

    “Glu-SCZ genes displayed widespread expression among many Glu neuronal subclasses, with relatively higher expression signatures in L3/4 neurons (Ex2) and L5/6 corticothalamic projection neurons (Ex7).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [8] Lake, B. B. et al. Neuronal subtypes and diversity revealed by single-nucleus RNA sequencing of the human brain. Science 352, 1586–1590 (2016). source

      Not checked in the free run.

  • May combine two claims.

    Findingown data

    143 BD candidate risk genes are identified from Glu and 101 from GABA H-MAGMA.

    “We also identified 143 and 101 BD candidate risk genes from Glu (Glu-BD genes) and GABA (GABA-BD genes) H-MAGMA, respectively (Supplementary Data 7).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Integrating Hi-C data from sorted neurons with H3K27ac peaks identified in Glu and GABA neurons refined maps of neuronal chromatin architecture and led to a successful delineation of neuronal subtype-specific regulatory relationships.

    “We further refined maps of neuronal chromatin architecture by integrating Hi–C data from sorted neurons with H3K27ac peaks identified in two major neuronal subtypes, Glu and GABA neurons, which led to a successful delineation of neuronal subtype-specific regulatory relationships.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Glu-SCZ genes are implicated in synaptic organization, ion channel activity, and neuron projection, while GABA-SCZ genes are associated with dendrite, axon, and hormonal response.

    “Glu-SCZ genes were implicated in synaptic organization, ion channel activity, and neuron projection, while GABA-SCZ genes were associated with dendrite, axon, and hormonal response (Supplementary Data 7).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingcitation

    Common genetic risk for SCZ and BD indicated shared pathways coalescing onto parvalbumin-positive interneurons, paralleling robust pathologic findings in SCZ and BD.

    “It is also of note that common genetic risk for SCZ and BD also indicated shared pathways coalescing onto parvalbumin-positive interneurons, paralleling what has been among the most robust pathologic findings in SCZ and BD63,64.”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [63] Enwright Iii, J. F. et al. Transcriptome alterations of prefrontal cortical parvalbumin neurons in schizophrenia. Mol. Psychiatry 23, 1606–1613 (2018). source

      Not checked in the free run.

    • cites [64] Mellios, N. et al. Molecular determinants of dysregulated GABAergic gene expression in the prefrontal cortex of subjects with schizophrenia. Biol. Psychiatry 65, 1006–1014 (2009). source

      Not checked in the free run.

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

    Interpretationcitation

    The high similarity of NeuN+ cells with fetal brain may reflect extensive neurogenesis at midgestation, the developmental stage at which the fetal brain was obtained.

    “The fact that NeuN+ cells show high similarity with fetal brain may reflect extensive neurogenesis at midgestation, the developmental stage at which the fetal brain was obtained22.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [22] Polioudakis, D. et al. A single-cell transcriptomic atlas of human neocortical development during mid-gestation. Neuron 103, 785–801.e8 (2019). source

      Not checked in the free run.

  • May combine two claims.

    Interpretationown data

    Gene expression is coordinately controlled by physical interactions of promoters with enhancers, indicative of transcriptional regulators that underlie cell-type-specific gene regulation.

    “Together, our results confirmed that gene expression was coordinately controlled by physical interactions of promoters with enhancers, indicative of transcriptional regulators that underlie cell-type-specific gene regulation.”
    Quote in the paperSubstantial claim

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

    Premisecitation

    Multiple enhancers can interact with one promoter.

    “consistent with the previous findings that multiple enhancers can interact with one promoter3,7.”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [3] Wang, D. et al. Comprehensive functional genomic resource and integrative model for the human brain. Science 362, eaat8464 (2018).

      Not checked in the free run.

    • cites [7] Jung, I. et al. A compendium of promoter-centered long-range chromatin interactions in the human genome. Nat. Genet. 51, 1442–1449 (2019). source

      Not checked in the free run.

  • May combine two claims.

    Study design and conductown data

    Hi-C libraries from NeuN+ and NeuN− cells show high reproducibility, with samples clustering by cell type rather than individual and high correlation between same cell types from different individuals (NeuN+ = 0.95-0.97; NeuN− = 0.86-0.91).

    “Notably, we found that samples were clustered by cell types, not individuals, and showed high correlation between the same cell types from different individuals (NeuN+ = 0.95–0.97; NeuN− = 0.86-0.91; Supplementary Fig. 3a), demonstrating the reproducibility across Hi–C libraries.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [19] Yang, T. et al. HiCRep: assessing the reproducibility of Hi-C data using a stratum-adjusted correlation coefficient. Genome Res. 27, 1939–1949 (2017). source

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Premisecitation

    Promoter interaction profiles have been inferred from neurons, oligodendrocytes, and microglia obtained from the postnatal human cerebral cortex.

    “Recently, promoter interaction profiles were inferred from three types of brain cells (neurons, oligodendrocytes, and microglia) obtained from the postnatal human cerebral cortex16.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [16] Nott, A. et al. Brain cell type–specific enhancer–promoter interactome maps and disease-risk association. Science 366, 1134–1139 (2019). source

      Not checked in the free run.

  • May combine two claims.

    Interpretationreasoning

    In AD brains, many neuronal genes are downregulated due to their hypoacetylation in neurons, whereas many glial genes are upregulated due to their hyperacetylation in glia.

    “These results collectively suggest that in AD brains, many neuronal genes are downregulated due to their hypoacetylation in neurons, whereas many glial genes are upregulated due to their hyperacetylation in glia.”
    Quote in the paperSubstantial claim

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

    Premisecitation

    The IRF family contains key regulators of neural immune pathways expressed in glial cells.

    “which contains key regulators of neural immune pathways expressed in glial cells31.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [31] Tamura, T., Yanai, H., Savitsky, D. & Taniguchi, T. The IRF family transcription factors in immunity and Oncogenesis. Annu. Rev. Immunol. 26, 535–584 (2008). source

      Not checked in the free run.

  • May combine two claims.

    Interpretationreasoning

    Differential enrichment of genetic risk factors and epigenetic regulation underscores different expression signatures in AD with distinct cellular specificities and likely causal relationships.

    “Since genetic risk resides at the initiation of the causal chain in disease pathogenesis, this differential enrichment of genetic risk factors and epigenetic regulation underscores different expression signatures in AD with distinct cellular specificities and likely causal relationships.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Genetic risk resides at the initiation of the causal chain in disease pathogenesis.

  • May combine two claims.

    Premisecitation

    Recent single-cell sequencing studies identify genetic risk enrichment and gene expression changes within microglia in AD, accompanied by significant gene expression changes in other cells, including astrocytes.

    “This is consistent with recent single-cell sequencing studies identifying genetic risk enrichment and gene expression changes within microglia in AD, accompanied by significant gene expression changes in other cells, including astrocytes50–52, which is where we observe epigenetic signature enrichment.”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites [50] Mathys, H. et al. Single-cell transcriptomic analysis of Alzheimer’s disease. Nature 570, 332–337 (2019). source

      Not checked in the free run.

    • cites [51] Grubman, A. et al. A single-cell atlas of entorhinal cortex from individuals with Alzheimer’s disease reveals cell-type-specific gene expression regulation. Nat. Neurosci. 22, 2087–2097 (2019). source

      Not checked in the free run.

    • cites [52] Sala Frigerio, C. et al. The major risk factors for Alzheimer’s disease: age, sex, and genes modulate the microglia response to Aβ plaques. Cell Rep. 27, 1293–1306.e6 (2019). source

      Not checked in the free run.

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

    Interpretationown data

    The results highlight the importance of obtaining cell-type-specific epigenetic landscapes in diseases.

    “These results highlight the importance of obtaining cell-type-specific epigenetic landscapes in diseases.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationreasoning

    Assigning AD-associated H3K27ac peaks to target genes on the basis of linear distance is likely to be underpowered and inaccurate.

    “so this initial assignment is likely to be underpowered and inaccurate.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationreasoning

    AD-associated H3K27ac peaks obtained from brain homogenates would mask opposing interactions and obscure the cell-type-specific substrate of these changes.

    “Furthermore, these peaks were obtained from the brain homogenates, which would mask opposing interactions and obscure the cell-type-specific substrate of these changes.”
    Quote in the paperSubstantial claim

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

    Interpretationreasoning

    Neuronal subtype-specific gene regulatory networks help refine the cellular etiology of disease and potentially identify discrete molecular neuropathology across psychiatric disorders.

    “Here, we reasoned that neuronal subtype-specific gene regulatory networks would help refine the cellular etiology of disease and potentially identify discrete molecular neuropathology across psychiatric disorders.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationreasoning

    The findings suggest cellular substrates for molecular convergence (Ex7 and In6 shared between SCZ and BD) and divergence (Ex1 BD-specific, Ex2 SCZ-specific) among two highly genetically correlated disorders.

    “these findings suggest cellular substrates for molecular convergence (Ex7 and In6 are shared between two disorders) and divergence (Ex1 is BD-specific, while Ex2 is SCZ-specific) among two highly genetically correlated disorders.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim.

    Premisecitation

    Previous work indicates neurons as the central cell type harboring genetic risk for the majority of psychiatric disorders.

    “In contrast to AD in which multiple glial cells are implicated, previous work has indicated neurons as the central cell-type harboring genetic risk for the majority of psychiatric disorders3,22,46,58–61.”
    Quote in the paperSubstantial claim0 of 7 sources checked
    • cites [3] Wang, D. et al. Comprehensive functional genomic resource and integrative model for the human brain. Science 362, eaat8464 (2018).

      Not checked in the free run.

    • cites [22] Polioudakis, D. et al. A single-cell transcriptomic atlas of human neocortical development during mid-gestation. Neuron 103, 785–801.e8 (2019). source

      Not checked in the free run.

    • cites [46] Sey, N. Y. A. et al. A computational tool (H-MAGMA) for improved prediction of brain-disorder risk genes by incorporating brain chromatin interaction profiles. Nat. Neurosci. https://doi.org/10.1038/s41593-020-0603-0 (2020). source

      Not checked in the free run.

    • cites [58] Satterstrom, F. K. et al. Large-scale exome sequencing study implicates both developmental and functional changes in the neurobiology of autism. Cell. https://doi.org/10.1016/j.cell.2019.12.036 (2020). source

      Not checked in the free run.

    • cites [59] Gandal, M. J. et al. Transcriptome-wide isoform-level dysregulation in ASD, schizophrenia, and bipolar disorder. Science 362, eaat8127 (2018). source

      Not checked in the free run.

    • cites [60] Ruzzo, E. K. et al. Inherited and de novo genetic risk for autism impacts shared networks. Cell 178, 850–866.e26 (2019). source

      Not checked in the free run.

    • cites [61] Skene, N. G. et al. Genetic identification of brain cell types underlying schizophrenia. Nat. Genet. 50, 825–833 (2018). source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Marzi et al. reported AD-associated H3K27ac peaks that are either hypo- or hyperacetylated in AD-affected individuals compared to age-matched low-pathology controls and linked these peaks to target genes on the basis of linear distance.

    “In their original work, Marzi et al.38 reported AD-associated H3K27ac peaks that are either hypo- or hyperacetylated in AD-affected individuals compared to age-matched low-pathology controls38 and linked these peaks to target genes on the basis of linear distance.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Marzi et al.38 Marzi, S. J. et al. A histone acetylome-wide association study of Alzheimer’s disease identifies disease-associated H3K27ac differences in the entorhinal cortex. Nat. Neurosci. 21, 1618–1627 (2018). source

      Not checked in the free run.

  • May combine two claims.

    Interpretationown data

    The study provides a foundation for future study by characterizing cell-type-specific 3D chromatin structure in the adult human brain, which can be used to improve understanding of the gene regulatory landscape in brain disorders.

    “Our study provides a foundation for such future study by characterizing cell-type-specific 3D chromatin structure in the adult human brain, which we show can be used to improve our understanding of gene regulatory landscape in brain disorders.”
    Quote in the paperSubstantial claim
57 of 57

  • The paper relies on this without support.

    Changes in gene expression or histone modifications may be a consequence or compensatory in disease and not necessarily causal.

    “Changes in gene expression or histone modifications may be a consequence or compensatory in disease and not necessarily causal.”

  • The paper relies on this without support.

    How AD-associated H3K27ac changes relate to cell-type-specific vulnerability is not known.

    “Studies in AD have revealed changes in gene regulation manifested by differences in H3K27ac in bulk tissue38, but how this relates to cell-type-specific vulnerability is not known.”

  • The paper relies on this without support.

    Hi-C data from neuronal subtypes is unavailable.

    “Since Hi–C data from neuronal subtypes is yet unavailable, we constructed Glu and GABA H-MAGMA maps by integrating Glu- and GABA-enhancers with NeuN+ Hi–C data, respectively (“Methods”)”

  • The paper relies on this without support.

    Genetic risk resides at the initiation of the causal chain in disease pathogenesis.

    “Since genetic risk resides at the initiation of the causal chain in disease pathogenesis, this differential enrichment of genetic risk factors and epigenetic regulation underscores different expression signatures in AD with distinct cellular specificities and likely causal relationships.”
4 of 4

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?

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.

  • NeuN− FIRE criterion selects NeuN+-high regions

    Minor

    The methods say NeuN− differential FIREs are defined in the opposite way from NeuN+ FIREs, but both stated criteria require high NeuN+ scores and low NeuN− scores. As written, the NeuN− criterion cannot identify regions with preferentially high NeuN− FIRE scores.

  • SCZ-specific neuronal subtype changes from Ex2 to Ex5

    Minor

    The results identify Ex2 as SCZ-specific relative to BD, but the discussion identifies Ex5 as the subtype distinguishing SCZ from BD.

  • Compartment-switch direction is reversed for the 2,098 regions

    Minor

    The text first describes the B-to-A switch as occurring from NeuN+ to NeuN−, then describes that same direction of switch as occurring from NeuN− to NeuN+.

3 of 3

Cites

Flagged first, then by citations

Research works printed but not linked: 1

  1. Stark, R. et al. DiffBind: differential binding analysis of ChIP-Seq peak data. R. Package Version 100, 4–3 (2011).

Cited by

Sorted by citations

Related

Authors
Hu, B., Won, H., Mah, W., Park, R. B., Kassim, B., Spiess, K., Kozlenkov, A., Crowley, C. A., Pochareddy, S., The PsychENCODE Consortium, Ashley-Koch, A. E., Crawford, G. E., Garrett, M. E., Song, L., Safi, A., Johnson, G. D., Wray, G. A., Reddy, T. E., Goes, F. S., Zandi, P., Bryois, J., Jaffe, A. E., Price, A. J., Ivanov, N. A., Collado-Torres, L., Hyde, T. M., Burke, E. E., Kleiman, J. E., Tao, R., Shin, J. H., Girdhar, K., Jiang, Y., Kundakovic, M., Brown, L., Wiseman, J. R., Zharovsky, E., Jacobov, R., Devillers, O., Flatow, E., Hoffman, G. E., Belmont, J., Del Valle, D., Francoeur, N., Hadjimichael, E., Pinto, D., van Bakel, H., Roussos, P., Fullard, J. F., Bendl, J., Hauberg, M. E., Charney, A. W., Haroutunian, V., Lipska, B. K., Lewis, D. A., Hahn, C.-G., Mangravite, L. M., Peters, M. A., Chae, Y., Peng, J., Niu, M., Wang, X., Webster, M. J., Beach, T. G., Chen, C., Jiang, Y., Dai, R., Wang, Y., Xia, Y., Shieh, A. W., Liu, C., Grennan, K. S., Vadukapuram, R., Giase, G., Fitzgerald, D., Cheng, L., Brown, M., Brown, M., Brunetti, T., Goodman, T., Alsayed, M., White, K. P., Ray, M., Polioudakis, D., Wamsley, B., Yin, J., De La Torre Ubieta, L., Gandal, M. J., Swarup, V., Sanders, S. J., State, M. W., Werling, D. M., An, J.-Y., Sheppard, B., Willsey, A. J., Kefi, A., Mattei, E., Purcaro, M., Weng, Z., Moore, J., Pratt, H., Huey, J., Borrman, T., Sullivan, P. F., Giusti-Rodriguez, P., Kim, Y., Szatkiewicz, J., Rhie, S. K., Armoskus, C., Camarena, A., Farnham, P. J., Spitsyna, V. N., Witt, H., Schreiner, S., Evgrafov, O. V., Knowles, J. A., Gerstein, M., Liu, S., Navarro, F. C. P., Warrell, J., Clarke, D., Emani, P. S., Gu, M., Shi, X., Xu, M., Yang, Y. T., Kitchen, R. R., Gürsoy, G., Zhang, J., Carlyle, B. C., Nairn, A. C., Li, M., Skarica, M., Li, Z., Sousa, A. M. M., Santpere, G., Choi, J., Zhu, Y., Gao, T., Miller, D. J., Cherskov, A., Yang, M., Amiri, A., Coppola, G., Mariani, J., Scuderi, S., Szekely, A., Vaccarino, F. M., Wu, F., Weissman, S., Wang, D., Roychowdhury, T., Abyzov, A., Li, Y., Dracheva, S., Sestan, N., Akbarian, S., & Geschwind, D. H.
Affiliations
  • University of North Carolina at Chapel Hill (US)
  • Icahn School of Medicine at Mount Sinai (US)
  • James J. Peters VA Medical Center (US)
  • Yale University (US)
  • Duke University (US)
  • Johns Hopkins University (US)
  • Karolinska Institutet (SE)
  • Lieber Institute for Brain Development (US)
  • Central South University (CN)
  • National Institutes of Health (US)
  • University of Pittsburgh (US)
  • University of Pennsylvania (US)
  • Sage Bionetworks (US)
  • St. Jude Children's Research Hospital (US)
  • Stanley Medical Research Institute (US)
  • Banner Sun Health Research Institute (US)
  • SUNY Upstate Medical University (US)
  • University of Chicago (US)
  • University of California, Los Angeles (US)
  • University of California, San Francisco (US)
  • University of Illinois Chicago (US)
  • University of Massachusetts Chan Medical School (US)
  • University of Southern California (US)
  • SUNY Downstate Health Sciences University (US)
  • University of Wisconsin–Madison (US)
  • Mayo Clinic (US)
Corresponding author
  • Won, Hyejung
  • Geschwind, Daniel H.
ORCID
Venue
Nature Communications
Publisher
Springer Science and Business Media LLC
Published
2021-06-25
Volume
12
Issue
1
Pages
3968-3968
ISSN
2041-1723 (ISSN-L)
DOI
10.1038/s41467-021-24243-0
PMC
PMC8233376
OpenAlex
W3173670503
Semantic Scholar
723f2286bfab54bcc9d0127b5899e81ffa957786
Cited by
103
References
86
Citations by year
2026: 6 · 2025: 21 · 2024: 24 · 2023: 22 · 2022: 24 · 2021: 5 · 2020: 1
FWCI
5.96
Type
journal article
Language
en
Open access
Licence
Locations
  • Nature Communications (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • eScholarship@UMassChan (submittedVersion, cc-by)
  • RePEc: Research Papers in Economics (submittedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • eScholarship (California Digital Library) (submittedVersion)
  • eScholarship (California Digital Library) (submittedVersion, cc-by)
  • Repositori digital de la UPF (Universitat Pompeu Fabra) (submittedVersion, cc-by)
Funders
  • U.S. Department of Health & Human Services | NIH | National Institute of Mental Health (R00MH113823, DP2MH122403, P50MH106438, R01MH094714, U01MH103339, R01MH110927, R01MH100027)
  • U.S. Department of Health & Human Services | NIH | National Institute on Drug Abuse (R21DA051921)
  • U.S. Department of Health & Human Services | NIH | National Institute on Aging (R01AG066871)
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Genomics and Chromatin Dynamics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Genetics › Genetic Associations and Epidemiology
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Epigenetics and DNA Methylation
Keywords
3D genome organization, chromosome conformation, neurons, glia, oligodendrocytes, microglia, Alzheimer's disease, schizophrenia, bipolar disorder, enhancers, non-coding variants, parvalbumin interneurons, glutamatergic neurons, GABAergic neurons, epigenetic dysregulation
MeSH terms
Chromatin Immunoprecipitation Sequencing, Acetylation, Alzheimer Disease, Bipolar Disorder, Chromatin, Enhancer Elements, Genetic, Gene Expression Regulation, Histones, Humans, Lysine, Neuroglia, Neurons, Promoter Regions, Genetic, Schizophrenia, Epigenesis, Genetic, Genome-Wide Association Study, GABAergic Neurons

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 2026-10-10.

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

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