henhaobio
Sign in

Scientific ReportsNeurosciencePeer-reviewed journal article

Multi-omics analysis reveals chromatin and transcriptomic remodeling in hippocampal CA1 following adolescent social isolation

Azusa KubotaiD, Yoshiaki ShinoharaiD, Akihiro GotoiD, Atsushi TajimaiD

PDFDOIPreprintData

In-depth summary

Claude Opus 5.5 · xhigh effort · 7 Nov 2026 · from the full text of version 2 (10 Sept 2026) · CC BY 4.0

Two weeks of post-weaning isolation in male mice altered expression of 106 CA1 genes, enriched for myelination, while chromatin accessibility shifted only at nominal significance and tracked expression weakly.

Earlier rodent work tied social isolation to histone-linked gene silencing in prefrontal cortex and midbrain and to synaptic and astrocytic changes in CA1, but how post-weaning isolation reshapes CA1 chromatin accessibility, and how that relates to transcription, was poorly characterized. This study pairs ATAC-seq with RNA-seq from the same CA1 nuclear preparations and uses a public hippocampal scRNA-seq reference to place the expression changes in cellular context.

Key results

  1. ShownIsolation from P24 to P38 changed CA1 expression of 106 genes. Fosl2, Hdac9 and myelin genes went up; Fbxw7, Tanc1 and circadian genes went down. GSEA favoured myelination and sterol synthesis.

    edgeR quasi-likelihood test, 12 libraries (n = 6 per group, left and right CA1 of 3 mice each): 64 up, 42 down at FDR < 0.05 (myelin: Ugt8a, Opalin; circadian: Per3, Ciart, Dbp, Nr1d2). GSEA: 91 GO-BP, 48 GO-CC, 36 GO-MF terms at FDR < 0.05, positive for axon ensheathment, myelin assembly, sterol biosynthesis. RT-qPCR on the same RNA matched direction for 4 DEGs. · Transcriptomic signatures in the CA1 following…; RNA-seq data processing and analysis; RT-qPCR; Figure legends

  2. SuggestedChromatin accessibility changes were weak. No peak passed FDR < 0.05. The nominal candidate regions split into gains biased toward enhancers (CTCF motif) and losses biased toward promoters (AP-1 and Egr1 motifs).

    n = 4 per group after QC; at nominal p < 0.01, 158 UP and 200 DOWN candidate DARs. UP: 42.4% distal intergenic, 29.1% intronic, 23.4% promoter; DOWN: 41.5% promoter. HOMER de novo motifs: CTCF in UP (-log10P = 47.77), AP-1 (36.67) and Egr1 (32.98) in DOWN. GREAT: no FDR-significant terms; nominal calcium transport (UP enhancer peaks) and synaptic organization (DOWN promoter peaks). · Alterations of chromatin accessibility in the…; Biological characterization of candidate regions…; ATAC-seq data processing and analysis

  3. SuggestedAccessibility and expression changes barely correlate genome-wide, but most DEGs pair with a nearby peak that changes in the same direction. Fbxw7 lies near an AP-1-motif region that loses accessibility.

    Nearest-TSS peak-gene log2FC: Spearman ρ = 0.048. Genes with decreased candidate DARs within 1 kb of the TSS showed a slight, non-significant shift (Wilcoxon p = 0.179). Pairing each of 101 DEGs with its smallest-p peak within 100 kb: 75 of 101 (74.26%) concordant, 26 (25.74%) discordant. Fbxw7: DOWN candidate DAR with AP-1 motif (ATTAGTCA) at -68 kb. · Associations between chromatin dynamics and…; Multi-omics integration of chromatin…; Figure legends

  4. SuggestedOn a public hippocampal scRNA-seq reference, genes upregulated by isolation score highest in oligodendrocytes and astrocytes. Downregulated genes score highest in excitatory neurons (and astrocytes).

    Seurat module scores across five annotated lineages. In the global subcluster ranking, the top UP-signature subclusters were mainly oligodendrocyte, astrocyte and microglial, while all top 15 DOWN-signature subclusters were excitatory neuronal. Myelin genes (Mag, Ugt8a, Fa2h, Opalin) led UP DEGs in oligodendrocyte subclusters; Hdac9 had high baseline expression in microglia and interneurons. · Potential cell type-specific signatures of…; Estimation of cell type-associated…

  5. SuggestedMuSiC deconvolution estimated nominally more oligodendrocyte and less excitatory-neuron signal in isolated CA1, and no change for interneurons, astrocytes or microglia. Neither difference survived FDR correction.

    n = 6 CA1 samples per group; reference built from the top 50 markers per lineage in the Allen hippocampal scRNA-seq data. Wilcoxon rank-sum p < 0.05 for oligodendrocytes (higher) and excitatory neurons (lower), not significant after FDR adjustment. The reference oligodendrocyte cluster expressed mature markers (Mbp, Mog), with some precursor (Pdgfra, Cspg4) and premyelinating (Enpp6, Itpr2) marker expression. · Estimation of SI-associated differences in…; Cell type deconvolution analysis; Figure legends

Methods, data and reproducibility

Male C57BL/6J mice were housed in groups of four or alone from P24 to P38; left and right CA1 per mouse were profiled separately by ATAC-seq and RNA-seq, analysed with hemisphere-adjusted edgeR quasi-likelihood tests. Peaks were linked to genes by TSS proximity; bulk signatures were mapped onto an Allen Brain Atlas hippocampal scRNA-seq reference with Seurat module scores and MuSiC deconvolution.

  • Hippocampal CA1 ATAC-seq, isolation-housed vs group-housed male C57BL/6J mice (P24 to P38) · mouse · n = n = 4 per group after QC (6 libraries per group sequenced, 3 mice per group) · GSE333471
  • Hippocampal CA1 poly(A) RNA-seq from the same nuclear isolations · mouse · n = 12 libraries, n = 6 per group (left and right CA1 of 3 mice per group) · GSE333472
  • Allen Brain Atlas whole cortex and hippocampus 10x scRNA-seq reference (hippocampal cells extracted, five lineages annotated) · mouse (Allen Brain Atlas mouse whole cortex and hippocampus dataset)
Methods & tools
Bowtie2, MACS3, ataqv, edgeR, ChIPseeker, HOMER, rGREAT, Salmon, clusterProfiler, Seurat, MuSiC, Click-qPCR
Reagents & environment
R 4.4.2 with tool versions listed (e.g. fastp 0.23.4, Bowtie2 2.5.4, MACS3 3.0.3, edgeR 4.4.2, Salmon 1.10.3, HOMER 4.10, Seurat 5.5.0, MuSiC 1.0.0); mm10 genome with GENCODE vM25. No container or requirements file is mentioned.
Reproducibility gaps
No code or analysis scripts are shared; result tables are in supplementary Data S1 to S5. Preprocessing, downsampling and cell-type annotation of the reference scRNA-seq data are described only briefly.
Limitations (6)
  • No ATAC-seq peak reached FDR < 0.05. Candidate DARs rest on a nominal p < 0.01 cut-off and have not been experimentally validated. The authors treat them as preliminary until higher-powered studies test them. · stated by the authors
  • Chromatin data are bulk, and cell-type links come from a reference scRNA-seq atlas. So same-cell coordination and causality are not established, nor is whether Hdac9 rises specifically in interneurons or microglia. · stated by the authors
  • Deconvolution gives only relative composition estimates and cannot show cell loss or new cell generation. The oligodendrocyte and excitatory-neuron differences were nominal and did not survive FDR correction. · stated by the authors
  • The study covers one time point (P38) and male mice only, with no behavioural, body-weight or corticosterone measurements in these animals. Links to phenotype, timing and sex therefore remain open. · stated by the authors
  • There are three mice per group: left and right CA1 enter as separate samples (n = 6), and the edgeR design adjusts for hemisphere but not for animal. RT-qPCR reused the sequenced RNA rather than an independent cohort. · noted in this summary
  • The 74.26% directional concordance between DEGs and nearby peaks is not compared with a null (e.g. randomly paired peaks). Each DEG's peak is the lowest-p peak within 100 kb, so the figure is hard to interpret. · noted in this summary

Read this if you study how adolescent social experience shapes hippocampal gene regulation and want paired CA1 ATAC-seq and RNA-seq data (GSE333471, GSE333472) with candidate AP-1 and oligodendrocyte-linked signatures to follow up.

Citations
0
FWCI
0.00
Field percentile
31th percentile

Abstract

Social isolation (SI) during adolescence is associated with neurobiological alterations; however, its effect on the hippocampal epigenome and transcriptome remains unclear. We performed ATAC-seq and RNA-seq of hippocampal CA1 in an adolescent mouse SI model, with single-cell RNA-seq reference mapping and cell type deconvolution. ATAC-seq identified candidate SI-associated chromatin accessibility alterations, with nominal functional enrichment linking increased enhancer-associated regions to calcium transport and decreased promoter regions to synaptic organization. Motif analysis showed enrichment of the activator protein 1 (AP-1) motif among candidate regions with reduced accessibility. RNA-seq identified 106 differentially expressed genes (DEGs), including upregulated Fosl2 and Hdac9 and downregulated Fbxw7, with enrichment of myelination-related pathways. Although global concordance between chromatin accessibility and transcriptional changes was limited, integrated analysis showed directional concordance between expression and nearby candidate accessibility changes for a subset of DEGs. Cell type analyses suggested that SI-downregulated transcriptional signatures were linked to excitatory neurons, whereas SI-upregulated signatures were enriched in non-neuronal populations, particularly oligodendrocytes. Deconvolution suggested higher estimated oligodendrocyte representation and lower estimated excitatory neuron representation in isolated mice. These findings suggest that adolescent SI is associated with significant transcriptomic remodeling and candidate chromatin accessibility changes within hippocampal CA1, accompanied by distinct neuronal- and glial-associated transcriptional signatures.

Reproducibility ledger

Nobody has reported trying to reproduce this paper yet. Did you replicate an experiment or run the code? Your report saves the next lab a week.

Sign in to file a reproducibility report.

Related papers

Similarity Map
SelectedSame fieldCross-field
THISBilgic '25You '26McDiarmid '26Achiro '24Traunmüller '25Laighneach '23
65% semantic matchNeuroscience
Merve Bilgic, Rinka Obata, Vlada‐Iuliana Panfil+4 more·Aging Cell·Sep 28, 2025· 4 citations