Dysregulated adult hippocampal neurogenesis in major depressive disorders, 2026, Peng et al

[N_Dina]

Spoiler: Authors:

Peng, Madeleine S.; Jiang, Jialin; Polizzi, Lucia; Shi, Tiancheng; Ramkumar, Rakshitha; Anosike, Victor O.; Guasoni, Giulia; Wamalwa, Alexandra M.; Mariani, Madeline B.; Sissoko, Cheick A.; Tartt, Alexandria N.; Fulmore, Camille; Rosoklija, Gorazd B.; Huang, Yung-yu; Arango, Victoria; McDonald, Shujuan T.; Bitoljanu, Natasha; Mann, J. John; Nguyen, Phi T.; Dwork, Andrew J.; Brown, Lewis M.; Hen, René; Galfalvy, Hanga; Dupont, Maura B.

Major depressive disorder (MDD) is associated with reduced hippocampal volume, altered connectivity and negative memory bias, suggesting disrupted hippocampal plasticity. Dysregulated adult hippocampal neurogenesis is a potential contributor, but its relevance in humans and role in MDD remain unclear. Here we investigated the molecular basis of hippocampal dysfunction in nonmedicated individuals with MDD by integrating analyses of neurogenic trajectories, cell-type- and subfield-specific gene expression, chromatin accessibility and protein expression. We identify a neurogenic lineage in the adult human hippocampal subgranular zone and provide evidence for a stalled neurogenic process in MDD, associated with transcriptional regulation, stress-related reprogramming and interferon signaling across developmental stages. Excitatory and inhibitory neurons show dysregulation of transcription factor networks affecting cell states. Cellular stress, excitatory–inhibitory imbalance, impaired synaptic plasticity, reduced metabolic capacity and immune activation, underlie impaired neurogenesis and reduced hippocampus circuit plasticity. Findings indicate genetic and epigenetic regulation of gene expression in MDD, and overlapping pathogenetic mechanisms with autoimmune, neurodevelopmental and neurodegenerative diseases. This work provides a new understanding of the pathogenesis of hippocampus-dependent cognitive symptoms in MDD and suggests potential therapeutic targets.

DOI: 10.1038/s41591-026-04571-8

 

[N_Dina]

Main​

MDD is a major global health burden presenting with mood, neurovegetative, and cognitive symptoms, including memory recall bias toward negative content, a cognitive marker of depression severity. Pattern separation is dependent on adult hippocampal neurogenesis (AHN) in rodents and was shown to become transiently deficient in patients undergoing focal hippocampus irradiation (likely depleting AHN) to treat benign cavernous meningioma. Despite increasing evidence that AHN might occur in human adult brain, its clinical relevance remains debated and a molecular atlas of hippocampus circuit cell functional states in MDD is missing.

To resolve neurogenic trajectories, hippocampal cell states and circuit function in MDD, we integrated single-nucleus transcriptomic and chromatin accessibility sequencing, spatial transcriptomics and regional proteomics in nonmedicated individuals with MDD and neurotypical nonpsychiatric controls (CTRLs; Fig. 1a, Extended Data Table 1 and Methods). This multimodal characterization of the largest human hippocampus dataset (Extended Data Table 2) identified an adult human hippocampal neurogenic lineage and evidence of stalled neurogenesis in MDD. The hippocampal trisynaptic memory circuit, involving dentate gyrus (DG) and cornu ammonis (CA) regions, showed disrupted excitatory–inhibitory balance, synaptic plasticity and neurotransmission driven by altered transcriptional and epigenetic regulation, while all cell types exhibited dysregulated long noncoding RNAs, cellular stress, impaired intracellular trafficking, serotonergic and glutamatergic dysfunction, and neuroinflammation.

Spoiler: Fig. 1:

Fig. 1: Multimodal profiling of hippocampus cell types and spatial clusters.
a, Schematic of study design. b–d, UMAP visualization plots of 495,037 human brain nuclei in 11 MDD and 19 CTRL samples, retained after sequencing quality control testing (Supplementary Table 1; out of the initial 14 MDD and 24 CTRL, with 1–4 technical replicates each; total 44 CTRL and 30 MDD samples; replicates were used as covariates in the analyses), using batch-, sample- and donor-corrected Harmony components. b, UMAP derived from single-nucleus RNA sequencing data, colored by cluster. c, UMAP derived from ATAC with sequencing data, colored by cluster. d, UMAP derived from integration of the two modalities (snMultiome) using weighted nearest neighbor (WNN) analysis, colored by cluster. e, Dot plot depicting the expression of marker genes for each cell type: PDFGRA and OLIG1 for oligodendrocyte progenitor cells (OPCs); PLP1 and MOBP for oligodendrocytes (Oligo); PECAM1 and CEMIP for vasculature cells (Vasc); APBB1IP and P2RY12 for immune cells (Immune); CFAP54 for ependymal cells (Epe); TTR and HTR2C for choroid plexus (CP); AQP4 and GFAP for astrocytes (Astro); RBFOX3 for neurons; CAMK2A and SV2B for excitatory neurons; PROX1 and CALB1 for GCs; EPHA5, NEUROD6 and FNDC1 for CA excitatory neurons (ExN1.CA1-4, ExN2.CA2); TSHZ2 for other excitatory neurons (ExN3–7); GAD1 and GAD2 for GABAergic InNs. For a detailed plot with the 31 cell clusters and specific marker genes see Extended Data Fig. 2a. f, Pearson’s correlation of natural log1p of average gene expression per nuclei and predicted gene activity based on ATAC data in each cell type. g, Anchor-based integration of spatial Visium v1 whole transcriptome and single-nuclei gene expression provides a probabilistic transfer of annotations from reference (single-nucleus) to query (spatial) database annotating the 55-μm-diameter spatial spots. Color shows probability score. h–j, Anchor-based integration of single-nucleus transcriptomic data (reference) with Xenium single-cell in-situ hybridization (query) showing predicted anatomical mapping for clusters of excitatory neurons (h), astrocytes (i) and InNs (j). Panel a created in BioRender; Ramkumar, R. https://BioRender.com/6qbckag(2026).