Thalamocortical transcriptional gates coordinate memory stabilization
Article meta
Article Date: 26 November 2025
Article URL: https://www.nature.com/articles/s41586-025-09774-6
Article Image: none supplied in the paper
Summary
This Nature study by Terceros, Chen et al. combines behaviour, optogenetics, single-cell transcriptomics (scRNA-seq), ATAC-seq, ChIP–seq and in vivo CRISPR to map how anteromedial thalamus (ANT) and anterior cingulate cortex (ACC) coordinate memory stabilisation. Using a high- vs low-repetition training paradigm in mice and detailed pseudotime analyses, the authors show that ANT and ACC engage distinct, time‑dependent transcriptional programmes: ANT recruits rapid synaptic‑plasticity gene programmes during early consolidation, while ACC engages chromatin and histone‑methylation modules that persist into remote memory. They identify and validate key transcriptional regulators (TRs) — including Camta1, Tcf4 and Ash1l among others — and show that perturbing these regulators (via in vitro screening and in vivo CRISPR knockouts) alters gene regulation, chromatin marks and behavioural memory stability. ATAC and ChIP data indicate sustained accessibility and histone modifications at regulator target modules, supporting a model in which thalamocortical transcriptional “gates” select and stabilise long‑term memories.
Key Points
- ANT and ACC show region‑specific transcriptional programmes during different stages of memory consolidation: ANT for early synaptic plasticity, ACC for later chromatin/histone‑methylation driven stabilisation.
- Pseudotime and single‑cell analyses reveal distinct macrostates associated with memory persistence and link gene expression changes to behavioural performance.
- Specific transcriptional regulators (e.g. Camta1, Tcf4, Ash1l, Creb1, Mef2c, Myt1l, Kmt2a) are implicated; targeted CRISPR knockouts alter target gene expression and memory behaviour.
- ATAC‑seq and ChIP‑seq show persistent accessibility and histone mark changes at TR target sites, indicating epigenetic mechanisms for long‑term memory stabilisation.
- Extensive data and methods (scRNA‑seq, ATAC, ChIP) are provided (GEO accessions listed) and analysis code will be released on the RajasethupathyLab GitHub.
Content summary
The authors trained mice in a virtual‑reality task with high‑repetition (HR) and low‑repetition (LR) conditions and collected tissue from ANT and ACC at training, recent, mid and remote retrieval time points. Large‑scale scRNA‑seq identified neuron subtypes and time‑dependent transcriptional responses; pseudotime trajectories captured early vs late consolidation branches. ANT HR neurons upregulated synaptic plasticity and immediate‑early genes during early/mid stages, while ACC HR neurons showed increasing enrichment for histone methylation and chromatin regulation modules at late‑remote stages. Perturbation experiments (in vitro sgRNA screen, in vivo CRISPR knockouts) validated candidate TRs; ChIP‑seq and ATAC‑seq demonstrated mechanistic links between TR binding, chromatin state and transcriptional outcomes. Behavioural deficits after TR knockouts support causal roles for these factors in memory stabilisation.
Context and relevance
This work extends the memory stabilisation narrative beyond hippocampus to a thalamocortical circuit mechanism. It ties transient synaptic programmes to longer‑lasting epigenetic programmes across regions, suggesting a coordinated gating system where thalamic transcriptional activity helps select memories that are then cemented in cortex through chromatin modifications. The findings sit squarely within current trends emphasising circuit‑level transcriptional control, single‑cell resolution of memory engrams and epigenetic contributions to long‑term memory and psychiatric disease risk.
Author’s note (punchy)
Big picture: this paper maps how thalamus and cortex split the work of making memories stick — fast synaptic switches in ANT, slow chromatin locking in ACC — and nails down specific regulators you can actually perturb. If you work on memory, circuits or epigenetics, this should be on your short reading list.
Why should I read this?
Because it’s a neat, well‑controlled study that actually connects behaviour to single‑cell transcription, chromatin biology and causal manipulation. If you care about how short‑term changes become long‑term memory (or how to break that process in disease), these folks did the heavy lifting so you don’t have to skim dozens of separate papers — they bundled circuits, regulators and mechanisms into one story.