Back

Single-cell multi-omics, spatial transcriptomics and systematic perturbation decode circuitry of neural crest fate decisions

Hu, Z.; Mayes, S.; Wang, W.; Santos-Pereira, J. M.; Theis, F.; Sauka-Spengler, T.

2024-09-17 developmental biology
10.1101/2024.09.17.613303 bioRxiv
Show abstract

The cranial neural crest (NC) is a migratory embryonic population ideal for studying cell plasticity, motility, and fate establishment. Although NC migration has been linked to changes in cell adhesion, polarity, and signaling, the gene regulatory circuitry governing these processes remained obscure. Using time-resolved single-cell multi-omics, spatial transcriptomics, and gene regulatory network reconstruction, we identified ten programs underlying 23 NC cell states and three spatial trajectories. Using in silico perturbation and systematic CRISPR/Cas9-mediated Perturb-seq, we uncovered novel lineage drivers and an endothelial-like program controlling NC migration, distinct from the epithelial-to-mesenchymal transition (EMT) program. We show that endothelial-like regulons (fli1a, elk3) drive migration through direct or tiered activation via the "FOX:ETS-Ebf3a-targets" axis, while ETS suppressors (erf, erfl3) maintain cell plasticity. Using the newly developed SyncReg tool, we identify functional redundancy among ETS regulons, which has thus far obscured their critical roles in NC migration, and we quantify their synergy with retinoic acid receptors, also essential for this process. Our GRN model, combined with novel velocity-embedded simulations, accurately predicted the functions of all major regulons, which were confirmed by in vivo functional perturbations. This study provides a comprehensive, validated cranial NC regulatory landscape, resolving heterogeneous regulatory circuits underlying NC cell motility and plasticity.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.