Back

Stoichiometric transcription factor partnerships specify GABAergic neuron subtype identity

Dvoretskova, E.; Lynch, C.; Bright, A. R.; Babal, Y. K.; Isaev, S.; Kharchenko, P.; Adameyko, I.; Mayer, C.

2026-05-27 developmental biology
10.64898/2026.05.25.727662 bioRxiv
Show abstract

Neuronal diversity arises when common progenitors deploy distinct gene expression programs to generate specialized cell types. The developing ventral telencephalon, which gives rise to forebrain GABAergic neurons, has recently been identified as a convergence point for neurodevelopmental disease (NDD) risk genes, many of which encode transcription factors (TFs). Yet how these TFs are directed to promote one fate while restraining another remains poorly understood. By combining sparse in vivo CRISPR perturbation with heritable lineage barcoding in the embryonic mouse ganglionic eminence, we show that loss of the NDD-associated zinc-finger TF Sp9 shifts cell production away from D2 medium spiny neurons, and toward D1 medium spiny neurons and intercalated cells of the amygdala. Mechanistically, SP9 operates through two distinct modes: at distal regulatory elements bound by DLX homeobox TFs, SP9 is recruited via zinc-finger-mediated interaction with these factors and engages chromatin-modifying complexes to repress their activity; at a smaller set of GC-rich promoters, SP9 directly binds DNA at SP-family motifs to activate transcription. Reporter assays and single-cell analyses show that the relative levels of SP9 and DLX TFs at shared regulatory elements, combined with SP9 self-activation, determine the regulatory outcome. These findings establish stoichiometric TF partnerships as a quantitative principle of cell-fate specification, with implications for NDDs affecting the developing ganglionic eminence.

Matching journals

The top 5 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.