Back

Identification of a neural circuit that enables safe, long-term torpor in mice

Guo, F.; Tong, K.; Yang, J.; Yuan, F.; Tang, M.; Gan, Y.; Wu, S.; Tong, X.; Luo, P.; Chen, S.; Ji, H.

2026-04-23 neuroscience
10.64898/2026.04.22.719582 bioRxiv
Show abstract

Torpor, a state of regulated hypothermia and hypometabolism, is a critical survival strategy in certain mammals. Previous studies established the primacy of neuron control in initiating acute torpor. However, whether neuromodulation can safely sustain torpor over the long term, and the specific governing pathways, remain unanswered challenges limiting its translational potential. Here we identify a distinct preoptic neuronal population expressing General control nonderepressible 2 ( G neurons) in mice that is essential and sufficient for torpor induction. Strikingly, sustained and selective activation of G neurons induces a stable, weeks-long torpid state (G neuron-driven long-term torpor, GLT), from which animals arouse without obvious behavioral deficits or tissue pathology. In contrast, long-term torpor driven by pan-neuronal activation of the same brain region (PLT) causes multiple post-recovery damages, underscoring the unique safety profile of selective pathway of G neurons. In a mouse cancer model, GLT directly suppressed tumor proliferation and markedly sensitized tumors to chemotherapy, achieving profound therapeutic outcomes. Together, we delineate a distinct neuronal population safely enabling prolonged torpor. This establishes a specific paradigm for inducing a stable and sustained hypometabolic state, providing a transformative platform for fundamental physiological research and pioneering clinical strategies against chronic pathologies such as cancer.

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.