Autumn has a way of making evenings feel like the brain is closing tabs one by one. This paper suggests that during REM sleep, your cortex is not simply "on" or "off" anyway - it is more like a city where some neighborhoods dim the lights while others keep arguing in the kitchen.

REM Is Not One Big Vibe

Most people hear "REM sleep" and picture one thing: dreams, twitchy eyes, weird plots, and a brain that looks suspiciously awake on EEG. Fair enough. REM has a strong reputation. It is the theater kid of sleep stages.

But this new mouse study shows that REM is not a single blanket state spread evenly across cortex. Campelo and colleagues recorded the membrane potentials of layer 2/3 pyramidal neurons in living, sleeping mice using in vivo whole-cell patch clamp, which is about as intimate as neuroscience gets without the neurons filing a complaint. In the prefrontal cortex, those neurons became less depolarized during REM. In the retrosplenial cortex, they became more depolarized. Same sleep stage. Different cortical zip code. Very different electrical mood. [1]

That matters because subthreshold depolarization is the electrical setup before a neuron fires. Shift that baseline and you change what information a circuit is ready to process, amplify, or ignore.

The Cortex Has Opinions

The prefrontal cortex is the part people love to cast as the adult in the room - planning, organizing, pretending to enjoy spreadsheets. The retrosplenial cortex is less famous outside neuroscience, but it helps with navigation, context, and stitching experiences into something coherent. In this paper, REM sleep pushed these regions in opposite directions. [1]

That is a useful correction to the old cartoon version of sleep science, where cortex just rides along while deeper structures do the switching. A 2023 study in Nature Neuroscience showed that excitatory neurons in the medial prefrontal cortex can actively promote REM sleep and shape its phasic features. So cortex is not just a passenger. At times, it is grabbing the wheel. [2]

This paper goes one level closer to the cell. Not just which region is active, but what the membrane potential is doing inside identified pyramidal neurons while the animal moves from wake to NREM to REM.

Tiny Electrical Nudges, Big Theories

The authors also probed mechanism. Blocking NMDA receptors selectively reduced REM-related subthreshold depolarization, pointing to glutamatergic currents as part of the reason these neurons sit where they do electrically during REM. Diazepam also reduced depolarization in prefrontal pyramidal neurons, likely through changes in inhibitory circuitry involving parvalbumin interneurons. [1]

Translation: REM is not just the brain passively replaying old material. It is actively tuning excitability through a mix of excitation and inhibition, and the balance depends on region.

That idea fits nicely with earlier work from the same group. In 2022, Aime and colleagues reported that REM sleep supports cortical plasticity through a kind of somatodendritic decoupling - dendrites and soma stop acting like one perfectly synchronized unit, which helps shape later learning and emotional discrimination. [3] The new study adds a finer electrical explanation for how REM may create those plasticity-friendly conditions in some regions while doing something else in others.

Other recent work points the same way: sleep states contain local, layered structure, with different networks doing different jobs at the same time. [4,5] Which sounds exactly like your group chat at 1:12 a.m.

Why You Should Care, Even If You Are Currently Holding a Drink

If these findings hold up and extend to humans, they could help explain why REM sleep is tied to memory, emotional recalibration, and psychiatric vulnerability without forcing all of that work onto one generic button.

A region-specific REM cortex suggests a more realistic model. Some circuits may be easing off the gas while others stay primed. Some may be protecting stored information. Others may be revising it.

It also raises practical questions about drugs that alter sleep. If diazepam changes subthreshold dynamics in prefrontal cortex during sleep, then sedating someone is not the same thing as preserving the natural computational benefits of sleep. Knocked out and well-slept are not synonyms.

The obvious caveat is that this is a mouse study, in head-restrained animals, with a very specific recording approach. It does not mean human dreams come from one cortex area dimming while another hums like a refrigerator. But it does mean REM deserves to be treated less like a monolithic state and more like a coordinated argument among circuits.

That is a better story. Also a weirder one. Which, for the brain, is usually how you know you are getting warmer.

References

1. Campelo T, Aime M, Rusterholz T, Gutierrez Herrera C, Adamantidis A. REM sleep reduces subthreshold depolarization in cortical pyramidal neurons in a region-specific manner. Cell Reports. 2025;44(11):116506. DOI: https://doi.org/10.1016/j.celrep.2025.116506. PubMed: https://pubmed.ncbi.nlm.nih.gov/41201914/
2. Hong J, Lozano DE, Beier KT, Chung S, Weber F. Prefrontal cortical regulation of REM sleep. Nature Neuroscience. 2023;26(10):1820-1832. DOI: https://doi.org/10.1038/s41593-023-01398-1. PubMed: https://pubmed.ncbi.nlm.nih.gov/37735498/
3. Aime M, Calcini N, Borsa M, Campelo T, Rusterholz T, Sattin A, Fellin T, Adamantidis A. Paradoxical somatodendritic decoupling supports cortical plasticity during REM sleep. Science. 2022;376(6594):724-730. DOI: https://doi.org/10.1126/science.abk2734. PubMed: https://pubmed.ncbi.nlm.nih.gov/35549430/
4. Boscher F, Jumel K, Dvořáková T, Gentet LJ, Urbain N. Thalamocortical dynamics during rapid eye movement sleep in the mouse somatosensory pathway. Journal of Neuroscience. 2024;44(25):e0158242024. DOI: https://doi.org/10.1523/JNEUROSCI.0158-24.2024. PMCID: https://pmc.ncbi.nlm.nih.gov/articles/PMC11209666/
5. Tononi G, Boly M, Cirelli C. Consciousness and sleep. Neuron. 2024;112(10):1568-1594. DOI: https://doi.org/10.1016/j.neuron.2024.04.011. PMCID: https://pmc.ncbi.nlm.nih.gov/articles/PMC11105109/

Disclaimer: The image accompanying this article is for illustrative purposes only and does not depict actual experimental results, data, or biological mechanisms.