Do MCH neurons during REM sleep drive active forgetting of memories?

Evidence suggests MCH neurons active during REM sleep can weaken hippocampus-dependent memories, but the picture is more about modulation than simple erasure.

Direct answer

Yes, but with an important qualifier: the evidence supports the idea that melanin-concentrating hormone (MCH) neurons active during REM sleep can promote forgetting of hippocampus-dependent memories, and this appears to be an active process rather than passive decay [1][2][5]. In mice, removing these neurons improved memory performance, while artificially activating them impaired memory, and their fibers suppressed activity in hippocampal cells [5]. However, other REM-active hypothalamic neurons support memory consolidation, so REM sleep is not uniformly a forgetting state [3].

5sources cited

This article was generated with WisPaper-powered search and paper analysis.

What MCH neurons do during REM sleep

MCH neurons sit in the lateral hypothalamic area, a brain region involved in sleep, wakefulness, and metabolism. They are especially active during REM sleep, and they help the brain transition from non-REM sleep into REM sleep [2]. Their axons reach the hippocampus, a structure central to forming and recalling facts and events, which is why researchers have asked whether they influence memory there [1][5].

The proposed function is not that MCH neurons erase memories indiscriminately. Instead, the evidence points to an active forgetting mechanism: when these neurons fire during REM sleep, they can weaken hippocampus-dependent memories, possibly as part of normal memory regulation [1][2]. The word 'active' matters because it implies a biological process that does something to memories, rather than memories simply fading on their own.

The mouse evidence: removing MCH neurons improved memory, activating them impaired it

The most direct evidence comes from experiments in mice. Genetically ablating MCH neurons increased memory performance, while pharmacogenetic activation of these neurons impaired memory [5]. In brain slices, activating MCH fibers in the hippocampus suppressed the spiking activity of pyramidal cells, the main output neurons of that region [5]. That suppression offers a plausible cellular route by which MCH signaling could weaken memory traces.

A 2025 review places this within a broader framework: MCH neurons regulate active forgetting via hippocampal neural circuits, and they do so alongside orexin neurons, which promote wakefulness and memory encoding [2]. The review describes the two systems as antagonistic but coordinated, with MCH promoting sleep and REM sleep while also influencing memory filtering through hippocampal theta rhythms [2]. This is a review-level synthesis, so it should be read as a proposed framework rather than a single experimental result.

REM sleep is not only for forgetting

A key complication is that other REM-active hypothalamic neurons do the opposite. In mice, neurons from the supramammillary nucleus that project to hippocampal area CA2 were highly active during REM sleep but not during non-REM sleep or quiet wakefulness. Silencing these neurons selectively during REM sleep impaired social memory, while silencing a different group of REM-active neurons projecting to the dentate gyrus had no effect on social memory [3]. That is causal evidence that some REM-active hypothalamic neurons are required for consolidating freshly encoded social memory.

So the answer is not that REM sleep is a global forgetting state. Different REM-active cell populations can push memory in different directions: MCH neurons appear to promote forgetting of hippocampus-dependent memories [1][5], while supramammillary neurons projecting to CA2 support social-memory consolidation [3]. The balance between these systems may determine what is kept and what is let go.

How firm is the conclusion?

The core claim rests mainly on mouse experiments using ablation and activation of MCH neurons, plus slice recordings showing hippocampal suppression [5]. That is strong causal evidence in animals, but it does not establish the same mechanism in humans. The 2025 review extends the framework to clinical possibilities, such as MCH receptor antagonists for memory reconsolidation in PTSD, but those are proposed therapeutic directions rather than demonstrated treatments [2].

There is also a technical caveat: one supplied abstract contains a fragment suggesting that MCH neurons in REM 'can not contribute to forgetting,' which appears to conflict with the paper's own verification summary that they do contribute to active forgetting [1]. Because only the abstract is available, the exact experimental conditions behind that fragment cannot be checked here. The weight of the available evidence still supports a role for MCH neurons in active forgetting, but the precise circumstances and boundaries of that role remain an active research question.

Sources used in this answer

1

REM sleep–active MCH neurons are involved in forgetting hippocampus-dependent memories

MCH neurons active during REM sleep contribute to active forgetting of hippocampus-dependent memories.

2

The role of orexin and MCH neurons in the hypothalamus in sleep-wake regulation and learning-forgetting balance

MCH neurons facilitate the NREM-to-REM transition and regulate active forgetting via hippocampal circuits, in coordinated antagonism with orexin neurons.

3

REM sleep-active hypothalamic neurons may contribute to hippocampal social-memory consolidation

REM-sleep-active supramammillary neurons projecting to hippocampal CA2 are required for consolidating social memory in mice.

4

Orexin/hypocretin receptor 2 signaling in MCH neurons regulates REM sleep and insulin sensitivity

Orexin receptor 2 signaling in MCH neurons regulates NREM-to-REM transitions and insulin sensitivity in female mice, without addressing memory.

5

A brain pathway for active forgetting

Ablating MCH neurons improved memory, activating them impaired memory, and activating MCH fibers suppressed hippocampal pyramidal cell spiking.