Activity in serotonergic axons in visuomotor areas of cortex is modulated by the recent history of visuomotor coupling
This article has been Reviewed by the following groups
Discuss this preprint
Start a discussion What are Sciety discussions?Listed in
- Evaluated articles (Peer Community in Neuroscience)
Abstract
Visuomotor experience is necessary for the development of normal function of visual cortex (Attinger et al., 2017) and likely establishes a balance between movement-related predictions and sensory signals (Jordan and Keller, 2020). This process depends at least in part on plasticity in visual cortex (Widmer et al., 2022). Key signals involved in driving this plasticity are visuomotor prediction errors (Keller et al., 2012; Keller and Mrsic-Flogel, 2018). Ideally however, the amount of plasticity induced by an error signal should be a function of several variables - including the total prediction error across all of cortex at that moment, the animal's experience in the current environment or task, stability of the current environment, and task engagement - for optimal computational performance. Candidates for regulators of visuomotor prediction error driven plasticity are the three major neuromodulatory systems that innervate visual cortex in the mouse: acetylcholine, noradrenaline, and serotonin. While visuomotor mismatch acutely triggers activity in noradrenaline (Jordan and Keller, 2023) but not acetylcholine (Yogesh and Keller, 2023) axons in visual cortex, how serotonergic axons in cortex respond to visuomotor mismatch is unknown. Here, we characterized the activity of serotonergic axons in visual cortex (V1) and in area A24b, a motor cortical area in anterior cingulate cortex (ACC), of awake head-fixed mice using two-photon calcium imaging. Our results reveal cortical region-specific responses to visuomotor stimuli in serotonergic axons, but no evidence of a response to visuomotor mismatch. However, average activity in serotonergic axons was modulated by the recent history of visuomotor coupling. We speculate that serotonin could function to regulate visuomotor plasticity as a function of the predictability of the environment with a slow integration time constant.
Article activity feed
-
-
The ability to adapt sensory representations based on motor context is central to predictive processing frameworks, where perception is shaped by comparisons between internal predictions and external input (Keller and Mrsic-Flogel, 2018). Several neuromodulatory systems, particularly noradrenaline and acetylcholine, have been implicated in modulating these computations, either by signalling surprise or contextual state. However, the role of serotonin in this context has remained largely unclear. In their new study, Yogesh and Keller (2025) explore this question by recording activity from serotonergic axons in two cortical areas, primary visual cortex (V1) and anterior cingulate cortex (A24b), during changes in visuomotor coupling.
The authors use axon-targeted GCaMP6s in dorsal raphe serotonin neurons combined with two-photon calcium …
The ability to adapt sensory representations based on motor context is central to predictive processing frameworks, where perception is shaped by comparisons between internal predictions and external input (Keller and Mrsic-Flogel, 2018). Several neuromodulatory systems, particularly noradrenaline and acetylcholine, have been implicated in modulating these computations, either by signalling surprise or contextual state. However, the role of serotonin in this context has remained largely unclear. In their new study, Yogesh and Keller (2025) explore this question by recording activity from serotonergic axons in two cortical areas, primary visual cortex (V1) and anterior cingulate cortex (A24b), during changes in visuomotor coupling.
The authors use axon-targeted GCaMP6s in dorsal raphe serotonin neurons combined with two-photon calcium imaging in head-fixed mice navigating a virtual environment. This design allows them to measure serotonergic axon responses to several manipulations: closed-loop coupling between locomotion and visual flow, open-loop conditions (visual flow replayed independently of movement), transient visuomotor mismatches (brief halts in flow), and variations in visuomotor gain.
The main result is that serotonergic axons in both V1 and A24b do not respond acutely to visuomotor mismatch events per se. Instead, their activity during subsequent quiescent periods is modulated by the preceding sensorimotor context: activity is higher following open-loop conditions and higher visuomotor gain. These effects are most robust in V1. In contrast to previous findings on noradrenaline — which responds in a phasic manner to mismatch events (Jordan and Keller, 2023) — and acetylcholine — which signals locomotion state (Yogesh and Keller, 2023) — serotonergic activity appears to reflect more slowly integrated features of recent experience.
The authors interpret this modulation as consistent with serotonin encoding aspects of environmental predictability or task context, rather than short-timescale prediction errors. This aligns with prior studies suggesting serotonergic involvement in regulating learning rates (Grossman et al., 2022), behavioural flexibility (Matias et al., 2017), or expected value (Harkin et al., 2025). Importantly, the paper presents this interpretation with appropriate caution, noting that their correlational data cannot distinguish between potential drivers such as uncertainty, arousal (inferred via pupil diameter), or prospective value.
I recommend this study because it addresses an underexplored question in cortical neuromodulation, specifically the role of serotonin in visuomotor contexts, using a targeted and appropriate experimental design. The findings are interpretable within broader theoretical frameworks, including predictive coding and neuromodulatory control of plasticity, and the authors are careful not to overextend their claims. Moreover, the study contributes comparative insights to an emerging model in which acetylcholine, noradrenaline, and serotonin play distinct and complementary roles in shaping cortical responses to sensorimotor signals, corresponding to internal state, surprise, and environmental predictability, respectively. While some open questions remain, such as whether the observed effects reflect environmental volatility, subjective confidence, or value estimation, the study lays important groundwork for future experiments that could address these possibilities using causal manipulations or receptor-specific approaches.
Overall, this is a well-executed and thoughtful contribution that adds clarity to the role of serotonergic input in cortical sensorimotor integration. It will be of interest to researchers studying neuromodulation, predictive processing, and the control of cortical plasticity.
References
Grossman CD, Bari BA, Cohen JY (2022) Serotonin neurons modulate learning rate through uncertainty. Current Biology 32:586-599.e7. https://doi.org/10.1016/j.cub.2021.12.006
Harkin EF, Grossman CD, Cohen JY, Béïque J-C, Naud R (2025) A prospective code for value in the serotonin system. Nature 641:952–959. https://doi.org/10.1038/s41586-025-08731-7
Jordan R, Keller GB (2023) The locus coeruleus broadcasts prediction errors across the cortex to promote sensorimotor plasticity. eLife 12:RP85111. https://doi.org/10.7554/eLife.85111.3
Keller GB, Mrsic-Flogel TD (2018) Predictive Processing: A Canonical Cortical Computation. Neuron 100:424–435. https://doi.org/10.1016/j.neuron.2018.10.003
Matias S, Lottem E, Dugué GP, Mainen ZF (2017) Activity patterns of serotonin neurons underlying cognitive flexibility. eLife 6:e20552. https://doi.org/10.7554/eLife.20552
Yogesh B, Keller GB (2025) Activity in serotonergic axons in visuomotor areas of cortex is modulated by the recent history of visuomotor coupling. bioRxiv, ver.2 peer-reviewed and recommended by PCI Neuroscience. https://doi.org/10.1101/2025.03.11.642559
-
