Neural networks representing temporal expectation in mice

T. Adrian Wendlandt,Patricia Wenk,Julia U. Henschke,Annika Michalek,Toemme Noesselt,Janelle M.P. Pakan,Eike Budinger

DOI: https://doi.org/10.1101/2024.04.25.589683

2024-04-26

Abstract:The ability to attend to specific moments in time is crucial for survival across species facilitating perception and motor performance by leveraging prior temporal knowledge for predictive processing. Despite its importance, the neural mechanisms underlying the utilization of macro-scale and meso-scale neural resources during temporal processing and their relationship to behavioural strategies and motor responses remain largely unexplored. To investigate the capacity for predictive temporal structure of multisensory stimuli to optimize motor behaviour, we established a behavioural paradigm, in which mice were trained to an auditory-cue and visual-target presented at expected or unexpected temporal delays. Using a combination of stimulus-evoked and resting-state functional magnetic resonance imaging, we examined task-related evoked activity in brain-wide networks and found that that the formation of temporal expectations relying on accumulated sensory information and combined multisensory input involves plasticity across large macro-scale cortical networks comprised of primary sensory systems, sensory association areas including posterior parietal cortex, retrosplenial cortex, prefrontal top-down executive control centres of the brain, as well as hippocampal networks. Additionally, employing in vivo two-photon calcium imaging, we explored local single-cell dynamics within the posterior parietal cortex during this task and found that temporal expectation could be decoded directly from neuronal activity within this brain region. Overall, our study provides insights into the neural correlates underlying the formation of multisensory temporal expectations in the mouse brain and highlights the recruitment of neural resources across temporally-driven statistical learning processes.

Neuroscience

What problem does this paper attempt to address?

This paper aims to explore how the temporal structure of multi - sensory stimulation optimizes the motor behavior of mice. Specifically, it attempts to address the following core issues:

1. **Formation mechanism of temporal expectation**: Researchers hope to understand through experimental paradigms how mice form expectations for specific time points through statistical learning. This expectation is based on the integration of accumulated sensory information and multi - sensory inputs, involving extensive functional networks in the brain, especially the primary sensory systems, sensory association areas (such as the posterior parietal cortex, posterior cingulate cortex), the prefrontal top - down executive control center, and plastic changes in the hippocampal network.

2. **Relationship between temporal expectation and behavioral strategies**: Researchers also hope to explore the relationship between different behavioral strategies and brain functional networks during the process of temporal processing. Specifically, they are concerned with how the use of these macro - scale neural resources is associated with specific behavioral strategies to support the formation of temporal expectation, thereby optimizing motor responses.

3. **Neural basis of temporal expectation**: By combining functional magnetic resonance imaging (fMRI) and in - vivo two - photon calcium imaging techniques, researchers aim to reveal the neural correlates of multi - sensory temporal expectation formation, especially the local single - cell dynamics in the posterior parietal cortex (PPC). They found that temporal expectation can be directly decoded from the neural activity patterns in the posterior parietal cortex of mice, indicating that this brain region plays an important role in planning movement, attention, and sensory association.

4. **Differences in neural representations of different behavioral strategies**: Researchers explored the differences in brain activation patterns before and after training in mice with two behavioral strategies, target - attending and target - invariant. The results showed that mice with the target - attending strategy exhibited less demand for neural resources in multiple brain regions, while mice with the target - invariant strategy required more neural resources to process the temporal structure.

In conclusion, through a series of experimental designs and technical means, this paper deeply explores the formation mechanism of multi - sensory temporal expectation and its role in optimizing the motor behavior of mice, and reveals the neural basis behind different behavioral strategies.