Periodic Reporting for period 1 - CereCode (Neuronal computations and population dynamics in the Cerebellar Nuclei during motor behaviours)
Reporting period: 2022-10-01 to 2025-03-31
A precise control of movements is crucial for survival, from simple reflexes to complex coordinated actions. While many brain regions contribute to motor control, the cerebellum plays a particularly critical role by integrating both higher brain signals and direct sensory inputs from the body. However, a major gap exists in our understanding: while the cerebellar cortex has been extensively studied, we know far less about its output stage - the cerebellar nuclei (CN).
These small nuclei form a crucial computational bottleneck where all cerebellar processing converges before influencing other brain regions. The CN can simultaneously modulate multiple aspects of movement: adjusting reflexes, controlling voluntary actions through brainstem centers, and participating in movement planning via connections with the motor cortex. Despite their importance, technical limitations have long impeded detailed study of these deep brain structures.
The CereCode project aims to overcome these limitations using state-of-the-art technologies. We combine innovative viral strategies for precise circuit mapping, advanced imaging techniques for recording deep brain activity, and newly developed behavioral systems. Our research addresses three fundamental questions: 1) How are CN circuits organized and how do they integrate diverse inputs? 2) How do CN neuron populations represent different motor behaviors? 3) What specific roles do different input pathways play during movement?
This work will provide crucial insights into both normal cerebellar function and potential therapeutic targets for movement disorders. Beyond advancing our understanding of motor control, the project will develop new tools for studying deep brain structures, benefiting the broader neuroscience community and potentially improving our ability to treat cerebellar disorders affecting movement coordination.
These small nuclei form a crucial computational bottleneck where all cerebellar processing converges before influencing other brain regions. The CN can simultaneously modulate multiple aspects of movement: adjusting reflexes, controlling voluntary actions through brainstem centers, and participating in movement planning via connections with the motor cortex. Despite their importance, technical limitations have long impeded detailed study of these deep brain structures.
The CereCode project aims to overcome these limitations using state-of-the-art technologies. We combine innovative viral strategies for precise circuit mapping, advanced imaging techniques for recording deep brain activity, and newly developed behavioral systems. Our research addresses three fundamental questions: 1) How are CN circuits organized and how do they integrate diverse inputs? 2) How do CN neuron populations represent different motor behaviors? 3) What specific roles do different input pathways play during movement?
This work will provide crucial insights into both normal cerebellar function and potential therapeutic targets for movement disorders. Beyond advancing our understanding of motor control, the project will develop new tools for studying deep brain structures, benefiting the broader neuroscience community and potentially improving our ability to treat cerebellar disorders affecting movement coordination.
During this initial period, the project has made substantial progress in understanding cerebellar nuclear circuits, while simultaneously developing key technological tools for the rest of the project. The work advanced along several complementary directions:
First, we successfully validated anatomical tracing strategies. We established reliable protocols for visualizing inputs from different brain regions to the cerebellar nuclei, particularly from the red nucleus and pontine nuclei. This anatomical work defined the input streams that we now focus on to study integration in the cerebellar nuclei.
After optimizing our preparation methods for studying adult mouse cerebellar nuclear tissue, traditionally a significant technical challenge in the field, we observed through electrophysiology recordings and imaging approaches previously unreported diversity in the properties of synaptic inputs to cerebellar nuclear neurons. This suggests a more sophisticated computational organization than previously thought. We are now investigating the anatomical origin of these different inputs.
On the technological front, we developed an advanced motorized wheel system for behavioral experiments. This device incorporates several innovative features: dynamic control of running resistance to investigate motor control under loads, a system for presenting varying textures during locomotion to see how the motor system uses the sensory context, and a novel mechanism for introducing obstacles to see how how the system learns and adapt to perturbations. These features will enable precise studies of motor learning during locomotion in head-fixed animals.
Finally, we developed comprehensive analysis pipelines for behavioral and imaging data. We implemented 3D behavioral tracking capabilities and adapted our data management to implement standardized data formats, ensuring our results will be readily shareable with the broader neuroscience community.
First, we successfully validated anatomical tracing strategies. We established reliable protocols for visualizing inputs from different brain regions to the cerebellar nuclei, particularly from the red nucleus and pontine nuclei. This anatomical work defined the input streams that we now focus on to study integration in the cerebellar nuclei.
After optimizing our preparation methods for studying adult mouse cerebellar nuclear tissue, traditionally a significant technical challenge in the field, we observed through electrophysiology recordings and imaging approaches previously unreported diversity in the properties of synaptic inputs to cerebellar nuclear neurons. This suggests a more sophisticated computational organization than previously thought. We are now investigating the anatomical origin of these different inputs.
On the technological front, we developed an advanced motorized wheel system for behavioral experiments. This device incorporates several innovative features: dynamic control of running resistance to investigate motor control under loads, a system for presenting varying textures during locomotion to see how the motor system uses the sensory context, and a novel mechanism for introducing obstacles to see how how the system learns and adapt to perturbations. These features will enable precise studies of motor learning during locomotion in head-fixed animals.
Finally, we developed comprehensive analysis pipelines for behavioral and imaging data. We implemented 3D behavioral tracking capabilities and adapted our data management to implement standardized data formats, ensuring our results will be readily shareable with the broader neuroscience community.
The most notable advance in our project so far has been to identify highly heterogeneous synaptic properties in cerebellar nuclear inputs. While some diversity was expected, the extent of this heterogeneity and its functional role is now under investigation using optogenetic approaches. This should lead to a better understanding of the contribution of sensory information in motor coordination and motor learning.
The work so far has strongly focused on developing in vitro parts of the project. We implemented imaging protocols to visualize neurotransmitter release in cerebellar nuclei. While molecular sensors for neurotransmitters are widely used in other brain regions, their application in cerebellar nuclei presented unique challenges. We are now using these tools to investigate synaptic transmission using optical methods, enabling us to understand the properties of different information streams contacting the cerebellar nuclei.
While our imaging setup is being installed, we focused our work on developing technological solutions for the rest of the project. For example, we developed a motorized wheel system integrating dynamic friction control, programmable sensory textures, and mobile obstacles that creates unprecedented possibilities for studying sensorimotor integration. This system enables translation of experimental paradigms previously limited to other behaviours (like eyeblink conditionning) into more complex locomotor behaviors.
Finally, adhering to open science principles, we have developed solutions for sharing our complex datasets with the broader scientific community. Our work on standardizing data formats for our 3D acousto-optic lens imaging will use the best current practices for storing and sharing recordings. The adaptation of the Neurodata Without Borders format for arbitrary-trajectory imaging data, combined with our commitment to releasing hardware designs and analysis code through public repositories, provides a comprehensive template that will benefit researchers that will use our technology in the future, or that would like to use our dataset for computational neuroscience projects.
The work so far has strongly focused on developing in vitro parts of the project. We implemented imaging protocols to visualize neurotransmitter release in cerebellar nuclei. While molecular sensors for neurotransmitters are widely used in other brain regions, their application in cerebellar nuclei presented unique challenges. We are now using these tools to investigate synaptic transmission using optical methods, enabling us to understand the properties of different information streams contacting the cerebellar nuclei.
While our imaging setup is being installed, we focused our work on developing technological solutions for the rest of the project. For example, we developed a motorized wheel system integrating dynamic friction control, programmable sensory textures, and mobile obstacles that creates unprecedented possibilities for studying sensorimotor integration. This system enables translation of experimental paradigms previously limited to other behaviours (like eyeblink conditionning) into more complex locomotor behaviors.
Finally, adhering to open science principles, we have developed solutions for sharing our complex datasets with the broader scientific community. Our work on standardizing data formats for our 3D acousto-optic lens imaging will use the best current practices for storing and sharing recordings. The adaptation of the Neurodata Without Borders format for arbitrary-trajectory imaging data, combined with our commitment to releasing hardware designs and analysis code through public repositories, provides a comprehensive template that will benefit researchers that will use our technology in the future, or that would like to use our dataset for computational neuroscience projects.