Home - Research Teams - The neural bases of sensorimotor learning

The neural bases of sensorimotor learning

The sensorimotor learning lab

Humans and other animals’ survival depends on their ability to learn new adaptive behaviors, to adjust them to changes in their environment or occurring in their own body, and, in some cases, to become extremely skilled at what they do. All these functions depend on the intricate interaction between sensory, motor, and “cognitive” processes. Our team tries to delineate the contribution(s) of the cortico-striatal system to adaptive behavior. We tackle this challenging question by combining a wide range of system-neuroscience techniques in rodents (rats and mice) and computational/theoretical approaches.

Understanding the function(s) of the cortico-striatal system is important because its dysfunction results in several prevalent brain diseases such as Parkinson’s disease, or hyperactivity disorders that, interestingly enough, are characterized by a mixture of sensory, motor, and cognitive deficits.

To learn more +

Research interests

Associative and procedural forms of learning require the detection and integration of sensory information and the generation, through trial-and-error, of adaptive motor sequences. Such processes rely on changes in neuronal activity in the primary sensory cortex and in motor regions such as the basal ganglia. The goal of the team is to understand the neuronal mechanisms at play during learning, with a focus on sensory processing, motor control, and the sensorimotor transformation. For this, we design and engineer new behavioral tasks and use in vivo electrophysiology (tetrodes, silicon probes, optrodes), manipulation of neuronal activity (close-loop optogenetic stimulation, brain lesions, pharmacological inactivation), ex vivo electrophysiology (patch-clamp, mapping with laser scanning photostimulation) and behavioral/neural statistical analysis to draw correlations and causal links between features of the animal behavior and the dynamics of neuronal population, single cells or identified circuits.

Our team is constantly looking for enthusiastic undergraduate students, Ph.D. students, postdoctoral fellows, and engineers. Applications should be sent to David Robbe, Ingrid Bureau or Elodie Fino

 

Specific projects

  • aFunctional changes within the cortical networks of sensory perception
  • bEncoding of motor learning within cortico-striatal networks
  • cRole of local inhibitory microcircuits in striatal dynamics
  • dRole of the motor cortex and basal ganglia in motor learning and control
  • eRole of striatal cell-types in motor control
  • fSensorimotor coupling between barrel cortex and striatum as a mechanism for action monitoring and modulation
a
a- In vivo, monitoring of whisker-evoked activity in columns of cortex. b- Ex vivo, LSPS circuit mapping.

Functional changes within the cortical networks of sensory perception

P.I.: Ingrid Bureau. With Simona Plutino.
The aim here is to understand how neural representation in the primary somatosensory cortex changes while an animal learns to associate a whisker stimulus with a rewarding motor response. We work at identifying the neuronal circuits underlying the transformation of the whisker cortical map and at finding neural correlates to whisker-triggered behaviors. To this aim, we combine behavioral studies and electrophysiological recordings in and ex vivo.

We also work on models of developmental diseases such as epilepsy to investigate the consequences of cellular alterations on functions of sensory circuits and on sensory-guided behaviors.

b
a) Two-photon calcium imaging of striatal neurons. b) Tracing studies

Encoding of motor learning within cortico-striatal networks

P.I.: Elodie Fino. With Nagham Badreddine and Gisela Zalcman.
We all said “It’s like riding a bike!” to convince someone that a task is really easy. Once we have learned how to ride a bike, we do it automatically and never forget. However, the neural mechanisms making it so simple remains extremely complex. The memory of motor skills, procedural memory, is a fundamental adaptive mechanism, which provides efficiency for common behaviors and set cognitive resources free for other tasks. But how does procedural memory build up? How do neuronal circuits transform the activities of individual neurons into such complex sensorimotor sequences?
The neural substrates underlying procedural memory involve the basal ganglia, and particularly their input stage, the striatum, key player for the integration and selection of functionally distinct cortical information. The aim of our project is to understand how the procedural memory is encoded within the striatum and which mechanisms within cortico-striatal networks allow the formation and the maintenance of the memory. For this purpose, we are using a combination of technics including behavior, multi-photon imaging, tracing studies, opto-/chemogenetics and electrophysiology.

c
PV and SOM interneurons expressing ChR2 and light-induced activity

Role of local inhibitory microcircuits in striatal dynamics

P.I.: Elodie Fino. With Gisela Zalcman.
Another level of complexity to understand the striatal dynamics arises from the heterogeneity of striatal networks. The striatum is mainly composed of GABAergic cells, with strong and diverse local inhibitory circuits formed by different subtypes of interneurons (parvalbumin, somatostatin, …). We are studying the organization of the different striatal GABAergic microcircuits and the role they play in shaping striatal population dynamics.

d

Role of the motor cortex and basal ganglia in motor learning and control

Multi-channel recording with silicon probe of neuronal activity in a behaving rat

P.I.: David Robbe. With Mostafa Safaie, Stephania Sarno and Masoud Aghamohamadian.

When observing animals in their natural environment, one is often fascinated by their capacity to execute fine-tuned behaviors adapted to a variety of complex and challenging situations. Even if the importance of innate capacity should not be undermined, adaptive behaviors are often acquired and improved through extensive practice and prolonged trial-and-error interaction with the environment. The biological algorithms and the neural implementations underlying this type of learning are largely unknown. We tackle these questions with an interdisciplinary approach combining behavior, electrophysiological and brain lesion experiments, data analysis and modeling (reinforcement learning theory).


Left, spike waveforms of 6 well-isolated units recorded in the striatum with a silicon probe (center) . Right, neuronal sequence in the striatum during motor sequence performance.

We have developed a new behavioral paradigm in which rats try to maximize reward collection while saving energy during a multi-trial time estimation task. Interestingly, during learning, all animals progressively converged toward a similar embodied strategy: they perform a stereotyped running sequence whose duration match the time to estimate. By combining behavioral testing (alteration of task rules and environments) and modeling we aimed at understanding this learning process at an algorithmic level. Using lesion and chronic extracellular electrophysiology (tetrode and silicon probes) we also examining how the basal ganglia and motor cortex contribute to the implementation of these embodied motor strategy.


e
Top, online video tracking of a rat running on a motorized treadmill. Bottom, high-throughput training unit for motor control study in mice

Role of striatal cell-types in motor control

P.I.: David Robbe. With Anass Tinakoua and Corane Karoutchi.

The striatum is the main entry point of the basal ganglia (BG) and there is a general agreement that the dorsal region of striatum (DS) is implicated in motor control. A long-standing hypothesis is that the DS contributes to the selection of actions via the concurrent activation of striatal neurons forming the so-called direct and indirect basal ganglia pathways. While this hypothesis is appealing its validity is far from being confirmed experimentally.

To address this question we developed a new locomotion-based task in which head-restrained mice are trained to start, maintain and stop running according to external cues to obtain rewards they can collect by licking. We combined electrophysiological and close-looped optogenetic-based perturbation to understand the contribution of striatal neurons forming the direct and indirect pathways in the control of locomotion and licking.

f
Labelling of cortical neurons projecting to striatum (from the Allen brain atlas)

Sensorimotor coupling between barrel cortex and striatum as a mechanism for action monitoring and modulation

P.I.: David Robbe and Ingrid Bureau. With Kenza Amroune

The primary somatosensory cortex projects heavily onto striatum. In this project we study the properties and the encoding activity of corticostriatal projections during a whisker-guided motor task.

External cool links

  • A post defining scientific attitude versus dogmatic: 

http://waitbutwhy.com/2015/11/the-cook-and-the-chef-musks-secret-sauce.html

  • A french excellent scientific radio show on the life of Marie-Curie.

http://www.franceinter.fr/emission-sur-les-epaules-de-darwin-a-la-decouverte-de-la-radioactivite

  • A great free tool for data analysis.  Help with transparency, collaboration and reproducibility :

http://jupyter.org/

https://try.jupyter.org/

 

Our publications

voir toutes les publications +