The overall objective of the group is to unravel principles of neural computations underlying sensory-motor integration in thevertebrate brain. We use the zebrafish larva as a model since it currently constitutes the only vertebrate system amenable towhole-brain recording with cellular resolution. Using one- or two-photon light-sheet microscopy, we are able to monitor the long-termactivity of the quasi-entirety of the 100,000 neurons that comprise the animal brain, as it performs basic sensory-motor tasks. Wecombine these experimental developments with a strong theoretical and computational effort aiming at interpretating the largedatasets produced by whole-brain imaging approaches. Dimensionality reduction and neural circuit inference methods areimplemented in order to extract from these continuous recordings information regarding the neural circuits architecture. In the nearfuture, we wish to combine brain-wide recordings with spatially-resolved optogenetics techniques in order to directly interrogate theneural circuits and thus test neuronal integrator models.
We currently focus on four different topics :
Phototaxis: this behavior drives the animal to swim towards a light source. We study how the closed-loop coupling between visual stimuli, eyes saccades and tail beats allows the animal to perform such a task. We seek to further identify the neural integrator circuitthat subserves this process.
Rheotaxis: This behavior, shared by most fish and amphibians, drives the animal to swim against an oncoming water flow. It is drivenby both the visual and lateral line system. We studied, from a behavioral view-point how these two sensory inputs contribute to themotor response. We now wish to identify the associated integrator centers and to dissect the neural circuit controling thismulti-sensory process.
Gaze stabilization: This other multi-sensory reflex allows the animal to compensate for self-motion by eliciting compensatory eyemovements. It is driven by both the vestibular and the visual systems. We developed an original experimental platform which allowsone to indepently address both the sensory systems in a controled way during whole-brain recording. This device allows us to studythe cross-modal integration when the two stimuli cues are either consistent or in conflict.
Chemical sensing: The question here is to understand how odors and tastants are coded in the brain. As a first step, we developed amicrofluidic-based chip that enables delivering brief pulses of chemicals with 10ms time resolution, directly towards the face of apartially restrained larva. This device allows us to probe the effect of stimulus duration on the neuronal response along the sensorypathways.