A lexical approach for identifying behavioural action sequences.

Animals display characteristic behavioural patterns when performing a task, such as the spiraling of a soaring bird or the surge-and-cast of a male moth searching for a female. Identifying such recurring sequences occurring rarely in noisy behavioural data is key to understanding the behavioural response to a distributed stimulus in unrestrained animals. Existing models seek to describe the dynamics of behaviour or segment individual locomotor episodes rather than to identify the rare and transient sequences of locomotor episodes that make up the behavioural response. To fill this gap, we develop a lexical, hierarchical model of behaviour. We designed an unsupervised algorithm called "BASS" to efficiently identify and segment recurring behavioural action sequences transiently occurring in long behavioural recordings. When applied to navigating larval zebrafish, BASS extracts a dictionary of remarkably long, non-Markovian sequences consisting of repeats and mixtures of slow forward and turn bouts. Applied to a novel chemotaxis assay, BASS uncovers chemotactic strategies deployed by zebrafish to avoid aversive cues consisting of sequences of fast large-angle turns and burst swims. In a simulated dataset of soaring gliders climbing thermals, BASS finds the spiraling patterns characteristic of soaring behaviour. In both cases, BASS succeeds in identifying rare action sequences in the behaviour deployed by freely moving animals. BASS can be easily incorporated into the pipelines of existing behavioural analyses across diverse species, and even more broadly used as a generic algorithm for pattern recognition in low-dimensional sequential data.

Spinal sensory neurons project onto the hindbrain to stabilize posture and enhance locomotor speed.

In the spinal cord, cerebrospinal fluid-contacting neurons (CSF-cNs) are GABAergic interoceptive sensory neurons that detect spinal curvature via a functional coupling with the Reissner fiber. This mechanosensory system has recently been found to be involved in spine morphogenesis and postural control but the underlying mechanisms are not fully understood. In zebrafish, CSF-cNs project an ascending and ipsilateral axon reaching two to six segments away. Rostralmost CSF-cNs send their axons ipsilaterally into the hindbrain, a brain region containing motor nuclei and reticulospinal neurons (RSNs), which send descending motor commands to spinal circuits. Until now, the synaptic connectivity of CSF-cNs has only been investigated in the spinal cord, where they synapse onto motor neurons and premotor excitatory interneurons. The identity of CSF-cN targets in the hindbrain and the behavioral relevance of these sensory projections from the spinal cord to the hindbrain are unknown. Here, we provide anatomical and molecular evidence that rostralmost CSF-cNs synapse onto the axons of large RSNs including Mauthner cells and V2a neurons. Functional anatomy and optogenetically assisted mapping reveal that rostral CSF-cNs also synapse onto the soma and dendrites of cranial motor neurons innervating hypobranchial muscles. During acousto-vestibular evoked escape responses, ablation of rostralmost CSF-cNs results in a weaker escape response with a decreased C-bend amplitude, lower speed, and deficient postural control. Our study demonstrates that spinal sensory feedback enhances speed and stabilizes posture, and reveals a novel spinal gating mechanism acting on the output of descending commands sent from the hindbrain to the spinal cord.

Somatostatin 1.1 contributes to the innate exploration of zebrafish larva.

Pharmacological experiments indicate that neuropeptides can effectively tune neuronal activity and modulate locomotor output patterns. However, their functions in shaping innate locomotion often remain elusive. For example, somatostatin has been previously shown to induce locomotion when injected in the brain ventricles but to inhibit fictive locomotion when bath-applied in the spinal cord in vitro. Here, we investigated the role of somatostatin in innate locomotion through a genetic approach by knocking out somatostatin 1.1 (sst1.1) in zebrafish. We automated and carefully analyzed the kinematics of locomotion over a hundred of thousand bouts from hundreds of mutant and control sibling larvae. We found that the deletion of sst1.1 did not impact acousto-vestibular escape responses but led to abnormal exploration. sst1.1 mutant larvae swam over larger distance, at higher speed and performed larger tail bends, indicating that Somatostatin 1.1 inhibits spontaneous locomotion. Altogether our study demonstrates that Somatostatin 1.1 innately contributes to slowing down spontaneous locomotion.

ZebraZoom: an automated program for high-throughput behavioral analysis and categorization.

The zebrafish larva stands out as an emergent model organism for translational studies involving gene or drug screening thanks to its size, genetics, and permeability. At the larval stage, locomotion occurs in short episodes punctuated by periods of rest. Although phenotyping behavior is a key component of large-scale screens, it has not yet been automated in this model system. We developed ZebraZoom, a program to automatically track larvae and identify maneuvers for many animals performing discrete movements. Our program detects each episodic movement and extracts large-scale statistics on motor patterns to produce a quantification of the locomotor repertoire. We used ZebraZoom to identify motor defects induced by a glycinergic receptor antagonist. The analysis of the blind mutant atoh7 revealed small locomotor defects associated with the mutation. Using multiclass supervised machine learning, ZebraZoom categorized all episodes of movement for each larva into one of three possible maneuvers: slow forward swim, routine turn, and escape. ZebraZoom reached 91% accuracy for categorization of stereotypical maneuvers that four independent experimenters unanimously identified. For all maneuvers in the data set, ZebraZoom agreed with four experimenters in 73.2-82.5% of cases. We modeled the series of maneuvers performed by larvae as Markov chains and observed that larvae often repeated the same maneuvers within a group. When analyzing subsequent maneuvers performed by different larvae, we found that larva-larva interactions occurred as series of escapes. Overall, ZebraZoom reached the level of precision found in manual analysis but accomplished tasks in a high-throughput format necessary for large screens.