Logo Logo
Help
Contact
Switch language to German
Signatures of motion processing and decisions in the larval zebrafish brain
Signatures of motion processing and decisions in the larval zebrafish brain
This thesis traces the investigation of representations of whole field motion, a highly behaviorally relevant stimulus across species and contexts, in the brain of the larval zebrafish. The studies both cover the whole brain, as well as focus on specific brain structures (the cerebellum and the interpeduncular nucleus) and its processing in different contexts, in particular evidence accumulation and decision making. In the first manuscript, we preformed a comprehensive characterisation of sensory and motor responses in the whole granule cell population of the larval zebrafish cerebellum. We found a responses to both neutral and behavior-eliciting stimuli (such as whole field motion), multimodality and dense, temporally-correlated activation of the population. The lack of sparseness and temporally-uniform activity were surprising in the context of prevailing theories of cerebellar function, however, the activity patterns we describe cover a rich range of sensorimotor signals that can support cerebellar learning. The second manuscript presents an adaptation of the classical random-dot kinematogram stimulation paradigm to zebrafish, by using the stimulus to elicit an optomotor response. This enabled us to examine decision and evidence accumulation together with whole-brain imaging. We have found that the optomotor behavior in this condition exhibits characteristic properties of evidence accumulation-based decision making: uncertainty-dependent latencies and error rates, as well as history depenence. We analyzed whole-brain recordings, finding signatures of different parts of the evidence-accumulation process distributed thought the brain. Responses with properties indicative of final stages of evidence accumulation - bidirectional modulation and long time constants - were localized in several midbrain structures, most prominently in nuclei raphe and the interpeduncular nucleus (IPN). In the third manuscript we investigated the motion-response properties of the IPN, identified in the previous study as a possible nexus of motioninformation. We characterized its anatomy with confocal imaging, and using functional imaging in different transgenic lines discovered precise geometric patterning of the responses to different motion directions thought the structure. Complementing this data with a traced electron microscopy dataset we found structural correspondences that partly explain the spatial distribution of the responses, in particular potential axo-axonal inhibition. Finally, we present Stytra, the software system for stimulation and behavioral tracking built to perform most of the studies included in the thesis, as well as many other studies in the lab. In addition to Stytra, I describe the associated open-source ecosystem we built alongside in the lab to acquire and analyse behavioral and imaging data.
whole-brain imaging, cerebellum, granule cells, decision making, open-source software
Štih, Vilim
2021
English
Universitätsbibliothek der Ludwig-Maximilians-Universität München
Štih, Vilim (2021): Signatures of motion processing and decisions in the larval zebrafish brain. Dissertation, LMU München: Graduate School of Systemic Neurosciences (GSN)
[img]
Preview
PDF
Stih_Vilim.pdf

51MB

Abstract

This thesis traces the investigation of representations of whole field motion, a highly behaviorally relevant stimulus across species and contexts, in the brain of the larval zebrafish. The studies both cover the whole brain, as well as focus on specific brain structures (the cerebellum and the interpeduncular nucleus) and its processing in different contexts, in particular evidence accumulation and decision making. In the first manuscript, we preformed a comprehensive characterisation of sensory and motor responses in the whole granule cell population of the larval zebrafish cerebellum. We found a responses to both neutral and behavior-eliciting stimuli (such as whole field motion), multimodality and dense, temporally-correlated activation of the population. The lack of sparseness and temporally-uniform activity were surprising in the context of prevailing theories of cerebellar function, however, the activity patterns we describe cover a rich range of sensorimotor signals that can support cerebellar learning. The second manuscript presents an adaptation of the classical random-dot kinematogram stimulation paradigm to zebrafish, by using the stimulus to elicit an optomotor response. This enabled us to examine decision and evidence accumulation together with whole-brain imaging. We have found that the optomotor behavior in this condition exhibits characteristic properties of evidence accumulation-based decision making: uncertainty-dependent latencies and error rates, as well as history depenence. We analyzed whole-brain recordings, finding signatures of different parts of the evidence-accumulation process distributed thought the brain. Responses with properties indicative of final stages of evidence accumulation - bidirectional modulation and long time constants - were localized in several midbrain structures, most prominently in nuclei raphe and the interpeduncular nucleus (IPN). In the third manuscript we investigated the motion-response properties of the IPN, identified in the previous study as a possible nexus of motioninformation. We characterized its anatomy with confocal imaging, and using functional imaging in different transgenic lines discovered precise geometric patterning of the responses to different motion directions thought the structure. Complementing this data with a traced electron microscopy dataset we found structural correspondences that partly explain the spatial distribution of the responses, in particular potential axo-axonal inhibition. Finally, we present Stytra, the software system for stimulation and behavioral tracking built to perform most of the studies included in the thesis, as well as many other studies in the lab. In addition to Stytra, I describe the associated open-source ecosystem we built alongside in the lab to acquire and analyse behavioral and imaging data.