Kaempf, F. | Search bioRxiv

bioRxiv2025

Parallel and convergent pathways for multifeature visual processing in larval zebrafish sensorimotor decision-making

Animals continuously extract and evaluate diverse sensory information from the environment to guide behavior. Yet, how neural circuits integrate multiple, potentially conflicting, inputs during decision-making remains poorly understood. Here, we use larval zebrafish to address this question, leveraging their robust optomotor response to coherent random dot motion and phototaxis towards light. We demonstrate that animals employ an additive behavioral algorithm of three visual features: motion coherence, luminance level, and changes in luminance. Using brain-wide two-photon imaging, we identify the loci of these computations, with the anterior hindbrain emerging as a multifeature sensory integration hub. Through single-cell neurotransmitter and morphological analyses of functionally identified neurons, we characterize potential connections within and across computational nodes. These experiments reveal three parallel and converging pathways, matching our behavioral results. Our study provides a mechanistic brain-wide account of how a vertebrate brain integrates multiple features to drive sensorimotor decisions, bridging the algorithmic bases of behavior and its neural implementation.

neuroscience↗

bioRxiv2025

Correlative light and electron microscopy reveals the fine circuit structure underlying evidence accumulation in larval zebrafish

AbstractEvidence accumulation is a fundamental neural computation essential for adaptive behavior, yet its synaptic implementation remains unclear. Addressing this challenge critically depends on linking neural dynamics to circuit structure within the same brain. Here, we combine functional calcium imaging with large-scale ultrastructural electron microscopy (EM) to uncover the wiring logic of visual evidence accumulation in larval zebrafish. In a functionally imaged EM dataset of the anterior hindbrain, we identify conserved morphological cell types whose activity patterns define distinct computational roles. Bilateral inhibition, disinhibition, and recurrent connectivity emerge as key circuit motifs shaping these dynamics. To generalize our findings across animals, we develop a photoconversion-based pipeline to label and reconstruct functionally characterized neurons, enabling us to train a classifier that predicts functional identity from morphology alone. Applying this classifier to a second, whole-brain EM dataset lacking functional data reveals matching connectivity patterns, significantly augmenting its applicability for detailed circuit dissections. Based on these results, we develop and constrain a biophysically realistic neural network model that captures observed dynamics and yields predictions we tested and confirmed experimentally. Our work illustrates how hypothesis-driven connectomics can uncover the synaptic basis of sensory-motor computations and establishes a novel framework for cross-animal circuit dissection in the vertebrate brain.

neuroscience↗