Visual–acoustic thigmotaxis in zebrafish larvae: a high throughput NAM for neurotoxicity assessment

Introduction: Current regulatory neurotoxicity guidelines do not include behavioral endpoints that capture stress-related responses. Zebrafish larvae prior to independent feeding offer a promising vertebrate model for developing new approach methodologies (NAMs) because they combine neurobiological relevance with high-throughput potential. In this study, we developed and evaluated a larval thigmotaxis assay to detect behavioral alterations induced by neuroactive substances. Methods: Zebrafish larvae at 120 hpf were exposed for 1 h to model compounds and then challenged with visual (light/dark) and acoustic (tapping/silence) stimuli. Thigmotaxis, defined as edge-preference behavior, and locomotor activity were assessed. To increase throughput, we compared the conventional 24-round-well format with a 96-square-well format. Assay performance was evaluated using caffeine and diazepam as reference compounds, followed by additional neuroactive substances (chlorpyrifos, nicotine, dexamethasone, ethylenethiourea) and low-neuroactivity comparators (saccharin, amoxicillin). Benchmark dose modeling was used to compare the sensitivity of thigmotaxis and locomotor endpoints. Results: The 24-well and 96-well formats produced equivalent results, supporting use of the higher-throughput system. Reference compounds confirmed assay performance, with caffeine increasing thigmotaxis and diazepam decreasing it under specific stimulus conditions. Additional neuroactive substances produced stimulus-dependent behavioral responses, whereas saccharin and amoxicillin caused little or no effect. Across compounds, benchmark dose modeling showed that thigmotaxis was generally more sensitive than traditional locomotor activity endpoints. Discussion: This multiplexed visual-acoustic thigmotaxis assay is reproducible, scalable, and sensitive for detecting neuroactive effects in zebrafish larvae. It can be used either as a stand-alone behavioral NAM or integrated into a broader test battery for neurotoxicity assessment. The method provides a practical and ethical tool to support chemical safety assessment in both ecotoxicology and human toxicology.