A High-Content Imaging Pipeline to Investigate Subcytotoxic Effects in RTgill-W1 Cells

High-content screening offers great potential for in vitro toxicology, yet its application in environmental test systems remains limited by the lack of open, standardized workflows and biologically interpretable analysis frameworks. Here, we established and optimized a reproducible pipeline for image-based phenotypic profiling of rainbow trout gill cells (RTgill-W1) exposed to environmental chemicals, using Tebuthiuron as a case study. The workflow integrates multiple organelle dyes and systematically evaluates critical data processing steps (quality control, standardization, outlier removal, concentration-response-guided feature selection, and dimensionality reduction) to identify robust, biologically meaningful end points. Our systematic comparison demonstrated remarkable robustness: even minimal processing detected concentration-response relationships, while optimized standardization and feature selection substantially improved signal clarity. The framework identified 94 morphological features capturing Tebuthiuron's concentration- and time-dependent cellular responses, with lysosomal redistribution and mitochondrial fragmentation emerging as early stress indicators. Maximum mean discrepancy and Uniform Manifold Approximation and Projection (UMAP) visualizations provided biologically interpretable maps of phenotypic divergence. This proof-of-concept demonstrates that our quality-controlled, transparent HCS pipeline can resolve concentration- and time-dependent subcellular responses in fish cells, establishing a methodological basis for future multichemical phenotypic profiling.