Colonoid-Based Transcriptomics Reveals Conserved and Model-Specific Mechanisms of Doxorubicin-Induced Intestinal Toxicity
Saad Lodhi, Marcel Van Herwijnen, Colette Kelly, Harvey Fowler-Williams, Carrie A. Duckworth, D. Mark Pritchard, Florian Caiment, Theo M. C. M. de Kok, Marcha C. T. Verheijen, Danyel G. J. JennenAnimal models are standard for safety evaluation, yet physiological differences limit human translation. Doxorubicin, a chemotherapeutic agent, can cause off-target gastrointestinal toxicity and treatment discontinuation. This study evaluated human colonoids as a controlled human-derived epithelial model for doxorubicin-induced gastrointestinal toxicity by comparing transcriptomic responses across human colonoids, mouse colonoids, and male C57BL/6J mouse colon tissue. Within each dataset, doxorubicin-treated conditions were pooled across model-specific exposure levels and time points to estimate broad doxorubicin-associated transcriptional signatures. Using a parallelogram approach, differential expression, co-expression, pathway mapping, and Comparative Toxicogenomics Database benchmarking were applied to compare model concordance and identify doxorubicin-responsive mechanisms. Shared responses converged on cell-cycle regulation, DNA damage response, DNA repair, and apoptosis. Concordance in differentially expressed genes was highest between mouse colonoids and mouse colon, while pathway mapping showed similarities between colonoid systems. A core set of p53-associated genes, including BAX, INKA2, and ZMAT3, was shared across datasets, with additional apoptotic and DNA damage response features observed in colonoids. Comparative Toxicogenomics Database benchmarking supported doxorubicin biology and highlighted underrepresented gastrointestinal toxicity-relevant signals, indicating gaps in intestinal toxicogenomic annotations. Colonoid-based transcriptomics supports human colonoids as controlled epithelial models for mechanistic gastrointestinal toxicity assessment, although missing vascular, immune, and systemic context means clinical validation remains necessary.