Investigating the muscle cell responses to myotoxicants: a multi-omics study of C2C12 myotubes and myoblasts

This thesis addresses the significant knowledge gap in myotoxicity, an under-researched area of toxicology, by investigating the muscle cell responses to a panel of 30 toxicants. Using a multi-omics approach on C2C12 myoblasts and myotubes, this research aimed to develop robust in vitro and in silico models to better predict and understand drug-induced muscle injury. The core methodologies included high-content phenotypic profiling with the Cell Painting Assay (CPA), transcriptomic analysis via TempO-Seq, and the development of computational models. Key findings demonstrate that the CPA is a valuable tool for detecting myotoxicants and differentiating their mechanisms of action, revealing unexpected phenotypic similarities between statins and certain tyrosine kinase inhibitors. Transcriptomics confirmed that statins significantly modulate the PI3K/Akt/mTOR pathway, a finding consistent with previous studies on statin-induced myotoxicity. Furthermore, machine learning models successfully predicted cellular cytotoxicity from morphological data, while advanced deep learning models accurately classified raw microscopy images by treatment class. A Quantitative Structure-Property Relationship (QSPR) model was also developed and validated, proving its ability to predict a compound's potential to induce rhabdomyolysis based solely on its chemical structure. In conclusion, this research provides a more comprehensive understanding of the mechanisms involved in myotoxicity and establishes a foundation for new predictive models for skeletal muscle toxicity, thereby contributing valuable tools and data for improved drug safety assessment.

Funder

European Union’s Horizon 2020

Publisher

University of Galway

Rights

CC BY-NC-ND

cbn

Collections

University of Galway Theses (PhD Theses)