Thermal proteome profiling and machine learning modeling for dissecting chemical-protein interactions in environmental toxicology: A mini-review and perspectives

Abstract

High throughput in vitro assays for screening chemical hazards focus primarily on specific receptors that are linked with certain adverse outcome pathways, neglecting potential novel endpoints or pathways induced by emerging pollutants. Identifying target proteins that interact with pollutants contributes to finding potential molecular initiating events under the adverse outcome pathways framework. Mass spectrometry-based thermal proteome profiling (TPP) assays have permitted uncovering binding targets of pollutants across the whole proteome. Based on the principle that proteins are thermally stabilized after binding with chemicals, TPP differentiates protein targets by determining the soluble fraction of proteins that remain stable after heat stress. Thus, TPP facilitates qualitative and quantitative measurements of chemical-protein interactions (CPIs) without modifications on chemical structures or immobilization of target proteins. In this mini-review, we introduced the principles, development and procedures of TPP, and summarized its applications in identifying protein targets and speculating toxicity pathways for emerging pollutants in environmental toxicological studies. Additionally, since measurements of CPIs using TPP for multiple chemicals could be labor- and cost-intensive, machine learning-based modeling is a feasible alternative to dissect CPIs due to its capability to mine intrinsic properties determining CPIs. Therefore, the recent development of machine learning models for CPI prediction was reviewed. Lastly, we envisioned prospects of combining TPP data with machine learning for CPI prediction, and the possibility of applying TPP to interpret toxicity pathways and phenotypes generated from multi-omics data, to inform future environmental toxicological research on forecasting targets and outcomes for emerging pollutants.

Graphical Abstract

Keywords:

• Chemical-protein interactions
• emerging pollutants
• machine learning
• molecular initiating events
• thermal proteome profiling

Handling Editors:

• Peng Gao/Scott Bradford

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This study was supported by the National Natural Science Foundation of China (22276020) and the National Key R&D Program of China (2022YFC3902100). We also acknowledge the funding from Open Foundation of Key Laboratory of Industrial Ecology and Environmental Engineering, MOE (KLIEEE-21-01).