Integrating Spectroscopy, Molecular Docking, and Machine Learning to Decipher Mycotoxin-Protein-Binding Mechanisms.

Area of Science:

• Biochemistry
• Toxicology
• Computational Chemistry

Background:

• Mycotoxins like ochratoxin A (OTA) and citrinin (CIT) pose significant health risks.
• Understanding their interactions with globular proteins, such as human serum albumin (HSA) and lysozyme (LYS), is crucial for assessing toxicity.

Purpose of the Study:

• To investigate the molecular binding mechanisms of OTA and CIT with HSA and LYS.
• To compare the binding affinities and interaction forces of these mycotoxins with different proteins.
• To evaluate the efficacy of integrated spectroscopic, computational, and machine learning (ML) approaches for studying mycotoxin-protein interactions.

Main Methods:

• Integrated spectroscopic techniques (fluorescence and UV-visible) to analyze mycotoxin-protein binding.
• Computational methods including molecular docking to predict binding modes and affinities.
• Machine learning (ML) models (random forest, Gaussian process regression) to predict quenching behavior.

Main Results:

• OTA exhibits strong, enthalpy-driven binding to HSA via static quenching, stabilized by hydrogen bonding and van der Waals forces.
• CIT shows weaker, polarity-driven interactions with both HSA and LYS.
• ML models accurately predicted quenching behavior (R² ≈ 0.985), outperforming traditional methods.
• Molecular docking confirmed stronger affinity between OTA and HSA compared to other complexes.

Conclusions:

• The combined approach of spectroscopy, ML, and docking provides quantitative and mechanistic insights into mycotoxin-protein binding.
• These findings are valuable for developing strategies for mycotoxin toxicity assessment and detoxification.
• OTA-HSA interaction is a key binding event with significant toxicological implications.