Prediction of Terpenoid Toxicity Based on a Quantitative Structure-Activity Relationship Model

Area of Science:

• Natural product chemistry
• Toxicology
• Computational chemistry

Background:

• Terpenoids are diverse plant-derived secondary metabolites with varied chemical structures.
• Assessing terpenoid toxicity is crucial for food safety and product development.
• Existing methods for toxicity assessment can be time-consuming.

Purpose of the Study:

• To develop a quantitative structure-activity relationship (QSAR) model for predicting terpenoid toxicity.
• To identify chemical structure features that influence terpenoid toxicity.
• To establish a rapid method for evaluating terpenoid toxicity in food products.

Main Methods:

• Toxicity of 27 terpenoid standards was assessed in real-time using the bioluminescent bacterium Vibrio fischeri.
• Quantitative structure-activity relationship (QSAR) models were developed using CODESSA software.
• Models correlated terpenoid toxicity with various molecular descriptors derived from chemical structures.

Main Results:

• Terpenoid toxicity varied significantly with concentration and functional groups (alcohol > aldehyde ~ ketone > ester > hydrocarbons).
• QSAR models identified geometric (e.g., asphericity) and electronic (e.g., partial charge) descriptors as key predictors of toxicity.
• A four-parameter QSAR model demonstrated strong predictive power (R² > 0.810, Q² > 0.689).

Conclusions:

• The developed QSAR models are effective and rapid tools for predicting terpenoid toxicity.
• The study provides insights into structure-toxicity relationships of terpenoids.
• This approach can aid in ensuring the safety of food products containing terpenoids.