Compound toxicity screening and structure-activity relationship modeling in Escherichia coli

Area of Science:

• Synthetic biology and metabolic engineering
• Microbial biotechnology
• Computational toxicology

Background:

• Cellular toxicity limits production of valuable compounds in engineered microorganisms.
• Understanding compound toxicity is crucial for both high-titer production and antimicrobial development.
• Existing toxicity prediction tools are primarily for eukaryotes, leaving a gap for prokaryotic systems.

Discussion:

• A library of 166 diverse compounds was screened for toxicity in Escherichia coli.
• A clustering algorithm ensured maximal chemical diversity within the compound library.
• Assay data was used to build a quantitative structure-activity relationship (QSAR) model for toxicity prediction.

Key Insights:

• The developed toxicity predictor assesses metabolite toxicity across the entire metabolome.
• This tool aids in fine-tuning heterologous gene expression for optimized compound production.
• The first E. coli toxicity prediction web server, EcoliTox, was created using QSAR models.

Outlook:

• Integration of the toxicity predictor into computational frameworks for metabolic pathway design.
• Facilitating the development of more efficient engineered microbial cell factories.
• Advancing the design of novel antibacterial agents through predictive toxicology.