Foldome database mining for novel crop pore-forming proteins with insecticidal potential

Overview

This solution leverages bioinformatic fold recognition algorithms to mine the AlphaFold structural database and uncover novel, previously uncharacterized pore-forming proteins in major crop species including maize, soybean, rice, and Arabidopsis. Pore-forming proteins are a proven mechanism for pest control in agriculture, and this approach enables the systematic discovery of new candidates that may serve as plant immunity proteins with native insecticidal or nematocidal activity. The platform supports both the identification of new pore-forming agents and the engineering of their specificity through mutagenesis and receptor-binding domain modifications, allowing activity to be redirected toward specific pests.

Technical specifications

Discovery and identification:

• Bioinformatic fold recognition algorithms scan the full AlphaFold structure database for distantly related homologs
• Structure-based mining has already identified dozens of novel pore-forming proteins in maize, soybean, and rice
• Sequence database searches using identified hits as templates expand the candidate pool
• Candidates are curated based on homology to known plant immunity proteins

Modification and engineering:

• Mutagenesis and sequence grafting approaches to reprogram receptor binding domains
• Previous work demonstrated that modifying receptor binding domains of fungal pore-forming proteins alters lipid specificity
• Engineered specificity enables targeting of specific insect or nematode pests

Validation pipeline:

• Candidate genes expressed and purified from a gene library
• Screening for cytotoxic activity against various cell lines and model organisms
• Structural investigation of successful candidates using cryoEM

Technology readiness level

The bioinformatic discovery pipeline is already operational, with dozens of novel pore-forming protein candidates identified in maize, soybean, and rice. The fold-mining method has been validated through prior success in identifying novel pore-forming proteins in humans. The team has demonstrated the ability to modify receptor binding domains to alter lipid specificity in related fungal proteins. The bioinformatics analysis phase requires approximately 4-8 weeks, while expression, purification, and screening of candidate molecules is estimated at 6-8 months. The next phase of work will focus on expanding the candidate list, curating the best candidates for biochemical and structural investigation, and engineering insect-specific variants of the most promising pore-forming proteins.