Predicting differential antigen-antibody contact regions based on solvent accessibility

Area of Science:

• Computational biology
• Immunology
• Protein structure prediction

Background:

• Botulinum neurotoxins (BoNT/A-G) and tetanus toxin (TeTX) are distinct and potent toxins.
• Understanding toxin structure is crucial for developing effective vaccines and therapeutics.
• The tertiary fold of these toxins is currently unknown, posing a challenge for epitope prediction.

Purpose of the Study:

• To develop and validate a novel computational approach for predicting epitopes from primary protein sequences.
• To analyze the heavy chains of TeTX and BoNT/A-G to identify immunologically relevant regions.
• To provide a theoretical framework for the design of genetically engineered vaccines.

Main Methods:

• Utilized an artificial neural network to estimate residue solvent accessibilities from multiple aligned sequences.
• Employed a similar neural network trained for secondary structure prediction.
• Validated the algorithm's accuracy using avian egg-white lysozyme and known lysozyme-antibody cocrystal data.

Main Results:

• The algorithm accurately predicted secondary structures (80% accuracy) and identified highly solvent-exposed residues in lysozyme.
• Analysis of TeTX and BoNT/A-G heavy chains revealed clustered, highly exposed regions at the N- and C-termini.
• Predicted exposed regions on heavy chains align with experimentally identified protective fragments and regions interfering with toxin binding.

Conclusions:

• The computational approach provides a theoretical basis for predicting toxin epitopes and potential vaccine targets.
• Nonconserved, highly exposed regions may explain type-specific antibody binding differences.
• The findings support the development of genetically engineered vaccines targeting specific toxin regions.