Prediction of Single-Mutation Effects for Fluorescent Immunosensor Engineering with an End-to-End Trained Protein...

Area of Science:

• Biotechnology and Biosensing
• Protein Engineering
• Computational Biology

Background:

• Quenchbodies (Q-bodies) are homogeneous immunosensors utilizing fluorophore quenching by tryptophan residues near the antigen-binding site.
• Developing Q-bodies on demand is challenging due to the vast sequence space of complementarity-determining regions (CDRs) influencing antigen binding and quenching.
• Existing methods lack efficient strategies for predicting and optimizing Q-body performance.

Purpose of the Study:

• To pioneer a strategy combining high-throughput screening and protein language models (pLMs) for predicting Q-body fluorescence quenching with single amino acid resolution.
• To enhance Q-body performance by rationally designing mutations that improve quenching efficiency.
• To enable the on-demand development of Q-bodies for immunosensor applications.

Main Methods:

• Collected yeast-displayed nanobodies with varying TAMRA fluorophore quenching properties from a large synthetic library.
• Trained a pretrained pLM, coupled with a single-layer perceptron, on enriched CDR sequences for end-to-end quenching prediction.
• Utilized the trained model for in silico prediction of mutations affecting quenching in anti-SARS-CoV-2 nanobodies (RBD1i13 and RBD10i14) and validated through yeast surface display.

Main Results:

• The developed quenching prediction model, focusing on CDR1 + 3, demonstrated high performance in evaluation.
• In silico scanning successfully predicted mutations enhancing quenching in both RBD1i13 and RBD10i14 nanobodies.
• Experimental validation confirmed enhanced fluorescence quenching in engineered Q-bodies, leading to improved responses.

Conclusions:

• The developed strategy allows accurate prediction of fluorescence responses based solely on antibody sequence.
• This approach facilitates the rational selection and design of antibodies for creating high-performance immunosensors.
• The method is essential for advancing the development of Q-bodies and other antibody-based biosensors.