Sequence-developability mapping of affibody and fibronectin paratopes via library-scale variant characterization

Area of Science:

• Biotechnology
• Protein Engineering
• Molecular Biology

Background:

• Protein developability is crucial for therapeutic, diagnostic, and industrial applications.
• Current low-throughput assays limit developability assessment to later stages of protein discovery.
• Novel high-throughput methods offer potential for broader applicability across protein families and metrics.

Purpose of the Study:

• To evaluate three library-scale assays for predicting protein developability outcomes.
• To assess the predictive power of on-yeast protease, split green fluorescent protein (GFP), and non-specific binding assays.
• To determine the utility of these assays for small protein scaffolds like affibody and fibronectin.

Main Methods:

• Library-scale screening using on-yeast protease, split GFP, and non-specific binding assays.
• Experimental assessment of thermal stability and recombinant expression for affibody and fibronectin scaffolds.
• Analysis of assay predictive capabilities using linear correlation and machine learning models.

Main Results:

• The on-yeast protease assay demonstrated high predictability for thermal stability in both affibody and fibronectin scaffolds.
• The split-GFP assay provided informative predictions for affibody thermal stability and expression.
• Library-scale data facilitated the mapping of sequence-developability landscapes for binding paratopes.

Conclusions:

• Library-scale assays can effectively predict key protein developability characteristics.
• These high-throughput methods accelerate the design and evolution of functional protein therapeutics.
• Mapping sequence-developability landscapes guides the rational design of improved protein variants and libraries.