From candidate ranking to lead identification - Hydrophobic interaction chromatography, PAIA, and structure-based calculations for hydrophobicity determination during developability assessment of mAbs and a mAb-derived fusion protein

Antibody therapeutics have gained critical relevance in the treatment of numerous health conditions, yet their development is resource-intensive and bears a non-negligible failure risk. Consequently, developability assessment to predict a protein's likeliness to evolve into an approved drug has become a pivotal measure. Surface hydrophobicity is a crucial developability feature because of its relation to self-interaction and aggregation. In our study, we determined the hydrophobicity of 21 antibodies and one antibody-derived fusion protein by hydrophobic interaction chromatography (HIC), the emerging PAIA assay, and structure-based calculations. We derived data analysis concepts for PAIA and calculated surface descriptors by taking well-established HIC as reference. Based on these concepts and considering each technique's capabilities and limitations, we deduce how HIC, PAIA, and surface descriptors for hydrophobicity can efficiently be combined throughout the different developability assessment stages to meet material availability and required throughput and characterization depth. For discovery and early development, we illustrate the definition of in silico thresholds to identify low-risk candidates and reduce the initial candidate pool. We then show how data from 384-well plate-based PAIA screening can be used for the classification into low-, medium-, and high-hydrophobicity molecules, followed by the high-resolution hydrophobicity ranking of final lead candidate(s) by HIC.