Abstract
Next-generation sequencing (NGS) has been applied successfully to the field of therapeutic antibody discovery, often outperforming conventional screening campaigns which tend to identify only the more abundant selective antibody sequences. We used NGS to mine the functional nanobody repertoire from a phage-displayed camelid immune library directed to the recepteur d'origine nantais (RON) receptor kinase. Challenges to this application of NGS include accurate removal of read errors, correct identification of related sequences, and establishing meaningful inclusion criteria for sequences-of-interest. To this end, a sequence identity threshold was defined to separate unrelated full-length sequence clusters by exploring a large diverse set of publicly available nanobody sequences. When combined with majority-rule consensus building, applying this elegant clustering approach to the NGS data set revealed a wealth of >5,000-enriched candidate RON binders. The huge binding potential predicted by the NGS approach was explored through a set of randomly selected candidates: 90% were confirmed as RON binders, 50% of which functionally blocked RON in an ERK phosphorylation assay. Additional validation came from the correct prediction of all 35 RON binding nanobodies which were identified by a conventional screening campaign of the same immune library. More detailed characterization of a subset of RON binders revealed excellent functional potencies and a promising epitope diversity. In summary, our approach exposes the functional diversity and quality of the outbred camelid heavy chain-only immune response and confirms the power of NGS to identify large numbers of promising nanobodies.