A Novel Synthetic Antibody Library with Complementarity-Determining Region Diversities Designed for an Improved Amplification Profile

• Published in: International journal of molecular sciences Vol. 23; no. 11; p. 6255
• Main Authors: Bai, Xuelian, Jang, Moonseon, Lee, Nam Ju, Nguyen, Thi Thu Ha, Jung, Mooyoung, Hwang, Jeong Yeon, Shim, Hyunbo
• Format: Journal Article
• Language: English
• Published: Switzerland MDPI AG 02.06.2022; MDPI
• Subjects: amplification efficiency; Antibody libraries; antibody library; Antigens; Binding; Cloning; Combinatorial analysis; Complementarity-determining region; Depletion; Design; E coli; Efficiency; machine learning; Monoclonal antibodies; Next-generation sequencing; Panning; Phage display; Phages; Protein A; Proteins; Sequences; amplification efficiency; machine learning; antibody library; phage display; complementarity-determining region
• ISSN: 1422-0067; 1661-6596; 1422-0067
• DOI: 10.3390/ijms23116255

Antibody discovery by phage display consists of two phases, i.e., the binding phase and the amplification phase. Ideally, the selection process is dominated by the former, and all the retrieved clones are amplified equally during the latter. In reality, the amplification efficiency of antibody fragments varies widely among different sequences and, after a few rounds of phage display panning, the output repertoire often includes rapidly amplified sequences with low or no binding activity, significantly diminishing the efficiency of antibody isolation. In this work, a novel synthetic single-chain variable fragment (scFv) library with complementarity-determining region (CDR) diversities aimed at improved amplification efficiency was designed and constructed. A previously reported synthetic scFv library with low, non-combinatorial CDR diversities was panned against protein A superantigen, and the library repertoires before and after the panning were analyzed by next generation sequencing. The enrichment or depletion patterns of CDR sequences after panning served as the basis for the design of the new library. Especially for CDR-H3 with a higher and more random diversity, a machine learning method was applied to predict potential fast-amplified sequences among a simulated sequence repertoire. In a direct comparison with the previous generation library, the new library performed better against a panel of antigens in terms of the number of binders isolated, the number of unique sequences, and/or the speed of binder enrichment. Our results suggest that the amplification-centric design of sequence diversity is a valid strategy for the construction of highly functional phage display antibody libraries.