Synthetic Antibodies Designed on Natural Sequence Landscapes

We present a method for synthetic antibody library generation that combines the use of high-throughput immune repertoire analysis and a novel synthetic technology. The library design recapitulates positional amino acid frequencies observed in natural antibody repertoires. V-segment diversity in four heavy (VH) and two kappa (Vκ) germlines was introduced based on the analysis of somatically hypermutated donor-derived repertoires. Complementarity-determining region 3 length and amino acid designs were based on aggregate frequencies of all VH and Vκ sequences in the data set. The designed libraries were constructed through an adaptation of a novel gene synthesis technology that enables precise positional control of amino acid composition and incorporation frequencies. High-throughput pyrosequencing was used to monitor the fidelity of construction and characterize genetic diversity in the final 3.6 × 1010 transformants. The library exhibited Fab expression superior to currently reported synthetic approaches of equivalent diversity, with greater than 93% of clones observed to successfully display both a correctly folded heavy chain and a correctly folded light chain. Genetic diversity in the library was high, with 95% of 7.0 × 105 clones sequenced observed only once. The obtained library diversity explores a comparable sequence space as the donor-derived natural repertoire and, at the same time, is able to access novel recombined diversity due to lack of segmental linkage. The successful isolation of low- and subnanomolar-affinity antibodies against a diverse panel of receptors, growth factors, enzymes, antigens from infectious reagents, and peptides confirms the functional viability of the design strategy.

Graphical AbstractFigure optionsDownload high-quality image (125 K)Download as PowerPoint slideResearch Highlights
► A novel method for synthesizing diverse gene libraries with precise control over positional amino acid frequencies.
► A library of synthetic antibodies designed to mimic amino acid usage in natural repertoires.
► Evidence that natural sequence landscapes confer expression advantages to synthetic libraries.
► Evidence that synthetic assembly confers combinatorial diversity advantages to natural repertoire designs.
► A method for analyzing theoretical clone frequencies in large antibody populations.