(19h) Correlations of Antibody Response Phenotype and Genotype Revealed By Immunoglobulin Repertoire Sequencing

It has long been possible to measure the phenotype of antibody responses (antigen-specific titers) through conventional serological assays (e.g., ELISA). In contrast, the ability to measure the genotype of antibody responses has only recently become possible through the advent of high-throughput antibody repertoire sequencing (Ig-seq), which provides quantitative molecular information on clonal expansion, diversity and somatic hypermutation. However, Ig-seq is compromised by the presence of bias and errors introduced during library preparation and sequencing and thus prevent reliable immunological conclusions from being made. By using synthetic antibody spike-in genes, we determined that Ig-seq data overestimated antibody diversity measurements by up to 5000-fold and was less than 60% accurate in clonal frequency measurements. To overcome this, we developed molecular amplification fingerprinting (MAF), which uses unique molecular identifier-tagging before and during multiplex-PCR amplification, enabling tagging of transcripts while accounting for PCR efficiency. Our method allowed us to develop a novel algorithm for bias correction, which improved clonal frequency measurements to an accuracy of up to 99%. Furthermore, we demonstrated nearly absolute error correction (98-100%) of full-length antibody diversity measurements. In order to probe the relationship between antibody genotype and phenotype, we have performed extensive in vivo experiments from immunized mice. Specifically, we modulated titer (phenotype) by varying the number of immunizations (boosts) and in parallel performed MAF Ig-seq to assess genotype through a series of quantitative metrics. These metrics describe clonal selection, expansion, and somatic hypermutation. Importantly, we also used different protein antigens, this allowed us to determine how epitope complexity impacted the genotype of antibody repertoires. Finally, by applying multivariate modeling using MAF-corrected data, we were able to predict the immune status of individual antibody clones. This extensive systems-based analysis demonstrates how accurate Ig-seq provides insight on the genotype and phenotype of humoral immunity.