135 Deep-mining antibody repertoires with sensitive, high-throughput single cell BCR and whole transcriptome sequencing

Background Antibody repertoires are highly diverse and individual-specific, necessitating large-scale sequence sampling to gain meaningful insights into the adaptive immune response. Both primary human samples and mouse models serve as rich sources of native diversity for studying antigen- or disease-specific antibody responses. Current antibody discovery methods lack the scale required to fully screen the antibody repertoire and uncover maximum desirable antibody hits per discovery campaign with high sensitivity. Here, we demonstrate our novel combinatorial barcoding approach to enable high-throughput single cell B cell receptor (BCR) and whole transcriptome profiling of human, mouse, or transgenic mouse samples with unprecedented sensitivity, scalability, and flexibility.Methods Parse single cell combinatorial barcoding method was used to profile both BCRs and whole transcriptomes of human, mouse, or transgenic mouse samples. For human, we profiled over 1 million cells from 12 healthy and 12 autoimmune human donors in a single experiment. For mouse, we performed a spike-in titration experiment with a fixed and known clonotype from a mouse model into wildtype mouse B cells. For transgenic mice with chimeric transcripts of human V-gene loci, we profiled B cells from a naive vs. immunized experiment. All samples were sequenced on a Illumina Novaseq and data was processed through Parse Biosciences Splitpipe. Analysis was done in Seurat v5.Results To highlight the high-throughput capabilities of this technology, we profiled over 1 million cells from 12 healthy and 12 autoimmune human donors in a single experiment resulting in over 900,000 unique paired clonotypes. For sensitivity benchmarking in mouse samples, we detected a known clonotype down to 0.1% frequency with high reproducibility in a spike-in titration experiment. In a transgenic naive vs. immunized experiment, we detected all human V-gene elements and found shifts in V-gene use between conditions. Using the whole transcriptome data, we detected all B cell subpopulations including plasma cells at single cell resolution and found an increase in plasma cell proportion in the immunized animal. We also found specific isotype switching to IgG in these cells, suggesting that these BCR sequences are potential hits for functional characterization.Conclusions This method paves the way for high-throughput screening of single cell BCR sequences to allow for deep mining of repertoires for improved antibody discovery and therapeutic outcomes from human, mouse, or transgenic mouse samples.