Methods to Study Antigen-Specific B cells

Abstract

The adaptive immune system enables us to quickly respond to infections in a targeted manner through the generation of highly specific B and T lymphocytes. B cells achieve this through somatic diversification of the B cell receptor which results in an unimaginably large repertoire of cells that can recognise a vast array of antigens. Studying the antigen-specific B cell repertoire is important for our understanding of infection, vaccination, and autoimmunity as it can help us gain insights into how protective or auto-reactive antibodies are generated. Furthermore, monoclonal antibodies have proven to be invaluable tools for therapeutic and diagnostic purposes. However, finding the cells that produce specific antibodies against a certain antigen target is challenging because they are buried in a pool of irrelevant cells. To find this proverbial ‘needle in a haystack’, high-throughput methods that enable the screening of millions of cells are urgently needed. With the goal of expanding the experimental toolkit for the study of antigen-specific B cells, I explore two complementary methods in this dissertation: a microfluidic platform for studying antibody secreting cells (ASCs) and a reporter mouse model that enables fate-mapping of stimulated B cells. Firstly, I introduce a screening platform for studying single ASCs. Studying ASCs with high throughput is inherently difficult due to the problem of linking the secreted antibody to the cell it originated from. To overcome this challenge, I combined microfluidic encapsulation in a capture hydrogel together with a FACS-based readout. Cell-antibody pairs of interest are selected by flow cytometric sorting using fluorescently labelled antigen and are then sequenced to correlate antigen binding with antibody sequence. For a proof-of-concept, I applied this microfluidic platform to the study of mouse and human antibody secreting cells. In mice, I generated anti-hen egg ovalbumin antibodies through immunisation and compared outputs generated by spleen and bone marrow plasma cells. I then applied the technology to produce a set of monoclonal antibodies against the SARS CoV 2 receptor binding domain (RBD) and characterised these antibodies in terms of their binding affinity and neutralisation capacity. In comparison to traditional hybridoma-based antibody generation, this direct plasma cell sorting and sequencing can speed up the antibody discovery process from several months to just weeks. For a proof-of-concept in humans, I studied the antigen-specific plasmablast response after vaccination with the SARS-CoV-2 mRNA vaccine BNT162b2. Specifically, I sorted RBD- or S1-specific plasmablasts after the second vaccine dose. Finally, I devised an Illumina-based sequencing and analysis strategy to generate antibody sequences from multiple plates of sorted hydrogels. To complement the ASC screening platform, I characterised an antigen receptor signalling reporter mouse which enables the tracking of cells that have received stimulation through their BCR at a defined time point of the immune response. This models enables tracking of cells that were activated early prior to entry into the germinal centre. I used this model to study the clonal expansion of cells that were stimulated after immunisation with hen egg ovalbumin and created a single cell transcriptomics dataset of fate-mapped cells which can serve as a reference for further studies with this mouse model.