Our laboratory has a long-standing interest in B cell antigen receptor (BCR) signaling and how BCR dependent processes regulate specific cell fate decisions and contribute to disease pathogenesis.
Molecular Mechanisms of B Cell Development
In B cell development, a canonical two-receptor model, the IL7 receptor and the pre-BCR, were thought to orchestrate developmental transitions between proliferation and Ig kappa recombination. However, this model required the pre-BCR to do multiple functions over time, including driving Ig kappa recombination in small pre-B cells while it is itself being repressed. We discovered that the pre-BCR initiates an IRF4- CXCR4 feed-forward loop, and it is CXCR4 that directly signals to open both Ig kappa and Ig lambda to recombination. Elucidation of the pathways downstream of the IL-7R, pre-BCR and CXCR4, have revealed coordinate regulation of the epigenetic reader BRWD1. These findings fundamentally rewrite the canonical model of B lymphopoiesis and challenge current paradigms of chemokine receptor function.
BRWD1 remodels nucleosomes relative to DNA GAGA motifs, which is required for assembling the recombination center at Jk. Genome-wide, it represses the proliferative transcriptional programs of early B cell development and induces those required for Igk recombination. By determining accessibility at enhancers, BRWD1 controls the activation of developmentally regulated genes by key transcription factors. We have found that these BRWD1-mediated chromatin topological changes complement those mediated by transcription factors such as Ikaros to drive B cell development. Currently, we are developing in vivo and in vitro models to define the role of BRWD1 epigenetic regulation in germinal center responses, and to further elucidate the ways that CXCR4 directs late B lymphopoiesis.
In situ Immune Mechanisms in Human Autoimmunity
Our human translational studies have focused on in situ adaptive immune responses in humans and how they contribute to both autoimmunity and cancer. For these studies, we have used deep machine learning to develop methods based on high-dimensional imaging that can be applied to human tissue samples to identify when any antigen-presenting cell population is presenting peptides to T cells. Most notably, we have developed a new image analysis technology, Cell Distance Mapping (CDM), which allows, for the first time, identification of cognate adaptive cell networks in human tissue. Remarkably, this bioinformatics platform approaches the sensitivity and specificity of two-photon excitation microscopy (TPEM). However, unlike TPEM, it can be applied to the study of human disease. We are now using CDM to quantify adaptive cell networks across several autoimmune diseases. In addition to quantitative imaging, we also use single cell technologies and high-throughput antibody expression to understand B cell selection at sites of inflammation and determine the interrelationships between transcriptional state and antigenic specificity. Overall, our goal is to obtain a functional understanding of how human adaptive immunity evolves at sites of disease, using ML and high-dimensional imaging.
