MSACL 2024 Abstract

The focus of our work is to understand how normative pathways for immune tolerance are hijacked in disease. Progression to metastatic cancer, delivery of a healthy full-term infant, and pulmonary tuberculosis (TB) each occur when the immune system acquires tolerance to foreign antigens. These programs are mediated through an evolving spatiotemporal network involving dozens of cell types and multiple organ sites. Consequently, understanding them requires technical approaches to identify these cell types and their functions within intact human tissue. To unravel this connection between tissue structure and function, we have developed a spatial biology framework for studying these processes. In 2017, we created a new type of ion microscope for high-dimensional subcellular imaging of intact archival human tissue from medical center biobanks (MIBI-TOF). Combining this technology with new reagents, statistical tools, and deep learning algorithms permitted us to generate the first comprehensive spatial representations of human tissue where the lineage and functional state of every cell was mapped and quantified. We have since used this capability to understand how immune function and tissue structure are coupled in cancer, infectious disease, and pregnancy. In these studies, we discovered spatial rulesets for predicting immune cell function and clinical outcome in pulmonary tuberculosis, ductal carcinoma in situ (DCIS), and triple negative-breast cancer (TNBC). Remarkably, we have found significant redundancy and conservation in these programs. This approach has revealed a shared ontology in pregnancy, tuberculosis granulomas, and metastatic cancer where innate immune suppression, stromal desmoplasia, and regulatory T-cells sustain a tolerogenic niche that deters cytotoxic T-cell infiltration. With this in mind, we are now working to create a universal spatial ontology of immune tolerance. Just as red, green, and blue can be mixed to create any color, we think the vast spectrum of tissue immune responses can be broken down into weighted combinations of a much smaller set of regulatory programs. Using complementary techniques for spatial transcriptomics and MALDI mass spectrometry imaging, we are expanding our spatial biology framework to map the spatial distribution of transcripts, lipids, and glycans. In doing so, our goal is to enable a new era of anatomic pathology where a single unified spatial representation can be used to discover regulatory pathways that can be targeted for diagnostic and therapeutic purposes.