Project Summary
Increased total serum IgE levels are associated with coronary artery disease (CAD). However, the causal role
of antigen-specific IgE in CAD remains largely unexplored. Recent work from our group and others provide
evidence that humans with IgE sensitization to the mammalian oligosaccharide allergen a-gal have larger
coronary artery plaques and unstable plaque features signifying increased CAD compared to those without IgE
to a-gal. Despite these compelling human associative findings, no study to date has investigated the role of
antigen-specific IgE as a driver of CAD severity and the molecular and cellular mechanisms mediating IgE
sensitization to a-gal linked to atherosclerosis. We recently reported that humans with IgE sensitization to a-gal
had a higher frequency of CCR6+ switched memory (SWM) B cells. Notably, consistent with the association of
the IgE sensitization to a-gal and CAD, the amount of CCR6 on SWM B cells was associated with CAD severity.
Transcriptomic analysis demonstrated that CCR6+ SWM B cells expressed higher IL-4R and STAT6 in subjects
that were IgE a-gal+ compared to IgE a-gal-. Interestingly, IL-4 and STAT6 are important for B cell class switch
recombination to IgE, suggesting that cells that make IL-4 may be important in IgE to a-gal production.
Preliminary data using a novel mouse model deficient in NKT cells that are early producers of IL-4 show reduced
levels of IgE to a-gal and implicates invariant NKT cells in the regulation of IgE antibody production to a-gal.
Based on these human and murine data, the overarching objective of the parent grant is to investigate whether
these factors and cells play a causal role in atherosclerosis development due to IgE sensitization to a-gal. A
major component of these studies is to analyze single cell mass cytometry data derived from PBMCs of a second,
independent and larger cohort of humans with CAD sensitized to a-gal allowing for more robust multivariate
analysis and deeper interrogation of immune cell phenotypes that mark those at greatest risk. The overall goal
of this research supplement is to make the mass cytometry data AI-ready with associated datasheets that contain
the cell subtype annotations, protein type and activation state markers, cohort statistics, CAD severity measures
and other relevant clinical variables (e.g., age, sex, smoking status, etc.). We will prepare noise filtered,
normalized, batch corrected, cell type annotated CAD patient single cell mass cytometry data for AI applications
and use an explainable machine learning framework to predict measures of CAD severity and identify cell types
driving the predictions. AI-ready data will be shared and serve as a model of how to prepare patient single cell
data to be AI-ready for precision medicine and other applications.
