Innate immune sentinels serve to both safeguard the organism from foreign invaders and appropriately manage internal damage signals. These cells are distributed throughout the body – circulating the blood, lining the mucosal membranes and the blood vessels, surveying the connective tissue. Their role is to sense the environment, and then decide whether to respond and how best to respond, to maintain the health of the organism. While sentinels include several different cell types, including fibroblasts, endothelial cells, macrophages, and dendritic cells, we here focus on macrophages and examine two properties that may define their sentinel function: response-specificity and epigenomic programmability.
Immune sentinel macrophages initiate the immune response and interact with T-cells, B-cells, NK cells, etc. to coordinate downstream immune system activation. Considering the thousands of potential human pathogens, hundreds of cell damage molecules, and dozens of host-cytokines that may signify inflammatory danger, it should be evolutionarily unsurprising that macrophages respond differently to each signal. However, quantifying this specificity requires single cell data that reveals response distributions. In Chapter 2, we measured single-cell transcriptomic profiles of macrophages responding to diverse bacterial, viral, and cytokine stimuli to quantify the response-specificity of macrophages. Using information theory, we uncovered distinct functional gene groups that differed in their degree of response-specificity. We found that response-specificity was context-dependent, as some genes lost specificity when macrophages were responding from IFNγ or IL4 cytokine contexts. Examining peritoneal macrophages from old and obese mice, which may harbor inflammatory microenvironments, revealed diminished response-specificity compared to healthy mice. These findings suggest that macrophage response-specificity quantified by single cell mRNA measurements after stimulation may be the basis for an innate immune health score.
Beyond responding to an immediate threat by deploying the appropriate transcriptomic programs, immune sentinels may also retain a stimulus-specific memory of the ligand encountered. Prior experimental studies suggested that stimulus-activated transcription factors when bound to DNA could provide the trigger for epigenomic reprogramming, but the mechanisms remained unclear. In Chapter 3, we developed mathematical models of the nucleosome to investigate mechanisms by which the spontaneous kinetics of nucleosomal-DNA interactions interfaced with the stimulus-specific dynamics of transcription factor activity. First focusing on the signal-dependent transcription factor NFκB, our model predicted that presence or absence of oscillations in NFκB signaling was the primary determinant of alterations to epigenomic state. We next developed more detailed Markov models of the nucleosome with base-pair resolution. Fitting the models to time-course stimulus-response ATAC-seq data, we found that kinetic rates of nucleosome unwrapping were slower in vivo than reported in vitro rates, and that the propensity for nucleosome eviction was greatest if transcription factor binding sites sat close to but not on the nucleosome dyad. This work uncovered biophysical principles governing nucleosome dynamics in vivo, enabling the prediction of epigenomic alterations during inflammation at single nucleosome resolution.