Intercellular Model Predicts Macrophage Signaling As A Driver Of Cardiac Fibroblast Response Post Myocardial Infarction

Introduction: Post-myocardial infarction (MI), cardiac fibroblasts and macrophages work together to regulate tissue homeostasis and infarct repair. Macrophage-fibroblast interactions in healthy tissue are stable and resistant to perturbations. However, this robustness post-MI has not been assessed. This study designs and implements an intercellular communication model of macrophage-fibroblast crosstalk to determine drivers of infarct repair. Methods: An ordinary differential equation model of post-MI cellular dynamics was developed ( Figure 1A ). Model inputs are time courses of cardiomyocytes, neutrophils, and monocytes. These cells, along with simulated macrophages and fibroblasts, secrete chemokines and cytokines which directed cell proliferation, removal, and chemical secretion. The outputs are macrophage and fibroblast densities, secreted factor dynamics, and produced collagen. Model validation was done using published data in post-MI mice. A sensitivity analysis was conducted by knocking down individual parameters to identify key drivers of fibroblast collagen production. Results: The simulated trends matched the validation time courses. Of the 28 validation relationships, 12 were input-output, 7 were knockouts, and 9 were inhibitor relationships. The validation passed at 78.5%; the 6 failed validations were due to the independent nature of the input curves. Sensitivity analysis ( Figure 1B ) identified macrophage differentiation rate, removal rate, and transforming growth factor-beta (TGFB) secretion rate as pro-fibrotic. Prolonged exposure to TGFB and granulocyte-macrophage colony stimulating factor was anti-fibrotic. Several clustered parameters differentially regulated macrophages and collagen production. Conclusions: The multicellular model identified macrophage density as a pro-fibrotic driver in the healing infarct. Differential regulation of macrophages and collagen production was predicted by the model.