Modeling cellular network dynamics of liver homeostatic renewal

Abstract

Recent experimental studies identified specific subpopulations of self‐renewing hepatocytes with ability to populate the liver during homeostatic maintenance of the tissue. It remains an open question as to how this process of precursor transformation into mature hepatocytes is regulated by molecular pathways and tissue microenviroment to yield robust homeostatic renewal. We explored the contribution of putative control mechanisms that yield robustness to the renewal process by developing and analyzing a computational model of liver homeostatic renewal that considers a cellular network of precursor cells and mature hepatocytes. We demonstrate that the renewal network without feedback and nonlinear dynamics cannot maintain liver tissue in response to repeated injury. We modified the model to consider multiple alternatives including capacity constraints on proliferation, inhibition of precursor proliferation and transformation by mature hepatocytes, and putative precursor replenishment from non‐hepatocyte sources. We followed a Design of Experiments approach to investigate the system dynamics under all possible combinations of feedback configurations, specifically characterizing recovery of precursor and mature hepatocyte populations in response to a wide variety of disturbances. This unbiased exploration led us to identify a model configuration that is most robust in maintaining liver homeostasis. We compared the dynamics of the most robust model configuration with the available experimental data on the effects of hepatocyte senescence and stem cell transplantation on liver tissue renewal. We investigated the dynamics of the homeostatic renewal network in response to chronic repeated injury. Our results suggest that increasing frequency and magnitude of chronic repeated injury destabilizes the liver homeostasis within several months, which is inconsistent with epidemiological observations of a much slower decay of liver function over several years. We interpret these findings as suggestive of the presence of additional stabilizing controls potentially by the non‐parenchymal cells. We modified the model to include manipulation of the liver microenvironment signals by the liver non‐parenchymal cells as an additional layer of feedback control over hepatocyte levels. Our simulations and model analyses lead us to predict that robust homeostatic balance of hepatocytes is achieved by a combination of intrinsic system stability and control by non‐parenchymal cells. Taken together, our computational modeling study suggests that homeostatic renewal and maintenance of liver tissue is achieved through a coordinated regulation of the process by a combination of cell‐intrinsic and intercellular feedback control mechanisms.Support or Funding InformationResearch Support: R01 AA018873, T32 AA007463, F31 AA023445