Many cells have the ability to break symmetry. Polarity signaling networks, underlying these processes, can be highly complex and typically involve many component and feedback interactions. We have focused on understanding design principles underlying the polarity networks in budding yeast and chemotaxing neutrophils. Our approach combines experiment, image analysis, and mathematical modeling. Highlights of this research program include the discovery of a fundamentally new polarity mechanism, referred to as the "neutral drift model," in which positive feedback drives recurrent stochastic clustering of signaling molecules, and a new approach for uncovering information flow among cytoskeletal signaling modules in rapidly polarizing human neutrophils.
Genetically and morphologically identical cells can exhibit very different biochemical states. Cellular plasticity, the ability of cells to transition among states, is highly controlled in the homeostasis and regeneration of healthy tissues. Yet, when control mechanisms are lost, cellular plasticity can drive disease progression and promote drug resistance, notably in cancer. We currently focus on the origins and functional consequences of how tissue microenvironment and oncogenic progression lead to diversification of cellular phenotypes. We apply quantitative imaging approaches to investigate cellular plasticity and heterogeneity within model systems, including colonic stem cells, liver lobules, and adenocarcinomas of the colon and lung.
Recent advances in high-content fluorescence microscopy have accelerated progress in many areas of cell biology. Approaches from computer vision are essential for analyzing image data sets that are too large to examine by human eye. Over the last decade, our lab has pioneered machine-learning approaches for extracting informative signatures of cellular perturbations and identifying stereotyped subpopulations. These approaches are being used to study cell individuality and to reverse-engineer models of network behaviors.