This is a proposal to establish a new paradigm for the study of cellular signal transduction. Most of our current understanding of signal transduction pathways relies on solution biochemical analyses of protein interactions and on bulk measurement of select pathways averaged over large cell populations. The sensitivity of these measurements is severely limited by the averaging of heterogeneous signaling states in asynchronous cell populations. To obtain robust responses strong stimuli are usually applied to homogenize and synchronize the signaling activities across cell populations. However, such stimulation tends to activate pathways far outside the normal range of operation. As a result, the balance between different network branches is distorted and feedback/feedforward interactions between pathway nodes are obfuscated. Hence, for many pathways, our current knowledge lacks the level of detail required to pinpoint key differences in signaling between physiological and pathophysiological conditions. Many pathways, especially those controlling cell morphogenic functions, are regulated over a time scale of seconds and on subcellular length scales. Thus, single cell measurements, e.g. by multi-parameter flow cytometry, tend to be incomplete as well. We argue that the use of emerging biosensor technology, capable of indicating protein activity at resolution levels matching the spatiotemporal regulation of cellular signal transduction, would be key to producing conclusive models of cellular signaling. However, at present biosensor imaging is employed mostly to visualize the activity of an isolated node in a signaling network and with little quantitation of the image dynamics. Coupled transformational changes in biosensor engineering and image analysis are required to provide more than a phenomenological view of one aspect of a pathway. In recognizing this need, we bring here together the expertise of the Hahn lab in biosensor design and live cell imaging and the expertise of the Danuser lab in image analysis of dynamic cellular processes, to establish systematic and quantitative imaging of large signal transduction networks. We propose developments using uniform, engineered biosensor scaffolds to generate new biosensor approaches that enable sensitive multiplex imaging, in living cells, of currently inaccessible network nodes. We also propose developments of computational methods to extract from these data the direction, efficiency, and kinetics of signal transduction between concurrently imaged network nodes and to compile the data from many experiments into a single concise pathway model, despite cell to cell heterogeneity. Hence pathways with tens to hundreds of nodes can be probed despite the spectral constrictions of biosensors, which can be foreseen to image at most 4 - 5 nodes simultaneously (current technology is reaching towards imaging only two activities). Furthermore, the new computational methods will provide the ability to identify feedback/feedforward interactions between observed nodes, determine spatial cues in signal transduction, and predict feedbacks involving as yet unobserved nodes. Central to our approach is exploiting the high sensitivity of the new biosensor technology to probe networks based on the propagation of constitutive signaling fluctuations between nodes. Thus, we can avoid massive stimulation of the imaged network and instead generate a relevant reconstruction of signal transduction at physiological activation levels. Our developments will be driven by investigation of Rho GTPase coordination, a network with many suspected feedback interactions centrally implicated in the regulation of cell migration.
This project will bundle expertise of the Hahn lab at UNC Chapel Hill in biosensor design and live cell imaging with expertise of the Danuser lab at the Scripps Research Institute in image analysis, to produce a transformative approach to studying cellular signal transduction and decision processes. Novel, versatile biosensor technology will enable direct visualization of a wide range of signaling events in living cells. Novel mathematical methods will enable inference from live cell images of the wiring and dynamics of signal transduction through large networks, including the identification of feedback/feedforward interactions between network nodes and of subcellular regions with distinct transduction regimes. In combination, these advances promise a breakthrough in our ability to probe complex, spatially and temporally organized signaling processes in cell models of normal and diseased physiology. Our developments will be driven by investigations of the balance between Rho-family GTPases in multiple nested feedback interactions, a network centrally implicated in the regulation of the mechanics of cell migration.
Showing the most recent 10 out of 15 publications
