227
Jin Zhang et al. (eds.), Fluorescent Protein-Based Biosensors: Methods and Protocols, Methods in Molecular Biology, vol. 1071,
DOI 10.1007/978-1-62703-622-1_18, © Springer Science+Business Media, LLC 2014
Chapter 18
Integrating Fluorescent Biosensor Data Using
Computational Models
Eric C. Greenwald, Renata K. Polanowska-Grabowska,
and Jeffrey J. Saucerman
Abstract
This book chapter provides a tutorial on how to construct computational models of signaling networks for
the integration and interpretation of FRET-based biosensor data. A model of cAMP production and PKA
activation is presented to provide an example of the model building process. The computational model is
defined using hypothesized signaling network structure and measured kinetic parameters and then simu-
lated in Virtual Cell software. Experimental acquisition and processing of FRET biosensor data is discussed
in the context of model validation. This data is then used to fit parameters of the computational model
such that the model can more accurately predict experimental data. Finally, this model is used to show how
computational experiments can interrogate signaling networks and provide testable hypotheses. This simple,
yet detailed, tutorial on how to use computational models provides biologists that use biosensors a powerful
tool to further probe and evaluate the underpinnings of a biological response.
Key words Computational modeling, FRET biosensors, Virtual cell, Cell signaling, cAMP
1 Introduction
Fluorescence microscopy of genetically encoded fluorescence
resonance energy transfer (FRET)-based biosensors is a powerful
technique for understanding signaling kinetics in living cells. These
biosensors allow researchers to visualize and quantify the spatiotem-
poral distribution of signaling molecules within the cell. As discussed
in previous chapters, a variety of FRET biosensors have been devel-
oped to directly detect changes in intracellular concentrations or
activation of signaling molecules in real time [1]. It is often the goal
to understand how these biochemical dynamics are modulated by
the overall signaling network, particularly in response to pharmaco-
logic or genetic perturbations. The complexity of signaling networks
often hinders attempts to relate biosensor data directly to the molec-
ular mechanisms underlying dynamic cell responses. Computational
models allow integration of diverse biochemical and biosensor data