Abstract

Increasingly, interaction networks of genes and proteins are being put together in order to make sense of the flood of data coming from various model systems, as well as highly parallel measurements, due to technological advances. In addition to strong interactions, those that are transient, and localization dependent, are being recognized as an important component of such networks. Thus measuring mobility and transient interactions in living cells has become a critical part of the effort to understand how molecular components of cells function together.

I will discuss a simple system of protein localization in mammalian cells, mapped onto a compartment model to represent the exchange of molecules, mathematically represented as ordinary differential equations (ODE). The set-up is based on published literature, the "knowledge" of the investigator and 3D visualization of the data. The molecule Topoisomerase IIß (TopoIIß) was chosen for this, since it is a nuclear protein that modifies DNA topology. It localizes in high concentration in the nucleolus and lower concentrations in the nucleoplasm, contrasted with nuclear green fluorescent protein (GFP) which is homogeneously distributed. We initially set up the simplest model of such non-membranous compartmentalized distribution, designed fluorescence recovery after photobleaching (FRAP) experiments for protein (GFP) labelled TopoIIß mobility measurements and analyzed the data. In the process of fitting the model, we found it necessary to add components to improve the fit. This model was then validated by comparing it with alternative scenarios and shown to have predictive power.

At a different scale, we simulated the effect of gene-protein interaction networks and their sub-cellular localization patterns on cell behaviour. This was additionally a part of a multi-cellular simulation of tumour growth. I will present the framework of a simulation platform, which encapsulates the multi-scale patterns from the sub- to multi-cellular environment. We use this platform to make predictions about our system of brain-tumor growth at different scales and use them to interpret experimental data as well as aid experimental design.