Modelling Nuclear Body Dynamics in Living Cells by 4-D Microscopy, Image Analysis and Simulation

Abstract

The work presented here demonstrates rules of and validates models for nuclear body (NB) dynamics. Simulation tools developed in the course of this work can be used in future work to generate hypotheses about related aspects of nuclear architecture. Initially I examined the mobility of vimentin nuclear bodies bodies (VNB) in interphase by single particle tracking and analysis of fluorescence images from 4-D confocal laser scanning microscopy (CLSM). These synthetic nuclear bodies are observed in cells transfected with labelled nuclear-targeted Xenopus laevis vimentin. Analysis shows that VNBs undergo anomalous diffusion in the nuclei, independent of metabolic energy. Individual bodies display either one of the three modes of diffusion -- directed, restricted or simple. The consistency of modes and magnitudes of diffusion constants between VNBs and bona fide nuclear bodies points to a generic mechanism that mediates and regulates the mobility of nuclear bodies. Since the results of diffusion analysis of VNBs did not agree with a simple diffusion model, I tested the alternative interchromosomal domain (ICD) compartment model. The ICD model predicts that in interphase cell nuclei, individual decompacted chromosomes do not intermingle, but are separated by a significant interchromatin space forming a network of channels. These networks could affect the mobility of nuclear bodies. Monte Carlo simulations that predict the effects of channels and other obstructions on NB diffusion were tested, but they could not explain deviation from ideal behaviour. Fitting an empirical model of `critical diffusion' produced similar results. Therefore the ICD model as a purely obstructing network of channels needs modification, to possibly include binding. To examine the role of chromatin density in intra-nuclear diffusion, I employed multidimensional fluorescence recovery after photobleaching (FRAP) in living cells. The influence of chromatin density on diffusive mobility of the nuclear yellow fluorescent protein (YFP) appears marginal. A 2-D diffusion simulation to better characterize the experiment provides a tool to produce `diffusion maps' of the nucleus. The related aspect of nuclear body integrity and dynamics was examined for the distribution of topoisomerase II beta (TopoIIb), which localizes preferentially in the nucleolus. The experimentally observed diffusion and binding dynamics were formulated as a compartment model and fitted to the experiments. The model topology, flux constants and residence times estimates could be validated, providing a predictive model of TopoIIb dynamics By demonstrating that VNB diffusion is anomalous and consistent with other bona fide NBs, I have revealed a mechanism that regulates NB mobility. The diffusion of these bodies deviates however from ideal diffusion, and can be explained by neither the effect of chromatin density on molecular diffusion, nor the different models of NB diffusion. I have shown that binding rather than diffusion appears to determine nuclear body localization and dynamics, as in the case of TopoIIb. Nuclear bodies and nuclear architecture has been recently hypothesized as emerging from simple local interactions. The predictive model for TopoIIb distribution dynamics provides evidence for this. The models presented here, are in keeping with the increasing trend to abstract nuclear dynamics as mathematical models. It is hoped that the work presented here will contribute to the effort of arriving at an integrated model for nuclear bodies and therefore better understanding nuclear architecture.