Analyzing motion and deformation of the cell nucleus for studying co-localizations of nuclear structures

In cell biology, modern imaging techniques using fluorescence microscopy allow to visualize specific nuclear structures in situ in the same cell nucleus. Hence, distances between these structures can be evaluated, in particular co-localization can be investigated. When the nucleus alters its global shape, especially if the structures are imaged sequentially, the distances are changing as well and the global movements have to be compensated. In this paper we present an image processing based quantitative evaluation system comprising: (i) semi-automatic non-rigid registration with a specific model for motion and deformation compensation, (ii) automatic detection and localization of the imaged structures, (iii) quantitative evaluation and statistical assessment of their proximity. We applied our approach to analyse the binding behaviour (indicated by co-localization) of telomeric DNA and TRF 1, a protein important for the regulation of the cellular lifespan. Our evaluation on real data shows that our motion model reproduces reliably the global movements of the nucleus and that protein and telomere co-localization could be identified by our pipeline where this was not possible before