Assessing the distribution of cancer stem cells in tumorspheres

In previous theoretical research, we inferred that cancer stem cells (CSCs), the cells that presumably drive tumor growth and resistance to conventional cancer treatments, are not uniformly distributed in the bulk of a tumorsphere. To confirm this theoretical prediction, we cultivated tumorspheres enriched in CSCs, and performed immunofluorecent detection of the stemness marker SOX2 using a confocal microscope. In this article, we present a method developed to process the images that reconstruct the amount and location of the CSCs in the tumorspheres. Its advantage is the use of a statistical criterion to classify the cells in stem and differentiated instead of setting an arbitrary threshold. From the analysis of the results of the methods using graph theory and computational modeling, we concluded that the distribution of Cancer Stem Cells in an experimental tumorsphere is non-homogeneous. This method is independent of the tumorsphere assay being useful for analyzing images in which several different kinds of cells are stained with different markers.