Quantification of telomere features in tumor tissue sections by an automated 3D imaging-based workflow
Authors: Gunkel M, Chung I, Wörz S, Deeg KI, Simon R, Sauter G, Jones DT, Korshunov A, Rohr K, Erfle H, Rippe K
CellNetworks People: Erfle Holger, Rippe Karsten, Rohr Karl
Journal: Methods. 2017 Feb 1;114:60-73. doi: 10.1016/j.ymeth.2016.09.014

The microscopic analysis of telomere features provides a wealth of information on the mechanism by which tumor cells maintain their unlimited proliferative potential. Accordingly, the analysis of telomeres in tissue sections of patient tumor samples can be exploited to obtain diagnostic information and to define tumor subgroups. In many instances, however, analysis of the image data is conducted by manual inspection of 2D images at relatively low resolution for only a small part of the sample. As the telomere feature signal distribution is frequently heterogeneous, this approach is prone to a biased selection of the information present in the image and lacks subcellular details. Here we address these issues by using an automated high-resolution imaging and analysis workflow that quantifies individual telomere features on tissue sections for a large number of cells. The approach is particularly suited to assess telomere heterogeneity and low abundant cellular subpopulations with distinct telomere characteristics in a reproducible manner. It comprises the integration of multi-color fluorescence in situ hybridization, immunofluorescence and DNA staining with targeted automated 3D fluorescence microscopy and image analysis. We apply our method to telomeres in glioblastoma and prostate cancer samples, and describe how the imaging data can be used to derive statistically reliable information on telomere length distribution or colocalization with PML nuclear bodies. We anticipate that relating this approach to clinical outcome data will prove to be valuable for pretherapeutic patient stratification.