An open-source platform to quantify subnuclear foci and protein colocalization in response to replication stress

Nuclear reorganization, including the localization of proteins into discrete subnuclear foci, is a hallmark of the cellular response to DNA damage and replication stress. These foci are thought to represent transient environments or repair factories, in which the lesion is sequestered with molecules and co-factors that catalyze repair. For example, nuclear foci contain signaling proteins that recruit transducer proteins. One important class of transducers is the structure-selective endonucleases, such as SLX1-SLX4, MUS81-EME1, and XPF-ERCC1, which remove branched DNA structures that form during repair. The relocalization of structure-selective endonucleases into subnuclear foci provides a visual read-out for the presence of direct DNA damage, replication barriers, or DNA entanglements and can be monitored using fluorescence microscopy. By simultaneously probing for two or more fluorescent signals, fluorescence microscopy can also provide insights into the proximal association of proteins within a local environment. Here, we report an open-source and semi-automated method to detect and quantify subnuclear foci, as well as foci colocalization and the accompanying pixel-based colocalization metrics. We use this pipeline to show that pre-mitotic nuclei contain a basal threshold of foci marked by SLX1-SLX4, MUS81, or XPF. Some of these foci colocalize with FANCD2 and have a high degree of correlation and co-occurrence. We also show that pre-mitotic cells experiencing replication stress contain elevated levels of foci containing SLX1-SLX4 or XPF, but not MUS81. These results point towards a role for SLX1-SLX4 and XPF-ERCC1 in the early cellular response to replication stress. Nevertheless, most of the foci that form in response to replication stress contain either FANCD2 or one of the three endonucleases. Altogether, our work highlights the compositional heterogeneity of subnuclear foci that form in response to replication stress. We also describe a user-friendly pipeline that can be used to characterize these dynamic structures.