During tumorigenesis, the activation of oncogenes generates DNA replication stress, which leads to replication-associated genome rearrangements. This stress is not felt equally across the genome, but instead is concentrated at certain ‘difficult-to-replicate’ loci. The most prominent of these are multiple conserved loci called common fragile sites (CFSs). It was while studying CFSs that we identified an unusual DNA structure that had escaped detection previously despite being present in virtually all human cell anaphases; the ultra-fine anaphase bridge (UFB). UFBs are thin threads of DNA that connect the separating sister DNA masses in anaphase. They cannot be stained with DNA dyes and do not contain histones, making their detection problematic and dependent on immunofluorescence for associated proteins such as PICH or BLM. UFBs arise from specific loci that are characterized by the unusual structure and replication program; most notably, centromeres, CFSs, the rDNA, telomeres. We are conducting a detailed analysis of two aspects of UFB biology: modeling the association of proteins to UFBs in vitro using optical tweezers, and investigating how unresolved UFBs affect cell division. In addition, through analysis of the underlying basis of UFB formation in human cells, we identified that CFSs and telomeres delay the completion of their replication program until early mitosis - in a process that we term MiDAS (mitotic DNA synthesis). MiDAS depends upon a subset of homologous recombination factors, such as RAD52, and appears to be a form of break-induced replication previously characterized in detail only in yeast. Inhibition of MiDAS leads to major chromosome segregation abnormalities. The latest progress on these projects will be presented.