Repair at single targeted DNA double-strand breaks in pluripotent and differentiated human cells.

Differences in ex vivo cell culture conditions can drastically affect stem cell physiology. We sought to establish an assay for measuring the effects of chemical, environmental, and genetic manipulations on the precision of repair at a single DNA double-strand break (DSB) in pluripotent and somatic human cells. DSBs in mammalian cells are primarily repaired by either homologous recombination (HR) or nonhomologous end-joining (NHEJ). For the most part, previous studies of DSB repair in human cells have utilized nonspecific clastogens like ionizing radiation, which are highly nonphysiologic, or assayed repair at randomly integrated reporters. Measuring repair after random integration is potentially confounded by locus-specific effects on the efficiency and precision of repair. We show that the frequency of HR at a single DSB differs up to 20-fold between otherwise isogenic human embryonic stem cells (hESCs) based on the site of the DSB within the genome. To overcome locus-specific effects on DSB repair, we used zinc finger nucleases to efficiently target a DSB repair reporter to a safe-harbor locus in hESCs and a panel of somatic human cell lines. We demonstrate that repair at a targeted DSB is highly precise in hESCs, compared to either the somatic human cells or murine embryonic stem cells. Differentiation of hESCs harboring the targeted reporter into astrocytes reduces both the efficiency and precision of repair. Thus, the phenotype of repair at a single DSB can differ based on either the site of damage within the genome or the stage of cellular differentiation. Our approach to single DSB analysis has broad utility for defining the effects of genetic and environmental modifications on repair precision in pluripotent cells and their differentiated progeny.