Megabase-scale loss of heterozygosity provoked by CRISPR-Cas9 DNA double-strand breaks.

Harnessing DNA double-strand breaks (DSBs) is a powerful approach for gene editing, but it may provoke loss of heterozygosity (LOH), which predisposes to tumorigenesis. To interrogate this risk, we developed a two- color flow cytometry-based system (Flo-LOH), detecting LOH in ∼5% of cells following a DSB. After this initial increase, cells with LOH decrease due to a competitive disadvantage with parental cells, but if isolated, they stably propagate. Segmental loss from terminal deletions with de novo telomere addition and nonreciprocal translocations is observed as well as whole chromosome loss, especially following a centromeric DSB. LOH spans megabases distal from the DSB, but also frequently tens of megabases centromere-proximal. Inhibition of microhomology-mediated end joining massively increases LOH, which is synergistically increased with concomitant inhibition of canonical nonhomologous end joining. The capacity for large-scale LOH must therefore be considered when using DSB-based gene editing, especially in conjunction with end joining inhibition.

Noncanonical Contributions of MutLγ to VDE-Initiated Crossovers During Saccharomyces cerevisiae Meiosis.

In Saccharomyces cerevisiae, the meiosis-specific axis proteins Hop1 and Red1 are present nonuniformly across the genome. In a previous study, the meiosis-specific VMA1-derived endonuclease (VDE) was used to examine Spo11-independent recombination in a recombination reporter inserted in a Hop1/Red1-enriched region (HIS4) and in a Hop1/Red1-poor region (URA3). VDE-initiated crossovers at HIS4 were mostly dependent on Mlh3, a component of the MutLγ meiotic recombination intermediate resolvase, while VDE-initiated crossovers at URA3 were mostly Mlh3-independent. These differences were abolished in the absence of the chromosome axis remodeler Pch2, and crossovers at both loci became partly Mlh3-dependent. To test the generality of these observations, we examined inserts at six additional loci that differed in terms of Hop1/Red1 enrichment, chromosome size, and distance from centromeres and telomeres. All six loci behaved similarly to URA3: the vast majority of VDE-initiated crossovers were Mlh3-independent. This indicates that, counter to previous suggestions, levels of meiotic chromosome axis protein enrichment alone do not determine which recombination pathway gives rise to crossovers during VDE-initiated meiotic recombination. In pch2∆ mutants, the fraction of VDE-induced crossovers that were Mlh3-dependent increased to levels previously observed for Spo11-initiated crossovers in pch2∆, indicating that Pch2-dependent processes play an important role in controlling the balance between MutLγ-dependent and MutLγ-independent crossovers.

Local chromosome context is a major determinant of crossover pathway biochemistry during budding yeast meiosis.

The budding yeast genome contains regions where meiotic recombination initiates more frequently than in others. This pattern parallels enrichment for the meiotic chromosome axis proteins Hop1 and Red1. These proteins are important for Spo11-catalyzed double strand break formation; their contribution to crossover recombination remains undefined. Using the sequence-specific VMA1-derived endonuclease (VDE) to initiate recombination in meiosis, we show that chromosome structure influences the choice of proteins that resolve recombination intermediates to form crossovers. At a Hop1-enriched locus, most VDE-initiated crossovers, like most Spo11-initiated crossovers, required the meiosis-specific MutLγ resolvase. In contrast, at a locus with lower Hop1 occupancy, most VDE-initiated crossovers were MutLγ-independent. In pch2 mutants, the two loci displayed similar Hop1 occupancy levels, and VDE-induced crossovers were similarly MutLγ-dependent. We suggest that meiotic and mitotic recombination pathways coexist within meiotic cells, and that features of meiotic chromosome structure determine whether one or the other predominates in different regions.