Meiotic chromosomes are organized into arrays of loops that are anchored to the chromosome axis structure. Programmed DNA double-strand breaks (DSBs) that initiate meiotic recombination, catalyzed by Spo11 and accessory DSB proteins, form in loop sequences in promoters, whereas the DSB proteins are located on chromosome axes. Mechanisms bridging these two chromosomal regions for DSB formation have remained elusive. Here we show that Spp1, a conserved member of the histone H3K4 methyltransferase Set1 complex, is required for normal levels of DSB formation and is associated with chromosome axes during meiosis, where it physically interacts with the Mer2 DSB protein. The PHD finger module of Spp1, which reads H3K4 methylation close to promoters, promotes DSB formation by tethering these regions to chromosome axes and activating cleavage by the DSB proteins. This paper provides the molecular mechanism linking DSB sequences to chromosome axes and explains why H3K4 methylation is important for meiotic recombination.

During the first meiotic prophase, programmed DNA double-strand breaks (DSBs) are distributed non randomly at hotspots along chromosomes, to initiate recombination. In all organisms, more DSBs are formed than crossovers (CO), the repair product that creates a physical link between homologs and allows their correct segregation. It is not known whether all DSB hotspots are also CO hotspots or if the CO/DSB ratio varies with the chromosomal location. Here, we investigated the variations in the CO/DSB ratio by mapping genome-wide the binding sites of the Zip3 protein during budding yeast meiosis. We show that Zip3 associates with DSB sites that are engaged in repair by CO, and Zip3 enrichment at DSBs reflects the DSB tendency to be repaired by CO. Moreover, the relative amount of Zip3 per DSB varies with the chromosomal location, and specific chromosomal features are associated with high or low Zip3 per DSB. This work shows that DSB hotspots are not necessarily CO hotspots and suggests that different categories of DSB sites may fulfill different functions.

Human cytomegalovirus (HCMV) primary infections of pregnant women can lead to congenital infections of the fetus that could have severe impacts on the health of the newborn. Recent studies have shown that 10–100 billion DNA fragments per milliliter of plasma are circulating cell-free. The study of this DNA has rapidly expanding applications to non-invasive prenatal testing (NIPT). In this study, we have shown that we can detect viral specific reads in the massively parallel shotgun sequencing (MPSS) NIPT data. We have also observed a strong correlation between the viral load of calibration samples and the number of reads aligned on the reference genome. Based on these observations we have constructed a statistical model able to quantify the viral load of patient samples. We propose to use this new method to detect and quantify circulating DNA virus like HCMV during pregnancy using the same sequencing results as NIPT data. This method could be used to improve the NIPT diagnosis.