The MRE11-RAD50-NBS1 complex both starts and extends DNA end resection in mouse meiosis

Nucleolytic resection of DNA ends is critical for homologous recombination, but its mechanism is not fully understood, particularly in mammalian meiosis. Here we examine roles of the conserved MRN complex (MRE11, RAD50, and NBS1) through genome-wide analysis of meiotic resection in mice with various MRN mutations, including several that cause chromosomal instability in humans. Meiotic DSBs form at elevated levels but remain unresected if Mre11 is conditionally deleted, thus MRN is required for both resection initiation and regulation of DSB numbers. Resection lengths are reduced to varying degrees in MRN hypomorphs or if MRE11 nuclease activity is attenuated in a conditional nuclease-dead Mre11 model. These findings unexpectedly establish that MRN is needed for longer-range extension of resection, not just resection initiation. Finally, resection defects are additively worsened by combining MRN and Exo1 mutations, and mice that are unable to initiate resection or have greatly curtailed resection lengths experience catastrophic spermatogenic failure. Our results elucidate multiple functions of MRN in meiotic recombination, uncover unanticipated relationships between short- and long-range resection, and establish the importance of resection for mammalian meiosis.

dashed lines align the different elements of the cartoons by the 3¢ end of the DSB ssDNA tail.SPO11 cuts DNA with a 2-nt 5¢ overhang and stays covalently bound to the overhang until released from DSB ends by resection.The existence of a SPO11-bound (capped) recombination intermediate was proposed previously 14,15 .During library preparation, deproteination by proteinase K removes the DNA-bound SPO11 from capped ends, leaving a small peptide adduct (dashed magenta ellipses).(i) The endonuclease activity of S1 cleaves ssDNA from resected and capped breaks in addition to D-loop recombination intermediates (pink arrowheads) and generates duplex ends suitable for sequencing adaptor ligation (pale green circle).During later steps of library preparation, the short duplex oligonucleotides produced from capped ends (as well as from D-loops) are not recovered 15 .Therefore, S1-seq primarily generates reads from resection endpoints, which are distributed approximately 1 kb away from the center of hotspots with the expected polarity.In addition, S1-seq generates reads from recombination intermediates near the hotspot center (referred to here as the central signal) that have opposite polarity.(ii) Exonuclease VII is a ssDNA specific exonuclease that degrades overhangs in either the 3'-to-5' or 5'-to-3' direction from a DNA terminus (green arrowheads), producing 4-10 nt oligonucleotides 115,116 .Exonuclease VII can also remove peptide blocks such as SPO11 or aborted topoisomerase II from DNA ends and can generate blunt DNA ends when combined with exonuclease T in the END-seq procedure 14,117 .Neither exonuclease VII nor exonuclease T is expected to be able to digest D-loops or similar structures, so the central signals from Exo7/T-seq are thought to be instead from SPO11-oligo-capped structures.Therefore, Exo7/T-seq gives a weaker central signal that is more congruent with DSB positions defined by SPO11 oligos.(A) Schematic illustration of the originally anticipated DSB resection defects in Mre11-cHN.If the wild-type MRE11 protein is fully depleted by the time DSBs are formed and MRE11 is the only endonuclease that can initiate resection, then DSBs should remain unresected in Mre11-cHN (left).Alternatively, if residual wild-type MRE11 protein is retained for long enough after Cre-mediated excision to initiate resection, and if MRE11 is only involved in that initiation step, then any DSBs that are resected will be processed to the normal extent (right).Counter to these expectations, however, Mre11-cHN mice showed only a small apparent fraction of unresected DSBs plus a large population of resected DSBs with processing lengths that were much shorter than normal.(E) Distinct patterns at hotspot centers for autosomal vs. sex chromosome hotspots depending on whether DSB resection is initiated.Left, S1-seq signals from autosomal and sex chromosome hotspots in wild type (14.5 dpp).In these mice, all DSBs are resected and the central signal (which is only from recombination intermediates) is seen only on autosomes.Right, S1-seq and Exo7/T-seq signals at hotspot centers in Mre11-cKO (5 wk old).In these mice, essentially all of the DSBs remain unresected and the central signal (which is now from unresected DSBs) is seen on both autosomes and sex chromosomes.Plots were generated as described in panel D, except that S1-seq signals from 14.5-dpp Spo11 -/-mice were used for background subtraction of the S1-seq profiles.(F) Heatmaps (data in 40-bp bins) of MRE11 ChIP-seq signals around DSB hotspots.Each line is a hotspot, ranked from strongest at the top (based on SPO11-oligo read count).The ChIPseq signal at each hotspot was locally normalized by dividing by the total signal in a 4001-bp window around that hotspot's center.Each hotspot thus has a total value of 1, so that spatial patterns can be compared between hotspots of different strengths.Note that the central enrichment of MRE11 ChIP-seq signal is more pronounced the stronger the hotspot is.(G) Good agreement between this study and previous data 14 for MRE11 ChIP-seq coverage around hotspots in Atm -/-mice.(K-N) Premeiotic depletion of spermatogonia in Ctip conditional deletion mice (Ctip-cKO).CtIP is an essential cofactor that promotes MRN nuclease activity 3 and its yeast orthologs (Sae2 in S. cerevisiae and Ctp1 in S. pombe) are needed for meiotic DSB resection 4,5,92,93,118 .Because Ctip is essential for viability, we attempted to evaluate its function in meiosis by combining a floxed Ctip allele 101,102 with Ngn3-Cre.
(K) Drastic depletion of CtIP protein in immunoblots of whole-testis extracts from Ctip-cKO mice (5 wk old).The faster migrating band in samples from Ctip-cHet (Ctip wt/flox Ngn3-Cre + ) or Ctip-cKO (Ctip flox/-Ngn3-Cre + ) may have originated from an internal translation start site on the transcript from the deleted allele; if so, the truncated form of CtIP appears to be stable only in the presence of the full-length protein.
(L) Reduced testis size in Ctip-cKO (5-8 wk old).Error bars indicate mean ± SD.The P value is from a Student's t test.(M) Germ cell depletion in Ctip-cKO.Left, Bouin's fixed, H&E-stained seminiferous tubule sections (7 wk old).Right, PFA-fixed sections stained for DDX4 to mark germ cells.Note that most of the tubules had few or no DDX4-positive cells, unlike in Mre11-cKO (compare with Figure S3A).We conclude that CtIP is essential for germ cell maintenance in the mouse testis.(N) Average Exo7/T-seq profiles at hotspots.Data are smoothed with a 151-bp Hann window.We observed only extremely weak Exo7/T-seq signals with resection tracts similar to wild type.We infer that this residual signal comes from a small fraction of phenotypically normal germ cells that escaped Cre-mediated excision.The non-normalized wild-type profile is reproduced from panel C. (O) Normal resection in mice homozygous for Rad50S (K22M mutation; 14.5 dpp).Averaged S1-seq profiles around hotspots are shown.The Rad50S samples are indistinguishable from wild type, consistent with a previous report that this mutant has unperturbed meiotic progression 21 .Wild type is reproduced from Figure 2A.Data are smoothed with a 151-bp Hann window and normalized to the peak height of resection endpoints as described in Figure 4D.
following a 1-Gy IR exposure in wild-type cells, but was only detected at sharply reduced levels 30 min after an even higher dose (3 Gy) in the mutant.Note that levels of RAD50, MRE11, and NBS1 proteins were normal in the mutant, indicating that the mutation does not destabilize RAD50 or interfere with formation of the MRN complex.(B) Reduced G2/M cell cycle checkpoint in Rad50-D69Y MEFs.After 1 hr recovery following irradiation with 3 Gy of IR, mitotic cells were quantified by measuring mitosis-specific phosphorylation of histone H3 Ser10 by flow cytometry.Pretreatment with an ATM inhibitor (KU55933, 10 μM) before irradiation served as an Atm-deficient control 119 .The mitotic indices of two independent Rad50-D69Y MEF lines were twofold higher than those of wild-type cells, consistent with reduced ATM activity.Error bars indicate means ± SD of three replicates.(C) Increased DNA damage sensitivity in Rad50-D69Y MEFs when ATR is inhibited.Wild-type and Rad50-D69Y MEFs were treated with increasing doses of camptothecin (CPT) with and without concomitant inhibition of ATR (ATRi) with 50 nM VE822.In the absence of ATRi, the mutant showed no increase in CPT sensitivity compared to wild type.Wild-type cells were modestly more sensitive to CPT when ATR was inhibited than without ATRi, but the Rad50-D69Y mutant cells were substantially more CPT-hypersensitive in the presence of ATRi.These findings are consistent with the mutant cells being less able to compensate for ATR loss because they fail to activate ATM normally.(D) S1-seq signals from autosomal and sex chromosome hotspots, from 14.5dpp mice, plotted as described in  S1

Supplemental Table
(E) Schematics illustrating presumed cleavage by nuclease S1 or exonuclease VII of unresected DSBs.Panels D and E are adapted from 32 under a CC-BY license.Supplementary Figure 2. Further analysis of resection defects in MRE11-depleted spermatocytes (A) The central signal from recombination intermediates in wild type is distinct from the signal from unresected DSBs in Mre11-cKO.The genome-wide average of strand-specific S1-seq around hotspot centers in wild type is shown here, for comparison with the Mre11-cKO pattern in Figure 2D.(B) Greatly elevated amounts of SPO11-oligo complexes in Mre11-cKO.A biological replicate is shown for the experiment in Figure 2E.(C) MRN protein levels in juvenile Mre11-cKO mice.Immunoblots of whole-testis extracts from 14.5-dpp mice are shown as in Figure 1C.(D) S1-seq signal at a representative DSB hotspot in 14.5-dpp Mre11-cKO mice (same hotspot as in Figure 1F).Wild type is reproduced from Figure 1F to facilitate comparison.(E) Heatmaps of S1-seq signals around DSB hotpots from 14.5-dpp Mre11-cKO mice.Plots were generated as described in Figure 1G.(F) Global average S1-seq profiles around hotspots from 14.5-dpp mice.Note the mix of unresected and resected DSBs in juvenile Mre11-cKO mice.We report modal resection lengths here instead of the mean values reported elsewhere because the high signal from unresected DSBs in Mre11-cKO distorts the estimate of resection length.Wild type is reproduced from Figure 2A; three biological replicates for wild type and two for Mre11-cKO were averaged.Data are smoothed with a 151-bp Hann window.(B) PFA-fixed seminiferous tubule sections at 7 wk of age stained with TUNEL to show apoptotic cells.(C) Frequencies of tubules with the indicated number of TUNEL-positive cells.Error bars indicate mean ± range for the indicated number of animals.(D) Bouin's fixed seminiferous tubule and epididymis sections stained with H&E (first three columns) and PFA-fixed seminiferous tubule sections stained for DDX4 (right column).Animals were 16 wk old.(E) Representative spermatocyte spreads stained for SYCP3 and SYCP1.
(B) Heatmaps of Exo7/T-seq signals around DSB hotpots from 5-wk-old mice.Two biological replicates each for wild type and Mre11-cHN were averaged.(C) The non-normalized plot of Figure 4D.(D) Exo7/T-seq signals from autosomal and sex chromosome hotspots.The profiles were generated by subtracting the Exo7/T-seq signal obtained in the 14.5-dpp Spo11 -/-mutant and then averaging top and bottom strand reads after co-orienting them around hotspot centers.Data are smoothed with a 151-bp Hann window.The bottom panels show zoomed views (smoothed with a 51-bp Hann window) into the region around hotspot centers.
(H) Reproducibility of resection profiles across a range of young adult ages for Mre11-cHN mice.Data are smoothed with a 151-bp Hann window and normalized to the peak height of resection endpoints.Resection profiles of wild type and 5 wk old Mre11 HN-het or Mre11-cHN are reproduced from Figure 4D.The profiles for 7, 9, and 11 wk old HN-cHN were each from a single library.(I) Biological replicate of the experiment in Figure 4F, showing that both the reduction in the number of SPO11-oligo complexes and their altered electrophoretic mobility are reproducible in Mre11-cHN mice.(J) Comparison of Exo7/T-seq signal strengths for wild type (n=2; 5 wk old) versus Mre11 HNhet (n=1; 5 wk old) or Mre11-cHN (n=5; 5-11 wk old) quantified as in Figure 2B.The P value is from a two-tailed Student's t test.

. Summary of SIM imaging Nbs1 ∆B/+ Nbs1 ∆B/∆B
*Co-foci were defined as axis-associated foci that were within 320 nm of a focus of the other protein. *