lncRNA transcription induces meiotic recombination through chromatin remodelling in fission yeast

Noncoding RNAs (ncRNAs) are involved in various biological processes, including gene expression, development, and disease. Here, we identify a novel consensus sequence of a cis-element involved in long ncRNA (lncRNA) transcription and demonstrate that lncRNA transcription from this cis-element activates meiotic recombination via chromatin remodeling. In the fission yeast fbp1 gene, glucose starvation induces a series of promoter-associated lncRNAs, referred to as ﻿metabolic-stress-induced lncRNAs (mlonRNAs), which contribute to chromatin remodeling and fbp1 activation. Translocation of the cis-element required for mlonRNA into a well-characterized meiotic recombination hotspot, ade6-M26, further stimulates transcription and meiotic recombination via local chromatin remodeling. The consensus sequence of this cis-element (mlon-box) overlaps with meiotic recombination sites in the fission yeast genome. At one such site, the SPBC24C6.09c upstream region, meiotic double-strand break (DSB) formation is induced in an mlon-box-dependent manner. Therefore, mlonRNA transcription plays a universal role in chromatin remodeling and the regulation of transcription and recombination. Senmatsu et al. show that transcription of long-noncoding RNAs (lncRNA) previously identified in the fission yeast fbp1 gene and found to induce chromatin opening, also activates meiotic recombination by inducing local chromatin remodeling at a meiotic recombination hotspot. This study furthers our understanding of the connection between lncRNA transcription and DNA break activity in yeast meiosis.

In meiosis, homologous recombination between homologous chromosomes is pivotal for meiotic progression and contributes to proper segregation of homologous chromosomes during gametogenesis [14][15][16][17] . This process is initiated by the formation of doublestrand breaks (DSBs) catalyzed by the Spo11 protein 18 . Distribution of meiotically induced DSBs is not uniform but is clustered at "hotspots" where meiotic recombination is frequently induced 19,20 . The global distribution of DSB has been extensively studied in the highly diverged yeasts Saccharomyces cerevisiae (budding yeast) and Schizosaccharomyces pombe (fission yeast) [21][22][23][24] . High resolution mapping of DSB sites in budding yeast has provided insights into DSB sites and hierarchical context, such as chromosome structures, chromatin, transcription factors, and local sequence composition 25 . Most DSB hotspots are located in promoter regions for mRNA 26 or ncRNA 13 , which are characterized by transcription factor binding, histone modifications and low nucleosome density, suggesting a close relationship between meiotic recombination and transcription. In fission yeast, the ade6-M26 hotspot, which contains a G-to-T substitution in the ade6 open reading frame (ORF) has been intensively investigated as a model locus of meiotic recombination hotspot 27,28 . The M26-mutation is a nonsense mutation creating a cyclic adenosine monophosphate (cAMP)-responsive element (CRE)-like heptanucleotide sequence 5′-ATGACGT-3′, to which the transcription factors Atf1-Pcr1 bind and activate meiotic recombination [29][30][31][32] . Atf1-Pcr1 binding induces M26 transcription from the M26-mutation site in the ade6 ORF, followed by histone acetylation and chromatin remodeling around the M26 site, thereby inducing meiotic recombination [31][32][33] . Taken together, these observations suggest that local chromatin configuration, histone modifications and transcriptional activity play key roles in the location of meiotic DSB sites in budding yeast and fission yeast 13,[32][33][34][35][36] . However, mechanisms underlying the determinants of this DSB distribution have not been fully elucidated.
Previously, we found a stepwise transcription of lncRNAs in the upstream region of fission yeast fbp1 gene 37 . The fbp1 gene is activated upon glucose starvation 38 , and this activation is mediated by two transcription factors, Atf1 and Rst2 39,40 . These transcription factors bind to critical cis-acting binding sequences located upstream of the fbp1 promoter, (upstream activating sequences 1 [UAS1] and 2 [UAS2]) 39 (Fig. 1a). During glucose starvation stress, several species of lncRNAs, referred to as metabolic stress-induced lncRNAs (mlonRNAs), are transcribed from the upstream region of the fbp1 promoter 37,41 .
Stepwise mlonRNA transcription (mlonRNA-a, mlonRNA-b, and mlonRNA-c initiating progressively closer to the fbp1 transcription start site [TSS]) induces chromatin remodeling at the fbp1 promoter (Fig. 1a), and subsequently facilitates transcription factor binding to activate fbp1 expression 37 . These observations indicate that mlonRNA transcription mediates chromatin remodeling and thereby contribute to the robust induction of fbp1 transcription upon glucose starvation. However, whether or not the role played by this lncRNA transcription is limited to the fbp1 locus has not been addressed. In this study, we investigated whether mlonRNA transcription plays a universal role in chromatin remodeling in the fission yeast genome, leading to the discovery that this lncRNA transcription induces meiotic recombination through inducing chromatin remodeling.
Results mlonRNA initiation element (mlon-IE) works at other loci and conditions in addition to its role in the fbp1 upstream region. We previously identified lncRNAs expressed in the fission yeast fbp1 upstream region in the process of massive transcriptional activation, naming these RNAs mlonRNA-a, mlonRNA-b, and mlonRNA-c (Fig. 1b). These transcripts initiate from far upstream of the fbp1 promoter and the initiation site progressively shifts closer to the fbp1 TSS (Fig. 1a, b). The chromatin configuration was analyzed by the distribution of micrococcal nuclease (MNase) sensitive site in fbp1 upstream region (Fig. 1b). MNase-sensitive sites appear 10 min after glucose starvation (Fig. 1b, arrowheads), coincident with the appearance of mlonRNA-b transcription. Another MNase-sensitive region appears around UAS1-UAS2 at 20 to 30 min after glucose starvation (Fig. 1b, dashed lines), when mlonRNA-c transcription is first observed. Finally, intense MNase-sensitive sites appear around the TATA box at 60 to 180 min after glucose starvation (Fig. 1b, thick lines), when massive fbp1 mRNA transcription occurs. Our previous study including these observations indicated that mlonRNA transcription induces chromatin remodeling in the transcribed tract in fbp1 upstream during the transcriptional activation processes 37 . We further identified the cis-element (mlonRNA initiation element [mlon-IE]) (5′-ATCTTATGTA-3′) required for initiation of mlonRNA-c transcription 42 . To investigate the generality of the role of mlonRNA transcription in chromatin modulation and to further assess the role of this lncRNA transcription in other aspects of chromosomal regulation, we inserted this cis-element into the ade6-M26 meiotic recombination hotspot (Fig. 1c) 27,28 . The M26-mutation is a nonsense mutation creating a CRE-like Atf1-Pcr1 bind site [29][30][31][32] . Since mlon-IE is located 200 bp downstream from UAS1 comprising an Atf1-Pcr1 binding sequence in fbp1, and Atf1 is required for mlonRNA-c expression 37 , the 65 bp fbp1 upstream sequence containing mlon-IE and a flanking sequence was placed 200 bp downstream from the M26 mutation site by replacing the same length of the ade6 ORF sequence (M26::mlon-IE cells). During meiosis, an additional shorter transcript is induced in M26::mlon-IE cells, but it does not appear in the M26::mlon-IEmut control carrying mutated mlon-IE (Fig. 1c, d). The TSS of this shorter transcript is +426 from the ade6 ORF ATG, located 94 bp downstream of the mlon-IE core 10 bp ( Supplementary  Fig. 1a). These findings show that mlon-IE works in other loci and conditions, in addition to its role in the fbp1 locus under glucose starvation stress.
We also translocated the cis-element into a control mutant allele of ade6, ade6-M375, which has a similar G-to-T substitution without creating a CRE-like sequence (Fig. 1c). M375::mlon-IE cells do not show transcription from mlon-IE (Fig. 1d). Similar to what is observed during meiosis, osmotic stress induces a shorter transcript downstream of the mlon-IE insertion site in M26:: mlon-IE cells but not in M26::mlon-IE-mut or M375::mlon-IE cells ( Supplementary Fig. 1b, c). These findings suggest that mlon-IE can also work during osmotic stress, and depends on the binding of the Atf1-Pcr1 transcription factors upstream of mlon-IE. This result is consistent with what is observed in the mlonRNA-c initiation during fbp1 activation, which is dependent on UAS1 comprising Atf1-Pcr1 binding sequence 37 .
mlon-IE-induced transcription activates meiotic recombination via inducing local chromatin remodeling. We next examined whether transcription from mlon-IE activates meiotic recombination. Briefly, we measured the recombination rate in the ade6 locus using tester strain ade6-469, which has a nonsense mutation in the 3' region of the ade6 gene ( Supplementary Fig. 2a) 32 . M26::mlon-IE cells and parental ade6-M26 cells exhibit a similar number of recombinant Ade + colonies in which a functional ade6 + gene is reconstituted by homologous recombination (Fig. 2a). Unfortunately, we cannot directly compare the number of Ade + recombinants between ade6-M26 and M26::mlon-IE cells for the following reasons. Insertion of mlon-IE drastically changes the amino acid sequence of the Ade6 protein and thereby disrupts the protein function ( Supplementary  Fig. 2b). This insertion makes the distances between M26::mlon-IE/469 shorter than that between M26/469 ( Supplementary  Fig. 2a). Moreover, the M26::mlon-IE allele has multiple mutations that would require longer recombination tracts for the reconstitution of a functional ade6 + gene than needed for the ade6-M26 allele. However, M26::mlon-IE cells show a significantly higher meiotic recombination rate compared to M26:: mlon-IE-mut cells (Fig. 2a). In contrast, M375::mlon-IE and M375::mlon-IE-mut cells show a similar recombination rate (Fig. 2a). These findings indicate that mlon-IE induces mlonRNA transcription even at 200 bp downstream of the M26 mutation point in the ade6-M26 gene and that this expression may augment meiotic recombination in comparison to mutated mlon-IE inserted cells.
Considering the role mlon-IE-induced transcription plays in meiotic recombination, we examined meiotic DSBs, which are introduced by Rec12 (Spo11 ortholog in fission yeast) and initiate meiotic recombination 43 . In M26 and M26::mlon-IE-mut cells, we   Supplementary Fig. 3a, b). Remarkably, M26::mlon-IE cells show increased intensity of DSBs at the mlon-IE site, and the DSB distribution changes with marked bias toward the mlon-IE site ( Fig. 2b and Supplementary Fig. 3a, b). As genome-wide DSBs assessed by pulse-field gel electrophoresis (PFGE) in these strains are indistinguishable, global DSB formation activity is not affected by the insertion of mlon-IE ( Supplementary Fig. 4). To determine whether the effect of mlon-IE-induced transcription on DSB formation is due to transcription-coupled chromatin remodeling, we analyzed the chromatin configuration by MNase digestion as in Fig. 1b. In all cells carrying the M26 mutation, the chromatin at M26 is protected from MNase before meiosis ( mlon-IE works downstream of transcription factor-binding sequences. Similar to M26 sequence, other mutations at ade6 that create transcription factor-binding sequences, including CCAAT, oligo-C, 4095, and 4156 motifs, activate meiotic recombination in the ade6 locus 44 . Next, we tested whether mlon-IE works together with these transcription factor-binding sequences instead of the M26 sequence. In ade6-4002 (CCAAT-motif) and ade6-4099 (oligo-C-motif), hotspot activity depends on transcription factors, CCAAT-binding factor CBF (heterotrimer composed of Php2, Php3, and Php5), and C 2 H 2 zinc finger protein (Rst2), respectively 45 . Regulators of ade6-4095 and ade6-4156 hotspot activity are unclear, although the ade6-4156 sequence has the potential to Recombination between leu1-32 and his3-D1 (indicated as leu + /his + percentage) was measured as the control (right). Error bars represent standard deviations. n = 3 or 5 biologically independent experiments. P values were calculated using the unpaired one-sided Student's t-test: *P < 0.05, n.s. (not significant). b Indicated haploid cells possessing the pat1-114 rad50s genotype were cultured to induce meiosis, and DNA was prepared as described in the Methods section. Meiotic DSBs were detected by Southern blotting. The dotted line indicates DSBs around M26 and inserted mlon-IE. c Meiotic chromatin remodeling in indicated cells. Diploid cells were cultured as in Fig. 1c. The black arrowhead and the dotted line indicate MNase-sensitive sites at the M26 mutation and the inserted mlon-IE, respectively. bind the Rds1 protein 19 . We inserted mlon-IE 200 bp downstream of ade6-4002, 4099, 4095, and 4156 sequences (Fig. 3a). These sequences induce a shorter transcript from inside the ade6 ORF, similarly to that seen in ade6-M26 cells (Fig. 3b). More importantly, mlon-IE inserted cells (4002::mlon-IE, 4099::mlon-IE, 4095::mlon-IE, and 4156::mlon-IE cells) show an additional shorter transcript (Fig. 3b), indicating that mlon-IE also works in the presence of other transcription factor-binding site, in addition to Atf1-Pcr1 binding site. We observed the additional shorter transcript in 4002::mlon-IE or 4099::mlon-IE cells only in meiosis (2 h), while that in 4095::mlon-IE or 4156::mlon-IE cells was strongly transcribed before the onset of meiosis (0 h; Fig. 3b). This difference might be due to the different binding pattern of transcription factors to the mutated ade6 loci (4002, 4099, 4095, and 4156). i.e., 4002 and 4099 are bound by transcription factors in a meiosis-specific manner, while 4095 and 4156 are constitutively bound by some transcription factors. Similar to M26:: mlon-IE cells, additional transcriptional initiation from inserted mlon-IE activates recombination in ade6-4002 and ade6-4099 (Fig. 3c). In addition, the DSB distribution changes with a bias toward the mlon-IE-inserted site in 4002::mlon-IE and 4099:: mlon-IE cells (Fig. 3d, e and Supplementary Fig. 5a-d). Again, DSB activity on a genome-wide scale is not affected by the insertion of mlon-IE in these strains ( Supplementary Fig. 4).
Natural mlonRNA transcription from an intergenic region activates meiotic DSB formation via inducing local chromatin remodeling. Having established the role of mlonRNA transcription from mlon-IE in the activation of meiotic recombination in the ade6 locus, we next asked if mlonRNA transcription plays a role in the regulation of meiotic recombination through chromatin modulation in the S. pombe genome. To this end, we determined the consensus sequence of mlon-IE by comprehensively mutating each of the 10 nucleotides existing in mlon-IE (Fig. 4a). By excluding mutations that decrease mlonRNA transcription <50% (Fig. 4b and Supplementary Fig. 6), we defined a consensus sequence for mlon-IE as 5′-N A/T A/C T T/G A T/C/G G T A/G-3′, and refer to this consensus element as the mlon-box (Fig. 4c). Next, we searched for mlon-box in the S. pombe genome and analyzed the positional relationship between mlon-box sites and meiotic recombination sites by using a previously reported Rec12 covalently associated oligonucleotide sequence for identifying meiotic DSB sites 24 (Fig. 5a). This analysis demonstrates that meiotic recombination tends to occur near mlon-box sequences. The strongest recombination site that overlapped with the mlon-box sequence is located in the intergenic region between SPBC24C6.09c and SPNCRNA.1506 (Fig. 5b). In this region, several meiotic DSBs are observed near the mlon-box (Fig. 5c). We inactivated the mlon-box by replacing this sequence with a 10 nucleotide sequence from act1 ORF and to generate mlon-box-replacement cells (Fig. 5b) 42 . The act1 ORF sequence is here used as neutral sequence carrying no DSB activity. In these mutant cells, DSBs closest to the mlon-box site are selectively lost (Fig. 5c dotted line, Supplementary Fig. 7a), indicating that the natural mlon-box induces meiotic DSB formation. In addition, SPBC24C6.09c transcription, but not SPNCRNA.1506 transcription, is activated early after the onset of meiosis in an mlon-box-dependent manner (Fig. 5d and Supplementary Fig. 7b). Also, an MNase-sensitive site appears at the mlon-box, and the pattern of MNase-sensitive bands surrounding the mlon-box changes in 0.5 and 1 h after the onset of meiosis ( Supplementary Fig. 7c). These results indicate that the chromatin configuration around the mlon-box changes during meiosis ( Supplementary Fig. S7c arrowhead and arrows) and that these changes are dependent on the presence of the mlon-box ( Supplementary Fig. 7c dotted line). Consistent with these observations, the amount of histone H3 binding at the mlon-box decreases in meiosis, and this histone eviction depends on the presence of the mlon-box (Fig. 5e). These results indicate that transcriptional initiation from the natural mlon-box in the intergenic region activates meiotic DSB formation via chromatin opening.
Previous genome-wide transcriptome analysis has shown that Atf1 regulates SPBC24C6.09c transcription 46 . However, we did not find a canonical Atf1-binding motif (TGACGT) in ±1 kbp from SPBC24C6.09c-mlon-box. We noticed a similar palindromic Atf1binding-like motif (TTACGT: a single G-to-T substitution) in SPBC24C6.09c-mlon-box ( Supplementary Fig. 8a). In fact, Atf1 binding at SPBC24C6.09c-mlon-box increases approximately threefold during meiosis, while the replacement of the entire mlon-box (10 bp) results in a pronounced decrease in Atf1 binding ( Supplementary Fig. 8b). To selectively disrupt the mlon-box while preserving this Atf1-binding sequence, the first three nucleotides (ATC) in the mlon-box were replaced with CCT, generating mlonbox-3bp-replacement cells (Supplementary Fig. 8a). This strain harboring a mutated mlon-box shows little Atf1 binding ( Supplementary Fig. 8b). Consistent with this, the transcription level in mlon-box-3bp-replacement cells is significantly decreased compared to wild-type cells (Supplementary Fig. 8c). These results indicate the importance of the mlon-box in Atf1 binding and later transcriptional activation of SPBC24C6.09c. The SPBC24C6.09c-mlon-box might enhance Atf1 affinity for the Atf1-binding motif-like sequence. The possible mechanism underlying mlon-box-mediated strengthening of Atf1 binding is that mlonRNA molecules initiated from the mlon-box bind to the Groucho-Tup1-type transcriptional corepressors Tup11-Tup12 and thereby locally antagonize their repressive functions on Atf1 binding as described previously 47 . An mlonRNA transcription amplification circuit such as this might play a critical role in the massive induction of SPBC24C6.09c from an incomplete Atf1binding site.   Supplementary Fig. 6 was quantified. n = 2 biologically independent experiments. c mlon-box, a consensus sequence of mlon-IE.

Discussion
In this study, we identified a novel cis-element, the mlon-box, which is required for mlonRNA transcriptional initiation. mlonRNA transcription plays a critical role in the massive induction of fbp1 via chromatin opening. In addition, mlonbox-mediated transcription activates meiotic DSB formation and recombination in ade6-M26, ade6-4002, and ade6-4099 via chromatin opening. Such mlon-box transcription-associated meiotic DSB formation is observed in the natural meiotic recombination hotspot located in the SPBC24C6.09c upstream region.
It remains to be determined how mlonRNA transcription induces chromatin remodeling. Three mlonRNAs (mlonRNA-a to mlonRNA-c) are initiated far upstream from the fbp1 promoter and stepwise chromatin remodeling is induced in their transcribed tract 37 (Fig. 1). Each transcriptional initiation efficiently induces local chromatin remodeling only within 290 bp of the initiation site 42 , and thus three consecutive mlonRNA transcription events induce stepwise chromatin remodeling over the entire regulatory upstream region of fbp1, thereby allowing the binding of transcription factors as well as RNA polymerase II (RNAPII). An intriguing possibility is that the RNAPII that initiates mlonRNA transcription binds unique accessory subunit (s) that possess histone acetyltransferase activity to induce local histone acetylation around these transcription-start sites and dissociate from the initiation complex after promoter clearance 48 .
Such histone modifications might recruit an ATP-dependent chromatin remodeler [49][50][51] . This hypothesis is supported by the observation that histone acetylation is gradually induced from upstream of fbp1 during glucose starvation and a histone acetyltransferase, Gcn5, and an ATP-dependent chromatin remodeler, Snf22, are required for chromatin remodeling in the upstream from fbp1 37,52 . If such novel RNAPII complexes do exist, it will be important to learn how they are selectively targeted to sites of mlon-RNA transcription initiation.
Many lncRNAs are located in the promoter region of genes and referred to as promoter-associated RNAs 53 . mlon-box sequences are enriched upstream of TSSs in the S. pombe genome (Supplementary Fig. 9), suggesting that mlon-box-induced transcription might regulate neighboring gene expression. Consistent with this, several studies have also reported the contribution of ncRNA transcription in gene expression control 7,54,55 . In addition, meiotic recombination sites in fission yeast are directed preferentially to loci expressing ncRNA 13 . It is thus possible that these pervasive lncRNA expressions are involved in the DSB formation as well as with later stages of meiotic recombination via facilitating chromatin remodeling. Taken together, this current study suggests a model that numerous human lncRNAs identified in association with developmental processes 11,12 and diseases including cancer 10,56 , are traces of mlonRNA-mediated regulation of genome function. This study provides important insights into the function of widespread lncRNA transcription in the  (Rec12-oligos). The X axis indicates the distance from the nearest mlon-box, and the Y axis indicates copy numbers of the Rec12-oligos-signal. The strongest plot (arrowhead) represents the meiotic recombination hotspot in the SPBC24C6.09c upstream region. b Schematic representation of the mlon-box (open circle) in the intergenic region between SPBC24C6.09c and SPNCRNA.1506. The entire mlon-box sequence comprising 10 nucleotides was replaced with a 10 nucleotide sequence from act1 ORF in mlon-boxreplacement cells (bold italic). Location of probes for northern blot was indicated by bold bar. c Detection of meiotic DSBs in the intergenic region between SPBC24C6.09c and SPNCRNA.1506. Indicated cells were cultured and analyzed, as described in Fig. 2b. The dotted line indicates DSB sites around the mlonbox. d Northern analysis to detect the SPBC24C6.09c transcript in wild-type and mlon-box-replacement cells. The left probe indicated in Fig. 4b was used. Indicated cells were cultured to induce meiosis, as described in Fig. 2b. The cam1 transcript is shown as a loading control. e, ChIP analysis to examine histone H3 binding at the mlon-box in indicated cells. Cells were cultured to induce meiosis, as described in Fig. 2b. The relative increase in the ratio at the indicated time after the onset of meiosis is indicated. n = 2 biologically independent experiments. P-values were calculated using Student's t-test: *P < 0.05. regulation of DNA-associated reactions on chromatin geometry. A further understanding of the roles of lncRNA transcription might provide important clues to treating lncRNA-associated diseases.

Methods
Fission yeast strains, genetic methods, and cell culture. Supplementary  Tables 1 and 2 list the fission yeast strains and primers used in this study, respectively. YEL medium (0.5% yeast extract and 2% glucose), synthetic dextrose (SD) medium, and minimal medium (MM) were used for cell culture 57 . YER medium (yeast extract containing 6% glucose) and YED medium (yeast extract containing 0.1% glucose and 3% glycerol) were used for glucose repression and starvation, respectively. To induce meiosis, diploid cells were grown to a density of 0.5 × 10 7 cells/mL in MM medium supplemented with 5 mg/mL of NH 4 Cl and 1% YEL medium, then diluted with three times the amount of YEL medium, and further cultured for 4 h. Next, the cells were collected and washed with distilled water twice and transferred to MM medium lacking a nitrogen source. To detect meiotic DSBs, synchronous meiosis using a pat1-114 mutation was employed together with a rad50S mutation to efficiently detect DSBs 58 . The haploid pat1-114 rad50S cells were precultured in MM medium containing 5 mg/mL of NH 4 Cl at 25°C, washed with distilled water twice, transferred to MM medium lacking a nitrogen source for arresting the cell cycle at the G1 phase, and cultured further at 25°C for 16 h. G1-arrested cells were collected by centrifugation, suspended in prewarmed MM medium containing 0.5 mg/mL of NH 4 Cl and 1% YEL medium, and cultured at 34°C to induce meiosis. The transformation was performed using the lithium acetate method as follows 59  After centrifugation, the cell pellet was resuspended in distilled water and plated on an appropriate selective medium. Mating and sporulation were performed on sporulation agar (SPA) medium 57 , followed by random spore analysis.
Construction of mutant strains carrying mlon-IE 200 bp downstream of ade6-M26, ade6-M375, ade6-4002, ade6-4009, ade6-4095, and ade6-4156. The ade6 gene region carrying M26, M375, 4002, 4099, 4095, or 4156 mutations was amplified using primer sets (p1/p2 and p3/p4). p2 and p4 were flanked by the fbp1 upstream sequence comprising 65 nucleotides containing mlon-IE (−870 to −806 from the first ATG of fbp1) and its complementary sequence, respectively. The resultant products were purified using the QIAquick gel extraction kit (Qiagen, Germany). These fragment pairs were combined at the complementary sequence flanking in p2 and p4 by polymerase chain reaction (PCR) using the p1/p3 primer set and then cloned at ApaI and SpeI sites in the pBluescript plasmid. Fission yeast cells carrying the ura4 marker gene in the HindIII site in the ade6 ORF were transformed, and the transformants were selected for uracil auxotrophy using SD plates containing 5-fluoroorotic acid (5-FOA) and uracil.
Construction of strains with single-base substitution in mlon-IE upstream at fbp1. Comprehensive mlon-IE replacement strains were constructed by PCR amplification according to the method of Senmatsu et al. 42 . Briefly, fbp1 upstream region was amplified using primers p5/p6 to p35 and p36/p37 to p66. The resultant products were purified using the QIAquick gel extraction kit (Qiagen, Germany). Pairs of fragments were used as templates for PCR amplification using the primer set p5 and p36. Fission yeast cells carrying the ura4 maker gene in the HpaI site in the fbp1 promoter were transformed, and the transformants were selected for uracil auxotrophy using SD plates containing 5-fluoroorotic acid (5-FOA) and uracil.
ChIP analysis. Chromatin immunoprecipitation (ChIP) analyses were performed according to the method of Senmatsu et al. 42 with slight modifications as follows. Fifty ml of culture was incubated with 1.4 ml of 37% formaldehyde solution for 20 min at room temperature, and then 2.5 ml of 2.5 M glycine was added. After centrifugation, collected cells were washed twice with cold TBS buffer (150 mM NaCl, 20 mM Tris HCl [pH 7.5]). The cells were mixed with 400 µl of lysis 140 buffer (0.1% Na-deoxycholate, 1 mM EDTA, 50 mM HEPES-KOH [pH 7.5], 140 mM NaCl, 1% Triton X100) supplemented with protease inhibitor, cOmplete (Roche), and 0.6 ml of zirconia beads were added. After disruption of the cells using a multi beads shocker (Yasuikikai, Osaka), the suspension was sonicated six times for 30 s each on ice to shear chromosomal DNA into around 500 bp fragment, and centrifuged at 4°C. The supernatant was collected as a whole-cell extract. Two microliters of Anti-Histone H3 antibody (abcam) or 1 μl of Anti-Atf1 antibody (abcam), and 20 µl or 10 µl of Dynabeads Protein A (Invitrogen) were mixed at 4°C to conjugate antibody and beads, and then washed twice with PBS (138 mM NaCl, 2.7 mM KCl, 10 mM Na 2 HPO 4 , 1.8 mM KH 2 PO 4 ) containing 0.1% BSA. Finally, 300 μl of the whole-cell extract was mixed with pretreated beads and allowed to immunoprecipitate at 4°C for 16 h. The precipitates were washed twice with lysis 140 buffer, once with lysis 500 buffer (0.1% Na-deoxycholate, 1 mM EDTA, 50 mM HEPES-KOH [pH 7.5], 500 mM NaCl, 1% Triton X100), and further washed once with wash buffer (0.5% Na-deoxycholate, 1 mM EDTA, 250 mM LiCl, 0.5% NP-40, 10 mM Tris-HCl [pH 8.0]) followed by once with TE (10 mM Tris-HCl [pH 8.0], 1 mM EDTA). The well-washed precipitates were mixed with 40 μl of elution buffer (10 mM EDTA, 1% SDS, and 50 mM Tris-HCl [pH 8.0]) and allowed to elute the immunoprecipitated protein-DNA complexes at 65°C for 15 min (immunoprecipitation [IP] sample). To elute IP sample completely, 100 µl of elution buffer and 150 µl of TE containing 0.67% SDS in remaining beads, then incubated at 65°C for 15 min again. Three microliters of the whole-cell extract was mixed with 97 µl of lysis 140 buffer and 400 μl of TE buffer containing 1% SDS (Input sample). IP and Input sample were added 84 µg of proteinase K (Merck, Darmstadt, Germany), and incubated at 37°C for 16 hr. After incubation, the temperature was shifted to 65°C and the sample was further incubated for 6 h. After incubation, DNA was phenol/chloroform extracted from each of the samples and quantified by quantitative PCR using Thermal Cycler Dice Real Time (Takara Bio, Shiga, Japan) and the THUNDERBIRD® SYBR qPCR Mix (TOYOBO, Osaka, Japan). Primer sets p81/p82 at SPBC24C6.09c-mlon-box and p83/p84 at the prp3 locus were used for quantitative PCR analysis.
Micrococcal nuclease digestion assay. Analysis of chromatin structure by indirect end-labeling was performed according to the method of Asada et al. 60 with modifications as follows. 5 × 10 8 cells from 100 ml of the culture was harvested. Cells were incubated in 0.25 ml of preincubation solution (20 mM Tris-HCl at pH 8.0, 0.7 M 2-mercaptoethanol, 3 mM EDTA) at 30°C for 10 min and washed once in 1 ml of ice-cold 1 M sorbitol with 10 mM EDTA. Cells were then centrifuged and resuspended in 1 ml of freshly prepared Zymolyase solution (37.5 mM Tris-HCl at pH 7.5, 0.75 M sorbitol, 1.25% glucose, 0.1% (wt/vol) Zymolyase-100T, 6.25 mM EDTA). Cells were incubated for 5 min at 30°C and resulting spheroplasts were pelleted. All subsequent steps were done at 4°C. The spheroplasts were then washed once in 1 ml of ice-cold 1 M Sorbitol and resuspended well by pipetting in 1 ml of freshly prepared lysis buffer (18% Ficoll-400, 10 mM KH 2 PO 4 , 10 mM K 2 HPO 4 [pH 6.8], 1 mM MgCl 2 , 0.25 mM EGTA, 0.25 mM EDTA, 1 mM PMSF). After centrifugation at 13,000 rpm for 40 min, the crude nuclear pellet was resuspended well in 1.5 ml of buffer A (10 mM Tris-HCl at pH 8.0, 150 mM NaCl, 5 mM KCl, 1 mM EDTA, 1 mM PMSF). 0.5 ml aliquots of the crude chromatin suspension were digested with different amounts of MNase (0, 20, and 50 U/ml) for 5 min at 37°C in the presence of 5 mM CaCl 2 . The reaction was terminated by adding 40 mM EDTA, 1% SDS, and 200 μg of proteinase K and incubated at 50°C for 16 h. Insoluble material was removed by centrifugation at 4°C. The supernatants were extracted with phenol/chloroform, digested with RNase A, and then extracted once with phenol/chloroform. The extracted DNA was precipitated by 2-propanol, rinsedin 70% ethanol, and resuspended in 50 μl of TE buffer. DNA samples were digested by ClaI, XhoI or ApaI for analysis of fbp1, ade6 or SPBC24C6.09c region respectively, and separated by electrophoresis on 1.5% for fbp1or 1.2% for ade6 or SPBC24C6.09c agarose gel in Tris base, acetic acid and EDTA (TAE) buffer (0.48% Tris base, 0.11% acetic acid and 0.037% EDTA) at 60 V for 20 h, followed by Southern blotting using the same probe as in the detection of fbp1, ade6 or SPBC24C6.09c transcripts. The methods for hybridization and signal detection were the same as in Northern blot analysis.
Detection of meiotic DSBs. DNA samples were prepared in agarose plugs from cells from a synchronous meiotic culture according to the method by Hirota et al. 61 with modifications as follows. Briefly, 50 ml of culture was harvested and washed twice using 500 µl of 20 mM citrate-phosphate (pH 5.6), 1.2 M sorbitol, and 40 mM EDTA (CSE) buffer. After resuspension with 500 µl of CSE buffer containing 1.5 mg/mL Zymolyase 20T, cells were incubated at 37°C for 60 min, centrifuged, and resuspended in 500 µl of 10 mM Tris-Cl (pH 7.5), 0.9 M sorbitol, and 45 mM EDTA (TSE) buffer. An equal volume of low-melting-point agarose (1%) in TSE buffer was added, and the mixture was poured into agarose plug molds (Bio-Rad Laboratories, Hercules, CA, USA). After 5 min on ice, the agarose plugs were collected into a buffer (0.25 M EDTA, 50 mM Tris-Cl [pH 7.5], and 1% sodium dodecyl sulfate [SDS]) and incubated at 55°C for 90 min. Then, the buffer was removed, and a second buffer (1% lauryl sarcosine, 0.5 M EDTA [pH 9.5], and 1 mg/mL of proteinase K) was added. The mixture was incubated at 55°C for 2 days. Next, agarose plugs were soaked in 1 ml of Tris-EDTA (TE) containing 0.5 mM phenylmethylsulfonyl fluoride (PMSF) at room temperature for 60 min and washed twice with TE at room temperature for 60 min. The agarose plugs were equilibrated using CutSmart buffer (New England Biolabs, Ipswich, MA, USA) overnight. After changing CutSmart buffer, the agarose plugs were incubated at room temperature for 60 min. DNA in the agarose plugs was digested by AflII or StuI and PstI-HF (New England Biolabs) to detect meiotic DSBs in ade6 or SPBC24C6.09c, respectively. The digested DNA fragments were separated by electrophoresis on 0.8% agarose gel in TAE buffer at 60 V for 20 h, followed by Southern blotting. DNA probes for ade6 and SPBC24C6.09c were amplified by PCR using primer sets p85/p86 and p87/p88, respectively. To detect whole meiotic DSBs, the agarose plugs were washed with TE for 60 min, and the chromosomal DNA and DSBs in agarose plugs were separated by pulse-field gel electrophoresis (PFGE) and stained by ethidium bromide. PFGE was carried out in 0.8% chromosomal grade agarose (Biorad) on a Biorad CHEF-DRIII system and re-circulated at 14°C. Electrophoresis was performed for 48 h at 2 V/cm in 1 × TAE (Tris-acetate-EDTA) buffer, with a switch time of 30 min at an included angle of 100°.
Determination of the recombination rate. The recombination rate was determined according to the method by Hirota et al. 32 with modifications as follows. Briefly, each strain was crossed with a tester strain (ade6-469) on SPA medium at 30°C for 2 days. Sporulating asci were suspended in 1 ml of distilled water containing 115 units of β-glucuronidase and agitated at 30°C for 30 min to release spores from asci and kill vegetative cells. After incubation, 500 µl of ethanol was added, and incubation was again performed at room temperature for 15 min. Spores were collected by centrifugation and resuspended in 1 mL of distilled water. Then, 10 4 , 10 3 , and 200 spores were spread onto SD-lacking adenine, SD-lacking leucine and histidine, and YE-containing adenine, respectively. The recombination rate in ade6 was represented as the colony number on SD-lacking adenine normalized by the colony number on YE-containing adenine. The recombination frequency between leu1 and his3 was calculated as the percentage of the colony number on SD-lacking leucine and histidine normalized by the colony number on YE-containing adenine.
5′ rapid amplification of cDNA ends experiment. The rapid amplification of complementary DNA (cDNA) ends (RACE) experiment was performed using the SMARTer® RACE 5′/3′ Kit (Takara Bio) according to the manufacturer's instructions. The 5′ ends of the transcripts were amplified by PCR using the universal primer mix included in the kit and the gene-specific primer p89. PCR products were purified using the QIAquick gel extraction kit (Qiagen) and cloned into pCR-BluntII-TOPO (Invitrogen, Carlsbad, CA, USA). Finally, the sequences were determined using the M13 primer p90.
Statistics and reproducibility. Throughout the manuscript, all individual data points are plotted. Sample mean and standard deviation are also shown on each bar graph. All statistical analyses were conducted using a one-sided Student's t-test.
Reporting summary. Further information on research design is available in the Nature Research Reporting Summary linked to this article.
Rec12 covalently associated oligonucleotide sequence can be found as Rec12-oligo (WT_SOLID_313) under accession No. G7E49977. All uncropped image data used for all figures in the paper are present in Supplementary Fig. 10. Raw data used for all histograms in this paper are present in Supplementary data 1.