P31comet, a member of the synaptonemal complex, participates in meiotic DSB formation in rice

Jianhui Jia,b,c,1, Ding Tanga,b,1, Yi Shena,b,1, Zhihui Xuea,b, Hongjun Wanga,b, Wenqing Shia,b, Chao Zhanga,b, Guijie Dua,b, Yafei Lia,b, and Zhukuan Chenga,b,2

aState Key Laboratory of Plant Genomics, Institute of Genetics and Developmental Biology, Chinese Academy of Sciences, Beijing 100101, China; bNational Center for Plant Research, Institute of Genetics and Developmental Biology, Chinese Academy of Sciences, Beijing 100101, China; and cJiangsu Key Laboratory for Eco-Agriculture Biotechnology around Hongze Lake, Huaiyin Normal University, Huaian 223300, China

Edited by James A. Birchler, University of Missouri-Columbia, Columbia, MO, and approved July 26, 2016 (received for review May 11, 2016) The mitotic arrest-deficient 2 (Mad2) binding p31comet Besides those SPO11 family , four members [i.e., Arabidopsis participates in the spindle checkpoint and coordinates cell cycle events thaliana putative recombination initiation defect (AtPRD)1, AtPRD2, in mitosis although its function in meiosis remains unknown in all AtPRD3, and Arabidopsis thaliana DSB formation (AtDFO)] are organisms. Here, we reveal P31comet as a synaptonemal complex (SC) required for DSB formation in Arabidopsis.AtPRD1sharesse- comet proteininrice(Oryza sativa L.). In p31 , homologous pairing and quence similarity with MEI1. AtPRD2 seems to be the homolog of synapsis are eliminated, leading to the homologous nondisjunction the yeast ScMei4 and SpRec24. AtDFO and AtPRD3 are plant- and complete sterility. The failure in loading of histone H2AX phos- specific proteins (12). In rice, only OsSPO11-1, OsSPO11-4, comet phorylation (γH2AX) in p31 , together with the suppressed chro- OsSDS, and CENTRAL REGION COMPONENT 1 (CRC1) mosome fragmentation in rice completion of meiotic recombination – comet comet have been found to be essential for DSB formation (13 16). 1(com1) p31 and radiation sensitive 51c (rad51c) p31 double Previous studies have established that synapsis can be either in- comet mutants, indicates that P31 plays an essential role in double-strand dependent of, or dependent upon, recombination. The former break (DSB) formation. Interestingly, the dynamic colocalization pat- is mainly described in Caenorhabditis elegans and Drosophila tern between P31comet and ZEP1 (a transverse filament protein of SC) comet melanogaster females, in which the formation of synaptonemal by immunostaining, as well as the interaction between P31 and complexes (SCs) and recombination are independent (17, 18). The CELL BIOLOGY CENTRAL REGION COMPONENT 1 (CRC1) in yeast two-hybrid assays, latter is found in most organisms, including plants (19), in which suggests possible involvement of P31comet in SC installation. Together, comet synapsis is dependent upon recombination. SC is a highly ordered, these data indicate that P31 plays a key role in DSB formation and ladder-like proteinaceous structure that promotes intimate connec- SC installation, mainly through its cooperation with CRC1. tion along the homologs’ entire lengths. SC assembly starts from early prophase I with the formation of two parallel axial elements rice | meiosis | synaptonemal complex | DSB formation | p31comet (AEs) connected by the central elements (CEs). Rice contains two AE components, HOMOLOGOUS PAIRING ABERRATION IN uring meiosis, homologous recombination plays a crucial RICE MEIOSIS 2 (PAIR2) and HOMOLOGOUS PAIRING Drole in ensuring faithful and precise segrega- ABERRATION IN RICE MEIOSIS 3 (PAIR3) (20, 21), and one tion, as well as in diversifying genetic information. Meiotic re- TF protein, ZEP1 (a transverse filament protein of SC) (22). Re- combination is induced by the programmed formation of DNA cently, CRC1 has been recognized as a new SC central region double-strand breaks (DSBs), catalyzed by Spo11, a functional homolog of subunit A of an archaeal topoisomerase (TopoVIA) Significance (1). In Saccharomyces cerevisiae, nine associated proteins, in- cluding meiotic recombination 11 (Mre11), radiation sensitive 50 (Rad50), X-ray sensitive 2 (Xrs2), superkiller 8 (Ski8), Rec102, The programmed double-strand break (DSB) formation is a Rec104, Rec114, meiosis-specific 4 (Mei4), and Mer2, are required prerequisite for homologous recombination. The synaptonemal for facilitating Spo11’s catalytic reaction (2). In Schizosaccharomyces complex (SC), known as the meiotic-specific proteinaceous structure, provides an appropriate framework for homologous pombe, seven DSB formation-related proteins were identified (3). recombination. Both are indispensable in meiosis. p31comet has Rec6 has been recognized as an essential component for ensuring been proved to play an essential role in metaphase-to-anaphase the meiotic DSB formation, and it forms a complex with Rec12 and transition during mitosis by indirectly activating anaphase- Rec14, the orthologs of Spo11 and Ski8, respectively (4). Further- promoting complex/cyclosome status, but the underlying func- more, the Mde2 protein, which lacks an ortholog in S. cerevisiae,is tion in meiosis remains elusive. Here, we demonstrate that the comet also required for DSB formation (5). Nevertheless, Rad50, Mre11, homolog gene in rice, P31 , is involved in meiosis. Our find- and Xrs2 orthologs are dispensable for DSB formation in S. pombe, ings uncover a role for P31comet in regulating meiotic DSB for- as well as in mice and plants, but they are required for meiotic DSB mation and SC assembly by cooperation with CENTRAL REGION processing (6). In mice, except for the conserved protein Spo11, COMPONENT 1 (CRC1). Our work sheds light on P31comet func- Mei4 and Rec114 were also identified as key components involved tions in rice meiosis and thereby may reveal a special role of this in DSB formation (7). In addition, Mei1 was recognized as a re- protein in plants. combination related factor, without an ortholog in S. cerevisiae or S. pombe (8). Recently, mouse TOPOVIBL protein was identified Author contributions: J.J., D.T., and Z.C. designed research; J.J., Y.S., Z.X., H.W., and Y.L. to form the complex with Spo11 and take part in meiotic DSB performed research; J.J., W.S., C.Z., and G.D. analyzed data; and J.J., Y.S., and Z.C. wrote formation (9). the paper. In plants, the Arabidopsis AtSPO11-1, AtSPO11-2, and AtSPO11-3 The authors declare no conflict of interest. share sequence similarity with other Spo11/topo VIA proteins. This article is a PNAS Direct Submission. AtSPO11-1 and AtSPO11-2, as subunits of a heterodimer, are Freely available online through the PNAS open access option. essential for DSB formation whereas AtSPO11-3 is not required 1J.J., D.T., and Y.S. contributed equally to this work. for DSB formation (10). Recently, an Arabidopsis homolog of the 2To whom correspondence should be addressed. Email: [email protected]. archaeal topo VIB subunit (MTOPVIB) was found to interact with This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. the SPO11-1/2 heterodimer and participate in DSB formation (11). 1073/pnas.1607334113/-/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.1607334113 PNAS Early Edition | 1of6 Downloaded by guest on September 23, 2021 component in rice (13). It encodes a conserved AAA-ATPase and essential roles in rice meiosis, especially in DSB formation and is the homolog of Pch2/thyroid receptor interacting protein 13 SC installation. (TRIP13), which was first recognized as a pachytene check- point gene in budding yeast (23), as well as in C. elegans and Results D. melanogaster (24, 25). However, the mouse TRIP13-deficient Characterization of a Sterile Mutant in Rice. A sterile rice mutant meiocytes undergo synapsis (26). was obtained by ethyl methane sulfonate (EMS) mutagenesis Mitotic arrest-deficient 2 (Mad2) acts as a pivotal com- from the indica rice variety, Zhongxian 3037. The mutant plant ponent of the spindle assembly checkpoint to delay the ini- showed normal vegetative growth but was completely sterile tiation of anaphase-promoting complex/cyclosome (APC/C) (Fig. 1A). The pollen grains of the mutant were shrunken and comet activity. P31 was first identified as a Mad2-interacting protein empty as revealed by staining with I2-KI solution (Fig. 1A). The in HeLa cells and named caught by MAD two (CMT2). Over- female gametes of the mutant were also inviable. No seed was set expression of CMT2 resulted in premature destruction of securin when the mutant spikelets were pollinated with WT pollen and permitted the exit from mitosis whereas the silence of CMT2 grains. The progeny of a self-fertilized heterozygous plant was induced cell death, accompanied with a transient anaphase delay examined. We found that the segregation ratio of plants with (27). CMT2 was renamed p31comet because of its comet tail-like normal fertility to sterile ones was about 3:1 (180 fertile and 62 localization pattern during mitosis. The major role of CMT2/ sterile plants), implying that the sterile phenotype is controlled p31comet is to counteract the function of Mad2 and subsequently by a single recessive nuclear gene (χ2 = 0.0496; P > 0.05). promote the activation of APC/C, resulting in degradation of securin and cyclin B (28). The crystal structure of Mad2-p31comet Map-Based Cloning of the P31comet Gene in Rice. Map-based cloning dimer was analyzed, and p31comet was considered to mimic the was used to isolate P31comet. The target gene was mapped to the structure of Mad2 to inhibit the dimerization of Mad2, allowing the short arm of chromosome 5 close to the centromere, which was metaphase/anaphase transition during mitotic checkpoint inactivation further delimited to a region of about 300 kb. Within this region, (29). Recent studies have shown that TRIP13 links p31comet to dis- one candidate gene (LOC_Os05g16250) was found to lack a G at assemble the mitotic checkpoint complex (MCC) (30, 31). nucleotide position 57 of the coding sequence (CDS), resulting Although p31comet acts as an interacting partner of Mad2 in premature stop codon (TAG) (Fig. 1B). We named the mu- during the spindle assembly checkpoint (SAC) in human cells, it tant p31comet-1 based on the protein (see is unclear whether CMT2/p31comet takes any roles in meiosis. Here, below). We also obtained another allele with the similar phe- we identified the p31comet homolog in rice and demonstrated its notype as p31comet-1 induced by 60Co γ-ray radiation. Sequencing

Fig. 1. Identification of the p31comet mutant. (A, Left and Center) Comparison of a WT plant (Left) and a p31comet-1 mutant plant (Center). (A, Right) Pollen comet comet grains with 1% I2-KI solution in a WT plant (Upper Right) and a p31 -1 plant (Lower Right). (Scale bars: 50 μm.) (B) Schematic representation of P31 gene and the mutation sites in P31comet, including p31comet-1 and p31comet-2. Coding regions are shown as black boxes, and untranslated regions are shown as gray boxes. The triangles indicate the mutated sites in P31comet.(C) The alignment of full-length amino acids of P31comet homologs among Oryza sativa, Arabidopsis thaliana, Drosophila willistoni, Homo sapiens, and Mus musculus. Residues identical to P31comet are boxed in red.

2of6 | www.pnas.org/cgi/doi/10.1073/pnas.1607334113 Ji et al. Downloaded by guest on September 23, 2021 Fig. 2. Meiotic chromosome behaviors and FISH analysis in both WT and p31comet PMCs. (A)Meiotic chromosome behaviors in both WT and p31comet mu- tant. (B) Telomere bouquet formation revealed by FISH assay, using pAtT4 (red) as probe in both WT and p31comet mutant. Meiotic are stained with DAPI (blue). (C) Chromosome pairing was ana- lyzed by FISH assay, using a 5S rDNA (red) probe in WT and p31comet mutant at pachytene. A single 5S rDNA focus was observed on the paired chromosomes in WT PMC, whereas two separate signals were detected in p31comet PMC. (Scale bars: 5 μm.)

revealed that this mutant contains a Ty1-copia–like retrotransposon of these meiocytes. Univalents segregated randomly at anaphase (LOC_Os12g03850) in the first intron of the same gene (Fig. 1B), I, leading to the homologous nondisjunction. The subsequent comet comet which was then designated p31 -2. p31 -1 was selected for second meiotic division occurred, and tetrads with an unequal most subsequent studies. Using National Center for Biotechnology number of chromosomes were generated. Additionally, the same Information (NCBI) BLASTp, we found that the candidate protein comet comet defects were observed in p31 -2 and p35S::P31 RNAi CELL BIOLOGY shares similarity (∼22% identity and 34.7% similarity) with the p31comet in (Fig. 1C). The phylogenetic tree further con- firmed that the predicted protein encoded by Os05g16250 is the closest homolog of p31comet in rice (Fig. S1).

The Functional Verification of P31comet in Rice. To determine whether the sterility of the mutant plants indeed results from the mutation of P31comet, an RNA interference (RNAi) experiment was performed. Among the 22 p35S::P31comet RNAi transgenic plants, 15 were shown to be completely sterile. RT-PCR analysis revealed that the transcription of P31comet was severely down- regulated in p31comet-2 panicles whereas, in p31comet-1 panicles, there was no obvious difference compared with the WT (Fig. S2). It is reasonable because p31comet-1 has a 1-bp deletion in the first exon whereas p31comet-2 contains a large insertion in the first intron, leading to a critical disruption of its transcription. Cytological defects of p31comet-2 and P31comet-RNAi plants were similar to that observed in p31comet-1 (Fig. S3). Altogether, we propose that mu- tations in the P31comet gene cause the sterility phenotype. By transformation of the full-length P31comet sequence into the genome of p31comet-1, the mutant phenotype was successfully complemented. Five independent transformants showed a vir- tually complete rescue of all phenotypic traits (Fig. S3 E and F). This complementation assay further demonstrated that the mu- tant phenotype is indeed caused by a 1-bp deletion in p31comet-1.

comet Abnormal Chromosome Behaviors in p31 Pollen Mother Cells. To investigate the meiotic defects in p31comet-1,weexamined the behaviors of 4,6-diamidino-2-phenylindole (DAPI)–stained chromosomes in pollen mother cells (PMCs). In the WT (Fig. 2A), SC formation was complete at pachytene, and 12 bivalents became highly condensed at diakinesis. The homologous chro- mosomes segregated equally to the opposite poles of the cell at anaphase I. During meiosis II, the sister chromatids separated Fig. 3. P31comet is essential for meiotic DSB formation. (A) Immunofluorescent comet investigation of histone H2AX phosphorylation (γH2AX) signals on the meiotic to produce tetrads. In p31 -1 (Fig. 2A), no obvious homolo- comet gous chromosome pairing and synapsis were observed at early spreads of the WT and p31 PMCs. OsREC8 signals were used to indicate the meiotic chromosomes (green); the γH2AX signals are in red. γH2AX presents as prophase I. From diakinesis to metaphase I, 24 univalents numerous dot-like signals in the WT, whereas no obvious signals are detected in were clearly observed, scattered throughout the entire nucleus. p31comet.(B) The loading pattern of P31comet signals in both WT and p31comet comet We investigated 125 p31 -1 PMCs at metaphase I and con- PMCs at leptotene. P31comet proteins present as punctate foci (green) in the WT firmed that there were no associated homologs (bivalent) in any PMCs, whereas no obvious signals are detected in p31comet. (Scale bars: 5 μm.)

Ji et al. PNAS Early Edition | 3of6 Downloaded by guest on September 23, 2021 Fig. 4. Meiotic chromosome behaviors of Osrad51c, Osrad51c p31comet, Oscom1, and Oscom1 p31comet. Chromosomes were stained with DAPI. The chro- mosome fragments in the Osrad51c and Oscom1 single mutant are suppressed in the Osrad51c p31comet and Oscom1 p31comet double mutants, respectively. (A, D, G,andJ)Pachytene.(B, E, H,andK) Diakinesis. (C, F, I,andL) Anaphase I. (Scale bars: 5 μm.)

plants (Fig. S3). In conclusion, the complete failure of chiasmata (Osrad51c) p31comet and Oryza sativa completion of meiotic re- formation, followed by unequal segregation of chromosomes, led combination 1 (Oscom1) p31comet double mutants and examined to the sterile phenotype observed in p31comet mutants. meiotic chromosome behaviors in these plants. OsRAD51C is re- quired for meiotic DSB repair (33). In Osrad51c, homologous comet Homologous Pairing Is Defective in p31 -1 PMCs. Telomere pairing and synapsis were affected at pachytene, and massive chro- bouquet clustering is a particular structure that may promote ini- mosomal fragments were observed at anaphase I, indicative of the tiation of homologous pairing in early prophase I (32). To explore defective DSB repairing (Fig. 4 A–C). In Osrad51c p31comet,the the bouquet formation in p31comet, we conducted fluorescence in meiotic chromosome behaviors mimic the chromosome dynamics in situ hybridization (FISH) analysis using a telomere-specific probe p31comet-1. Twenty-four univalents were present at diakinesis and (pAtT4) in WT and p31comet-1 (Fig. 2B). In WT PMCs, almost all then segregated randomly at anaphase I (Fig. 4 D–F), indicating that of the pAtT4 signals were gathered within a limited region at early the chromosome fragmentations in Osrad51c were suppressed by the zygotene stage, showing a typical telomere bouquet configuration. p31comet mutation. OsCOM1 has been demonstrated to be essential In p31comet-1, the telomere clustering was also observed, which was for DSB processing. In rice meiosis, loss of OsCOM1 ledtonon- similar to that in the WT. The number of telomere foci outside the homologous entanglements and massive fragments that were de- clustered region was quantified. Further statistical tests revealed rived from the failure of faithful DSB repair (34). Here, the that there was no significant difference [t(112) = 0.731, P = 0.466, abnormal chromosome associations in Oscom1 (Fig. 4 G–I)wereno unpaired t test] in the number between WT (2.3 ± 0.2, n = 52) and longer observed in Oscom1 p31comet, being replaced by 24 univalents p31comet-1 (2.5 ± 0.2, n = 62) (Fig. S4). at diakinesis (Fig. 4 J–L). Both assays provided forceful evidence to FISH analysis using 5S rDNA as the probe was performed to demonstrate that P31comet isaprerequisiteforthedefectsinDSB monitor the homologous pairing status in p31comet PMCs. The 5S repair in either Osrad51c or Oscom1. ribosomal DNA (rDNA) is located on the short arm of chro- mosome 11 and often is used to indicate chromosome pairing P31comet Overlaps with ZEP1 During both SC Installation and and segregation. In the WT, paired 5S rDNA signals were de- Uninstallation. To gain further insight into the spatial and tem- tected on one bivalent at pachytene, indicating well-paired ho- poral localization of P31comet, immunostaining was conducted in mologous chromosomes at this stage (Fig. 2C). In p31comet-1, rice PMCs using the antibody against P31comet. P31comet proteins unpaired 5S rDNA signals were constantly detected in all ob- appeared as punctate foci on chromosomes at leptotene in WT served pachytene PMCs (n = 103, Fig. 2C), indicating that (Fig. 3B) whereas no P31comet staining was detected in the cor- P31comet is required for homologous pairing. responding stage in p31comet-1 PMCs, confirming the specificity of the anti-P31comet antibody. Here, PAIR3 was available as a P31comet Is Essential for Normal DSB Formation. Because 24 univa- good marker to indicate chromosome axes during early prophase lents were present in p31comet PMCs, we wanted to know whether I (20). During zygotene, numerous foci and short stretches of DSB formation was disrupted in p31comet mutants. γH2AX, the P31comet signals were observed and rapidly developed into linear, phosphorylated form of the histone variant H2AX, has been but not intact, structures along the chromosome axes (Fig. 5A). At widely used as a cytological marker for DSB formation in dif- early pachytene, P31comet staining assembled perfectly along the ferent organisms. Here, in WT PMCs, dot-like γH2AX signals entire chromosome and exhibited an almost overlapping, yet slightly were observed at zygotene (Fig. 3A). By contrast, no obvious narrower, localization pattern with PAIR3. From late pachytene to γH2AX signals were detected at the corresponding stage in early diplotene, P31comet was observed as intermittent linear signals p31comet-1, implying that P31comet is crucial for meiotic DSB between the two parallel PAIR3 axes. P31comet proteins were then formation. Furthermore, P31comet protein signals displayed as released from the chromosomes quickly, with only some residual numerous foci and localized on meiotic chromosomes in lepto- dispersed signals remaining around chromosomes (Fig. 5A). tene (Fig. 3B), implying its role during early prophase I. The dynamics of P31comet localization were similar to that of To further confirm the role of P31comet in meiotic DSB for- ZEP1 and CRC1, both of which have been regarded as the mation, we constructed Oryza sativa radiation sensitive 51c central SC components in rice (13, 22). Thus, a triple immunostaining

4of6 | www.pnas.org/cgi/doi/10.1073/pnas.1607334113 Ji et al. Downloaded by guest on September 23, 2021 Fig. 5. The localization pattern of P31comet and the interaction between P31comet and CRC1 by Y2H as- say. (A) Dual immunolocalization of PAIR3 (red) and P31comet (green) in WT PMCs. PAIR3 was used to in- dicate the meiotic chromosomes. P31comet proteins present as foci and short linear signals at zygotene and quickly form the entire threads along homolo- gous chromosomes at pachytene. P31comet proteins are dissociated and remain as several diffused signals around chromosomes at diplotene. (B) Triple immu- nolocalization of OsREC8 (blue), P31comet (green), and ZEP1 (red) in the WT PMCs. The merged signals are shown in pink. (Scale bars: 5 μm.) (C)P31comet interacts with CRC1 in Y2H assays. SD/DDO (SD-Leu- Trp) was used to test the cotransformation effi- ciency. Interactions between the full length of CRC1 and P31comet were verified by the growth and blue color of yeast on selective medium SD/QDO/X/A (SD-Leu-Trp-His-Ade+X-α-Gal+Aureobasidin A). AD, prey vector; BD, bait vector.

assay for OsREC8, P31comet, and ZEP1 was conducted in WT PMCs P31comet Is Dispensable for Establishing Normal Spindle Morphology. (Fig. 5B). From zygotene to diplotene, P31comet and ZEP1 displayed Highly conserved among eukaryotes, SAC proteins regulate chromosome segregation by sequestering CDC20 and promote

very well colocalization patterns on chromosomes (Fig. 5B and Fig. CELL BIOLOGY S5), indicating that P31comet was constantly together with ZEP1 the transition from metaphase to anaphase (36). Previous find- during both SC installation and uninstallation processes. However, ings also reported that, in rice SAC protein Bub1-related kinase they were always between the two layers of lateral elements, as in- 1(brk1) and Arabidopsis cell division cycle 20.1 (cdc20.1) mu- dicated by OsREC8. tants, the spindle lengths were a little bit longer than in the WT at metaphase I (37, 38). Because p31comet acts as an interacting comet partner of SAC effector molecule Mad2 in human (29), we thus P31 Interaction with CRC1 Other than ZEP1 Revealed by Yeast comet Two-Hybrid Assay. Previously we have shown that CRC1 is a investigated whether the loss of P31 is associated with the conserved SC component and also is involved in DSB formation aberrant spindle assembly. We examined spindle morphology through immunolocalization using an anti–ɑ-tubulin antibody. In (13). Consistent with this finding, we found that P31comet probably WT, bivalents were aligned well on the equatorial plate at meta- also participate in these events. This shared function prompted us phase I, due to the equal drawing force by bundles of microtubule to investigate the possible interaction between the two proteins. fibers generated from two opposite poles (Fig. S7A). At metaphase Yeast two-hybrid (Y2H) assays confirmed a direct interaction be- comet II, microtubules were detected as a similar organization pattern, tween full-length P31 and CRC1 proteins (Fig. 5C). However, surrounding two clusters of segregated homologous chromosomes comet no interaction was detected between P31 and ZEP1. Consid- (Fig. S7B). In p31comet-1 PMCs, from metaphase I to metaphase II, ering the previous result that CRC1 interacted with N-terminal of we observed a spindle morphology similar to that in WT, except for ZEP1, we propose that P31comet performs its possible function in the abnormal chromosome morphology, with univalents at meta- SC installation through the interaction with CRC1. phase I (Fig. S7 C and D). Thus, we conclude that P31comet is dispensable for normal spindle assembly in rice meiosis. Organization of the Axial Elements and Central Elements in p31comet-1. To further verify the function of P31comet in rice meiosis, Discussion immunostaining assays were performed to investigate the local- In this study, we present a detailed characterization of the rice ization patterns of chromosome axis proteins and SC components P31comet gene and propose its specific roles in meiosis. We found comet in p31comet PMCs. OsREC8 is essential for chromatid cohesion and that rice P31 is essential for SC installation because of the comet has been used to indicate chromosomes during early prophase I colocalization patterns of P31 and ZEP1 and also due to the comet (35). PAIR2, the homolog of yeast HOP1 and Arabidopsis ASY1, is interaction between P31 and CRC1 in Y2H assays. We also found that P31comet plays an essential role in meiotic DSB formation. an AE-associated protein in rice (21). PAIR3 seems to be the comet functional homolog of Red1 in budding yeast and of ASY3 in Although we report P31 playing crucial roles in meiotic pro- gression, its detailed function needs a more in-depth investigation. Arabidopsis and plays pivotal roles in homologous pairing and SC Here, we found that P31comet shares the sequence homology formation in rice (20). Here, immunostaining results revealed the with its homologs, indicating that these proteins might share a normal localization patterns of all three of these chromosome axis- comet comet common ancestry. However, the rice p31 mutant did not show related proteins in p31 -1 PMCs, implying that the organization obvious defects, either in meiotic spindle assembly or in a delay of of AEs is probably not affected in the mutant (Fig. S6A). meiotic progression from diakinesis to anaphase I. One important Immunostaining for CRC1 and ZEP1, two SC components in comet comet function of P31 in rice meiosis is to take part in DSB forma- rice, was conducted in p31 . Neither CRC1 nor ZEP1 signals tion, mimicking its Y2H-interacting protein, CRC1. CRC1 inter- comet comet were observed in p31 . We did not detect any P31 signals acts with HOMOLOGOUS PAIRING ABERRATION IN RICE in either crc1 or zep1 PMCs (Fig. S6B). Taken together, we MEIOSIS 1 (PAIR1) and probably builds a bridge between these conclude that installation of AEs occurs normally whereas the DSB formation-related proteins (13). Although no direct interac- assembly of central elements is disrupted in p31comet mutants. tion between PAIR1 and P31comet was detected through the Y2H

Ji et al. PNAS Early Edition | 5of6 Downloaded by guest on September 23, 2021 assay, we speculate that CRC1/P31comet may coordinate with that p31comet is involved in SAC function. In addition, TRIP13 PAIR1 to promote DSB formation. and p31comet promoted the inactivation of the mitotic checkpoint In rice, the CRC1 signals were localized on the central region of (30). However, the function of SAC-related proteins in mitosis the SC, and CRC1 interacted with the N terminus of ZEP1, sug- has not yet been tested in plants, and several SAC proteins have gesting that CRC1 may act as the central element of the SC (13). In been found to be involved in plant meiosis, such as MPS1, Au- this study, we also found that the linear P31comet signals localize in rora kinases, AtPCH2, and CDC20.1 in Arabidopsis, as well as the central region between the paired lateral elements. This local- CRC1 and BRK1 in rice (12, 37). In rice brk1-1 meiocytes, the comet ization pattern prompted us to propose that CRC1/P31 col- spindle length was slightly elongated (38). In Arabidopsis cdc20.1, comet laborates in SC installation. Interestingly, the interaction of p31 meiocytes exhibited atypical spindle morphology and had a with TRIP13 has also been reported in mice (39). The joint action longer average spindle length at metaphase I (37). Although comet of TRIP13 and p31 promotes the release of Mad2 from the p31comet acts to monitor the spindle checkpoint in humans, we MCC, participates in the complete disassembly of the MCC, and ’ comet didn t found any defects in meiotic spindle assembly in the rice abrogates checkpoint inhibition of APC/C (30). CRC1 and P31 p31comet mutant. Further studies to explore the role of P31comet are both colocalized with the ZEP1 protein, implying an intimate in meiosis will help us to fully understand its biological functions. relationship among these three proteins. Some differences were still uncovered between CRC1 and P31comet. For example, PAIR2 Materials and Methods comet loaded in p31 PMCs successfully, but not in crc1, showing that The p31comet-1 was characterized from an indica rice variety Zhongxian 3037, CRC1 is required for recruitment of PAIR2 onto the chromosome, and the p31comet-2 was isolated from a japonica variety Yandao 8. Details on comet whereas P31 is not required for this function. Secondly, no the following are available in SI Materials and Methods: plant materials, comet interaction between P31 and ZEP1 was detected whereas phylogenetic tree construction and gene structure presentation, full-length CRC1 interacted with the N-terminal domain of ZEP1 in Y2H cDNA cloning of P31comet, real-time PCR for transcript expression assays, assays. This clue may hint at the underlying mechanism of these RNAi knockdown of P31comet and complementation test, antibody pro- three proteins in SC installation. CRC1 seems to act as a bridge to duction, meiotic chromosome preparation and FISH analysis, immunofluo- connect P31comet withZEP1.ItispossiblethatP31comet/CRC1 act rescence, and Y2H assay analysis. The primers used in this study are listed in as SC components or that they just function as primary sensors to Table S1. signal the state of SC assembly. The detailed interaction patterns of these proteins remain to be determined. ACKNOWLEDGMENTS. This work was supported by the National Key Re- comet search and Development Program of China Grant 2016YFD0100901, National Human p31 physically interacted with the SAC compo- Natural Science Foundation of China Grants 31230038 and 31400169, and nent Mad2 through structural mimicry, leading to the conclusion Natural Science Foundation of Jiangsu Province Grant BK20140454.

1. Bergerat A, et al. (1997) An atypical topoisomerase II from Archaea with implications 20. Wang K, et al. (2011) PAIR3, an axis-associated protein, is essential for the recruitment for meiotic recombination. Nature 386(6623):414–417. of recombination elements onto meiotic chromosomes in rice. Mol Biol Cell 22(1): 2. Edlinger B, Schlögelhofer P (2011) Have a break: Determinants of meiotic DNA double 12–19. strand break (DSB) formation and processing in plants. J Exp Bot 62(5):1545–1563. 21. Nonomura K, Nakano M, Eiguchi M, Suzuki T, Kurata N (2006) PAIR2 is essential for 3. Steiner S, Kohli J, Ludin K (2010) Functional interactions among members of the homologous chromosome synapsis in rice meiosis I. J Cell Sci 119(Pt 2):217–225. meiotic initiation complex in fission yeast. Curr Genet 56(3):237–249. 22. Wang M, et al. (2010) The central element protein ZEP1 of the synaptonemal complex – 4. De Veaux LC, Hoagland NA, Smith GR (1992) Seventeen complementation groups regulates the number of crossovers during meiosis in rice. Plant Cell 22(2):417 430. of mutations decreasing meiotic recombination in Schizosaccharomyces pombe. 23. San-Segundo PA, Roeder GS (1999) Pch2 links chromatin silencing to meiotic check- point control. Cell 97(3):313–324. Genetics 130(2):251–262. 24. Bhalla N, Dernburg AF (2005) A conserved checkpoint monitors meiotic chromosome 5. Miyoshi T, et al. (2012) A central coupler for recombination initiation linking chro- synapsis in Caenorhabditis elegans. Science 310(5754):1683–1686. mosome architecture to S phase checkpoint. Mol Cell 47(5):722–733. 25. Joyce EF, McKim KS (2009) Drosophila PCH2 is required for a pachytene checkpoint 6. de Massy B (2013) Initiation of meiotic recombination: How and where? Conservation that monitors double-strand-break-independent events leading to meiotic crossover and specificities among eukaryotes. Annu Rev Genet 47:563–599. formation. Genetics 181(1):39–51. 7. Kumar R, Bourbon HM, de Massy B (2010) Functional conservation of Mei4 for meiotic 26. Li XC, Schimenti JC (2007) Mouse pachytene checkpoint 2 (trip13) is required for DNA double-strand break formation from yeasts to mice. Dev 24(12): completing meiotic recombination but not synapsis. PLoS Genet 3(8):e130. – 1266 1280. 27. Habu T, Kim SH, Weinstein J, Matsumoto T (2002) Identification of a MAD2-binding 8. Libby BJ, et al. (2002) The mouse meiotic mutation mei1 disrupts chromosome syn- protein, CMT2, and its role in mitosis. EMBO J 21(23):6419–6428. apsis with sexually dimorphic consequences for meiotic progression. Dev Biol 242(2): 28. Xia G, et al. (2004) Conformation-specific binding of p31(comet) antagonizes the 174–187. function of Mad2 in the spindle checkpoint. EMBO J 23(15):3133–3143. 9. Robert T, et al. (2016) The TopoVIB-Like protein family is required for meiotic DNA 29. Yang M, et al. (2007) p31comet blocks Mad2 activation through structural mimicry. Cell double-strand break formation. Science 351(6276):943–949. 131(4):744–755. 10. Stacey NJ, et al. (2006) Arabidopsis SPO11-2 functions with SPO11-1 in meiotic re- 30. Eytan E, et al. (2014) Disassembly of mitotic checkpoint complexes by the joint action of combination. Plant J 48(2):206–216. the AAA-ATPase TRIP13 and p31(comet). Proc Natl Acad Sci USA 111(33):12019–12024. 11. Vrielynck N, et al. (2016) A DNA topoisomerase VI-like complex initiates meiotic re- 31. Miniowitz-Shemtov S, Eytan E, Kaisari S, Sitry-Shevah D, Hershko A (2015) Mode of combination. Science 351(6276):939–943. interaction of TRIP13 AAA-ATPase with the Mad2-binding protein p31comet and 12. Mercier R, Mézard C, Jenczewski E, Macaisne N, Grelon M (2015) The molecular bi- with mitotic checkpoint complexes. Proc Natl Acad Sci USA 112(37):11536–11540. ology of meiosis in plants. Annu Rev Plant Biol 66:297–327. 32. Scherthan H (2007) Telomere attachment and clustering during meiosis. Cell Mol Life 13. Miao C, et al. (2013) Central region component1, a novel synaptonemal complex Sci 64(2):117–124. component, is essential for meiotic recombination initiation in rice. Plant Cell 25(8): 33. Tang D, et al. (2014) OsRAD51C is essential for double-strand break repair in rice 2998–3009. meiosis. Front Plant Sci 5:167. 14. Wu Z, et al. (2015) OsSDS is essential for DSB formation in rice meiosis. Front Plant Sci 34. Ji J, et al. (2012) The role of OsCOM1 in homologous chromosome synapsis and re- combination in rice meiosis. Plant J 72(1):18–30. 6:21. 35. Shao T, et al. (2011) OsREC8 is essential for chromatid cohesion and metaphase I 15. Yu H, et al. (2010) OsSPO11-1 is essential for both homologous chromosome pairing monopolar orientation in rice meiosis. Plant Physiol 156(3):1386–1396. and crossover formation in rice. Chromosoma 119(6):625–636. 36. Pesin JA, Orr-Weaver TL (2008) Regulation of APC/C activators in mitosis and meiosis. 16. An XJ, Deng ZY, Wang T (2011) OsSpo11-4, a rice homologue of the archaeal TopVIA Annu Rev Cell Dev Biol 24:475–499. protein, mediates double-strand DNA cleavage and interacts with OsTopVIB. PLoS 37. Niu B, et al. (2015) Arabidopsis Cell Division Cycle 20.1 is required for normal meiotic One 6(5):e20327. spindle assembly and chromosome segregation. Plant Cell 27(12):3367–3382. 17. Lake CM, Hawley RS (2012) The molecular control of meiotic chromosomal behavior: 38. Wang M, et al. (2012) BRK1, a Bub1-related kinase, is essential for generating proper – Events in early meiotic prophase in Drosophila oocytes. Annu Rev Physiol 74:425 451. tension between homologous kinetochores at metaphase I of rice meiosis. Plant Cell 18. Dernburg AF, et al. (1998) Meiotic recombination in C. elegans initiates by a con- 24(12):4961–4973. served mechanism and is dispensable for homologous chromosome synapsis. Cell 39. Tipton AR, et al. (2012) Identification of novel mitosis regulators through data mining 94(3):387–398. with human centromere/kinetochore proteins as group queries. BMC Cell Biol 13(1):15. 19. Pawlowski WP, et al. (2004) Coordination of meiotic recombination, pairing, and 40. Cheng Z (2013) Analyzing meiotic chromosomes in rice. Methods Mol Biol 990: synapsis by PHS1. Science 303(5654):89–92. 125–134.

6of6 | www.pnas.org/cgi/doi/10.1073/pnas.1607334113 Ji et al. Downloaded by guest on September 23, 2021