Plants have multiple functional SPO11 candidates (see Introduction), and the SPO11 candidate(s) responsible for meiotic DSB formation has not been identified. To answer this question, we developed a Drosophila bioassay for the quantitative evaluation of the DSB activity of SPO11 candidates expressed from transgenes in fly oocytes. The interspecies bioassay that we developed in this study was effective to detect the DSB signals generated by the trans-expression of SPO11s and to identify the SPO11s with DSB-forming activities, among the candidates in plants. It was shown for the first time that both AtSPO11-1 and AtSPO11-2 exhibit DSB activity alone, in the absence of species-specific (i.e. Arabidopsis) interacting proteins for SPO11 functions. Arabidopsis has three genes encoding SPO11 homologues. Two of them (AtSPO11-1 and AtSPO11-2) are required for meiosis. Our results are surprising, since previous genetic studies showed that both AtSPO11-1 and AtSPO11-2 are required for meiosis, and interacting proteins, such as PRD1, are required for the SPO11 functions (see Introduction section). Note that our results do not mean that AtSPO11-1 and AtSPO11-2 work independently in Arabidopsis. It is likely that in Arabidopsis, SPO11-interacting proteins coordinate AtSPO11-1 and AtSPO11-2 to induce functional meiotic DSBs, leading to chiasma formation.
By using this method, we further investigated the three rice SPO11 homologues and a novel one, OsSPO11D. OsSPO11A and OsSPO11B are considered as the counterparts of Arabidopsis AtSPO11-1 and AtSPO11-2, respectively, from their amino acid-sequence similarities. The bioassay revealed that OsSPO11A has DSB activity, as in the case of its Arabidopsis counterpart (Figure ). However, we obtained another unexpected result from other tests: our assay did not detect any significant DSB-forming activity of OsSPO11B, unlike the case of its Arabidopsis counterpart, AtSPO11-2 (Figure ). This was surprising, since OsSPO11B was expressed in the fly at a similar level as OsSPO11A, and at two- to five-fold higher levels than AtSPO11-1, AtSPO11-2 and OsSpo11D, which all showed DSB signals in the oocyte nuclei (Table and Figure ). This observation does not support the suggestion from the phylogenetic analyses (Figure ) that OsSPO11B is an orthologue of AtSPO11-2. OsSPO11D displayed robust DSB activity (Figures and ).
All of the results obtained from the analysis of the γ-H2Av signals were supported in parallel by the analysis of the frequencies of oocyte nuclei with defective karyosome morphology in the DSB-repair defective fly, which represent the ectopically expressed plant SPO11-induced DSBs in Drosophila oocyte nuclei (Figure ).
In order to obtain additional evidence for the DSB activity of SPO11s, we performed genetic tests, using the dmspo11-deficient, but mus301-proficient (DSB-repair proficient), flies. We assumed that if the DSBs induced by a plant SPO11 are not repaired, then the DSBs caused the X chromosome loss in Drosophila oocytes, and if they are repaired by a normal recombination process leading to chiasma formation, then the DSBs restored the normal meiotic X chromosome disjunction. The expression of DmSPO11 from the transgene fully restored the meiotic deficiencies of the dmspo11 mutant flies (Table ). The expression of the rice SPO11 transgenes, as well as the Arabidopsis SPO11s, in oocytes bearing the dmspo11 mutation caused an increase in nullo-X eggs, without associated diplo-X eggs (Table ). The aberrant X-chromosome segregation, which was induced by the expression of the OsSPO11D, AtSPO11-1 and AtSPO11-2 transgenes in dmspo11 mutant flies, correlated well with the abilities of the expressed OsSPO11D, AtSPO11-1 and AtSPO11-2 proteins to induce DSBs, as shown by the γ-H2Av signals in the oocyte nuclei of the dmspo11 mus301 double mutant flies (Table and Figure ). This profile is explained by the X chromosome abnormalities caused by aberrant DSBs, as shown in Figure . Therefore, we conclude that OsSPO11D, as well as AtSPO11-1 and AtSPO11-2, has DSB activity in oocytes.
Then, one may wonder why the DSBs induced by the plant SPO11s did not complement the meiotic defect of the dmspo11
mutant flies. The importance of the timing of DSB formation in meiotic disjunction was clearly shown by Bhagat et al
]: X-ray irradiation of prophase I oocytes of dmspo11
-deficient mutant flies efficiently induced meiotic exchanges and suppressed meiotic nondisjunction of the mutant flies. However, pre-meiotic or post-meiotic irradiation did not induce meiotic exchange and caused more severe nondisjunction in the mutants. The S. cerevisiae
Spo11 protein, expressed by the ubiquitous ADH1
promoter in the host, exhibited DSB activity only in meiotic cells [55
]. In addition, the DSB activity of Spo11 required meiosis-specific and meiosis-nonspecific interactors [56
]. These findings indicated the presence of regulatory factors for Spo11 to exhibit the proper timing of the DSB activity. In this study, we found that DmSPO11
, expressed by the ubiquitous hsp83
promoter in the Drosophila dmspo11
mutant, allowed normal progression through meiosis. This suggested that, in Drosophila
, a regulator conferring the stage specificity of DmSPO11-induced DSB formation prevents the DNA scission by the DmSPO11 protein in the premeiotic or postmeiotic stage, but in early pachynema, another regulatory factor induces the active form of the DmSPO11 complex, leading to DSB formation for crossing-over. It is likely that the Drosophila
regulatory factors are unable to regulate the Arabidopsis
SPO11s. Thus, the DSB induction by plant SPO11s, expressed by the ubiquitous promoter hsp83
in the absence of plant SPO11-interacting protein factors, is unregulated and causes untimely DSBs in Drosophila
oocytes. The DSBs induced at inappropriate times are repaired in the mus301
-proficient fly as described, probably through inter-sister chromatid-homologous recombination or non-homologous end-joining, rather than inter-homologous chromosome-recombination, which is required for chiasma formation.
A recent paper reported that RNA interference for OsSPO11D (described as OsSPO11-2 in this paper) reduced pollen viability and seed setting rates in rice [57
]. Our finding of the specific expression of OsSPO11D
at the transcriptional level in the anthers containing meiotic pollens (Figure ) further strengthens the specific role of OsSPO11D in meiotic recombination, as in the case of SPO11 in the yeast S. cerevisiae
, which is expressed specifically at meiosis I under strict transcriptional regulation [58
More studies on OsSPO11D are necessary to prove this hypothesis. The bioassay described in this study would also be helpful to identify species-specific SPO11-interacting proteins that stimulate the SPO11 functions, the DSB forming activity specific for prophase I pachytene chromosomes, and the following interactions with repair enzymes. The biochemical characterization of purified SPO11s is necessary. We purified AtSPO11-1 [32
], and other plant SPO11s (Y. S. unpublished observations) in soluble forms, but they lacked detectable endonuclease activity. A recent report claimed that OsSPO11D exhibited DSB forming activity by itself in vitro
]. Further studies are needed to determine whether the purified SPO11s introduce the DSBs with either site- or sequence-specificity and are attached covalently at the termini of the cleavage sites, and which amino acid substitutions inactivate the SPO11s.