- Split View
-
Views
-
Cite
Cite
Satoru Moritoh, Daisuke Miki, Masahiro Akiyama, Mihoko Kawahara, Takeshi Izawa, Hisaji Maki, Ko Shimamoto, RNAi-mediated Silencing of OsGEN-L (OsGEN-like), a New Member of the RAD2/XPG Nuclease Family, Causes Male Sterility by Defect of Microspore Development in Rice, Plant and Cell Physiology, Volume 46, Issue 5, May 2005, Pages 699–715, https://doi.org/10.1093/pcp/pci090
- Share Icon Share
Abstract
We have cloned a new member of the RAD2/XPG nuclease family, OsGEN-L (OsGEN-like), from rice (Oryza sativa L.). OsGEN-L possesses two domains, the N- and I-regions, that are conserved in the RAD2/XPG nuclease family. Database searches and phylogenetic analyses revealed that OsGEN-L belongs to class 4 of the RAD2/XPG nuclease family, and OsGEN-L homologs were found in animals and higher plants. To elucidate the function of OsGEN-L, we generated rice OsGEN-L-RNAi transgenic plants in which OsGEN-L expression was silenced. Most of the OsGEN-L-RNAi plants displayed low fertility, and some of them were male-sterile. OsGEN-L-RNAi plants lacked mature pollen, resulting from a defect in early microspore development. A OsGEN-L–green fluorescent protein (GFP) fusion protein was localized in the nucleus, and the OsGEN-L promoter was specifically active in the anthers. Furthermore, a recombinant OsGEN-L protein possessed flap endonuclease activity and both single-stranded and double-stranded DNA-binding activities. Our results suggest that OsGEN-L plays an essential role in DNA metabolism required for early microspore development in rice.
Introduction
Nucleases play critical roles in nucleic acid metabolism in all organisms, and are involved in a variety of basic cellular and genetic processes (Mishra 2002). The RAD2/XPG nuclease family was the first eukaryotic nuclease family to be established, and is conserved among eukaryotes (Lieber 1997). It is characterized by two conserved domains, the N-terminal region (N-region) and the internal region (I-region) (Lieber 1997, Szankasi and Smith 1995). To date, three subgroups of the RAD2/XPG nuclease family, RAD2/XPG (class 1), FEN1/RAD27 (class 2) and EXO1 (class 3), have been functionally characterized (Nicolet et al. 1985, Madura and Prakash 1986, Scherly et al. 1993, Harrington and Lieber 1994b, Murray et al. 1994, Reagan et al. 1995, Szankasi and Smith 1995). It has been reported that they mainly function in DNA repair, DNA replication and recombination (Lieber 1997). These three nuclease subgroups have different substrate specificities and biological functions, but they have partially overlapping functions. Two new members of the RAD2/XPG nuclease family, OsSEND-1 (class 4) and DmGEN (class 5), have recently been isolated (Furukawa et al. 2003, Ishikawa et al. 2004).
RAD2/XPG (class 1) functions in nucleotide excision repair (NER) (Clarkson 2003). The XPG gene encodes a structure-specific endonuclease that nicks damaged DNA 3′ to the lesion during NER (O’Donovan et al. 1994). xpg-deficient humans suffer from the inherited disorder xeroderma pigmentosum, and some affected individuals also exhibit Cockayne’s syndrome (Nouspikel et al. 1997, Clarkson 2003). xpg-deficient mice exhibit postnatal growth failure and die prematurely (Harada et al. 1999). In mice, the nuclease activity of XPG is required only for NER (Tian et al. 2004), while the nuclease-independent function of the C-terminal region is important for the onset of Cockayne’s syndrome (Shiomi et al. 2004). However, the basis of Cockayne’s syndrome has remained unclear. Recently, RAD2, the Saccharomyces cerevisiae counterpart of XPG, was reported to be involved in promoting efficient RNA polymerase II transcription, suggesting that transcriptional defects are the underlying cause of Cockayne’s syndrome (Lee et al. 2002). The C-terminal region of RAD2 is important for transcription, because transcription was impaired by C-terminal truncation of RAD2, which is analogous to the human Cockayne’s syndrome mutation of XPG (Lee et al. 2002).
FEN1/RAD27 (class 2) functions in maturation of Okazaki fragments, base excision repair (BER) and maintenance of genome stability (Henneke et al. 2003, Liu et al. 2004). FEN1 has DNA flap endonuclease activity (Harrington and Lieber 1994a, Harrington and Lieber 1994b, Lieber 1997). Budding yeast rad27 mutants show various phenotypes, including sensitivity to genotoxic stress, conditional lethality, increased mutation frequency and trinucleotide repeat expansion, due to DNA repair and replication defects (Reagan et al. 1995, Freudenreich et al. 1998). fen1-deficient mice exhibit embryonic lethality (Kucherlapati et al. 2002, Larsen et al. 2003), and haploinsufficiency of FEN1 leads to the rapid progression of tumors in mice (Kucherlapati et al. 2002). As well as the function of FEN1/RAD27 in DNA repair and DNA replication, the worm FEN1 homolog CRN-1 is involved in apoptotic DNA degradation during normal developmental processes (Parrish et al. 2003).
EXO1 (class 3) was isolated as an exonuclease that is induced during meiosis of fission yeast (Szankasi and Smith 1992, Szankasi and Smith 1995). EXO1 is involved in mismatch repair and recombination (Szankasi and Smith 1995, Fiorentini et al. 1997, Tishkoff et al. 1997, Tran et al. 2001), and exo1-deficient mice show increased cancer susceptibility, male and female sterility due to a meiotic defect (Wei et al. 2003), and deficiencies in somatic hypermutation and class switch recombination (Bardwell et al. 2004).
OsSEND-1 (class 4) and DmGEN (class 5) were isolated based on the sequence similarity of the conserved N- and I-regions of their nuclease domains (Furukawa et al. 2003, Ishikawa et al. 2004). Detailed biochemical analysis of recombinant proteins revealed that OsSEND-1 has single-stranded endonuclease activity (Furukawa et al. 2003), while DmGEN has endo-exonuclease activities on single-stranded DNA, double-stranded DNA, gapped double-stranded DNA and nicked double-stranded DNA (Ishikawa et al. 2004).
Plants also encode RAD2/XPG family nucleases, and several genes belonging to this family have been studied previously. In rice, two FEN-1 homologs have been reported (Kimura et al. 2000, Kimura et al. 2003). OsFEN-1a has the same enzymatic properties as other eukaryotic FEN1 nucleases (Kimura et al. 2000), and can complement a budding yeast rad27 mutant (Kimura et al. 2003). The intracellular distribution of plant FEN1 protein was analyzed in tobacco BY-2 cells and meiotic lily cells (Kimura et al. 2001). A recently isolated class 4 member of the RAD2/XPG family, OsSEND-1, is preferentially expressed in proliferating tissues such as meristem, and is induced by UV and other DNA-damaging agents (Furukawa et al. 2003). However, the biological functions of these rice genes have not been reported. The Arabidopsis UV-sensitive mutant uvh3 has a mutation in a RAD2/XPG homolog (Liu et al. 2001), and the uvr1 mutant, which is allelic to uvh3, is also defective in the repair of 6–4 UV photoproducts (Britt et al. 1993, Liu et al. 2001), suggesting that the plant RAD2/XPG homolog has a conserved function among eukaryotes ranging from yeast to human.
Most male-sterile mutants are controlled by monogenic recessive genes and have defects in sporogenous tissues, tapetal cells, pollen mother cells, microspores and pollen, which appear at the pre-meiotic, meiotic and post-meiotic stages of anther and pollen development (Twell 2002, Singh 2003). Many male-sterile genes act during post-meiotic stages, particularly immediately after the tetrad stage (Singh 2003). The diploid tapetum surrounding haploid microspores plays an important role in microspore development by transporting nutrients to developing microspores (Twell 2002, Shivanna 2003). Therefore, cel–cell communications between diploid sporophytic cells and haploid gametophytic cells are important for anther and pollen development in plants. Recently, Wang et al. (2003) reported that RAFTIN, a sporophytically produced structural protein, is essential for pollen development and is targeted to microspore. Suppression of RAFTIN in rice by RNA interference (RNAi) nearly eliminates the seed set and this is due to male defect (Wang et al. 2003). However, the molecular mechanisms of microspore development are not yet understood in detail.
Previously we found a RAD2/XPG-like sequence during an analysis of flanking sequences of Ac-tagged rice lines (Enoki et al. 1999), and named this gene OsGEN-L (OsGEN-like). Sequence comparisons indicated that OsGEN-L and its homologs constitute a new subclass within the RAD2/XPG nuclease family. However, no functional information about these genes has been reported. To study the biological function of OsGEN-L, we generated OsGEN-L-silenced transgenic rice plants using an RNAi construct. We found that RNAi-mediated silencing of OsGEN-L causes male sterility in rice as a result of defective microspore development. The OsGEN-L promoter was specifically active in anthers. Biochemical analysis of recombinant OsGEN-L protein indicated that it possessed flap endonuclease activity and single-stranded and double-stranded DNA-binding activity. Our results suggest that OsGEN-L has an essential role in DNA metabolism that is required for male microspore development in rice.
Results
Cloning and characterization of the rice OsGEN-L gene
From analysis of Ac flanking sequences of rice Ac-tagged lines, we found a RAD2/XPG-related gene (Enoki et al. 1999) and named it OsGEN-L based on its sequence similarity to the recently reported DmGEN (Drosophila melanogaster XPG-like endonuclease) gene (Ishikawa et al. 2004). By screening cDNA and genomic DNA libraries using the Ac flanking sequence of the OsGEN-L gene, we isolated the full-length cDNA and corresponding genomic DNA of japonica rice cv. Toride 1. The OsGEN-L gene has eight exons and seven introns (Fig. 1A), and encodes a 629 amino acid protein (Fig. 1B). Southern blot analysis showed that OsGEN-L exists as a single-copy gene in rice (data not shown). Database searches revealed that the OsGEN-L gene is present in the Syngenta bacterial artificial chromosome (BAC) clone CL034874.88 and the fosmid clone OSJNOA273B05 (accession No. AP006859). The fosmid clone has been mapped at 78.5 cM on chromosome 9 by the Rice Genome Research Program (http://rgp.dna.affrc.go.jp). A full-length cDNA sequence from japonica rice cv. Nipponbare, which is almost identical to our OsGEN-L clone, has been registered in GenBank (accession No. AK063534), and a partial expressed sequence tag (EST) sequence (CA765363) from indica rice cv. IR64 has been registered.
Database searches revealed that the N-terminal region of OsGEN-L shows similarities to the N- and I-regions of the RAD2/XPG nuclease family, which are important for structure-specific nuclease activity (Fig. 1B). The N-region of OsGEN-L shows 32, 30, 23, 30 and 40% identity to the those of human XPG (class 1), human FEN1 (class 2), human EXO1 (class 3), rice OsSEND-1 (class 4) and Drosophila DmGEN (class 5), respectively (Fig. 1B); the I-region shows 30, 32, 22, 37 and 27% identity to those of human XPG (class 1), human FEN1 (class 2), human EXO1 (class 3), rice OsSEND-1 (class 4) and Drosophila DmGEN (class 5), respectively (Fig. 1B). In Fig. 1C, the amino acid sequences of the N- and I-regions of OsGEN-L, human XPG and FEN1 are aligned for comparison. Amino acid residues required for substrate cleavage by XPG and FEN1 (Shen et al. 1997, Constantinou et al. 1999) are also conserved in the OsGEN-L protein (marked by asterisks in Fig. 1C). These results indicate that the rice OsGEN-L gene encodes a member of the RAD2/XPG nuclease family.
OsGEN-L is a new member of the RAD2/XPG nuclease family
Among the completely sequenced genomes of various organisms, putative homologs of OsGEN-L are found in human, mouse, Drosophila and Arabidopsis. Full-length cDNAs of human (accession No. NM_182625), mouse (NM_177331), Drosophila (NM_139686), which is the same as DmGEN (Ishikawa et al. 2004), chicken (XM_419963) and Ciona intestinalis (AK116543) homologs have been deposited in the database, and Arabidopsis (At1g01880, NM_100069) and rat (XM_233966) homologs are present in annotated genomic sequences. However, no biological information concerning any of these genes has been reported. The RAD2/XPG nuclease family contains a conserved helix–hairpin–helix (HhH) class 2 DNA-binding motif (Pol1-type) (Doherty et al. 1996). A search of the NCBI conserved domain database (CDD v2.01) revealed that the I-region of OsGEN-L and its homologs is divided into two parts, namely the XPG-I (smart00484) and the HhH class 2 (cd00080) domains (Fig. 2A). The XPG-I domain of OsGEN-L showed 51 and 64% sequence identity with its human and Arabidopsis homologs, respectively (Fig. 2A). This domain is also conserved in the RAD2/XPG nuclease family. Outside the N- and I-regions, OsGEN-L also has moderate sequence identity to its human homolog (Fig. 2A, B). Phylogenetic analysis using the highly conserved XPG-I domain of the RAD2/XPG nuclease family suggested that, together with OsSEND-1, OsGEN-L and its homologs constitute class 4 of the RAD2/XPG nuclease family (Fig. 2C). Within the XPG-I domain, OsGEN-L showed the highest identity (46%) to OsSEND-1 among the known rice RAD2/XPG family members. Since OsSEND-1 belongs to class 4 of the RAD2/XPG nuclease family (Furukawa et al. 2003), OsGEN-L is a new member of this class (Fig. 2C). DmGEN was reported to be a class 5 member of the RAD2/XPG nuclease family (Ishikawa et al. 2004). Because the XPG-I domain of DmGEN is incomplete, we could not include DmGEN in phylogenetic comparisons of the RAD2/XPG nuclease family. However, phylogenetic analysis using the XPG-N domain classified DmGEN into class 4 of the RAD2/XPG nuclease family (data not shown). Interestingly, homologs of class 4 RAD2/XPG nucleases have not been found in S. cerevisiae, Schizosaccharomyces pombe or Caenorhabditis elegans.
Plant homologs of OsGEN-L were identified in tomato (Lycopersicon esculentum) (AW222911), soybean (Glycine max) (BU926894), sugarcane (Saccharum officinarum) (CF569989), Medicago truncatula (AW584311) and Brasssica napus (CD837439) (Fig. 2D). From the deduced amino acid sequences of partial ESTs and Arabidopsis genomic sequences, the XPG-N domain of OsGEN-L showed 91, 67, 65, 64, 63 and 55% identity to its homologs in sugarcane, soybean, B. napus, M. truncatula, Arabidopsis and tomato, respectively. These results suggest that OsGEN-L is conserved in higher plants.
RNAi-mediated silencing of OsGEN-L causes male sterility
To study the function of OsGEN-L in rice, we first examined OsGEN-L Ac homozygous insertion plants. The expression of OsGEN-L in Ac homozygous plants was not reduced, probably because the insertion is in an intron, and normally processed OsGEN-L mRNA was detected in these plants (Fig. 1A and data not shown). Therefore, we generated OsGEN-L-RNAi transgenic cell cultures and plants. To make the RNAi construct, a 683 bp OsGEN-L gene-specific fragment was used (Fig. 3A). RNA gel blot analysis of OsGEN-L-RNAi plants indicated that the expression of OsGEN-L in leaf was silenced in most of these transgenic lines (Fig. 3B). Furthermore, short-interfering RNA (siRNA) of OsGEN-L, a hallmark of RNAi, was detected in leaf RNA of OsGEN-L-RNAi lines (Fig. 3C). To determine whether the RNA silencing induced by the OsGEN-L-RNAi construct is specific to OsGEN-L, we first examined the expression of other members of the RAD2/XPG nuclease family in rice. Reverse transcription–polymerase chain reaction (RT–PCR) analysis revealed that OsGEN-L expression was suppressed, but expression of other RAD2/XPG nucleases was not suppressed, compared with wild-type plants (Fig. 3D). Secondly, we searched the rice genomic sequence database RiceBLAST (http://RiceBLAST.dna.affrc.go.jp/) using the 683 bp RNAi trigger region, and found that 19 genomic loci other than the OsGEN-L locus have short sequence identities to the 683 bp RNAi trigger region. Identities were 23 bp/24 bp–18 bp/18 bp. Using the RiceGAAS program (Rice Genome Automated Annotation System; http://RiceGAAS.dna.affrc.go.jp/rgadb/), we found that four loci hit exon sequences, six loci hit intron sequences and nine loci hit intergenic regions of the rice genome. RT–PCR analysis revealed that these four genes were not suppressed in OsGEN-L-RNAi plants (Fig. 3E). Among the six loci that match intron sequences, three possess full-length cDNAs. We examined their expression and found that these three genes were not suppressed in OsGEN-L-RNAi plants (data not shown). These results showed that the OsGEN-L-RNAi construct specifically suppressed the OsGEN-L gene in rice.
Vegetative growth of the OsGEN-L-RNAi plants appeared to be normal (data not shown), but the anther morphology at the time of flower opening was frequently abnormal (Fig. 4A). Iodine staining of these abnormal anthers revealed an absence of mature pollen grains (Fig. 4B). Many of the OsGEN-L-RNAi plants showed low fertility (Table 1), and two severely affected OsGEN-L-RNAi lines (i1 and i4) were almost sterile. Fertility in these two lines could be recovered by pollination with wild-type pollen (Table 1), suggesting that the sterility was due to male defects. Most of the OsGEN-L-RNAi lines showed suppression of OsGEN-L expression in leaves, and siRNA was detected in all lines, except i13 (Fig. 3C). However, two lines (i6 and i7) showed normal fertility. We performed real-time RT–PCR analysis using anther RNA of the OsGEN-L-RNAi lines to clarify further the relationship between OsGEN-L mRNA level and fertility. Anther RNA was isolated from the panicles of +10 to +11 cm of auricle length (distance between the auricles of the last two leaves). Low fertility lines (i1, i4 and i14) showed reduction of OsGEN-L mRNA expression in anthers (Table 1). In contrast, those OsGEN-L-RNAi lines having normal fertility (i6 and i7) showed OsGEN-L expression comparable with that of wild type (Table 1). The reduction of OsGEN-L mRNA expression in the anthers correlated with reduced fertility of the OsGEN-L-RNAi lines (R2 = 0.7451, Fig. 5). These results indicate that silencing of OsGEN-L expression in rice anthers causes male sterility.
OsGEN-L-RNAi plants have defects in early microspore development
To analyze the developmental defects that cause male sterility in these OsGEN-L-RNAi plants, we compared anther and pollen development in the empty vector control plants and OsGEN-L-RNAi plants. In early meiosis, meiocytes and anther layers including tapetum, middle layer and endothecium were formed normally in the OsGEN-L-RNAi plants (Fig. 4C, F). Tapetal cells appeared to be cytologically normal at the early microspore stage (Fig. 4D, G), and later disintegrated as in the tapetal cells of control plants (Fig. 4E, H). Before flowering, pollen was absent from the anthers in the OsGEN-L-RNAi plants (Fig. 4E, H). We examined male meiosis in the OsGEN-L-RNAi plants and observed normal metaphase I and normal dyad stages (Fig. 4I, J). Most male meiocytes proceeded to the tetrad stage with apparently normal distributions of DNA content (Fig. 4K). These results indicate that male meiosis is not defective in OsGEN-L-RNAi plants. The first abnormality appeared in the early uninucleate microspore stage. Early microspores failed to develop and were degraded (Fig. 4L). These OsGEN-L-RNAi phenotypes suggest that OsGEN-L may play a critical role in early microspore development in rice.
OsGEN-L is expressed in anthers
RT–PCR analysis showed that OsGEN-L mRNA was expressed in the roots, leaves and flower buds (Fig. 6A). To study the spatial pattern of OsGEN-L expression, a fusion of a 1.4 kb putative OsGEN-L promoter fragment to the β-glucuronidase (gus) marker gene was constructed and introduced into wild-type rice plants. Expression of OsGEN-L:gus was detected specifically in the anthers (Fig. 6B). GUS activity in pre-meiotic stages of anthers was also observed (data not shown). Subsequently, the GUS signals were undetectable during the meiotic stage but were later up-regulated during the post-meiotic stages (Fig. 6C). GUS activity was not detected in the anther filaments, female organs, leaves or roots in four independent transgenic lines (data not shown). Cross-sections of GUS-stained anthers showed that OsGEN-L was expressed in both anther layers and anther locules (Fig. 6D). These results indicate that the OsGEN-L promoter is active in rice anthers and is up-regulated during the post-meiotic stages of microspore development. The discrepancy between OsGEN-L expression in RT–PCR analysis and that in OsGEN-L:gus transgenic plants may be due to low-level expression in vegetative cells. Real-time RT–PCR analysis showed that the relative OsGEN-L mRNA abundance normalized by ubiquitin levels in anthers from the panicles of +10 cm auricle length was four times higher than that in mature leaves (Fig. 6E). However, it is also possible that some cis-elements required for expression in vegetative cells were not present in the 1.4 kb promoter region of our construct.
OsGEN-L–green fluorescent protein (GFP) fusion protein is localized in the nucleus
To study the subcelluar localization of OsGEN-L protein, a 35S: OsGEN-L-GFP fusion construct was made and transiently introduced into onion epidermal cells by particle bombardment. OsGEN-L–GFP was localized in the nucleus (Fig. 7B, D), in contrast to GFP alone (Fig. 7A, C), suggesting that OsGEN-L functions in the nucleus.
OsGEN-L protein possesses flap endonuclease activity and single-stranded and double-stranded DNA-binding activities
To investigate the biochemical functions of OsGEN-L, we cloned the OsGEN-L coding sequence using the vector pET30b (Novagen, Madison, WI, USA) and overexpressed a His-tagged recombinant OsGEN-L protein in Escherichia coli. Fig. 8A shows a schematic representation of the recombinant OsGEN-L protein, which has His and S tags at the N-terminus and a His tag at the C-terminus. The calculated molecular mass of the protein was 76.6 kDa. Although induction of a 76.6 kDa polypeptide was not observed by Coomassie brilliant blue staining (Fig. 8B), an 86 kDa protein band was detected in soluble cell extracts by Western blot using anti-His antibody (Invitrogen, Carlsbad, CA, USA), which specifically recognizes the polyhistidine at the C-terminus (data not shown). When the extracts were analyzed on blots using the S-protein (Novagen), which binds to the S tag sequences, 86 and 63 kDa protein bands were detected (data not shown). This indicated that the extracts contained full-length and C-terminally truncated OsGEN-L proteins. It is known that some proteins migrate differently from their calculated molecular weight on SDS–polyacrylamide gels. Both His-tagged proteins were eluted from a nickel affinity chromatography column, and the intact 86 kDa protein was purified further to near homogeneity by heparin affinity chromatography and Mono Q chromatography as shown by SDS–PAGE analysis (Fig. 8B).
Since the OsGEN-L protein has the conserved amino acids required for nuclease activity in the RAD2/XPG family, we examined the nuclease activities of the recombinant OsGEN-L protein with various kinds of DNA substrates: single-stranded M13 DNA and double-stranded plasmid DNA for endonuclease activities; single-stranded oligonucleotide and double-stranded oligonucleotide with overhangs for exonuclease activities; and bubble DNA and 5′-flap DNA for structure-specific endonuclease activities. When we examined flap endonuclease activity using 5′-flap DNA substrates, under the same reaction conditions used for OsFEN-1 (Kimura et al. 2000), OsGEN-L had weak flap endonuclease activity compared with OsFEN-1 (Fig. 8C). In equimolar to 100 ng OsFEN-1 protein (2 pmol), 250 ng of OsGEN-L (3 pmol) showed weak flap endonuclease activity (Fig. 8C). When we added excess OsGEN-L protein (500 and 1,000 ng) to the reaction, OsGEN-L clearly showed flap endonuclease activity (Fig. 8C). In our assay conditions (see Materials and Methods), other nuclease activities on our tested substrates were not detected in this recombinant OsGEN-L preparation (data not shown). When we used double-stranded plasmid DNA in the assay and assessed nuclease activity by native agarose gel electrophoresis, gel mobility-shifted patterns were observed instead of any endonucleolytic digestion (Fig. 8D). The shifted patterns depended on the concentration of the protein, and disappeared when 1% SDS (final concentration) was added to the reaction before electrophoresis. A weak mobility shift was also observed with single-stranded M13 DNA on agarose gel electrophoresis (data not shown). These results suggested that the OsGEN-L protein might bind to DNA in a sequence-non-specific manner. To confirm further the DNA-binding activity of the OsGEN-L protein, we examined the affinity between OsGEN-L and DNA cellulose resin. The protein bound to both single-stranded and double-stranded DNA cellulose resin at 0.1 M NaCl, and was released from the resin at salt concentrations higher than 0.2 M NaCl (Fig. 8E). These results show that the OsGEN-L protein possesses flap endonuclease activity and both single-stranded and double-stranded DNA-binding activities. Together with the nuclear localization of the protein, these biochemical data support the idea that the OsGEN-L protein may function in DNA metabolism for microspore development in rice.
Discussion
OsGEN-L encodes a new class 4 member of the RAD2/XPG nuclease family
In this study, we have isolated the OsGEN-L gene, which belongs to the RAD2/XPG nuclease family (Fig. 1B). Previously, OsSEND-1 was the only member of class 4 of the RAD2/XPG nuclease family (Furukawa et al. 2003). Our phylogenetic analysis and database searches indicated that OsGEN-L also belongs to class 4 (Fig. 2C). While OsSEND-1 may be a plant-specific member of class 4 of the RAD2/XPG nuclease family, OsGEN-L is conserved in eukaryotes ranging from plants to human. To our knowledge, this is the first report of an OsGEN-L subclass member in class 4 of the RAD2/XPG nuclease family in plants.
The Drosophila homolog (CG10670) of OsGEN-L, which was recently reported as DmGEN (Ishikawa et al. 2004), was initially described as one of two FEN1 homologs in Drosophila (Sekelsky et al. 2000). The closest member to OsGEN-L in other classes of the RAD2/XPG nuclease family is vertebrate FEN1, which belongs to the FEN1/RAD27 family (class 2). Therefore, we introduced the OsGEN-L gene into the budding yeast rad27 disruptant to test functional complementation. OsGEN-L did not complement the phenotypes of high-temperature lethality and methylmethane sulfonate (MMS) sensitivity of the rad27 mutant cells (data not shown), indicating that OsGEN-L may have a function that is distinct from the class 2 members. However, purified recombinant OsGEN-L has weak flap endonuclease activity. As well as FEN1 (class 2), XPG (class 1) and EXO1 (class 3) also have flap endonuclease activity. Functional differences between OsGEN-L and the other classes of the family are not clear.
The RAD2/XPG family nucleases participate in DNA repair processes. To study the function of OsGEN-L in DNA repair, we first checked the MMS sensitivity of OsGEN-L-RNAi cell cultures. In our experiment, these cell cultures did not show MMS sensitivity when compared with the wild type (data not shown), suggesting that OsGEN-L may not be involved in the repair of alkylating DNA damage. It has been reported for the other member of the class 4 family, OsSEND-1, that its mRNA was induced by MMS (Furukawa et al. 2003). Thus, it cannot be excluded that OsSEND-1 might be functionally redundant with OsGEN-L in alkylating DNA damage repair.
Essential role of OsGEN-L in early microspore development in rice
Many male-sterile mutants have been reported in rice (Kinoshita 1997). Tamaru and Kinoshita (1985) classified 29 rice male-sterile mutants into seven types of abnormalities based on histological observations. Rice male-sterile mutants ms24, ms25, ms34 and ms36 exhibit abnormalities at tetrad or early microspore stage (Kinoshita 1997, Tamaru and Kinoshita 1985), similar to OsGEN-L-RNAi plants. However, the genes responsible for these male-sterile mutant phenotypes are not known. Lee et al. (2004) recently reported that a T-DNA-tagged mutant of a rice cysteine protease gene, OsCP1, shows a low fertile phenotype that is similar to the OsGEN-L-RNAi plants, and the OsCP1 promoter is active in anthers. OsGEN-L and OsCP1 may therefore function in the same or a related pathway in anther and pollen development.
Severely affected OsGEN-L-RNAi plants showed almost complete sterility, attributable to defects in the early microspore stage (Table 1 and Fig. 4). This suggests that OsGEN-L may be a sporophytic gene controlling male fertility (McCormick 2004). In accordance with this idea, the OsGEN-L promoter is active in parental sporophytic tissues which are important for haploid microspore development. Interestingly, OsGEN-L-RNAi plants showed no meiotic and tapetal abnormalities. There is no effect in the female, and thus OsGEN-L probably is not a general meiotic factor. While it is possible that some subtle alterations in male meiosis and tapetal cells were present but undetectable in our experiments, silencing of the OsGEN-L gene does not appear to affect male meiosis or the overall anther layer structures.
GUS activity in the anthers of OsGEN-L:gus plants was up-regulated during post-meiotic stages. The timing appears to coincide with the onset of defects in early microspore development of OsGEN-L-RNAi plants. Interestingly, GUS activity was also observed in pre-meiotic stages of the anthers, but was then absent during the meiotic stage (Fig. 6C and data not shown). The OsGEN-L promoter was active in most of the diploid and haploid cells of the anthers, including the pollen mother cells, microspores and anther layers (Fig. 6D and data not shown). However, how sporophytic expression of OsGEN-L functions in haploid microspore development is still unclear. OsGEN-L mRNA or protein may move from diploid cells to haploid microspores, and the OsGEN-L protein may function directly in the microspore nucleus. Alternatively, expression of OsGEN-L in diploid cell nuclei may be important, and affect haploid microspore development indirectly through the interaction between diploid sporophytic cells and haploid gametophytic cells. Another possibility is that OsGEN-L protein expressed in pollen mother cells might function later in the early microspore nucleus. The OsGEN-L gene may also have a gametophytic function, because its promoter is also active in the haploid microspores and pollen. Arabidopsis Bcp1 has both sporophytic and gametophytic functions, and controls male fertility (Xu et al. 1995). Further studies on mRNA and protein expression patterns are needed to understand the molecular function of OsGEN-L in anther and pollen development.
RT–PCR analysis showed that OsGEN-L is expressed in vegetative tissues, including the leaves and roots (Fig. 6A). However, OsGEN-L-RNAi plants showed no abnormalities in vegetative cells, suggesting that OsGEN-L function may not be required for vegetative cells or perhaps that other genes of the rice RAD2/XPG family may have functional redundancy in vegetative cells. Some meiotic genes, such as SYN1, DIF1, Asy1, AHP2 and PAIR2, which have essential functions during meiosis, are also known to be expressed in vegetative tissues (Bai et al. 1999, Bhatt et al. 1999, Caryl et al. 2000, Schommer et al. 2003, Nonomura et al. 2004), although their functions in vegetative tissues are not clear.
Biochemical function of OsGEN-L protein
Expression of full-length RAD2/XPG nuclease family proteins in E. coli has not been reported, except for the class 2 FEN1/RAD27 proteins and the class 5 DmGEN protein. For example, full-length OsSEND-1 (class 4) is insoluble in E. coli (Furukawa et al. 2003). We successfully purified full-length recombinant OsGEN-L protein (class 4) from soluble fractions by using a low temperature (16°C) for cell growth in the bacterial pET expression system. The purified OsGEN-L protein possessed both single-stranded and double-stranded DNA-binding activities. The HhH class 2 motif, conserved in OsGEN-L, is known to be a structural basis for non-specific DNA binding (Doherty et al. 1996) and could be responsible for these DNA-binding activities. The other domains conserved in the RAD2/XPG family and found in the amino acid sequence of OsGEN-L were the N- and I-regions. We detected weak flap endonuclease activity (Fig. 8C), indicating that OsGEN-L possesses the conserved function of structure-specific nuclease activities of the RAD2/XPG nuclease family. DmGEN, the Drosophila homolog of OsGEN-L, was reported to have endo-exonuclese activities (Ishikawa et al. 2004). However, OsGEN-L did not have these nuclease activities in this study, and DmGEN did not have structure-specific endonuclease activities such as bubble structure and 5′-flap structure (Ishikawa et al. 2004). Further studies of OsGEN-L nuclease activities are needed to understand the biochemical function of OsGEN-L protein.
Through its interaction with DNA, OsGEN-L may exert some functions required for early microspore development. One of these could be a nuclease activity, possibly specialized for a certain substrate of anther cells. DNA metabolic events that need OsGEN-L nuclease activity in anther cells may be important for microspore development in rice. One of the substrates of OsGEN-L may be 5′-flap DNA. Identification of the true in vivo substrate of OsGEN-L is the next important question. OsGEN-L might also act non-catalytically, as its homolog XPG (class 1) has nuclease-independent functions. XPG is a well-known structure-specific nuclease involved in NER (O’Donovan et al. 1994), but other functions for XPG have also been proposed (Clarkson 2003). XPG has structural roles in the protein–DNA complex for the human NER reaction (Wakasugi et al. 1997), and human nuclease-deficient XPG acts as a cofactor for the NTH1 DNA glycosylase in BER (Bessho 1999, Klungland et al. 1999). It would therefore be intriguing to screen for proteins interacting with OsGEN-L.
Unique DNA metabolic events are observed in anther and pollen development. Pollen mother cells undergo DNA replication during pre-meiotic S phase, before meiotic division occurs. After meiosis, the individual cells of the tetrad are released as free microspores, and mitosis then occurs in the microspores to yield mature pollen (McCormick 2004). Tapetal cells are known to increase their DNA content by endomitosis and endoreduplication (Shivanna 2003, Scott et al. 2004). During the tapetum degeneration, programmed cell death occurs. Since OsGEN-L-RNAi plants did not display any abnormalities in meiosis and tapetum in our observations, it seems that OsGEN-L might play a role in the mitotic process in early microspores, after the pollen mother cells complete meiosis. Although the plant endonuclease synthesized by barley uninucleate microspores has been partially purified (Marchetti et al. 2001), it is not known if this nuclease is related to OsGEN-L. Recently, the Arabidopsis HhH DNA glycosylase DEMETER has been reported to activate maternal allele expression of MEDEA (MEA) in the central cell of the female gametophyte by nicking the MEA promoter region (Choi et al. 2002, Choi et al. 2004). OsGEN-L may regulate the expression of male genes which are important for microspore development through interactions with DNA. RAD2, the budding yeast homolog of XPG, has been shown to facilitate efficient transcription of some genes by RNA polymerase II, independent of its nuclease activity (Lee et al. 2002).
Many higher plants possess conserved OsGEN-L homologs, as shown in this study. Further studies of the molecular functions of this new protein will help us to understand the DNA metabolism required for early microspore development in higher plants, about which little is currently known.
Materials and Methods
Plant materials and growth conditions
Japonica rice cv. Toride 1 (Oryza sativa L.) was used as wild type. Transgenic rice plants were generated using Agrobacterium-mediated transformation of rice calli, by the method of Hiei et al. (1994). Plants were regenerated from transformed calli by selecting for hygromycin resistance. Regenerated transgenic plants were grown in a greenhouse at 28°C during the day and 23°C at night. Wild-type plants for RNA isolation were grown in a field in Ikoma, Nara, Japan or in climate chambers at 30°C under continuous light conditions.
Isolation of the OsGEN-L gene
A cDNA library made from Toride 1 root RNA (kindly provided by Dr. Tomoyuki Yamaya, Tohoku University, Japan) and a Toride 1 genomic library made using the λBlueSTAR Vector System (Novagen, Madison, WI, USA) were screened with an OsGEN-L-specific probe amplified from genomic DNA of an OsGEN-LAc insertion line using primers Ac5(0) (5′-CCCATCCTACTTTCATCCCTG-3′) and RAD2-1 (5′-CCAACAAGAAGTGCCATTGCTACC-3′) (Enoki et al. 1999). Isolated genomic clones were sequenced using the Kilo-Sequence Deletion Kit (TaKaRa, Shiga, Japan).
Sequence analysis
DNA sequences were determined by ABI Prism 373, 310 or 3100 sequencers (Applied Biosystems, Foster City, CA, USA) using dye-teminator chemistry. Sequence data were analyzed using GeneWorks software version 2.5.1 (Intelligenetics, Mountain View, CA, USA). For phylogenetic analysis, predicted amino acid sequences were entered into the NCBI conserved domain search database (CDD v2.01, http://www.ncbi.nlm.nih.gov/Structure/cdd/wrpsb.cgi) to identify the XPG-I domain (smart00484). The dendrogram of XPG-I domains of the RAD2/XPG nuclease family was generated using the CLUSTALW program.
Vector constructs
To make the OsGEN-L-RNAi construct, a 683 bp fragment of OsGEN-L cDNA was amplified using primers RAD2-31 (5′-ATTATATGAAATTGGTAAAG-3′) and RAD2-32 (5′-ATTCTGTCTGCTTGCTAGGT-3′), and subcloned into pGEM-T (Promega Corporation, Madison, WI, USA). This construct was denoted OsGEN-L i pGEM-T and its sequence was verified. A SacII–SpeI fragment and a SalI–ApaI fragment of the OsGEN-L i pGEM-T were inserted into the pGUS27 vector (Miki and Shimamoto 2004). A SacI–KpnI fragment of this construct, in which 683 bp of OsGEN-L cDNA sequences were placed upstream and downstream of the gus linker fragment in inverse orientations, was cloned into the p2K-1+ binary vector harboring the maize Ubq1 promoter and the nos terminator (Miki and Shimamoto 2004).
To make the OsGEN-L promoter:gus fusion construct, we amplified a 1.4 kb sequence upstream of the OsGEN-L gene, based on the sequence of the Syngenta BAC clone CL034874.88, from japonica rice cv. Nipponbare genomic DNA using primers RAD2-51 (5′-GCTCTAGACCACAGCTCATTACCACACATCTG-3′) and RAD2-50 (5′-CGGGATCCCTCTCTTCTTCCCCGGCGAC-3′) containing XbaI and BamHI restriction sites (underlined in primer sequences) and cloned the product into the binary vector (Yokoi et al. 1997).
To make the OsGEN-L-sGFP fusion construct, a HindIII–SacI fragment containing the 35S promoter and the Gateway cassette (Invitrogen, Carlsbad, CA, USA) of the pGWB5 binary vector (kindly provided by Dr. Tuyoshi Nakagawa, Shimane University, Japan) was cloned into pUC12. The resulting construct was denoted 35S-attR-sGFP. An OsGEN-L open reading frame (ORF) fragment, which was amplified using primers RAD2-56 (5′-CACCATGGGGGTGGGGGGAAGCTT-3′) and RAD2-57 (5′-GTCGAAGAGGAGGCGTCGTC-3′), was cloned into pENTR/D-TOPO (Invitrogen) and a sequence-verified clone was transferred into 35S-attR-sGFP by the LR Clonase (Invitrogen) reaction according to the manufacturer’s protocol. The resulting construct was denoted p35S-OsGEN-L-sGFP. Control p35S-GFP vector was used as previously described (Miki and Shimamoto 2004) after removing the intron sequences by blunt-end ligation of BamHI restriction sites.
To construct an expression plasmid for the His-tagged OsGEN-L protein, the OsGEN-L coding fragment was amplified from the OsGEN-L cDNA clone by PCR using RAD2-47 (5′-CCATGGGGGTGGGGGGAAGC-3′) and RAD2-48 (5′-CTCGAGGTCGAAGAGGAGGCGTCG-3′) primers containing NcoI and XhoI restriction sites (underlined), respectively. The amplified OsGEN-L fragment was ligated to the pGEM-T vector (Promega) having single T-overhangs, and sequenced to avoid PCR errors. The verified plasmid clone was digested with both NcoI and XhoI, and the OsGEN-L fragment was inserted into the corresponding sites of the pET30b vector (Novagen) which are located downstream of the lactose promoter under the control of LacI. The resulting plasmid was denoted OsGEN-L-pET30b.
RNA gel blot analysis and siRNA detection
RNA gel blot analysis and siRNA detection were performed as described previously (Miki and Shimamoto 2004). For gel blot analysis of OsGEN-L mRNA, 20 µg of total RNA was transferred after electrophoresis to a Hybond N+ nylon membrane (Amersham Biosciences, Little Chalfont, UK) and hybridized with 32P-radiolabeled C-terminal OsGEN-L cDNA probes amplified by primers RAD2-11 (5′-CGCAATGTACATCCATTGGCCGA-3′) and RAD2-40 (5′-GCTAGTGAATGCTGCATTTCC-3′) using standard procedures (Sambrook and Russell 2001). siRNA was detected according to a published protocol (Hamilton and Baulcombe 1999). Probes used for detection of OsGEN-L siRNA were DNA fragments amplified by primers RAD2-31 (5′-ATTATATGAAATTGGTAAAG-3′) and RAD2-32 (5′-ATTCTGTCTGCTTGCTAGGT-3′) and labeled with [32P]dCTP. Probes used for detection of control 5S rRNA were 89mer sense (5′-GATCCCATCAGAACTCCGAAGTTAAGCGTGCTTGGGCGAGAGTAGTACTAGGATGGGTGACCTCCTGGGAAGTCCTCGTGTTGCATCCC-3′) and antisense (5′-GGGATGCAACACGAGGACTTCCCAGGAGGTCACCCATCCTAGTACTACTCTCGCCCAAGCACGCTTAACTTCGGAGTTCTGATGGGATC-3′) oligonucleotides labeled with [32P]dCTP. All hybridization signals were detected by a BAS-2500 Bioimaging analyzer (Fuji, Tokyo, Japan).
Histological analysis
Young panicles or flowers were fixed with 2% glutaraldehyde in 0.1 M sodium phosphate buffer, pH 7.4, overnight at 4°C, and dehydrated in an ethanol series. For plastic sections, the samples were embedded in Technovit 7100 resin (Hereaus Kuzer, Wehrheim, Germany), and cut into 1 µm sections using an ULTRACUT UCT ultramicrotome (Leica, Heidelberg, Germany). Sections were stained with toluidine blue O and photographed using a BX50 microscope (Olympus, Tokyo, Japan). Rice anthers were stained with iodine according to a published method (Terada et al. 2000), and photographed under an Axioskop microscope (Zeiss, Jena, Germany) using dark-field illumination. Flowers were fixed in ethanol : acetic acid solution (3 : 1) and stored at 4°C. Tetrads and male meiocytes from fixed anthers were stained with 10 mg ml–1 4′,6-diamidino-2-phenylindole (DAPI), mounted on a glass slide with Vectorshield (Vector Laboratories, Burlingam, CA, USA), and photographed using a microscope. Microspores from fixed anthers were stained with acetocarmine solution (Wako Pure Chemical, Osaka, Japan). For GUS staining, plant materials were infiltrated with a staining solution [100 mM phosphate buffer (pH 7.0), 100 mM EDTA, 0.5 mM potassium ferrocyanide, 0.5 mM potassium ferricyanide, 0.1% Triton X-100] under vacuum for a total of 15 min, and X-Gluc was added to a final concentration of 0.4 mM. Incubation was carried out for 48 h at 37°C and the samples were fixed in a solution of 50% ethanol, 5% acetic acid and 3.7% formaldehyde, and stored at 4°C until use. GUS-stained immature flowers were pre-cleared in a trichloroacetaldehyde monohydrate : glycerol : water solution (8 g : 1 ml : 2 ml) or directly observed under a SZX7 or SZH10 microscope (Olympus). For sections of GUS-stained materials, the fixed samples were embedded in Technovit 7100 resin (Hereaus Kuzer), and cut into 5 µm sections using an ultramicrotome. Sections were photographed under a microscope using dark-field illumination. All photos were taken by DXM1200 microscopic digital camera (Nikon, Tokyo, Japan).
RT–PCR and real-time RT–PCR
Total RNA was isolated from root, leaf and flower buds, according to a published method (Kawasaki et al. 1996). A 1 µg aliquot of total RNA treated with RNase-free DNase I Amplification Grade (Invitrogen) was reverse-transcribed using Random 9mer Primer (TaKaRa) and RAV2 reverse transcriptase (TaKaRa). Of the synthesized first-strand cDNA, 20% was used for PCR analysis with different sets of gene-specific primers: for OsGEN-L, RAD2-21 (5′-CGCGAAATGGCTTCATCTCA-3′) and RAD2-11 (5′-CGCAATGTACATCCATTGGCCGA-3′); for actin, ActF (5′-TGGGTTCGCCGGAGATGAT-3′) and ActR (5′-CTTGGGTACGACCACTGGC-3′). The PCR amplification conditions were: after incubation 94°C for 1 min/94°C, 30 s/58°C, 30 s/72°C, 1 min; 28 cycles for OsGEN-L; and 23 cycles for actin.
For RT–PCR analysis of OsGEN-L-RNAi construct specificity, total RNA was isolated using the RNeasy Plant Mini Kit (QIAGEN, Hilden, Germany). A 1–2 µg aliquot of total RNA treated with RNase-free DNase I Amplification Grade (Invitrogen) was reverse-transcribed using a poly(T) primer and SuperScript II (Invitrogen). Of the synthesized first-strand cDNA, 1.25–2.5% was used for PCR analysis with different sets of gene-specific primers: for OsGEN-L, RAD2-21 and RAD2-11 primers were used; for ubiquitin, UBQ-F (5′-CCAGGACAAGATGATCTGCC-3′) and UBQ-R (5′-AAGAAGCTGAAGCATCCAGC-3′) primers were used; for OsXPG, OsXPG-1 (5′-CCTCCATCAATGCAGTTGG-3′) and OsXPG-2 (5′-GATGAAGAGAACTCCGCAGC-3′) primers were used; for OsFEN-1a, OsFEN1a-1 (5′-GAAGCAGAAGCAGAATGTGC-3′) and OsFEN1a-2 (5′-AAGAAGGATTCGAGCCTTCC-3′) primers were used; for OsFEN-1b, OsFEN1b-1 (5′-TTCCAGACTTCCTGTGGACC-3′) and OsFEN1b-2 (5′-TCACCGTTCAACACTCAAGC-3′) primers were used; for OsEXO1, OsEXO1-5 (5′-CACACAACACAGAGCCATCC-3′) and OsEXO1-6 (5′-TCAGCGCAGCAAACTTATCC-3′) primers were used; for OsSEND-1, OsSEND1-1 (5′-ATTGCTGAGAGAGAACTTCGG-3′) and OsSEND1-2 (5′-TACCAGCCACATTGACATCC-3′) primers were used; for gene 1, AK071791–1 (5′-CGATGAGAGCAGCAAGAAGC-3′) and AK071791–2 (5′-CAATCCAATGCACGATTGG-3′) primers were used; for gene 2, AK107043–1 (5′-CGATGAGAGCAGCAAGAAGC-3′) and AK107043–2 (5′-CAATCCAATGCACGATTGG-3′) primers were used; for gene 3, AK111811–1 (5′-TGTGACATTGCTTCTGGAGC-3′) and AK111811–2 (5′-GCACAACTAGAGTGGAAGGC-3′) primers were used; and for gene 4, Pulipase-1 (5′-GACATTCACATATGAATCAC-3′) and Pulipase-2 (5′-CATCCAACGAGGAAGACGTG-3′) primers were used. The PCR amplification conditions were: after incubation 94°C for 2 min/94°C, 30 s/55°C, 30 s/72°C, 1 min; 28 cycles for OsGEN-L, OsXPG and gene 3; 24 cycles for ubiquitin; 27 cycles for OsFEN-1a and OsFEN-1b; 32 cycles for OsSEND-1 and OsEXO1; 34 cycles for gene 4; 37 cycles for gene 1 and gene 2.
For real-time RT–PCR analysis, SYBR Green PCR master mix (Applied Biosystems) and the ABI Prism 7000 Sequence Detection System (Applied Biosystems) were used according to the manufacturer’s instructions. RNA isolation from anthers was carried out using the RNeasy Plant Mini Kit (QIAGEN). A 1 µg aliquot of total RNA treated with DNase I (Invitrogen) was reverse-transcribed using Random 9-mer Primer (TaKaRa) and SuperScript II (Invitrogen). Of the synthesized first-strand cDNA, 5% was used for PCR analysis with different sets of gene-specific primers: for OsGEN-L, RAD2-real-F (5′-CGGAGGCCAGTGCGG-3′) and RAD2-real-R (5′-CTCCACGTATCGCGTGAACTT-3′) primers were used; for ubiquitin, real-ubq-F (5′-AACCAGCTGAGGCCCAAGA-3′) and real-ubq-R (5′-ACGATTGATTTAACCAGTCCATGA-3′) primers were used.
Transient expression in onion epidermal cells
A transient expression assay was performed by particle bombardment of onion epidermal cells using the PDS/He biolistic particle delivery system (Bio-Rad, Hercules, CA, USA). A 6 µg aliquot of control p35S-GFP plasmids and 15 µg of p35S-OsGEN-L-sGFP plasmids were bombarded. Plasmid DNA isolation was carried out using a Plasmid Maxi Kit (QIAGEN). After an overnight incubation at 30°C, the onion epidermis was examined under an LSM510 confocal laser scanning microscope (Zeiss).
Overexpression and purification of recombinant OsGEN-L
All manipulations were carried out at 0–4°C unless otherwise noted. Escherichia coli BL21-CodonPlus (DE3)-RIL cells (Stratagene, La Jolla, USA) carrying OsRAD-pET30b plasmid DNA were grown in LB broth containing 50 µg ml–1 kanamycin and 34 µg ml–1 chloramphenicol at 16°C. When the absorbance of the culture at 600 nm was about 0.7, 1 mM isopropyl-β-d-thiogalactopyranoside IPTG (final concentration) was added to induce OsGEN-L gene expression. After further incubation at 16°C for 4 h, the cells were harvested by centrifugation. The cell pellets (4 g wet cells) were suspended in 20 ml of buffer A [50 mM Na2HPO4 (pH 7.2), 0.5 M NaCl, 10 mM mercaptoethanol] and disrupted by sonication. The sonicated cell suspension was incubated on ice in 0.04% Nonidet P-40 for 20 min. To obtain cleared cell extracts, the solution was centrifuged at 12,000×g for 15 min followed by ultracentrifugation at 100,000×g for 10 min. Imidazole (pH 7.0) was added to the cleared extracts to a final concentration of 80 mM. The extracts were loaded onto a Ni2+-charged 1 ml HiTrap chelating column (Amersham Biosciences) equilibrated in buffer B [20 mM Na2HPO4 (pH 7.2), 0.5 M NaCl, 0.04% Nonidet P-40, 1 mM mercaptoethanol] containing 80 mM imidazole. The column was washed with 5 ml of buffer B containing 80 mM imidazole. Bound proteins were eluted with buffer B containing 300 mM imidazole, and the pooled protein fractions were dialyzed against buffer C [20 mM Tris-HCl (pH 7.4), 0.3 M NaCl, 10% glycerol, 0.4% Nonidet P-40, 1 mM dithiothreitol (DTT)]. Half of the dialysate (1.8 mg of protein) was loaded onto a 1 ml Hitrap heparin column (Amersham Biosciences) equilibrated with buffer D [20 mM Na2HPO4 (pH 7.2), 10% glycerol, 0.04% Nonidet P-40, 1 mM DTT] containing 0.3 M NaCl. The column was washed with 10 ml of buffer D containing 0.3 M NaCl, and the proteins were eluted with a 10 ml linear gradient of 0.3–1.0 M NaCl in buffer D. The pooled fractions were dialyzed against buffer E [40 mM Tricine-HCl (pH 8.8), 10% glycerol, 0.04% Nonidet P-40, 1 mM DTT] containing 0.2 M NaCl. The dialysate (1 mg of protein) was loaded onto a Mono Q HR5/5 column (Amersham Biosciences) equilibrated with buffer E containing 0.2 M NaCl. After washing the column with 5 ml of buffer E containing 0.2 M NaCl, the proteins were eluted with a 20 ml linear gradient of 0.2–1.0 M NaCl in buffer E. Peak fractions containing the intact protein were pooled and dialyzed against buffer F [40 mM HEPES-KOH (pH 7.4), 10% glycerol, 0.04% Nonidet P-40, 1 mM DTT] containing 0.2 M NaCl. Aliquots of the purified OsGEN-L protein (0.2 mg) were frozen in liquid nitrogen and stored at –80°C.
Endonuclease and DNA binding assays
The reaction mixture (20 µl) contained 20 mM MOPS (pH 7.4), 100 mM NaCl, 10 mM MgCl2, 0.1 mg ml–1 bovine serum albumin, 1 mM DTT and 100 ng of double-stranded plasmid DNA. His-tagged OsGEN-L proteins were added as indicated. After incubation at 30°C for 2 h, half of the reaction (10 µl) was terminated with 2 µl of stop buffer (0.05% bromophenol blue, 0.05% xylene cyanol, 30% glycerol, 60 mM EDTA) and the other half was terminated with stop buffer containing 6% SDS. The reactions were loaded on a 0.8% agarose gel, and electrophoresis was carried out in TAE (40 mM Tris-acetate, 1 mM EDTA) at room temperature. DNA was stained with ethidium bromide and visualized by ultraviolet light. When the 5′ end-labeled oligonucleotide was used as a substrate, the reaction was terminated with 40% formamide and analyzed on 15% sequencing gels, followed by detection with a BAS-2500 Bioimaging analyzer (Fuji). The flap endonuclease activity assay was carried out as described previously (Kimura et al. 2000), as was the bubble structure cutting activity assay (O’Donovan et al. 1994).
Single-stranded and double-stranded DNA cellulose resin were purchased from Sigma (St Louis, MO, USA). DNA cellulose resin was washed with buffer F containing 0.1 M NaCl. The resin (20 µl) was mixed with 40 µl of OsGEN-L protein (10 µg) in the same buffer and incubated for 30 min on ice, centrifuged at 400×g, and then suspended sequentially in 20 µl of buffer F containing 0.1, 0.2, 0.4, 0.6 and 0.8 M NaCl. OsGEN-L protein in each supernatant was separated on an 8% SDS–polyacrylamide gel and visualized by Coomassie brilliant blue staining.
Acknowledgments
We thank Ms. Sawako Kohashi for technical assistance, Dr. Chieko Saito for suggestions on histological techniques, Dr. Tomoyuki Yamaya for providing the rice cDNA library, Dr. Junko Kyozuka for the RNAi construct, Dr. Ken-Ichi Nonomura and Dr. Nori Kurata for technical advice about meiosis and the histological staining method, Dr. Norihiko Tamaru for advice about microspore development, Dr. Tsuyoshi Nakagawa for providing the pGWB5 vector, Mr. Yoshihiko Yagi for the nuclease assay, Dr. Seisuke Kimura, Mr. Yukinobu Uchiyama and Dr. Kengo Sakaguchi for providing OsFEN-1 protein and technical advice about the nuclease assay, Dr. Tsukasa Matsunaga for providing material for the nuclease assay, Dr. Ikuko Yonamine and Dr. Keiko Umezu for the yeast rad27 complementation experiment, Dr. Anne Britt for valuable suggestions, Dr. Tsutomu Kawasaki, Dr. Masayuki Isshiki, Dr. Shuji Yokoi and Dr. Ian Smith for critical reading of the manuscript, and Dr. Hann Ling Wong for laser confocal microscopy and manuscript preparation. We also thank the members of the Laboratory of Plant Molecular Genetics at Nara Institute of Science and Technology for technical assistance, comments and participation in discussions.
Present address: Division of Molecular Genetics, National Institute for Basic Biology, National Institute of Natural Sciences, 38 Nishigonaka, Myodaiji, Okazaki, Aichi, 444-8585 Japan.
Present address: Laboratory of Applied Plant Genomics, National Institute of Agrobiological Sciences, Tsukuba, Ibaraki, 305-8602 Japan.
The nucleotide sequences reported in this paper have been submitted to EMBL/GenBank/DDBJ under accession numbers AB158320 and AB194139 for OsGEN-L cDNA and OsGEN-L genomic DNA, respectively.
Line | Fertility (%) | Seeds/flowers | Cross-pollination (seeds/flowers) | OsGEN-L mRNA level (%) |
Wild-type a | 82.1 | 358/435 | 100 | |
Wild-type b | 63 | 63/100 | ||
i1 | 0.88 | 5/570 | (22/44) | 50 |
i2 | 0 | 0/133 | ||
i4 | 2.2 | 20/893 | (15/46) | 7 |
i5 | 0 | 0/190 | ||
i6 | 67.2 | 314/467 | 155 | |
i7 | 69.4 | 302/435 | 102 | |
i8 | 29.4 | 68/231 | 73 | |
i11 | 0 | 0/63 | ||
i13 | 83.5 | 81/97 | 114 | |
i14 | 0 | 0/101 | 31 | |
i15 | 0 | 0/91 |
Line | Fertility (%) | Seeds/flowers | Cross-pollination (seeds/flowers) | OsGEN-L mRNA level (%) |
Wild-type a | 82.1 | 358/435 | 100 | |
Wild-type b | 63 | 63/100 | ||
i1 | 0.88 | 5/570 | (22/44) | 50 |
i2 | 0 | 0/133 | ||
i4 | 2.2 | 20/893 | (15/46) | 7 |
i5 | 0 | 0/190 | ||
i6 | 67.2 | 314/467 | 155 | |
i7 | 69.4 | 302/435 | 102 | |
i8 | 29.4 | 68/231 | 73 | |
i11 | 0 | 0/63 | ||
i13 | 83.5 | 81/97 | 114 | |
i14 | 0 | 0/101 | 31 | |
i15 | 0 | 0/91 |
Fertility was determined in 2002. To determine the fertility of OsGEN-L-RNAi plants, the number of seeds was divided by the total number of flowers. Cross-pollination between OsGEN-L-RNAi female plants and wild-type pollen was done in 2001 and 2002. The OsGEN-L mRNA level was determined by real-time RT–PCR analysis from anther RNA in 2003. Wild-type and T0 generation OsGEN-L-RNAi plants were maintained from 2001 to 2003. Some OsGEN-L-RNAi lines (i2 and i11) were lost in 2003, but the same tendency of fertility of other OsGEN-L-RNAi lines was confirmed for 3 years from 2001 to 2003.
Line | Fertility (%) | Seeds/flowers | Cross-pollination (seeds/flowers) | OsGEN-L mRNA level (%) |
Wild-type a | 82.1 | 358/435 | 100 | |
Wild-type b | 63 | 63/100 | ||
i1 | 0.88 | 5/570 | (22/44) | 50 |
i2 | 0 | 0/133 | ||
i4 | 2.2 | 20/893 | (15/46) | 7 |
i5 | 0 | 0/190 | ||
i6 | 67.2 | 314/467 | 155 | |
i7 | 69.4 | 302/435 | 102 | |
i8 | 29.4 | 68/231 | 73 | |
i11 | 0 | 0/63 | ||
i13 | 83.5 | 81/97 | 114 | |
i14 | 0 | 0/101 | 31 | |
i15 | 0 | 0/91 |
Line | Fertility (%) | Seeds/flowers | Cross-pollination (seeds/flowers) | OsGEN-L mRNA level (%) |
Wild-type a | 82.1 | 358/435 | 100 | |
Wild-type b | 63 | 63/100 | ||
i1 | 0.88 | 5/570 | (22/44) | 50 |
i2 | 0 | 0/133 | ||
i4 | 2.2 | 20/893 | (15/46) | 7 |
i5 | 0 | 0/190 | ||
i6 | 67.2 | 314/467 | 155 | |
i7 | 69.4 | 302/435 | 102 | |
i8 | 29.4 | 68/231 | 73 | |
i11 | 0 | 0/63 | ||
i13 | 83.5 | 81/97 | 114 | |
i14 | 0 | 0/101 | 31 | |
i15 | 0 | 0/91 |
Fertility was determined in 2002. To determine the fertility of OsGEN-L-RNAi plants, the number of seeds was divided by the total number of flowers. Cross-pollination between OsGEN-L-RNAi female plants and wild-type pollen was done in 2001 and 2002. The OsGEN-L mRNA level was determined by real-time RT–PCR analysis from anther RNA in 2003. Wild-type and T0 generation OsGEN-L-RNAi plants were maintained from 2001 to 2003. Some OsGEN-L-RNAi lines (i2 and i11) were lost in 2003, but the same tendency of fertility of other OsGEN-L-RNAi lines was confirmed for 3 years from 2001 to 2003.
Abbreviations
- BAC
bacterial artificial chromosome
- BER
base excision repair
- DAPI
4′,6-diamidino-2-phenylindole
- DTT
dithiothreitol
- EST
expressed sequence tag
- GFP
green fluorescent protein
- GUS
β-glucuronidase
- HhH
helix–hairpin–helix
- IPTG
isopropyl β-d-thiogalactopyranoside
- MMS
methylmethane sulfonate
- NER
nucleotide excision repair
- RNAi
RNA interference
- RT–PCR
reverse transcription–polymerase chain reaction
- siRNA
short-interfering RNA
References
Bai, X., Peirson, B.N., Dong, F., Xue, C. and Makaroff, C.A. (
Bardwell, P.D., Woo, C.J., Wei, K., Li, Z., Martin, A., Sack, S.Z., Parris, T., Edelmann, W. and Scharff, M.D. (
Bessho, T. (
Bhatt, A., Lister, C., Page, T., Fransz, P., Findlay, K., Jones, G.H., Dickinson, H.G. and Dean, C. (
Britt, A.B., Chen, J.-J., Wykoff, D. and Mitchell, D. (
Caryl, A.P., Armstrong, S.J., Jones, G.H. and Franklin, F.C. (
Choi, Y., Gehring, M., Johnson, L., Hannon, M., Harada, J.J., Goldberg, R.B., Jacobsen, S.E. and Fischer, R.L. (
Choi, Y., Harada, J.J., Goldberg, R.B. and Fischer, R.L. (
Constantinou, A., Gunz, D., Evans, E., Lalle, P., Bates, P.A., Wood, R.D. and Clarkson, S.G. (
Doherty, A.J., Serpell, L.C. and Ponting, C.P. (
Enoki, H., Izawa, T., Kawahara, M., Komatsu, M., Koh, S., Kyozuka, J. and Shimamoto, K. (
Fiorentini, P., Huang, K.N., Tishkoff, D.X., Kolodner, R.D. and Symington, L.S. (
Freudenreich, C.H., Kantrow, S.M. and Zakian, V.A. (
Furukawa, T., Kimura, S., Ishibashi, T., Mori, Y., Hashimoto, J. and Sakaguchi, K. (
Hamilton, A.J. and Baulcombe, D.C. (
Harada, Y.N., Shiomi, N., Koike, M., Ikawa, M., Okabe, M., et al. (
Harrington, J.J. and Lieber, M.R. (
Harrington, J.J. and Lieber, M.R. (
Henneke, G., Friedrich-Heineken, E. and Hubscher, U. (
Hiei, Y., Ohta, S., Komari, T. and Kumashiro, T. (
Ishikawa, G., Kanai, Y., Takata, K-I., Takeuchi, R., Shimanouchi, K., Ruike, T., Furukawa, T., Kimura, S. and Sakaguchi, K. (
Kawasaki, T., Mizuno, K., Shimada, H., Satoh, H., Kishimoto, N., Okumura, S., Ichikawa, N. and Baba, T. (
Kimura, S., Furukawa, T., Kasai, N., Mori, Y., Kitamoto, H.K., Sugawara, F., Hashimoto, J. and Sakaguchi, K. (
Kimura, S., Suzuki, T., Yanagawa, Y., Yamamoto, T., Nakagawa, H., Tanaka, I., Hashimoto, J. and Sakaguchi, K. (
Kimura, S., Ueda, T., Hatanaka, M., Takenouchi, M., Hashimoto, J. and Sakaguchi, K. (
Kinoshita, T. (
Klungland, A., Hoss, M., Gunz, D., Constantinou, A., Clarkson, S.G., Doetsch, P.W., Bolton, P.H., Wood, R.D. and Lindahl, T. (
Kucherlapati, M., Yang, K., Kuraguchi, M., Zhao, J., Lia, M., et al. (
Larsen, E., Gran, C., Saether, B.E., Seeberg, E. and Klungland, A. (
Lee, S., Jung, K.H., An, G. and Chung, Y.Y. (
Lee, S.K., Yu, S.L., Prakash, L. and Prakash, S. (
Lieber, M.R. (
Liu, Y., Kao, H.I., and Bambara, R.A. (
Liu, Z., Hall, J.D. and Mount, D.W. (
Madura, K. and Prakash, S. (
Marchetti, S., Zaina, G., Chiaba, C., Pappalardo, C. and Pitotti, A. (
Miki, D. and Shimamoto, K. (
Mishra, N.C. (
Murray, J.M., Tavassoli, M., Al-Harithy, R., Sheldrick, K.S., Lehmann, A.R., Carr, A.M. and Watts, F.Z. (
Nicolet, C.M., Chenevert, J.M. and Friedberg, E.C. (
Nonomura, K.-I., Nakano, M., Murata, K., Miyoshi, K., Eiguchi, M., Miyao, A., Hirochika, H. and Kurata, N. (
Nouspikel, T., Lalle, P., Leadon, S.A., Cooper, P.K. and Clarkson, S.G. (
O’Donovan, A., Davies, A.A., Moggs, J.G., West, S.C. and Wood, R.D. (
Parrish, J.Z., Yang, C., Shen, B. and Xue, D. (
Reagan, M.S., Pittenger, C., Siede, W. and Friedberg, E.C. (
Sambrook, J. and Russell, D.W. (
Scherly, D., Nouspikel, T., Corlet, J., Ucla, C., Bairoch, A. and Clarkson, S.G. (
Schommer, C., Beven, A., Lawrenson, T., Shaw, P. and Sablowski, R. (
Scott, R.J., Spielman, M. and Dickinson, H.G. (
Sekelsky, J.J., Brodsky, M.H. and Burtis, K.C. (
Shen, B., Nolan, J.P., Sklar, L.A. and Park, M.S. (
Shiomi, N., Kito, S., Oyama, M., Matsunaga, T., Harada, Y.N., Ikawa, M., Okabe, M. and Shiomi, T. (
Shivanna, K.R. (
Singh, R.J. (
Szankasi, P. and Smith, G.R. (
Szankasi, P. and Smith, G.R. (
Tamaru, N. and Kinoshita, T. (
Terada, R., Nakajima, M., Isshiki, M., Okagaki, R.J., Wessler, S.R. and Shimamoto, K. (
Tian, M., Jones, D.A., Smith, M., Shinkura, R. and Alt, F.W. (
Tishkoff, D.X., Boerger, A.L., Bertrand, P., Filosi, N., Gaida, G.M., Kane, M.F. and Kolodner, R.D. (
Tran, P.T., Simon, J.A. and Liskay, R.M. (
Twell, D. (
Wakasugi, M., Reardon, J.T., and Sancar, A. (
Wang, A., Xia, Q., Xie, W., Datla, R. and Selvaraj, G. (
Wei, K., Clark, A.B., Wong, E., Kane, M.F., Mazur, D.J., et al. (
Xu, H., Knox, R.B., Taylor, P.E. and Singh, M.B. (