| Edited by Yoshibumi Komeda * Corresponding author. E-mail: takayama@bs.naist.jp** Corresponding author. E-mail: nabe@iwate-u.ac.jp |
Many flowering plants have mechanisms to avoid self-fertilization. One such sophisticated mechanism is self-incompatibility (SI), which is defined as the inability of a fertile hermaphrodite plant to produce zygotes after self-pollination. In the Brassica SI system, self-pollen is rejected at the surface of the papilla cells due to recognition of self and non-self pollen. This highly regulated system is sporophytically controlled by a single S locus with multiple alleles (Bateman 1955). At the Brassica S-locus region, three highly polymorphic genes, SRK (a gene for S receptor kinase), SLG (a gene for S locus glycoprotein), and SP11/SCR (a gene for S locus protein 11 or S locus cysteine-rich protein), are located (Suzuki et al. 1999; Takayama et al. 2000). The SLG gene encodes a secreted glycoprotein (Takayama et al. 1987). The SRK gene encodes a transmembrane receptor-like kinase, which consists of an extracellular SLG-like domain (S domain), a transmembrane domain, and a cytoplasmic kinase domain (Stein et al. 1991). Both SLG and SRK are specifically expressed in stigma tissues (Stein et al. 1991; Watanabe et al. 1994). Based on a gain-of-function experiment, it has been demonstrated that SRK is a female determinant of the SI recognition (Takasaki et al. 2000). For SLG’s function, it has been reported that SLG enhanced the SI reaction in SRK9-introduced transgenic Brassica (Takasaki et al. 2000), and co-expression of SLG stabilized SRK in transgenic plants (Dixit et al. 2000). SP11/SCR encodes a small cysteine-rich protein (Schopfer et al. 1999; Suzuki et al. 1999), and was specifically expressed in anther tapetum and immature pollen (Takayama et al. 2000; Shiba et al. 2001; Iwano et al. 2002). From gain-of-function and bioassay experiments, it has been demonstrated that SP11/SCR is a male determinant of the SI recognition (Schopfer et al. 1999; Takayama et al. 2000; Shiba et al. 2001). In addition, recent studies showed that SP11/SCR was released from a self-pollen coat, and could bind SRK on the stigma surface to activate its kinase in an allele-specific manner (Kachroo et al. 2001; Takayama et al. 2001; Iwano et al. 2002).
To date, more than 50 S haplotypes have been identified in Brassica species (Nou et al. 1993; Ockendon 2000; Ruffio-Chable and Gaude 2001). Based on the sequence diversity of SI genes, S haplotypes are classified into class I and class II; the amino acid sequence similarity of SLG and SRK is almost 65% between classes and 80–90% within classes (Watanabe et al. 2001; Watanabe et al. 2003). This classification has also been applied to highly polymorphic SP11/SCR, because SP11/SCR seems to have co-evolved with SLG and SRK (Watanabe et al. 2000, Shiba et al. 2002). As for the SI phenotype, class-I S haplotypes are generally dominant over class-II S haplotypes on the pollen side (Thomson and Taylor 1966; Hatakeyama et al. 1998b). In general, class-I S haplotypes exhibit a strong SI phonotype, and class-II S haplotypes show a weak or leaky SI phenotype (Nasrallah and Nasrallah 1993, Cabrillac et al. 1999).
In B. oleracea, three class-II S haplotypes (S2, S5, and S15) were identified (Thompson and Taylor 1966), and two different forms of SLG (SLGA and SLGB) were associated with these class-II S haplotypes (Cabrillac et al. 1999; Miege et al. 2001). SLGA encodes both soluble and membrane-anchored forms (Tantikanjana et al. 1993; Cabrillac et al. 1999). SLGB encodes a soluble form (Cabrillac et al. 1999), which has also been characterized in class-II S haplotypes of B. rapa (Hatakeyama et al. 1998a). It has been reported that two groups of the S2 haplotypes could be distinguished depending on whether SRK was associated with SLGA or SLGB (Miege et al. 2001), and that S5 haplotype carried only SLGB (Cabrillac et al. 1999). It is noteworthy that both SLGA and SLGB were identified in the self-compatible B. oleracea line, P57Sc, carrying the S15 haplotype, and was thought to have evolved via a duplication event (Gaude et al. 1993; Cabrillac et al. 1999). Although SLG is not necessary to recognize self-pollen in the SI system (Okazaki et al. 1999; Takasaki et al. 2000; T Suzuki et al. 2000; G Suzuki et al. 2003), the duplication which occurred in the S15 hap-lotype is interesting from the viewpoint of the evolutionary history of the SI genes and its weak SI phenotype. In this study, we characterized the genomic organization and transcription of the SP11 gene in the self-incompatible S15 haplotype of B. oleracea, and found that duplicated S15-SP11 genes produced a long alternative transcript containing three SP11 sequences.
B. oleracea S15 homozygote, which was classified as a class-II S haplotype, was used. The self-incompatible line of S15 was derived from inbred line #102 (NongWoo Seed Co., Seoul, Korea) and inbred line TCS15-S (Takii Seed Co. Ltd, Kyoto, Japan).
Total DNA was extracted from young leaf tissues of B. oleracea according to Park et al. (2002). To determine the genomic structure of SP11s, a set of polymerase chain reaction (PCR) primers (5’-GCGAAAATCTTATATACTCATAAG-3’ and 5’-TTCGTTGATCAATTATGATT-3’) designated from the 5’- and 3’-ends of the coding region of S60-SP11, respectively, was used (Shiba et al. 2002). In the case of the determination of the 5’-upstream region of SP11, another set of PCR primers (C2SP11pro-2F: 5’-CGCAGGAAAAGAAGAATTGGAT-3’, C2SP11pro-2R: 5’-CAGGGCATTTCCAAGCT-CCCAC-3’) was used. The 3’-end of the SP11 was determined by 3’-rapid amplification of the cDNA ends (RACE) method (Shiba et al. 2002). To analyze the transcript of SP11, reverse transcription (RT)-PCR and 5’-RACE were performed. In the RT-PCR, a set of primers (S15-SP11F: 5’- ACTAGATGTGGGAGCTTGGAAATG-3’, S15-SP11R: 5’- CTCGAGTTCGTTGATCAATTATGATT-3’), which corresponded to the 2nd exon and 3’-untlansrated region (UTR), respectively, were used. PCR condition and 5’-RACE were done according to Shiba et al. (2002). The amplified fragments were subcloned into the vector pGEM-T Easy (Promega, Madison, USA), and sequenced by a dideoxy-chain termination method (Watanabe et al. 2000). Sequence data were analyzed with GENETYX (Software Development, Tokyo, Japan) and DNASIS (Hitachi Software, Tokyo, Japan).
Isolation of total RNA, electrophoresis of denatured RNA samples and blotting to a nylon membrane (Hybond N+; Amersham-Pharmacia) were performed as described by Shiba et al. (2002). Five μg of total RNA was loaded in each lane. 32P-labeled S15-SP11a probe was prepared as described above. After hybridization of each probe, membranes were washed in 0.1x SSPE, 0.1% SDS at 65°C for 30 min, and exposed on X-ray film. Equal loading of total RNA was assessed by ethidium bromide staining of rRNA bands.
Two kinds of genomic fragments (0.4 and 0.9 kb in length) containing S15-SP11 were amplified from genomic DNA of the B. oleracea S15 homozygote by polymerase chain reaction (PCR) with a set of primers, which could specifically amplify class-II SP11 genes (Shiba et al. 2002). We cloned and sequ-enced the two fragments, and identified that the 0.4-kb fragment contained one SP11 gene (S15-SP11a), and the 0.9-kb fragment contained two SP11 genes (S15-SP11a and S15-SP11b). The two amplified fragments were derived from the same genomic region, because they contained the same S15-SP11a sequences. Thus, we found that two SP11 genes, S15-SP11a and S15-SP11b, occurred in tandem with the same orientation at the S-locus region of the S15 haplotype (Fig. 1). From the comparison of the nucleotide sequence of S15-SP11a with that of S15-SP11b, the sequence identities of the regions corresponding to exon 1, intron 1, and exon 2 were 65.2%, 100%, and 98.5%, respectively. The low similarity observed in exon 1 was due to the deletion in S15-SP11b, but not sequence diversity between S15-SP11a and S15-SP11b, as described below. The N-terminal hydrophobic amino acid sequ-ences of S15-SP11a and S15-SP11b should be corresponded to their putative signal peptides. Eight cysteine residues, which were characteristics of other SP11s, were also conserved in S15-SP11a and S15-SP11b (Fig. 1). Furthermore, the 5’-upstream region of S15-SP11a was amplified and sequenced; a putative TATA-box element and sequ-ences similar to the 5’-promoter region of class-II S60-SP11 were identified (Shiba et al. 2002, Fig. 1, Fig. 2), suggesting that S15-SP11a was normally transcribed in anthers. A 239-bp intergenic region, separating S15-SP11a and S15-SP11b, did not show significant homology to the 5’-promoter regions of S15-SP11a and other class-II SP11s, indicating that this intergenic region did not contain the promoter elements for S15-SP11b.
 View Details | Fig. 1. Nucleotide sequence and deduced amino acid sequence of the SP11 genes in the S15 haplotype of B. oleracea. The coding regions are shown in uppercase letters. The non-coding regions are shown in lowercase letters. Nucleotides are numbered from the transcription start site (a in boldface), which was determined in a 5’-RACE experiment. A putative TATA box is denoted in underlined boldface letters. Direct AACATC repeats observed at the position of insertion is denoted in underlined italic letters. The enclosed C with square is conserved cysteine residue in other SP11s. The deduced signal peptide of SP11 is denoted in underlined boldface letters. The nucleotide sequence of S15-SP11 has been deposited in GenBank, EMBL, and DDBJ databases (accession no. AB176545). |
 View Details | Fig. 2. The 5’-promoter region of S15-SP11a is significantly similar to that of class-II S60-SP11. Nucleotide sequences cor-responding to 5’-upstream regions of S15-SP11a (class II) of B. oleracea and S60-SP11 (class II) of B. rapa are aligned. Nucleotides are numbered from the transcription start site. A putative TATA box and inverted sequence homologs of the CAAT motif and the LAT52/56 box are denoted in underlined boldface letters. B.o., Brassica oleracea; B.r., Brassica rapa. |
In order to analyze transcription of the duplicated SP11 genes in the S15 haplotype, RNA gel blot with an S15-SP11a probe was performed. Two kinds of SP11 transcripts (1.4 kb and 0.65 kb in length) were specifically expressed in anther tissues, and were developmentally up-regulated from 3 to 6 days before anthesis (Fig. 3). The amount of the 0.65-kb transcript was significantly higher than that of the 1.4-kb transcript.
 View Details | Fig. 3. Two types of SP11 transcripts observed in the S15 haplotype. RNA gel blot analysis of S15-SP11 in an SI line of the S15 homozygote of B. oleracea. The total RNAs of anther (A), stigma (S), and leaf (L) were used. Developmental stages of the anther were classified as follows: stage 2, 6–7 days before anthesis; stage 3, 5–6 days before anthesis; stage 4, 4–5 days before anthesis; stage 5, 3–4 days before anthesis; stage 6, 3days before anthesis. The bottom gel of each blot shows ethidium bromide (EtBr)-stained rRNA bands. |
Furthermore, we characterized the SP11 transcripts by sequencing analysis of reverse transcription (RT)-PCR and 5’-rapid amplification of the cDNA ends (5’-RACE) products. The amplified SP11 transcripts were classified into two groups based on the length of the products. In the first group of the ~0.65-kb transcript, only an S15-SP11a region was transcribed, and the intron 1 was spliced out (Fig. 4). The length of the 3’-untlansrated region (UTR) sequence varied among the cDNA clones; interestingly, the exon-1 region of S15-SP11b, encoding the signal peptide, was also co-transcribed in some cases (Fig. 4). These ~0.65-kb S15-SP11a transcripts might have been mainly transcribed, as shown in the RNA gel blot analysis (Fig. 3). In fact, the cDNA sequence identical to S15-SP11a has been reported as BoSP11-15 (Sato et al. 2003). In the second group of the 1.4-kb transcript, both full-length S15-SP11a and S15-SP11b were co-transcribed, and their intron-1 regions were spliced out (Fig. 4). The 1.4-kb transcripts contained the region corresponding to the exon-1 region of another SP11 as the 3’-UTR sequence which was also the case of the ~0.65-kb transcripts (Fig. 4), suggesting that the third SP11 gene (S15-SP11b’) was located 182-bp downstream of S15-SP11b with the same orientation (Fig. 1). The exon 1 and a part of the intron 1 of S15-SP11b’ observed in the 1.4-kb transcript were completely identical to those of S15-SP11b, suggesting that a duplication event occurred to produce S15-SP11b and S15-SP11b’ (described below). The existence of the full-length S15-SP11b’ in the S15 genome was not confirmed in this study.
 View Details | Fig. 4. Schematic representation of the S15-SP11 transcripts observed in sequencing analysis of RT-PCR and 5’-RACE products. Filled and open boxes represent exon-1 (without 5’-UTR) and exon-2 (without 3’-UTR) regions, respectively. Arrows indicate direction of transcription of the SP11 genes. |
Transcripts containing only S15-SP11b could not be observed in our RT-PCR and 5’-RACE analyses. This may have been due to the lack of promoter elements in the intergenic region between S15-SP11a and S15-SP11b as described above. The S15-SP11b might have been a pseudogene whose transcript is only transcribed as the 1.4-kb mRNA with S15-SP11a. Similarly, S15-SP11b’ might also have been a pseudogene, because there were no promoter elements in the intergenic region between S15-SP11b and S15-SP11b’. Thus, we concluded that the SI response in the S15 haplotype was regulated by the SRK15 and S15-SP11a.
Sequence similarity within the S15-SP11 region was analyzed by Harr plot (Fig. 5). The plot clearly showed the long stretch of similar sequences within the S15-SP11 region, suggesting that gene duplication events occurred in the S15 haplotype. This finding is quite interesting because duplicated SLGs were observed in a self-compatible S15 line (Cabrillac et al. 1999). The 611-bp region encompassing from the exon-1 of S15-SP11a to the intron-1 of S15-SP11b was quite similar to the 554-bp region encompassing from the exon-1 of S15-SP11b to the intron-1 of S15-SP11b’; only three nucleotide differences were observed between the two sequ-ences except for a 57-bp insertion associated with direct AACATC repeats in the 611-bp region (Fig. 1). In the S15 haplotype, a 502-bp region of 100% sequence identity is shared by SRK15 and SLGA15 (Cabrillac et al. 1999), suggesting that the two genes have been involved in a gene conversion event. Similarly, the quite similar ca. 600-bp sequences observed in the SP11 region might have been produced by recent duplication of the SP11 genes or gene conversion of the duplicated SP11 genes, as discussed by Suzuki et al. (1997) and Cabrillac et al. (1999). These evidences indicated that the gene duplication and/or conversion events have frequently occurred in the S15 haplotype. The duplicated SP11 genes in the S15 haplotype were shown to produce 1.4-kb alternative transcript containing full-length S15-SP11a, S15-SP11b, and a part of S15-SP11b’. Although there remains the possibility that this alternative transcript might be able to produce functional SP11 protein as observed in certain transcripts containing short open reading frames (Campalans et al. 2004; Casson et al. 2002), this unusual form of transcript is expected to be non-functional. Generally, the plants with class-II S-haplotype are known to exhibit weak or leaky SI phenotype. In B. oleracea, some S15 homozygous lines (e.g. P57Sc) with complete self-compatible phenotype have been generated from SI founder plants (Gaude et al. 1993). It should be an interesting experiment to analyze the relationship between the amount of alternative SP11 transcript and SI phenotype in these lines.
 View Details | Fig. 5. Duplication in the S15-SP11 region. Sequence similarities within the S15-SP11 region is shown by Harr plot which was made with DNASIS V3.7 software. Within the plot, each dot represents a match of >90% within a window of 10 nucleotides. |
This work was supported in a part by Grants-in-Aid for Special Research on Priority Areas (B) (Nos. 11238021 and 11238025), for Scientific Research (A) (No. 15208013), for Scientific Research (B) (Nos. 14360002 and 14360066), for Young Scientists (B) (No. 14760218) from the Ministry of Education, Culture, Sports, Science and Technology (MEXT), Japan, by a grant from the Research for the Future Program (JSPS-RFTF 00L01605) from the Japan Society for the Promotion of Science, by a grant from Intelligent Cosmos Academic Foundation, Japan, and R & D Promotion Center for Agriculture and Forestry, Korea. The authors are grateful to Koji Sakamoto, Takii Seed Co. Ltd., for providing the S15 inbred line. We thank Kanako Iwasaki, Naoko Wada, Hitomi Ichikawa, Hanae Sugita, Taduru Ueda, and Hiroko Sato (NAIST), Ayako Chiba, Hiroyuki Ishikawa, and Yukiko Ohyama (Iwate University) for their help.
|
Index