Index References

Self-incompatibility systems prevent self-pollination and promote outbreeding. In Brassica, this system is sporophytically controlled by a single S locus with multiple alleles, and recognition of pollen is controlled by S haplotypes (designated S¹, S²,…, Sⁿ) (Bateman 1955). The molecules controlling the self-incompatibility have been identified as SP11 (S-locus protein 11, also known as SCR; Suzuki et al. 1999; Schopfer et al. 1999; Takayama et al. 2000) and SRK (S receptor kinase; Takasaki et al. 2000). An allele-specific interaction between SP11 and SRK has been proposed to cause activation of signaling cascade in the stigma (Kachroo et al. 2001; Takayama et al. 2001). Because the S locus appears to be a multigene complex, “S allele” is referred to as “S haplotype” (Nasrallah and Nasrallah 1993).

Based on the sequence diversity of the SP11 and SRK (S genes), S haplotypes are classified into two classes, class-I and class-II. The amino acid sequence identities among class-II SP11s are 63.2 to 94.6%, rather high compared with those of class-I SP11s which range from 19.5 to 76.1% (Shiba et al. 2002). Interestingly, class-I S haplotypes are dominant over those of class-II on the pollen (Nasrallah and Nasrallah 1993; Hatakeyama et al. 1998b). Thus, the pollen grain derived from S heterozygotes having the class-I and class-II S haplotypes shows S phenotype of class-I S haplotype. SP11 alleles frequently show dominance/recessive relationships. This relationship has been shown to be due to reduced mRNA abundance of the recessive SP11 allele in B. rapa (Shiba et al. 2002; Kakizaki et al. 2003) and in Arabidopsis lyrata (Kusaba et al. 2002). However, the mechanism of monoallelic expression of the SP11 gene has not been defined. It is important to compare genomic organization of dominant S haplotypes and recessive S haplotypes precisely, because dominance relationships are determined by the S locus itself. The detailed genomic organization of the S locus has already been reported in five class-I S haplotypes in B. rapa (Suzuki et al. 1999; Kimura et al. 2002; Shiba et al. 2003) and in B. napus (Cui et al. 1999). In the class-I S haplotypes, two S genes (SP11 and SRK) are located close to each other on the S locus, and the physical distance between genes is within 15-kb. However, their direction of transcription of S genes is diverted among S haplotypes (Takayama et al. 2000). On the other hand, the report of the S locus organization of the class-II S haplotypes is limited to S⁶⁰ haplotype of B. rapa (Fukai et al. 2003). The S²⁹, S⁴⁰ and S⁴⁴ haplotypes have been classified into class-II S haplotypes, and linear dominance hierarchies have been observed by reciprocal test pollinations (S⁴⁴ > S⁶⁰ > S⁴⁰ > S²⁹) (Hatakeyama et al. 1998b; Kakizaki et al. 2003). Moreover, in accordance with the linear dominance hierarchies observed among class-II S haplotypes, SP11 alleles can be dominant or recessive, with corresponding changes in expression, dependent on their allelic partner (Kakizaki et al. 2003). However, little is known about the genomic organization of the S-intergenic region in class-II S haplotypes without S⁶⁰ haplotype. To clarify the relation between dominance relationships and genomic organization around SP11, we investigated the S-intergenic region of three additional class-II S haplotypes.

At first, to characterize the S-intergenic region of three class-II S haplotypes S²⁹, S⁴⁰ and S⁴⁴, we screened the genomic libraries that were composed of lambda phage vector with each SP11 cDNA probe. We obtained the clones #29-4, #40-20, and #44-4, containing the 5’-upstream region of the SP11 gene of S²⁹, S⁴⁰, and S⁴⁴ haplotypes, respectively. The length of the DNA inserts in #29-4, #40-20, and #44-2 were estimated to be approximately 18-kb, 15-kb, and 14-kb, respectively, by long-PCR. From the shot-gun sequencing analysis of the three clones, the #29-4 clone contained the SP11-29 gene and the 5’-genomic region of the SRK²⁹ gene. The #29-4 clone contained the 5’-upstream region of SRK²⁹, whose nucleotide sequence was completely overlapped to the SRK²⁹ genomic clone determined by Hatakeyama et al. (1998a). In the case of the #40-20 clone, it contained the SP11-40 gene and a part of the regions encoding the S-domain and transmembrane domain of the SRK⁴⁰ gene. The #44-2 clone contained SP11-44 and a part of the region encoding the S-domain of SRK⁴⁴ (Fig. 1).

View Details

Fig. 1. Comparative map of the S-intergenic region among class-II S haplotypes. The genomic organization of the S-intergenic regions of the S29, S40, S44 and S60 haplotypes is represented. S60 haplotype was referred from Fukai et al. (2003; accession nos. AB097116). Arrows indicate the orientation of transcription. The restriction sites indicated are as follows: B, BamHI; E, EcoRI; S, SacI.

The full-length cDNA for SRK⁴⁰ and SRK⁴⁴ was obtained by RT-PCR using the information of the 5’-UTR sequences on the lambda clones. The nucleotide sequence of SRK⁴⁰ and SRK⁴⁴ has been deposited in GenBank, EMBL, and DDBJ databases (accession nos. AB211197 and AB211198, respectively). It was confirmed that the nucleotide sequences of cDNA clones for SRK⁴⁰ and SRK⁴⁴ completely corresponded to those of each genomic clone, respectively. From the comparison of the genomic sequence and cDNA sequence, sites of intron/exon junction were estimated (Fig. 1).

The promoter sequence elements that are required for sporophytic expression have already been identified in the class-I S⁹ haplotype (Shiba et al. 2001). In situ hybridization analysis revealed that class-II SP11 was expressed predominantly in anther tapetum (Shiba et al. 2002). We compared precisely the nucleotide sequences of the promoter region of four class-II S haplotypes of SP11 from the S²⁹, S⁴⁰, S⁴⁴ and S⁶⁰ haplotypes of B. rapa (designated SP11-29, SP11-40, SP11-44 and SP11-60 respectively) (Fig. 2). These sequences are highly conserved (72.2 to 85.8% identity) in the SP11 promoter from -250 to -1 bp. However, there is little similarity in the region upstream beyond -250 bp. Therefore, it is considered that the region of about -250 bp is important for the expression of SP11. However, it is difficult to identify the sequence similarity between class-I and class-II promoter regions (data not shown).

View Details

Fig. 2. Multiple alignment of the nucleotide sequences of the 5’-region of class-II SP11 (A) and short repetitive sequence (B). (A) The open box represents the putative TATA box. The arrow indicates the transcription start site of the SP11-29. The position of the translation start site is assigned as +1. (B) The open boxes indicate three repetitive sequence units (REP-A, REP-B, and REP-C). The number indicates the distance from initiation codon. The nucleotide sequence that has similarity between 5’-upstream of SP11 and short repetitive sequence is shown as a shaded box. Conserved sequences are indicated by asterisks. (C) Genomic structure of short repetitive sequence and SP11 promoter. The black arrows indicate the direction of repetitive sequence. The white arrow indicates the orientation of transcription of SP11.

As a next step, the sequence similarity of the S-intergenic region was surveyed by Harr plot analysis between four S haplotypes. The representative results between S²⁹ and S⁴⁰ haplotypes are shown in Fig. 3. The plots shows that the short repetitive sequences were conserved within S haplotypes at 1-kb upstream of the SP11 gene (REP29, REP40, REP44 and REP60; Fig. 3). The unit of the 64-bp sequence was contained three times in this repetitive sequence directly (REP-A, -B and -C; Fig. 2). This repetitive sequence had 63.8 to 91.0% similarity among four class-II S haplotypes. There was another direct repeat over 1 kb in S⁴⁰ and S⁴⁴ haplotypes, however, this repetitive sequence did not exist in the other two S haplotypes (S²⁹ and S⁶⁰; Fig. 1). The distance between S genes (SP11 to SRK) was 18.4-kb, 8.9-kb and 11.5-kb in S²⁹, S⁴⁰ and S⁴⁴ haplotypes, respectively. In the case of the S⁶⁰ haplotype, the distance has been determined as 6.6-kb (Fukai et al. 2003). From these data, the physical distance between S genes was relatively conserved among class-II S haplotypes. On the other hand, the distance between S genes was intensely diverted among class-I S haplotypes (Takayama et al. 2000). Except for the two S genes (SP11 and SRK) and the short repetitive sequences, we could not identify the sequence similarity among class-II S haplotypes. The sequence diversity throughout the SP11 and SRK region should contribute to the suppression of recombination within the S-intergenic region.

View Details

Fig. 3. Harr plot analysis on the S-intergenic region nucleotide sequences of S29 and S40 haplotypes. Arrows indicate the orientation of transcription. Plots were made with GENETYX-WIN software (Software development Co., Tokyo, Japan). The program was conditioned with a minimum length of 20, and matching of 18/20 or more. Each dot represents nucleotide shared by S29 and S40 haplotypes.

In the S-intergenic region of the S⁶⁰ haplotype, anther-specific non-coding RNA, called SAN1 (S locus Anther-expressed Non-coding RNA like-1), was predominantly expressed in anther (Fukai et al. 2003). In several cases, the expression of the non-coding RNA from one allele correlates with the repression of the flanking genes on the same allele (Sleutels et al. 2002), it is considered that SAN1 might be related to the dominance relationship on the pollen (Fukai et al. 2003). Therefore, we precisely surveyed the S-intergenic region of three additional S haplotypes to identify the homologous sequence to SAN1. However, we could not identify the orthologus gene of SAN1 in other class-II S haplotypes. Thus, this SAN1 might not be related to the dominance relationships in the pollen S gene.

Interestingly, the high sequence similarity was observed between the short repetitive sequence and 5’-upstream region of SP11 (Fig. 2). Especially in the case of S²⁹ haplotype, a part of REP-C shares 81.6% identity with the 5’-upstream region (-136 to -102 bp) of SP11-29. It has been known that such a repetitive element is modified by DNA methylation, representing a transposable element (Okamoto et al. 2001). Moreover, in one of the cruciferous plants, Arabidopsis suecica, which was an allotetraploid hybrid of A. thaliana and A. arenosa, the rRNA genes inherited from A. thaliana were transcriptionally silenced by selective DNA/histone methylation in spacer elements, whereas rRNA genes derived from A. arenosa were transcribed (Chen et al. 1998; Lawrence et al. 2004). Because the REP-C that exists in the S-intergenic region is similarly included in the SP11 promoter region, such a repetitive cluster might be the target of recessive allele specific modification as is the case as well in rRNA repression. Therefore, it is considered that it is important to analyze the state of modification of DNA and/or histone in the recessive allele of the SP11 promoter and repetitive sequences. From these analyses, the molecular mechanisms of the dominance relationship of SP11 in pollen will be discovered in the near future.

This work was supported in part by Grants-in-Aid for Special Research on priority Areas (Nos. 11238201 and 14360002) and Grants-in-Aid for the 21st Century Center of Excellence Program to Nara Institute of Science and Technology and Iwate University from the Ministry of Education, Culture, Sports, Science, and Technology of Japan and a Grant-in-Aid for Creative Scientific Research (No. 16G3016) from Japan Society for Promotion of Science (JSPS). T. K. and Y. T. are the recipients of a Research Fellowship of the JSPS for Young Scientists. The authors are grateful to Ayako Chiba, Yukiko Ohyama, and Hiroyuki Ishikawa (Iwate University) for technical assistance.