The dominance of alleles controlling self-incompatibility in Brassica pollen is regulated at the RNA level

Abstract

Self-incompatibility (SI) in Brassica is controlled sporophytically by the multiallelic S-locus. The SI phenotype of pollen in an S-heterozygote is determined by the relationship between the two S-haplotypes it carries, and dominant/recessive relationships often are observed between the two S-haplotypes. The S-locus protein 11 (SP11, also known as the S-locus cysteine-rich protein) gene has been cloned from many pollen-dominant S-haplotypes (class I) and shown to encode the pollen S-determinant. However, SP11 from pollen-recessive S-haplotypes (class II) has never been identified by homology-based cloning strategies, and how the dominant/recessive interactions between the two classes occur was not known. We report here the identification and molecular characterization of SP11s from six class II S-haplotypes of B. rapa and B. oleracea. Phylogenetic analysis revealed that the class II SP11s form a distinct group separated from class I SP11s. The promoter sequences and expression patterns of SP11s also were different between the two classes. The mRNA of class II SP11, which was detected predominantly in the anther tapetum in homozygotes, was not detected in the heterozygotes of class I and class II S-haplotypes, suggesting that the dominant/recessive relationships of pollen are regulated at the mRNA level of SP11s.

Figure 1.

Genomic Structure of the SRK Flanking Region of the S₆₀-Haplotype of B. rapa and the DNA Sequence of S₆₀-SP11. (A) Genomic organization of the SRK flanking region of the S₆₀-haplotype. Thick bars represent the λ phage clones covering the S₆₀-SP11/SRK₆₀ region. Arrows indicate the direction of transcription of each gene. Exons are indicated by solid boxes, and introns are indicated by dips. (B) Nucleotide sequence of S₆₀-SP11 and its deduced amino acid sequence. The noncoding regions are shown in lowercase letters, with the intron splice donor/acceptor sequences (gt and ag in boldface letters) demarcating the intron. The coding regions are shown in uppercase letters, with the underlined TAA sequence indicating the stop codon. Nucleotides are numbered from the transcription start site (a in boldface), which was determined in a 5′ rapid amplification of cDNA ends experiment. A putative TATA box and inverted sequence homologs of the CAAT motif and the LAT52/56 box are denoted in underlined boldface letters.

Figure 2.

Alignment of the Predicted Amino Acid Sequences of Six Allelic Variants of Class II SP11 and Class I S₉-SP11. Gaps (hyphens) were introduced to optimize the alignment. Alternative signal peptide cleavage sites of the two alternative transcriptional forms for class II SP11s are indicated by arrows. Conserved amino acid residues, eight cysteine residues (C1 to C8), a glycine residue (#), and an aromatic amino acid residue (†) are boxed. B.r., B. rapa; B.o., B. oleracea.

Figure 3.

Alternative Transcripts of Class II SP11s. (A) Exon-intron structure and alternative transcripts of S₆₀-SP11. The alternative splicing produces S₆₀-SP11 proteins with or without an alanine-leucine (AL) in the sequence. They are designated AL(+) and AL(−), respectively. Coding regions are shown in uppercase letters, and the amino acid residues they encode are shown below. Introns are shown in lowercase letters, and the intron splice donor/acceptor sequences (gt and ag) are shown in boldface letters. The expected signal peptide cleavage sites are indicated by arrows. (B) Alignment of the exon/intron junction of six class II SP11s. The intron splice donor/acceptor sequences (gt and ag) and an alternative possible acceptor sequence (AG) are shown in boldface letters. The alternative transcripts were detected experimentally in the S₆₀-, S₂₉-, and S₄₀-haplotypes, but not in the S₄₄-haplotype, of B. rapa. The S₂- and S₅-haplotypes of B. oleracea were not tested.

Figure 4.

DNA Gel Blot Analyses of the Class II SP11s. (A) DNA gel blot analyses of SP11s using the corresponding SP11 cDNAs as probes. Genomic DNA (1 μg each) of the S₂₉-, S₄₀-, S₄₄-, or S₆₀-haplotypes was digested with EcoRI (E), BamHI (B), or HindIII (H) and analyzed. The length of the DNA fragments in kilobases is indicated at left. (B) Restriction fragment length polymorphism linkage analysis of an F2 population segregating for S₂₉- and S₄₀-haplotypes. Genomic DNAs isolated from parental (P) plants homozygous for either the S₂₉- or S₄₀-haplotype, their F1 heterozygote, and F2 plants were digested with HindIII and hybridized with a mixed-SP11 probe (a mixture of S₂₉-SP11 and S₄₀-SP11 cDNA probes). The incompatibility phenotype of each plant (H, S₂₉S₄₀-heterozygote; 29, S₂₉S₂₉-homozygote; 40, S₄₀S₄₀-homozygote) was determined by pollination tests (Hatakeyama et al., 1998a). (C) PFGE gel blot analysis of three S-haplotypes of B. rapa. High molecular mass genomic DNA was digested with mluI and separated by PFGE. The DNA was hybridized with an SRK₂₉ kinase domain (SRK₂₉-KD) probe, an SLG₂₉ probe, or SP11-mix probes (a mixture of the S₂₉-SP11 and S₆₀-SP11 probes or the S₂₉-SP11 and S₄₄-SP11 probes). DNA size markers are shown at left in kilobases.

Figure 5.

Phylogenetic Tree of Eight Alleles of SP11 from Brassica Species. Bootstrap probabilities for clusters are shown as percentages. The bar under the tree indicates the number of amino acid substitutions per site. B.r. B. rapa; B.o., B. oleracea.

Figure 6.

RNA Gel Blot Analyses of the Class II SP11s. (A) RNA gel blot analysis of S₆₀-SP11 in an S₆₀S₆₀-homozygote of B. rapa. The total RNAs of anther (A), stigma (S), and leaf (L) were used. Numbers represent developmental stages of anther classified by bud sizes, where 5 = 4 to 5 mm, 6 = 5 to 7 mm, and 7 = 7 to 10 mm in length (Takayama et al., 2000a). (B) RNA gel blot analysis of S₆₀-SP11 and S₅₂-SP11 in an S₆₀S₆₀-homozygote (60) and an S₅₂S₆₀-heterozygote (H) of B. rapa. The same amount of total RNA of anthers (mixture or stages 5 to 7) was loaded in each lane. Similar results were obtained on three independent S₆₀S₆₀-homozygotes and S₅₂S₆₀-heterozygotes, and the results of a representative experiment are shown. (C) RNA gel blot analysis of S₄₀-SP11 in an S₄₀S₄₀-homozygote (40) and an S₃₅S₄₀-heterozygote (H) of B. rapa. The same amount of total RNA of anthers (stages 5 to 7, mixture) was loaded in each lane. The blot shown is representative of two independent experiments. The bottom gel of each blot shows ethidium bromide (EtBr)–stained rRNA bands.

Figure 7.

Analysis of S₆₀- and S₅₂-SP11 Expression in Anther Using in Situ Hybridization. Anther sections (stages 5 and 7) of an S₆₀S₆₀-homozygote or an S₅₂S₆₀-heterozygote were hybridized with S₆₀-SP11 or S₅₂-SP11 antisense riboprobes or their sense riboprobes. (A) and (B) S₆₀S₆₀-homozygote hybridized with an antisense S₆₀-SP11 probe. (C) and (D) S₆₀S₆₀-homozygote hybridized with a sense S₆₀-SP11 probe. (E) and (F) S₅₂S₆₀-heterozygote hybridized with an antisense S₆₀-SP11 probe. (G) and (H) S₅₂S₆₀-heterozygote hybridized with an antisense S₅₂-SP11 probe. Bar in (H) = 50 μm for (A) to (H).