Molecular characterization of the self-incompatibility (S-)locus of Petunia inflata

Self-incompatibility (SI) in the Solanaceae is controlled by the highly polymorphic S-locus. Two tightly linked genes at the S-locus control specific SI interactions: S-RNase determining pistil S-haplotype specificity and the yet unidentified pollen S-gene determining pollen S-haplotype specificity. To characterize the Solanaceae S-locus and identify the pollen S-gene by map-based cloning, a high-resolution genetic map of the S-locus was constructed by recombination analysis of 1205 segregating plants using 13 S-linked pollen-expressed genes as markers. The P. inflata S-locus was mapped to within a 0.25-cM region delimited by markers 3.16 and G221 . Nine of the markers were tightly linked to the S-locus. Allelic cDNA sequence comparison or genomic DNA blotting revealed that these nine marker genes exhibited low degrees of allelic sequence diversity, suggesting that none of them was likely to be the pollen S-gene. Chromosome walking in the S2-locus region yielded ten separate contigs, which collectively spanned 4.4 Mb. To characterize the S -locus, S1- and S 2-alleles of genomic regions encompassing 3.16 and G221 (located outside the S-locus), as well as G261 and S-RNase (located at the S-locus), were sequenced. In contrast to S-RNase, these three genes exhibited similarly low degrees of allelic sequence diversity in the introns and flanking regions as well as the coding regions. To identify additional genes at the S-locus, a 328-kb sequence containing S2-RNase was analyzed. Repetitive sequences, including transposons, constituted ∼76% of the region. Except for S 2-RNase, the only predicted non-transposon-like gene whose deduced amino acid sequence was similar to known proteins in the database was an F-box gene, located 161 kb downstream of S2-RNase . This gene was named PiSLF2 ( S2-allele of P. inflata S-locus F-box gene). In addition to S2-RNase and PiSLF 2, 11 genes were identified by two cDNA selection methods in an 881-kb BAC contig covering the 328-kb region. Expression profiles of these 11 genes were examined by RT-PCR, and seven of them were verified to be located in the 881-kb contig. Both sequence analysis of the 328-kb region and cDNA selection in the 881-kb region suggested that the S-locus was rich was repetitive sequences but deficient in genes.