A dual role for the S-locus receptor kinase in self-incompatibility and pistil development revealed by an Arabidopsis rdr6 mutation

Abstract

The coordinate evolution of self-incompatibility (SI) and stigma-anther separation, two mechanisms that promote cross-pollination in plants, has been a long-standing puzzle in evolution and development. Using a transgenic self-incompatible Arabidopsis thaliana model, we performed screens for mutants exhibiting a modified SI response. A mutation in the RNA-dependent RNA polymerase RDR6, which functions in trans-acting short interfering RNA (ta-siRNA) production, was found that simultaneously enhances SI and causes stigma exsertion, without associated increases in SRK transcript levels. While rdr6 mutants had been previously shown to exhibit stochastic stigma exsertion, our results demonstrate that the S-locus receptor kinase (SRK) gene further enhances pistil elongation and stigma exsertion in this mutant background, a process that requires SRK catalytic activity and correlates with SRK transcript levels. These results suggest that positive regulators or effectors of SI and pistil development are regulated by ta-siRNA(s). By establishing complex connections between SI and stigma exsertion through the sharing of a ta-siRNA-mediated regulatory pathway and the dual role of SRK in SI and pistil development, our study provides a molecular explanation for the coordinate evolution of these processes.

Figure 1.

Enhanced SI and Stigma Exsertion in m014∷SRKb-SCRb Caused by a Mutation in RDR6. (A)to (D) Self- and cross-pollination assays. Self-pollination of +3-stage flowers from Col-0∷SRKb-SCRb (A) and m014∷SRKb-SCRb(B) and reciprocal crosses to wild-type Col-0 showing that stigma (C) and pollen (D) of m014∷Sb are functional. The numerical designation of developing flower buds adopted in studies of Brassica SI is used: a floral bud at flower opening (anthesis) is floral bud stage 0 (equivalent to stage 13 of Smyth et al., 1990), with progressively younger buds along the inflorescence designated by negative numbers (−1, −2, etc.), and progressively older buds designated by positive numbers (+1, +2, etc). Bars = 50 μm in (A), (C), and (D) and 20 μm in (B). (E) and (F) floral characteristics of Col-0∷SRKb-SCRb (E) and m014∷SRKb-SCRb plants (F) showing stigma exsertion in mutant plants. (G) The RDR6 gene and the positions of the rdr6-en missense mutation and other previously reported rdr6 mutant alleles (Peragine et al., 2004). (H) Self-pollen inhibition on a +3-stage stigma of an F1 plant derived from a cross between an rdr6-en∷SRKb-SCRb and an rdr6-11 homozygote. Bar = 20 μm.

Figure 2.

Seed Production in Col-0∷Sb, rdr6-11, and rdr6-11∷Sb Plants and Expression of SRKb in Developing Stigmas. (A)and (B) Seed production in the three lines as a measure of the degree and uniformity of successful pollination. Images of inflorescences (A) and counts of autonomous seed set (B) show abundant and uniform seed production in Col-0∷Sb (all siliques are seed filled and expanded), stochastic reduction in seed set in rdr6-11 (mixture of expanded and unexpanded siliques containing few or no seed), and the dramatic and uniformly reduced seed production in rdr6-11∷Sb (absence of fully expanded siliques and only occasional partially expanded siliques containing few seeds). In (B), the number of seeds per silique derived from autonomous self-pollination (i.e., in the absence of pollinators) was obtained by analysis of 40 siliques (10 siliques from each of four plants) from each line. In Col-0∷Sb plants, among the siliques that contained >20 seeds, 32 siliques contained >30 seeds and seven siliques contained 21 to 30 seeds. (C) Gel blot analysis of poly(A)+ RNA from stigmas at the −1, +1, and +3 developmental stages. The blot was hybridized first with a probe derived from the SRKb first exon, which hybridizes with a 3.0-kb full-length transcript (asterisk) and an alternative 1.6-kb transcript (circle) corresponding to exon 1 and subsequently with an actin probe. The normalized values of the 3.0-kb SRKb transcript (with Col-0∷Sb set at 1.0) are indicated. (D) Seed production after manual self- and cross-pollination. The histogram shows the average number of seeds per silique (±1.96 se) produced after manual application of pollen on +1-stage stigmas. The average number of seeds in each pollination experiment was obtained by analysis of 30 siliques (10 siliques from each of plants from each line).

Figure 3.

Stigma Exsertion and Expression of SRKb in Pistil Tissues. (A) Extent of stigma exsertion in +1-stage flowers of Col-0∷Sb, rdr6-11, and rdr6-11∷Sb plants. The four floral organs are labeled in the Col-0∷Sb panel and the anther and stigma, which cap the stamen and pistil, respectively, are labeled in the rdr6-11∷Sb panel. (B) Relative heights (in millimeters) of stigma and anther in each of 40 +1-stage flowers from the three lines. The anther height shown is the average height of the four long anthers in each flower. (C) Average stigma-anther separation in the three lines. For each of the 40 flowers analyzed per line, the distance between stigma and anther was assigned a positive value if the stigma was exserted above the anthers and a negative value if the stigma was recessed below the anthers. The stigma-anther separation values, presented as means ±1.96 se, were significantly different among the three lines (P value < 0.01). (D)Gel blot analysis of poly(A)+ RNA purified from stigmas and the remainder of the pistil (style + ovary) from –1-stage buds of the three A. thaliana lines and A. lyrata (A.l.) Sb plants carrying the native SRKb allele. Hybridization and analysis of hybridization signals were as in Figure 2C. The blot was hybridized first with a probe derived from the SRKb first exon, which hybridizes with a 3.0-kb full-length transcript (asterisk) and an alternative 1.6-kb transcript (circle) corresponding to exon 1, and subsequently with an actin probe.

Figure 4.

Differential Elongation in Col-0, Col-0∷Sb, rdr6-11, and rdr6-11∷Sb Pistils. (A) Relative pistil and stamen lengths in Col-0 and Col-0∷Sb plants. Neither pistil nor stamen length show significant differences between the two strains (P = 0.5262 and 0.3047). (B) Relative pistil and stamen lengths in Col-0∷Sb, rdr6-11, and rdr6-11∷Sb. Pistil length differs significantly among the three lines (P < 0.01). Stamen length in rdr6-11 was significantly shorter than that in Col-0∷Sb (P < 0.01), whereas stamen length in rdr6-11∷Sb was not significantly different from rdr6-11 (P = 0.254). (C)and (D) Relative lengths of the stigma+style region (C) and ovary (D) in Col-0∷Sb, rdr6-11, and rdr6-11∷Sb plants. For each line, the average length (±1.96 se) of each structure was calculated from measurements performed as described in Figure 3 on 40 flowers at the +1 stage (10 flowers from each of four plants).

Figure 5.

Scanning Electron Micrographs of +1-Stage Pistils from Col-0∷Sb and rdr6-11∷Sb Plants. The higher magnification micrographs show close-ups of the style (top) and a region in the middle of the ovary (bottom) from the same pistil of each genotype.

Figure 6.

Correlation of SI, Stigma Exsertion, and Seed Set with SRKb Transcript Levels in rdr6-11∷Sb Plants. (A) Enhancement of stigma exsertion by SRKb. Only rdr6-11∷SRKb flowers, but not rdr6-11∷SCRb flowers, exhibited enhanced stigma exsertion relative to untransformed rdr6-11 flowers. (B) Effect of SRKb expression levels on SI, stigma-anther separation, and seed set. Four independent rdr6-11∷SRKb T1 plants (designated as numbers 1 through 4), which exhibited different degrees of seed set, were used for quantitative real-time RT-PCR analysis. The values (±sd) were derived from analysis of three biological replicates. Ten flowers from each T1 plant were used to assess the strength of SI [average number of pollen tubes (±1.96 se) after manual application of SCRb-expressing pollen on −1-stage stigmas] and the enhancement of stigma exsertion [average stigma-anther separation (±1.96 se)]. The average number of seeds (±1.96 se) reflects the extent of stigma-anther separation.

Figure 7.

Requirement of a Catalytically Active SRKb for Stigma Exsertion in rdr6 Plants. (A) Floral morphology of rdr6-11∷SRK_bmu (kinase-negative mutant) and untransformed rdr6-11 plants. (B) Comparison of relative SRK_b mRNA levels measured by real-time quantitative RT-PCR in different rdr6-11∷SRK_bmu transgenic lines and the transgenic line shown in Figure 3A, which contains a functional SRK_b and exhibits enhanced stigma exsertion.

Figure 8.

Enhanced SI and Stigma Exsertion in ago7/zip∷Sb Plants. (A) and (B) Self-pollination assays of Col-0∷Sb and ago7/zip∷Sb flowers showing large numbers of elongated pollen tubes, indicative of complete breakdown of SI, in +1-stage Col-0∷Sb flowers (A) and lack of elongated pollen tubes, indicative of enhanced SI, in ago7/zip∷Sb flowers (B) after prolonged stigma-pollen contact (9 h). (C) and (D) Floral characteristics of ago7/zip (C) and ago7/zip∷Sb (D) plants showing enhanced stigma exsertion in ago7/zip∷Sb. (E) Average stigma-anther separation (±1.96 se) in ago7/zip and ago7/zip∷Sb obtained from analysis of 40 +1-stage flowers from each line. The stigma-anther separation values are significantly different between the two lines (P < 0.01).