Constitutive signaling activity of a receptor-associated protein links fertilization with embryonic patterning in Arabidopsis thaliana

Abstract

In flowering plants, the asymmetrical division of the zygote is the first hallmark of apical-basal polarity of the embryo and is controlled by a MAP kinase pathway that includes the MAPKKK YODA (YDA). In Arabidopsis, YDA is activated by the membrane-associated pseudokinase SHORT SUSPENSOR (SSP) through an unusual parent-of-origin effect: SSP transcripts accumulate specifically in sperm cells but are translationally silent. Only after fertilization is SSP protein transiently produced in the zygote, presumably from paternally inherited transcripts. SSP is a recently diverged, Brassicaceae-specific member of the BRASSINOSTEROID SIGNALING KINASE (BSK) family. BSK proteins typically play broadly overlapping roles as receptor-associated signaling partners in various receptor kinase pathways involved in growth and innate immunity. This raises two questions: How did a protein with generic function involved in signal relay acquire the property of a signal-like patterning cue, and how is the early patterning process activated in plants outside the Brassicaceae family, where SSP orthologs are absent? Here, we show that Arabidopsis BSK1 and BSK2, two close paralogs of SSP that are conserved in flowering plants, are involved in several YDA-dependent signaling events, including embryogenesis. However, the contribution of SSP to YDA activation in the early embryo does not overlap with the contributions of BSK1 and BSK2. The loss of an intramolecular regulatory interaction enables SSP to constitutively activate the YDA signaling pathway, and thus initiates apical-basal patterning as soon as SSP protein is translated after fertilization and without the necessity of invoking canonical receptor activation.

Keywords: Arabidopsis thaliana; BRASSINOSTEROID SIGNALING KINASE; MAP kinase signaling; embryogenesis; evolution.

Fig. 1.

Functional importance of BSK family kinases in early embryogenesis. (A) Nomarsky images of transition-stage embryos in whole-mount cleared seeds of wild-type Col-0, yda, bsk1-2, bsk2-2, and ssp-2 single mutants, as well as bsk1 bsk2 double mutants (b1 b2) and bsk1 bsk2 ssp triple mutants (b1 b2 ssp). (Scale bars: 10 μm.) (B) Box plot diagram of suspensor length measurements for >100 embryos. The total number of analyzed embryos is depicted above the x axis. The box plot diagram shows the median as center lines, and the 25th and 75th percentiles are indicated by box limits. Whiskers show 1.5× interquartile distance. Outliers are represented by dots. Statistically significant differences from wild type were determined by the Mann–Whitney U test (**P < 0.001; *P < 0.01). Statistically significant differences in other pairwise comparisons are indicated by brackets. (C) Nomarsky images of embryos at the one-cell stage in whole-mount cleared seeds. The average sizes of the apical (yellow) and basal (green) daughter cells with the SD and number of analyzed embryos are given below the image. Furthermore, significant differences in pairwise comparisons to the indicated reference (Ref) determined by the Mann–Whitney U test (*P < 0.001) are indicated below the image (gray). Apical cells are false-colored in yellow, and basal cells are shown in green. (Scale bars: 10 μm.)

Fig. 2.

Phenotypic analysis of YDA-dependent plant development. (A) Inflorescence development of wild-type Col-0 in comparison to single-, double-, and triple-mutant combinations of ERf kinases and BSK family kinases. The bsk1 bsk2 double mutant (b1 b2) and bsk1 bsk2 ssp triple mutant (b1 b2 ssp) show more compact inflorescences very similar to erecta (er) single mutants and erecta erecta-like1 erecta-like2 (er erl1 erl2) triple mutants. (Scale bar: 5 mm.) (B) Rosette leaves at the time of bolting. No obvious difference in rosette diameter can be seen in bsk1-2, bsk2-2, or ssp single mutants in comparison to wild type. The bsk1 bsk2 double mutant, however, shows a reduced rosette size similar to plants with reduced ERf receptor signaling (SI Appendix, Fig. S4). (Scale bar: 2 cm.) (C) Epidermal phenotype of 10 d-old seedlings of wild-type Col-0, bsk1, bsk2, ssp, bsk1 bsk2 double mutants (b1 b2), and bsk1 bsk2 ssp triple mutants (b1 b2 ssp), as well as bsk1 bsk2 ssp triple mutants carrying a constitutively active variant of yda (b1 b2 ssp yda-CA). Loss-of-function yda mutants and a constitutively active variant of YDA (yda-CA) are shown as controls. (Scale bar: 50 μm.)

Fig. 3.

Genetic rescue of ssp mutants by sperm-specific BSK gene expression. A box plot diagram of suspensor size measurements 75 h after pollination is shown. Homozygous transgenic lines carrying pMGH3::SSP or pMGH3::BSK1 in a bsk1 bsk2 ssp triple-mutant background were used as pollen donors to pollinate wild-type Col-0 plants. Control Col-0 and bsk1 bsk2 ssp triple mutants were used as pollen donors in parallel experiments. Without genetic rescue, the paternal effect of the mutant ssp allele causes reduced suspensor size in heterozygous embryos. The box plot diagram shows the median as center lines, and the 25th and 75th percentiles are indicated by box limits. Whiskers show 1.5× interquartile distance. Outliers are represented by dots. The total number of embryos for each genotype is given above the x axis. Statistically significant differences from the control crosses with homozygous bsk1 bsk2 ssp triple mutants as pollen donors are indicated by asterisks (Mann–Whitney U test: *P < 0.01).

Fig. 4.

SSP-specific properties can be mapped to the C-terminal TPR motif. (A) To map the structural features underlying the functional difference between the SSP and BSK1 protein, chimeric constructs (described in Materials and Methods) were tested for their ability to genetically rescue the ssp mutant phenotype. For each construct, 12 independent transgenic lines were analyzed for genetic rescue of the embryonic ssp phenotype (>100 embryos each line) in comparison to a full-length SSP rescue construct as a positive control and a nonfunctional SSP construct [SSP G2 > A (12)] as a negative control. Functional protein domains/motifs are depicted as larger rectangular boxes. A, kinase activation loop; K, kinase domain; M, myristoylation motif; T, TPR domain. Each predicted TPR repeat is shown as an individual rectangular box, and the first cryptic repeat is indicated by a dashed line. SSP sequences are depicted in gray, and BSK1 sequences are depicted in white. (B) Genetic rescue of ssp mutants based on morphological criteria by expression of chimeric SSP and BSK1 variants shown in A under control of the SSP regulatory sequences. The bar graph shows the number of wild-type embryos as an average percent value with the SD for 12 independent transgenic lines for each construct. Statistically significant differences from the negative control (ssp) are indicated (determined by χ² test: *P < 0.0001) (details of the analysis and statistical test are provided in Dataset S1, and the heritability of these traits is shown in Dataset S2). (C) Transcript levels of CYP82C in Arabidopsis protoplasts in response to the expression of SSP and BSK1 and chimeric versions determined by qRT-PCR. As controls, a constitutively active variant (YDA-CA) and a nonfunctional version of YDA (YDA-KD) were used. Transcript levels were normalized to EF2 expression as a reference. Expression levels are given as relative values, with the negative control (YDA-KD) arbitrarily set to 1. Significant differences from the negative control (YDA-KD) are indicated (Student’s t test: *P < 0.05).

Fig. 5.

Activation of the ERf/YDA pathway by ectopic SSP expression. (A) Transcript levels of EPF2 in Arabidopsis protoplasts in response to expression of SSP and BSK1 determined by qRT-PCR. As controls, a constitutively active variant (YDA-CA) and a nonfunctional version of YDA (YDA-KD) were used. Transcript levels were normalized to EF2 expression as a reference. Expression levels are given as relative values with the negative control (YDA-KD) arbitrarily set to 1. Significant differences from the negative control (YDA-KD) are indicated by asterisks (Student’s t test: *P < 0.05). (B) Phosphorylation of MPK3 and MPK6 in response to YDA activation detected by Western blotting with a phosphorylation-sensitive antibody against the activation loop of both kinases. As loading controls, phosphorylation-independent antibodies against MPK6 and tubulin were used. Note that the polyclonal antibody against MPK6 also detects MPK3 (black arrowhead). The MPK3 and MPK6 proteins have expected sizes of 42.7 kDa (black arrowhead) and 45.1 kDa (white arrowhead), respectively. The bar graph gives the relative amount of phosphorylated MPK6 in relation to total MPK6 as inferred by the signal intensity of the corresponding Western blots. Average values and SDs of three Western blots are shown. Significant differences from the negative control (YDA-KD) are indicated (Student’s t test: *P < 0.05).

Fig. 6.

SSP has lost the competence for intramolecular interaction. (A) Yeast two-hybrid assay to test for interaction between the TPR and kinase domains of SSP and BSK1. Serial dilutions of yeast cells containing the GAL4-DNA binding domain (BD) fused to the TPR domain of SSP or BSK1 plus the GAL4 activation domain (AD) fused to the kinase domains of SSP or BSK1, respectively, were dropped on vector-selective [complete supplement medium without leucine and tryptophan (CSM L- W-)] and interaction-selective medium (CSM L- W- H- Ade-). Growth was monitored after 3 d at 30 °C. pGADT7 empty vector was used as a negative control (“vector”). (B) Interaction tests using rBiFC. (Top) Confocal images of transiently transformed N. benthamiana leaf epidermal cells representing the median value of RFP and YFP fluorescence for each experiment are shown. (Bottom) Box plot diagram depicts the ratio of the mean fluorescence of YFP against RFP from 25 different, randomly chosen leaf sections. Center lines of boxes represent the median, with outer limits at the 25th and 75th percentiles. Notches indicate 95% confidence intervals. Tukey whiskers extend to 1.5× the interquartile ratio.