ARABIDOPSIS DEHISCENCE ZONE POLYGALACTURONASE1 (ADPG1), ADPG2, and QUARTET2 are Polygalacturonases required for cell separation during reproductive development in Arabidopsis

Abstract

Cell separation is thought to involve degradation of pectin by several hydrolytic enzymes, particularly polygalacturonase (PG). Here, we characterize an activation tagging line with reduced growth and male sterility caused by increased expression of a PG encoded by QUARTET2 (QRT2). QRT2 is essential for pollen grain separation and is part of a small family of three closely related endo-PGs in the Arabidopsis thaliana proteome, including ARABIDOPSIS DEHISCENCE ZONE POLYGALACTURONASE1 (ADPG1) and ADPG2. Functional assays and complementation experiments confirm that ADPG1, ADPG2, and QRT2 are PGs. Genetic analysis demonstrates that ADPG1 and ADPG2 are essential for silique dehiscence. In addition, ADPG2 and QRT2 contribute to floral organ abscission, while all three genes contribute to anther dehiscence. Expression analysis is consistent with the observed mutant phenotypes. INDEHISCENT (IND) encodes a putative basic helix-loop-helix required for silique dehiscence, and we demonstrate that the closely related HECATE3 (HEC3) gene is required for normal seed abscission and show that IND and HEC3 are required for normal expression of ADPG1 in the silique dehiscence zone and seed abscission zone, respectively. We also show that jasmonic acid and ethylene act together with abscisic acid to regulate floral organ abscission, in part by promoting QRT2 expression. These results demonstrate that multiple cell separation events, including both abscission and dehiscence, require closely related PG genes.

Figure 1.

Activation Tagging of At3g07970. (A) Genomic region of the T-DNA insertion (activation tagging) site in line AT3. Boxes and lines represent exons and introns/intergenic regions, respectively. LB indicates left border sequence of the activation tagging vector (pSKI015; Weigel et al., 2000), and the four closed ovals represent the 4x 35S enhancer sequence. (B) RT-PCR analysis, using RNA extracted from flower buds and rosette leaves, of genes flanking the T-DNA insertion site in AT3. rRNA was used as the control. (C) Schematic diagram of the 35S enhancer-ProQRT2:QRT2 construct used to determine if the AT3 phenotype is caused by overexpression of the At3g07970 gene. The solid and dotted lines 5′ of the 1st exon indicate the presumed At3g07970 promoter sequence (2085 bp) and sequence (630 bp) from the pMN19 vector (Weigel et al., 2000), respectively. The four gray ovals and box represent the 4x 35S enhancer sequence (from pMN19) and nopaline synthase terminator (NOS) from the pGWB1 vector, respectively. (D) Phenotype of 6-week-old plants. Wild-type Columbia, AT3 (backcrossed into Columbia [Col-0] twice), and a representative strong line containing the construct in (C) are shown. Bar = 3cm. (E) Flower bud phenotype at floral developmental stage 14 (Smyth et al., 1990) of the wild-type, AT3 (backcrossed into Col-0 twice), and representative examples of the weak and strong phenotype in plants containing the construct in (C).

Figure 2.

adpg1, adpg2, and qrt2 Loss-of-Function Phenotypes. (A) Position of the T-DNA insertions in At3g07970 (QRT2) and the related genes, ADPG1 and ADPG2. Boxes and lines represent exons and introns, respectively. Triangles indicate positions of T-DNA insertions, and the position of the point mutation in qrt2-1 is shown. Open boxes represent predicted 5′ and 3′ untranslated regions. There is no information available regarding the 5′ untranslated region of QRT2. The exon shown in red is missing in transcript from the adpg1-1 allele. (B) Silique indehiscence phenotype of the adpg1-1 and adpg1-2 mutants. Fully matured wild-type siliques break easily, while those of the adpg1 mutants require external mechanical stress (C) Transverse sections of wild-type and adpg1 adpg2 qrt2 triple mutant siliques (stage 17) stained with Toluidine Blue. The red boxes indicate where the DZs form in the wild-type and fail to form in the mutant. V, silique valve; S, septum. Bar = 50 μm. (D) Flowers of the wild-type (stage 14: pollination) and the adpg1 adpg2 qrt2 triple mutant with delayed anther dehiscence. (E) Scanning electron microscopy image of a mature pollen tetrad from the adpg1 adpg2 qrt2 triple mutant. (F) Transverse sections of wild-type and adpg1 adpg2 qrt2 triple mutant anthers (stage 13) stained with Toluidine Blue. Arrowheads indicate where stomium separation has occurred in the wild-type anther and has not yet occurred in the adpg1 adpg2 qrt2 anther.

Figure 3.

ADPG1 and ADPG2 Are Required for Pod Shatter. (A) to (C) Toluidine Blue–stained transverse sections of wild-type ([A] and [B]) and adpg1 adpg2 qrt2 triple mutant (C) siliques at stage 18. V, silique valve; S, septum. (B) shows an enlarged image of the section indicated by the rectangle in (A). The wild-type silique has begun to dehisce, and broken cell walls are visible where the valves have separated from the replum. By contrast, the replum and valve remain intact in the triple mutant. (D) and (E) TEMs of transverse sections of comparable wild-type (D) and adpg1 adpg2 qrt2 triple mutant (E) siliques, immediately after pod shatter in wild-type siliques, labeled with the JIM5 monoclonal antibody against the unesterified region of pectin. A broken cell wall from the wild-type is shown in (D), while cell separation has not occurred in the triple mutant (E). Bars = 50 μm in (A) and (C) and 0.2 μm in (D) and (E).

Figure 4.

ADPG2, QRT2, Ethylene, JA, and ABA Are Required for Normal Floral Organ Abscission. (A) The number of flowers on a single inflorescence with visible petals and floral organs remaining attached after firm pressing (at least nine plants per genotype; means ± se are shown). (B) Typical wild-type inflorescence with two flowers with visible petals that retained floral organs after pressing. Flowers were numbered down the main inflorescence starting from the youngest with visible petals for scoring floral organ abscission. (C) Typical adpg2 qrt2 inflorescence with three flowers with visible petals that retained floral organs after pressing. (D) Typical ein2 aos aba2 inflorescence showing the youngest four flowers with visible petals that retained floral organs after pressing. All floral organs remained turgid and healthy while still attached. The arrowhead indicates the youngest flower (flower 1) with visible petals: more immature flowers (not counted) are present compared with AOS inflorescences because JA is required for normal flower maturation. Self-pollination of ein2 aos aba2 triple mutant flowers (with pollen from JA-treated triple mutants) failed to restore floral organ abscission (inset). (E) Relationship between the number of flowers retaining visible petals after firm pressing and the youngest flower with QRT2:GUS expression in the floral AZ. Plants with a range of floral organ abscission phenotypes were selected from the F2 progeny of a cross between a representative QRT2:GUS line and the ein2 aos aba2 triple mutant. The values at the top of the graph show the mean ± se for the different genotypes as shown in (A). Flowers were numbered as shown in the inset image, which shows a plant with three flowers retaining visible petals after firm pressing. Flower 3 is the youngest flower with detectable GUS expression in the floral AZ (arrowhead). For all genotypes, GUS staining was first observed in either the youngest flower that lost its floral organs or in the next youngest flower, just about to shed its organs (as in the inset). Floral organ abscission was not observed prior to GUS expression in any plant. Thirty-two F2 progeny were examined, of which 26 contained the QRT2:GUS transgene.

Figure 5.

Tissue-Specific Expression of ADPG1, ADPG2, and QRT2. The relative expression levels of ADPG1, ADPG2, and QRT2 in different tissues were determined using qRT-PCR and gene-specific primers. The lowest expression level (ADPG1 in roots) was adjusted to 1, and the relative expression levels of each gene in all tissues tested was calculated relative to this value. Data represent the means and se from three independent biological samples and are shown on the figure (with se in parentheses) where too small to see. ND, not detected. (A) RNA was extracted from roots and rosette leaves from 3-week-old plants just before bolting, from flower buds, and from siliques at three developmental stages: (i) young siliques (1 to 3 d after pollination), (ii) developing siliques (about 1 week after pollination), and (iii) fully matured siliques just after turning yellow but before pod shatter (2 to 3 weeks after pollination). (B) RNA was also isolated separately from entire siliques with all seeds removed, but including the receptacle, placenta, replums, funiculi, and style (Silique) and the removed seeds (Seed) from mature siliques just after turning yellow (2 to 3 weeks after pollination).

Figure 6.

ADPG1, ADPG2, and QRT2 Are Expressed at Sites of Cell Separation. (A) ADPG1:GUS expression in the silique DZ and seed AZ (arrowhead) from mature siliques immediately prior to pod shatter. (B) ADPG2:GUS expression in anthers and the floral organ AZ (arrowhead) prior to floral organ abscission. In anthers, expression occurs during early development (1), is absent at later stages (2), and returns just prior to dehiscence (3). (C) QRT2:GUS expression in anthers and in the floral organ AZ (arrowhead) prior to floral organ abscission. In anthers, expression occurs during early development (1), is absent at later stages (2), and returns just prior to dehiscence (3). (D) ADPG1:GUS expression at the apical tip of the stamen filament and in the anther DZ. (E) ADPG2:GUS expression in the silique DZ at pod shatter. (F) QRT2:GUS expression and anther dehiscence was restored after aos plants were sprayed to runoff with 1 mM methyl jasmonate. No GUS activity was detected in untreated aos anthers. (G) QRT2:GUS expression in the floral organ AZ at the abscission scars of the stamens, petals, and sepals. (H) QRT2:GUS expression surrounding an emerging lateral root. (I) ADPG2:GUS expression at the site of radicle emergence in germinating seeds 24 h after stratification. No expression was observed immediately after stratification finished. (J) Transverse section of a flower at stage 13 showing ADPG1:GUS expression where stomium separation has occurred and at the apical tip of the stamen filament. (K) Transverse section of a dehiscing silique showing ADPG1:GUS expression where the valves have separated from the replum. (L) Dark-field image showing ADPG2:GUS expression in the anther tapetum at flower stage 11. (M) Dark-field image showing QRT2:GUS expression in the anther DZ at stage 13. (N) Dark-field image showing QRT2:GUS expression in the anther tapetum at flower stage 10, when microspore separation occurs. (O) ADPG1:GUS expression in the valve margin is present in the wild-type and hec3 but absent in the nonshattering ind mutant (top). Scanning electron micrograph of mature wild-type, ind, and hec3 seeds showing the constrictions (arrowheads) in the wild-type and ind funiculi where abscission will occur. This structure is absent in hec3 plants, and seed abscission does not occur (bottom). (P) ADPG1:GUS expression in mature siliques with the valve removed. IND is required for expression in the silique DZ (1), while HEC3 is required for expression in the seed AZ (2) and ovule funiculus (3).

Figure 7.

Complementation of the Pod Shatter Defect. The cDNAs of ADPG1, ADPG2, and At1g48100 were each placed under the control of the presumed ADPG1 promoter and transformed separately into the adpg1 adpg2 qrt2 triple mutant. The three constructs were tested for their ability to complement the pod shatter defect due to reduced adpg1 and adpg2 activity. The bottom panel shows the number of independent lines for each construct in each complementation category: full (+++), moderate (++), weak (+), and none (−). Transgene-specific primers were used to determine the expression levels of the different cDNAs in three lines of each construct with various degrees of complementation (top panel). Values represent the mean ± se from triplicate qRT-PCR reactions. ND, not detected.