The Arabidopsis thaliana E3 Ubiquitin Ligase BRIZ Functions in Abscisic Acid Response

Abstract

The ubiquitin system is essential for multiple hormone signaling pathways in plants. Here, we show that the Arabidopsis thaliana E3 ligase BRIZ, a heteromeric ligase that consists minimally of BRIZ1 and BRIZ2 proteins, functions in abscisic acid (ABA) signaling or response. briz1 and briz2 homozygous mutants either fail to germinate or emerge later than wild-type seedlings, with little cotyledon expansion or root elongation and no visible greening. Viability staining indicates that briz1 and briz2 embryos are alive but growth-arrested. Germination of briz mutants is improved by addition of the carotenoid biosynthetic inhibitor fluridone or gibberellic acid (GA₃), and briz mutants have improved development in backgrounds deficient in ABA synthesis (gin1-3/aba2) or signaling (abi5-7). Endogenous ABA is not higher in briz2 seeds compared to wild-type seeds, and exogenous ABA does not affect BRIZ mRNAs in imbibed seeds. These results indicate that briz embryos are hypersensitive to ABA and that under normal growth conditions, BRIZ acts to suppress ABA signaling or response. ABA signaling and sugar signaling are linked, and we found that briz1 and briz2 mutants excised from seed coats are hypersensitive to sucrose. Although briz single mutants do not grow to maturity, we were able to generate mature briz2-3 abi5-7 double mutant plants that produced seeds. These seeds are more sensitive to exogenous sugar and are larger than seeds from sibling abi5-7 BRIZ2/briz2-3 plants, suggesting that BRIZ has a parental effect on seed development. From these data, we propose a model in which the BRIZ E3 ligase suppresses ABA responses during seed maturation and germination and early seedling establishment.

Keywords: ABA2; ABI5; Arabidopsis; E3 ligase; abscisic acid; germination; hormone signaling; ubiquitin.

Figure 1

briz1 and briz2 mutants are defective in seed coat rupture (SCR) and germination and have arrested growth. (A–E, top) Top, image of a portion of a GM plate with seeds from plants of the following genotypes after ~7 days at 20°C under constant light: (A) wild-type Col, (B) BRIZ1/briz1-1, (C) BRIZ2/briz2-1, (D) BRIZ2/briz2-2, and (E) BRIZ2/briz2-3. Blue boxes surround briz seeds with non-germination or slow germination phenotypes. Scale bars represent 1 cm. (A–E, bottom) The fraction of total seeds at each of three different stages of germination was scored after 24 and 48 h, 5 and 7 days at 20°C constant light, and graphed for each genotype. Graphs show means of three independent experiments with approximately 150 seeds per genotype per experiment. Bars represent ±SD. Examples of each germination stage are shown in (F). The three stages are (1) no SCR and no radicle emergence; (2) SCR but no radicle emergence; and (3) both SCR and radicle emergence (germination). After more than 1 week, some briz embryos emerge from their seed coats (germinate), but remain pale and growth arrested, examples in (G) and (H). Scale bars in (F–H) represent 0.25 mm.

Figure 2

briz1 and briz2 mutants are alive despite arrested growth. Seeds from BRIZ/briz plants were plated on GM for 15 days (A–I) or 30 days (J–R) before embryos were treated with viability stains. Embryos from Col seeds soaked for 1 h in water were used as controls (S–V). briz embryos were either heat-treated for 5 min at 98°C or untreated, followed by Sytox staining (A–C, J–L, S or D–F, M–O, T, respectively) or fluorescein diacetate (FDA) only staining (G–I, P–R, U,V) before confocal microscopy. Heat-treated (S) or untreated Col after Sytox (T) or FDA staining (U). Heat-treated Col after FDA staining (V). White bar (V) represents 70 μm. All were visualized at the same scale.

Figure 3

Fluridone reduces the fraction of briz1-1 and briz2-2 mutants with a typical briz phenotype, which returns upon addition of abscisic acid (ABA). (A) Seeds from heterozygous BRIZ1/briz1-1 plants were plated on agar GM plates or on GM plates lacking sucrose or containing 1% mannitol, 1 μM 2,4-D, 50 μM aminocyclopropane carboxylic acid (ACC), 10 μM brassinolide (BL), 100 μM fluridone (Flur.), or 100 μM fluridone + 0.1 μM ABA. Plates with 0.1% EtOH were used as solvent controls for the 2,4-D and ACC plates. Plates with 0.1% DMSO were used as solvent controls for the BL and fluridone plates. For the dark treatment, GM plates were wrapped in foil. Seedlings with a typical briz phenotype were scored after 12–15 days and represented as a fraction of the total seeds plated. Asterisks (****) represent p < 0.0001, ANOVA analysis (Graph Pad Prism). Bars represent ±SD. Data represent 3–4 independent experiments with 30–50 seeds each. (B) Seeds from BRIZ2/briz2-2 plants, analyzed as described in (A).

Figure 4

briz seeds do not hyper-accumulate ABA, and ABA does not affect BRIZ mRNA levels. (A) Anti-Myc immunoblot of extracts from 24 h imbibed seeds from untransformed control Col (left), a complemented briz2-1 line expressing Myc-BRIZ2 with a wild-type phenotype (right), and the homozygous S1 briz2-1 line with the silenced Myc-BRIZ2 transgene, whose progeny failed to germinate (S1, middle lane). Below, total protein staining with Coomassie blue (CB). Asterisk denotes a non-specific immunoreactive band. Arrow denotes position of Myc-tagged BRIZ2. (B) ABA content [ng/gm fresh weight GFW)] in 1 h imbibed seeds from wild-type (Col) and two silenced lines (S1 and S2) derived from independent T1 briz2-1 Myc-BRIZ2 transgene-containing complementation lines (Hsia and Callis, 2010). S1-A and S1-B represent siblings of same initial transformation event. Mean of three replicates except for S2, which was measured twice, ±SD. ANOVA was performed for S1 compared to Col (GraphPad Prism, with Tukey’s multiple comparisons test). (C) Expression of BRIZ1, BRIZ2, and RD29A mRNAs after 6 h of ABA or mock treatment as assayed by quantitative RT-PCR normalized to UBQ10. Results are shown as mean ± SD (n = 9) from three independent experiments. RD29A serves as a positive control for ABA response.

Figure 5

briz2-2 mutants have increased germination, greening, and growth in the gin1-3 background. Seeds from plants homozygous for gin1-3 and heterozygous for BRIZ2/briz2-2 were plated on plain GM. (A) Plate with seeds from a gin1-3 plant heterozygous for BRIZ2/briz2-2. Red circles are briz2-2 gin1-3 seedlings. (B) Fraction of homozygous briz2-2 mutants with green expanded cotyledons was recorded after 14 days. N = 3 plates with approximately 50 seeds per plate. Bars represent ±SD. Asterisks indicate a significant difference (** for p < 0.01) according to a t-test with Welch’s correction in GraphPad Prism. (C) Photos of homozygous briz2-2 seedlings at 18 days. Bars, 1 mm. (D) PCR genotyping of seedlings in panel (C) for BRIZ2, with Col as a WT control. GSP = PCR with gene-specific primers flanking the T-DNA insertion site. T-DNA = PCR with T-DNA primer and gene-specific primer. Primer sequences in Supplementary Table S1.

Figure 6

Gibberellic acid addition improves germination of briz1-1, briz2-1, and briz2-3 mutants. (A) Seeds from plants heterozygous for BRIZ1/briz1-1 were plated on GM with 0, 10, 100, or 200 μM GA₃. Germination of briz seeds was scored after 1 week. Bars represent ±SD. Data from 2 to 4 independent experiments with 120–150 seeds per experiment. (B) Seeds from plants heterozygous for BRIZ2/briz2-1. (C) Seeds from plants heterozygous for BRIZ2/briz2-3. (D) Seeds were plated on GM with or without 1 mM GA₃.

Figure 7

briz embryos are hypersensitive to sucrose and glucose. Embryos were excised from Col, briz1-1, or briz2-3 seeds, placed on nylon filters on top of the indicated solid growth media (GM), and photographed at 0 days (top three rows) and 5 days (bottom three rows) after incubation at 20°C constant light. Four types of GM were used in the experiment, +/− 10 μM gibberellic acid (GA₃₎: (A,B) GM which contains 1% sucrose, (C,D) GM without sugar, (E,F) GM with 1% mannitol instead of sucrose, or (G,H) GM with 1% glucose instead of sucrose. Bars, 1 mm. At least three groups of embryos were plated for each genotype and type of growth media, and representative images are shown. A similar experiment with older seeds is shown in Supplementary Figure S7. Note: to identify briz1-1 and briz2-3 seeds, which are indistinguishable from wild-type at the dry seed stage, seeds from heterozygous BRIZ1/briz1-1 or BRIZ2/briz2-3 plants were plated on the appropriate growth media, and seeds that had not germinated after 48 h were selected for the experiment. PCR genotyping after the above experiment confirmed that these seeds were briz1-1 or briz2-3 homozygous mutants (shown in Supplementary Figure S6).

Figure 8

briz mutants have more ABI5 protein than wild-type seedlings at 72 h post-imbibition. Total proteins were extracted from Col seeds at various timepoints (dry, stratified at 4°C for 24 h, or stratified at 4°C for 24 h followed by incubation at RT under lights for 12, 18, 24, or 72 h), and from briz1-1 and briz2-3 seeds stratified at 4°C for 24 h followed by incubation at RT under lights for 72 h (seeds from heterozygous BRIZ1/briz1-1 or BRIZ2/briz2-3 plants were plated on GM, and seeds that had not germinated after 72 h were selected for the experiment). Around 25 μgs of total protein per sample were separated by SDDS-PAGE, and (A, top panel) anti-ABI5 western blotting was used to visualize the endogenous ABI5 protein in each sample. Dry abi5-7 seeds (left lane) are a negative control for the anti-ABI5 antibody. (A, middle panel) As a loading control, an anti-actin western blot was performed on the same membrane. (A, lower panel) Ponceau staining of total proteins on the membrane. Note the decrease in seed storage proteins (< the 35 kDa marker) in the Col samples across timepoints, and the appearance of the Rubisco large subunit (near the 55 kDa marker) in the 72 h sample. Asterisk indicates non-specific band visible in 3-day-old seedings, but not dry seeds, from abi5-7 null line. (B) ABI5 was normalized to actin, and the fraction of ABI5 protein relative to the amount in Col dry seeds was graphed for each sample. Graph shows the means of three independent experiments (see Supplementary Figure S8 for images of the other two experiments). Bars represent ±SD.

Figure 9

briz1-1 mutants exhibit increased germination and greening in the abi5-7 background. (A) A plate with seeds from an abi5-7 plant heterozygous for BRIZ1/briz1-1 incubated for 14 days constant light. Red circles are briz1-1 abi5-7 seedlings with green cotyledons. Blue circles are typical non-germinated briz1-1 abi5-7 seeds. (B) The fraction of germinated homozygous briz1-1 seeds was recorded after 14 days. N = 4 experiments with 13 total plates with ~50 seeds per plate. Bars represent ±SD. (C) The fraction of homozygous briz1-1 seedlings with green cotyledons was recorded after 14 days. N = 3 experiments with nine total plates with ~50 seeds per plate. Bars represent ±SD. (D) Photos of one briz1-1 abi5-7 plant, 11 and 14 weeks after plating. Bars, 1 cm. (E) PCR genotyping of the plant in panel (D) for BRIZ1. Wt = wild-type control. Mu = the plant in (D). Top panel, PCR with gene-specific primers flanking the T-DNA insertion site. Lower panel, PCR with a T-DNA primer and a flanking gene-specific primer. Asterisks indicate significant differences (* for p < 0.05) according to t-tests in GraphPad Prism. For (B) Welch’s correction was used.

Figure 10

Multiple briz2 mutants exhibit increased germination, greening, and growth in the abi5-7 background. (A) Photos of homozygous briz2-3 abi5-7 seedlings after 14 days compared to a phenotypically wild-type sibling on the same plate (top left, WT sib). Bars, 1 mm. Lower left, PCR genotyping of the pictured seedlings for BRIZ2 (lane numbers correspond to numbers in photos). Top panel, PCR with gene-specific primers flanking the T-DNA insertion site. Lower panel, PCR with T-DNA primer and a flanking gene-specific primer. (B) Fraction of germinated homozygous briz2-3 seeds after 14 days. N = 4 experiments with nine total plates, with ~70 seeds per plate. (C) Fraction of homozygous briz2-3 seedlings with green cotyledons after 14 days. N = 4 experiments with nine total plates, with ~70 seeds per plate. (D) Fraction of germinated homozygous briz2-1 seeds (a different allele than in panels A–C) after 14 days. N = 3 experiments with seven total plates, with ~50 seeds per plate. Bars represent ±SD (B–D). Asterisks indicate significant differences (* for p < 0.05, ** for p < 0.005) according to t-tests in GraphPad Prism. For (B–D), Welch’s correction was used.

Figure 11

briz2-3 abi5-7 mutants are hypersensitive to sucrose and glucose. Seeds from Col, BRIZ2-3/briz2-3, abi5-7, or briz2-3 abi5-7 plants were plated on GM containing either 1% sucrose (left), no sucrose (2nd from left), 1% mannitol (2nd from right), or 1% glucose (right). All data shown are after 10 days of growth in constant light (A) Photos of plates. (B) Closer view of the briz2-3 abi5-7 section on same plates. (C) Photos of individual seeds or seedlings from the briz2-3 abi5-7 section from same plates. (D) Germinated (radicle emergence) seeds (white graphs) and germinated seeds with green expanded cotyledons (green graphs) were scored 10 days after plating. N = 3 independent experiments, with 20 seeds of each genotype per experiment. Bars represent ±SD. Asterisks indicate significant differences (* for p < 0.05, ** for p < 0.005, and **** for p < 0.0001, according to ANOVAs in GraphPad Prism) compared to the same genotype on GM (1% sucrose).