A unique short-chain dehydrogenase/reductase in Arabidopsis glucose signaling and abscisic acid biosynthesis and functions

Abstract

Glc has hormone-like functions and controls many vital processes through mostly unknown mechanisms in plants. We report here on the molecular cloning of GLUCOSE INSENSITIVE1 (GIN1) and ABSCISIC ACID DEFICIENT2 (ABA2) which encodes a unique Arabidopsis short-chain dehydrogenase/reductase (SDR1) that functions as a molecular link between nutrient signaling and plant hormone biosynthesis. SDR1 is related to SDR superfamily members involved in retinoid and steroid hormone biosynthesis in mammals and sex determination in maize. Glc antagonizes ethylene signaling by activating ABA2/GIN1 and other abscisic acid (ABA) biosynthesis and signaling genes, which requires Glc and ABA synergistically. Analyses of aba2/gin1 null mutants define dual functions of endogenous ABA in inhibiting the postgermination developmental switch modulated by distinct Glc and osmotic signals and in promoting organ and body size and fertility in the absence of severe stress. SDR1 is sufficient for the multistep conversion of plastid- and carotenoid-derived xanthoxin to abscisic aldehyde in the cytosol. The surprisingly restricted spatial and temporal expression of SDR1 suggests the dynamic mobilization of ABA precursors and/or ABA.

Figure 1.

Genetic and Phenotypic Analyses of Arabidopsis gin1 and gin4 Mutants. (A) Glc-insensitive mutants. Three wild-type (Col, Ler, and Ws), three gin1 (gin1-1, gin1-2, and gin1-3), and four aba2 (aba2-1, aba2-2, aba2-3, and aba2-4) seedlings were grown on 6% Glc Murashige and Skoog (1962) (MS) plates under light for 10 days. All experiments were repeated twice with consistent results. (B) gin1-3 and aba2-1 are allelic. Reciprocal crosses between gin1-3 and aba2-1 (g/a and a/g) were performed. Wild-type (WT) and F1 seedlings were grown on 6% Glc MS plates. (C) ABA restores the Glc-sensitive phenotype. gin1-1 (g1-1), gin1-3 (g1-3), gin5 (g5), and abi4 seedlings were grown on 4% Glc MS plates without (top row) or with (bottom row) 100 nM ABA. (D) Mannitol and Glc are distinct signals. Wild-type, gin1-3 (gin1), and gin5 seedlings were grown on 4% mannitol MS plates without (top row) or with (bottom row) 100 nM ABA. (E) The phenotype of gin4 resembles that of ctr1. Wild-type, gin4-1 (g4-1), gin4-2 (g4-2), and ctr1 seedlings were grown on 2% (top row) and 6% Glc (bottom row) MS plates. (F) gin4 and ctr1 are allelic. Wild-type (W), gin4-1 (g4), F1 gin4-1/ctr1-1 (g/c), and ctr1-1 (c1) seedlings were grown on 2% Glc MS plates for 5 days in the dark. The scale bars in (B) to (F) = 1 cm.

Figure 2.

Map-Based Cloning of ABA2/GIN1. (A) Physical map of chromosome I in the ABA2/GIN1 region. Ten new SSLP markers are marked as lines in six of the BAC clone boxes. The number of recombinants (Rec) found in 5627 DNA samples is shown. (B) The gin1-1 and gin1-3 alleles lack exon 2 sequences. RT-PCR was performed using primers derived from the exon 2 sequence for GIN1. A control experiment was performed with primers for the Suc synthase gene (SUS) transcript. (C) Scheme of the genomic and cDNA structure of ABA2/GIN1. The deduced polypeptide is predicted to be 285 amino acids in length with a molecular mass of 30.2 kD, a pI of 5.85, and an α-β structure. Comparison of the cDNA and genomic DNA sequences (in the F19K6 BAC clone; 41428 to 43279 bp) indicated that ABA2/GIN1 contains two exons of 168 and 808 bp and one intron of 672 bp. The black box represents the coding region. (D) Genomic and cDNA sequences of GIN1/ABA2. The 5′ and 3′ ends of the cDNA (uppercase letters) were determined by rapid amplification of cDNA ends and sequencing. (E) Molecular basis of aba2/gin1 mutants. TGXXXGXG is the conserved NAD binding site, and YXXSK is the catalytic site. The mutation sites of 11 aba2/gin1 alleles are shown. Names of the same alleles are shown in parentheses. a.a., amino acids.

Figure 6.

Tissue, Stress, and ABA Regulation of Gene Expression. (A) Tissue expression pattern. ABA2/GIN1 expression was examined in the aba2-2 null mutant and in wild-type (Col) seedlings (Sl) and adult roots (Rt), rosettes (Rs), stems (St), flowers (Fl), siliques (Si), and dry seeds (Sd). The rosettes were separated further to petioles (Pe) and blades (Bl). Seedlings were grown on MS agar plates for 14 days, and adult plants were grown in pots for ∼40 days. Each lane contained 10 μg of total RNA. A 366-bp fragment of ABA2/GIN1 cDNA digested with AluI was used as a probe for hybridization. rRNA was used as a loading control. The experiment was repeated twice with consistent results. (B) Gene expression under dehydration. The expression of ABA2 and NCED3 was determined at various time points after dehydration in rosettes of 40-day-old soil-grown plants. Samples with water (W) were used as a control. The experiment was repeated twice with consistent results. (C) ABA regulation of ABA biosynthesis genes in shoots and roots. Wild-type plants (Ler) grown on agar plates for 14 days were transferred to filter paper soaked with 100 μM ABA and incubated for 0 and 3 h. The experiment was repeated twice with consistent results.

Figure 3.

Arabidopsis ABA2/GIN1 Encodes a Unique Member of the SDR Superfamily. The ABA2/GIN1 coding sequence (217 of 285 residues) covering the six conserved motifs was aligned with sequences of putative SDRs from seven other model organisms using the Clustal alignment program. Arabidopsis ABA2/GIN1 (AtSDR1) is grouped with tobacco (NtSDR) and cucumber (CsSDR) homologs in the phylogenetic tree at bottom. AtSDR2 is closely related to maize ZmTS2 and different from AtSDR3, AtSDR4, and AtSDR5. AtHSD (11-β-hydroxysteroid dehydrogenase) is closely related to the human (HS) and mouse (MM) HSDs. The human and mouse HPGDs (15-hydroxyprostaglandin dehydrogenase) and human and mouse RDHs (retenoid dehydrogenase) also are shown. The scale value of 0.1 indicates 0.1 amino acid substitutions per site.

Figure 4.

Interplay of Glc, ABA, and Ethylene Signaling Pathways. (A) gin1 is epistatic to ein2. Single-mutant (gin1 and ein2) and double-mutant (gin1 ein2) seedlings were grown on 4% Glc MS plates under light for 12 days. All experiments were repeated twice with consistent results. (B) Glc repression of PDF1.2 expression requires ABA. Plants were grown on MS plates without (0) and with 2 and 6% Glc (2G and 6G) or 2 and 6% mannitol (2M and 6M) for 14 days. RT-PCR was repeated twice with consistent results. WT, wild type. (C) Growth defects in gin1. Wild-type and gin1 plants were grown on soil under a 12/12-h light/dark photoperiod for 14 days. The two top arrows point to cotyledons with different cell sizes. The two bottom arrows indicate developing true leaves with similar cell sizes but different cell numbers. (D) Growth defects in gin1. Wild-type and gin1 plants were grown on soil under a 12/12-h light/dark photoperiod for 50 days with 40% (gin1; middle plant) or 85% (gin1; right plant) RH. (E) Wilty phenotype in gin1. Wild-type and gin1 plants were grown on 2% Glc MS medium under a 16/8-h light/dark photoperiod and 100% RH for 24 days. The plastic cover was removed for 5 min. (F) Uncoupling of shoot and root growth regulation in gin1. Single-mutant (gin1 and ein2) and double-mutant (gin1 ein2) plants were grown on soil under a 12/12-h light/dark photoperiod for 16 days.

Figure 5.

Complementation. (A) A genomic ABA2/GIN1 clone restores Glc sensitivity. The T2 segregation of gin1/GIN1 and gin1 is shown. Plants were grown on 6% MS plates under light for 10 days. WT, wild type. (B) Genomic and cDNA clones restore cotyledon and leaf growth defects. The T2 segregation of TG (transgenic with the genomic ABA2/GIN1 clone), TC (transgenic with the 35S::GIN1 cDNA clone), and gin1 plants is shown. Plants were grown on 2% Glc MS plates under light for 14 days. (C) A genomic ABA2/GIN1 clone restores the gin1 lateral root defects. (D) A genomic ABA2/GIN1 clone restores the gin1 defects in leaves, stems, and siliques. Plants were grown under constant light and watering for 37 days. (E) A genomic ABA2/GIN1 clone restores the gin1 defect in silique size. (F) A genomic ABA2/GIN1 clone restores the gin1 defects in embryos and fertility. Arrows indicate three normal sized seeds in the gin1 silique. (G) A genomic ABA2/GIN1 clone eliminates the wilty phenotype of gin1 (28 days). (H) A genomic ABA2/GIN1 clone restores seed dormancy in gin1. Dry seeds (50 to 80) were allowed to imbibe in water for 12 h and then germinated on water-saturated filter paper (7 cm in diameter) under constant light. The values shown (percent germination) are averages from three independent experiments.

Figure 7.

Glc Regulation of Gene Expression. Glc regulation of ABA biosynthesis and signaling genes. Plants were grown on MS plates without (0) and with 2 and 6% Glc or 2 and 6% mannitol for 14 days under a 16/8-h photoperiod. Mannitol treatment was included as an osmotic control. RT-PCR analyses were performed four times with consistent results. The numbers below the ABA2 gels indicate the band intensity (quantified with 1D Gel Analysis Software; Phoretix, Durham, NC) relative to the loading control, UBQ5, and are normalized to the “0” sample. The PCR product of ABA2 in gin1-3 is 53 bp shorter than that in the wild type as a result of a deletion in the mutant.

Figure 8.

Spatial and Temporal Expression Patterns of ABA2/GIN1. (A) GUS staining in the hypocotyl of a wild-type plant grown on soil for 10 days. Three independent transgenic lines displayed similar patterns. The experiment was repeated twice with consistent results. (B) GUS staining in the hypocotyl of a gin1 plant grown on soil for 10 days. A cross-section of a hypocotyl showing GUS staining in the vascular bundle is shown in the inset. The experiment was repeated twice with consistent results. (C) GUS staining in a wild-type root showing strong signals at the branching point. (D) GUS staining in a wild-type plant grown on soil for 20 days. (E) GUS staining in the cotyledon and first and second true leaves (left to right) excised from the plant shown in (D). (F) GUS staining in a gin1 plant grown on soil for 30 days. The cross-sections of a petiole (left inset) and a leaf vein (right inset) showing GUS staining in the vascular tissues are shown. Wild-type plants displayed similar GUS staining patterns. (G) GUS staining in the pollen of opened flowers in wild-type plants. Young flowers and flower buds did not show GUS staining. (H) GUS staining at the junctions of pedicels and young siliques in wild-type plants. (I) GUS staining in a mature wild-type silique at 57 days after germination. (J) GUS staining in a wild-type seedling (grown on soil for 10 days) after cold treatment (4°C) for 6 h. (K) GUS staining in a gin1 seedling (grown on soil for 10 days) after cold treatment (4°C) for 6 h. (L) GUS staining in a wild-type seedling (grown on soil for 10 days) after NaCl treatment (250 mM) for 6 h. (M) GUS staining in a gin1 seedling (grown on soil for 10 days) after NaCl treatment (250 mM) for 6 h. (N) GUS staining in a wild-type seedling grown on 2% Glc MS medium for 10 days. (O) GUS staining in a gin1 seedling grown on 2% Glc MS medium for 10 days. (P) GUS staining in wild-type (left) and gin1 (right) plants grown on 6% Glc MS medium for 14 days. (Q) GUS staining in a wild-type seedling grown on 2% mannitol MS medium for 10 days. (R) GUS staining in a gin1 seedling grown on 2% mannitol MS medium for 10 days. (S) GUS staining in a wild-type seedling grown on 6% mannitol MS medium for 10 days. (T) GUS staining in a gin1 seedling grown on 6% mannitol MS medium for 10 days.

Figure 9.

Glc Regulation of ABA2/GIN1 Promoter Activity. GUS activity was measured in 16-day-old wild-type and gin1 mutant plants carrying GIN1::GUS and grown on MS plates without (0) or with 2 and 6% Glc (2G and 6G) or 2 and 6% mannitol (2M and 6M). Results from duplicate samples in two independent experiments are shown with error bars. MU, methylumbelliferone; WT, wild type.

Figure 10.

ABA2/GIN1 Subcellular Localization and Biochemical Activities. (A) GIN1-GFP localization. Arabidopsis mesophyll protoplasts expressing a cytosolic GFP marker (CATGFP) or GIN1-GFP were examined by confocal microscopy. Chloroplasts show red autofluorescence. These mesophyll protoplasts have large vacuoles. (B) Glc induces ABA accumulation. Wild-type (WT), gin1, and complemented (gin1/GIN1) seedlings were grown without (−Glc) or with (+Glc) 6% Glc for 10 days. Glc-induced ABA accumulation was abolished in gin1. Results shown are averages of two experiments. FW, fresh weight. (C) Purification of the recombinant ABA2/GIN1/SDR1 protein. His-tagged ABA2/GIN1/SDR1 protein expressed in P. pastoris was purified by ammonium sulfate fractionation (0 to 45% saturation) (lane 1), Ni²⁺ column (stepwise) chromatography (lane 2), and DEAE column chromatography (lane 3). Protein fractions from each step were subjected to SDS-PAGE and stained with Coomassie blue. The arrow indicates the ABA2/GIN1/SDR1 protein. (D) Mass spectra of authentic ABAld and the product from xanthoxin catalyzed by ABA2/GIN1/SDR1. GC-MS fragment patterns of ABAld (top spectrum) and the reaction product (relative abundance) of xanthoxin by ABA2/GIN1/SDR1 (bottom spectrum) are shown.

Figure 11.

Model of the Interplay of the Glc, ABA, and Ethylene Signaling Pathways. This model is based on postgermination development phenotypes of Arabidopsis seedlings.

Figure 12.

Proposed ABA Biosynthesis Pathway and Functions. The ABA biosynthesis steps that are expressed preferentially in leaves are highlighted in the box. It is likely that early ABA precursors have distinct functions in photosynthesis and light stress protection. It is not clear whether these ABA precursors are transported from leaves to petioles, stems, and roots. The expression of NCED3 and SDR1 is regulated differently. Xanthoxin needs to be exported from plastids to be converted to ABAld by SDR1 in the cytosol. The active sites of the GIN1 promoter are listed. New ABA functions presented in this study are listed in boldface.