Overexpression of an Arabidopsis cysteine-rich receptor-like protein kinase, CRK5, enhances abscisic acid sensitivity and confers drought tolerance

Abstract

Receptor-like kinases (RLKs) have been reported to regulate many developmental and defense process, but only a few members have been functionally characterized. In the present study, our observations suggest that one of the RLKs, a membrane-localized cysteine-rich receptor-like protein kinase, CRK5, is involved in abscisic acid (ABA) signaling in Arabidopsis thaliana Overexpression of CRK5 increases ABA sensitivity in ABA-induced early seedling growth arrest and promotion of stomatal closure and inhibition of stomatal opening. Interestingly, and importantly, overexpression of CRK5 enhances plant drought tolerance without affecting plant growth at the mature stages and plant productivity. Transgenic lines overexpressing a mutated form of CRK5, CRK5 (K372E) with the change of the 372nd conserved amino acid residue from lysine to glutamic acid in its kinase domain, result in wild-type ABA and drought responses, supporting the role of CRK5 in ABA signaling. The loss-of-function mutation of the CRK5 gene does not affect the ABA response, while overexpression of two homologs of CRK5, CRK4 and CRK19, confers ABA responses, suggesting that these CRK members function redundantly. We further showed that WRKY18, WRKY40 and WRKY60 transcription factors repress the expression of CRK5, and that CRK5 likely functions upstream of ABI2 in ABA signaling. These findings help in understanding the complex ABA signaling network.

Keywords: ABI2; CRK5; WRKY18; WRKY40; WRKY60.; abscisic acid; drought tolerance; receptor-like kinase.

Fig. 1.

Overexpression of CRK5, but not its mutated form CRK5 ^K372E, results in an ABA-hypersensitive phenotype in early seedling growth. (A) Real-time PCR analysis of the transgenic lines overexpressing CRK5 (OE-1 and OE-2) or a mutated form of CRK5 encoding CRK5^K372E with a point mutation at its kinase domain (CRK5^K372E OE-1 and OE-2). Expression level of CRK5 or CRK5 ^K372E was normalized to that of Actin2/8, and the expression level of CRK5 in Col-0 was set to 1. Values are the mean±SE of three independent biological determinations, and different letters represent significant differences at P<0.05 (Duncan’s multiple range test). (B) Real-time PCR analysis of the CRK5 expression level in wild-type Col-0, and crk5-1 and crk5-2 T-DNA insertion mutant plants. Values are the mean±SE of three independent biological determinations, and different letters represent significant differences at P<0.05 (Duncan’s multiple range test). (C, D) Root growth of wild-type Col-0, crk5-1, crk5-2, OE-1, OE-2 (C) or Col-0, OE-1, OE-2, CRK5^K372EOE-1 and CRK5^K372EOE-2 (D) growing on ABA-free (0 μM) or (±)ABA-containing (0.3 and 0.5 μM) MS medium. Seeds were directly planted in the medium for a 72-h stratification and germinating seeds/young seedlings continued to grow for 10 d before investigation. The experiments were repeated three times with similar results. (E, F) Statistical analysis of absolute (top) and relative values (bottom) of root length of different genotypes described in (C) and (D), respectively. Relative values of the root length of each genotype grown on MS medium containing 0.3 and 0.5 μM (±)ABA are normalized relative to the value of the corresponding genotype at 0 μΜ (±)ABA, which is taken as 100%. Values are the mean±SE of three biological determinations, and different letters represent significant differences at P<0.05 (Duncan’s multiple range test).

Fig. 2.

Overexpression of CRK5, but not its mutated form CRK5 ^K372E, results in an ABA-hypersensitive phenotype in ABA-induced inhibition of cotyledon greening. (A) Cotyledon greening of wild-type Col-0, CRK5-transgenic lines OE-1 and OE-2, and CRK5 ^K372E-transgenic lines CRK5^K372EOE-1 and CRK5^K372EOE-2 in ABA-free (0 μM, top) or (±)ABA-containing (0.5 μM, bottom) MS medium. (B) Percentages of green cotyledons of the different genotypes as described in (A). Green cotyledons were scored 5 days after stratification. Values are the mean±SE of three biological determinations, and different letters represent significant differences at P<0.05 (Duncan’s multiple range test).

Fig. 3.

Overexpression of CRK5, but not CRK5 ^K372E, results in ABA hypersensitivity in stomatal movement and increases plant drought tolerance. (A) ABA-induced inhibition of stomatal opening (left) and promotion of stomatal closure (right) in wild-type Col-0 plants, CRK5-transgenic lines OE-1 and OE-2, and CRK5 ^K372E-transgenic lines CRK5^K372EOE-1 and CRK5^K372EOE-2. The experiments were repeated five times with similar results. The values are the mean±SE from 60 stomata for each time point, and different letters represent significant differences at P<0.05 (Duncan’s multiple range test). (B) Test of drought tolerance of the different genotypes described above. Plants were well watered (control, ‘Well water’) or drought stressed by withholding water (‘Drought’) for 16 d (D) and then re-watered (‘Rewater’). The experiments were repeated five times, and at least 30 plants per individual line were used for each experiment. (C) Survival rates of the plants described in (B). The values are the mean±SE of three biological determinations, and different letters represent significant differences at P<0.05 (Duncan’s multiple range test).

Fig. 4.

Effects of over-expression of CRK5 on plant growth and productivity. Two-week-old (A; bar=1cm), 4-week-old (B; bar=3cm), and 6-week-old seedlings (C; bar=4cm), and siliques (D; bar=1cm) are shown for wild-type Col-0 plants, CRK5-transgenic lines OE-1 and OE-2, and CRK5 ^K372E-transgenic lines CRK5^K372EOE-1 and CRK5^K372EOE-2. (E) Statistics of the plant height, total silique number per plant, silique length and total seed weight (dry weight) per plant of the different genotypes as described above. Each value is the mean±SE of three biological determinations, and different letters represent significant differences at P<0.05 (Duncan’s multiple range test).

Fig. 5.

Test of genetic interaction of CRK5 with ABI2 involved in ABA signaling or with ABA2 involved in ABA biosynthesis. (A) ABI2 is genetically epistatic to CRK5. Seeds of wild-type Col-0, CRK5-overexpression line OE-1, ABI2-overexpression line ABI2OE and CRK5/ABI2-double-overexpression line OE1×ABI2OE were directly planted on the ABA-free (0 μM) or 0.8 μM-ABA-containing MS medium, and the growth status was recorded 10 d after stratification. The experiments were repeated three times with similar results. (B) Statistical analysis of absolute (top) and relative values (bottom) of root length of different genotypes described in (A). Relative values of the root length of each genotype grown on ABA-containing medium are normalized relative to the value of the corresponding genotype at 0 μM ABA, which is taken as 100%. Values are the mean±SE of three biological determinations, and different letters represent significant differences at P<0.05 (Duncan’s multiple range test). (C) Loss-of-function of ABA2 (aba2) does not affect ABA-hypersensitive response of the CRK5-overexpression line OE-2. Seeds of wild-type Col-0, aba2 mutant, CRK5-overexpression line OE-2 and CRK5-overexpression line OE2 in the aba2 mutant background (OE-2×aba2) were directly planted on the ABA-free (0 μM) or 0.5 μM-ABA-containing MS medium, and the growth status was recorded 10 d after stratification. The experiments were repeated three times with similar results. (D) Statistical analysis of absolute (top) and relative values (bottom) of root length of different genotypes described in (C). Relative values of the root length of each genotype grown on ABA-containing medium are normalized relative to the value of the corresponding genotype at 0 μM ABA, which is taken as 100%. Values are the mean±SE of three biological determinations, and different letters represent significant differences at P<0.05 (Duncan’s multiple range test).

Fig. 6.

Subcellular localization of CRK5 protein and expression profile of CRK5 gene. (A) CRK5 is localized to plasma membrane. The Col-0 plants were transformed with the construct carrying CRK5-GFP or empty GFP, respectively, driven by CaMV 35S promoter, and the roots of transgenic plants were investigated by a confocal laser scanning microscope. (a) CRK5–GFP localization in the mature root zone. (b) FM4-64 staining of the CRK5-GFP-transgenic plant in the mature root zone. (c) The corresponding bright field of (a) and (b). (d) Merged imagine of (a), (b) and (c). (e) Empty GFP localization in the mature root zone. (f) FM4-64 staining of GFP-transgenic plants in the mature root zone. (g) The corresponding bright field (e) and (f). (h) Merged imagine of (e), (f) and (g). Bars=20 μm. (B) Expression of the CRK5-promoter–GUS in transgenic lines. (a) Dry seed. (b) Young seedling 48h after stratification. (c) Young seedling 72h after stratification. (d) Young seedling 14 d after stratification. (e) Young seedling 21 d after stratification. (f) Rosette leaves and stomata (shown at bottom, right). (g) Flower. (h) Silique. (C) Relative expression levels of CRK5 in different tissues/organs determined by real-time PCR analysis.

Fig. 7.

Expression of some ABA-responsive genes in CRK5-transgenic line OE-1 and CRK5 ^K372E-transgenic line CRK5^K372EOE-1. The seeds were germinated and grown on ABA-free (–ABA) or 0.5 μM-ABA-containing (+ABA) MS medium for 4 days before sampling for RNA extraction. Transcription levels of these genes were assayed by real-time PCR. Expression level of each gene is normalized to that of Actin2/8, and the relative expression level of each gene is normalized relative to the level of this gene of the wild-type Col grown in ABA-free medium, which is taken as 1. Values are the mean±SE of three independent biological determinations, and different letters represent significant differences at P<0.05 (Duncan’s multiple range test).

Fig. 8.

Phenotypes of the transgenic lines overexpressing CRK4, CRK19 or CRK20 homologous to CRK5 in ABA-induced early seedling growth arrest. (A) Phylogenic analysis of Arabidopsis CRK4, CRK5, CRK19 and CRK20 using the neighbor-joining method with MEGA version 4 by alignment of the amino acid sequences with ClustalW. (B) Real-time PCR analysis of the transgenic lines. C4OE-1 and C4OE-2 denotes CRK4-overexpression lines; C19OE-1 and C19OE-2, CRK19-overexpression lines; C20OE-1 and C20OE-2, CRK20-overexpression lines. Values are the mean±SE of three independent biological determinations. (C) Phenotypes of ABA-induced inhibition of early seedling growth in different transgenic lines as described in (B). Seeds were directly planted in ABA-free (0 μM ABA, top) or 0.6 μM-ABA-containing MS medium, and the growth status was recorded 10 d after stratification. The experiments were repeated three times with similar results. (D) Statistical analysis of absolute (top) and relative values (bottom) of root length of different genotypes described in (C). Relative values of the root length of each genotype grown on ABA-containing medium are normalized relative to the value of the corresponding genotype at 0 μM ABA, which is taken as 100%. Values are the mean±SE of three biological determinations, and different letters represent significant differences at P<0.05 (Duncan’s multiple range test).

Fig. 9.

Overexpression of CRK4, but not CRK19 or CRK20, results in ABA hypersensitivity in stomatal movement and increases plant drought tolerance. (A) ABA-induced inhibition of stomatal opening (left) and promotion of stomatal closure (right) in wild-type Col-0 plants, CRK4 (C4OE1, C4OE2), CRK19 (C19OE1, C19OE2) and CRK20 (C20OE1, C20OE2) transgenic lines. The experiments were repeated five times with similar results. The values are the mean±SE from 60 stomata for each time point, and different letters represent significant differences at P<0.05 (Duncan’s multiple range test). (B) Test of drought tolerance of the different genotypes described above. Plants were well watered (control, ‘Well water’) or drought stressed by withholding water (‘Drought’) for 16 d and then re-watered (‘Rewater’). The experiments were repeated five times, and at least 30 plants per individual line were used for each experiment. (C) Survival rates of the plants described in (B). The values are the mean±SE of three biological determinations, and different letters represent significant differences at P<0.05 (Duncan’s multiple range test).

Fig. 10.

Test of the interaction of WRKY18, WRKY40 and WRKY60 with the promoter of the CRK5 gene. (A) Promoter diagram of the CRK5 gene. W1–W10 represent the W-box numbered from left to right with their location sites relative to the start codon (ATG). The segments marked with ProCRK5-1, ProCRK5-2 and ProCRK5-3 indicate the probe fragments used in the gel shift assays described in Fig. 11. (B) Yeast one-hybrid assays to test the interaction of WRKYs with the CRK5 promoter. Yeast cells were co-transformed with pGADT7 prey vector containing WRKY18, WRKY40, or WRKY60 and pHIS2 bait vector containing the promoter of CRK5. The corresponding transformation with pGADT7 prey vector containing WRKY18, WRKY40, or WRKY60 and pHIS2 bait vector containing p53 promoter fragment were used as negative controls. Co-transformation of pHIS2-p53 and pGADT7-p53 was used as positive control, and co-transformation of pGADT7-p53 and empty pHIS2 was used as its own negative control. Three 10-fold series dilutions were dropped vertically for each assay on SD-2 medium (synthetic dropout medium lacking Leu, Trp) and SD-3 medium (synthetic dropout medium lacking Trp, Leu, His) supplemented with 40mM 3-aminotriazole (3-AT). All the experiments were repeated five times with the same results. (C) WRKY18, (D) WRKY40 and (E) WRKY60 inhibit the transcription activity of the CRK5 promoter in tobacco system, assayed with luciferase (LUC) imaging. Tobacco leaves were co-transformed with the constructs ProCRK5-LUC plus WRKY18-FLAG or ProCRK5-LUC plus empty FLAG (C), or with the constructs ProCRK5-LUC plus WRKY40-FLAG or ProCRK5-LUC plus empty FLAG (D), or with the constructs ProCRK5-LUC plus WRKY60-FLAG or ProCRK5-LUC plus empty FLAG (E). Top panels in (C), (D) and (E): LUC fluorescence imaging. Bottom panels in (C), (D) and (E): optical densities calculated with the ImageJ software. The experiments were repeated three times with similar results. Each value for the columns in (C), (D) and (E) are the mean±SE of three biological determinations, and different letters represent significant differences at P<0.05 (Duncan’s multiple range test).

Fig. 11.

Gel shift assays to test interaction of WRKY18, WRKY40 or WRKY60 with the promoter of CRK5 gene. (A) WRKY18 binds to the ProCRK5-1, ProCRK5-2 and ProCRK5-3 fragments. 6×His-W18 indicates the purified 6×His-WRKY18 fusion protein; 6×His, the 6×His tag peptide served as negative control; Biotin-Probe, the biotin labeled CRK5 promoter fragments ProCRK5-1, ProCRK5-2, ProCRK5-3; mW1 and mW2/W3, the two mutant forms in W-boxes (W1 single mutation and W2/W3 double mutation) of biotin labeled ProCRK5-1 with W-box mutation at W1: TTGACT→TTAAAT, and W-box mutation at W2/W3: TTGACCTTGACT→TTAAACTTAAAT; Cold-Probe, the three unlabeled CRK5 promoter fragments; 50× and 200×, 50-fold or 200-fold cold-probe relative to the labeled probes added, respectively, for binding competition; Free-Probes, the labeled probes that do not bind the WRKY protein; WRKY18/ProCRK5, the shift bands of the complex of WRKY18 protein with corresponding ProCRK5 fragments. Each experiment was repeated three times with the same results. (B) WRKY40 binds to the ProCRK5-1, ProCRK5-2 fragments, but does not bind to the ProCRK5-3 sequence. 6×His-W40 denotes the purified 6×His-WRKY40 fusion protein; WRKY40/ProCRK5, the shift bands of the complex of WRKY40 protein with the corresponding ProCRK5 fragments. Other symbols are the same as described above in (A). Each experiment was repeated three times with the same results. (C) WRKY60 does not bind to any of the three promoter segments of CRK5 gene. 6×His-W60 denotes the purified 6×His-WRKY60 fusion protein, and other symbols are the same as described above in (A). Each experiment was repeated three times with the same results.

Fig. 12.

Expression of CRK4, CRK5, CRK19 and CRK20 genes in wrky loss-of-function mutants. (A–D) Two-week-old seedlings grown on MS medium (A, C) or rosette leaves of 4-week-old seedlings (B, D) were sampled for RNA extraction. The transcription levels of CRK4, CRK5, CRK19 and CRK20 were assayed in the wrky single, double (wrky40 wrky18, wrky18 wrky60, and wrky40 wrky60), and triple (wrky40 wrky18 wrky60) mutants by real-time PCR. Actin2/8 was used as internal control. Each value is the mean±SE of three independent experiments, and the letters indicate significant differences at P<0.05 (Duncan’s multiple range test).