OsWDRP3-RNAi knockdown transgenic plants display erect leaves
To investigate the biological function of OsWDRP3, an OsWDRP3 mutant (wdrp3, 2A-20340) was isolated from a T-DNA insertion library with the cv Dongjin background (Jeon et al. 2000; Jeong et al. 2006). wdrp3 plants were short with a small leaf angle, but the number of seeds was very small; thus, we developed OsWDRP3-overexpressing and OsWDRP3-silenced lines to characterize the roles of OsWDRP3 in rice plants. Over 20 independent T0 OsWDRP3 RNAi lines were obtained. We then selected the RNAi14 line with the lowest OsWDRP3 gene expression to analyze the gene functions (Fig. 1A). We also generated transgenic rice plants overexpressing OsWDRP3 (OE) under the control of a maize ubiquitin promoter. The expression level of OsWDRP3 was higher in OE line 9 than in wild-type plants. We selected OE line 9 to analyze the phenotype (Fig. 1B). In this study, all the transgenic lines, including the OE and RNAi lines, were in the background of Hejiang19, a japonica variety.
OsWDRP3-RNAi lines showed a reduction in plant height (to 70%-75% of the wild-type height, n = 15, P < 0.01). OsWDRP3-OE plants were slightly higher than the wild ones, but there was no significant difference (Fig. 1C). Internode elongation was inhibited in all of the RNAi plants. We measured the lengths of the individual internodes of plants. The results showed that each internode of OsWDRP3-RNAi plants had a different length than the wild type. The first, second, third and fourth internodes were shortened by 17%, 41%, 48% and 53%, respectively (Fig. 1D). Based on the proportion of each internode to the total internode length, OsWDRP3-RNAi plants belong to type D6 dwarfism, which is usually related to BR synthesis or signal transduction correlation (Takeda 1974). Additionally, although the plant height of the overexpressed transgenic plants was similar to that of the wild type, its internode length changed, in which the first internode extended relative to the wild type. It was 39% longer in the first internode, 2% shorter in the second, 43% shorter in the third and 88% shorter in the fourth compared with the wild type. Some overexpressed transgenic plants did not even have the fourth internode.
In wild-type rice, the leaf blade bends away from the vertical axis of the leaf sheath, whereas the leaves of OsWDRP3-RNAi seedlings were erect at the seedling and adult stages (Fig. 1). The second lamina joint angle of 9-day-old wild-type seedlings was 17.49 ± 4.09°, whereas that of the OsWDRP3-RNAi was 5.80 ± 1.63° (Fig. 1E), representing a 3-fold increase in the lamina angles of wild-type relative to that of the OsWDRP3-RNAi plants. At the heading and harvest stages, the flag leaf angle, second leaf angle and third leaf angle of OsWDRP3-RNAi plants were reduced by 36%, 28% and 32%, respectively, compared with the wild type (Fig. 1F). Since BR regulates the bending of the lamina joint (Yamamuro et al. 2000; Hong et al. 2003), these results suggest that OsWDRP3 may be involved in BR biosynthesis or signal transduction in rice.
OsWDRP3 participates in the regulation of BR synthesis and transduction
We checked the expression of OsWDRP3 in wild-type rice under BR treatment by quantitative PCR. Seven DAG wild-type rice seedlings were treated with 1 μM 24-epibrassinolide (24-eBL, the active BL), and leaves treated with 0 h, 2 h, 4 h, 8 h, 12 h and 24 h were extracted. Total RNA was extracted, and cDNA was synthesized by reverse transcription in vitro. OsWDRP3 gene expression increased after 2 h of treatment with 1 μM 24-eBL and decreased after 4 h of treatment (Fig. 2A). The results showed that BR induced OsWDRP3 gene expression, suggesting that the OsWDRP3 gene may be involved in the BR signaling pathway.
BR biosynthesis and signal transduction are modulated in an intricate manner in plants (Tanaka 2005). OsBRI1 and OsBZR1 have been identified, and their loss-of-function mutants display erect leaves (Yamamuro et al. 2000; Bai et al. 2007). OsBZR1 directly regulates the expression of the DWARF AND LOW TILLERING (DLT) gene, an important BR response factor in rice (Tong et al. 2009). To investigate whether OsWDRP3 was involved in the BR signaling pathway, we examined the expression of BR-related genes in wild type and OsWDRP3-RNAi. The expression of OsBRI1 was downregulated in wild-type plants under BR treatment, which was consistent with a previous report (Yamamuro et al. 2000). In RNAi transgenic plants, the expression of OsBRI1 was downregulated, and there was little change in the expression level after BR treatment. The results also showed that the BR signaling pathway regulation gene OsBAK1 upregulated expression in wild-type plants under BR treatment. Additionally, the expression of OsBAK1 was upregulated in RNAi plants but decreased after BR treatment, which was different from wild-type plants. The expression levels of BZR1, which plays positive roles in BR signaling, were upregulated when BR treatment was applied to wild-type plants, but no significant change occurred in OsWDRP3-RNAi plants (Fig. 2B).
We therefore analyzed the expression of BR biosynthetic genes in WT and OsWDRP3-RNAi (Tanabe et al. 2005; Hong et al. 2003). It was previously shown that the expression of CPD, which is involved in BR biosynthesis, is downregulated by BL treatment (Tanaka et al. 2005). Hence, the expression level of CPD can be used as an indicator of BL perception. We studied the effect of eBL on the expression of BR-specific biosynthetic genes in wild-type and OsWDRP3-RNAi plants. When the plants were treated with eBL, the expression of these genes in the wild-type plants was visibly reduced but not affected in OsWDRP3-RNAi plants (Fig. 2C).
qRT–PCR results suggested that BR signal transduction was impaired in the OsWDRP3-RNAi plants. BR signaling represses BR synthesis in a negative feedback manner, and impaired BR signaling usually promotes the transcript accumulation of BR biosynthesis genes (Bai et al. 2007; Tong et al. 2009). These results suggest that OsWDRP3 might be a positive regulator in the BR signaling pathway.
OsWDRP3-RNAi plants show BR-insensitive phenotypes
To examine whether the leaf-bending morphology of OsWDRP3-RNAi was related to BR signaling, we tested the effect of exogenous BR on wild-type plants and RNAi plants.
First, we examined the coleoptile elongation of wild-type and OsWDRP3-RNAi seedlings in response to eBL. Rice seeds were germinated in MS medium in the dark with or without 1 μM eBL. In wild-type plants, coleoptile elongation was promoted when 1 μM eBL was added to the medium, but the same experimental phenomenon was not seen in OsWDRP3-RNAi plants (Fig. 3A).
Second, we tested the response of root growth of the wild-type and OsWDRP3-RNAi seedlings to eBL. In the absence of eBL, the root growth of OsWDRP3-RNAi plants was obviously shorter than that of the wild type. (Fig. 3 B). In the presence of 1 μM eBL, root elongation in wild-type seedlings was inhibited, whereas OsWDRP3-RNAi showed the same growth status before and after treatment (Fig. 3B).
eBL can promote bending at the lamina joint, and BR-deficient and BR-insensitive mutants show decreased bending of the lamina joint (Cao and Chen 1995; Tanabe et al. 2005). We analyzed the sensitivity of OsWDRP3-RNAi and OsWDRP3-OE to eBL with a lamina joint assay (Yamamuro et al. 2000). The lamina inclination assay revealed that the OsWDRP3-RNAi plants displayed an insensitive phenotype to BR. After treatment with 100 nM 24-eBL, the second lamina joint angle of the 7-day-old wild-type and OsWDRP3-RNAi seedlings increased by 8.12 ± 0.18 and 4.83 ± 0.14 times, respectively (Fig. 3C). The lamina angles of wild-type plants increased significantly after eBL treatment, whereas those of OsWDRP3-RNAi had no obvious response. There was no significant difference between OsWDRP3-OE and wild-type plants (Fig. 3C).
These results suggested that the OsWDRP3-RNAi seedlings were less sensitive to exogenous BR than wild-type seedlings.
OsWDRP3 is a WD40 repeat protein and widely express in rice
The deduced amino acid sequence of OsWDRP3 consists of 345 residues. The primary sequence contains seven WD40 domains, one WDxR motif and one putative DWD box (Fig. 4A). WD40 domains often fold into the most ideal and stable seven-bladed β-propellers with a funnel-like shape (Jain and Pandey 2018). We analyzed the WD40 domain in OsWDRP3 (https://swissmodel.expasy.org/). There were seven WD40 repeat domains and a funnel-like shape, as displayed in Figure 1A. This result suggests that OsWDRP3 may be involved in intracellular protein interactions similar to other WD40 proteins.
Comparison of amino acid sequences indicated that OsWDRP3 shared high identities with three WD repeat proteins in Arabidopsis (At3g15610, 79% sequence identity; At1g52730, 79% sequence identity; At1g15470/XIW1, 74% sequence identity) (Fig. 4B). These data clearly suggested that OsWDRP3 is one of the closest relatives of WD repeat proteins in Arabidopsis.
We performed RT–qPCR assays using total RNA from various tissues of plants at the seedling stage and reproductive stage, including panicles, different internodes, nodes, leaf blades, leaf sheaths, and roots. OsWDRP3 expression was detected in all tissues examined but primarily in dividing tissues, including the uppermost internode, node, and leaf sheath. The abundance of transcripts encoding OsWDRP3 was highest in the internode at 60 days after germination (DAG) and lower in the seedling, stem and lamina joints of the japonica rice variety Hejiang19 (Fig. 5A).
To analyze the expression pattern of OsWDRP3 in detail, we cloned a 2-kb promoter region of OsWDRP3 to drive the GUS gene and then transformed rice plants. GUS staining was observed in various organs in transgenic plants. Overall, the histochemical staining results showed that OsWDRP3 expressed in most tissues throughout plant growth and development. A high level of expression was primarily observed in the root, pollen and internodes (Fig. 5B). We examined the lamina joint region, leaf blade and sheath in 9-d-old transgenic seedlings (Fig. 5B). The veins in the leaf blade and sheath were well stained (Fig. 5B). In the reproductive stage, staining was observed in the vascular bundles of the lemma and palea in particular (Fig. 5B), and the anther was also stained (Fig. 5B). Thus, OsWDRP3 functions throughout plant growth and development.
To investigate the subcellular localization of OsWDRP3, we constructed a pUbi::OsWDRP3-GFP construct. We observed the fluorescence of the OsWDRP3-GFP fusion protein and found that GFP was distributed in the cytoplasm (Fig. 5C), indicating cytoplasmic localization of OsWDRP3.
OsWDRP3 interacts with OsBAK1 and regulates its degradation in rice
Based on the bioinformatics analysis, OsWDRP3 is a serine-threonine kinase receptor-associated protein. Combined with the previous phenotypic analysis results, we speculated that OsWDRP3 might interact with BR-related serine-threonine kinases.
Brassinosteroid insensitive 1 (BRI1) is the BR receptor in plants. BRI1 heteromerizes with BRI1-associated kinase 1 (BAK1) and forms a receptor complex to activate BR signaling. We chose OsBRI1 and OsBAK1 to analyze the interactions with OsWDRP3.
To determine whether WDRP3 and BAK1 or BRI1 interact in plant cells, we performed bimolecular fluorescence complementation (BiFC) in Agrobacterium tumefaciens-infiltrated tobacco (N. benthamiana) and rice protoplasts. We fused WDRP3 to the N-terminal yellow fluorescent protein (YFP) fragment (OsWDRP3-N-YFP) and the intracellular kinase region of OsBAK1 (BAK1K) or OsBRI1 (BRI1K) to the C-terminal YFP fragment (BAK1K-C-YFP, BRI1K-C-YFP). In control experiments, WDRP3-N-YFP coexpressed with unfused C-YFP or unfused N-YFP coexpressed with BAK1K-C-YFP (Fig. 6 A and B). When fused WDRP3-N-YFP was coexpressed with BAK1K-C-YFP in tobacco leaves or rice protoplasts, fluorescence signals were detected in transformed cells (Fig. 6 A and B), but no signal was detected in WDRP3-N-YFP/BRI1K-C-YFP coexpressed cells.
OsBAK1-GFP fusion protein is principally located on the plasma membrane. However, OsBAK1-GFP was abundant in the cytoplasm during plasmolysis experiments and distributed in a spot-like way (Li et al., 2009). This result suggested that BAK1 might break off above the membrane and enter the cytoplasm under treatment conditions. Our BiFC results showed that OsBAK1 was also localized in the cytoplasm when it interacted with OsWDRP3 in rice protoplasts (Fig. 6B).
Coimmunoprecipitation assays showed that OsWDRP3 interacts with OsBAK1 in planta. Rice seedling leaves isolated from the Ubi::OsWDRP3-Myc transgene plant were ground in protein extraction buffer. After centrifugation, the supernatant was used as a total protein extract. Protein extracts (input) were immunoprecipitated with anti-MYC magnetic beads. Immunoblots were developed with anti-BAK1 antibody to detect BAK1 and with anti-MYC to detect OsWDRP3 (Fig. 6C).
We used in vitro degradation experiments to investigate whether OsWDRP3 promotes the degradation of the OsBAK1 protein. Purified OsBAK1-6His was incubated with protein extracts from wild-type, OsWDRP3-RNAi, and OsWDRP3-OE plants and detected with anti-6 × His, anti-OsBAK1, and anti-plant actin antibodies. As shown in Figure 7, when the purified OsBAK1259-624 and OsBAK1307-624 proteins were incubated with the wild-type protein extract, only OsBAK1259-624 proteins were degraded, while OsBAK1307-624 proteins were not degraded (Fig. 7A). Subsequently, we examined the effect of OsWDRP3 on the degradation of OsBAK1259-624 protein. OsBAK1259-624 protein was incubated with protein extracts from OsWDRP3-OE, wild-type and OsWDRP3-RNAi plants, and the degradation rate of OsBAK1259-624 protein was slowed down in turn (Fig. 7B).
To verify the effect of OsWDRP3 on the degradation of OsBAK1259-624 protein, we added purified OsWDRP3 into the protein extract of wild-type plants and then incubated OsBAK1259-624 protein. The degradation rate of OsBAK1259-624 protein in the reaction system with OsWDRP3 added was faster than that in the system without OsWDRP3 added (Fig. 7B and 7C).