Overexpression of the maize WRKY114 gene in transgenic rice reduce plant height by regulating the biosynthesis of GA

ABSTRACT

WRKYs represent an important family of transcription factors that are widely involved in plant development, defense regulation and stress response. Transgenic rice that constitutively expressed ZmWRKY114 had shorter plant height and showed less sensitivity to gibberellic acid (GA₃). Further investigation proved that transgenic rice accumulated lower levels of bioactive GAs than that in wild-type plants. Application of exogenous GA₃ fully rescued the semi-dwarf phenotype of ZmWRKY114 transgenic plants. Transcriptome and qRT-PCR analyses indicated that the expression of OsGA2ox4, encoding the repressor of GA biosynthesis, was markedly increased. Electrophoretic mobility shift assay and dual-luciferase reporter assay indicated that ZmWRKY114 directly binds to a W-box motif in the OsGA2ox4 promoter. Taken together, these results confirm that ZmWRKY114 is a GA-responsive gene and is participated in the regulation of plant height in rice.

KEYWORDS:

• GA biosynthesis
• plant height
• rice
• RNA-seq
• ZmWRKY114

Introduction

Plant height is an important agronomic trait that significantly affect crop yield. Crop with a suitable height often showed improved lodging resistance in unfavorable environments and increased total biomass attributing to enhanced nitrogen-use efficiency.¹

Gibberellic acids (GAs), a large class of tetracyclic diterpenoid phytohormone, were proved to play vital role in the regulation of plant height.² Usually, plants with a dwarf or semi-dwarf phenotype are deficient in GA synthesis. For example, the reduced internode length was often considered to be the most prominent phenotype of GA biosynthesis mutants in Arabidopsis.³ Several of the identified dwarf mutants in rice are also characterized as GA-deficient or GA-insensitive mutants.^4–6 Therefore, it is essential for plants to produce and maintain right levels of bioactive GAs (mainly GA₁, GA₃, GA₄, and GA₇) to guarantee normal growth and development.⁷

WRKY transcription factors are one of the biggest families of regulatory proteins in higher plants. In different species, the WRKY family can be classified into three major structural groups, based both on the number of WRKY domains and the types of their zinc finger motifs.⁸ Recent research has supplied some evidence to prove that WRKY members also play important roles in the regulation of plant height. For example, OsWRKY6-overexpressing rice lines shows a growth-retardation phenotype at the seedling and tillering stages.⁹ Mutation of rice WRKY11 gene leads to a semi-dwarf and late flowering phenotype.¹⁰ OsWRKY78 RNAi lines also exhibited a semi-dwarf and small kernel phenotype because of reduced cell length in these transgenic plants.¹¹ In cotton, silence of WRKY42 gene in VIGS (virus-induced gene silencing) plants significantly reduced plant height.¹²

Our previous results showed that ZmWRKY114 encoding a WRKY transcription factor involved in plant response to salt stress.¹³ In this study, we surveyed the functions of this gene in rice growth and development. Our data proved that ZmWRKY114 is participated in regulating plant height through GA signaling.

Materials and methods

Plant materials and growth conditions

The ZmWRKY114-overexpressing transgenic rice lines reported in our previous article were used in this study.¹³ Zhong Hua 11 (ZH11) rice was used as the control plant for phenotypic examination. The transgenic and wild type (WT) plants were planted in a blue plastic basin which contained field soil and grown in a greenhouse with controlled environmental conditions (28°C, 60% relative humidity and 16 h-light /8 h-dark photoperiod).

Phenotypic analysis

The heights of at least 10 plants were measured at the heading stage in the field. For paraffin section, the penultimate internodes were collected from WT and two transgenic lines and fixed in FAA solution (50% alcohol: formalin (37–40% formaldehyde): glacial acetic acid;18:1:1). Next, the internodes were dehydrated in a series of ethanol solutions (70% (v/v) ethanol, 80% (v/v) ethanol, 85% (v/v) ethanol, 90% (v/v) ethanol, 95% (v/v) ethanol, and anhydrous ethanol) and detained in a series of xylene solutions (3:1 ethanol: xylene, 1:1 ethanol: xylene, 1:3 ethanol: xylene, and pure xylene). The samples were dissected and then observed by light microscope (Leica). Parts of these experiments were performed by Servicebio company (Servicebio technology CO, Wuhan, China).

The GA sensitivity assay of transgenic seedlings

In order to investigate the effect of ZmWRKY114 on plant growth, GA sensitivity assay was firstly performed. The surfaced-sterilized rice seeds of WT and two transgenic lines (OE8 and OE6) were soaked in sterile water at 28°C for 2 days under darkness. Seeds with the same size were inoculated into a sterile box containing 0.5 × MS medium supplemented with 0, 0.1, 1, 10 or 100 μM GA₃ + 200 μM Paclobutrazol (PAC, a GA biosynthetic inhibitor, to exogenous GA₃ treatments) and 200 μM PAC alone. After grown in a growth chamber for 14 days, the seedling height and root length of 30 seedlings for each sample were measured. Each treatment was repeated three times.

Quantification of endogenous GAs

The 4-week-old seedlings of transgenic and WT plants were sampled for quantitative GA analysis. The samples were frozen in liquid nitrogen and ground into fine powder. Hormones were extracted by 10 mL methanol containing 20% water (v/v) for 12 h at 4°C. After centrifugation at 10,000 × g (4°C) for 5 min, the extracts were collected and passed through a C18 SPE cartridge, then dried in N2. The HPLC-MS/MS analysis was performed using an AB SCIEX QTRAP 6500 System (AB SCIEX, USA). The quantification of endogenous GA₁ and GA₃ was performed as described by.¹⁴

Growth chemical treatments

The 4-week-old transgenic seedlings showing a semi-dwarf phenotype were sprayed with 20 mL of 10 μM GA₃ (in water). The treatments were repeated every 15 days for 2 times. Untreated WT and transgenic plants were used as the control. The plant height was measured at 30 days of treatment.

RNA sequencing analysis

Total RNA was extracted from 4-week-old seedlings of WT and two transgenic lines. The cDNA library construction and transcriptome sequencing were performed by Geneseeq company (Nanjing, China) using an Illumina Hiseq 2000 platform. Gene expression levels were estimated based on fragments per kilobase of exon per million fragments mapped (FPKM) using the Cuffinks software (version: 2.2.1). According to the FPKM value, differentially expressed genes (DEGs) between WT and transgenic line were identified. FDR ≦0.05 and ∣log₂ (FPKM-transgenic/FPKM-WT)∣≧1 and FPKM ≧2 were regarded as thresholds to determine the significance of differences in gene expression.

Quantitative real-time PCR (qRT-PCR) analysis

Four plant height related genes were selected to verify the RNA-seq analysis using qRT-PCR. qRT-PCR was performed using the ABI 7300 real-time PCR system (Applied Biosystems), and the actin gene was used as an internal control. The PCR reaction system and thermal cycle conditions were referred to previous study.¹⁵ The primers used for qRT-PCR analysis are listed in Table S1, and the relative expression level was calculated using the 2^−ΔΔCT method.¹⁶ Each PCR reaction was performed with three biological replicates and three technical repeats.

Electrophoretic mobility shift assay (EMSA)

Based on RNA-seq and qRT-PCR data, biotin labeled probes containing W-boxes existed in the promotor of OsGA2ox4 were synthesized (Sangon Biotech, Shanghai, China). The EMSA assays were conducted using Thermo Fisher Chemiluminescent EMSA Kits, according to the manufacturer’s instructions (Thermo Fisher Scientific). In brief, probes and proteins were mixed and transferred to Hybond nylon membranes using a sandwich method, and they were cross-linked to the membrane by a UV-light cross-linker. Then, the membranes were transferred to blocking solutions for 15 min with gentle shaking. The membranes were then incubated in another blocking solution with 1:2000 diluted streptavidin-HRP conjugate antigens at RT (15 min) with gentle shaking, and the membranes were washed 4 times with a washing buffer. Finally, 1 mL Luminol/Enhancer Solution and 1 mL Stable Peroxide Solution were added to the membrane and the signals were detected using the Gel Doc EZ gel image analysis system (Bio-Rad, Hercules, CA, USA).

Dual-luciferase reporter assay

The full-length coding sequence of ZmWRKY114 cDNA was inserted into the constructed pGreenII-62-SK vector as the effector, driven by CaMV 35S promoter, and the empty vector was used as a control. The 256 bp OsGA2ox4 promoter fragment containing W-box was ligated into the transient expression pGreenII0800-LUC vector. The primers W1-F, W1-R, W2-F and W2-R are listed in Table S1. The fusion vectors and pSoup helper plasmid were co-introduced into A.tumefaciens strain GV3101. The infiltrated leaves were sprayed with luciferin at 2–3 days post-infiltration and coexpression proteins were monitored using an In Vivo Plant Imaging System (Berthold Technologies, Germany). Cells were lysed and luciferase activity was measured using the dual-luciferase reporter assay kit (TransGen Biotech). The promoter activities were indicated by the LUC/REN ratio.

Results

Overexpression of ZmWRKY114 led to a semi-dwarf phenotype in rice

Our previous study has proved that transgenic plants overexpressing a maize WRKY114 gene displayed a semi-dwarf phenotype at the seedling stage.¹³ In fact, the transgenic plants were shorter than the WT almost in the whole developmental process of rice (Figure 1a and S1). At the heading stage, the average plant height of WT was 91.46 cm; the average height of two ZmWRKY114 overexpression lines, OE8 and OE6 was 63.67 cm and 57.87 cm respectively (Figure 1b). In addition, as shown in Figure 1c and d, OE8 and OE6 were significant shorter than WT in top three internodes.

Figure 1. Phenotypes characterization of the WT and ZmWRKY114-overexpressing plants. (a) WT and ZmWRKY114-overexpressing plants at the heading stage. Bars = 20 cm. (b) Plant height at heading stage. Data are mean ± SD from at least 10 plants. (c) Flag leaves, panicles and internode phenotypes of WT and transgenic plants. fl., flag leaf; P, panicle, I to Ⅳ indicate internodes from top to bottom. Bar, 5 cm. (d) Panicle and internode lengths of WT and transgenic plants. Values are means ± SD of 15 plants. (e) Longitudinal view of parenchyma cells in penultimate internodes of WT and transgenic plants. Bar, 50 μm. (f-h) cell numbers, Cell lengths and cell widths of penultimate internodes in WT and transgenic plants. Values are means ± SD (n = 3). Significant difference was checked by Student’s t-test (* P < .05, **P < .01)

In order to determine the main reason for the semi-dwarf phenotype of ZmWRKY114 transgenic lines, the penultimate internodes of WT and transgenic lines were used for paraffin section observation. Compared to WT, the cell number of OE8 and OE6 was decreased for 34.2% and 28.8%, and the cell width was increased for 52.9% and 44.2% respectively. In addition, there was no significant differences in cell length between WT and transgenic lines (Figure 1e-h). These results showed that a decrease in cell number is the major cause of the dwarfism phenotype in ZmWRKY114 transgenic plants.

ZmWRKY114 overexpression reduced sensitivity to GA₃ in rice

To further determine the factors of height reduction in transgenic rice, GA₃ and PAC treatments were performed on WT and transgenic rice. There was no difference in seedling heights between transgenic lines and WT plants when treated with 200 μM PAC, but OE8 and OE6 were shorter than WT plants when 0.1, 1, and 10 μM GA₃ were added (Figure 2a-c). These results indicated that overexpression of ZmWRKY114 could reduce the GA sensitivity of transgenic plants.

Figure 2. Effects of GA₃ on plant height development of ZmWRKY114 transgenic plants. (a) The phenotype of seedlings treated with different concentrations of GA₃ and 200 μM PAC. (b and c) Comparison of plant height and root length between WT and two transgenic lines. Data are means ± SD from 30 seedlings. (d) WT and two transgenic lines seedlings aged 4 weeks before GA₃ treatment. (e) WT and two transgenic line seedlings were sprayed with 20 mL water for 30 days. (f) WT seedings were sprayed with 20 mL water while two transgenic line seedlings were sprayed with 20 ml 10 μM GA₃ for 30 days. Bar, 3.5 cm. (g) Measurement of plant height before and after GA₃ treatment. Data are mean ± SD from 15 plants. DAT, day(s) after treatment. (h) GA contents of the 4-week-old seedlings of WT and transgenic plants. Data are means ± SD from two biological samples. Asterisks mean a signifiant difference between WT plants and different transgenic lines by Student’ t-test, * P < .05, **P < .01

Exogenous GA₃ treatment rescued the dwarfism phenotype of ZmWRKY114 transgenic plants

To test whether the semi-dwarf phenotype of ZmWRKY114 transgenic plants was the result of GA deficiency, the 4-week-old transgenic plants were subjected to GA₃ treatments by spraying with 10 μM GA₃ for 30 days and measuring their heights. As shown in Figure 2d, significant difference in plant height was observed between WT and transgenic plants before GA₃ treatment. After 30 days, the average plant height of the wild plants was 1.5 times higher than that of the untreated ZmWRKY114 transgenic strain (Figure 2e and g). However, after treatment with 10 μM GA₃, the ZmWRKY114 overexpression plant height was fully restored and was the same as the average plant height of the wild type (Figure 2f and g). This result indicated that GA synthesis is possibly repressed in ZmWRKY114-OE plants.

Subsequently, GA₁ and GA₃ contents of WT and transgenic lines were measured using HPLC-MS/MS. As shown in Figure 2h, both the GA₁ content and GA₃ content were lower in ZmWRKY114-overexpressing plants compared with WT. This result strongly implies that overexpression of ZmWRKY114 has a negative effect on GA biosynthetic pathway.

ZmWRKY114 overexpression altered the expression of plant height related genes in rice

To further examine whether the semi-dwarf phenotype was caused by ZmWRKY114 overexpression, RNA-seq analysis was conducted for WT and two transgenic lines. Each sample was biologically repeated twice. 5799 DEGs in OE8 and 4828 DEGs in OE6 compared with the WT were found (Table S2 and Table S3). Among these DEGs, 1722 up-regulated and 1661 down-regulated genes had significantly different transcript abundance (Figure 3a and b), many genes were reported to involve in plant height control through regulating the GA synthesis, such as OsEUI1, OsLOL2, OsYAB1, OsGI and OsGA2ox4 (Table 1).

Table 1. Plant height related genes altered in two transgenic lines compared to WT plants

	Gene Name	Gene ID	Fold changes of candidate DEGs compared to WT
			OE8	OE6
Up-regulated	OsEUI1	Os05g0482400	6.20	5.00
	OsGA2ox4	Os05g0514600	23.52	4.49
	OsYAB1	Os07g0160100	3.35	6.00
	OsJMT1	Os05g0102000	25.32	60.75
	OsGI	Os01g0182600	3.58	3.26
	OsGhd2	Os02g0731700	3.98	4.55
	OsLOL2	Os01g0612700	2.81	2.70
Down-regulated	OsEIN2	Os07g0155600	4.49	3.14
	OsNAL9	Os03g0411500	3.11	3.00
	OsJar3	Os05g0586200	2.27	2.04
	OsCPK4	Os02g0126400	3.36	2.78

Figure 3. Venn diagram and qRT-PCR analyses. (a) Overlapping of up-regulated genes. (b) Overlapping of down-regulated genes. The overlap of genes that were up- or down-regulated in the different comparisons. (c) qRT-PCR analyses of expression of plant height related genes including OsGA2ox4 (AK107211), OsYAB1 (NM_001065472), OsGhd2 (XM_015767173), OsEUI1 (AK109526) and OsJar3 (NM_001062995) in WT and two transgenic lines. OsActin gene was used as an internal control. Data indicate mean ± SD of three replicates. Asterisks mean a signifiant difference by Student’ t-test, * P < .05, **P < .01

The expression levels of plant height related genes, such as OsEUI1, OsGA2ox4, OsYAB1, OsGhd2 and OsJar3 genes were compared between WT and ZmWRKY114-overexpressing plants using qRT-PCR method. The results showed that ZmWRKY114 overexpression indeed alters the expression of OsGA2ox4, OsYAB1, OsGhd2, OsEUI1 and OsJar3 (Figure 3c). The expression levels of OsEUI1, OsGA2ox4 and OsYAB1 were up-regulated but OsGhd2 and OsJar3 expression was down-regulated.

ZmWRKY114 can bind to the W-box motif in OsGA2ox4 promoter

RNA-seq and qRT-PCR data showed that the OsGA2ox4 expression was altered significantly in the ZmWRKY114-overexpressing plants, suggesting that OsGA2ox4 may be a direct target of ZmWRKY114 transcription factor. To further test whether ZmWRKY114 can bind to the W-box motif in the promoter region of OsGA2ox4, EMSAs were used with ZmWRKY114-MBP protein. We identified W-box motif by examining the promoter region of OsGA2ox4. As shown in Figure 4a, the potential ZmWRKY114-binding W-box site, TTGACT at −1136 to − 1142 bp upstream of the ATG position of OsGA2ox4, was discovered. In Figure 4b, shifted band was detected when probes harboring TTGACT in the OsGA2ox4 promoter region were incubated with ZmWRKY114 protein (lane 2). Conversely, there was no shift band was observed when the solution harboring the probe alone (lane 1). In addition, the ability of ZmWRKY114 binding to the OsGA2ox4 promoter was alleviated along with increased concentrations of competitive probes (lane 3; 4). However, the mPr with mutated W-box elements did not change significantly (lane 5; 6).

Figure 4. EMSA analysis showing binding of recombinant ZmWRKY114 protein to the promoter of OsGA2ox4. (a) The W-box was predicted in the promoter region of OsGA2ox4. (b) The black arrow indicates the binding of ZmWRKY114 and the biotin-labeled W-box probe. Pr: TTGACTTTGACTTTGACT; mPr: TTTACTTTTACTTTTACT. The + and − signs represent the presence and absence of corresponding components, respectively. (c) Schematic diagrams of the effector and reporter constructs for dual luciferase assay. REN, Renilla luciferase; LUC, firefly luciferase. (d) Luciferase assay of the OsGA2ox4 promoter. proOsGA2ox4::LUC reporter constructs were transiently expressed in leaves of N.benthamiana together with either the empty vector or 35S::ZmWRKY114 effector, and the LUC signal was imaged at 72 h post-transfection. (e) The LUC/REN ratio as determined using a dual luciferase assay. Data are means (±SD) of at least three biological replicates. Statistical significance was determined using Student’s t-test: **P < .01

We then performed dual luciferase assays to determine whether ZmWRKY114 was an activator or a repressor of OsGA2ox4 transcription. A 256 bp OsGA2ox4 promoter fragment containing the TTGACT sequence was introduced into a vector to generate a reporter construct, which was then co-transformed with either the effector or the control (Figure 4c). Subsequent analysis showed that the co-expression of 35S::ZmWRKY114 with proOsGA2ox4::LUC led to a 1.5-fold rise in firefly luciferase (LUC) activity compared with control (Figure 4d and e), suggesting that ZmWRKY114 directly binds to the OsGA2ox4 promoter and acts as a transcriptional activator.

Discussion

Since the important roles of WRKY TFs in plant development and stress response were demonstrated, members of this gene family have been identified from various higher plants.^17–22 At present, Wei et al.²³ identified 119 WRKY genes from the complete genome of maize B73 inbred line, and only a few members have been functionally characterized. Most of the identified maize WRKY members are involved in plant abiotic stress response, including ZmWRKY17, ZmWRKY58 and ZmWRKY114.^24,25 Another maize WRKY protein ZmWRKY79 has been found to play an important role in plant biotic stress response via participation in phytohormone metabolism and ROS scavenging.²⁶

In this study, the reduced plant height of ZmWRKY114 transgenic lines implied this gene could involve in the regulation of plant growth, largely because of its highest expression in stem.¹³ In particular, the phytohormone GA is an important factor in determining plant height through affecting the length and number of internode cells.²⁷ Light microscope of internode tissues indicated that the semi-dwarf phenotype of transgenic plants was caused by reduced cell numbers (Figure 1e-h). ZmWRKY114-overexpressing plants were more insensitive to GA₃ and exogenous GA₃ treatment could rescue the dwarfism phenotype of these transgenic plants, suggesting that ZmWRKY114 overexpression affected the homeostasis of GA in two transgenic lines, which was confirmed by the reduced bioactive GA₁ and GA₃ content in transgenic plants compared with that of WT.

To further uncover the molecular mechanisms underlying the decreased plant height of ZmWRKY114 transgenic lines, RNA-seq and qRT-PCR experiments were performed. The data showed that ZmWRKY114 overexpression could alter the expression of some GA related genes including OsEUI1, OsGA2ox4, OsYAB1 and OsLOL2. OsEUI1 was reported to encode a cytochrome P450 monooxygenase that deactivates bioactive GAs and transgenic plants overexpressing this gene exhibited a severe dwarf phenotype.²⁸ Gibberellin 2-oxidases (GA2oxs) control plant growth by inactivating endogenous bioactive GAs through 2β-hydroxylation.³ Functional studies have testified that increased expression of these GA2oxs leads to reduced levels of bioactive GAs and corresponding dwarf phenotypes.^3,7 GA2ox4 has been shown to act exclusively as a C19-GA 2-oxidase.²⁹ Overexpression of OsYAB1 also resulted in a semi-dwarf phenotype by impairing GA-mediated repression of GA3ox2.³⁰ In addition, as a positive regulator in the process of GA biosynthesis, OsLOL2 promotes the accumulation of bioactive GA through activating the expression of OsKS1, a gene encoding for GA biosynthetic enzyme.³¹ In the light of the significant change of OsGA2ox4 expression in the transgenic plants reflected by RNA-seq and qRT-PCR assays, we scan the promoter region of OsGA2ox4 and a W-box cis-element was found in its promoter. Data from EMSA assay and dual-luciferase reporter assay showed that OsGA2ox4 is a direct target of ZmWRKY114 (Figure 4). The result further proved that semi-dwarf phenotype in ZmWRKY114-overexpressing plants result from GA deficiency. All these results suggest that overexpression of the maize WRKY114 gene in transgenic rice regulates the expression level of OsGA2ox4 by binding to the W-box element in the promoter, and that increased expression of OsGA2ox4 leads to reduced bioactive GAs levels and a corresponding dwarf phenotype.

In addition, A phylogenetic analysis was carried out to evaluate the evolutionary relationships between ZmWRKY114 and other WRKY proteins. As shown in Fig. S2, ZmWRKY114 displayed a close relationship with OsWRKY51 and OsWRKY42 in rice. It has been reported that OsWRKY51 and OsWRKY71 physically interact in the nuclei to synergistically suppress the induction of Amy32b by GA.³² Ectopic expression of OsWRKY42 results in reduced expression of cell wall damage and salt stress induced jasmonic acid biosynthesis and response genes.³³ It is worth mentioning that we found that the expression level of Ghd2, a novel CO-like gene which has been reported to be a regulator in the control of grain number, heading date, and plant height, was decreased approximately 4-fold in transgenic plants compared to WT.³⁴ Yeast two-hybrid and pull-down assays confirmed that OsGhd2 could interact with ZmWRKY114 (Fig. S3). Previous study showed that OsGhd2 also specifically interacts with three 14-3-3 proteins,³⁴ which have been reported to function as the WRKY-interacting partners.³⁵ The results implied that ZmWRKY114 involved in the control of plant height through a complex regulatory network. In addition, the decreased expression of OsGhd2 might affect the GA biosynthesis through altering the expression level of GA2-oxidase genes (OsGA2ox4 in this study), just like its homolog OsGhd7³⁶ to influence the plant height.

In conclusion, this study indicates that ZmWRKY114 functions as a negative regulator in GA biosynthesis during growth and development in rice.

Author contributions

RC and QM conceived and designed the research; XF, CB, MW, HY, WL and HC performed the experiments; XF and RC wrote the manuscript. All authors had read and approved the final manuscript.

Acknowledgments

This work was supported by the China Postdoctoral Science Foundation (2018M642506) and the National Natural Science Foundation of China (31801365 and 31971895), the Natural Science Foundation of Anhui Province (1808085QC88).

Disclosure statement

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Supplementary material

Supplemental data for this article can be accessed on the publisher’s website

Additional information

Funding

This work was supported by the Program of China Postdoctoral Science Foundation 2018M642506; National Natural Science Foundation of China [31801365]; National Natural Science Foundation of China [31971895]; Anhui Provincial Natural Science Foundation [1808085QC88].