WRKY42 Modulates Phosphate Homeostasis through Regulating Phosphate Translocation and Acquisition in Arabidopsis

The Arabidopsis WRKY transcription factor family has more than 70 members, and some of them have been reported to play important roles in plant response to biotic and abiotic stresses. This study demonstrates that WRKY42 regulated phosphate homeostasis in Arabidopsis . The WRKY42- overexpressing lines, similar to the pho1 mutant, were more sensitive to low-inorganic phosphate (Pi) stress and had lower shoot Pi content compared with wild-type plants. The PHO1 expression was repressed in WRKY42 -overexpressing lines and enhanced in the wrky42 wrky6 double mutant. The WRKY42 protein bound to the PHO1 promoter under Pi-sufficient condition, and this binding was abrogated during Pi starvation. These data indicate that WRKY42 modulated Pi translocation by regulating PHO1 expression. Furthermore, overexpression of WRKY42 increased root Pi content and Pi uptake, while the wrky42 mutant had lower root Pi content and Pi uptake rate compared with wild-type plants. Under Pi-sufficient condition, WRKY42 positively regulated PHT1;1 expression by binding to the PHT1;1 promoter, and this binding was abolished by low-Pi stress. During Pi starvation, the WRKY42 protein was degraded via the 26S proteasome pathway. Our results demonstrated that At WRKY42 modulated Pi homeostasis by regulating the expression of PHO1 and PHT1;1 to adapt to environmental changes in Pi availability.

Another important pathway controlling Pi homeostasis involves PHO1, which plays an important role in Pi translocation from roots to shoots (Poirier et al., 1991;Hamburger et al., 2002;Wang et al., 2004). The pho1 mutant is deficient in loading Pi acquired by roots to the xylem vessel and only accumulates 24-44% as much total phosphate as wild-type plants in shoots (Poirier et al., 1991). PHO1 is located primarily in the root stelar cells and has a role in Pi efflux out of root stelar cells for xylem loading (Hamburger et al., 2002). There are 11 members of the PHO1 gene family in the Arabidopsis genome, and only PHO1 and PHO1;H1 can complement the pho1 mutant (Wang et al., 2004), indicating that PHO1 and PHO1;H1 are involved in long-distance Pi transport from roots to shoots. The increased transcript level of PHO1;H1 during Pi starvation is mainly controlled by the PHR1 transcription factor (Stefanovic et al., 2007), while expression of PHO1 is independent of PHR1 regulation (Stefanovic et al., 2007).
The PHO1 expression is directly down-regulated by the WRKY6 transcription factor under Pi-sufficient condition (Chen et al., 2009). WRKY42, a homolog of WRKY6, could bind to the PHO1 promoter in vivo and repressed the PHO1 promoter activity in Nicotiana benthamiana (Chen et al., 2009), indicating that WRKY42 also negatively regulates PHO1 expression. PHO2 can modulate PHO1 degradation (Liu et al., 2012).
In this paper, we report that Arabidopsis WRKY42 plays important roles in modulating Pi homeostasis in Arabidopsis. WRKY42 modulates Pi uptake and translocation by directly regulating PHT1;1 and PHO1 expression under Pi-sufficient conditions. During Pi starvation, WRKY42 expression is repressed and the WRKY42 protein is degraded via a proteasome pathway, and then the binding of WRKY42 to the The pho1 mutant has a defect in Pi transfer from roots to shoots, which results in reduced Pi content in shoots (Poirier et al., 1991;Hamburger et al., 2002). Therefore, a role for WRKY42 in translocating Pi was hypothesized. To test this, the shoot Pi was measured in the 10-d-old WRKY42-overexpressing lines, wrky42 mutant, pho1 mutant and wild-type seedlings grown under Pi-sufficient condition. The WRKY42-overexpressing lines had similarly reduced shoot Pi contents to the pho1 mutant, and the reduced level of shoot Pi content was closely related to WRKY42 expression ( Fig. 4), indicating that WRKY42 negatively modulated Pi translocation in Arabidopsis.

WRKY42 Directly Down-Regulates PHO1 Expression
Because WRKY42-overexpressing lines and the pho1 mutant had similar low-Pi sensitive phenotypes and lower shoot Pi contents (Figs 3 and 4), it was hypothesized that WRKY42 negatively regulated PHO1 expression. The transcription level of PHO1 gene was evaluated in the roots of WRKY42-overexpressing lines, the wrky42 mutant and wild-type plants, since PHO1 is mainly expressed in roots (Hamburger et al., 2002). The transcription of PHO1 was repressed in the WRKY42-overexpressing lines (Fig. 5A), and the repression level of PHO1 expression was consistent with WRKY42 expression in the WRKY42-overexpressing lines, with the strongest repression in Super:WRKY42-5 and the weakest in Super:WRKY42-3.
Because WRKY42 is a typical WRKY transcription factor that can bind to W-box motif (Fig. 2C), and sequence analysis showed that there are several W-boxes within between pho1 mutant and wild-type plants (Fig. 7A), indicating that the induced root Pi content in WRKY42-overexpressing lines was not due to the repression of PHO1 caused by WRKY42 overexpression.
The Pi uptake rate was measured to determine the effect of WRKY42 on Pi acquisition. The 10-d-old seedlings were transferred into a Pi uptake solution containing 500 μM Pi supplemented with 32 P orthophosphate, and Pi uptake was measured over a 4-h period. Consistent with the root Pi content, the WRKY42-overexpressing lines had a significantly (P < 0.05) higher Pi uptake rate compared with wild-type seedlings, and that of the wrky42 mutant was lower than that of wild-type (Fig. 7B). Arsenate is an oxyanion structurally analogous to phosphate (Asher and Reay, 1979) and is taken up mainly via Pi transporter PHT1;1 (Catarecha et al., 2007). When grown on medium containing arsenate, the pht1;1 mutant showed an arsenate-tolerant phenotype, and the PHT1;1-overexpressing line was more sensitive to arsenate than wild-type plants (Supplemental Fig. S1;Catarecha et al., 2007;Wang et al., 2014). To gain further insight into the role of WRKY42 in Pi acquisition, the phenotype of WRKY42-overexpressing lines, the wrky42 mutant and wild-type seedlings were tested with arsenate. When grown on Pi-sufficient medium without arsenate [-As(V)], there were no obvious phenotypic differences among the WRKY42-overexpressing lines, the wrky42 mutant and wild-type seedlings ( Fig. 7C and Supplemental Fig. S1). When grown on Pi-sufficient medium with 200 μM arsenate [+As(V)], although the toxic effect of arsenate was evident in the growth of WRKY42-overexpressing lines, the wrky42 mutant and wild-type plants, their degree of sensitivity varied. The WRKY42-overexpressing lines had a much more arsenate-sensitive phenotype, similar to the phenotype of PHT1;1-overexpressing lines, compared with wild-type seedlings ( Fig. 7C and Supplemental Fig. S1). There were no obvious differences between the wrky42 mutant and wild-type seedlings when grown on Pi-sufficient medium with 200 μM arsenate ( Fig. 7C and Supplemental Fig. S1).
Together, these data indicate that overexpression of WRKY42 enhanced Arabidopsis Pi accumulation.
Therefore we examined expression of PHT1;1 in roots of WRKY42-overexpressing lines, the wrky42 mutant and wild-type plants under Pi-sufficient conditions. Transcription of PHT1;1 was obviously elevated in the WRKY42-overexpressing lines (Super:WRKY42-40 and Super: , and repressed in the wrky42 mutant compared with wild-type plants (Fig. 8A). The PHT1;1 expression was also tested in the pho1 mutant. The expression level of PHT1;1 in the pho1 mutant was similar to that in wild-type plants (Fig. 8B) (Fig. 8D). These data indicated that WRKY42 positively regulated PHT1;1 expression.
Promoter sequence analysis showed that there were several W-boxes within the PHT1;1 promoter ( Fig. 9A; Martín et al., 2000;Wang et al., 2014), thus we hypothesized that WRKY42 directly regulates PHT1;1 expression by binding to the W-box within the PHT1;1 promoter. The in vivo interaction between WRKY42 and the W-box motifs within PHT1;1 promoter was investigated using ChIP-qPCR analysis. The 7-d-old wild-type seedlings were transferred to Pi-sufficient (+P) or Pi-deficient (-P) medium for another 7 d, and then the roots were harvested for ChIP-qPCR assay. When wild-type plants were grown in Pi-sufficient condition, the chromatin immunoprecipitated with the anti-WRKY42 antibody was enriched in the P2 fragment of the PHT1;1 promoter, while no interaction was observed between WRKY42 and the PHT1;1 promoter containing P1, P3 or P4 fragments (Fig. 9B). During Pi starvation, the interaction between WRKY42 and the P2 fragment within the PHT1;1 promoter was abolished (Fig. 9B). Furthermore, the EMSA assay was also performed to detect whether WRKY42 could bind to the P2 fragment of the PHT1;1 promoter in vitro. The WRKY42-His fusion protein could bind to P2 within the PHT1;1 promoter, and the binding was effectively reduced by adding increasing amounts of unlabeled competitors with the same P2 sequence (Fig. 9C). In contrast, the WRKY42-His fusion protein could not bind to the mutation probe (mP2) which has two mutated W-boxes (Fig. 9C). As the negative control, the His protein alone did not bind to the PHT1;1 promoter (Fig. 9C). These data demonstrate that WRKY42 positively regulated PHT1;1 expression.

WRKY42 is Degraded during Phosphate Starvation
Because the interaction between WRKY42 and the promoters of PHO1 or PHT1;1 was abolished during Pi starvation ( Fig. 5C and 9B protein was degraded under Pi-deficient stress. In order to determine the relationship between the WRKY42 degradation and Pi status, the cell-free degradation analysis was conducted. The recombinant WRKY42-His protein was purified from E. coli, and incubated with the total protein extracts from the 7-d-old wild-type seedlings cultured under Pi-sufficient (MS medium with 1.25 mM Pi, +P) or Pi-deficient (low-phosphate medium with 10 μM Pi, -P) conditions for another 5 d. When incubated with +P total protein extract, the WRKY42 protein showed very faint degradation (Fig. 10A). When the WRKY42 protein was incubated with -P total protein extract, the WRKY42 protein was obviously degraded. This degradation of WRKY42 was inhibited by the MG132, a 26S proteasome inhibitor (Fig. 10A), indicating that Pi starvation induced the proteasome-dependent degradation of WRKY42.
To further confirm the degradation of WRKY42 during Pi starvation in vivo, the Super:WRKY42-GFP and Super:GFP transgenic lines were generated. The 7-d-old Super:WRKY42-GFP and Super:GFP seedlings were transferred to Pi-sufficient (MS) or Pi-deficient (LP) medium, and then harvested at the indicated time for protein gel blot analysis using anti-GFP. The WRKY42 protein decreased much more rapidly in Super:WRKY42-GFP exposed to Pi starvation compared with Pi-sufficient condition ( Fig. 10B). To further confirm that reduction of WRKY42 protein level was due to the proteasome-dependent degradation in vivo, the 7-d-old Super:WRKY42-GFP seedlings were also transferred to LP medium with 10 μM MG132. The addition of MG132 clearly inhibited WRKY42 degradation under Pi starvation condition (Fig. 10B). Super:GFP was used as a control, and no GFP degradation was detected in Pi-deficient or -sufficient condition (Fig. 10B). Taken together, these data demonstrated that the WRKY42 protein was degraded via the proteasome pathway during Pi starvation and was stabilized by abundant Pi. Phosphate plays important roles in regulation of many biochemical and physiological processes and is an essential building block of cell components. The intracellular concentration of Pi in plants is tightly regulated to maintain Pi homeostasis. To achieve this, plants have evolved a series of strategies, such as enhancing Pi acquisition and remobilizing internal Pi (Raghothama, 1999;Vance et al., 2003). Arabidopsis PHO1 encodes a membrane protein and is involved in Pi loading from roots to shoots (Hamburger et al., 2002). The pho1 mutant has lower shoot Pi (Poirier et al., 1991) and

WRKY42 is a Key Regulator in Phosphate Homeostasis in Plants
shows a low-Pi sensitive phenotype due to defective Pi loading in the xylem (Poirier et al., 1991;Chen et al., 2009). In the present study, the WRKY42-overexpressing lines showed a reduced shoot Pi and low-Pi sensitive phenotype, similar to the pho1 mutant (Figs 3 and 4), suggesting that WRKY42 played a role in regulating Pi translocation. As a typical WRKY transcription factor, WRKY42 directly bound to the W-boxes within the PHO1 promoter and repressed PHO1 expression under Pi-sufficient condition (Fig. 5).
These data demonstrate that the WRKY42 transcription factor negatively regulated Pi translocation.
Interestingly, our data also showed that WRKY42 positively regulated Pi acquisition.
Overexpression of WRKY42 enhanced Pi uptake and root Pi content, and During Pi starvation, transcription of WRKY42 was repressed (Fig. 1C), and the WRKY42 protein was degraded in a proteasome-dependent manner ( Fig. 10 repressed PHO1 expression, and WRKY42 bound to the Y and Z sites within the PHO1 promoter (Fig. 5C), demonstrating that WRKY42 directly down-regulated PHO1 expression. The PHO1 expression was enhanced in the wrky42 or wrky6 single mutants compared with wild-type plants (Fig. 6B). And the expression level of PHO1 in the wrky42 wrky6 double mutant was much higher than those in wild-type or single mutant  In the present study, PHT1;1 expression was elevated in the WRKY42-overexpressing lines and repressed in the wrky42 mutant compared with wild-type plants (Fig. 8), and WRKY42 could bind to the PHT1;1 promoter ( Fig. 9) was degraded and then regulation of PHO1 and PHT1;1 by WRKY42 ceased.

Plant Materials and Growth Conditions
The wild-type plants were the Col-0 ecotype. The Super:PHT1;1, pho1 and pht1;1 plants used in the study were described previously (Chen et al., 2009;Wang et al., 2014).
The Arabidopsis seeds were surface sterilized and cold treated at 4°C for 3 d. Then, the seeds were plated on MS medium containing 1.25 mM Pi, 3% (w/v) Suc, 0.8% (w/v) agar and grown at 22°C with illumination of 100 μmol m -2 s -1 for a 16-h daily light period, unless otherwise indicated.
For Pi starvation treatment, 7-d-old seedlings were transferred to MS or LP medium.
The LP medium was made by modifying MS medium to contain 10 μM Pi, and the agar was replaced by agarose (Promega).

Phosphate Content and Phosphate Uptake Assay
The Arabidopsis plants were germinated and grown on MS medium for 10 d, and then the shoots and roots were harvested for Pi content measurement. The Pi content in the samples was quantified as described previously (Ames, 1966;Chiou et al., 2006). For the Pi uptake assay, 10-d-old seedlings grown on MS medium were transferred to the Pi uptake solution containing 500 μM Pi supplemented with 0.2 μCi 32 P orthophosphate. A group of 15 seedlings was used as one biological sample. The 7-d-old seedlings were transferred to MS or LP medium for another 10 d, and then the seedlings were harvested for anthocyanin measurement. Anthocyanin was determined as described by Lu et al. (2014).

Plasmid Construction and Plant Transformation
To construct Super:WRKY42, the coding sequence of WRKY42 was cloned into the The primers used are listed in Supplemental Table 1 online.

Subcellular Localization
For the subcellular localization assay, WRKY42 fused to GFP was cloned into a modified

Transient Expression Assays in Nicotiana benthamiana
The transient GUS expression assays were performed as described (Chen et al. 2009).
The constructs ProPHT1;1:GUS, Super:WRKY42 and pCAMBIA1300-ProSuper were transformed into Agrobacterium strain GV3101 separately. For every infiltration sample, Super:LUC was added as an internal control. Agrobacterium cells were harvested by centrifugation and suspended in induction buffer to OD 600 of 0.4. After 2 hr at 22°C, Agrobacterium cells were infiltrated into leaves of 7-week-old Nicotiana benthamiana leaves, and the infiltration ratio of Super:WRKY42 and ProPHT1;1 or pCAMBIA1300-ProSuper and ProPHT1;1 was 9:1(v/v). After infiltration for 36 hr, leaf discs were harvested for GUS and LUC proteins extraction. The GUS and LUC activities of the infiltrated leaves were quantitatively determined, and the GUS/LUC ratio was used to quantify the promoter activity.

ChIP-qPCR Assay
To generate the anti-WRKY42 antibody, the whole coding sequence of WRKY42 was cloned into the pET30a vector. The recombinant WRKY42-His protein was expressed in E. coli and purified. The polyclonal anti-WRKY42 antibody was generated by inoculating a mouse with the recombinant WRKY42. For ChIP-qPCR assay, 7-d-old seedlings were transferred to MS or LP medium for another 7 d, and then the roots were harvested for ChIP assay. The ChIP-qPCR assay was conducted as previously described (Chen et al., 2009;Feng et al., 2014), and the primers used are listed in Supplemental Data are mean values of three replicates ± SE from one experiment.

EMSA Assay
The EMSA assay was conducted using a LightShift Chemiluminescent EMSA Kit (Pierce) following the manufacturer's protocol. The recombinant WRKY42-His protein and His protein were purified from E. coli. The fragments of the PHT1;1 promoters were obtained by PCR using biotin-labeled or unlabeled primers (Supplemental Table 1 online). Biotin-unlabeled fragments of the same sequences were used as competitors, and the His protein alone was used as the negative control.

Protein Extraction and Cell-Free Degradation
Seven-day-old Arabidopsis seedlings were transferred to MS medium (+P) or LP (-P) medium for 5 d, and then the seedlings were harvested and ground into fine powder in liquid nitrogen. Total proteins were extracted in degradation buffer containing 25 mM Tris-HCl, pH 7.5, 10 mM NaCl, 10 mM MgCl 2 , 4 mM PMSF, 5 mM DTT and 10 mM ATP as described by Wang et al. (2009). The total protein concentration was determined by Bio-Rad protein assay. The total protein extracts prepared were adjusted to equal concentrations in the degradation buffer for each assay. Then, exogenous MG132 was added to the total proteins extracted from -P plants, the final concentration was 10 μM.
The 250 ng of recombinant WRKY42-His protein was incubated in 20-μL extracts (containing 50 μg of total proteins) for the individual assays. The extracts were incubated at 22 °C, and samples were taken at indicated times for determination of WRKY42 protein abundance by immunoblots with anti-His.

Immunoblot Analysis
Total proteins were extracted according to Saleh et al. (2008), and 80 μg of proteins of each sample were separated on a 10% SDS-PAGE and transferred to polyvinylidene fluoride membranes. MG132 treatment was conducted as described by Chen et al.   and (c).

(C) qRT-PCR analysis of WRKY42 expression in Arabidopsis under Pi starvation.
Seven-day-old wild-type seedlings were transferred to Pi-sufficient condition (MS medium, +P) or Pi-deficient condition (LP medium with 10 μM Pi, -P) for 3 d, and then the roots were harvested for RNA extraction. Transcript level of WRKY42 was quantified relative to ACTIN2/8. The data represent the mean values of three replicates ± SE. Biotin-labeled Pchn0 probe incubated with His protein served as the negative control.   (C) ChIP-qPCR assay to detect the association between WRKY42 and the PHO1 promoter. Seven-day-old seedlings were transferred to Pi-sufficient (MS) or Pi-deficient (LP) condition for another 7 d, and then the roots were harvested for ChIP-qPCR.
Chromatins were immunoprecipitated with anti-WRKY42 antibody, and amount of indicated DNA in immune complex was tested by qRT-PCR. The ratio of IP DNA over the input was presented as the percentage of input (% IP). The experiments were repeated three times, and three replicates were included for each sample in each experiment. The data are presented as means ± SE (n = 3).    (A) Cell-free degradation assay. Seven-day-old wild-type seedlings were transferred to Pi-sufficient medium (+P) or Pi-deficient medium (-P) for another 5 d, and then the seedlings were harvested for protein extraction. The plant protein extracts were incubated with recombinant WRKY42-His for the indicated time, and then WRKY42 abundance was determined by immunoblotting with anti-His.

(B) Immunoblot analysis of WRKY42 protein. Seven-day-old Super:WRKY42-GFP and
Super:GFP transgenic seedlings were transferred to MS medium, LP medium or LP medium with 10 μM MG132 (LP+MG132), and the seedlings were harvested at the indicated time for protein extraction. Protein extracts were analyzed by immunoblots using anti-GFP. Actin was used as the loading control.   . The adenine residue of the translational start codon ATG was assigned position +1, and the numbers flanking the sequences of the PHO1 promoter fragments were counted based on this number. The W-boxes are marked by gray rectangles, and relative positions and sizes of the different PCR amplified fragments are indicated by black lines under the W-box. (C) ChIP-qPCR assay to detect the association between WRKY42 and the PHO1 promoter. Sevenday-old seedlings were transferred to Pi-sufficient (MS) or Pi-deficient (LP) conditions for another 7 d, and then the roots were harvested for ChIP-qPCR. Chromatins were immunoprecipitated with anti-WRKY42 antibody, and amount of indicated DNA in immune complex was tested by qRT-PCR. The ratio of IP DNA over the input was presented as the percentage of input (% IP). The experiments were repeated three times, and three replicates were included for each sample in each experiment. The data are presented as means ± SE (n = 3).     qRT-PCR analysis of PHT1;1 expression in the roots of the WRKY42-overexpressing lines, wrky42 mutant and wild-type plants. The plants were germinated and grown on MS medium for 10 d, and then the roots were harvested for RNA extraction. Each data bar represents the means  SE (n = 3). Asterisks indicate significant differences compared with wild-type plants (paired test): *, P < 0.05; **, P < 0.01. Wild-type plants (WT) were used as the control (#). (B) qRT-PCR analysis of PHT1;1 expression in the roots of the pho1 mutant and wildtype plants. Transcript level of PHT1;1 was quantified relative to ACTIN2/8. Each data bar represents the means  SE (n = 3). (C) Transient overexpression of WRKY42 fused to ProPHT1;1:GUS in Nicotiana benthamiana leaves. Each data bar represents the means  SE (n = 5). Asterisks indicate significant differences: *, P < 0.05; # is the control.  . The adenine residue of the translational start codon ATG was assigned position +1, and the numbers flanking the sequences of the PHT1;1 promoter fragments were counted based on this number. (B) ChIP-qPCR assay to detect the association between WRKY42 and the PHT1;1 promoter. Sevenday-old seedlings were transferred to Pi-sufficient (MS) or Pi-deficient (LP) conditions for another 7 d, and then the roots were harvested for ChIP-qPCR assay with anti-WRKY42. The ratio of IP DNA over the input was presented as the percentage of input (% IP). The data are presented as means ± SE (n = 3). (C) EMSA assay to analyze the binding of WRKY42 to P2 fragment of PHT1;1 promoter. Each biotinlabeled DNA probe was incubated with WRKY42His protein. An excess of unlabeled probe was added to compete with labeled promoter sequence. Biotin-labeled probe incubated with His protein served as the negative control.   (A) Cell-free degradation assay. Seven-day-old wild-type seedlings were transferred to MS medium (P) or LP medium (P) for another 5 d, and then the seedlings were harvested for protein extraction. The plant protein extracts were incubated with recombinant WRKY42-His for the indicated time, and then WRKY42 abundance was determined by immunoblotting with anti-His. (B) Immunoblot analysis of WRKY42 protein. Seven-day-old Super:WRKY42-GFP and Super:GFP transgenic seedlings were transferred to MS medium, LP medium or LP medium with 10 M MG132 (LPMG132), and the seedlings were harvested at the indicated time for protein extraction. Protein extracts were analyzed by immunoblots using anti-GFP. Actin was used as the loading control.
www.plantphysiol.org on August 27, 2017 -Published by Downloaded from Copyright © 2015 American Society of Plant Biologists. All rights reserved. Under high-Pi condition, the WRKY42 directly represses PHO1 expression and activates PHT1;1 expression by binding to the W-box motifs within the promoters of PHO1 and PHT1;1, in order to maintain phosphate homeostasis. Under low Pi stress, the WRKY42 protein is degraded, and then the regulation of PHO1 and PHT1;1 by WRKY42 ceased.