Overexpression of ThVHAc1 and its potential upstream regulator, ThWRKY7, improved plant tolerance of Cadmium stress

As one of the most toxic heavy metals in the environment, cadmium (Cd) poses a severe threat to plant growth. We previously reported that overexpression of the Tamarix hispida V-ATPase c subunit (ThVHAc1) improved the Cd tolerance of Saccharomyces cerevisiae. In the current study, we further explored the Cd tolerance conferred by ThVHAc1 in Arabidopsis and T. hispida. ThVHAc1 transgenic Arabidopsis had higher seed germination, biomass, and chlorophyll content under CdCl2 treatment. In Cd-stressed plants, overexpression of ThVHAc1 significantly improved V-ATPase activity and affected the expression of other V-ATPase subunit-encoding genes. Intriguingly, the lower level of ROS accumulation in ThVHAc1-overexpressing lines under CdCl2 treatment demonstrated that ThVHAc1 may modulate Cd stress tolerance by regulating ROS homeostasis. Transient expression of ThVHAc1 in T. hispida further confirmed these findings. Furthermore, promoter analysis and yeast one-hybrid assay revealed that the transcription factor ThWRKY7 can specifically bind to the WRKY cis-element in the ThVHAc1 promoter. ThWRKY7 exhibited similar expression patterns as ThVHAc1 under CdCl2 treatment and improved Cd tolerance, suggesting that ThWRKY7 may be an upstream regulatory gene of ThVHAc1. Therefore, our results show that the combination of ThVHAc1 and its upstream regulator could be used to improve Cd stress tolerance in woody plants.

Scientific RepoRts | 6:18752 | DOI: 10.1038/srep18752 membrane-peripheral domain) and V 0 (260 kDa membrane-integral domain). The V 1 domain contains eight different subunits (A-H) and is responsible for ATP hydrolysis, while the V 0 domain includes six different subunits (a, d, c, c′ , c″ and e) and is responsible for proton translocation. There are few reports on the mechanism of V-ATPase regulation in response to various adverse conditions. However, some results have indicated that the expression levels of V-ATPase subunits are involved in various abiotic stresses. For example, overexpression of the wheat (RH8706-49) V-ATPase B subunit (TaVB) in Arabidopsis conferred higher V-ATPase activity and overall salt tolerance than were observed in the wild type (WT) 16 . Therefore, the cloning and characterization of the V-ATPase subunits may be an effective way to understand the regulation of this molecule and its response mechanism during abiotic stress.
The V-ATPase c subunit (VHAc) participates in the formation of a proton channel that is responsible for proton transport and is an essential factor in the assembly of V 1 -V 0 15 . Mutated yeast lacking VHAc fail to assemble V 1 into the membrane 17 . Some studies have demonstrated the salt regulation capacity of the VHAc gene. For example, a transcript analysis of Pennisetum glaucum PgVHA-c1 showed that the expression of PgVHA-c1 was increased in response to salinity stress 18 . Under salt stress conditions, overexpression of LbVHA-c1 (Limonium bicolor) in tobacco led to higher activity of superoxide dismutase (SOD) and peroxidase (POD) and lower levels of malondialdehyde (MDA) than in the WT 19 . Despite these findings, there are few reports on gene expression in response to heavy metals, especially related to the VHAc gene in a woody halophyte.
We found that V-ATPase activity and ThVHAc1 protein expression in T. hispida under CdCl 2 treatment were much higher than those under NaCl, PEG, or CuSO 4 treatments (data not shown), indicating that ThVHAc1 may play a key role in Cd stress tolerance in T. hispida. Consistently, we confirmed that the expression of ThVHAc1 was induced by CdCl 2 treatment in T. hispida roots, stems, and leaves and that the overexpression of ThVHAc1 in yeast improved Cd tolerance 20 . However, the role of ThVHAc1 in response to Cd stress and the mechanisms of ThVHAc1 regulation under Cd stress remain far from being fully elucidated. In this study, we identified a potential upstream regulator of ThVHAc1, ThWRKY7, which showed expression patterns similar to those of ThVHAc1 and which improved Cd stress tolerance. In addition, the regulation of plant Cd tolerance by the ThVHAc1 gene and the relationship between ThVHAc1 and V-ATPase activity under CdCl 2 treatment were further analyzed. Our results demonstrate that ThVHAc1 may participate in Cd tolerance through the reactive oxygen species (ROS) scavenging system to alleviate cell damage. The heterologous expression of ThVHAc1 effectively improved V-ATPase activity and affected the expression of other subunits and related genes. This study expands our knowledge of the response of T. hispida to Cd stress and the relationship of the c subunit with the entire V-ATPase enzyme. Further, our findings provide new insights into the role and regulatory mechanism of ThVHAc1 upon exposure to Cd stress, which will be beneficial for providing candidate genes to genetically improve tolerance of Cd stress in woody plants.

Materials and Methods
Plant materials and treatments. Two-month-old T. hispida seedlings were grown in a greenhouse on a 14 h light/10 h dark cycle, with 70-75% relative humidity and an average temperature of 24 °C. The seedlings were well watered at the roots with 150 μ M CdCl 2 for 6, 12, 24, 48, and 72 h, as indicated. Well-watered seedlings were used as the control. The roots, stems and leaves of every treated seedling (sample size of 20 seedlings) were harvested for quantitative real-time PCR (qRT-PCR) analyses. All treatments were applied at least three times (as biological replicates).
Cloning and expression analysis of the ThVHAc1 promoter. The ThVHAc1 promoter was PCR-amplified from T. hispida genomic DNA using a genome walking kit (Takara, Japan). The cis-elements in the ThVHAc1 promoter were analyzed using the PLACE database (http://www.dna.affrc.go.jp/PLACE) 21 . The 35S promoter in pCAMBIA1301 was replaced with the ThVHAc1 promoter to drive the expression of β-glucuronidase (GUS) (Fig. S1a), and this construct (named pThVHAc::GUS) was transferred into Arabidopsis 22 . The T 3 generation seedlings were used to study the ThVHAc1 temporal and spatial expression patterns through GUS staining. The 30 d transgenic Arabidopsis and T. hispida transiently expressing pThVHAc::GUS were independently treated with 100 μ M CdCl 2 or H 2 O (as a control) for 0 h, 12 h, or 24 h. Samples were then collected and labeled, and GUS activities were used to analyze the expression of the ThVHAc1 promoter. The GUS activity was measured according to Hunter and Watson (2008) 23 . When the protein concentrations were high, the samples were diluted with sterile water to maintain A 595 at less than 2.0. Thus, the determination of GUS activity is precise and avoids the errors generated by saturated staining.
RNA isolation and qRT-PCR. Total RNA was isolated from each sample using the CTAB method. qRT-PCR was carried out using an MJ Opticon TM2 machine (Bio-Rad, Hercules, CA, USA) with the reaction system and procedures from Gao et al.(2011) 20 , and α-Tubulin (FJ618518), β-actin (FJ618517), and β-tubulin (FJ618519) were used as internal controls. The primer sequences are listed in Table S1. Relative expression levels were calculated using the Δ Δ Ct method 24 . Identification of the upstream regulator of ThVHAc1. The WRKY motif ("GTGACA") was identified in the ThVHAc1 promoter (Fig. S1b, S2). A yeast one-hybrid assay was used to find the transcription factors that recognize the WRKY motif. Three tandem copies of the WRKY motif were cloned into a pHis2 vector (pHis2-WRKY) (Fig. S1c). WRKY TFs were identified from seven T. hispida libraries and cloned into pGADT7-Rec2 (Clontech, Palo Alto, CA, USA) to produce a cDNA library for use in the one-hybrid assay 25 .
To confirm the interactions between the motif and positive clones, the WRKY core "TGAC" was mutated to "GTCA" and cloned into pHis2 (pHis2-WRKY-M). Fragments of the ThVHAc1 promoter, including the WRKY motif (pHis2-WRKY-Seg), excluding the WRKY motif (pHis2-WRKY-Seg-M1), or including the mutated WRKY motif (pHis2-WRKY-Seg-M2) were separately cloned into pHis2 (Fig. S1c). The pHis2 plasmid containing three copies of the p53 DNA element (p53His2) was used as a control vector in the yeast one-hybrid assay. All primers are listed in Table S2.
To further confirm the above-mentioned interactions, WRKY, WRKY-M, WRKY-Seg, WRKY-Seg-M1, and WRKY-Seg-M2 were each fused with a CaMV35S-46 minimal promoter and cloned into pCAMBIA1301 to drive the GUS gene (acting as a reporter) (Fig. S1d). The ORF of ThWRKY7 (interaction TF) was cloned into the prokII vector under control of the 35S promoter (prokII-ThWRKY7) (Fig. S1d) to act as an effector. The prokII-ThWRKY7 construct was transferred into Arabidopsis using the floral dip method 23 . All reporters were transiently transformed into ThWRKY7 transgenic Arabidopsis using the agrobacterium-mediated transformation method, and all co-transformed Arabidopsis leaves were used to measure and stain GUS activity 26 . ThVHAc1 transgenic Arabidopsis. The ORF of ThVHAc1 was amplified and cloned into a prokII vector (35S::ThVHAc1). The primers are shown in Table S1. 35S::ThVHAc1 was transferred into Arabidopsis using the Agrobacterium-mediated transformation method 23 . An empty prokII plasmid was also transferred into Arabidopsis and used as a control (ck). Kanamycin-resistant lines were detected by PCR using vector-specific primers. The expression level of ThVHAc1 was confirmed by qRT-PCR, and two transgenic lines with intermediate expression levels (c1#10 and c1#17) were selected for further analysis.
Stress tolerance analysis. Seeds from the control and T 4 transgenic Arabidopsis were sown on 1/2 Murashige & Skoog (MS) agar medium with 100 μ M CdCl 2 . The germination and fresh weight were recorded after 8 d. Six-day-old seedlings sown on 1/2 MS were transferred to 1/2 MS agar plates with an additional 100 μ M CdCl 2 for another 12 d to compare the fresh weight and root length between lines. All experiments were performed three times.
Five-week-old WT and transgenic plants were used to determine the stress tolerance of the transgenic lines. SOD activity, POD activity, glutathione transferase (GST) activity, total chlorophyll content (Tcc), H 2 O 2 content, proline content, MDA content and electrolyte leakage (EL) were measured after treatment with 100 μ M CdCl 2 for 6 d 27,28 . The fresh weights of the aerial parts of seedlings placed on clean filter paper were measured to compare their water loss. The Cd content was determined using atomic florescence spectrometry 29 . Seedlings watered with 1 μ M CdCl 2 for 6 d were used as the control. Leaves sampled from the above lines and treated with 100 μ M CdCl 2 for 0 (control), 1, and 2 h were stained with nitroblue tetrazolium (NBT), 3,3′-diaminobenzidine (DAB), and Evans blue to analyze the in vivo accumulation of O 2− , H 2 O 2 , and cell death in leaves, respectively. ROS produced by intact guard cells and roots were stained with 3 μ M 2,7-dichlorofluorescein diacetate (H 2 DCF-DA, Fluka) 30 , and dead cells in the main roots were stained by propidium iodide 31 . ROS and cell death were visualized using a confocal laser-scanning microscope (CLSM) featuring an LSM410 microscope (Zeiss, Jena, Germany) with excitation at 488 nm and emission at 525 nm. Images were acquired using the ZEN 2009 "lite" edition 32 .
Five-week-old transgenic and control Arabidopsis treated with 100 μ M CdCl 2 for 0, 3, 6, 9, 12, and 24 h were collected for isolation of total RNA. The expression of 28 V-ATPase subunits, other V-ATPase-related genes, and stress-related genes was examined by RT-PCR or qRT-PCR. The gene names and primer sequences are listed in Table S3.
Transient expression of ThVHAc1 in T. hispida. 35S::ThVHAc1, RNAi::ThVHAc1 and empty Agrobacterium (used as control, labeled as T-ck) were transiently transformed into the aerial parts of five-week-old T. hispida seedlings 33 . The transformed seedlings were stained with DAB and Evans blue to visualize the ROS and cell death after treatment with 100 μ M CdCl 2 for 0 (control), 1, or 2 h. ThVHAc1 expression was analyzed using qRT-PCR. Meanwhile, the SOD, POD, GST, and glutathione peroxidase (GPX) activities, as well as MDA and EL, were assayed. Furthermore, the expression levels of ThSOD, ThPOD, ThGSTZ1, and ThGPX, as well as 15 subunits of V-ATPase were examined using qRT-PCR. Two other V-ATPase-related genes and five stress response genes were also examined using RT-PCR. The primers are shown in Table S1. Meanwhile, 35S::ThWRKY7 was transiently transformed into T. hispida for analysis of Cd tolerance.
Tonoplast isolation, SDS-PAGE and western blotting. Using a modification of the method of Ma et al. (2002) 34 , tonoplasts were isolated from 200 g of aerial tissue from five-week-old transgenic and WT Arabidopsis either treated with 100 μ M CdCl 2 or well watered (as a control) for 6 d on a 0-25% sucrose solution plate. Similarly, tonoplasts from T. hispida seedlings with transient expression of 35S::ThVHAc1, empty prokII (named T-ck), or RNAi::ThVHAc1 treated with 100 μ M CdCl 2 for 0 (control) or 2 h were also isolated. SDS-PAGE of purified V-ATPase (100 μ g tonoplast protein) was conducted using 15% polyacrylamide gels according to a previously published procedure 34,35 . The blots were performed based on the recommendations of Ma et al. (2002) 34 and Burnette et al. (1981) 36 . Meanwhile, the relative activity levels of P-ATPase and F-ATPase from the 25%-50% sucrose solution plate were also tested 34,35 . Proton pumping assays, ATPase hydrolysis activity, ATPase activity, and protein concentration. A total of 60 μ g of the membrane protein from each sample was used to monitor proton pumping activity.

Results and Discussion
Cloning and analysis of ThVHAc1 promoter. A 1,164 bp promoter fragment (from − 1 to − 1164) was amplified by TAIL-PCR, and pThVHAc::GUS transgenic Arabidopsis was generated. GUS staining in Arabidopsis revealed GUS activity in mature seeds, cotyledons, leaves, stems, roots, petals, stamen, stigma, pistils, and anthers but not in fresh pods or immature seeds ( Fig. 1a-q), indicating that ThVHAc1 expression is tissue-specific. Consistent with this finding, Padmanaban et al. 39

(2004) detected strong GUS activity in the root cap in
AtVHA-c3 promoter::GUS transgenic plants, and AtVHA-c3 dsRNA-mediated mutant lines exhibited decreased root length and diminished salt tolerance. These results indicated that the roots play a role in the regulation of abiotic stresses, especially Cd tolerance, by ThVHAc1. Furthermore, GUS activity was more obvious in roots and leaves than in stems ( Fig. 1a-l). Upon exposure to CdCl 2 for 12 h, the GUS activity increased 2.04-fold in aerial parts and 1.77-fold in roots compared to control conditions. When treated for 24 h, the increase was 1.87-fold in aerial parts and 1.22-fold in roots (Fig. 2a,b), indicating that the ThVHAc1 promoter confers a tissue-specific Cd stress response. Furthermore, GUS staining also revealed the transient expression of pThVHAc::GUS in T. hispida, showing that the GUS activity increased after CdCl 2 treatment. The GUS activity increased 2.99-fold or 1.94-fold in leaves and 2.34-fold or 1.51-fold in roots when treated with CdCl 2 for 12 or 24 h, respectively (Fig. 2c,d). All GUS activities in both transgenic Arabidopsis and transgenic T. hispida increased after CdCl 2 treatment for 12 h and 24 h, and these activity levels were consistent with the transcription level of ThVHAc1 in T. hispida roots and leaves under the same treatments, indicating a correlation between the promoter activity and ThVHAc1 gene expression pattern in plants exposed to Cd stress.
A PLACE database (http://www.dna.affrc.go.jp/PLACE) comparison of the ThVHAc1 promoter revealed many abiotic stress-related cis-elements, such as ARR1AT, CAATBOX, DOFCOREZM, EBOXBNNAPA, GT1CONSENSUS, MYCCONSENSUSAT, NODCON1GM, WBOX, and WRKY710S ( Fig. S2), indicating that ThVHAc1 may be regulated by different types of transcription factors (TFs). In particular, the promoter contains eighteen WBOX and WRKY motifs that are found in many promoters of stress tolerance genes, such as TaeIF5A 40 , PROPEP2 and PROPEP3 41 , and GbDXS and GbGGPPS 42 . The promoter of pathogenesis-related protein (PR) in parsley also contains WRKY motifs to which WRKY transcript factors specifically bind, and these motifs were further used to identify PR proteins 43 . In bananas, ethylene-induced ripening induces the expression of both the PR gene and the V-ATPase c" subunit in fruit tissue, suggesting a probable interaction between the PR and V-ATPase c" genes 44 . Thus, WRKY genes may also regulate V-ATPase c" gene expression. Moreover, the V-ATPase c" subunit shares high homology with VHA-c, and their expression patterns under different stresses were highly similar in T. hispida roots, stems, and leaves (data not shown), suggesting a possible synergistic function in stress tolerance. These results led us to investigate whether WRKY motif-binding TFs bind specifically to the ThVHAc1 promoter and function as its upstream regulator to control ThVHAc1 expression during stress tolerance.
ThWRKY7 is an upstream regulator of ThVHAc1. A yeast one-hybrid assay was used to verify the interaction between TFs and the WRKY motif in the promoter. ThWRKY7 was found to bind to the WRKY motif, as shown by the interaction between pHis2-WRKY-Seg and ThWRKY7 and the absence of an interaction between pHis2-WRKY-M/M1/M2 and ThWRKY7 on the SD/-Trp-Leu-His/50 mM 3-AT (3-amino-1, 2, 4-triazole) solid medium (Fig. 3a). Furthermore, the effector construct prokII-ThWRKY7 was transferred into Arabidopsis, and T 3 seedlings were used to detect transient expression of reporter plasmids. The reporter plasmids were constructed in pCAMBIA1301 harboring the intact or mutated WRKY motif, harboring the ThVHAc1 promoter fragment with its intact or mutated WRKY motif, or without the WRKY motif, followed by a 46 bp minimal promoter. The leaves of prokII-ThWRKY7 transgenic T 3 seedlings were then transiently transformed with one of the above-mentioned reporter plasmids. GUS activity was clearly detected in prokII-ThWRKY7 transgenic leaves with the reporter plasmid containing the ThVHAc1 promoter fragment with its intact WRKY motif, whereas   other mutated reporters did not activate GUS expression (Fig. 3b,c). These results further demonstrate that ThWRKY7 may bind specifically to the WRKY motif in the ThVHAc1 promoter.
qRT-PCR analysis of ThVHAc1 and ThWRKY7 showed that the two genes displayed similar expression patterns under CdCl 2 treatment. Specifically, the relative expression levels of both genes in roots increased before 12 h after CdCl 2 treatment and then decreased, reaching their lowest levels at 24 or 48 h (Fig. 4a). In stems, the expression of ThWRKY7 was slightly higher than that of ThVHAc1 at every time point. However, the two genes exhibited the same expression pattern, with the expression levels being highest at 24 h and lowest at 72 h (Fig. 4b). In leaves, the two genes were downregulated at 6 h and exhibited peak expression at 48 h (Fig. 4c). This synchronized expression patterns indicate that ThWRKY7 may play a critical role in either regulating the ThVHAc1 expression or cooperating with ThVHAc1 to improve plant Cd stress tolerance. To further confirm this conclusion, ThWRKY7 was transiently overexpressed in T. hispida. The leaves transformed with 35S::ThWRKY7 showed lower levels of DAB and Evans blue staining as well as slower accumulation of MDA and EL than did leaves transformed with T-ck. Moreover, the SOD, POD, GST, and GPX activities of leaves transformed with 35S::ThWRKY7 were significantly higher than those of leaves transformed with T-ck (Fig. S3), suggesting that overexpression of ThWRKY7 also markedly improves Cd tolerance, further demonstrating that ThWRKY7 may fine-tune the ThVHAc1-mediated Cd stress response.
WRKY transcription factors are a complex family with previously reported relationships to the plant immune response, in which they function as either positive or negative regulators 45,46 . Various functions in the protection process are a basic feature of WRKY genes, and the redundant elements in the promoters of their target genes imply a regulatory capacity of WRKY 47 . WRKY TFs regulate plant tolerance to abiotic stress by binding to the WRKY cis-element present in many stress-related and co-regulated gene promoters in Arabidopsis 46 . Regarding our results, the binding of ThWRKY7 to the WRKY element in the ThVHAc1 promoter, as well as the similar expression profiles of these elements over time when exposed to Cd stress, suggest that ThVHAc1 and ThWRKY7 may co-regulate Cd tolerance and that ThWRKY7 may control ThVHAc1 in the improvement of abiotic stress tolerance.

Heterologous expression of ThVHAc1 improves Cd tolerance in Arabidopsis.
To study the function of ThVHAc1 in plants, ThVHAc1 was overexpressed in Arabidopsis, and two transgenic lines (c1#10 and c1#17) were subjected to CdCl 2 treatment. The results showed that the germination, fresh weight, and main root length did not differ among c1#10, c1#17, WT, and ck under normal conditions (Fig. 5). However, when exposed to the CdCl 2 treatment c1#10 and c1#17 showed better biomass accumulation than WT and ck. The average germination, fresh weight, and main root length of c1#10 and c1#17 were 3.3-, 1.5-, and 1.7-fold greater, respectively, than those of both WT and ck (P < 0.05) (Fig. 5), indicating that the heterologous expression of ThVHAc1 can improve biomass and germination of Arabidopsis under Cd stress.
The Cd content was determined in two transgenic and two control lines. Under the control condition, the levels of Cd accumulation did not differ among the four lines, but when treated with 100 μ M CdCl 2 , the Cd content in WT and ck was an average of 1.4-fold higher than that in c1#10 and c1#17, which is a statistically significant difference (Fig. 6a,b). This evidence suggests that the heterologous expression of ThVHAc1 can reduce Cd accumulation in plants.
Heavy metal stress always leads to the production of ROS and to disturbed cellular redox status. Plants respond by increasing the production of a series of enzymes such as V-ATPase, SOD, POD, GPX, ascorbate peroxidase (APX), and GST 48,49 . Therefore, we compared the V-ATPase activity between the control and ThVHAc1 transgenic lines under CdCl 2 treatment for 6 d. The tonoplasts of WT, ck, and transgenic lines were isolated. Western blotting indicated the successful isolation of tonoplast, as determined by the detection of V-ATPase (Fig. 6c). Assays for V-ATPase-related activities were performed, and all four Arabidopsis lines displayed similar hydrolytic activity, ATPase activity, and proton transport activity under normal conditions. After CdCl 2 treatment, all V-ATPase-related activities were increased in all lines, but the activities in the transgenic ThVHAc1 lines were increased more than those in the control WT and ck lines. The average V-ATPase activities, hydrolytic activities, and proton transport activities of c1#10 and c1#17 were 1.8-, 1.2-, and 1.4-fold greater than those of the control lines, respectively (Fig. 7). These results indicate that the heterologous expression of ThVHAc1 leads to an increase in V-ATPase-related activities in response to Cd stress.
Similarly to V-ATPase, F-ATPase and P-ATPase may also be involved in stress tolerance. Lemos et al. (2005) showed that F-ATPase functions in maintaining cytoplasmic pH, determining the acid tolerance of cariogenic Streptococci mutans 50 . P-ATPase was previously shown to be an important factor in salt tolerance 48,51,52 . In this   study, the activity of F-ATPase and P-ATPase in isolated tonoplast samples was measured after the addition of the corresponding inhibitors, and the results were very similar to those for V-ATPase activity (Fig. 7). Compared with WT and ck, the activity of F-ATPase and P-ATPase were increased 1.2-and 1.3-fold, respectively, in Arabidopsis lines overexpressing ThVHAc1. Taken together, this finding demonstrates that the regulation of Cd tolerance is complex, and the possible roles of F-ATPase and P-ATPase in Cd tolerance merit further study.
The activity of the antioxidants SOD, POD and GST also did not differ among the four lines under normal conditions. After exposure to CdCl 2 treatment, the SOD, POD, and GST activities in c1#10 and c1#17 were significantly higher than those in WT and ck. The SOD activity of c1#17 was 1.92-fold greater than that of WT, the POD activity of c1#17 was 1.77-fold greater than that of the control lines, and the average GST of the transgenic lines was 1.46-fold greater than that of the control lines (Fig. 7f-h). These results suggested that the heterologous expression of ThVHAc1 correlates with increased activities of protective regulatory enzymes under Cd stress. A previous study suggested that antioxidant activity, such as that of POD in Kandelia candel and lipid peroxidation in Bruguiera gymnorrhiza, can be used as a biomarker for heavy metal stress conditions 53 . In the current study, the activities of V-ATPase, SOD, POD, and GST were all increased and higher than the corresponding activities in the control lines exposed to Cd stress, indicating that the heterologous expression of ThVHAc1 increased the activities of the above enzymes, keeping the ROS level low in transgenic Arabidopsis. Consistent with the above results, the histochemical staining of the ROS level showed that under normal conditions, levels of DAB staining for H 2 O 2 and NBT staining for O 2− in leaves were similar among WT, ck, c1#10, and c1#17. After CdCl 2 treatment, c1#10 and c1#17 accumulated less H 2 O 2 and O 2− than did WT and ck (Fig. 8a,b). The H 2 DCF staining of ROS in intact guard cells and main roots also showed that ROS accumulation in WT and ck was higher than that in c1#10 and c1#17 under CdCl 2 treatment (Fig. 8d,e), suggesting a positive role for ThVHAc1 in regulating the ROS level in plants under Cd stress. In addition, Evans blue and propidium iodide staining for cell damage in leaves and main roots also revealed less cell damage in transgenic lines than in WT and ck (Fig. 8c,f). Meanwhile, the EL rate and MDA, H 2 O 2 and proline contents of WT and ck were also significantly higher than those of c1#10 and c1#17 (p < 0.05) (Fig. 9a-d), confirming that the Cd stress response involves ROS metabolism and that ThVHAc1 may play a positive role in Cd stress tolerance by controlling ROS homeostasis.
Plants exposed to various stresses need to maintain normal metabolic functions, such as growth 54 , water-holding capacity 55 and chlorophyll content 56 . Water and chlorophyll are requirements for photosynthesis. Our results showed that the transgenic lines had higher water-holding capacity than the control lines (Fig. 9e) and that after CdCl 2 treatment, the chlorophyll content of c1#10 and c1#17 was higher than that of WT and ck (Fig. 9f). Figure 6a also shows that growth for all four lines is similar under normal conditions; however, after exposure to CdCl 2 treatment for 6 d, c1#10 and c1#17 displayed greener leaves than WT and ck. All these results suggest that the heterologous expression of ThVHAc1 in Arabidopsis increased the activity of both V-ATPase and antioxidants, which may regulate ROS homeostasis, cell damage, and photosynthesis for better Cd stress tolerance.  of T-ck, and the expression of RNAi::ThVHAc1 was only 6.79% that of T-ck (Fig. 10a), indicating that the transient expression in these lines was successful.  DAB and Evans blue staining of these transient expression lines showed that the ROS accumulation in RNAi::ThVHAc1 was higher than that in T-ck, while the lowest ROS accumulation was observed in 35S::ThVHAc1 under CdCl 2 treatment (Fig. 10b,c). The EL and MDA levels in 35S::ThVHAc1 were also significantly lower than those in T-ck and RNAi::ThVHAc1. Specifically, the EL in 35S::ThVHAc1 was 73.6% of that in T-ck and 58.9% of that in RNAi::ThVHAc1, and the MDA content in 35S::ThVHAc1 was 78.2% of that in T-ck and 69.0% of that in and RNAi::ThVHAc1 (Fig. 10d,e). The tonoplasts of these lines were isolated, as confirmed by western blotting (Fig. 11a), and all control, RNAi::ThVHAc1, and 35S::ThVHAc1 lines showed similar V-ATPase activities before Cd stress. However, after treatment with CdCl 2 , 35S::ThVHAc1 displayed the highest V-ATPase activity and RNAi::ThVHAc1 the lowest. The V-ATPase activity, hydrolytic activity and proton transport activity of 35S::ThVHAc1 were 1.3-, 1.4-, and 1.6-fold greater than those in the RNAi::ThVHAc1 line, respectively (Fig. 11). The F-ATPase and P-ATPase activities showed tendencies similar to that of V-ATPase activity. In 35S::ThVHAc1, the F-ATPase and P-ATPase activities were 1.5-and 1.4-fold greater than those in RNAi::ThVHAc1, respectively, while the corresponding hydrolytic activities were 1.4-and 1.2-fold greater than those in RNAi::ThVHAc1 (Fig. 11).

Cd tolerance analysis in
Furthermore, the activities of protective enzymes, including SOD, POD, GST, and GPX, were significantly higher in 35S::ThVHAc1 than in RNAi::ThVHAc1 and T-ck after CdCl 2 treatment (Fig. 12). These results further suggest that ThVHAc1 participated in the regulation of Cd tolerance by increasing the activity of protective enzymes to maintain ROS homeostasis in cells. Taken together, these results indicate that ThVHAc1 may be an effective gene for improving plants' Cd tolerance.

Expression of ThVHAc1 influenced other related genes and V-ATPase subunits. To investigate
whether other genes were affected by the expression of ThVHAc1, the expression levels of five AHA (H + -ATPase), five ACA (auto-inhibited Ca 2+ -ATPase), and eight stress-related genes were analyzed by RT-PCR. AHA genes primarily participate in ATP binding, the biosynthetic process, protein phosphorylation-dependent regulation, and coupling with transmembrane ion movement 57 . One of the AHA genes, At3g42640, was reported to be induced during Arabidopsis pollen development and during fertilization in B. campestris subsp. Chinensis 58 . ACA genes function in ATP activity, calcium channel activity, catalytic activity, hydrolase activity, and carbonate dehydratase activity 59 . When exposed to a boron deficiency for 24 h, the transcriptional level of the ACA gene At1g27770 increased by 1.43-fold 60 . CSD (cytosolic copper/zinc superoxide dismutase, At1g08830) is involved in ROS accumulation 61 , and APX (ascorbate peroxidase, At1g07890) is an ascorbate peroxidase with increased activity under oxidative stress in DET2-mutant Arabidopsis 62 . The transcription levels of RBOHC (respiratory burst oxidase homolog c, At5g51060) and LOX1 (lipoxygenase, At3g45140) were markedly upregulated when Arabidopsis was exposed to Cd stress 63 .
The results of the present study revealed that three AHA genes (At2g07560, At1g80660, and At3g42640), three ACA genes (At1g27770, At1g08065, and At1g08080) and four stress-related genes [(LOX1 (At3g45140), CSD (At1g08830), APX (At1g07890), and RBOHC (At5g51060)] were expressed at higher levels in the transgenic Arabidopsis c1#10 and c1#17 lines than in the WT and ck lines. For example, the ThVHAc1 transgenic lines All data are displayed as the mean ± SD of three independent experiments, and significant differences between transgenic lines and WT (P < 0.05) are indicated by asterisks. expressed the AHA genes at levels more than 3-fold those of the WT. The highest expression level of the AHA gene was approximately 4.9-fold higher (relative to ck) in c1#10, while that of the CSD gene was 5.3-fold higher (relative to ck) in c1#10 (Fig. 13a-d).
In three transiently transformed T. hispida lines, four antioxidant genes were analyzed using qRT-PCR. ThSOD, ThPOD, ThGSTZ1 and ThGPX showed similar expression profiles, all of which were upregulated after CdCl 2 treatment. The expression levels of these genes were highest in 35S::ThVHAc1 and lowest in RNAi::ThVHAc1 (Fig. 12). We also characterized the expression of several genes associated with stress-related functions and V-ATPase activity in T. hispida. The upregulated genes are shown in Fig. 13e,f, including one vacuolar cation/proton exchanger isoform CAX2 gene, three chloroplast protease genes (CSB), one ATP-dependent protease proteolytic subunit (ADP), one glycoside hydrolase protein (GLH) and one NADPH gene. These results indicate that the expression of ThVHAc1 changed the expression of other stress-related and V-ATPase-related genes, suggesting a complex network of Cd tolerance regulation.
Consistent with this result, other researchers have also shown that the overexpression or suppression of some genes always affects other genes. For example, overexpression of a DREB gene affected the expression of SOD, GST, and other stress-related genes 27 . R740S mutation in the a3 subunit of V-ATPase decreased the expression of key osteoclast markers (TRAP, cathepsin K, OSCAR, DC-STAMP, and NFATc1) 64 . The overexpression of SaVHAc1 in rice upregulated many stress-related genes, such as cysteine synthase, the pathogenesis-related protein Bet vI family protein, and glutamine synthetase under salt stress 65 .
V-ATPase is a multi-subunit enzyme. Overexpression of the ThVHAc1 gene affected many aspects of V-ATPase activity under CdCl 2 treatment (Figs 7 and 11). To better understand whether other subunits of V-ATPase were also affected by the expression of ThVHAc1 under CdCl 2 treatment, the expression profiles of 28 subunits in Arabidopsis were analyzed by qRT-PCR in c1#10 and c1#17 at different times. Clustering analysis of the expression patterns of all 28 subunits in c1#10 showed that they were primarily clustered into three groups. All subunits in group 1, including AtVHA-E2, F, G1, G2 and E1, were upregulated. Meanwhile, most subunits in group 2, including AtVHA-B1, C, B3, a2, H, e2, A, E3, d1 and d2, were induced after 12 h of treatment. The remaining subunits, except a1 and c1, belong to group 3 and were primarily suppressed, especially after 24 h of treatment (Fig. 14). The expression of the five AtVHA-c subunits was unchanged except for AtVHA-c3, which was induced at 6 h ( Fig. 15). At the same time, ThVHAc1 showed much higher expression under the same conditions, especially at 12 h (Fig. 15), suggesting that expression of the exogenous VHA-c subunit may suppress the expression of intrinsic VHA-c genes. The expression patterns of all subunits were also similar in c1#17 (Fig. S4), indicating that expression of ThVHAc1 may cause other subunits to participate in V-ATPase regulation under Cd stress and that V-ATPase activity is controlled by a complex network.
Interestingly, transient overexpression of ThVHAc1 in T. hispida had a different effect. Fifteen subunits were amplified from the T. hispida cDNA library. All subunits except ThVHA-H in 35S::ThVHAc1 and ThVHAc1 in RNAi::ThVHAc1 showed positive expression levels under the control conditions, and most subunits showed greater expression in RNAi::ThVHAc1 than in 35S::ThVHAc1 (Fig. 16). However, when treated with CdCl 2 , although all subunits except ThVHA-e in 35S::ThVHAc1 were induced, their expression was higher in 35S::ThVHAc1 than in RNAi::ThVHAc1 (Fig. 16). These results suggest that ThVHAc1 responds to Cd stress and that all subunits may  participate in the regulation of V-ATPase activity. However, the mechanisms by which all subunits act in such a complex network of V-ATPase regulation require further study.
A previous study indicated that the expression patterns of different subunits of V-ATPase may differ under the same stress. The Mesembryanthemum crystallinum V-ATPase subunits A, B and c were all upregulated approximately 2-fold relative to the control plant in roots and young leaves when exposed to salt stress. However, when the leaves fully expanded, only the c subunit was induced in reaction to salt 66 . Sugar beet VHA-A and VHA-c were coordinately expressed during plant development and are induced in response to high salinity 67 . The subunit E was also induced after treatment with salt for 3 d in mature common ice plant leaves, but it was not induced in juvenile leaves under the same conditions 68 .

Conclusion
In the plant kingdom, the V-ATPase c subunit (VHAc) is an important component of V-ATPase, which mediates abiotic stress responses. Some studies have demonstrated the salt regulation capacity of the VHAc gene. However, there are few reports on VHAc gene function in response to heavy metal stresses in a woody halophyte. Because ThVHAc1 rapidly responded to Cd stress in T. hispida, in this study, we further investigated the role of ThVHAc1 in Cd tolerance regulation. Our results showed that overexpression of ThVHAc1 effectively enhanced the tolerance of the transgenic Arabidopsis and T. hispida plants to Cd stress and that ThVHAc1 may modulate Cd stress tolerance by improving protective enzyme activities and strengthening the reactive oxygen species (ROS) scavenging system to decrease the cell damage when exposed to CdCl 2 treatment. Moreover, we identified a potential upstream regulator of ThVHAc1, ThWRKY7, which also responded to Cd stress, showed expression patterns similar to those of ThVHAc1, and improved the Cd stress tolerance of transgenic T.hispida. Although it remains unclear whether ThWRKY7 and ThVHAc1 cooperate to participate in the regulation of tolerance to other stresses, the current study provides new insights into the role and regulatory mechanism of ThVHAc1 in the regulation of tolerance to Cd stress in T. hispida.