AtHAP5A modulates freezing stress resistance in Arabidopsis through binding to CCAAT motif of AtXTH21

Authors

  • Haitao Shi,

    1. Key Laboratory of Plant Germplasm Enhancement and Specialty Agriculture, Wuhan Botanical Garden, Chinese Academy of Sciences, Wuhan, China
    Search for more papers by this author
  • Tiantian Ye,

    1. Key Laboratory of Plant Germplasm Enhancement and Specialty Agriculture, Wuhan Botanical Garden, Chinese Academy of Sciences, Wuhan, China
    2. University of Chinese Academy of Sciences, Beijing, China
    Search for more papers by this author
  • Bao Zhong,

    1. Key Laboratory of Plant Germplasm Enhancement and Specialty Agriculture, Wuhan Botanical Garden, Chinese Academy of Sciences, Wuhan, China
    2. University of Chinese Academy of Sciences, Beijing, China
    Search for more papers by this author
  • Xun Liu,

    1. Key Laboratory of Plant Germplasm Enhancement and Specialty Agriculture, Wuhan Botanical Garden, Chinese Academy of Sciences, Wuhan, China
    2. University of Chinese Academy of Sciences, Beijing, China
    Search for more papers by this author
  • Rui Jin,

    1. Key Laboratory of Plant Germplasm Enhancement and Specialty Agriculture, Wuhan Botanical Garden, Chinese Academy of Sciences, Wuhan, China
    2. University of Chinese Academy of Sciences, Beijing, China
    Search for more papers by this author
  • Zhulong Chan

    Corresponding author
    1. Key Laboratory of Plant Germplasm Enhancement and Specialty Agriculture, Wuhan Botanical Garden, Chinese Academy of Sciences, Wuhan, China
    Search for more papers by this author

Summary

  • Several eukaryotic Heme-associated proteins (HAPs) have been reported to bind specifically to DNA fragments containing CCAAT-box; however, the physiological functions and direct targets of these HAP proteins in plants remain unclear.
  • In this study, we showed that AtHAP5A as a transcription factor interacted with CCAAT motif in vivo, and AtXTH21, one direct target of AtHAP5A, was involved in freezing stress resistance. The AtHAP5A overexpressing plants were more tolerant, whereas the loss-of-function mutant of AtHAP5A was more sensitive to freezing stress than wild-type plants. Chromatin immunoprecipitation (ChIP) assay demonstrated that AtHAP5A could bind to five fragments that contained CCAAT motifs in the AtXTH21 promoter.
  • Similarly, the AtXTH21 overexpressing plants exhibited improved freezing resistance, while xth21 knockdown mutants displayed decreased freezing resistance. Notably, the modulated freezing resistance of AtHAP5A overexpressing plants and knockout mutant could be reversed by the xth21 mutant and AtXTH21 overexpressing plants, respectively, indicating that AtHAP5A might act upstream of AtXTH21 in freezing stress. Additionally, modulation of AtHAP5A and AtXTH21 expression had the same effects on abscisic acid (ABA) sensitivity and reactive oxygen species (ROS) metabolism.
  • Taken together, these results demonstrated that AtHAP5A modulates freezing stress resistance in Arabidopsis through binding to the CCAAT motif of AtXTH21.

Introduction

Low temperature (cold stress), including chilling (< 20°C) and freezing (< 0°C), is a serious environmental stress that disrupts cellular homeostasis and limits plant growth (Doherty et al., 2009; Guo et al., 2013). To respond to seasonal variations in temperature, immotile plants have developed complex biochemical and physiological processes (Doherty et al., 2009; Qin et al., 2011; Guo et al., 2013). A number of plant hormones and genes, and several transcription factors in particular, play pivotal roles in plant cold stress response (Xiong et al., 2002; Vogel et al., 2005; Yamaguchi-Shinozaki & Shinozaki, 2005; Dong et al., 2006; Jeon et al., 2010; Hu et al., 2013). Plant transcription factors including basic domain-leucine zipper (bZIP) families, MYBs, MYCs and zinc finger proteins, not only serve as important mediators during hormone crosstalks under stress conditions, but also function in the early stress signaling transduction via directly activating or inhibiting the expression of several stress-responsive genes (Qin et al., 2011; Shi et al., 2012b).

In Arabidopsis, a widely known model involved in plant cold stress response is the C-repeat (CRT)/Dehydration responsive element (DRE) BINDING FACTORs (CBF/DREBs)-mediated pathway (Mäntylä et al., 1995; Thomashow, 1999, 2010; Agarwal et al., 2006; Dong et al., 2006). Briefly, INDUCER OF CBF EXPRESSION 1/2 (ICE1/2), encoding basic helix-loop-helix (bHLH) transcription factors, directly bind to CANNTG box in the promoter of CBF/DREBs which interact with CRT/DRE of downstream Cold-Regulated (COR) genes and regulate cold stress response (Gilmour et al., 1998; Jaglo-Ottosen et al., 1998; Chinnusamy et al., 2003; Zarka et al., 2003; Cook et al., 2004; Zhang et al., 2004; Canella et al., 2010). Recently, abscisic acid (ABA), cytokinin, ethylene and jasmonate (JA) were found to be involved in plant cold resistance through the CBFs-mediated pathway (Jeon et al., 2010; Shi et al., 2012b; Hu et al., 2013). More recently, Guo et al. (2013) found that Arabidopsis Lipid transfer protein 3 (AtLTP3) might act as a target of AtMYB96 to be involved in plant resistance to cold stress independent of the CBF pathway. Because plant cold stress response is a complex signaling pathway, many unknown mechanisms need to be further dissected, especially the in vivo roles of several transcription factors.

The CCAAT motif is a very common cis-element in the promoters of many eukaryotic genes (McNabb et al., 1995; Edwards et al., 1998; Kato, 2005; Yazawa & Kamada, 2007). Plant HAP (for histone- or haem-associated protein) complex has been widely confirmed to be required for CCAAT binding in yeast and in plants (McNabb et al., 1995; Edwards et al., 1998; Kato, 2005; Yazawa & Kamada, 2007). Edwards et al. (1998) have characterized Arabidopsis homologs of the HAP proteins of CCAAT-binding transcription factors, including AtHAP2/3/5. Yazawa & Kamada (2007) found that AtHAP proteins could bind specifically to DNA fragments containing the CCAAT motif via electrophoretic mobility shift assay (EMSA). Although Li et al. (2013) recently found that overexpression of the homologous HAP5 subunit from Picea wilsonii in Arabidopsis improved resistance to salt and decreased sensitivity to ABA, most previous plant HAPs studies have focused on their roles in developmental processes, including pollen tube guidance and fertilization (von Besser et al., 2006; Yu et al., 2011), flowering (Chen et al., 2007; Ito et al., 2011) and root elongation (Ballif et al., 2011). In Arabidopsis, HAP2 (also known as NUCLEAR FACTOR Y, SUBUNIT A, NF-YA), HAP3 (NF-YB) and HAP5 (NF-YC) were encoded by 10, 11 and 13 genes, respectively (Li et al., 2013). The HAP complex, composed of different subunits (HAP2, HAP3, and HAP5), is involved in a wide range of processes and gene transcription in plants (Yazawa & Kamada, 2007; Li et al., 2013).

Leyva-González et al. (2012) found that overexpression of NF-YA2, 7 and 10 enhanced tolerances to multiple types of abiotic stress including low Pi, flood, freezing and heat stresses. Additionally, transcriptomic analysis of transgenic plants that express miR169-resistant versions of NFYA2, 3, 7 and 10 under an estradiol inducible system, as well as a dominant-repressor version of NF-YA2, identified a set of differentially expressed genes whose promoters were enriched in CCAAT elements (Leyva-González et al., 2012). Recently, Petroni et al. (2012) reviewed the involvement of plant HAPs (NF-Ys) in plant–environment interactions. Several plant NF-Ys modulated plant resistance to drought and endoplasmic reticulum (ER) stress (Petroni et al., 2012), but no stress-related target of plant HAPs has been revealed, and the direct link between the CCAAT element and plant abiotic stress response is also unknown.

In this study, we investigated the in vivo role of AtHAP in plant cold stress by modulating expression of AtHAP5A (At3g48590). The results demonstrated that AtHAP5A overexpressing plants were more tolerant to freezing stress than wild-type (WT) plants, whereas the loss-of-function mutant of AtHAP5A was more sensitive to freezing treatment. AtXTH21 (At2g18800), which had enriched CCAAT motifs in the promoter region, was identified as a direct target of AtHAP5A through genomic analysis and chromatin immunoprecipitation (ChIP) assay. Moreover, genetic evidence indicated that AtHAP5A acts upstream of AtXTH21 in freezing stress response in Arabidopsis. These results revealed that AtHAP5A modulates freezing stress resistance through interaction with the CCAAT motif of AtXTH21 in Arabidopsis.

Materials and Methods

Plant materials and growth conditions

Arabidopsis thaliana (L.) Heynh ecotype Columbia was used in this study. Arabidopsis seeds were sterilized with 70% (v/v) ethyl alcohol, 10% (w/v) NaClO and deionized water. After stratification at 4°C for 3 d in darkness, Arabidopsis seeds were sown in soil or on Murashige and Skoog (MS) medium containing 1% sucrose (w/v) in the growth chamber. The growth chamber was controlled at an irradiance of c. 120 μmol quanta m−2 s−1, 23 ± 2°C, with 65% relative humidity under 16 h : 8 h, light : dark cycles. The nutrient solution was watered from below in pots with soil-grown plants twice every week. The mutants of hap5a (SALK_086334) and xth21 (SALK_057963) were obtained from the Arabidopsis Biological Resource Center.

In order to determine the effects of different stresses on the expressions of AtHAP5A and AtXTH21, 14-d-old seedlings grown on MS medium were transferred to fresh MS liquid medium containing 50 μM ABA or 200 mM NaCl, or subjected to dehydration (un-covering the plate in the growth chamber), 23 and 4°C stress treatments, and samples were collected at 0, 3, 6, 12 and 24 h after treatments.

Transgenic plasmid construction and plant transformation

For the overexpressing constructs, AtHAP5A and AtXTH21 cDNA were amplified and cloned into the pCAMBIA1301S vector (Hu et al., 2006) under the control of CaMV 35S promoter to generate 35S::HA-AtHAP5A and 35S::HA-AtXTH21 constructs. For the AtHAP5A and GFP co-expressing construct, AtHAP5A cDNA was amplified and cloned into pEGAD vector to generate 35S::GFP-AtHAP5A (Cutler et al., 2000). The corresponding primers for gene clone were listed in Supporting Information Table S1.

The above recombinant constructs were then transformed into Agrobacterium tumefaciens strain GV3101 and introduced into Arabidopsis WT (Col-0) plants, and hap5a (SALK_086334) and xth21 (SALK_057963) mutants using the floral dip method (Clough & Bent, 1998) to obtain AtHAP5A and AtXTH21 overexpressing and complement plants. Homozygous transgenic plants were selected on MS medium containing 3% (w/v) sucrose using hygromycin and basta resistance for pCAMBIA1301S and pEGAD vector constructs, respectively, and the transgenic plants were confirmed by PCR analyses. Additionally, the expression levels of corresponding genes were also determined by semi-quantitative reverse transcription (RT)-PCR and quantitative real-time PCR.

RNA isolation, Semi-quantitative RT-PCR and quantitative real-time PCR

Total RNA was extracted from 100 mg of plant samples (flower, leaf, root, silique and stem tissues for expression pattern analyses and leaf tissue for other expression analyses) using TRIzol reagent (Invitrogen) and treated by RQ1 RNase-free DNase (Promega) to avoid possible genomic DNA contamination. First-strand cDNA was then synthesized using reverse transcriptase (Toyobo, Osaka, Japan) from 2 μg of total RNA. Semi-quantitative RT-PCR and quantitative real-time PCR were performed as described in Shi et al. (2013a,b). Briefly, quantitative real-time PCR was performed using the CFX96 Real Time System (Bio-Rad) with iQ SYBR® Green Super mix (Bio-Rad) and diluted cDNA. All experiments were repeated at least three times, and the expression levels were standardized with EIF4A and UBQ10 using the comparative ΔΔCT method. The specific primers are listed in Table S1.

Transactivation activity assay for AtHAP5A in yeast

For the transactivation activity assay of AtHAP5A, the GAL4 System (Clontech, CA, USA) was used as described by Peng et al. (2012). Briefly, the coding region of AtHAP5A was fused to the GAL4 DNA-binding domain in the bait plasmid pGBKT7 to generate AtHAP5A-pGBKT7. The pCL-1-pGBKT7 and the empty vector pGBKT7 were used as positive and negative controls, respectively (Peng et al., 2012). The primers for vector construct were listed in Table S1. The above plasmids were transformed into the yeast strain AH109, and the transformants were then plated on selective SD medium (SD-Trp, SD-Trp-Ade-His with 10 mM 3-AT) and grown for 2 d at 30°C. To confirm the result, the a-galactosidase activity of yeast strains was assayed using p-nitrophenyl a-d-galactopyranoside as a substrate.

Subcellular localization analysis

For subcellular localization analysis, GFP signals in 7-d-old plant root elongation regions and root tips of 35S::GFP-HAP5A transgenic plants were detected using an Olympus FluoView 1000-confocal laser scanning microscope. Additionally, 20 μg ml−1 propidium iodide (PI) was used for cell wall staining in GFP lines.

Determination of freezing resistance

The freezing stress treatment was performed according to Shi et al. (2013a) with slight modifications. Briefly, 10-d-old MS-plate grown seedlings were nonacclimated at 23°C and cold acclimated at 4°C in the light for 3 d, and 14-d-old soil grown plants were nonacclimated at 23°C and cold acclimated at 4°C in the light for 7 d. Then the nonacclimated and cold acclimated plants were placed in a growth chamber set to −1°C for 16 h. Ice chips were sprinkled on these plants before the chamber was programmed to cool at 1°C h−1 until −8°C was reached. Plants were removed after 8 h of freezing, thawed at 4°C for 12 h in the dark, and then transferred to control condition at 22°C. The survival rate was assayed after 4 d of recovery.

Determination of electrolyte leakage (EL)

The assay of EL was performed as described in Shi et al. (2012a, 2013a,b,c,d). Briefly, c. 0.1 g of plant leaves were incubated in 20 ml deionized water, and shaken on a gyratory shaker at room temperature for 6 h at c. 150 rpm. The initial conductivity (Ci) was determined using a conductivity meter (DDS-307A; Shanghai Leici, Shanghai, China). The mixtures were then boiled for 20 min. After cooling to room temperature, the conductivity of the killed mixtures (Cmax) was measured. The relative EL (%) was calculated as (Ci/Cmax) × 100.

Chromatin immunoprecipitation (ChIP) assay

ChIP assays were performed as described previously (Bowler et al., 2004; Zong et al., 2013). Briefly, 4 g of 14-d-old 35S::HA-HAP5A transgenic plant leaves was harvested, and then immersed in 1% formaldehyde for cross-linking the DNA with DNA-binding proteins. Next, the chromatin pellets were extracted and sheared by sonication as described in Bowler et al. (2004). The anti-HA antibody (AH158; Beyotime, Haimen, China) was used to immunoprecipitate the DNA-HAP5A complexes. The DNA was released with proteinase K and then purified for further PCR analysis. The enrichment of DNA fragments was determined using quantitative real-time PCR with the specific primers listed in Table S1.

Transient protein expression in tobacco leaves

In order to confirm the interaction between AtHAP5A and the CCAAT box, the transient protein expression in tobacco leaves was assayed as described in Huang et al. (2013). The m35S-GUS, CCAAT-m35S-GUS and CATAT-m35S-GUS vectors were constructed by inserting PCR products of mini35S, 5×CCAAT-mini35S and 5×CATAT-mini35S from the minimal-100 CaMV 35S-pCAMBIA1391 plastid (Huang et al., 2013) into the HindIII and BamHI sites of the pCAMBIA1391Z vector. The corresponding primers for vector construct were listed in Table S1. For transient expression analysis, the pCAMBIA1301S and AtHAP5A-pCAMBIA1301S construct was used as the effector, the m35S-GUS (mini35S-pCAMBIA1391Z), CCAAT-m35S-GUS (5×CCAAT-mini35S-pCAMBIA1391Z) and CATAT-m35S-GUS (5×CATAT-mini35S-pCAMBIA1391Z) constructs were used as the reporters. The effector and reporter were transformed into Agrobacterium tumefaciens strain GV3101 and co-infected into tobacco leaves as Huang et al. (2013) described, and 35S::GFP was co-injected as an internal standard in each infection. Quantitative measurement of GUS activity and GFP signal was performed as described in Jefferson et al. (1987) and Shi et al. (2013a), respectively.

Determination of reactive oxygen species (ROS) accumulation and antioxidant enzyme activities

Plant leaves were harvested for the assays of ROS accumulation and antioxidant enzyme activities. For ROS accumulation, hydrogen peroxide (H2O2) and superoxide radical (O2•−) were determined using the titanium sulfate method and the Plant O2•− ELISA Kit (10-40-488; Bejing Dingguo, China) as described in Shi et al. (2012a, 2013a,b,c,d). The malondialdehyde (MDA) content was extracted using chilled thiobarbituric acid reagent and quantified via subtracting the nonspecific absorption at 450 nm and 600 nm from the specific absorption at 532 nm as described in Shi et al. (2012a, 2013a,c,d).

The activities of antioxidant enzymes (superoxide dismutase (SOD, EC 1.15.1.1), catalase (CAT, EC 1.11.1.6) and peroxidase (POD, EC 1.11.1.7)) were determined using the Total SOD Assay Kit (S0102; Beyotime), CAT Assay Kit (S0051; Beyotime) and Plant POD Assay Kit (A084-3; Nanjing Jiancheng, Nanjing, China), respectively, according to the manufacturer's instructions (see Shi et al. 2012a, 2013a,b,c,d).

Statistical analysis

All experiments in this study were repeated at least three times. Samples of every experiment were harvested from at least 20 seedlings per genotype. In all results, Student's t-test was used to determine the significant difference between WT and other lines, and asterisk symbols (*) indicate significant differences of < 0.05 compared with WT.

Results

The transactivation activity and subcellular localization of AtHAP5A

On the SD-Trp medium, yeast cells transformed with pCL-1-pGBKT7 (positive control) (Peng et al., 2012), AtHAP5A-pGBKT7 and pGBKT7 grew well (Fig. 1a). When transferred to SD-Trp-Ade-His medium containing p-nitrophenyl a-d-galactopyranoside and 10 mM 3-amino-1, 2, 4-triazole (3-AT), only the yeast AH109 strains with transformants of pCL-1-pGBKT7 (positive control) and HAP5A-pGBKT7 grew normally and exhibited the activity of a-galactosidase, while yeast cells transformed with empty plasmid pGBKT7 could not grow in these highly stringent conditions (Fig. 1a). These results indicated that HAP5A had transactivation activity in yeast.

Figure 1.

Transactional activation activity and subcellular localization of AtHAP5A. (a) Transactivation activity of AtHAP5A in yeast. The yeast strain AH109 containing plastids of AtHAP5A-pGBKT7, pCL-1-pGBKT7, pGBKT7 were plated on different selective SD mediums. The pCL-1-pGBKT7 was used as positive control. (b) Nuclear localization of AtHAP5A in 7-d-old Arabidopsis seedling root elongation regions and root tips of 35S::GFP-AtHAP5A transgenic plant. Bars, 200 μm. The red fluorescent indicated cell walls by propidium iodide (PI) staining, and the green fluorescent indicated the nuclear localization of AtHAP5A using confocal laser scanning microscope.

In order to investigate the subcellular localization of AtHAP5A, transgenic 35S::GFP-HAP5A plants were constructed and identified. 35S::GFP-HAP5A fluorescence was specifically localized in the nucleus, indicating that AtHAP5A is a nuclear protein (Fig. 1b).

Expression patterns of AtHAP5A and AtXTH21

First, we examined the expression levels of AtHAP5A and AtXTH21 in various organs of Arabidopsis plants by quantitative real-time PCR. Both AtHAP5A and AtXTH21 were highly expressed in Arabidopsis roots, and AtHAP5A exhibited higher expression levels in plant leaves than in flowers, siliques or stems, whereas AtXTH21 was weakly expressed in plant leaves, flowers, siliques and stems (Fig. 2a,b). Additionally, the expression levels of AtHAP5A and AtXTH21 in plant leaves were significantly increased after 4°C stress for 3, 6, 12 and 24 h, indicating the in vivo roles of AtHAP5A and AtXTH21 in plant response to cold stress (Fig. 2c,d). Moreover, the expression levels of AtHAP5A and AtXTH21 were also significantly induced after dehydration stress for 3, 6 and 24 h, after salt stress for 3, 6, 12 and 24 h, and after ABA treatment for 24 h (Fig. 2c,d). Interestingly, the expression levels of both AtHAP5A and AtXTH21 were not induced at 12 h after dehydration treatment (Fig. 2c,d), indicating the fluctuations in their expressions during dehydration stress treatment. Moreover, the same expression patterns under stress conditions further suggested the putative connection between AtHAP5A and AtXTH21.

Figure 2.

Expression patterns of AtHAP5A and AtXTH21. (a, b) The expressions of AtHAP5A (a) and AtXTH21 (b) in various organs of Arabidopsis plants by quantitative real-time PCR. (c, d) The expressions of AtHAP5A (c) and AtXTH21 (d) under control (23°C) condition and after 50 μM abscisic acid (ABA), cold stress (4°C), dehydration and 200 mM NaCl treatments by quantitative real-time PCR. The data represent the means of three independent experiments ± SD. Significant difference of < 0.05 between control and 4°C stress conditions: *, < 0.05.

Modulation of AtHAP5A expression affects freezing resistance

In order to investigate the in vivo role of AtHAP5A in freezing stress resistance, AtHAP5A overexpressing and knockout plants were isolated. The expression level of AtHAP5A was constitutively expressed in AtHAP5A overexpressing plants (Fig. 3a), but was undetectable in hap5a mutants (Figs 3b, S1a). The EL is a widely used indicator for membrane damage assay, and survival rate under stress conditions is an indicator of stress resistance. In the control and nonacclimated plants, no significant difference in plant growth, membrane injury (EL) and survival rate were obtained among WT and AtHAP5A overexpressing and knockout plants under control and freezing stress conditions (Figs 3c–e, S2a–c). In the acclimated plants, AtHAP5A overexpressing plants displayed improved freezing resistance with lower EL and higher survival rate than WT, while hap5a mutants exhibited decreased freezing resistance with higher EL and lower survival rate in comparison to WT (Fig. 3c–e). Additionally, the decreased freezing resistance of hap5a mutants could be restored in the complement plants (Fig. 3c–e). These results indicated that modulation of AtHAP5A expression affected freezing resistance in Arabidopsis.

Figure 3.

Freezing stress resistance of AtHAP5A overexpressing and knockout Arabidopsis plants. (a) The expression of AtHAP5A in AtHAP5A overexpressing plants by quantitative real-time PCR. (b) The AtHAP5A expressions of hap5a mutant and complement lines. (c, d) Photograph showing 14-d-old soil grown AtHAP5A overexpressing plants (c), hap5a mutant and complement lines (d) after freezing stress treatment and recovery. Briefly, 14-d-old soil grown plants were subjected to 23 and 4°C for nonacclimated and acclimated treatments for 7 d, respectively, and then transferred to −8°C for 8 h. (e, f) electrolyte leakage (EL) of nonacclimated and acclimated 14-d-old soil grown AtHAP5A overexpressing plants (e), hap5a mutant and complement lines (f) after freezing stress treatment. (g, h) The survival rate of AtHAP5A overexpressing plants (g), hap5a mutant and complement lines (h) after freezing stress treatment and 4 d of recovery as designed in (c). At least 78 plants in nine pots were counted for the survival rate assay in three independent experiments. The data represent the means of three independent experiments ± SD. Significant differences (Student's t-test) in comparison to wild-type (WT): *, < 0.05.

AtHAP5A can bind to CCAAT motif in vivo

In order to test the interaction of AtHAP5A and CCAAT in vivo, transient expression assays in tobacco leaves were performed using 35S-vector and 35S-AtHAP5A as the effectors, and m35S-GUS, CCAAT-m35S-GUS and CATAT-m35S-GUS as the reporters (Fig. 4a). As shown in Fig. 4(b) and Fig. S3, only those leaf pieces co-transformed with 35S-AtHAP5A and CCAAT-m35S-GUS largely activated the expression of GUS, while the leaf pieces co-transformed with 35S-AtHAP5A and CATAT-m35S-GUS could not activate the expression of GUS, indicating the in vivo interaction of AtHAP5A and CCAAT.

Figure 4.

AtXTH21 is a direct target of AtHAP5A. (a) Schematic diagrams of the effectors (35S-vector and 35S-AtHAP5A) and reporters (m35S-GUS, CCAAT-m35S-GUS and CATAT-m35S-GUS) used for transient expression analysis. (b) Transient expression of representative tobacco leaves infiltrated with different combinations of effector and reporter. 35S::GFP was co-injected as an internal standard in each infection. Significant difference between vector and AtHAP5A as the effector: *, < 0.05. (c) Schematic structure of five CCAAT motifs in the promoter of AtXTH21. (d) ChIP enrichment to show the in vivo binding ability of 35S::HA-HAP5A to the DNA fragments of AtXTH21 promoter by quantitative real-time PCR. The ChIP results were normalized to input chromatin, and a fragment in the ACTIN2 promoter was used as the negative control. Significant difference between nonantibody input and antibody input: *, < 0.05. (e, f) The effects of AtHAP5A overexpressing and knockout plants on the expressions of AtHAP5A (e) and AtXTH21 (f). Significant difference compared with wild-type(WT): *, < 0.05. The data represent the means of three independent experiments ± SD.

AtHAP5A directly regulates AtXTH21 expression

Through genome sequence analysis, five CCAAT motifs were enriched in AtXTH21 promoter region (Fig. 4c). Is AtXTH21 a direct target of AtHAP5A? To address this question, a ChIP experiment was performed to determine whether AtHAP5A was able to directly bind to the fragments that contained CCAAT motif in AtXTH21 promoter. For ChIP analysis, the HA-AtHAP5A fusion protein was triggered by the 35S promoter in AtHAP5A overexpressing plants. As shown in Fig. 4(d), the AtHAP5A protein strongly interacted with five fragments that contained CCAAT motifs under control condition (23°C), and the interaction was enhanced by 4°C treatment, indicating that the CCAAT motif is important for the binding of AtHAP5A to the AtXTH21 promoter.

Moreover, the expression of AtXTH21 was assayed in AtHAP5A overexpressing and knockout plants. After 4°C treatment for 1 d, the expression of AtHAP5A in WT plants showed a two-fold increase over the control condition (23°C) (Fig. 4e). Consistently, the expression of AtHAP5A was significantly induced in AtHAP5A overexpressing plants after 4°C treatment, while hap5a mutants had no AtHAP5A expression under control and 4°C stress conditions (Fig. 4e). Interestingly, the expression of AtXTH21 was significantly decreased in hap5a mutants, but increased in AtHAP5A overexpressing plants relative to WT (Fig. 4f). When exposed to 4°C stress treatment, the expression of AtXTH21 in WT, AtHAP5A overexpressing and knockout plants was largely induced. However, hap5a mutants exhibited lower AtXTH21 expression than WT, while AtHAP5A overexpressing plants displayed higher AtXTH21 expression than WT (Fig. 4f). These results indicated that AtHAP5A directly regulated AtXTH21 expression.

Modulation of AtXTH21 expression affects freezing resistance

In order to determine the possible role of AtXTH21 in freezing resistance, AtXTH21 overexpressing and knockdown plants were isolated and confirmed by quantitative real-time PCR (Figs 5a,b, S1b). Under control conditions, AtXTH21 overexpressing and knockout plants had the same growth pattern, and the same EL and survival rate as WT plants (Figs 5c–e, S2a–c). When freezing stress was applied, AtXTH21 overexpressing plants displayed improved freezing resistance with lower EL and higher survival rate than WT in both nonacclimated and acclimated plants, while xth21 mutants exhibited decreased freezing resistance with higher EL and lower survival rate in comparison to WT (Fig. 5c–e). Moreover, the decreased freezing resistance of xth21 mutants was restored in the complemented plants (Fig. 5c–e). These results indicated that modulation of AtXTH21 expression affected freezing resistance.

Figure 5.

Freezing stress resistance of AtXTH21 overexpressing and knockout plants. (a) The expression of AtXTH21 in AtXTH21 overexpressing plants by quantitative real-time PCR. (b) The AtXTH21 expressions of xth21 mutant and complement lines. (c, d) Photograph showing 14-d-old soil grown AtXTH21 overexpressing Arabidopsis plants (c), xth21 mutant and complement lines (d) after freezing stress treatment and recovery. (e, f) Electrolyte leakage (EL) of nonacclimated and acclimated 14-d-old soil grown AtXTH21 overexpressing plants (e), xth21 mutant and complement lines (f) after freezing stress treatment. (g, h) Survival rate of AtXTH21 overexpressing plants (g), xth21 mutant and complement lines (h) after freezing stress treatment and 4 d of recovery as designed in (c). At least 78 plants in nine pots were counted for the survival rate assay in three independent experiments. The data represent the means of three independent experiments ± SD. Significant differences (Student's t-test) in comparison to wild-type (WT): *, < 0.05.

Modulation of AtHAP5A and AtXTH21 expression affects ROS metabolism

Because ROS metabolism including ROS accumulation, lipid peroxidation and related antioxidant enzymes plays important roles in cold stress-triggered oxidative injury, the effects of AtHAP5A and AtXTH21 expressions on ROS metabolism were further determined. Under control condition, modulation of AtHAP5A or AtXTH21 expression had no significant effects on the accumulation of H2O2, O2•− and MDA and the activities of SOD, CAT and POD (Fig. 6a–f). When exposed to the 4°C stress condition, AtHAP5A and AtXTH21 overexpressing plants displayed significantly lower concentrations of H2O2, O2•− and MDA and higher activities of SOD, CAT and POD than WT, conferring less oxidative injury and a more effective antioxidant defense system (Fig. 6a–f). On the contrary, AtHAP5A and AtXTH21 mutants exhibited significantly higher concentrations of H2O2, O2•− and MDA and lower activities of SOD, CAT and POD than WT, conferring more oxidative damage (Fig. 6a–f). These results indicated that overexpressing AtHAP5A and AtXTH21 could alleviate 4°C stress-induced ROS accumulation and related oxidative damage in Arabidopsis, while AtHAP5A and AtXTH21 mutants had the opposite effect.

Figure 6.

Effects of AtHAP5A and AtXTH21 expressions on reactive oxygen species (ROS) metabolism. (a–f) Quantifications of H2O2 content (a), O2•− content (b), malondialdehyde (MDA) content (c), superoxide dismutase (SOD) activity (d), catalase (CAT) activity (e) and peroxidase (POD) activity (f) in 14-d-old soil grown wild-type (WT), AtHAP5A and AtXTH21 overexpressing and knockout Arabidopsis plants under control and 4°C stress conditions for 7 d. The data represent the means of three independent experiments ± SD. Significant differences (Student's t-test) in comparison to WT: *, < 0.05.

Mutation and ectopic expression of AtXTH21 restore the freezing resistance and ABA sensitivity in gain- and loss-of-function AtHAP5A plants

In order to further determine the genetic interaction of AtHAP5A and AtXTH21, we generated hap5a × XTH21-OE-19 and HAP5A-OE-12 × xth21 crossed transgenic plants and confirmed the crossed plants through PCR analysis. Notably, hap5a × 35S::AtXTH21 restored the decreased freezing resistance of hap5a mutants, and the improved freezing resistance of HAP5A-OE-12 transgenic plants was reversed in HAP5A-OE-12 × xth21 plants (Fig. 7a–c).

Figure 7.

Freezing stress resistance of athap5a×35S::AtXTH21 and 35S::AtHAP5A × atxth21 plants. (a) Photograph showing 14-d-old soil grown Arabidopsis plants after freezing stress treatment and recovery. (b) Survival rate of athap5a × 35S::AtXTH21 and 35S::AtHAP5A × atxth21 plants after 8 h of freezing stress treatment and 4 d of recovery as designed in (a). At least 78 plants in nine pots were counted for the survival rate assay in three independent experiments. The data represent the means of three independent experiments ± SD. Significant differences (Student's t-test) in comparison to wild-type (WT): *, < 0.05.

Additionally, in the presence of exogenous ABA, the germination rate and root elongation of both AtHAP5A and AtXTH21 overexpressing plant were much higher and longer than those of WT, respectively (Figs 8a–d, S4). Conversely, the hap5a and xth21 mutants showed sensitivity to ABA compared with the WT, and the complemented lines of hap5a and xth21 mutants rescued the ABA sensitivity phenotype of the mutants (Figs 8a–d, S4). Interestingly, mutation and ectopic expression of AtXTH21 restored the ABA sensitivity in gain- or loss-of-function AtHAP5A plants not only during seed germination stage, but also during the seedling growth stage (Figs 8a,d, S4). These results indicated that AtHAP5A might act upstream of AtXTH21 in ABA sensitivity and freezing stress response in Arabidopsis.

Figure 8.

Effects of AtHAP5A and AtXTH21 expressions on abscisic acid (ABA) sensitivity. (a) Comparison of germination of different Arabidopsis lines after 10 d of treatment on MS plates containing various concentrations of ABA. (b) The seed germination ratios of different lines in the presence of various concentrations of ABA. Germination rates were scored as emerged radicals and green cotyledon after 10 d in growth chamber. (c) Comparison of the inhibiting effect on root growth of different lines after 7 d of treatment on MS plate containing various concentrations of ABA. For the ABA sensitivity assay at seedling stage, plant seeds on MS medium were placed vertically at growth chamber for 3 d, then the 3-d-old seedlings were transferred to MS plates containing various concentrations of ABA for 7 d before the root length was recorded. (d) Relative root elongation of different lines in the presence of various concentrations of ABA. The data represent the means of three independent experiments ± SD. Significant differences < 0.05 (Student's t-test) in comparison to wild-type (WT): *, < 0.05.

Modulation of AtHAP5A and AtXTH21 did not affect expression of CBFs

In order to dissect the relationship between the AtHAP5A- and AtXTH21-mediated plant freezing stress responses and the CBFs-mediated pathway, the expressions of CBF1, CBF2 and CBF3 were examined when AtHAP5A and AtXTH21 expressions were modulated under control and 4°C stress conditions. However, no significant differences were observed in the expressions of CBF1, CBF2 and CBF3 among WT, AtHAP5A and AtXTH21 overexpressing and knockout/knockdown plants under control and 4°C stress conditions (Fig. S5a–c). Additionally, no appropriate motif responsible for CBFs binding was found in the promoter of AtXTH21 (data not shown). Thus, AtHAP5A and AtXTH21 might be involved in freezing stress resistance independent of the CBFs-mediated pathway.

Discussion

AtHAP5A functions as a transcription factor

Although the functional mechanisms of CCAAT motif binding factors of HAPs in yeast and vertebrates have been well characterized, little is known about the relationship between the CCAAT sequence and plant HAP proteins (McNabb et al., 1995; Edwards et al., 1998; Kato, 2005; Ben-Naim et al., 2006; Wenkel et al., 2006; Yazawa & Kamada, 2007). Yazawa & Kamada (2007) found that AtHAP2/5 were able to interact with C-LEC1 to form heterotrimeric complexes that could bind specifically to DNA fragments containing the CCAAT motif. In the present study, we found that AtHAP5A had transactivation activity in yeast (Fig. 1a). Additionally, transient expression assay in tobacco leaves showed the interaction of AtHAP5A and CCAAT in vivo when 35S-AtHAP5A and CCAAT-m35S-GUS were co-transformed (Fig. 4a,b). Moreover, green fluorescence of 35S::GFP-HAP5A transgenic plants was specifically localized in the nucleus (Fig. 1a). These results suggested that AtHAP5A serves as a novel transcription factor with transactivation activity on CCAAT motif in vivo. Thus, our results provided insights into the dissection of targets of AtHAP5A, and AtHAP5A might be involved in developmental and other processes via directly modulating downstream gene expression.

Involvement of AtHAP5A and AtXTH21 in ABA sensitivity and plant response to freezing stress

Plant HAPs have been widely confirmed to be required for CCAAT binding in yeast and in plants (McNabb et al., 1995; Edwards et al., 1998; Kato, 2005; Yazawa & Kamada, 2007; Maris et al., 2011); however, the direct targets in plants remain uncharacterized. Previous research has focused on their roles in developmental processes (von Besser et al., 2006; Chen et al., 2007; Ballif et al., 2011; Ito et al., 2011; Yu et al., 2011), but information about the involvement of HAPs in abiotic stress is very limited. Recently, Li et al. (2013) found that PwHAP5 positively modulated plant resistance to salinity, osmotic and ABA stresses via regulating stress-related genes. Plant XTHs, which encode xyloglucan endotransglucosylase/hydrolase, is widely known for modulating cell wall xyloglucan content and plant growth (Liu et al., 2007; Miedes et al., 2013). Moreover, constitutive expression of some plant XTHs improved resistance to drought, salt and aluminum stresses (Cho et al., 2006; Zhu et al., 2012; Han et al., 2013). Thus, the in vivo roles of HAPs and XTHs in Arabidopsis' response to abiotic stress need to be further dissected.

In this study, we found that the expressions of AtHAP5A and AtXTH21 were significantly induced by ABA and multiple abiotic stress treatments (Fig. 2c,d), indicating the possible roles of AtHAP5A and AtXTH21 in plant response to abiotic stress. Although AtXTH21 was mainly expressed in root tissue under the control of native promoter (Fig. 2b), constitutive overexpression of this gene using the 35S promoter enabled high expression in leaf tissue (Fig. 5a). The results in this study demonstrated that AtHAP5A and AtXTH21 overexpressing plants were more tolerant to freezing stress and more insensitive to ABA than WT plants, whereas AtHAP5A and AtXTH21 mutants were more sensitive to freezing stress and ABA (Figs 3c–e, 5c–e, 8a–d, S2a–c), which were consistent with the ABA sensitivity of PwHAP5 overexpressing plants demonstrated by Li et al. (2013).

Under cold stress conditions, there are numerous physiological and molecular changes in plants, including a transient increase in the ABA content and ABA-responsive gene expression, ROS metabolism and the expression of CBFs (CBF1, CBF2 and CBF3) (Xiong et al., 2002; Shi et al., 2012b). The ABA pathway plays important roles in cold stress response (Xiong et al., 2002; Qin et al., 2011; Shi et al., 2012b). The increased expression levels of ABA-synthetic genes (ABA2 and NCED3) and ABA-responsive genes (RAB18 and RD22) (Fig. S6) might contribute to the improved freezing resistance in AtHAP5A and AtXTH21 overexpressing plants, at least partially. ROS metabolism including lipid peroxidation, ROS accumulation and underlying antioxidant enzymes is largely consistent with cold stress resistance (Xiang et al., 2007; Ning et al., 2010; Tang et al., 2012). In this study, AtHAP5A and AtXTH21 overexpressing plants displayed significantly lower concentrations of lipid peroxidation, ROS accumulation and higher activities of SOD, CAT and POD than WT plants, whereas AtHAP5A and AtXTH21 mutants exhibited the opposite results (Fig. 6a–f). These results indicated the protective effects of AtHAP5A and AtXTH21 expressions on cold stress-induced ROS metabolism, resulting in less oxidative damage. Although the CBFs-mediated pathway plays critical roles in plant cold stress response (Mäntylä et al., 1995; Chinnusamy et al., 2003; Agarwal et al., 2006; Dong et al., 2006; Jeon et al., 2010), it has been reported that several physiological responses including ROS detoxification system and osmolyte accumulation (Huang et al., 2013) and AtMYB96-mediated freezing stress response via AtLTP3 (Guo et al., 2013) are independent of the CBF pathway. The expressions of three core genes (CBF1, CBF2 and CBF3) in the CBFs-mediated pathway exhibited no significant differences when the expressions of AtHAP5A and AtXTH21 were modulated under control and 4°C stress conditions (Fig. S5a–c); this finding, together with no appropriate motif responsible for CBFs binding being found in the promoter of AtXTH21, suggests that AtHAP5A and AtXTH21 might be involved in cold stress resistance independent of the CBF pathway.

Therefore, AtHAP5A- and AtXTH21-mediated freezing resistance appeared to be partially attributed to the modulation of ABA sensitivity and ROS metabolism but was uncoupled from the CBF pathway.

AtHAP5A acts as positive upstream regulator of AtXTH21 in plant response to ABA sensitivity and freezing stress

Besides the same positive effects on freezing resistance, the relationship between AtHAP5A and AtXTH21 in freezing stress response was further revealed. This study provided genetic evidence showing that AtXTH21 is a direct target of AtHAP5A in the positive regulation of plant response to freezing stress. First, ChIP assay demonstrated that AtHAP5A could bind to five fragments that contained CCAAT motifs in the AtXTH21 promoter (Fig. 4c,d). Second, AtXTH21 expression was upregulated in AtHAP5A overexpressing plants, but was downregulated in hap5a mutants under control and 4°C stress conditions (Fig. 4e,f). Third, consistent with the gene expression results, AtHAP5A and AtXTH2 overexpressing plants were much more tolerant to freezing stress with less ROS accumulation than WT plants, whereas AtHAP5A and AtXTH21 mutants were more sensitive to freezing stress with more ROS accumulation (Fig. 6a,f). Notably, the modulated ABA sensitivity and freezing resistance of AtHAP5A overexpressing plants and knockout mutant could be reversed by the xth21 mutant and AtXTH21 overexpressing plants, respectively (Figs 7, 8). Interestingly, the hap5a × XTH21-OE-19 and HAP5A-OE-12 × xth21 plants reversed the decreased freezing resistance of hap5a plants and improved freezing resistance of HAP5A-OE-12 plants, respectively, but displayed no significant difference in comparison to WT plants, unlike XTH21-OE-19 and HAP5A-OE-12 plants which showed increased freezing resistance (Fig. 7a–c). We speculate that the AtXTH21 may only be one target of AtHAP5A, and other unidentified targets of AtHAP5A may also contribute to AtHAP5A-mediated freezing stress response. Usually, transcriptional regulators have broad target groups. Similarly, AtLTP3 is not the only target of AtMYB96-mediated freezing and drought stress responses (Guo et al., 2013). To date, there are very few examples of the target of a cold-specific transcriptional regulator having a significant effect on tolerance when modulate its expression. The genetic evidence suggests that AtXTH21 as a direct target of AtHAP5A largely contributed to AtHAP5A-mediated freezing stress response. Furthermore, the dissection of other targets of AtHAP5A will shed more light on the in vivo role of AtHAP5A.

Taken together, this study demonstrated that AtHAP5A modulates freezing stress resistance in Arabidopsis through binding to the CCAAT motif of AtXTH21. Cold stress induces the expression of AtHAP5A, and the induced AtHAP5A activates the expression of AtXTH21 via binding to five fragments that contained CCAAT motifs in the AtXTH21 promoter, which in turn confers improved freezing resistance. AtHAP5A and AtXTH21 might be positive regulators of plant freezing resistance independent of the CBF pathway, partly through modulating ABA sensitivity and ROS metabolism.

Acknowledgements

We thank Professor Pingfang Yang for the help in quantitative real-time PCR, and we also thank Professor Jihong Liu and Dr Xiaosan Huang for providing the construct of minimal-100 CaMV 35S-pCAMBIA1391 plastid and Dr Jun You for providing the pCAMBIA1391Z vector. This research was supported by the National Natural Science Foundation of China (Grant no. 31370302) and ‘the Hundred Talents Program', the Knowledge Innovative Key Program of Chinese Academy of Sciences (Grant no. 54Y154761O01076 and Y329631O0263) to Z.C., and by the National Natural Science Foundation of China (Grant no. 31200194), Youth Innovation Promotion Association of Chinese Academy of Sciences and the Outstanding Young Talent Program of Key Laboratory of Plant Germplasm Enhancement and Specialty Agriculture (Grant no. Y452331O03) to H.S.

Ancillary