Plant materials and growth conditions
Seedlings of T. hispida were grown in a greenhouse under controlled conditions of 70% relative humidity, light/dark cycles of 14/10 h, and were maintained at 24°C. To induce abiotic stresses, the seedlings were at 4°C or watered on their roots with a solution of 0.4 M NaCl, 20% (w/v) PEG6000, 100 μM ABA or 50 μM MV for 3, 6, 9, 12 and 24 h; the seedlings watered with water were harvested at the corresponding time points as controls.
The ORF of ThbZIP1 was inserted into pROKII driven by CaMV35S promoter, designated as pROKII-ThbZIP1 (see Additional file 8: Table S5 for primers used). The pROKII-ThbZIP1 was transformed into Arabidopsis (ecotype Columbia) using the floral dip method.
Cloning and activity analysis of the ThbZIP1 promoter
Based on the sequence of ThbZIP1 (GenBank number: FJ752700), the promoter of ThbZIP1 (1,571 bp in length) was cloned using a Genome Walking Kit (Takara, Dalian, China). The potential cis-regulatory elements in the promoter were predicted using the software PLACE . The 35S promoter in pCAMBIA1301 was replaced with the ThbZIP1 promoter that fused with the 5′ UTR of ThbZIP1 to drive the β-glucuronidase (GUS) gene (ProThbZIP1::GUS; Additional file 1: Figure S1A). The ProThbZIP1::GUS construct was transferred into Arabidopsis plants by the floral dip method. The T3 seedlings were used for promoter activity analysis.
Analysis of the upstream regulators of ThbZIP1
Previously, eight transcriptomes from roots of T. hispida treated with NaHCO3 for 0, 12, 24 and 48 h (two biological replicates were set at each time point) were built . In total, 47,324 unigenes were generated after these transcriptomes de novo assembly using SOAPdenovo. The TFs from different families were identified, PCR amplified and cloned into pGADT7-Rec2 (Clontech, Palo Alto, CA, USA) to form a cDNA library (designed as TFs library) for yeast one-hybrid assay. There are seven E-box motifs (“CANNTG”) in the promoter of ThbZIP1. Three tandem copies of E-box motif were cloned into a pHIS2 vector (designed as pHIS2-E-box; see Additional file 8: Table S6 for primers used), and were screened with a TFs library for yeast one-hybrid assay (Clontech, Palo Alto, CA, USA). The interactions of pHIS2-p53 (three tandem copies of the cis-acting DNA consensus sequence in pHIS2, which is recognized by p53) with the tested TFs were used as negative controls.
Two MYC (ThMYC4 and ThMYC6 GenBank number: JN166788 and JN166790) were identified to bind to the E-box motif, of which ThMYC6 that bound with greater strength to the E-box was used for further study. The E-box core motif of “CANNTG” was mutated to “ACAATG”, “CAAACA”, and “ACCGCA” (designed as pHIS2-E-M1, -E-M2, -E-M3, respectively; see Additional file 8: Table S6 for primers used). The interactions of ThMYC6 with E-box motif and its corresponding mutated motifs were studied using the yeast one-hybrid analysis.
To determine if ThMYC6 is able to activate the expression of ThbZIP1 by interacting with the E-box motifs, the pHIS2 constructs that harbored the promoter fragments of ThbZIP1 which contained E-boxes (pHIS2-ProE(+)), or without E-boxes (pHIS2-ProE(-)) were respectively generated as reporter vectors (see Additional file 8: Table S6 for primers used). The interactions of these constructs with ThMYC6 were studied using yeast one-hybrid analysis.
To further verify these interactions, the three tandem copies of the E-box and its mutant E-M3 (ACCGCA) were respectively fused to the minimal 35S promoter (-46 to + 1) to drive GUS, and designed as pCAM-E-box and pCAM-E-M3 (see Additional file 8: Table S7 for primers used). The promoter fragments of ThbZIP1, which contained E-box motifs (named as pCAM-E-boxp+), and lacked E-box motifs (pCAM-E-boxp-) (Figure 1C) were respectively fused to the minimal 35S promoter to drive GUS as reporter vectors (see Additional file 8: Table S7 for primers used). The effector vector was constructed by cloning the full ORF of ThMYC6 into pROKII driven by the 35S promoter (named as pROKII-ThMYC6). Both of the reporter vectors and their corresponding effector vectors were co-transformed into tobacco leaves using the particle bombardment. The transformation of pCAMBIA1301 alone (CaMV35S) was used as positive control. The transformation of the reporter plasmids alone or effector plasmids alone was used as negative controls. All assays were repeated three times. GUS histochemical staining assay was performed as described by Jefferson , and the GUS stained leaves were scanned by using scanner (D4800,UNISPLENDOUR, China). GUS activity levels were determined according to Jefferson .
Real-time PCR analysis of gene expression
The real-time RT-PCR was performed using α-tubulin (XM_002301092) and actin 3 (XM_002308329) as internal controls (see Additional file 8: Table S8 for primers used). PCR was performed on a MJ Research OpticonTM2 instrument with the following conditions: 94°C for 30 s, 45 cycles of 94°C for 12 s, 58°C for 30 s, 72°C for 40 s, and 80°C for 1 s for a plate reading. The relative expression levels of the products were calculated according to the 2-ΔΔCt method . Relative gene expression levels were calculated as the transcription level under stress treatment divided by the transcription level of the controls (i.e., samples from plants grown under normal conditions and harvested at the same treatment time points).
Assay of ThbZIP1 bindings to C-box, G-box and A-box motifs
Three tandem copies of C-box, G-box and A-box, together with their mutants, CM1: CAAGTG, CM2: CAAAAG, CM3: CACATG and CM4: CACGAG, were cloned into a pHIS2 vector (see Additional file 8: Table S6 for primers used), respectively. Yeast one-hybrid screening analysis was performed to study their interactions with ThbZIP1. The yeast cells were grown on selective dropout media: SD/- Trp-Leu/-His (TDO) + 3-AT (3-AT concentration from 30 to 60 mM). Three tandem copies of the C-box, G-box and A-box and the mutant sequence CM2 (CAAAAG) were fused to the minimal 35S promoter (-46 to +1) for driving GUS (constructs containing C-box, G-box, A-box and mutant sequence CM2 named as pCAM-C, pCAM-G, pCAM-A, and pCAM-CM2, respectively). The effector vector was constructed by cloning the ORF of ThbZIP1 into pROKII driven by the 35S promoter (pROKII-ThbZIP1) (see Additional file 8: Table S7 for primers used). Both the reporter and effector vectors were co-transformed into tobacco leaves using particle bombardment. GUS staining and GUS activity assay were determined as above.
Analysis of ABA, salt and drought stress tolerances
The T3 generation of ThbZIP1 transgenic plants were used in ABA, salt and drought stress tolerance tests. The seeds were sown on MS medium for three days and germinated seeds were transferred into a 1/2 MS medium plus 0, 1, 2, 5, and 10 μM ABA; or 0, 50, 100, 150 and 200 mM NaCl; or 0, 50, 100, 150 and 200 mM Mannitol for two weeks, respectively. The root length and fresh weight were measured. The seeds were sown on a 1/2 MS medium plus 0, 1, 2, 5, or 10 μM ABA, or 0, 50, 100, 150 or 200 mM NaCl, or 0, 50, 100, 150 or 200 mM mannitol for one week, and the germination rates of each transgenic plant or Col-0 plant were measured.
Microarray experiments and data analysis
The four-week-old seedlings of Col-0 and ThbZIP1 transgenic plants without treatment or subjected to 150 mM NaCl for 3 h were used for the microarray analyses and three independent biological replications were performed. The Agilent Arabidopsis Oligo microarrays were employed. A Welch’s t-test was used for the parametric test, and the Benjamini and Hochberg false discovery rate for multiple testing corrections was used with a P-value of < 0.05 to filter and identify reliable genes. All genes that were considered to show significant expression level differences by these tests were then filtered by a fold change >2.0. For verification of microarray results, 24 differentially expressed genes identified by microarray were randomly selected for real-time RT-PCR analyses.
Searching for ThbZIP1-binding sequences in gene promoters
The genes up-regulated by ThbZIP1 under normal growth or salt stress conditions were randomly selected for ThbZIP1-binding sequences in their promoter regions. The promoter sequences (from -1 to -900) of these genes were derived from TAIR database (http://www.arabidopsis.org/). For identification of ThbZIP1-binding motifs, the sequences of C-, G- and A-box were searched in the promoter regions of these genes.
Detection of ROS and cell death
Arabidopsis leaves from the two transgenic lines and Col-0 subjected to the ABA, NaCl or Mannitol treatments were infiltrated with 3, 30-diaminobenzidine (DAB) solutions or nitroblue tetrazolium (NBT) following the procedures described by Zhang et al.. Cell death was examined by Evans blue staining as described by Kim et al.. ROS production in intact guard cells and root tips were detected using 2, 7-dichlorofluorescin diacetate (H2DCF-DA) as described by Pei et al.. H2O2 levels and GST activity were measured according to Thordal-Christensen et al. and Terada et al.. Five-day-old seedlings were transferred into MS medium or MS medium with l25 mM NaCl, 10 μM ABA and placed vertically. After stress for 24 h, at least 9 seedlings of each line were incubated with 1 mg/mL PI (Invitrogen) for 20 min. The root tips of stained seedlings were visualized by LSM710 microscope (Zeiss, Jena, Germany) with excitation at 488 nm and emission at 516 nm, respectively.
Measurement of electrolyte leakage and water loss rates
Electrolyte leakage was measured according to Liu et al.. For water loss rates measurements, leaves were detached and weighed immediately (fresh weight, FW), and were then left on the laboratory bench (humidity, 45–50%, 20–22°C) and weighed at designated time intervals (desiccated weights). Leaves were finally oven-dried for 24 h at 80°C to a constant dry weight (DW). Water contents (WC) were measured according to the formula: WC (%) = (desiccated weight – DW)/(FW – DW) × 100.
Statistical analyses were carried out using SPSS 16.0 (SPSSInc, Chicago, III, USA) software. Data were compared using Student’s t-test. Differences were considered to be significant if P < 0.05. ** represented 0.001 < P < 0.01 and * represented 0.01 < P < 0.05.