The bHLH transcription factor CgbHLH001 is a potential interaction partner of CDPK in halophyte Chenopodium glaucum

Plants have evolved different abilities to adapt to the ever-fluctuating environments for sessility. Calcium-dependent protein kinase (CDPK) is believed to play a pivotal role in abiotic stress signaling. So far, study on the specific substrates that CDPK recognized in response to adversity is limited. In the present study, we revealed a potential interaction between CDPK and a bHLH transcription factor under salt stress in Chenopodium glaucum. First, we identified a CgCDPK, which was up-regulated under salt and drought stress; then by Y2H screening, CgCDPK was detected to be involved in interaction with a bHLH TF (named as CgbHLH001), which also positively respond to salt and drought stress. Further computational prediction and experiments including GST-pulldown and BiFC assays revealed that potential interaction existed between CgCDPK and CgbHLH001, and they might interact on the plasma membrane. In addition, CgCDPK-overexpressed transgenic tobacco line could significantly accumulate transcripts of NtbHLH (a homolog of CgbHLH001 in N. tabacum), which provided another evidence of correlation between CgCDPK and CgbHLH001. Our results suggest that CgbHLH001 can interact with CgCDPK in signal transduction pathway in response to abiotic stress, which should provide new evidence for further understanding of the substrate specificity of plant CDPK signaling pathway.


Results
Molecular cloning and characterization of CDPK from halophyte C. glaucum. CgCDPK encodes a subgroup II CDPK. By RT-PCR method, we isolated a coding sequence (CDS) of CDPK gene (shorted as CgCDPK) from C. glaucum. CgCDPK CDS contains a 1 605 bp ORF sequence, which encodes 534 amino acids with a predicted molecular mass of 59.5 kDa and a pI of 6.11. Alignment of the amino acid sequences of CgCDPK and CDPKs from other plants showed that the CgCDPK contains all the four typical domains of CDPK: the N-terminal variable domain, a Ser/Thr kinase domain, an auto-inhibitory junction region and the regulatory calmodulin-like domain (Fig. 1a). In addition, analysis of the amino terminus of CgCDPK with the consensus sequence [Myristoylator (http://us.expasy.org/tools/myristoylator/myristoylator-ref.html) 31 and NMT Predictor (http://mendel.imp.ac.at/myriatate/SUPLpredictor.htm) 32 ] indicates a myristoylation sequence of CgCDPK at 2-GICASKDRDSKEQNGYS-18 region (Fig. 1a), which has been shown to be a signal for plasma membrane localization 33 . Under laser confocal microscope, we observed that red [DiI staining for plasma membrane (PM) marker] and green (CgCDPK-GFP overexpression) fluorescent signals were completely merged into yellow color on the PM of the epidermal cells of tobacco leaf (Fig. 1b), it revealed that CgCDPK located on the PM.
Analysis of the evolutionary relationships between CgCDPK and CDPKs from Arabidopsis, rice, maize, tobacco, tomato, etc. showed that CgCDPK belongs to subgroup II (Fig. 1c) out of four subgroups of CDPKs divided previously 34 . CgCDPK shared relatively higher similarity of amino acid sequence to BvCDPK from Beta vulgaris (94.12%) and NtCDPK1 (80.13%) from Nicotiana tabacum.
CgCDPK enhanced the positive response to salt and drought stress. After being exposed to varying concentrations of NaCl and PEG, C. glaucum plant was analyzed with the response of CgCDPK at transcriptional level (Fig. 2a). It was found that salt stress significantly enhanced the accumulation of CgCDPK transcripts shortly after treatment, it was about 8-fold greater at 0.5 h under different NaCl concentrations than that of the control (P < 0.001), after then (from 1 h to 48 h) which fell down sharply (except for 300 mmol/L at 1 h), but still significantly higher than the control ( Fig. 2a; left). For PEG treatment, the transcriptional level of CgCDPK increased much higher at 0.5 h for all PEG concentrations (P < 0.0001) and then fell down, which was similar to that of the NaCl treatment within a short time; however, the expression of CgCDPK rose again and reached to the maximum (about 30-fold more than control) at 5 h (except for 20%) (P < 0.0001) ( Fig. 2a; right), which was different with that of the NaCl stress. In order to further explore the function of CgCDPK in response to abiotic stress, we examined seed germination behavior of CgCDPK-overexpressed transgenic tobacco lines under salt and drought stress. Results showed that transgenic lines were less sensitive to NaCl or PEG in germination than NT plants, especially OE1 and OE3 at higher concentrations (Fig. 2b,c). In combination of these results, it suggests that CgCDPK gene can positively and quickly respond to abiotic stress.
Screening and verification of CDPK interaction components by yeast two-hybrid system. Construction of cDNA library in C. glaucum under salt stress. Total RNA was isolated from leaves of C. glaucum under 300 mmol/L NaCl treatment with the ratio of A 260 /A 280 = 2.05, and mRNA was properly purified with a smear distributed between 500 bp and 5000 bp. Then the cDNA library was constructed with a total colony-forming unit (CFU) as 1.26 × 10 7 , which meets the requirement of cDNA library construction. Twenty singular colonies from one plate were randomly selected to check the insert fragment distribution by PCR test, which suggests a 100% recombinant frequency, and the fragment size was distributed from 850 bp to 2000 bp ( Supplementary Fig. 1).
SCIEnTIfIC RePoRTs | 7:8441 | DOI:10.1038/s41598-017-06706-x Screening and verification of CDPK interaction components by split-ubiquitin based membrane yeast two-hybrid analysis. To identify the components that potentially interact with CgCDPK, we performed a split-ubiquitin based membrane yeast two-hybrid screening of a cDNA library of C. glaucum using CDPK as bait and obtained 51 clones that encode 32 potential candidate proteins (Fig. 3a), which were classified as transcription factors (TFs) and proteins with predicted functions in protein synthesis, photosynthetic pathway, stress tolerance and metabolism.
Subsequently, the positive clones were further screened by the reporter gene -His, Ade expression on SD/-TLHA medium for the first (Fig. 3b) and the second (Fig. 3c) verification; meantime, another reporter gene LacZ expression was detected by measuring β-galactosidase activity (Fig. 3d), finally we got 12 candidate clones which showed potential interaction with CgCDPK. Self-activation detection revealed that two candidate clones showed similar reaction with negative control (Fig. 3e). Based on above analysis, we chose No.3 clone which encodes a bHLH transcription factor (named as CgbHLH001) for the following study.  amino acid residues with a predicted molecular weight of 28.60 kDa and pI of 6.80. Structural analysis (http:// www.ebi.ac.uk/interpro/; Iterative Threading ASSEmbly Refinement) 35 suggests that the secondary structure of CgbHLH001 presents a typical MYC-type bHLH domain (IPR011598) and a coiled coil region (Fig. 4a). Furthermore, phylogenetic analysis suggests that plant bHLH proteins are monophyletic and constitute 26 subfamilies 36 . Seven well-characterized bHLHs in subfamily XII are listed in Fig. 4b, in which CgbHLH001 showed the highest homology to BvbHLH79, BvBPE of Beta vulgaris subsp. vulgaris, and to AtbHLH031 and AtbHLH079 of Arabidopsis thaliana. In addition, the amino acid sequence revealed a putative nuclear localization signal (NLS) within bHLH domain of CgbHLH001. Further verification by transgenic tobacco of P35S::CgbHLH001-GFP showed that strong green fluorescence was exclusively localized to the nucleus, which may support a role for CgbHLH001 as nuclear transcription factor (Fig. 4c).

Characterization of CDPK interaction component-CgbHLH001 in
CgbHLH001 expression is induced by abiotic stress. Transcriptional expression of CgbHLH001 in C. glaucum was analyzed after exposure to salt and drought stress. Under NaCl or PEG treatment, CgbHLH001 was upregulated with time increasing, the highest expression level was from 5 h to 12 h (for NaCl or PEG, P < 0.0001) (Fig. 5a), the higher the concentration was, the earlier the highest expression value was presented. CgbHLH001 shared similar transcriptional pattern under NaCl and PEG treatments, but its response to the former was much greater than that of the latter. These data indicate that CgbHLH001 can positively respond to salt or drought stress. To make further understanding of gene function in response to stress, the ectopic expression of CgbHLH001 in E. coli was analyzed. The growth performance of E. coli strain harboring with recombinant plasmid pET-28a-CgbHLH001 was examined under various stresses, i.e. different concentrations of NaCl, PEG, methyl viologen, different range of pH, and low temperature (−20 °C). Compared to the control strain, the recombinant strain could tolerate much broader range of different stresses (Fig. 5b) and grow much better (Fig. 5c), especially under 400 mmol/L NaCl, pH 9 and −20 °C treatment. The above data indicate that CgbHLH001 overexpression in prokaryote is able to improve the stress tolerance of the recombinant strain.

Analysis of interaction between CDPK and bHLH in C. glaucum. Prediction of interaction between
CgCDPK and CgbHLH001. Prediction of 3D structure of CgCDPK, CgbHLH001 and the probable complexes of CgCDPK-CgbHLH001 interaction based on the sequences revealed different models in Fig. 6, which showed that CgbHLH001 had a typical basic helix-loop-helix conserved domain (Fig. 6a), and CgCDPK had the structure with an N-terminal region, a protein kinase domain and several EF-hand domains (Fig. 6b). The three top ranking models of the CgCDPK-CgbHLH001 interaction complex were presented in Fig. 5c, which indicate the possible interaction sites between helix-loop-helix domain of CgbHLH001 and protein kinase domain or the

GST-pulldown verification of protein interaction in vitro.
In the present study, His-tagged CgbHLH001 was immobilized on Ni column and assayed for the ability to pull down the GST-CgCDPK fusion protein (Fig. 7a,b). Pulldown results were analyzed by immunoblotting with anti-GST antibody (Fig. 7c). As shown in the first lane of BiFC assay detects protein interaction in vivo. To further characterize interaction between CgCDPK and CgbHLH001 in vivo, bimolecular fluorescence complementation assay was performed by co-infiltration of recombinant strain combination − 35S::CgCDPK-nYFP + 35S::CgbHLH001-cYFP or 35S::CgCDPK-cYFP + 35S ::CgbHLH001-nYFP into N. benthamiana fresh leaf to observe the generation of fluorescence. As a result, a strong yellow fluorescent signal was observed on the PM of the epidermal cells when either of the above combination was delivered into the tobacco plant, compared to no fluorescent signal in cells with any of other combinations (Fig. 8). Meantime, the tobacco leaves were simultaneously treated with a PM fluorescent dye, which further verified that CgbHLH001 potentially interacted with CgCDPK on the PM.  Results showed that the transcripts of NtbHLH was significantly accumulated under NaCl or PEG treatment from 2 h to 24 h tested, the highest expression level was observed around 2-5 h with more than 2.0-fold in CgCDPK-overexpressed line than that of the NT plant (for NaCl or PEG, P < 0.0001) (Fig. 9a,b), which suggests that CgCDPK overexpression can induce expression of NtbHLH -a homolog of CgbHLH001 in N. tabacum under salt or drought stress.

Discussion
Calcium-dependent protein kinases (CDPKs) are key stress sensors and signal transducers of calcium signaling pathway in plants, which play important roles in response to environmental stimuli 37 . Identification of the interaction components is an efficient way to reveal CDPK functions 38 . So far, limited reports on the interaction between CDPK and its substrates have been documented. In the present study, we identified an interaction component of CDPK -a basic helix-loop-helix (bHLH) transcription factor (TF) from C. glaucum by yeast two hybrid (Y2H) screening, further in vitro pulldown and in vivo BiFC assay as well as transgenic plant verification all provided evidence that CgCDPK and CgbHLH001 could positively respond to abiotic stress and potentially interact in transduction of signal to enhance stress tolerance. In the present study, a subfamily II CgCDPK was identified from C. glaucum and was observed to be localized on the PM, it is consistent with the prediction that a myristoylation (with a Gly residue at position 2) and a palmitoylation (with a Cys residue at position 4) sites are harbored in the N-terminal sequence -MGICASKDRDSKEQNGYS of CgCDPK, which is associated to membrane localization 39,40 . It has been reported that protein myristoylation can also promote protein-membrane or protein-protein interaction, which may occur co-translationally and is irreversible 41,42 ; afterwards the reversible palmitoylation of the protein may further stabilize or regulate the interaction and enable CDPK to shuttle between membrane and the cytosol or nucleus in response to stress signals 40,43 , which has experimentally been verified with some CDPKs in Arabidopsis, rice, and other flowering plant species 43,44 .
CDPKs have long been considered to be involved in various abiotic stresses, e.g. AtCPK3, AtCPK4, AtCPK11, AtCPK23, AtCPK27 (subfamily II) in Arabidopsis 4, 16, 45 . By contrast, AtCPK12, the closest homolog of AtCPK4/ AtCPK11, plays the opposite role in stress tolerance 21 , which suggests that the commitment of the CDPK functions may depend on the downstream interaction components. In the present study, the positive responses of CgCDPK to salt and drought in C. glaucum were in line with AtCPK4 and AtCPK11 in Arabidopsis, and ZmCPK4 in maize 18,46 . The quick activation of CgCDPK transcription under NaCl and PEG treatment may suggest that this kind of kinase in the upstream responds stress earlier. Compared to the reported transcriptional level of ZmCPK4 after 3 h PEG treatment, that of CgCDPK at 5 h in the present study was a much greater increase, the details still remain to be explained by further experiments. In addition to the active response of CgCDPK to abiotic stress, seed germination of CgCDPK-overexpressed transgenic tobacco line was much less sensitive to salt and drought stress. All these data indicate that CgCDPK should be important in stress tolerance.
CDPKs are major signaling molecules that are involved in a variety of stress responses, however, the molecular mechanisms in the interaction between CDPK and the specific substrate are still largely unknown 11 . So far, a limited number of transcription factors or proteins have been characterized as the CDPK interaction components, e.g. StCDPK5 (Solanum tuberosum) and NtCDPK1 (Nicotiana tabacum), two homologs of CgCDPK in the present study, are dominantly localized to the PM, the former can activate StRBOHB (an NADPH oxidase) on the PM by phosphorylating the N-terminal region 14 ; the N-variable domain of NtCDPK1 can interact with a bZIP TF-RSG (repression of shoot growth) in a Ca 2+ -dependent manner and specifically phosphorylates Ser 114 of RSG 11,47 . In the present study, we identified a potential substrate of CgCDPK, a bHLH TF -CgbHLH001, by Y2H screening, GST-pulldown and BiFC assay. The predicted interaction complexes revealed that the N-variable domain of CgCDPK could interact with CgbHLH001 48 . A phosphorylation site in 61 -GKRLKS-66 (Ø-X-R/K-X-X-S, Ø is a hydrophobic residue and X is any residue) 49 out of thirty-nine predicted ones of CgbHLH001 is likely to be the specific action site of CDPK, which agreed with the predicted interaction pattern at N-variable domain. A tobacco CDPK1 (located on the PM) 47 was reported to interact with and phosphorylate RSG -a leucine zipper transcription factor, finally make it translocate from cytoplasm to nucleus and regulate the target gene expression 50 . In combination with the PM localization of CgCDPK and the potential interaction between CgbHLH001 and CgCDPK on the PM revealed by BiFC assay in the present study, we speculate that these two components should interact on the PM before CgbHLH001 can make further actions. In addition, our investigation with the transgenic tobacco line showed that overexpression of CgCDPK significantly accumulated the transcripts of NtbHLH (a homolog of CgbHLH001 in N. tabacum) under stress conditions, which may provide further evidence in vivo for the relationship existed between CgCDPK and CgbHLH001. So far, few reports have been documented on bHLH TF as the substrate of plant CDPKs. Our findings on the interaction between CgbHLH001 and CgCDPK should be an important contribution to this field. bHLH proteins are the second largest family of plant TFs which are classified into 26 subgroups by phylogenetic analysis 36 . They can function as activators of one set of genes and repressors of others 51 . In Arabidopsis or rice genome, 162 or 167 bHLH proteins have been identified, and about 30% of which have been functionally characterized, the role includes regulatory networks of plant growth, development and stress responses 52, 53 , however, only a few members have been shown to be involved in stress tolerance. In the present study, phylogenetic analysis revealed that CgbHLH001 belongs to typical XII subgroup of bHLH TF family, and the accumulation of CgbHLH001 transcripts under salt or drought stress was consistent with the performance of bHLH genes observed in Arabidopsis, rice, soybean, etc 54,55 . The best-studied bHLH TFs are members of subgroup III, which are largely involved in responses to abiotic stress, e.g. AtMYC2, AtbHLH17, AtbHLH92 and AtICE1 in Arabidopsis, OrbHLH001, OrbHLH002 in wild rice were shown to be upregulated by drought, salinity or cold stress [56][57][58][59][60][61] . So far, however, much limited reports on bHLH TF in XII subgroup have been recorded in stress tolerance. Our results indicate that ectopic expression of CgbHLH001 could confer E. coli strain with enhanced stress tolerance, which means that helix-loop-helix in CgbHLH001 may have similar function with helix-turn-helix in prokaryote in regulation of relevant gene expression 62 . Our preliminary investigation on CgbHLH001 may provide an evidence for understanding the role of bHLH TFs in XII subgroup in stress responses.

Figure 8. BiFC analysis of interaction between CgCDPK and CgbHLH001 in vivo.
The C-terminal half and the N-terminal half of YFP were fused to CgCDPK (CgCDPK-cYFP and CgCDPK-nYFP) and CgbHLH001 (CgbHLH001-cYFP and CgbHLH001-nYFP). YFP fluorescence was detected in N. benthamiana fresh leaves co-infiltrated with combinations of CgCDPK-cYFP + CgbHLH001-nYFP and CgCDPK-nYFP + CgbHLH001-cYFP. Combinations between any above fused construct and non-fused cYFP or nYFP, as well as cYFP + nYFP were used as controls. The plasma membrane was marked by DiI staining. Fluorescence was visualized by confocal microscope. The experiment was repeated two times with similar results. Bars = 50 µm.
SCIEnTIfIC RePoRTs | 7:8441 | DOI:10.1038/s41598-017-06706-x In conclusion, we showed that a potential interaction existed between CgCDPK and CgbHLH001 under stress conditions, which was supported by in vitro pulldown and in vivo BiFC assays, active expression patterns of CgCDPK and CgbHLH001 under stress, and the behavior of NtbHLH (a homolog of CgbHLH001) in CgCDPK-overexpressed transgenic tobacco line. Based on our data, a possible model for the interaction between CgCDPK and CgbHLH001 in response to salt and drought stress is proposed (Fig. 10). Plant may sense salt and drought through osmotic stress (or ABA signaling) to activate CgCDPK, which may in turn interact with CgbHLH001 (or other TFs) on the PM, and finally might enter the nucleus to regulate the target gene expression.   For analysis of the expression pattern of CgCDPK and CgbHLH001 in C. glaucum, or NtbHLH in CgCDPK-overexpressed tobacco transgenic line, the specific primers of each gene were used in qPCR ( Table 1). The relative amplification of β-actin of C. glaucum or β-actin of tobacco was used for normalization. The above amplification was performed in the following conditions: 95 °C 2 min followed by 40 cycles of 95 °C 5 s, 60 °C 30 s. qPCR was performed with QuantiNova SYBR Green PCR Kit (Cat. 208054; Qiagen, Germany) and ABI 7500 Real time PCR system (Applied Biosystem, USA). Relative quantification of specific mRNA level was calculated using the cycle threshold (Ct) 2 −∆∆Ct method 63 . Four samples (biological replicate) of each treatment were duplicated (technical replicate) in qPCR experiment. The final value of relative quantification was described as fold change of gene expression in the test sample compared to control.

Methods
Analysis of phylogenetic relationship. The phylogenetic relationship of CDPK or bHLH transcription factors among various plant species was analyzed by using protein sequences from C. glaucum, Arabidopsis thaliana, Glycine max, Zea mays, Oryza sativa, etc. Multiple alignments were conducted with full-length amino acid sequence by Clustal W tool of MEGA 6.06. The phylogenetic tree was constructed through neighbor-joining (NJ) method of MEGA 6.06. Molecular distance of the aligned sequences was calculated according to the p-distance parameter, the gaps and missing data were treated as pairwise deletions. Branch points were tested by bootstrap with 1000 replicates. All amino acid sequences used in the present study were acquired from GenBank and Phytozome database.  transformed into Agrobacterium tumefaciens strain EHA105. Leaves from 5-6 week-old tobacco (N. tabacum; NC89) were transformed with the recombinant A. tumefaciens strain via the leaf disk method 64 to generate the transgenic tobacco lines.

Inspection of GFP signal under confocal microscope.
To visualize the subcellular localization of CgCDPK or CgbHLH001, seeds of transgenic tobacco lines (35S::CgCDPK-GFP or 35S::CgbHLH001-GFP) were surface-sterilized and sown on MS medium containing 300 mg/L Kanamycin for two weeks, the survival seedlings were subjected to PCR identification, and the small fresh leaves of PCR positive seedlings (T1) were chose to treat with PM fluorescent marker -DiI (1,1′-dioctadecyl-3,3,3′,3′-tetramethylindocarbocyanine perchlorate) (Solarbio, Shanghai, China). The seedling leaves were gently scratched with scalpel and then stained in working solution (stock solution was prepared in ethanol at 5 mmol/L, and diluted into 10 µmol/L with PBS buffer for ready staining) of PM marker for 15 min at 37 °C. After washing for three times with distilled water, the leaves were inspected with fluorescent signal under confocal microscope (Zeiss LSM 800, Carl Zeiss, Jena, Germany). Seedlings with 35S::GFP was used as the positive control. All images were visualized and acquired with the ZEN Imaging Software (Version 2.3).

Split-ubiquitin based membrane yeast two-hybrid analysis. Construction of cDNA library of C. glau-
cum under 300 mmol/L NaCl treatment. First and double-strand (ds) cDNA were synthesized according to the protocol of EasyClone cDNA Library Construction Kit (Cat. P01010; Dualsystem Biotech, Switzerland). dscDNA synthesis was visualized by 1% agarose gel and then completely digested with Sfi I restriction endonuclease, cDNA fragments longer than 1000 bp were recovered from a low melting agarose gel, which were then ligated into Sfi I digested pPR3-N yeast expression vector, and the ligation mixture was transformed into the competent cells of E. coli by chemical method 65 . For CFU (colony forming unit) calculation of the library, 10 µL of the transformed cell culture (with brief cultivation) were diluted 100 times and spread on LB medium with ampicillin and cultivated overnight at 37 °C. When the colonies became clearly visible, CFU was calculated as follows: colonies/10 µL medium × 100 times × total volume of the library (µL). For recombination frequency calculation of the library, 20 randomly selected colonies were examined with the inserted fragments by PCR amplification (using the primers of pPR3-N and pPR3-C) and subjected to agarose gel electrophoresis, the size distribution of inserted fragments was evaluated and the recombination frequency of the library was calculated as follows: colonies with inserted fragments/randomly selected colonies number × 100%.
Detection of interaction between bait and preys. The DUALmembrane starter kit SUC (Cat. P01301-P01329; Dualsystems Biotech, Switzerland), a split-ubiquitin based membrane yeast two-hybrid (MYTH) system, was employed in screening the interaction components. For bait-vector construction, the coding sequence of CgCDPK containing Sfi I restriction sites on both ends was cloned into the pBT3-SUC yeast expression vector to yield pBT3-SUC-CgCDPK plasmid in which CgCDPK bait gene was fused to the ORF of Cub-LexA-VP16. To verify the expression and self-activation ability of bait on the reporter genes, the bait plasmid and the positive pNubG-Fe65 or negative pPR3-N control preys were co-transfected into the NMY32 yeast strain and cultured successively on the dual, triple and quadruple synthetic dropout nutrient medium (SD/-Trp-Leu, SD/-Trp-Leu-His and SD/-Trp-Leu-His-Ade; shorted as SD/-TL, SD/-TLH, SD/-TLHA). cDNA library of C. glaucum under 300 mmol/L NaCl treatment was constructed with pPR3-N yeast expression vector as above. pPR3-N-CgcDNA plasmid (25 μg) was transformed into the NMY32 yeast strain containing pBT3-SUC-CgCDPK, which was then screened on SD/-TLH medium with 5 mmol/L 3-amino-1,2,4-triazole (3-AT), a histidine analog and competitive inhibitor of the His3 gene product. The positive colonies were further screened by cultivating on SD-TLHA/X-α-gal medium, and measured with β-galactosidase activity (HTX β-galactosidase assay kit, Cat. P01002; Dualsystems Biotech, Switzerland) to test the expression of the reporter gene LacZ. The blue colonies were identified and the prey plasmids in which were then re-transfected into NMY32 yeast strain containing pBT3-SUC-CgCDPK plasmid for a second-round screening on SD-TLHA/X-α-gal medium. The yield positive colonies were analyzed with the inserted cDNA sequences, which were then compared to GenBank database by using BLAST program available in the NCBI (National Center for Biotechnology Information) and analyzed with the possible functions.

Detection of interactions in vivo.
Recombinant A. tumefaciens strains (GV3101) containing different constructs were incubated, harvested and resuspended in infiltration buffer (10 mmol/L MES, 0.2 mmol/L acetosyringone and 10 mmol/L MgCl 2 ) at a final concentration of OD 600 = 0.4, then allowed to stand at room temperature for 1-3 h. Equal amounts of the Agrobacterium suspension of each construct were mixed into a new 1.5 mL tube and vortexed for 10 sec to be ready for use. Five-to six-week-old N. benthamiana plants were prepared for infiltration. Placed the tip end of the syringe (without needle) against the underside of the leaf (avoiding the veins) by supporting with one finger on the upperside, then gently pressed the syringe to infiltrate the Agrobacterium mixture into the fresh leaf, followed by infiltration of DiI fluorescent PM marker (10 µmol/L working solution) nearby and labeled the infiltration area for future recognition. Treated plants were kept in darkness overnight, and then transferred to normal growth conditions for 48 h. The YFP and DiI fluorescent signals in the leaf of N. benthamiana were examined under the confocal microscope (Zeiss LSM 800, Jena, Germany) with excitation at 514 and 549 nm, respectively.
Statistical analysis. All data were analyzed by Microsoft Excel 2010 and GraphPad Prism 5.0 (GraphPad Software, San Diego, USA). For qPCR, three biological replicates with two technical replicates of each treatment were measured. For seed germination, four biological replicates with 30 seeds of each were applied. For E. coli stress treatment, three biological replicates of each treatment were assayed. Two-way ANOVA was used to test the significance of main effects. Differences were measured by Post Hoc Duncan multiple comparison test at 0.05, 0.01 or 0.001 significance level.