Introduction

Plant responses to abiotic and biotic stresses involve changes at the transcriptome, cellular and physiological levels. The interaction between biotic and abiotic stress is orchestrated by hormone signaling pathways (Atkinson and Urwin 2012). Responses to stress require the production of important functional proteins, such as those involved in the synthesis of osmoprotectants, and regulatory proteins, kinases, and transcription factors (TFs), operating in the signal transduction pathways (Saibo et al. 2009). Several different techniques can be used to study the transcriptome during multiple stress responses. Affymetric ATH1 microarrays are commonly used to investigate universal components of the plant’s response to different stress conditions (Mongkolsiriwatana et al. 2009; Swindell 2006). Quantitative trait analysis can be used for the identification of useful regions of genomes (Ashraf 2010), and next-generation high-throughput sequencing (Quail et al. 2012) offers whole plant transcriptome surveys (Wang et al. 2010), but gene functional analyses are still needed to study plant development and gain an understanding of responses to biotic and abiotic stress conditions.

Transcription factors are DNA-binding proteins that activate or repress transcription of downstream genes by binding to a consensus sequence in their promoters. The NAC TF family is one of the largest TF families in plants, with more than 100 members identified in both Arabidopsis and rice (Wang and Dane 2013) and 80 in Citrullus lanatus (Guo et al. 2013). The acronym NAC originates from the no apical meristem (NAM), Arabidopsis transcription activation factor 1 and 2 (ATAF), and cup-shaped cotyledon (CUC) genes. NAC proteins typically share a well conserved N-terminal NAC domain, which is divided into five conserved subdomains (A–E), and a diversified C-terminal transcription regulatory domain (Puranik et al. 2012). NAC TFs are specific to plants and associated with many biological functions during embryonic, floral and vegetative development, and stress-related processes (Olsen et al. 2005; Atkinson and Urwin 2012). A number of NAC proteins interact with pathogens, the hormones abscisic acid (ABA), JA, and salicylic acid (SA) and exhibit interactions with both biotic and abiotic stress responses. Signaling crosstalk among phytohormones in NAC-associated pathways regulate the protective responses in plants via synergistic or antagonistic actions (Tuteja and Sopory 2008). Different phytohormones, which can bind to the NAC recognition sequence, can regulate different NACs and further regulate stress-related genes. Many NAC genes are associated with stress, and some of the NAC genes have multiple functions. OsNAC6 in rice, for example, is involved in both abiotic and biotic stresses (Nakashima et al. 2007). CsNAC1 is induced by salt stress, cold and ABA (Oliveira et al. 2011). Overexpression of ONAC045 results in enhanced drought and salt tolerance (Zheng et al. 2009), while RD26 was induced by drought, ABA, and high salinity (Fujita et al. 2004). NACs have long been associated with stress signaling, and recent discoveries suggest that they may make excellent targets for improving broad-spectrum tolerance in crops through genetic engineering (Nakashima et al. 2007; Xu et al. 2011). Although quite a few NACs have been functionally characterized primarily in model plants like Arabidopsis, the functions of the majority of the members of the large NAC gene family remain unknown (Hu et al. 2010).

Citrullus colocynthis (L.) Schrad, closely related to domesticated watermelon (C. lanatus var. lanatus), is a non-hardy drought-tolerant perennial herbaceous species in the Cucurbitaceae family (Jeffrey 2008; Al-Zahrani and Al-Amer 2006). It can survive arid environments by maintaining its water content without wilting of the leaves or desiccation under severe stress conditions. C. colocynthis has a rich history as an important medicinal plant and as a source of valuable oil (Dane et al. 2006). Its seeds appear in several early Egyptian, Libyan, and Near Eastern sites from about 4000 BC (Zohary and Hopf 2000). The species grows in sandy areas throughout northern Africa, southwestern Asia, and the Mediterrranean region (Zamir et al. 1984; Burkill 1985; Jarret et al. 1997). Drought-tolerance studies in C. colocynthis pointed to several drought-inducible genes, including a partial NAC transcript (GenBank accession number GH626169), with complex adaptive transcriptional regulation (Si et al. 2009, 2010a, b). To further characterize the function of stress-tolerant genes in C. colocynthis, two novel plant-specific TFs, CcNAC1 (KC814686), and CcNAC2 (KC814687) and their promoter regions, were cloned using 5′RACE and the genome walker kit. Here, we report the characterization of the CcNAC1 and CcNAC2 genes to gain an understanding of their function under stress conditions. Manipulation of TFs represents a potential strategy for development of transgenic stress-tolerant plants. Results indicate that CcNAC1 and CcNAC2 may have multiple functions to regulate the plant’s defense responses to abiotic stress.

Materials and methods

Plant materials

Citrullus colocynthis seeds (No. 34256) from Israel with high tolerance to drought were sown in potting mix in the greenhouse with a 14 h photoperiod and temperatures ranging from 22 to 33 °C and ambient relative humidity and light conditions (600–720 μmol m−2 s−1). Arabidopsis seeds were planted for leaf protoplast isolation. Wild-type Arabidopsis (A. thaliana, ecotype Columbia) was grown in growth chambers for 5–6 weeks, as previously described (Jensen et al. 2007).

NAC isolation and phylogenetic analysis

NAC domain proteins from more than ten different species were used as query sequences for Blastx searches of the GenBank database (http://www.ncbi.nlm.nih.gov/) for analysis of the conserved domains of NAC proteins. Primers were designed based on conserved domain sequences for cloning of the CcNAC genes into T-easy-vector (Promega, Madison, WI), followed by sequence analysis and re-blasting into the NCBI database. The 5′ RACE cloning technique (Clontech Lab Inc, Mountain View, CA) was used to obtain full-length NAC sequences from C. colocynthis. Vector sequences and low-quality sequences were manually removed following sequencing. The non-redundant C. colocynthis sequences with the highest similarity to the query sequences were investigated as putative NAC domain genes. Evaluation of putative open reading frames (ORFs) of the identified sequences was based on (1) the sequences near the translation start site corresponding to the eukaryotic consensus sequence GCC(AG)CCATGG, (2) the sequence length and homology to the in vitro translated product, (3) the conserved eukaryotic polyadenylation signal 50-AATAAA-30 following the stop codon. The ORF Finder (http://www.ncbi.nlm.nih.gov/gorf/gorf.html) was used to detect ORFs and predict amino acid sequences. Primers described in Table 1 were used for cloning of the full length CcNAC1 and CcNAC2 genes. Con-F and Con-R were used for obtaining the conserved regions of CcNAC1 and CcNAC2; GSP1 and GSP2 for the 5′ region of CcNAC1; 2.1-GSP1 and 2.1-GSP2 for 5′ region of CcNAC2; CcNAC1R/F and CcNAC2 F/R were used separately to clone the ORF of CcNAC1 and ORF of CcNAC2.

Table 1 Oligonucleotide primer sequences for CcNAC1 cDNA cloning and CcNAC2 cDNA cloning
Full size table

Previously published plant NAC-like gene sequences were retrieved from the GenBank database: ATAF1 (X74755), AtNAC2 (AB049071), AtNAM (AF123311), CUC3 (AF54194), AtNAC3 (AB049070), ANAC (AY11722), ANAC019 (At1g52890), NAC1 (AF198054), CUC1 (AB049069), CUC2 (AB002560), TIP (AF281062),NAP (At1g69490), BnNAC5-11 (AY245884), BnNAC14 (AY245886), OsNAC19 (AY596808), OsNAC5 (AB028184), OsNAC4 (AB028183), OsNAC6 (AB028185), GmNAC8 (EU661911). Multiple sequence alignment of NAC proteins was performed using ClustalW2 (http://www.ebi.ac.uk/Tools/msa/clustalw2/). Phylogenetic analysis was based on neighbor-joining (NJ) using MEGA5 (Tamura et al. 2011) with 1,000 bootstrap replications. Jones Taylor Thornton (JTT) model was used as substitution model, while Gamma distribution with invariant sites (G + I) as range substitution pattern. Amino acid sequences of NAC genes with high homology to CcNAC1 and CcNAC2 (AEF80001, XP_004161162, AFY26893, ACS94038, XP_004149802, XP_004172335, ACS94038) were aligned to investigate conserved NAC domains.

Green fluorescent protein-conjugated plasmid construction

The cDNA was amplified with primers CcNAC1F/R and CcNAC2F/R (Table 1), and the resulting PCR product was fused into the pCR8/GW/TOPO entry vector. After a sequencing check, the insert was transferred into pMDC43 via the LR reaction (GatewayR Entry vector, Life Technologies). The resulting plasmids containing Pro35S: CcNAC-green fluorescent protein (GFP) insert were used in electroporation experiments to determine the subcellular localization of CcNAC.

Protoplasts of Arabidopsis were isolated and transformed essentially as previously described in Sheen et al. (1999) with minor modifications. The tissues for protoplasts isolation were Arabidopsis leaves. The leaves were collected before flowering, excised and cut into 1 mm strips and immediately placed into an enzyme solution for overnight digestion in the dark. The enzyme solution which contained 2 % cellulose R10, 0.5 % macerozyme R10, 0.5 % driselase, 2.5 % KCl, 0.2 % CaCl2, pH 5.7, was filter sterilized. After overnight incubation, leaf tissue was gently shaken for 30 min at 40 rpm to release protoplasts, followed by filtration through a 40 μm cell sifter to remove debris and centrifugation at 150 g to pellet the protoplasts. Protoplasts were washed twice with a washing solution (0.5 M mannitol, 4 mM MES pH 5.7, and 20 mM KCl) and re-centrifuged at 150g. The protoplasts were suspended in washing solution on ice for electroporation.

Protoplasts were transformed in a manner essentially as previously described (Sheen et al. 1999; Rashotte et al. 2006). Electroporation was typically carried out with 1–2 × 105 protoplasts in 200 μl of wash solution and about 1–2 μl 400 ng/μl of plasmid DNA. Protoplasts were electroporated at 300 V in a 0.1 mm cuvette using an Eppendorf Electroporator 2510 (Hauppauge, NY). After overnight incubation in the dark, protoplasts were examined under Accu-scope 3025 phase fluorescence microscope (New York Microscope Company, Inc.). A GFP filter was used to block the chlorophyll autofluorescence and a UV filter was used to detect Hoechst 33342 fluorescence under UV light. All photographs were taken with a Qimaging Fast 1394 digital camera (imaging).

Transcriptional activation activity

The yeast strain YPG-2 containing His3 and LacZ reporter genes was used as an assay system (Stratagene, La Jolla, CA, USA). The coding sequences of ccNAC1 and ccNAC2 and the ccNAC1 and ccNAC2 N-terminal and C-terminal fragments were obtained by PCR using primers described in Table 2. The PCR products were cloned into the vector containing the GAL4 DNA binding domain to obtain pBD-ccNAC1, pBD-ccNAC2, pBD-ccNAC1-N, pBD-ccNAC1-C, pBD-ccNAC2-N, pBD-ccNAC2-C. According to the protocol of the manufacturer (Stratagene), pBD-ccNAC1, pBD-ccNAC2, pBD-ccNAC1-N, pBD-ccNAC1-C, pBD-ccNAC2-N, pBD-ccNAC2-C and the positive control pGAL4 and the negative control pBD vector were all transformed into the yeast YPG-2 competent cells. PCR products were inserted into the SalI-PstI site of pDB vector containing CcNAC1/CcNAC2-F, -N, -C, respectively.

Table 2 Primers used for transcription activation assay of CcNAC1 and CcNAC2
Full size table

The transformed strains were confirmed by PCR and streaked on YPAD or SD/His-plates. The transcription activation activities of each protein were evaluated according to their growth status. The underlined nucleotide bases in Table 2 indicate restriction enzyme digestion sites.

Isolation of CcNAC1 and CcNAC2 promoters and in silico promoter analysis of CcNAC1 and CcNAC2 promoters

The promoters (1,585 and 1,299 bp) of CcNAC1 (KC814688) and CcNAC2 (KC814689) were obtained using the Genome walker universal kit (Clontech Cat NO.638904). Plant CARE (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/), a database of plant cis-acting regulatory elements and a portal of tools for in silico analysis of promoter sequences (Lescot et al. 2002), was used to identify consensus motifs in the promoter sequences of CcNAC1 and CcNAC2.

Abiotic treatments

Treatments were conducted on seedlings at the 5–6 leaf stage. Seedlings were placed in 20 % PEG8000 solution (−0.5 MPa osmotic potential) or 50 mM NaCl to induce drought or salt treatment or water as the control treatment. Hormone and hydrogen peroxide treatments were conducted on 14-day-old seedlings. Seedlings were treated with 100 μM MeJA, 200 μM ABA, 100 μM SA, 100 μm ethephon (ET), 20 mM H2O2, or 40 μM gibberellic acid (GA). The leaves were also wounded using a hemostat (wounding). Leaves were harvested following each treatment at specific time points: 0, 1, 2, 6, 12, and 24 h.

cDNA synthesis and relative quantitative (RQ) real-time RT-PCR

RNA was extracted from leaf material using the Trizol (Invitrogen Life Technologies, Grand Island, NY) method. To eliminate the remaining genomic DNA, RNA was treated with Dnase I (Ambion Life Technologies) according to the manufacturer’s instruction. cDNA was synthesized using RETROscript™ (Ambion).

qRT-PCR was carried out using an Bio-Rad, iCycler Real Time PCR (Hercules, CA) system and iCycler detection system software. The C. colocynthis-specific actin gene (ccActin154F/R), used as the reference gene, was amplified in parallel with the target gene, allowing normalization of gene expression and providing quantification. Primers were designed based on specific regions. Primers sequences of the CcNAC1 (Q-NAC1F/R), CcNAC2 (Q-NAC2F/R) and Actin (ccActin154F/R) are listed in Table 3. Detection of RQ real-time RT-PCR products was conducted using the SYBR® Green PCR Master mix kit (Applied Biosystems, Life Technologies) following the manufacturer’s recommendations. Quantification of the relative transcript levels was performed using the comparative C T method. The induction ratio was calculated as recommended by the manufacturer and corresponds to 2−∆∆CT, where ∆∆CT = (C T, target gene,−CT, actin) treatment−(CT, Target−CT, actin)control. Relative quantification relies on the comparison between expression of a target gene versus a reference gene and the expression of same gene in the target sample versus the reference sample (Pfaffl 2001).

Table 3 Q-RT PCR primers for detection of relative expression levels of CcNAC1 and CcNAC2
Full size table

Results

Cloning and sequence analysis of the CcNAC1 and CcNAC2 genes

Earlier research on gene expression changes in response to drought in C. colocynthis had identified a partial NAC transcript (GH626169), which showed high expression levels in shoots following drought (PEG) treatment and in response to different hormones such as ABA, JA, and SA (Si et al. 2009, 2010a). Following primer design based on conserved regions of NAC genes, two NAC genes from C. colocynthis were sequenced and cloned and designated as CcNAC1 and CcNAC2. CcNAC1 encodes a 900 bp (300 amino acids) long sequence. CcNAC2 encodes an 888 bp (296 amino acids) long sequence. Amino acid alignment of CcNAC1, CcNAC2, and other NAC proteins was used to construct a phylogenetic tree (Fig. 1a). Phylogenetic analysis indicated that NAC proteins can be classified into several subgroups based on similarities to published NACs (Ooka et al. 2003). Major subgroups are shown in Fig. 1. Both CcNAC1 and CcNAC2 align with proteins in the ATAF subgroup composed of ATAF1, BnNAC5-11, GmNAC8, OsNAC19, OsNAC5, OsNAC4, OsNAC6, and BnNAC14. Amino acid alignment analysis (Fig. 1b) indicated that CcNAC1 and CcNAC2 show high homology to NACs with conserved A–E domains (five N-terminal subdomains). Even though the CcNACs are variable, especially at the C-terminal region, several conserved amino acid domains were detected.

Fig. 1
figure 1figure 1

a Phylogenetic tree of CcNAC1, CcNAC2 with other NACs proteins in plants. Numbers at the nodes of the trees represent the bootstrap vales for the node (100 replicates). CcNAC1and CcNAC2 proteins are indicated by arrows. b Analyses of amino acid sequences of Cc NAC1 and CcNAC2. The five sub-domains (A–E) are underlined by dashes. A–E domains are conservative domains of NAC transcription factors

Full size image

CcNAC1 and CcNAC2 are localized to the nucleus

To identify the subcellular localization of the CcNACs, the following constructs were made: 35S::CcNAC1-GFP and 35S::CcNAC2-GFP. The constructs were used to transform Arabidopsis leaf protoplasts. Analysis of more than 20 protoplasts showed nuclear localization of the fusion protein of CcNAC1 and CcNAC2 with GFP, as illustrated in Fig. 2, whereas the GFP protein was distributed ubiquitously in protoplasts transformed with vector plasmid control, PMDC43 (Fig. 2a2). These results indicated that both CcNAC1 and CcNAC2 are nuclear proteins.

Fig. 2
figure 2

Subcellular localization of the CcNAC1 and CcNAC2 protein in Arabidopsis protoplasts. A representative example of 35S:CcNAC1:GFP, and 35S: CcNAC2:GFP fusion in a leaf mesophyll protoplast of Arabidopsis are (B1–B3 and C1–C3). The control (A1, A2, A3) protoplast was transformed with vector pMDC43. Protoplasts were visualized under white light (A1, B1, C1), GFP excitation (A2, B2, C2 with filter blocking UV and other wavelengths) and UV light (A3, B3, C3)

Full size image

CcNAC1-C terminal has transcription activation function

CcNAC1, CcNAC2, their N-terminal domain and C-terminal domain, and full length coding sequence were fused to the GAL4 DNA binding domain to investigate their transcription activation activity. The yeast strain YRG-2 was transformed with the fusion plasmids pBD-ccNAC1, pBD-ccNAC2, pBD-ccNAC1-N, pBD-ccNAC1-C, pBD-ccNAC2-N, pBD-ccNAC2-C, the positive control pGAL4 and the negative control pBD. As shown in Fig. 3, all transformed cells can grow well on YPAD medium, but only pBD-ccNAC1-C can grow on SD medium without histidine. The filter lift assay showed that the yeast cells that grew on the SD medium without histidine turned blue in the presence of 5 bromo-4-chloro-3-indolyl-β-d-galactopyranoside (X-Gal) due to the activation of another reporter gene LacZ. The results indicate that only the C terminal region of CcNAC1 has transcription activation activity, while the full length and N-terminal region of CcNAC1 do not have this activity. This phenomenon was also observed in GmNAC20 where the C-terminal has transcriptional activation ability (Hao et al. 2010). CcNAC2 did not show transcriptional activation activity using yeast assay, since blue color was not observed using the X-Gal assay (data not shown).

Fig. 3
figure 3

Transactivation activity of CcNAC1. a Diagram of arrangement of transformants. b pBD-ccNAC1-N, pBD-ccNAC1-C, pBD-ccNAC1-F, and pBD transformants were streaked on YPAD medium. c pBD-ccNAC1-N, pBD-ccNAC1-C, pBD-ccNAC1-F, and pBD transformants were streaked on SD-His− medium. d Filter lift assay results of pBD-ccNAC1-N, pBD-ccNAC1-C, pBD-ccNAC1-F, and pBD transformants

Full size image

CcNAC1 and CcNAC2 expression patterns

Expression analysis of CcNAC1 and CcNAC2 was conducted using semi-quantitative real-time PCR. Actin, used as an internal control for constitutive expression, was uniformly expressed in all organs. As shown in Fig. 4, CcNAC1 and CcNAC2 are expressed in every tissue of C. colocynthis. CcNAC1 and CcNAC2 showed the highest level of expression in male flowers, and low expression levels were detected in fruits. Comparisons between the two genes also indicated that CcNAC2 is highly expressed in the hypocotyl of C. colocynthis.

Fig. 4
figure 4

CcNAC1 and CcNAC2 relative expression in different plant tissues and different stages as determined by qRT-PCR. Results are relative to expression in fruits. Arrow bars show SE (n = 3).CcNAC1 expression was shown as solid black, and CcNAC2 was shown as gray dashed lines

Full size image

In silico identification of stress-related promoter motifs

CcNAC1 and CcNAC2 promoters were isolated using the genome walking method. A 1,585 bp region upstream of the CcNAC1 gene and a 1,298 bp region upstream of the CcNAC2 gene were cloned, which should contain most of the regulatory domains. For further analysis of stress-related motifs in the two promoters, Plant CARE was used. Table 4 shows the details of stress-related motifs detected in the two promoters, with in the attachment the promoter sequences and motifs in color (Fig. 5). The CcNAC1 promoter contains the ABA-response element (ABRE) (Yamaguchi-Shinozaka et al. 1989; Mundy et al. 1990; Michel et al. 1993; Giraudat et al. 1994; Barker et al. 1994), ARE motif (Manjunath and Sachs 2005), CE-3 (coupling element 3) (Hubo et al. 1999), CGTCA motif, TC-rich motif, TGA-box, and TGACG-motif, which are correlated with ABA response, anaerobic induction, JA response, and auxin response. The CcNAC2 promoter contains an ABRE motif, ARE motif, HSE (heat stress responsiveness), MBS (MYB-binding site), Box-W1, and TC-rich repeats, which are correlated with ABA response, drought stress, anaerobic stress, JA response, and auxin response. Promoter motifs provide evidence for the involvement of CcNAC1 and CcNAC2 in biotic and abiotic stresses. The two promoters contain several identical motifs, such as ABRE, ARE, TC-rich repeats, which indicate that both promoters might have similar functions. However, some key differences in the composition or distribution of putative stress-related cis-acting elements (Fig. 5) were observed. Five ABRE motifs are in the CcNAC1 promoter, but only two ABRE in the CcNAC2 promoter. CcNAC2 promoter contains three TC-rich repeats, while CcNAC1 promoter contains only one TC-rich repeats. Both contain some special motifs, for example, CcNAC1 contains CE3, CGTCA-motif and TGA-motif, whereas the CcNAC2 promoter contains Box-W1, HSE, and MBS motifs. These special characteristics indicate that although both promoters have similar regulatory domains, they might be regulated by different factors. It can also be deduced that the CcNAC1 and CcNAC2 TFs might have differential regulation.

Table 4 Details of stress-related elements in CcNAC1 promoter and CcNAC2 promoter
Full size table
Fig. 5
figure 5

Location of stress-related cis-regulatory elements in CcNAC1 and CcNAC2 promoters

Full size image

CcNAC gene expression during stress and plant hormone treatments

NAC TFs are known to play important roles in plant growth and development, and can be induced by multiple biotic and abiotic stresses (Wang and Dane 2013). The complex regulatory and interaction network occurring between hormone-signaling pathways allows the plant to activate responses to different types of stimuli (Bari and Jones 2009). CcNAC1 and CcNAC2 gene expression was studied under different stress (drought, wounding, and salt) and phytohormone (ABA, JA, SA, GA, ET) treatments as shown in Fig. 6. Results indicated that both genes, CcNAC1 and CcNAC2, show similar expression patterns under hormonal and stress treatments. Both genes are induced by stress factors and hormones, even though differences in the timing of induction and level of expression were observed. Hormones can thus regulate CcNACs expression levels. For example, treatment with GA resulted in a 10-fold up-regulation of CcNAC1 at 2 h, while much higher levels of CcNAC2, 40-fold increases, were observed at a later time point, 24 h. Similarly treatment with JA resulted in 30-fold up-regulation of CcNAC1 at 24 h, while the highest levels of CcNAC2 (40-fold) were detected following JA treatment at 12 h. Treatment with ET resulted in higher expression levels of CcNAC2 than CcNAC1, although both showed the highest level at 6 h. The highest levels of CcNAC1 were detected 10 h following ABA treatment, while CcNAC2 showed the highest levels already at 2 h (40-fold). Treatment with SA resulted in 6–8-fold up-regulation of CcNAC1 from 6 to 24 h, while CcNAC2 was expressed from 1 to 12 h, and reached very high levels (40×).

Fig. 6
figure 6figure 6

CcNAC1 and CcNAC2 expression profiles under different treatments. The gray figures are expression patterns of CcNAC1, while the figures with slanted lines show expression patterns of CcNAC2. Gene expression was normalized by comparing ∆∆CT to control (0 h) (n = 3). Y-axis shows the expression level of CcNAC1 and CcNAC2

Full size image

Abiotic stresses do regulate CcNACs expression. For example, H2O2 treatment did result in high expression of CcNAC1, more than 15-fold at 24 h, while it did result in up-regulation of CcNAC2 more than 40-fold at 40 h. PEG treatment did not induce changes in expression of CcNAC1, but did cause high CcNAC2 expression, 15–25-fold. The effect of salt treatment did effect the expression of CcNAC2 more than CcNAC1; both showed similar patterns. Wounding resulted in high (40×) up-regulation of CcNAC1 and CcNAC2 after 2 h, and 24 h and similar expression patterns.

Discussion

In this study, two members of the NAC gene family in C. colocynthis were identified and NAC gene expression in different tissues and under different treatments was examined. This is the first report of molecular characterization of NAC genes in C. colocynthis The Cucurbitaceae is a large and diverse family containing several domesticated species such as watermelon, melon and cucumber (Cucumis species), squashes, pumpkins, and gourds (Cucurbita species). C. colocynthis is a source of genetic improvement for drought resistance, since this species is widely distributed in the Sahara-Arabian desert areas and well adapted to drought stress (Dane et al. 2006). Tolerance to drought stress is a complex phenomenon, comprising a number of physio-biochemical processes at both the cellular and whole plant level which are activated during different stages of plant development. Molecular mechanisms involved in different stresses have been revealed in other plant species, and TFs are one of the promising players in stress signaling pathways (Fujita et al. 2006).

The plant-specific NAC proteins constitute a major TF family implicated in many developmental processes (Puranik et al. 2012). Like most NACs, CcNAC1, and CcNAC2 contain conserved NAC domains. NAC domains are N-terminal regions of NACs that can bind both DNA and other proteins (Ernst et al. 2004). N-terminal amino acid substitutions can abolish NAC DNA-binding or structural integrity (Olsen et al. 2005). CcNAC1 and CcNAC2, like most other NACs, contain conserved NAC N-terminal and variable C-terminal domains. Protoplast transformation experiments indicated that both CcNAC1 and CcNAC2 are localized in the nucleus, which is where most TFs function. Research has indicated that C-terminal regions of many NACs possess trans-activation activity (He et al. 2005; Peng et al. 2009). Yeast assay experiments showed that CcNAC1 has trans-activation activity, while CcNAC2 does not have that ability. Some NACs have been reported to function as transcriptional repressors. Hao et al. (2010) reported that NARD (NAC Repression domain) contributed to the transcriptional repression function of GmNAC20, with the LVFY motif essentially required for suppression.

Examination of CcNAC1 and CcNAC2 expression in different plant tissues pointed to similar expression patterns, although both genes were expressed mainly in male flowers and the hypocotyl. Tissue-specific expression of members of the NAC gene family has also been studied in other species. For example, ATAF2, which is a pathogenesis-related gene in Arabidopsis, showed expression mainly in roots, leaves and mature flowers (Delessert et al. 2005). ANAC036 which caused a dwarf phenotype in Arabidopsis thaliana, was expressed mainly in rosette leaves (Kato et al. 2010). AtNAC2 was expressed mainly in root tissues and involved in salt stress responses and lateral root development (He et al. 2005). ATAF1 in Arabidopsis showed expression in every tissue, but mainly in stems, flowers and seedlings. Its overexpression resulted in severe developmental defects in Arabidopsis (Kleinow et al. 2009).

Phytohormone and stress treatments induced CcNAC1 and CcNAC2 expression to different levels. Since the promoters of both genes contain the anaerobic related motif ARE, this indicates that both genes might be correlated with oxidative stress. Salt and drought stress are worldwide problems, effecting global crop production and quality. Both genes were regulated by salt and drought stress, and a drought related motif was also detected in the CcNAC2 promoter. Elevated levels of ABA, JA and anthocyanin are metabolic signatures of oxidative stress (Steppuhn et al. 2010). While the impact of the different treatments was similar on both genes, differences in gene expression were detected. NAC TFs are candidate molecules that potentially regulate aspects of both biotic and abiotic signaling (Fujita et al. 2006). In signaling pathways, different hormones play different parts and crosstalk with each other. Earlier experiments had indicated that many NACs play a role in phytohormone pathways (Kim et al. 2008). NTL8, a membrane-bound NAC TF, plays a role in GA-mediated salt signaling in Arabidopsis (Kim et al. 2008). Studies of gene expression in wild-type and mutant Arabidopsis genotypes in response to pathogens revealed interactions among SA, JA, and ethylene (Maleck et al. 2000; Tao et al. 2003; Salzman et al. 2005). It has been suggested that ethylene produced during wounding can activate JA biosynthesis, and ethylene can also interact with the JA pathway to induce a number of pathogenesis-related and defense genes (Laudert and Weiler 1998; Kunkel and Brooks 2002). JA operates in a distinct defense pathway, which interacts with the SA pathway. JA is known to effectively mediate the defense of necrotrophic pathogens, while SA is effective against biotrophic fungi, bacteria and viruses (Murphy and Carr 2002). SA is thought to be antagonistic to JA, indicating that SA can block the JA induction pathway (Doares et al. 1995). Ethylene and JA are also associated with pathogen-induced wounding (Kunkel and Brooks 2002). JA is one of the main components of the wound repair signal in plant tissues, and the formation of JA is activated by ABA, ethylene, hydrogen peroxide, UV, whereas SA and nitric oxide inhibit the synthesis of JA (Vasyukova et al. 2011). Both JA and wounding did induce CcNAC1 and CcNAC2 expression, which indicates that both genes might play a role in the signaling of pathogen resistance and wounding response. Similar to AtNAC2, CcNAC1, and CcNAC2 were also up-regulated by ethylene. SA is a major component of the systemic acquired resistance (SAR) response, which refers to induced resistance to pathogens (Bostock 2005). ANAC055, ANAC092, and GmNAC6 genes were identified as SA signaling components (Delessert et al. 2005; Faria et al. 2011). SA similarly induced changes in CcNAC1 and CcNAC2 expression, especially to a large degree in CcNAC2 expression. The fungal elicitor motif Box-W1 was detected in the CcNAC2 promoter, which is further evidence that CcNAC2 has a function in biotic stress responses. JAZ (JASMONATE-ZIM DOMAIN) family proteins are JA co-receptors and transcriptional repressors in JA signaling in Arabidopsis. Research has indicated that JAZ orchestrates the crosstalk between JA and other hormone signaling pathways such as ethylene, gibberellic acid, SA and auxin (Kazan and Manners 2012). Both CcNAC1 and CcNAC2 were regulated by GA as well. It is known that some NACs act as regulators in several phytohormone pathways. AtNAC2 is a TF downstream of the ethylene and auxin signaling pathway (He et al. 2005).

In conclusion, two NAC TFs CcNAC1 and CcNAC2 were identified in C. colocynthis. Different stresses and phytohormones did induce CcNAC1 and CcNAC2 gene expression, which may provide clues for a better understanding of NAC gene family in this drought-tolerant cucurbit species. The identification of novel TFs regulating abiotic stress tolerance will enable further enhancement of stress tolerance in cultivated cucurbit species.

Author contribution

Z. Wang designed, conducted the experiments and wrote the manuscript. A.G. Moss and A. Rashotte helped with the fluorescence microscopy experiments, and F. Dane supervised and edited the manuscript.