Introduction

Plants are regularly threatened by various abiotic stresses, such as drought, salinity and cold, which adversely affect plant growth and development and are responsible for major yield loss in agriculture (Bray et al. 2000). As a consequence, plants have developed multifaceted strategies at the physiological, biochemical, and molecular levels to cope with these adverse stress factors, thus enabling them to survive under various environmental conditions (Bohnert et al. 2006; Broun 2004).

Plant stress responses are modulated by complex gene regulatory systems, in which transcription factors play essential roles in stress responses by regulating their target genes through specific binding to cis-acting elements in their promoters (Liu et al. 1998; Tran et al. 2004; Uno et al. 2000). Several types of transcription factors (TFs), such as NAC (NAM, ATAF1/2 and CUC2), WRKY, zinc finger, AP2/EREBP, and MYB family members, are known to transcriptionally regulate downstream gene expression under abiotic stresses in plants (Chen et al. 2003; Eulgem et al. 1999; Fujita et al. 2004; Liu et al. 1998; Mukhopadhyay et al. 2004; Tran et al. 2004; Uno et al. 2000).

NAC proteins comprise one of the largest families of plant-specific transcription factors. NAC family genes have been identified by genome-wide analysis from various plant species, such as Arabidopsis (Ooka et al. 2003), rice (Fang et al. 2008; Ooka et al. 2003), poplar (Hu et al. 2010), soybean (Le et al. 2011), etc. NAC proteins share a common structure consisting of a conserved DNA-binding NAC domain in the N-terminal region and a highly diversified C-terminal domain that confers transcriptional activator or repressor activity (Olsen et al. 2005). Based on the motif distribution, the NAC domain can be further divided into five subdomains (A–E) (Ooka et al. 2003).

NAC proteins play essential roles in diverse aspects of plant development, such as pattern formation in embryos (Souer et al. 1996), lateral root development (He et al. 2005; Xie et al. 2000), leaf senescence (Guo and Gan 2006; Uauy et al. 2006), flowering (Sablowski and Meyerowitz 1998) and secondary wall formation (Mitsuda et al. 2005). There is increasing evidence demonstrating that NAC family transcription factors are involved in responses to various biotic and abiotic stresses, including drought, salinity, cold, bacterial and fungal pathogens, low-oxygen stress, etc. (for reviews, see (Nuruzzaman et al. 2013; Puranik et al. 2012). For example, the expression of three Arabidopsis NAC genes, ANAC019, ANAC055, and ANAC072 (RD26), are induced by drought, high salinity and abscisic acid (ABA), respectively. Overexpression of these three genes remarkably enhances tolerance to drought stress (Tran et al. 2004). Another Arabidopsis stress-inducible NAC gene, AtNAC2 (ANAC092/ORE1), functions at the downstream of ethylene and auxin-signaling pathways and is required for salt stress response and lateral root development (He et al. 2005).

In Arabidopsis, ATAF1 (ANAC002) and ATAF2 (ANAC081), together with ANAC102 and ANAC032 are phylogenetically classified into a small subfamily (ATAF) (Christianson et al. 2009; Ooka et al. 2003). ATAF1 was initially reported to play a negative role in response to drought stress by functional analysis of ataf1 null mutants (Lu et al. 2007). However, later on another group reported that it is the overexpression of ATAF1, not the null mutation conferred the enhanced drought tolerance, revealing a positive role of ATAF1 in plant drought response (Wu et al. 2009). In addition, ATAF1 also functions as a negative regulator in defense responses against pathogen attack, indicating that ATAF1 plays important roles in biotic stresses (Jensen et al. 2008; Wu et al. 2009). ANAC102 was found to play an important role in regulating seed germination under low-oxygen stress (Christianson et al. 2009).

In rice (Oryza sativa), a subset of NAC transcription factors have been well characterized to participate in stress responses. Overexpression of an ATAF-like NAC gene, SNAC2/OsNAC6, in rice resulted in enhanced tolerance to dehydration, high salinity, cold stresses and blast disease (Hu et al. 2008; Nakashima et al. 2007). Another stress-responsive rice NAC gene, SNAC1/OsNAC1, confers enhanced drought and salt tolerance under field conditions by promoting stomatal closure (Hu et al. 2006). Overexpression of OsNAC5 was shown to confer enhanced stress tolerance to rice via an ABA-dependent pathway (Song et al. 2011; Takasaki et al. 2010). Overexpression of OsNAC10 driven by a root-specific promoter improves drought tolerance and grain yield in rice under field conditions (Jeong et al. 2010). In addition, improved tolerances to multiple abiotic stresses such as drought and salinity in transgenic rice plants overexpressing OsNAC045, OsNAC052 or OsNAC063 have been reported (Gao et al. 2010; Yokotani et al. 2009; Zheng et al. 2009).

Although tremendous research on NAC proteins has focused primarily on model plant species such as Arabidopsis and rice, more and more stress-related NAC genes were functionally characterized in agronomy crops such as wheat, barley, maize and soybean (Chen et al. 2013; Hao et al. 2011; Jensen et al. 2007; Lu et al. 2012; Mao et al. 2014, 2012). As for Miscanthus lutarioriparius, a potential lignocellulosic plant for bioenergy production, little information of NAC genes is available. In a previous study, we reported stress-inducible expression patterns of 13 stress-related NAC genes from M. lutarioriparius (Ji et al. 2014). In the present study, we further fulfilled a detailed functional analysis of MlNAC5. MlNAC5 was demonstrated to function as a nucleus located transcriptional activator. Overexpression of the MlNAC5 in transgenic Arabidopsis led to enhanced tolerance to drought and cold stress. These results suggest that MlNAC5 functions as a stress-responsive transcription factor in response to abiotic stresses. Furthermore, MlNAC5 overexpression in Arabidopsis affected diverse aspects of plant development under normal growth conditions, suggesting that MlNAC5 not only played important roles in abiotic stress but also was involved in diverse plant development process.

Materials and methods

Plant material and growth conditions

Arabidopsis thaliana ecotype Columbia (Col-0) was used for gene transformation in this study. Seeds were surface sterilized with 10 % sodium hypochlorite, rinsed several times with sterile distilled water and sown onto 1/2 MS plates for germination. At 8 days (d) after germination, seedlings were transferred to soil and grown in growth chamber at 21 ± 1 °C with relative humidity of 55 % under long-day (LD) conditions (16 h light/8 h dark).

Multiple sequence alignment and phylogenetic analysis

Alignment of MlNAC5 and other NAC-domain protein sequences were performed with ClustalX program (ver 2.0) (Thompson et al. 2002) and adjusted by GeneDoc software (ver 2.5). The phylogenetic tree was constructed with MEGA program (ver 5.0) by Neighbor-Joining (NJ) algorithm (Tamura et al. 2011). The parameters pairwise deletion and p-distance model were used. Bootstrap test of phylogeny was performed with 1,000 replicates.

Subcellular localization analysis

The full-length coding sequence of MlNAC5 without a terminator codon was amplified by PCR and the fragment was cloned into pGWC-T (Chen et al. 2006) to generate an entry clone (pGWC-MlNAC5) and sequenced. LR reaction (Gateway, Invitrogen) was carried out with pGWC-MlNAC5 and the binary vector pEarleyGate101 (Earley et al. 2006), which resulted in a plasmid (pEarleyGate101-MlNAC5) consisting of MlNAC5-YFP under the control of CaMV 35S promoter. The resulting construct was then sequenced to confirm an intact in-frame fusion. The construct and negative control (pEarleyGate101) were transformed into Agrobacterium strain EHA105 and infiltrated into tobacco (Nicotiana benthamiana) leaves via Agrobacterium-mediated transformation. The YFP fluorescence signal was visualized and photographed with a Laser Scanning Confocal Microscope (Olympus, FluoView FV1000). Cells were labeled with the DNA dye 4,6-diamidino-2-phenylindole (DAPI) to visualize the nucleus.

Transactivation assay

The coding sequences of MlNAC5, the fragments encoding the N-terminus (1–477 bp) and the C-terminus (478–873 bp) were amplified by PCR. The primer sequences are listed in Supplemental Table 1. The PCR products were inserted into the NdeI and EcoRI sites of the pGBKT7 vector to obtain pGBKT7-MlNAC5-FL (1–306aa), pGBKT7-MlNAC5-N (1–135aa) and pGBKT7-MlNAC5-C (136–306aa). These three constructs and the pGBKT7 empty vector (negative control) were transformed into the yeast strain AH109. The transformed strains were confirmed by PCR and then were streaked on SD/Trp and SD/-Trp/-His plates. The transactivation activity of each protein was evaluated according to its growth status and β-galactosidase filter lift assay.

β-Glucuronidase assay

A 726 bp DNA fragment upstream from the translational initiation codon of the MlNAC5 was amplified by hiTAIL-PCR (Wang et al. 2013). The fragment was inserted upstream GUS reporter gene in pCXGUS-P vector. The construct was introduced to Agrobacterium strain EHA105 and transformed into Arabidopsis Col-0 by the floral dip method (Clough and Bent 1998). In situ GUS activity assays were performed for T1 plants as described previously (Beeckman and Engler 1994). Plant materials were stained in X-Gluc solution at 37 °C for 3 h. The stained materials were cleared with 75 % (v/v) ethanol and photographed using a dissecting microscope (Olympus BX51).

In silico promoter sequence analysis

For detection of putative cis-acting regulatory elements in the promoter sequence, the online search tool of PlantCARE (plant cis-acting regulatory elements, http://bioinformatics.psb.ugent.be/webtools/plantcare/html/) was utilized.

Generation of transgenic Arabidopsis plants

The coding sequences of MlNAC5 was amplified by PCR and cloned into the pGWC-T vector (pGWC-MlNAC5) to create an entry clone. The overexpression vector (pEarlayGate100-MlNAC5) under the control of the CaMV 35S promoter was constructed by LR recombination action (Gateway, Invitrogen). The construct was transferred into Agrobacterium strain EHA105 and then transformed into Arabidopsis Col-0 plants using the floral dip method (Clough and Bent 1998). Transgenic plant seeds were screened in soil by spraying BASTA (25 mg/L). T3 homozygous lines were used for further analysis.

Germination assay

Approximately 100 seeds of homozygous MlNAC5 overexpressing lines and the wild type (WT) were surface sterilized and sown on 1/2MS agar medium (0.8 %) containing different concentrations of NaCl (50, 100 and 150 mM) or ABA (0.5 and 1.0 μM). Plates were stratified at 4 °C in the dark for 3 days, and then transferred to a growth chamber to germinate. The germination rate was calculated based on radicle protrusion or the greening of expanded cotyledons at 10 days thereafter. Each experiment was performed in triplicates.

Root length measurement

Seedlings grown on 1/2 MS plates for 3 or 6 days were transferred vertically to 1/2 MS plates supplemented with 10 μM ABA or 100 mM NaCl. The seedlings grown on 1/2 MS plates without stress were employed as the control. Root length was measured 5 days after transplantation for each plant. Each experiment was performed in triplicates.

Abiotic stress tolerance assays

For dehydration treatment, eight-day-old seedlings of MlNAC5 overexpression lines and the WT were transferred from 1/2 MS plates to water-saturated soil, water was withheld until the plants showed evident drought-stressed phenotypes, then the plants were re-watered. To characterize the function of MlNAC5 in cold tolerance, 2-week-old potted plants were treated at −8 °C for 1.5 h. After that, the plants were incubated in a cold growth chamber (4 °C) for 3 h, then transferred to normal growth conditions (21 ± 1 °C) for recovery. For each stress assay, the survival rates of transgenic lines and the WT were calculated. Experiments were repeated independently three times. For each experiment, approximately 20 plants were used.

RNA isolation and quantitative real-time RT-PCR (RT-qPCR) analysis

Total RNA was extracted using the RNAiso Plus kit (TaKaRa) and treated with RNAase-free DNase I (TaKaRa). Reverse transcription was performed with 2 μg RNA using SuperScript III reverse transcriptase (Invitrogen) according to the manufacturer’s instructions. RT-qPCR was performed on the LightCycler 480 system (Roche) using SYBR Premix ExTaq (TaKaRa) as described previously (Ji et al. 2014). Each reaction was performed in triplicates, and products were verified by melting curve analysis. Actin2 was used as an internal control for normalization. Relative gene expression levels are represented by relative quantification values calculated using the 2−ddCt method. The primers are listed in Supplemental Table 1.

Statistical analysis

Statistical analysis was calculated in SPSS 11.5 (SPSS, Inc., IL, USA) by Student’s t test to determine significant differences.

Results

MlNAC5 belongs to ATAF subfamily of NAC proteins

In a previous study, we reported the expression analysis of 13 NAC stress-related NAC transcription factor genes in M. lutarioriparius (Ji et al. 2014). Phylogenetic analysis revealed that MlNAC5 was clustered together with Arabidopsis ATAF1 and rice orthologue OsNAC6/SNAC2. MlNAC5 had an overall sequence identity of 83.3 and 71.9 % with ZmSNAC1 and OsNAC6/SNAC2, respectively, at the amino acid level. Sequence alignments showed that MlNAC5 contained a conserved NAC domain in the N-terminal region, which was further divided into five subdomains (A–E) (Supplemental Fig. 1). To further reveal the divergence of MlNAC5 proteins during evolution, we analyzed the phylogenetic relationship of MlNAC5 with their ATAF1 orthologues as well as other NAC genes, which are functionally characterized in different plant species including Arabidopsis, rice, wheat and soybean (Supplemental Fig. 2). Phylogenetic analysis showed that all NACs surveyed were divided into four subgroups in the tree. MlNAC5 was located in the first subgroup of the tree, which included all the ATAF orthologues.

MlNAC5 is localized in the nucleus

To determine the subcellular localization of MlNAC5, the MlNAC5-YFP fusion construct and the YFP control in pEarleyGate101 driven by CaMV 35S promoter were transiently expressed in tobacco epidermal cells and visualized under a confocal microscope. The YFP signal was consistently observed within both the cytoplasm and nucleus, whereas the MlNAC5-YFP fusion protein was exclusively localized in the nucleus that was confirmed by DAPI staining (Fig. 1). This result indicated that the MlNAC5 encodes a nuclear protein.

Fig. 1
figure 1

Nuclear localization of MlNAC5. YFP and MlNAC5-YFP fusion proteins were transiently expressed under control of the CaMV 35S promoter in tobacco leaves and observed under a laser scanning confocal microscope. a, e: fluorescent images of YFP. b, f: DAPI stained. c, g: Differential interference contrast (DIC) images in bright field. d, h: the merged images

Full size image

MlNAC5 is a transcriptional activator

To investigate whether the deduced MlNAC5 protein had transcriptional activity, the entire coding region, N-terminal domain and C-terminal domain were cloned into pGBKT7, which contains the GAL4 DNA-binding domain, respectively, The constructs and empty vector pGBKT7 (negative control) were transformed into yeast strain AH109. All the transformed yeast cells grew well on SD medium lacking tryptophan (-Trp) (Fig. 2). The transformants containing the full-length MlNAC5 (pGBKT7-MlNAC5-FL) and the C-terminus of MlNAC5 (pGBKT7-MlNAC5-C) could grow on selection medium SD/-Trp/-His/5 mM 3-AT, whereas the cells with the N-terminus of MlNAC5 (pGBKT7-MlNAC5-N) and pGBDT7 control failed to grow. Furthermore, the yeast cells that grew well on the SD/-His medium turned blue in the presence of 5-bromo-4-chloro-3-indolyl β-D-galactopyranoside (X-Gal), indicating that the reporter gene LacZ was activated. These results indicated that MlNAC5 functions as a transcriptional activator and its transactivation domain is located in the C-terminus.

Fig. 2
figure 2

Transactivation assay of MlNAC5 in yeast cells. The full-length proteins (MlNAC5-FL), N-terminal fragment (MlNAC5-N) and C-terminal fragment (MlNAC5-C) were fused with GAL4 DNA-binding domains and expressed in yeast strain AH109. The pGBKT7 vector was used as a negative control. The transformed yeasts were streaked on the SD/-Trp and SD/-His medium. LacZ activity was assessed by β-galactosidase filter lift assay

Full size image

Morphological characteristics of MlNAC5-overexpressing Arabidopsis plants

To analyze the functions of MlNAC5, we generated transgenic Arabidopsis plants overexpressing MlNAC5 under the control of CaMV 35S promoter. More than 30 transgenic lines were obtained and the expression of MlNAC5 was examined by RT-qPCR. Two homozygous lines (MlNAC5-OX-4 and MlNAC5-OX-18) at T3 generation with relatively high expression levels were selected for further phenotypic analysis.

The MlNAC5 overexpressing lines displayed multiple phenotypic changes at various developmental stages (Fig. 3). The transgenic plants exhibited retarded growth compared to WT, and the size of seedling was much smaller than WT. The rosette leaf morphology was also significantly altered in MlNAC5-overexpressing plants. The transgenic lines had much smaller rosette leaf size compared to WT under long-day conditions, and length to width ratio dramatically reduced.

Fig. 3
figure 3

Phenotypic analysis of MlNAC5-overexpression lines. a Seven-week-old WT and MlNAC5-overexpression plants grown under long-day (LD) conditions. Two independent MlNAC5-overexpression lines (OX-4 and OX-18) display growth retardation and the plant height was significantly reduced compared to WT. b Five-week-old WT and MlNAC5-overexpression plants grown under LD conditions. The growth of MlNAC5-overexpression lines was retarded and the plants displayed precocious leaf senescence. c Detached leaves of WT and MlNAC5-overexpression plants grown for 5 weeks under LD conditions. d Siliques of WT and MlNAC5-overexpression plants, the siliques of MlNAC5-overexpression plants were shorter compared to WT. e Flower morphology of WT and MlNAC5-overexpression lines. Some petals and sepals were removed. The length of stamens of MlNAC5-overexpression lines was significantly decreased compared to that of WT. f Plant height of 8-week-old MlNAC5-overexpression and WT plants. Plant height was measured using 20 plants for each genotype. Asterisks indicate statistically significant differences compared with WT as determined by Student’s t test (P < 0.01). Bars indicate standard deviation. g Leaf length, width and petiole length of rosette leaves from 4-week-old MlNAC5 overexpression and WT plants. Measurement was performed using 30 leaves (leaf 5 and leaf 6) for each genotype. Asterisks indicate statistically significant differences compared with WT values based on Student’s t test (P < 0.01). Bars indicate standard deviation

Full size image

The inflorescence stem growth was also retarded in MlNAC5 overexpression lines, which resulted in a moderately dwarf phenotype at maturity. In addition, the flowering time was delayed in MlNAC5-overexpressing plants compared to WT. Moreover, the fertility of MlNAC5-overexpressing plants was significantly affected and few seeds were generated at maturity. Anatomy analysis revealed that pistils and stamens developed normally in MlNAC5-overexpressing plants. However, the length of stamens in MlNAC5 transgenic plants was dramatically reduced compared to that of WT, which may hinder the successful pollination thus leading to the higher sterility in transgenic line.

MlNAC5 promoter harbors stress-responsive related cis-acting elements

We analyzed the promoter fragment of the MlNAC5 for the presence of putative cis-acting regulatory elements. A number of stress response-related cis-acting elements were present in the MlNAC5 promoter. These include three ABA response elements (ABREs), two TGACG-motifs and two CGTCA-motifs involved in the MeJA responsiveness, one LTR motif involved in low-temperature stress, and two GC-motifs involved in anoxic-specific inducibility (Table 1). These stress-related cis-acting elements may be responsive for stress-regulated expression of MlNAC5.

Table 1 Stress-related cis-acting regulatory elements identified in the promoter region of MlNAC5
Full size table

Histochemical assay of MlNAC5::GUS expression in transgenic Arabidopsis

The expression pattern of MlNAC5 was determined by histochemical β-glucuronidase (GUS) staining of transgenic Arabidopsis plants that harbored an MlNAC5 promoter::GUS reporter construct. The results indicated that MlNAC5 was mainly expressed in the hypocotyl of 5-day-old seedlings. High level of GUS expression was also detected in the main veins of rosette leaves and at the junction of the stem and the petiole of the 2-week-old seedlings. When plants grew to maturation, strong GUS activity was observed in anthers and the abscission zone of mature siliques (Fig. 4).

Fig. 4
figure 4

Expression of GUS gene driven by MlNAC5 promoter in transgenic Arabidopsis. a 5-day-old seedling. b Stem. c Leaf. d Floral buds. e Close-up view of an opened flower. f Mature silique. Scale bars a, b, e: 500 µm; c, d, f: 1 mm

Full size image

MlNAC5 overexpression confers to ABA hypersensitivity in transgenic Arabidopsis

To elucidate the role of MlNAC5 in ABA signaling, two MlNAC5-overexpressing lines and WT seeds were germinated on 1/2 MS agar plates supplemented with 0.5 and 1.0 μM ABA for 10 days. Under normal growth conditions, the germination rate of WT reached to nearly 97 %, and the two MlNAC5-overexpressing lines exhibited comparable germination rates to WT. In the presence of 0.5 μM ABA, the generation rates of MlNAC5-overexpressing seeds ranged from 10 to 20 %. By contrast, the germination rate of WT was much higher, reaching to 90 %. In the presence of 1.0 μM ABA, the germination rate of WT seeds dropped to about 45 %, while only 4 % of MlNAC5 overexpression lines were germinated (Fig. 5).

Fig. 5
figure 5

Effects of ABA treatment on seed germination of MlNAC5-overexpression lines and WT plants. a Germination assay on 1/2 MS plate supplemented with different concentrations of ABA. Seeds were sowed on 1/2 MS plates supplemented with or without different concentrations of ABA as indicated and cultured under LD conditions for 10 days. b Quantification of germination rates under ABA treatment. The germination rates were scored 10 days after planting on the plates supplemented with the indicated ABA concentrations. Three independent experiments were performed with 100 seeds for each genotype per experiment. Asterisks indicate statistically significant differences compared with WT values based on Student’s t test (P < 0.01). Bars indicate standard deviation

Full size image

To further reveal the functional roles of MlNAC5 in ABA signaling pathway during the post-germination stage, root elongation inhibition was analyzed for these plants. Six-day-old seedlings were transferred to 1/2 MS agar medium with or without 10 μM ABA and root length was measured at 5 days. The results showed that exogenous ABA substantially inhibited the root elongation for all the plants, however, the root growth inhibition was more pronounced for MlNAC5-overexpressing lines than that of WT plants (Fig. 6). Taken together, these results indicate that MlNAC5 overexpression leads to ABA hypersensitivity in transgenic Arabidopsis during germination and the root elongation stage.

Fig. 6
figure 6

MlNAC5 overexpression led to enhanced ABA sensitivity in root growth. a Comparison of primary root length of transgenic and WT Arabidopsis seedlings with ABA treatment. Seedlings grown on 1/2 MS plates for 3 days were transferred vertically to 1/2 MS plates supplemented with or without 10 μM ABA. b Quantification of primary root length. Root length was measured in a triplicates experiment using 20 plants for each genotype per experiment. Asterisks indicate statistically significant differences compared with WT as determined by Student’s t test (P < 0.01). Bars indicate standard deviation

Full size image

Overexpression of MlNAC5 enhances drought tolerance in transgenic Arabidopsis

We further tested the effect of MlNAC5 overexpression on dehydration stress tolerance via a whole-plant drought assay in soil (Fig. 7). Ten-day-old MlNAC5-overexpressing and WT plants were transferred from 1/2 MS plates to water-saturated soil. Hereafter water was withheld to gradually reduced water availability in soil. After 14 days, the relative soil water content was decreased to about 10 %, and most of the WT plants had wilted because of the extreme water deficit. In contrast, the MlNAC5-overexpressing plants did not show serious wilting except for a slight leaf senescence phenotype. After re-watering, approximately 90 % of MlNAC5-overexpressing plants remained vigorous, whereas almost 65 % of WT plants did not recover from the drought stress. In addition, MlNAC5-overexpression plants showed significantly lower water loss rates than WT during a 2-h dehydration stress. These results indicated that overexpression of MlNAC5 could enhance plant drought tolerance.

Fig. 7
figure 7

Overexpression of MlNAC5 enhanced drought tolerance in Arabidopsis. a Phenotypes of MlNAC5-overexpression lines and WT during drought stress. MlNAC5-transgenic lines and WT plants were grown for 2 weeks in soil under LD conditions, subjected to 14 days drought stress followed by 2 days of re-watering. b Percentage of survival plants in drought-tolerance assays. Experiments were performed in triplicates with 20 plants for each genotype per experiment. Asterisks indicate significantly higher survival rates compared to WT control as determined by Student’s t test. Bars indicate standard deviation. c Water loss assay from 2-week-old seedlings of MlNAC5 transgenic and WT plants grown on 1/2 MS plates. The seedlings were excised from 1/2 MS plate, and exposed to dehydration stress for 2 h on the dry filter paper at room temperature. Each data point represents the means of triplicate measurements

Full size image

Overexpression of MlNAC5 confers hypersensitivity to salinity in transgenic Arabidopsis

The effect of MlNAC5 overexpression on salt tolerance was also investigated by measuring the germination rates under 50, 100 and 150 mM NaCl treatments. As shown in Fig. 8, there were no significant differences in terms of germination rates between WT and MlNAC5 overexpression seeds under normal growth conditions. Under 50 mM NaCl treatment, the germination rates of MlNAC5 overexpression lines and WT exhibited no significant differences. When exposed to 100 mM NaCl treatment, approximately 60–70 % of MlNAC5 overexpression seeds germinated, whereas the germination rate of WT seeds was much higher, reaching to 95 % at day 10. Under 150 mM NaCl treatment, only 10–15 % seeds of MlNAC5-overexpression lines germinated, while the germination rate of WT seed reached to approximately 90 %.

Fig. 8
figure 8

Effects of NaCl treatment on seed germination of MlNAC5-overexpression lines and WT plants. a Germination assay on 1/2 MS plate with different concentrations of NaCl. Seeds were sowed on 1/2 MS plates with or without different concentrations of NaCl as indicated and cultured under LD conditions for 10 days. b Quantification of germination rates under NaCl treatment. The germination rates were scored 10 days after planting on the plates with different NaCl concentrations. Three independent experiments were performed with 100 seeds for each genotype per experiment. Asterisks indicate statistically significant differences compared with WT values based on Student’s t test (P < 0.01). Bars indicate standard deviation

Full size image

To further reveal the functional roles of MlNAC5 in salt response during the post-germination stage, root elongation inhibition was analyzed for MlNAC5 overexpression and WT plants. Three-day-old seedlings were transferred to 1/2 MS agar medium with or without 100 mM NaCl and root length was measured at 5 days. NaCl treatment substantially inhibited the root elongation for all the plants. Although the root length of two MlNAC5-overexpressing lines was shorter than that of the WT plants under normal conditions, root growth retardation was more pronounced in MlNAC5-overexpressing lines compared to that of WT plants on 1/2 MS medium with 100 mM NaCl (Fig. 9).

Fig. 9
figure 9

MlNAC5 overexpression led to hypersensitivity to NaCl in root growth. a Comparison of primary root length of transgenic and WT Arabidopsis seedlings under NaCl treatment. Seedlings grown on 1/2 MS plates for 3 days were transferred vertically to 1/2 MS plates supplemented with or without different concentrations of NaCl as indicated. b Quantification of primary root length under NaCl treatment. The experiment was performed in triplicates using 20 plants for each genotype per experiment. Asterisks indicate statistically significant differences compared with WT based on Student’s t test (P < 0.01). Bars indicate standard deviation

Full size image

Overexpression of MlNAC5 improves cold tolerance of transgenic Arabidopsis

To characterize the function of MlNAC5 in cold tolerance, 2-week-old potted plants were subject to freezing at −8 °C for 1.5 h. The plants were incubated in a cold growth chamber (4 °C) for 3 h, then transferred to normal growth conditions (21 ± 1 °C) for recovery. The results showed that the survival ratios of the two MlNAC5 transgenic lines (approximately 80 %) were much higher than those of the WT (50 %) (Fig. 10). These results indicated that overexpression of MlNAC5 improves plant tolerance to low-temperature stress.

Fig. 10
figure 10

Overexpression of MlNAC5 enhanced cold tolerance in Arabidopsis. a Phenotypes of MlNAC5-overexpression lines and WT controls during cold stress. Two-week-old potted plants were treated at −8 °C for 1.5 h, then incubated at 4 °C for 3 h, and transferred to normal growth conditions (21 ± 1 °C) for recovery. b Quantitative analysis of the plant survival rate in cold-tolerance assays. Experiments were performed in triplicates with 20 plants for each genotype per experiment. Asterisks indicate significant differences compared to WT as determined by Student’s t test. Bars indicate standard deviation

Full size image

Overexpression of MlNAC5 induces expression of stress-responsive genes

To elucidate the possible molecular mechanisms underlying drought and cold tolerance conferred by overexpression of MlNAC5, we investigated the expression of several abiotic stress-responsive genes in transgenic plants under drought, cold, and ABA treatments (Figs. 11, 12). The RT-qPCR analysis showed that transcripts of all genes examined were significantly up-regulated in WT and MlNAC5-overexpressing plants following ABA treatment. However, the fold change of up-regulation for these genes after ABA treatment for 0.5 and 1 h was much higher in MlNAC5-overexpressing plants compared to that of WT. Generally, these genes had three to tenfold higher expression in the transgenic plants than that of the WT (Fig. 11). Similarly, all genes examined were significantly up-regulated in WT and MlNAC5-overexpression plants under drought treatments. The fold changes of up-regulation for all genes except COR47 were much higher in MlNAC5-overexpression plants compared to WT control (Fig. 11).

Fig. 11
figure 11

Expression analysis of stress-responsive genes in MlNAC5-overexpression and WT plants. Transcript levels of six stress-responsive genes were analyzed by RT-qPCR using actin2 as a reference. Seedlings before treatment were used as controls. Data represent means of three independent assays. Bars indicate standard deviation

Full size image
Fig. 12
figure 12

Time-course expression of cold-responsive genes in MlNAC5-overexpression and WT plants. Expressions of six cold-responsive genes were analyzed by RT-qPCR using actin2 as a reference. The plants were grown at 21 ± 1 °C for 10 days, and treated at 4 °C for 1, 3, 6, 12 and 24 h, respectively. Plants before treatment were used as controls. Data represent means of three independent assays. Bars indicate standard deviation

Full size image

The expressions of six cold-responsive marker genes were significantly up-regulated in MlNAC5-overexpression lines and WT plants during a 24-h cold treatment. Although the transcripts of these genes peaked at different time points after the cold treatment, the fold change of up-regulation of these genes was much higher at almost all the time points in two MlNAC5 transgenic lines compared to WT (Fig. 12).

Discussion

We previously reported stress-inducible expressions of 13 NAC transcription factor genes in Miscanthus lutarioriparius (Ji et al. 2014), a candidate bioenergy plant for lignocellulosic biomass production. In the present study, we overexpressed MlNAC5 in transgenic Arabidopsis to further identify its functional roles in response to abiotic stresses. Our results showed that the transgenic plants have improved drought and cold tolerance, as assessed by a whole-plant abiotic stress assay.

The N-terminal domain of the NAC protein is crucial for recognition of and binding to specific cis-elements in their target genes, while the C-terminal region is considered to be essential for transcriptional activation. We showed that MlNAC5 is exclusively located in the nucleus in a transient assay using tobacco leaves, which is in agreement with the characteristic of a transcription factor with DNA-binding activity. Furthermore, a transactivation assay was conducted in yeast, and the results showed that MlNAC5 functions as a transcriptional activator, with transcriptional activation domain located in the C-terminal region. These results were in accordance with other NAC members reported previously (Fujita et al. 2004; Hao et al. 2011; He et al. 2005; Jeong et al. 2010; Mao et al. 2012; Nakashima et al. 2007; Olsen et al. 2005; Wu et al. 2009).

Phylogenetic analysis further revealed that MlNAC5 was clustered into the ATAF subgroup of NAC protein family and was one of the closest orthologues of ATAF1. The expression of ATAF1 was induced by wounding, pathogen infection, drought, and ABA (Jensen et al. 2008; Lu et al. 2007; Wu et al. 2009). ATAF1 was shown to mediate efficient penetration resistance in Arabidopsis upon pathogen attack (Jensen et al. 2008; Wu et al. 2009). On the other hand, ATAF1 was also reported to function as a negative regulator in drought signal transduction pathways as ataf1-1 mutant showed increased tolerance to drought stress (Lu et al. 2007). However, overexpression of ATAF1 was later reported to enhance the tolerance to drought (Wu et al. 2009). ATAF1 has been thought to respond to both abiotic and biotic stress despite of a controversy role of ATAF1 in drought stress (Nuruzzaman et al. 2013).

The orthologues of ATAF1 were also reported in other plant species, including rice (OsNAC6/SNAC2 and OsNAC52) (Gao et al. 2010; Hu et al. 2008; Nakashima et al. 2007), barley (HvNAC6) (Chen et al. 2013; Jensen et al. 2007), soybean (GmNAC20) (Hao et al. 2011), maize (ZmSNAC1) (Lu et al. 2012) and chickpea (CarNAC5) (Peng et al. 2009). The expression of OsNAC6/SNAC2 was significantly induced by drought, salt, cold wounding, and pathogen attack (Hu et al. 2008; Nakashima et al. 2007). OsNAC6/SNAC2-overexpression plants not only showed improved tolerance to drought and salinity stresses, but also exhibited increased tolerance to blast disease (Hu et al. 2008; Nakashima et al. 2007). Overexpression of OsNAC52 conferred hypersensitivity to ABA and enhanced dehydration tolerance in transgenic Arabidopsis plants (Gao et al. 2010). Transient overexpression of HvNAC6 in barley enhanced the tolerance to powdery mildew attack, whereas knockdown of HvNAC6 by RNAi led to reduced basal resistance (Chen et al. 2013; Jensen et al. 2007). GmNAC20 was shown to be regulated by various stresses, and its overexpression improved salinity and freezing tolerance in transgenic Arabidopsis plants through activation of the DREB/CBF-COR pathway (Hao et al. 2011). Meanwhile, GmNAC20 was also shown to promote the lateral root formation probably via the auxin-signaling pathway (Hao et al. 2011). Overexpression of ZmSNAC1 conferred hypersensitivity to ABA at germination as well as enhanced tolerance to dehydration in transgenic Arabidopsis (Lu et al. 2012). CarNAC5 from chickpea was reported to be induced by various stress treatments including drought, heat, wounding, salicylic acid (SA), and indole-3-acetic acid (IAA) treatments (Peng et al. 2009). All these data suggest that ATAF1 and its orthologues share a conserved role in both biotic and abiotic stresses. Our results showed that overexpression of MlNAC5, one of the closest orthologues of ATAF1 in Miscanthus, significantly improved the drought and cold tolerance in transgenic Arabidopsis via an ABA-dependent signaling pathway, which further supports a positive role of ATAF1 and its orthologues in abiotic stresses. However, the functional roles of MlNAC5 in responses to biotic stresses remain to be elucidated.

Plant response to abiotic stresses via both ABA-dependent and ABA-independent signal transduction pathways. ABA acted as the signaling molecule in responses to abiotic stresses in plants. NAC TFs have been demonstrated to regulate the expression of both ABA-dependent and ABA-independent genes during abiotic stress responses (Tran et al. 2004; Hao et al. 2011; Jensen et al. 2008; Puranik et al. 2012). The expressions of ATAF1 and its orthologues reported previously and herein could be induced by exogenously applied ABA, which suggested that ATAF1 functions to abiotic stresses through an ABA-dependent pathway. MlNAC5 overexpression plants exhibited hypersensitivity to exogenous ABA and enhanced tolerance to dehydration stress. A higher ABA sensitivity may stimulate stomatal closure to retain water and increase drought tolerance in plants. Similar results have been reported in Arabidopsis and maize (Lu et al. 2012; Wu et al. 2009).

ABA response elements (ABRE) and Dehydration responsive element (DRE) are typical cis-acting elements that mediate ABA-dependent signaling. Consistent with a role for MlNAC5 in abiotic stress responses via ABA-dependent pathway, various cis-acting regulatory elements, including four ABRE motifs, are present in the MlNAC5 promoter region.

To reveal the underlying molecular mechanisms in abiotic stress tolerance, we examined the expression levels of six abiotic stress responsive genes in MlNAC5-overexpression lines in Arabidopsis, and found that these genes were significantly up-regulated. In Arabidopsis, these genes are ultimately induced by a drought signaling pathway (Ingram and Bartels 1996). The products of these genes are low molecular weight hydrophilic proteins and may be helpful for water retention under stressed conditions. However, further extensive analysis will be necessary to elucidate the specific regulatory mechanism of MlNAC5 in abiotic stress.

Abiotic stresses such as drought, salinity and cold, are the main adverse environmental factors and adversely affects all aspects of plant phenology and metabolism. To cope with these challenges, plants have evolved a range of complex regulatory mechanisms. Growth suppression is commonly deemed as such an adaptable strategy for plants to combat against abiotic stresses. In the present study, overexpression of MlNAC5 in Arabidopsis affected many aspects of plant developmental processes, such as leaf morphology, dwarfism, late flowering and sterility. These phenotypes are similar to the previously reported transgenic plants of ATAF1 and its orthologues under the control of the constitute CaMV 35S promoter. For instance, overexpression of the Arabidopsis ATAF1 reduced the rosette size and plant height, and caused yellowing of the rosette leaves (Jensen et al. 2013). The OsNAC6-overexpression plants exhibited growth retardation and lower productivity (Hu et al. 2008; Nakashima et al. 2007). Overexpression of TaNAC2 resulted in earlier flowering and longer primary roots (Mao et al. 2012). To circumvent this problem, stress-inducible promoters should be used to suppress the detrimental effects of NAC TFs on plant growth under normal growth conditions, as well as improve stress tolerance. Supporting this claim are studies on OsNAC6/SNAC2, in which overexpression of OsNAC6 driven by stress-inducible promoters, including the LIP9 promoter and especially its own promoter, simultaneously improved stress tolerance without growth retardation effects (Hu et al. 2008; Nakashima et al. 2007). However, whether stress-inducible promoters can be used in MlNAC5 overexpression to improve stress tolerance while simultaneously suppress adverse growth effects awaits investigation.

In conclusion, we characterized MlNAC5 as a stress-responsive NAC-type transcription activator in Miscanthus. Overexpression of MlNAC5 in Arabidopsis led to hypersensitivity to ABA and NaCl, and conferred enhanced tolerance to drought and cold stresses. In addition, the data indicated that MlNAC5 is also involved in various developmental processes.

Author contribution statement

Conceived and designed the experiments: RH ZY GZ. Performed the experiments: XY XW. Analyzed the data: XY RH. Contributed reagents/materials/analysis tools: LJ JR CF. Wrote the paper: RH GZ.