Introduction

The NAC (NAM, ATAF1,2, and CUC2) family is one of the largest gene families in plants genome (Olsen et al. 2005). Totally there are 117 NAC transcription factors in Arabidopsis, which are involved in diverse developmental programs and environmental stress responses (Shahnejat-Bushehri et al. 2016). Recently, NAC transcription factors have been found to play important roles in response to endoplasmic reticulum (ER) stress, in which the transcription factor NAC103, and the membrane-associated transcription factors NAC089 and NAC062 are all up-regulated by ER stress. NAC062 relocates from the plasma membrane to the nucleus and regulates the expression of ER stress-responsive genes (Yang et al. 2014a), and NAC089 relocates from the ER membrane to the nucleus to induce programmed cell death (PCD) (Yang et al. 2014b) following ER stress treatment. NAC103 is a typical NAC family transcription factor with its transcriptional activation domain located in the C-terminus. NAC103 localizes in the nucleus and regulates UPR downstream genes including CRT1, CNX1, PDI5 and UBC32 (Sun et al. 2013). Recently, NAC103 was reported to be involved in the expression of genes related to DNA damage response (Ryu et al. 2019).

Abscisic acid (ABA) is a kind of plant hormone widely existing in plants, which is involved in regulating the growth and development of plants, such as seed dormancy, seed germination, seedling growth, leaf senescence, and other physiological processes (Seok et al. 2017; Pelagio-Flores et al. 2019). ABA is also a stress hormone that is rapidly accumulated in response to biotic and abiotic stresses. Seed dormancy is very important for the survival of plants. It can ensure that seeds germinate only when the external environment conditions are suitable. At present, many transcription factor families have been reported to play regulatory roles in the ABA pathways, such as B3, AP2/ERF, bZIP (Zhou et al. 2015). ABI3 (Giraudat et al. 1992), ABI4 (Finkelstein 1994; Finkelstein et al. 1998; Finkelstein and Lynch 2000), and ABI5 (Finkelstein 1994; Finkelstein and Lynch 2000) are all positive regulators in the process of ABA signal transduction and their mutants are less sensitive to ABA in the process of seed germination and seedling growth. In addition, some members of other transcription factor families were also found to participate in ABAresponses, such as MYB/MYC (bHLH), NAC, HD-Zip and WRKY families. At least 31 members in the NAC family are strongly up-regulated by ABA or other environmental stresses. Among them, ANAC019 (At1g52890) (Jensen et al. 2010), ANAC055 (At3g15500), ANAC072 (At4g27410) (Tran et al. 2004) and VNI2/ANAC083 (At5g13180) (Yang et al. 2011), have been demonstrated to be positive regulators in the ABA signaling pathway. However, the role of NAC103 in the ABA pathway has not yet been reported.

In the current study, we examined the biological function of NAC103 in ABA responses during seed germination and seedling growth. Genetic experiments demonstrated that NAC103 is a positive regulator in the ABA pathway. RNA-seq and qRT-PCR results showed that NAC103 regulates several ABA-responsive downstream genes. Furthermore, NAC103 is induced by ABA treatment at both transcriptional and post-translational levels. Thus, NAC103 plays an important role in ABA responses, especially during seed germination.

Materials and methods

Plant materials and growth conditions

All Arabidopsis wild-type (WT), nac103 gene-edited mutants obtained by CRISPR-Cas9 system and NAC103 overexpression plants (Sun et al. 2013) in this study were in the Columbia (Col-0) ecotype background. Seeds were surface-sterilized for 5 s in 50% ethanol and for 10 min in 0.15% NaClO solution and then rinsed three times with sterile water. The seeds were cold-treated for 3 days at 4 °C in the dark and grown on agar plates containing half-strength MS (1/2 MS) medium including multiple vitamins, 0.05% MES buffer and 1.2% sucrose in an illuminated growth chamber with long-day conditions of 16 h light and 8 h dark at 22 °C.

Seed germination and seedling greening ratio analyses

Seeds were surface-sterilized and plated on 1/2 MS medium with different concentrations of ABA (Sigma, America) as shown in the experiments. Emergent radicles were counted at the indicated time points. The averaged seed germination ratio was obtained from three experimental replicates and about 50 seeds per each genotype were observed. Seedling greening ratio was obtained by counting the number of germinated seeds with green seedlings divided by the total number of tested seeds.

Plasmid construction for CRISPR-Cas9 system

According to the online search software (https://crispr.mit.edu/), the mutation targets (target 1:CCATTGCTGAGGTCGACATTTAC; target 2:TGTGGTTACTGGAAGACCACAGG) with high scores were found in sequences encoding the NAC domain of NAC103. The target 1 or 2 was constructed on AtU6-26-sgRNA-SK plasmid vector using BsaI enzyme, and the fragments recovered by NheI and SpeI were used as sgRNA cassette. At 37 °C, pCAMBIA1300-pYAO:Cas9 plasmids were digested with SpeI (Yan et al. 2015), and then dephosphorylated with alkaline phosphatase, connected with sgRNA cassette with T4 ligase, and a binary carrier with a single target was obtained. Binary vector containing target 1 digested with SpeI single enzyme was used as the carrier, the fragments from the binary vector containing target 2 digested with NheI and SpeI double enzyme were ligated into the carrier to form the binary vector containing two targets, which were transferred into Arabidopsis thaliana with the floral-dip method, as described previously (Clough and Bent 1998). Transgenic Arabidopsis seeds were planted on the 1/2 MS plates containing hygromycin (25 mg/L) to screen transgenic plants. Fragment of NAC103 gene was amplified by PCR and sequenced in T0–T2 generation transgenic plants. The sequencing results were compared with the information on the website of NCBI (www.ncbi.nlm.nih.gov) to screen for nac103 mutants.

RNA sequencing and gene expression analyses

About 7-day-old seedlings grown on 1/2 MS plates treated with or without 200 μM ABA were immediately collected for RNA-seq. Three biological replicates for each sample were used in this study. The total messenger RNA (mRNA) of the sample was extracted using mirVana™ miRNA Isolation Kit (Ambion-1561, Invitrogen, America) according to the manufacturer’s instructions. The cDNA libraries were constructed using multiple kits according to the manufacturer’s instructions. After the cDNA libraries passed the quality test in a bioanalyzer (Agilent 2100, America), then they were sequenced on the Illumina sequencing platform (HiSeq X Ten, America). Raw RNA-seq data were processed using Trimmomatic (Bolger et al. 2014). The clean reads were obtained by removing the reads containing ploy-N and the low-quality reads, and then mapped to Arabidopsis reference genome TAIR10.31 (ftp://ftp.ensemblgenomes.org/pub/plants/release-31/fasta/arabidopsis_thaliana/dna/Arabidopsis_thaliana.TAIR10.31.dna.genome.fa.gz) using hisat2 software (Kim et al. 2015). The gene expression was calculated by an fpkm method (Adam et al. 2011) (fragments per kb per million reads), which was the number of fragments per thousand base length of a gene per million fragments. The number of counts of each sample gene was standardized by DESeq (Anders et al. 2012) software (base mean value was used to estimate the expression amount), and the analysis of different expression genes (DEGs) was calculated based on fold change and p value. All of the raw sequencing data were deposited in the National Center for Biotechnology Information (NCBI). The related raw data were as follows: BioProject ID: PRJNA643876, BioSample accession: SAMN15437556, and SRA accession: SRR12158819-30. To analyze the gene expression, quantitative RT-PCR was performed to validate the accuracy of the RNA-seq data (Sun et al. 2013). 2 μg total RNA was used for the first-stand cDNA synthesis using PrimeScript RT Master mix (RR036A, TaKaRa, Japan) according to the manufacturer’s instructions. Actin and UBQ5 were used as internal controls for normalization in quantitative reverse transcription polymerase chain reaction (qRT-PCR) analyses. The primers used in this study were given in Table S1.

Results

NAC103 is induced by ABA at both transcriptional and protein levels

Previous studies showed that many ER stress regulators, such as S2P, are involved in ABA responses (Zhou et al. 2015), we were interested in the possible role of NAC103 in ABA response. The effects of ABA on the transcriptional expression of NAC103 were analyzed. Results showed that the transcriptional expression of NAC103 was up-regulated by ABA treatment in wild-type seedlings, compared with the expression of NAC103 under normal growth conditions (Fig. 1a). Moreover, the protein level of NAC103-YFP in NAC103 over-expressing seedlings (CaMV35S::NAC103-YFP) was also increased under ABA treatment conditions (Fig. 1b). Therefore, NAC103 was induced by ABA at both transcriptional and protein levels.

Fig. 1
figure 1

NAC103 is induced by ABA. a Up-regulation of NAC103 in the wild-type seedlings (8 days after stratification) by ABA treatment. Relative gene expression is the value of wild-type plants treated with ABA (200 μM, 3 h, 6 h or 9 h) divided by that of non-treated wild-type plants, both of which were normalized to the expression of Actin. Bars depict SEM (n = 3). b Protein accumulation of NAC103 following ABA treatment. NAC103-YFP was detected by western blot with anti-GFP in transgenic plants (CaMV35S::NAC103-YFP) (8 days after stratification) treated or untreated with ABA (200 μM) for the indicated processing time. Anti-tubulin was used as a loading control

Full size image

Generation of nac103 loss-of-function mutants using the CRISPR-Cas9 gene-editing system

To analyze the molecular function of NAC103 in the ABA pathway, we used CRISPR-Cas9 system to generate NAC103 loss-of-function mutants. After primary screening and secondary screening, we obtained two independent mutant lines, nac103-1 and nac103-2, in which the NAC domains of NAC103 are predicted to be truncated (Fig. S1). In nac103 mutants, either a nucleotide T or a G was inserted after 115 bases from the transcription initiation site, resulting in an open reading frame-shift and premature termination (Fig. S1B, C). Homozygous mutants were screened and used for subsequent functional analysis.

NAC103 is a positive regulator of ABA response during seed germination

To analyze the biological function of NAC103 under normal soil-grown conditions, phenotypic differences between the nac103 mutants and wild-type plants were compared (Fig. S2). Compared with the wild-type Arabidopsis, the nac103 mutants had no significant difference at seedling stage as well as at reproductive stage (Fig. S2). To analyze the biological function of NAC103 in ABA responses, the ABA sensitivity of wild-type, NAC103 overexpression plants (NAC103-OE1, NAC103-OE3), and nac103 mutant plants (nac103-1, nac103-2) were compared. The seed germination rates of these materials were examined on 1/2 MS plates supplemented without or with 0.5 μM, 1.0 μM, 1.5 μM, 3.0 μM ABA. On the 1/2 MS plates without additional ABA, the wild-type, NAC103-OE1 and NAC103-OE3 showed similar germination rates (85–90%) after 48 h of germination (Fig. 2). When 0.5 μM, 1.0 μM, 1.5 μM or 3.0 μM ABA was added, the seed germination rates of NAC103-OE1 and NAC103-OE3 were lower than that of the wild-type plants within 72 h, especially when 1.5 μM or 3.0 μM ABA was added (Fig. 2). Without adding ABA, nac103-1 and nac103-2 showed almost the same germination rates as WT during germination (Fig. 3). When 0.5 μM, 1.0 μM, 1.5 μM or 3.0 μM ABA was added, the germination rates of nac103-1 and nac103-2 were higher than that of the wild-type Arabidopsis plants at different time points, especially when the concentration of ABA was 3.0 μM, the germination rates of nac103-1 and nac103-2 were 38.9% and 29.6%, respectively, which were significantly higher than that of wild-type Arabidopsis (only 4.4%) (Fig. 3). These results demonstrated that NAC103 is a positive regulator of ABA response during seed germination.

Fig. 2
figure 2

The NAC103 overexpression plants are more sensitive to ABA during seed germination. a, b Seed germination image (a) and statistical seed germination ratio analysis (b) of wild-type (WT) and NAC103 overexpression plants on the 1/2 MS plates with or without ABA at indicated time. Representative photographs (a) were taken at 3 day post-imbibition (dpi). Radicle emergence of seeds is defined as germination. Emergent radicles were counted at the indicated time points post-imbibition. The averaged seed germination ratio was obtained from three experimental replicates and about 50 seeds per each genotype were observed. Bars depict SEM (n = 3)

Full size image
Fig. 3
figure 3

The nac103 mutant plants are less sensitive to ABA during seed germination. a, b Seed germination image (a) and statistical seed germination ratio analysis (b) of wild-type (WT) and nac103 gene-edited mutants on the 1/2 MS plates with or without ABA at indicated time. Representative photographs (a) were taken at 3 day post-imbibition (dpi). Radicle emergence of seeds is defined as germination. Emergent radicles were counted at the indicated time points post-imbibition. The averaged seed germination ratio was obtained from three experimental replicates and about 50 seeds per each genotype were observed. Bars depict SEM (n = 3)

Full size image

NAC103 regulates ABA response during seedling growth

To study the function of NAC103 in regulating seedling growth under high ABA conditions, seedling greening ratio of the wild-type, NAC103 overexpression plants (NAC103-OE1, NAC103-OE3), and nac103 mutants (nac103-1, nac103-2) were examined. As shown in the Fig. 4, on the 1/2 MS plates without ABA, the tested plants had similar growth rates and the seedling greening ratios were more than 97%; when 0.3 μM, 0.5 μM, or 0.7 μM ABA was added to the growth medium, the nac103-1 and nac103-2 mutants had similar seedling greening ratios to that of the wild-type plants, however, seedling greening ratios of NAC103-OE1 and NAC103-OE2 plants were significantly lower than that of the wild-type plants. These results suggested that overexpression of NAC103 inhibited seedling growth when plants were treated with ABA. The nac103 mutants had no obvious seedling growth phenotype under ABA treatment, most possibly due to the potential functional redundancy of other ABA pathways in Arabidopsis.

Fig. 4
figure 4

The NAC103 overexpression plants are more sensitive to ABA during seedling growth. a, b The phenotype of the wild-type (WT), nac103 mutants and NAC103 overexpression plants (25 days after stratification) on the 1/2 MS with or without ABA (a). The percentage of green seedlings were calculated on the plates supplemented with different concentrations of ABA (b). Bars depict SEM (n = 3)

Full size image

NAC103 regulates ABA downstream gene expression

To further analyze the biological mechanism of NAC103 in ABA response, seedlings of the wild-type and nac103-1 mutants were treated without or with ABA (200 μM) for 5 h and sampled for RNA-seq analyses. The sequencing depth, the gene annotations and counts of genes mapped on the Arabidopsis genome were analyzed for each biological sample (Table S2–3), and there was a significant correlation among the three replicates of each biological sample. The differential expression profiles of different samples were obtained (Fig. 5). Among the differential expression genes (DEGs), 2395 genes were up-regulated by ABA treatment both in the wild-type and nac103-1 mutants, 1016 genes were specifically up-regulated in nac103-1 mutants while 289 genes were specifically up-regulated in wild-type plants. In contrast, there were 1932 down-regulated DEGs both in the wild-type and nac103-1 mutants, 1012 down-regulated DEGs only in nac103-1 mutants, and 699 down-regulated DEGs only in the wild-type plants.

Fig. 5
figure 5

NAC103 regulates ABA-responsive genes. a, b The differentially up-regulated (a) and down-regulated (b) genes by ABA treatment in the wild-type (WT) and nac103-1 mutant plants (8 days after stratification)

Full size image

Based on the published results (Sun et al. 2013), NAC103 is a transcription factor with transcriptional activation activity. To explore downstream genes regulated by NAC103 in the ABA pathway, we focused on the specific up-regulated DEGs in wild-type Arabidopsis, and the common up-regulated DEGs with 1.5-fold higher ratio in the WT than that in the nac103-1 plants, and GO and KEGG function enrichment was analyzed. The results showed that many of the DEGs were concentrated in terpenoid anabolism, ABA synthesis and metabolism, and gibberellin metabolism. Most importantly, DEGs related to ABA responses and ABA metabolism were enriched in the NAC103-regulated downstream genes (Table 1). The relative expression of RGL2, ICMEL2 and NCED9 in the wild-type and nac103-1 mutant treated with or without ABA were validated by qRT-PCR method (Fig. 6). The results showed that the gene expression of RGL2, ICMEL2 and NCED9 increased after ABA treatment in the wild-type Arabidopsis, in contrast, the gene expression of RGL2, ICMEL2 and NCED9 did not increase after ABA treatment in the nac103-1 mutant, indicating that these genes might be regulated by NAC103 in response to ABA treatment.

Table 1 NAC103 regulates ABA-responsive genes
Full size table
Fig. 6
figure 6

Validation of RNA-seq by qRT-PCR analysis of the NAC103-regulated ABA-responsive downstream genes. Relative gene expression is the value of wild-type or nac103-1 mutant plants treated with ABA (200 μM, 5 h) divided by that of non-treated wild-type or nac103-1 mutant plants (8 days after stratification), which were normalized to the expression of actin. Bars depict means SEM (n = 3)

Full size image

Discussion

Recently, several studies showed that plant-specific NAC103 has a variety of biological functions in various stress responses. In 2013, Arabidopsis NAC103 was first proved to be a transcription factor with transcriptional activation activity (Sun et al. 2013). The expression of NAC103 is induced by ER stress, which is directly regulated by bZIP60 through recognizing the UPRE-III cis-acting elements. NAC103 protein enters nucleus to regulate UPR downstream genes, such as BiP3, CRT1, CNX1, PDI5 and UBC32. Overexpression of NAC103 significantly increased the transcription of ER stress-responsive (ERSR) genes, such as CRT1, CNX1, PDI5 and UBC32 as well as DDR genes. Interestingly, NAC103 was also induced by genotoxic reagent during DNA replication stress. In 2019, Tae Ho Ryu et al. (2019) reported that NAC103 was involved in DNA damage response (DDR) downstream of SOG1, a master regulator in genotoxic stress response. The nac103 mutants displayed substantial inhibition of DDR gene expression after gamma radiation or radiomimetic zeocin treatment, however, overexpression of NAC103 increased DDR gene expression without genotoxic stress and substantially rescued the phenotypic changes in the sog1-1 mutant after zeocin treatment. It is known that unmitigated ER stress also induces programed cell death (PCD) (Yang et al. 2014b). Since genotoxic reagents also trigger PCD. Therefore, during ER stress response and DDR, NAC103 may mediate or inhibit PCD in Arabidopsis. In this study, the Arabidopsis NAC103 was shown to be induced by ABA treatment at the transcriptional and protein level. Under the exogenous ABA, overexpression of NAC103 inhibited seed germination rates and seedling growth, in contrast, nac103 mutants displayed increased seed germination rates. Moreover, NAC103 regulated several ABA downstream gene expressions as revealed by RNA-seq and qRT-PCR. In Canola (Brassica Napus L.), BnaNAC103 (Niu et al. 2014) was reported to respond to multiple stress signals, including cold, salicylic acid (SA) and fungal pathogen Sclerotinia sclerotiorum, and reactive oxygen species (ROS). In conclusion, the plant-specific NAC103 plays important roles in response to multi-stresses including ABA, ER stress and DNA damage. ABA, ER stress and DNA damage are all closely related to various environmental stresses. ABA, as a hormone widely found in plants, is involved in the regulation of plant development and plant responses to adversity stress, such as drought, high salt, low temperature. Seki et al. (2002) identified 245 ABA-induced genes in Arabidopsis thaliana, among which 155 genes (63%) were induced by drought, 133 genes (52%) by high salinity, and only 25 genes (10%) by low temperature. Moreover, 114 genes (46%) were found to be induced by both drought and high salt, while only 20 (8%) were induced by both drought and low temperature. ER stress is induced in plants during certain developmental stages or under adverse environmental conditions, due to the accumulation of unfolded or misfolded proteins in the ER (Liu et al. 2016). Among the ER stress-induced genes, bZIP17 was also induced by salt stress (Liu et al. 2008), bZIP28 (Deng et al. 2011) and bZIP60 (Gao et al. 2008) were involved in high-temperature stress, and IRE1A and IRE1B were both induced by pathogen infection and salicylic acid (Humbert et al. 2012). bZIP60, bZIP17, S2P and IRE1, the important regulators in ER stress response, are all involved in ABA responses (Liu et al. 2008; Zhou et al. 2015; Xian et al. 2016). Although it is not clear how ABA responses links to ER stress responses, it is possible that ABA treatment induces a subset of proteins that has some protein folding defects in the ER, or ABA treatment affects the stability of ER membranes. DNA damage is caused by abiotic and biotic environmental stresses, including UV-B, ozone, pathogens and endogenous metabolic processes, by generating reactive oxygen species (ROS) in plant cells (Rold and Ariza 2009). Because ABA, ER stress and DNA damage are closely related to environmental stress, many genes, such as NAC103, bZIP17, bZIP28, bZIP60, have various functions in response to multi-stresses.

NAC family, as one of the largest plant-specific families, is involved in diverse developmental and signaling events. NAC transcription factors in ABA signaling have been gradually revealed. The Arabidopsis NAC Transcription Factor ANAC096 (Xu et al. 2013) is involved in a synergistic relationship with a subset of bZIP-type transcription factors, ABRE-binding factor and ABRE-binding protein (ABF/AREB) for the transcriptional activation of ABA-inducible genes in response to dehydration and osmotic stresses. The transcript of NAC072 (Li et al. 2016) is up-regulated by ABF3 in ABA response, and NAC072 protein interacts with ABF3. NAC072 and ABF3 cooperate to regulate RD29A expression, but are antagonistic when regulating RD29B expression. Therefore, NAC072 displays a dual function in ABF3-mediated ABA-responsive gene regulation. VND-INTERACTING2 (VNI2) (Yang et al. 2011), an ABA-responsive NAC transcription factor, integrates ABA-mediated abiotic stress signals into leaf aging by regulating a subset of COLD-REGULATED (COR) and RESPONSIVE TO DEHYDRATION (RD) genes. In this study, we demonstrate that NAC103 was induced by ABA at transcriptional and protein levels (Fig. 1). Overexpression of NAC103 inhibited seed germination and seedlings growth under exogenous ABA treatment (Fig. 2), but mutation of nac103 increased seed germination (Fig. 3), which suggests that NAC103 is a positive regulator in ABA response.

Through RNA-seq analysis, we found that NAC103 was involved in the regulation of some ABA-related gene expression, including PLP3 (Patatin-like protein 3), AMY1 (Alpha-amylase 1) (Wang et al. 2009), HVA22B (HVA22-like protein b), RGL2 (DELLA protein RGL2) (Liu et al. 2016), MYB family transcription factor MYB78 and MYB3 (Monika et al. 2018), and ETC3 (MYB-like transcription factor ETC3). The expression of these genes was significantly induced by ABA in the wild-type, however, the fold change was decreased in the nac103 mutants, which suggests that NAC103 partially regulates the expression of these genes (Table1; Fig. 6). Although the gene expression analysis was done with seedlings rather than germinating seeds due to technical difficulties, the NAC103-dependent genes identified in this study may also be important for seed germination in response to ABA treatment, since some of these downstream genes have already been reported to be involved in ABA responses during seed germination. It was reported that there was interaction between ANAC (abscisic acid-responsive NAC) proteins and RING-H2 (Greve et al. 2003). Both RHA2a (the small RING-H2 protein) and ANAC have nuclear localization signal (NLS) to enter the nucleus, which suggests that the interaction between these two proteins regulates downstream genes expression. Moreover, the conserved N-terminal NAC domains of NAC proteins consist of a twisted β-sheet surrounded by a few helical elements, but not possess a classical helix–turn–helix motif, which can bind both DNA and other proteins (Ernst et al. 2004). Therefore, the structure of NAC103 in the N-terminus may contribute to the regulation of the ABA downstream genes.

Author contribution statement

LS, M-JW and J-XL conceived and designed research. LS, L-PL, M-JW, Y-ZW and LY conducted experiments. LS, L-PL, and M-JW analyzed data. LS wrote the manuscript. J-XL and M-JW revised the manuscript. All authors read and approved the manuscript.