Introduction

Programmed cell death (or PCD) represents a cascade of events that lead to the programmed destruction of cells1. This process is accurate and genetically controlled due to the regulation of a large number of genes and related processes. Plant experiences a variety of responses to orchestrate PCD events such as the accumulation of reactive oxygen species (ROS), release of mitochondrial cytochrome c and the activation of caspase-like proteases2. Under unfavorable conditions, H2O2 accumulates, hence, stimulates the production of ROS or the occurrence of oxidative burst2. ROS comprise free radicals, e.g., superoxide (O2−), hydroxyl radical (OH•), and non-radicals, e.g., hydrogen peroxide (H2O2) and singlet oxygen (1O2)3. Detoxification of oxygen species, generated under adverse condition, is mandatory for the protection of plant cells and organelles against the toxic effects4.

Cells exposed to harsh conditions, e.g., pathogen attack or abiotic stresses, induce several processes including PCD. In our recent work, we confirmed that manipulating PCD-related genes resulted in differential levels of salt stress tolerance in plant5,6. Of which, knockout mutants of Bax Inhibitor 1 (BI-1) gene and its driving transcription factor namely ethylene responsive factor 109 (ERF109) gene in Arabidopsis confirmed the role of either gene in standing salt stress. Bax inhibitor-1 (BI-1) is a cell death suppressor conserved among eukaryotes7,8. BI-1 blocks Bax-induced cell death downstream of Bax action in the mitochondria. The protein (25–27 kDa) is trans-membrane residing in endoplasmic reticulum that exists in several tissues including leaf, root and stem and largely enhanced under stress conditions such as heat, cold, drought and salt. BI-1 serves in increasing the capacity of cellular homeostasis under oxidative stress condition, thus, blocks cell death2. Bahieldin et al.5,6 indicated that expression of such a protein is due to the action of ERF109. In tobacco (Nicotiana benthamiana), BI-1 gene was upregulated early before the onset of PCD (2 h of treatment with a PCD-inducing agent namely oxalic acid; Bahieldin et al.5. Action of BI-1 gene was also previously confirmed in Arabidopsis9. These results supports the idea that BI-1 gene overexpression is vital for basal suppression of cell death progression under adverse conditions.

Arabidopsis ERF109 was recently proposed to mediate cross-talking of jasmonic acid and auxins such as indole acetic acid10. The TF drives expression of two genes, e.g., Anthranilate synthase alpha subunit 1 (or ASA1) and YUC2 gene of the YUCCA gene family encoding flavin monooxygenase-like enzyme, important in two pathways namely tryptophan biosynthesis and tryptophan metabolism, respectively. Plants overexpressing ERF109 gene resulted in the overproduction of auxins, thus, overgrowth of roots, hairy roots and hypocotyls. Upon JA treatment, the gene is also overexpressed in shoots and roots, especially in the lateral root primordia11.

In the present study, transcriptomic analysis via RNA-Seq was conducted in leaves of Arabidopsis T-DNA insertion mutant knocked out for ERF109 gene (loss-of-function) and transgenic plants overexpressing ERF109 gene (gain-of-function) in a trial to detect the possible role of this TF in driving expression of other PCD- or plant growth-related genes under salt stress.

Results and Discussion

RNA-Seq of cDNA samples of Arabidopsis leaves of three genotypes, e.g., WT, KOERF109 and OEERF109, treated with NaCl (200 mM) for 2 and 12 h as well as the untreated control was done in order to search transcripts that might be driven by ERF109 other than those recently detected5,6,9. Two time points were selected based on previous evidence that this TF is upregulated after 2 h of exposure to PCD inducer (e.g., oxalic acid) or salt stress5,6. The analysis yielded ~22–24 million reads corresponding to an average of >2 billion nucleotides of cDNA per sample (Table 1). The raw RNA-Seq data indicated that over 90% of the reads mapped to the Arabidopsis reference nuclear genome in the exonic regions (Table 1).

Table 1 Statistics of Arabidopsis RNA-Seq numerical data analysis for three genotypes (WT, KOERF109 and OEERF109) across time of salt stress treatment (2 and 12 h as well as the untreated controls).
Full size table

Cluster analysis

Hierarchical cluster analysis of gene expression based on log ratio RPKM data and Multi-dimensional scaling plot for transcripts of the three genotypes of Arabidopsis indicated transcriptomic separation of the two time points of salt stress (Figs 1 and 2, respectively). Transcriptomes at 2 and 12 h time points were closer than either time point and control. Based on the transcriptomic data shown in Figs 1 and 2, we concluded that the distance between transcriptomes of 2 h salt-treated or 12 h salt-treated samples and transcriptomes of the control samples is almost the same. Therefore, we assume untreated samples at 12 h time point can be a general control of the experiment. The total number of generated clusters was 65 (Table S1) with ~1470 differentially expressed transcripts (Table S1). Of which, expression levels of transcripts of clusters 1, 2, 3 and 4 differed at the genotype (e.g., WT, KOERF109 or OEERF109) level. No regulation was shown across the four clusters for transcripts of the three genotypes of the control. Transcripts of OEERF109 showed the high expression levels at 2 and 12 h time points in cluster 1 and at 12 h only in cluster 2, while transcripts of KOERF109 showed low expression levels at the two time points in the two clusters (Figs 3 and 4). Transcripts of WT and OEERF109 in cluster 3 showed similar expression levels at 2 h time point, while transcripts of KOERF109 showed the lowest at this time point (Fig. 5). However, results of cluster 4 remain unexplained as transcripts of KOERF109 showed high expression levels at 12 h time point, while OEERF109 showed low expression levels (Fig. 6). Transcripts of these four clusters at selected time points are further called ERF109-regulated or ERF109-upregulated. Transcripts of other clusters were differentially expressed across time of salt stress treatment regardless of the genotype, e.g., ERF109-independent (Table S1). These clusters showed either salt-induced up- or downregulation at a given time point, e.g., 2 h (9 clusters), 12 h (28 clusters), 2 h/12 h (20 clusters) or with gradual up- or downregulation (4 clusters). RNA-Seq data was successfully validated via the use of semi-quantitative (sq)RT-PCR for five transcripts of cluster 2 (Table S2) with WT and OEERF109 genotypes unevenly upregulated at 12 h time point (Fig. S1). The four ERF109-regulated clusters were further analyzed as the focus of this study. These clusters contained 3, 54, 15 and 4 ERF109-regulated transcripts, respectively (Table S1).

Figure 1
figure 1

Hierarchical cluster analysis of gene expression based on log ratio RPKM data for leaf transcriptome of Arabidopsis genotypes (e.g., WT, KOERF109 and OEERF109) under high salt (200 mM NaCl) treatment for 2 and 12 h as well as the untreated control. WT = wild type, C = control untreated, KO = ERF109 knocked out mutant, OE = ORF109 overexpressed.

Full size image
Figure 2
figure 2

Multi-dimensional scaling plot of raw sequence counts of RNA-seq data for leaf transcriptome of Arabidopsis genotypes (e.g., WT, KOERF109 and OEERF109) under high salt (200 mM NaCl) treatment for 2 and 12 h as well as the untreated control. WT = wild type, C = control untreated, KO = ERF109 knocked out mutant, OE = ORF109 overexpressed. Each sample is displayed as a different color.

Full size image
Figure 3
figure 3

FPKM values for the transcripts upregulated at 2 h of salt (200 mM) stress treatment that were concordantly expressed with ERF109 transcript in leaves of Arabidopsis genotypes (e.g., WT, KOERF109 and OEERF109) across time (2 and 12 h) of salt stress treatment as well as the untreated controls. Transcripts showed relatively lower levels of expression in leaves of KOERF109 plants, while higher levels of expression in leaves of OEERF109 plants compared with WT at any given time point. WT = wild type, C = control untreated, KOERF109 = ERF109 knocked out mutant, OEERF109 = ORF109 overexpressed plant. Description of transcripts is shown in Table S1 (cluster 1).

Full size image
Figure 4
figure 4

FPKM values for the transcripts upregulated at 12 h of salt (200 mM) stress treatment with relatively lower levels of expression in leaves of Arabidopsis KOERF109 plants, while higher levels of expression in leaves of OEERF109 plants compared with WT. Gene expression levels in the three genotypes (e.g., WT, KOERF109 and OEERF109) were not changed at earlier time points (control and 2 h) of salt stress treatment. WT = wild type, C = control untreated, KOERF109 = ERF109 knocked out mutant, OEERF109 = ORF109 overexpressed plant. Description of transcripts is shown in Table S1 (cluster 2).

Full size image
Figure 5
figure 5

FPKM values for the transcripts upregulated at 2 h of salt (200 mM) stress treatment with relatively lower levels of expression in leaves of Arabidopsis KOERF109 plants compared with OEERF109 and WT. Gene expression levels in the three genotypes (e.g., WT, KOERF109 and OEERF109) were not changed at either earlier or later time point (control or 12 h, respectively) of salt stress treatment. WT = wild type, C = control untreated, KOERF109 = ERF109 knocked out mutant, OEERF109 = ORF109 overexpressed plant. Description of transcripts is shown in Table S1 (cluster 3).

Full size image
Figure 6
figure 6

FPKM values for the transcripts downregulated at 12 h of salt (200 mM) stress treatment with relatively higher levels of expression in leaves of Arabidopsis KOERF109, while lower levels of expression in leaves of OEERF109 plants compared with WT. Gene expression levels in the three genotypes (e.g., WT, KO and OE) were not changed at earlier time points (control and 2 h) of salt stress treatment. WT = wild type, C = control untreated, KOERF109 = ERF109 knocked out mutant, OEERF109 = ORF109 overexpressed plant. Description of transcripts is shown in Table S1 (cluster 4).

Full size image

Regulated cell death genes

One transcript in cluster 2 namely DAL1-like or DAL2-like, analogue of DIAP1 gene in Drosophila, encodes a death-associated inhibitor of ROS during either PCD or biotic/abiotic stress12,13. This gene naturally acts as a negative regulator of PCD in Arabidopsis12. Previous reports with knockout mutants of DAL1 or DAL2 gene resulted in higher accumulation of ROS upon infection with the avirulent strain of Pseudomonas syrinage pv. tomato (Pst) DC3000, thus, approached PCD earlier than the WT. The idea that this gene is regulated by ERF109 aligns with our speculation that ERF109 might regulate PCD-inhibitor genes other than the recently discovered BI-1 gene.

Regulated genes encoding ROS scavenging enzymes

The results in Fig. 5 indicate that some ERF109-regulated transcripts of cluster 3 belong to gene families encoding four ROS scavenging enzymes, namely ascorbate peroxidase (APXS), glutathione S-transferase (GST), phospholipid hydroperoxide glutathione peroxidase (PHGPX) and dehydroascorbate reductase (DHAR). As Bax is a PCD-inducing protein stimulated by ROS, its action can be blocked not only by the action of BI-1 or DAL1/2 protein but also by the action of ROS scavenging enzymes such as APXS, GST, PHGPX and DHAR2,14,15,16. The first enzyme namely APXS is a key enzyme in the peroxide-detoxification system in chloroplasts as it converts H2O2 into H2O with ascorbate used as an electron donor17. Previous reports indicated that this enzyme is regulated under both biotic and abiotic stresses17 and acts in scavenging ROS in several organelles and the cytosol18. Under salt stress, responses of APXS gene were previously reported to be tissue- and developmental stage-specific19. The second ROS scavenger namely GST has a major role in glutathione-dependent isomerization and the reduction of toxic hydroperoxides as GST attaches glutathione to electrophilic xenobiotics for subsequent sequestration in the plant vacuole20. The enzyme also has affinity for auxins, thus, contributes to hormone homeostasis under different environmental conditions21. The third enzyme namely PHGPX provides an enzymatic defense mechanism against oxidative destruction of biomembranes and acts in the removal of lipid hydroperoxides22. The latter compounds are toxic and naturally recovered during accumulation of ROS, e.g., hydroxyl radicals and singlet oxygen. Lipid hydroperoxidation results in the decreased membrane fluidity and damage of transmembrane proteins, hence, inactivates protein receptors, important enzymes and ion channels within the membrane. Recent reports indicated that lipid hydroperoxides can be scavenged by PHGPX22. The forth enzyme namely DHAR reduces dehydroascorbate to non-enzymatic antioxidants ascorbic acid. This action requires the use of a reduced form of glutathione as an electron donor23. The enzyme participates in regulating AA pool size in symplast and apoplast towards maintaining the plant cell’s redox state under stress conditions24. In general, the present data supports the involvement of ERF109 in retarding PCD under salt stress due to the regulation of processes participating in ROS inhibition.

Aligning with the latter conclusion, ERF109 gene was previously proven to express in Arabidopsis when plant leaves approach re-adjustment towards homeostasis under other types of abiotic stresses, ex., high light stress25.

KEGG analysis

Other ERF109-regulated transcripts of the four clusters mainly belong to three consecutive pathways, e.g., tryptophan biosynthesis26,27,28 (Fig. S2), tryptophan metabolism26,27,28 (Fig. S3) and the downstream plant hormone signal transduction26,27,28 (Fig. S4). The first pathway results in the biosynthesis of tryptophan that is used in the second pathway as a substrate for auxin (IAA) biosynthesis. In turn, biosynthesized auxin participates in the third pathway in inducing responses of genes related to cell enlargement and plant growth. KEGG analysis was conducted in order to detect the activated enzymes after 12 h of exposure to the target salt concentration (200 mM NaCl) as compared to those in the control plantlets whose samples were concurrently harvested. We can extract that ERF109 might act as a “master switch” mediator of a cascade of consecutive events across these three pathways by directly or indirectly regulating expression of a battery of genes in these pathways (Fig. 7).

Figure 7
figure 7

Activated steps in the pathways of tryptophan biosynthesis, tryptophan metabolism and plant hormone signal transduction due to the initial ERF109-mediated “master switch” (indicated by red asterisks) at the three pathways for ASA1, IGPS, YUC2 and LAX2 genes. Master switch subsequently triggered expression of other downstream genes and gene families of TRP1, PAI1, TSA1, TAA1, YUC2, AUX/IAA (or IAA), ARF, TIR1, GH3, PIN4, AUX/LAX (or LAX), LBD and SAUR. AS = anthranilate synthase alpha subunit 1, APRT = anthranilate phosphoribosyltransferase, PRAI = phosphoribosylanthranilate isomerase 1, IGPS = indole-3-glycerol phosphate synthase, TSAαC = tryptophan synthase alpha chain, TPA = L-tryptophan pyruvate aminotransferase, IPM = indole-3-pyruvate monooxygenase.

Full size image

Tryptophan biosynthesis and metabolism pathways

We speculate that the master switch is basically triggered by the ERF109-induced expression of ASA1 and YUC2 genes in tryptophan biosynthesis and metabolism pathways, respectively (Fig. 7). The downstream ERF109-regulated genes in the first pathway involve the consecutively functioning TRP1, PAI1, IGPS and TSA1 genes (Fig. S2). The ASA1 gene presents in cluster 1 (Fig. 3), while the rest of genes in this pathway present in cluster 2 (Fig. 4). The ERF109-regulated genes in tryptophan metabolism involve the consecutively functioning TAA1 and YUC2 genes (Fig. S3). The TAA1 gene in the present study presents in cluster 2 with upregulation 12 h after salt stress (Fig. 4), while YUC2 presents in cluster 1 with upregulation 2 h after salt stress (Fig. 3). ERF109 was recently reported to drive expression of ASA1 gene encoding anthranilate synthase alpha subunit 1 in tryptophan biosynthesis pathway and YUC2 gene encoding indole-3-pyruvate (or flavin) monooxygenase in tryptophan metabolism pathway9. Expression of these two genes participates in the biosynthesis of IAA that is required for triggering the plant hormone signal transduction pathway (Fig. 7). Auxin biosynthesis pathway with tryptophan as a precursor has more than one direction in Arabidopsis. The shortest is a two-step reaction due to the expression of TAA1 gene, encoding tryptophan pyruvate aminotransferase, and YUC2 gene, encoding flavin monooxygenase29. The latter gene catalyzes a rate-limiting step in the pathway30. We suggest that ASA1 gene is also a rate-limiting step in tryptophan biosynthesis pathway (Fig. 7). As ERF109 is driving expression of the latter two genes, it is assumed that expression of this TF is vital for the two pathways.

Plant hormone signal transduction pathway

The ERF109-regulated genes in the plant hormone signal transduction pathway involve a battery of gene families, e.g., AUX/IAA (or IAA), ARF, TIR1, PIN, GH3, AUX/LAX (or LAX), LBD and SAUR (Figs 7 and S4). These auxin-responsive genes mostly present in cluster 2 (Fig. 4) and pathway results in enlarged cells and improved plant growth. The influence of IAA is cell type-dependent resulting in differential responses at the transcriptional up to the post-translational levels31. Auxin signaling is ultimately controlled by the complex of auxin response factors (ARFs) and their interacting repressors AUX/IAA proteins. ARFs bind to promoter elements of auxin responsive genes (e.g., LAX, LBD and SAUR), while AUX/IAA proteins bind to ARFs to block its action, thus negatively regulate transcription of auxin-induced genes32,33. A large number of genes, e.g., 29 AUX/IAA and 23 ARF, encode these two heterogeneous, tissue-specific types of controlling elements in Arabidopsis31. These diverged elements justify the differential role in development processes and their controlling battery of genes34. Previous reports indicated that degradation of AUX/IAA proteins in Arabidopsis takes place due to the elevated levels of auxin and subsequent increased levels of ARFs35,36. AUX/IAA proteins differ in their expression patterns and tissue specificity. For example, five IAA genes, namely IAA9, IAA12, IAA18, IAA19 and IAA26 are expressed only in the lateral roots. Of which, only IAA19 was ERF109-upregulated in the present study (Fig. 4). At least, one IAA gene in each tissue is required to respond rapidly after exposure to auxin in order to securely act in the different tissues. The encoding/degradation process of AUX/IAA proteins is regulated to promote the continuous expression of the downstream responsive genes37. Among ARF genes, ARF19 is the most sensitive to auxin biosynthesis, while ARF16 is expressed after prolonged exposure. Unlike ARF19, ARF16 is a repressor of expression of downstream responsive genes. Albeit the different functions, the two genes were previously reported to co-express in different plant organs and during different developmental stages37. ARF19 gene is closely-related to ARF5, ARF7 and ARF834. The latter four genes act as activators of auxin-responsive gene expression in Arabidopsis. ARF19 is speculated to act as an amplifier of the auxin signal representing a positive-feedback signaling loop38. In the present study, these four genes were ERF109-upregulated 12 h after salt stress treatment (Fig. 4) supporting the indirect action of ERF109 in promoting plant growth under normal and adverse conditions. The E3 ubiquitin ligase SCFTIR1 (or TIR1) is also auxin-induced and is the main contributor to AUX/IAA protein degradation. This enzyme allows the activation of ARF, derepresses downstream auxin responsive pathways, thus mediates plant growth and development31. As the highly conserved domain II of AUX/IAA proteins is a target for degradation process, auxin promotes degradation of all types of AUX/IAA proteins by allowing the participation of auxin-induced TIR1. In the present study, TIR1 gene was also ERF109-upregulated 12 h after salt stress treatment (Fig. 4) to secure the activation of its target factors, e.g., ARFs, thus, allowing the auxin-responsive downstream genes, AUX/LAX, LBD and SAUR, to act (Fig. 7). Active form of auxin functions in its unconjugated form. Towards the homeostasis and regulation of active form of auxin in the cell, GH3 gene family encodes a class of auxin-induced conjugating enzymes that block action of excessively available auxin37. This gene family comprises 19 genes in Arabidopsis. Genes encoding group II enzymes, e.g., GH3.1, GH3.3, GH3.4, GH3.5 and GH3.6, and group III enzymes of GH3, e.g., GH3.14 and GH3.17, are upregulated in seedling and root, respectively, to decrease the hormone active form in these tissues. In the present study, only two GH3 genes, e.g., GH3.1 and GH3.3, of group II and one, e.g., GH3.17, of group III in addition to GH3.2 and GH3.9 genes were upregulated after 12 h of salt stress treatment. The most rapidly responding gene to auxin is GH3.3. Localization of auxin carriers occurs asymmetrically in which influx/efflux process are genetically regulated. Of which, PIN gene family acts as auxin efflux carrier. The gene family comprises as little as eight members39. For example, PIN4 gene generates a sink for auxin into columella cells40. This action secures the occurrence of auxin gradient and homeostasis to correct root patterning. The present study indicated that PIN4 gene was ERF109-upregulated after 12 h of salt stress treatment. Interestingly, previous reports indicated that the three PIN genes, e.g., PIN1, PIN3 and PIN7, are the only members that are upregulated by auxin37. The latter genes or gene families, e.g., TIR1, GH3 and PIN either participate in degrading AUX/IAA proteins or secure auxin homeostasis under normal and adverse conditions. Thus, these genes allow the activated form of ARF to stimulate expression of downstream genes of growth enhancement such as like LAX), LBD and SAUR. As PIN family members act as auxin efflux carriers, LAX genes act as auxin influx carriers41. The gene family mainly comprises three genes, e.g., LAX1, LAX2 and LAX3, with differential roles and tissue specificities by inducing several downstream genes towards promoting the overall plant growth. In the present study, these three genes were ERF109-upregulated 12 h after salt stress treatment (Fig. 4). The LBD genes act towards the induction of lateral root formation and enhancement of LR density9. LBD16 and LBD29 genes are considered the most sensitive to auxin presence as they overexpress after as little as 30 min of exogenous or indigenous auxin. Out of the 42 LBD genes in Arabidopsis, only LBD16 was ERF109-upregulated in the present study after 12 h of salt stress treatment (Fig. 4). Activated ARF also stimulates the induction of SAUR gene family encoding highly unstable mRNAs42. The family comprises over 70 genes. Of which, SAUR62 and SAUR64-68 were reported to be strongly upregulated by auxin, while clade including SAUR15, SAUR19-22, SAUR24 was moderately auxin-responsive. Previous reports indicated that SAUR36 and SAUR72 respond negatively to auxin. High expression levels of SAUR genes exist in leaves37 in accordance with the results of the present finding. SAUR genes act as regulators of cell elongation43 and stimulators of shoot elongation37. In the present study, all these SAUR genes were ERF109-upregulated after 12 h of salt stress treatment except for SAUR36.

Putative cis-acting promoter elements

The data shown in Table S3 indicates putative cis-acting promoter elements of several regulated genes of clusters 3 and 4 with binding sites for ERF109 and several other transcription factors of the AP2/ERF gene family. These genes include three GST genes (cluster 3) namely GSTU7, GSTU6 and GSTF8. As indicated earlier, these genes encode ROS scavenging enzymes and act in blocking Bax protein to avoid PCD and contributes to hormone homeostasis under different environmental conditions. The other two regulated genes (cluster 2) act during tryptophan biosynthesis (e.g., IGPS) and plant hormones and signal transduction (e.g., LAX2) pathways. The promoters of GSTF8 and LAX2 genes harbor four binding sites for ERF109, while GSTU7 gene harbors three binding sites. Promoter sequences of the GSTU6 and IGPS genes harbor only one binding site for ERF109. This data indicates that ERF109 might act as the master switch for IGPS and LAX2 genes in addition to the previously published ASA and YUC2 genes10. These four genes act during the three pathways under study.

In conclusion, we speculate that ERF109 acts under salt stress not only in regulating PCD inhibitors (e.g., BI-1 and DAL1/2) or inducing ROS scavengers, but also as a “master switch” mediator in promoting plant growth and re-adjustment to homeostasis due to the direct participation in auxin biosynthesis. These actions combine to increase the plant’s ability to tolerate salt stress. This conclusion is supported by our recent study in detecting the influence of ERF109 to confer salt stress tolerance in Arabidopsis6.

Materials and Methods

Plant materials

Arabidopsis WT (Col), the knockout T-DNA insertion (SALK_150614) mutant (namely KOERF109) and over-expression lines of ERF109 gene (CS2102255) (OEERF109) of locus AT4G34410 were provided by the SALK Institute, Genomic Analysis Laboratory (SIGnAL) (http://signal.salk.edu/tdnaprimers.2.html). Plantlets were grown from seed in a growth chamber for two weeks under the following growth conditions. First, seeds of the three genotypes were surface sterilized, sown in Petri dishes containing MS medium. The plates were kept in the dark at 4 °C for 2 days and then shifted to 21 ± 2 °C (day/night) under light intensity of ~175 umolm-2s-1 and a 16-h-light/8-h-dark cycle where plantlets were allowed to grow for 12 more days. Knockout mutant was screened for homozygosity by standard PCR approach, while seeds of the over-expression line were homozygous. Sequences of PCR primers and reaction conditions were recovered from Arabidopsis database (TAIR, http://www.arabidopsis.org/, Table S2).

Salt stress experiment

The experiment was conducted at the laboratories of the Department of Biological Sciences, KAU, Jeddah, Saudi Arabia. In order to harvest transcriptomes under salt stress, 2-wk-old plantlets of KOERF109 and OEERF109 lines as well as the WT (Col) with homogeneous performance were transferred to pots (9 cm) filled with soil mix (1 soil: 1 vermiculite), where salt stress experiment started (Fig. S5). Two-wk-old control untreated and salt-stressed plantlets were allowed to grow at the above mentioned growth conditions. Control plantlets were irrigated daily with deionized double distilled water and allowed to grow for two more weeks. While, salt-stressed plantlets were initially irrigated daily with salt concentration of 50 mM NaCl for one week. Then, two incremental increases of salt stress was made as recommended by Munns44. An increase of 75 mM NaCl was made for 3-wk-old plantlets, which were left to grow for one more week. Another increase of 75 mM NaCl was made for 4-wk-old plantlets which is the target salt concentration (200 mM NaCl) for RNA-Seq analysis. At the same day of reaching the target salt concentration, leaf samples of the 4-wk-old plantlets were harvested after 2 and 12 h. Leaf samples of the control unstressed 4-wk-old plantlets were harvested concurrently with those harvested 2 and 12 h after salt treatment.

RNA-Seq analysis

RNA was isolated from leaves of different genotypes across time of treatment. Total RNA was extracted from three similar-sized (10 mm2) leaf discs per plant (approximately 50 mg tissue) collected from upper leaves using Trizol (Invitrogen) and treated with RNase-free DNase (Promega Inc.). RNA samples were then shipped to BGI, China, for deep sequencing. However, the low quality RNAs isolated from control leaf samples at 2 h time point made them unsuitable for deep sequencing. Therefore, we had to rely only on the leaf samples of control plantlets harvested concurrently with those harvested 12 h after salt stress, only. Raw data were submitted to the NCBI for reviewing and receiving accession numbers. Analysis of the RNA-Seq datasets indicated the recovery of >100 million reads per sample. Adapter sequences were trimmed-off and high quality sequences were aligned (≥2 mismatches) to Arabidopsis thaliana genome (http://www.arabidopsis.org/, TAIR version 10) using RSEM v1.1.6 and Bowtie aligner (Bowtie v0.12.1). Then, differential expression and cluster analysis were done by EdgeR (version 3.0.0, R version 2.1.5). Blastx was performed (with an E-value cut off of 1e−5) and FPKM values of differentially expressed transcripts were measured against the actin used as the house-keeping gene. Significant Pearson correlation was determined during permutation analysis. Principal component analysis (PCA) was determined using trinity-v2.3.2 PtR module with default parameters. The generated clusters were analyzed for GO terms using Blast2GO (http://www.blast2go.org/). To identify the biological pathways that are active at 12 h time point of salt stress as compared to the untreated control at 12 h time point, the detected genes were mapped to reference canonical pathways in the Kyoto Encyclopedia of Genes and Genomes (KEGG) (http://www.genome.ad.jp/kegg/)26,27,28. Analysis of promoter cis-elements for the genes in the first four clusters was done using Promoter Analysis software (http://plantpan2.itps.ncku.edu.tw/promoter.php) consulting the promoter sequence (up to −500 nt) datasets available at TAIR in order to detect putative binding sites for ERF109 and other TFs in the AP2/ERF gene family.

RNA-Seq datasets were then validated via sqRT-PCR of selected transcripts. First-strand cDNA was synthesized using 2.5 ug of total RNA, 0.5 ug oligo (dT) primer and Superscript II reverse transcriptase (Invitrogen) to a final volume of 20 ul. sqRT-PCR was performed in 20-ul reactions using 1 ul cDNA, 1 × PCR buffer (with 1.5 mM MgCl2), 200 uM dNTPs, 200 nM of each gene-specific primers (Table S2) and 0.2 U of Taq DNA polymerase (Promega Inc.). Primers were designed using Netprimer software (http://www.premierbiosoft.com/netprimer/index.html) with the following criteria: length 20–22 bases, GC content ~40–50%, minimal secondary structures, comparable annealing temperatures of the primer pairs, and PCR products of 261–353 bp. Forty PCR cycles for each gene product include denaturation at 94 °C for 15 sec, annealing at appropriate temperature for 30 sec, and extension at 72 °C for 45 sec. Amplicons were analyzed on a 1.2% agarose gel stained with ethidium bromide and visualized using the Gel Doc XR from Bio-Rad Laboratories (Hercules, CA, USA).