Overexpression of Jatropha curcas ERFVII2 Transcription Factor Confers Low Oxygen Tolerance in Transgenic Arabidopsis by Modulating Expression of Metabolic Enzymes and Multiple Stress-Responsive Genes

Enhancing crop tolerance to waterlogging is critical for improving food and biofuel security. In waterlogged soils, roots are exposed to a low oxygen environment. The group VII ethylene response factors (ERFVIIs) were recently identified as key regulators of plant low oxygen response. Oxygen-dependent N-end rule pathways can regulate the stability of ERFVIIs. This study aims to characterize the function of the Jatropha curcas ERFVIIs and the impact of N-terminal modification that stabilized the protein toward low oxygen response. This study revealed that all three JcERFVII proteins are substrates of the N-end rule pathway. Overexpression of JcERFVII2 conferred tolerance to low oxygen stress in Arabidopsis. In contrast, the constitutive overexpression of stabilized JcERFVII2 reduced low oxygen tolerance. RNA-seq was performed to elucidate the functional roles of JcERFVII2 and the impact of its N-terminal modification. Overexpression of both wildtype and stabilized JcERFVII2 constitutively upregulated the plant core hypoxia-responsive genes. Besides, overexpression of the stabilized JcERFVII2 further upregulated various genes controlling fermentative metabolic processes, oxidative stress, and pathogen responses under aerobic conditions. In summary, JcERFVII2 is an N-end rule regulated waterlogging-responsive transcription factor that modulates the expression of multiple stress-responsive genes; therefore, it is a potential candidate for molecular breeding of multiple stress-tolerant crops.


Introduction
Waterlogging can damage most crops, creating one of the most significant problems in agriculture worldwide. During the heavy rainy season in the plain area, soil can quickly become waterlogged due to poor drainage, creating a low oxygen environment in the root area underground. Low oxygen stress leads to the induction of a particular set of genes involved in carbohydrate utilization, energy metabolism, and fermentation to sustain ATP production [1]. Over the long term, low oxygen stress means morphological adaptation is required to keep the level of oxygen under control [2]. proteins with 385, 253, and 314 amino acids with predicted molecular weights of 43,29, and 36 kD, respectively. The amino acid sequences of the JcERFVIIs were aligned with amino acid sequences from all five members of the Arabidopsis ERFVIIs, including RAP2.2, RAP2.3, RAP2.12, HRE1, and HRE2, and the phylogenetic relationship was evaluated. The results revealed that JcERFVII1 clustered with RAP2.2 and RAP2.12, JcERFVII2 clustered with RAP2. 3, and JcERFVII3 clustered with HRE2 ( Figure 1A). Based on the previously reported Arabidopsis ERFVII protein domain data [5], MEME assisted domain analysis also showed the similarity among each phylogenetic cluster ( Figure 1B).  [5].

Tissue-Specific and Waterlogging Expression Patterns of JcERFVIIs
We examined the expression pattern of the three JcERFVIIs in the tissue of Jatropha seedlings using qRT-PCR ( Figure 2). Under aerobic conditions, the expression of all three JcERFVIIs can be found in roots, leaves, apical buds, and petioles of Jatropha seedlings. JcERFVII1 exhibited the highest expression in apical buds and the lowest expression in leaves, while JcERFVII2 and JcERVII3 exhibited the highest expression in roots and the lowest expression in leaves. We also compared the expression levels of the three JcERFVIIs using the transcriptome data from roots, leaves, stems, and shoot apexes collected in a publically available J. curcas database (JCDB) [30]. We found that among the three JcERFVIIs, JcERFVII1, and JcERFVII3 displayed the highest and the lowest expression, respectively (Supplementary Materials Figure S1). Moreover, the expression levels of JcERFVII2 in Jatropha tissues were more uniform than those of the others (Supplementary Materials Figure S1).
To explore whether the JcERFVIIs are related to waterlogging response, we examined the expression patterns of JcERFVIIs in Jatropha seedlings subjected to 24 h soil waterlogging. In the waterlogged root, the expression of JcERFVII2 and JcERFVII3 was significantly increased, while the expression of JcERFVII1 remained unaffected ( Figure 2). Besides, waterlogging resulted in a significant reduction of JcERFVII1, JcERFVII2, and JcERFVII3 expression in apical buds ( Figure 2).

Tissue-Specific and Waterlogging Expression Patterns of JcERFVIIs
We examined the expression pattern of the three JcERFVIIs in the tissue of Jatropha seedlings using qRT-PCR ( Figure 2). Under aerobic conditions, the expression of all three JcERFVIIs can be found in roots, leaves, apical buds, and petioles of Jatropha seedlings. JcERFVII1 exhibited the highest expression in apical buds and the lowest expression in leaves, while JcERFVII2 and JcERVII3 exhibited the highest expression in roots and the lowest expression in leaves. We also compared the expression levels of the three JcERFVIIs using the transcriptome data from roots, leaves, stems, and shoot apexes collected in a publically available J. curcas database (JCDB) [30]. We found that among the three JcERFVIIs, JcERFVII1, and JcERFVII3 displayed the highest and the lowest expression, respectively (Supplementary Materials Figure S1). Moreover, the expression levels of JcERFVII2 in Jatropha tissues were more uniform than those of the others (Supplementary Materials Figure S1).
To explore whether the JcERFVIIs are related to waterlogging response, we examined the expression patterns of JcERFVIIs in Jatropha seedlings subjected to 24 h soil waterlogging. In the waterlogged root, the expression of JcERFVII2 and JcERFVII3 was significantly increased, while the expression of JcERFVII1 remained unaffected ( Figure 2). Besides, waterlogging resulted in a significant reduction of JcERFVII1, JcERFVII2, and JcERFVII3 expression in apical buds ( Figure 2). Relative expression was normalized to the abundance of UBQ10. Data represent mean ± SE (n = 3). Asterisks indicate p < 0.05 (t-test).

Stability of JcERFVIIs In Vitro
Since all three JcERFVII proteins possess a conserved N-degron, we hypothesized that they are targets of the N-end rule pathway. We used a previously established in vitro assay by which proteins are expressed in a rabbit reticulocyte system containing essential components for the N-end rule pathway [14]. Western blot analysis of in vitro translated JcERFVII proteins tagged with a haemagglutinin (3xHA) epitope demonstrated a single band with the migration pattern corresponding to their predicted molecular weight ( Figure 3). Our results demonstrated that mutation of cysteine to alanine at amino acid residue position 2 (MA) in all three JcERFVIIs increased protein stability after 60 and 120 min incubation periods ( Figure 3). We also showed that supplementation of MG132, a proteasome inhibitor, increased the accumulation of wildtype (MC) JcERFVII proteins in vitro ( Figure 3). These data strongly suggest that JcERFVIIs are substrates of the N-end rule pathway. Relative expression was normalized to the abundance of UBQ10. Data represent mean ± SE (n = 3). Asterisks indicate p < 0.05 (t-test).

Stability of JcERFVIIs In Vitro
Since all three JcERFVII proteins possess a conserved N-degron, we hypothesized that they are targets of the N-end rule pathway. We used a previously established in vitro assay by which proteins are expressed in a rabbit reticulocyte system containing essential components for the N-end rule pathway [14]. Western blot analysis of in vitro translated JcERFVII proteins tagged with a haemagglutinin (3xHA) epitope demonstrated a single band with the migration pattern corresponding to their predicted molecular weight ( Figure 3). Our results demonstrated that mutation of cysteine to alanine at amino acid residue position 2 (MA) in all three JcERFVIIs increased protein stability after 60 and 120 min incubation periods ( Figure 3). We also showed that supplementation of MG132, a proteasome inhibitor, increased the accumulation of wildtype (MC) JcERFVII proteins in vitro ( Figure 3). These data strongly suggest that JcERFVIIs are substrates of the N-end rule pathway.  Coomassie staining of a similar SDS-PAGE used for western blotting was used as a loading control.

Overexpression of the JcERFVII2 Enhanced Low Oxygen Tolerance in Arabidopsis
Based on previous studies, among the five members of the Arabidopsis ERFVIIs, the role of RAP2.3 in low oxygen responses has been less explored. Therefore, we aim to characterize JcERFVII2 function towards low oxygen response. To investigate the function of JcERFVII2 in providing tolerance to low oxygen stress and whether modulation of its stability could affect the stress tolerance, we generated transgenic Arabidopsis lines overexpressing MA-or MC-JcERFVII2 driven by the CaMV35S promoter.

Overexpression of the JcERFVII2 Enhanced Low Oxygen Tolerance in Arabidopsis
Based on previous studies, among the five members of the Arabidopsis ERFVIIs, the role of RAP2.3 in low oxygen responses has been less explored. Therefore, we aim to characterize JcERFVII2 function towards low oxygen response. To investigate the function of JcERFVII2 in providing tolerance to low oxygen stress and whether modulation of its stability could affect the stress tolerance, we generated transgenic Arabidopsis lines overexpressing MAor MC-JcERFVII2 driven by the CaMV35S promoter.  Figure S2A).
For submergence stress, the four transgenic lines and the wildtype A. thaliana Col-0 were grown until reaching the 10 leaf-stage and subjected to submergence stress for 3 d ( Figure 4A). While overexpression of MC-JcERFVII2 did not show any effect on the phenotype of the transgenic lines (MC-Line3 and MC-Line5) ( Figure 4A,B), it considerably improved submergence tolerance with respect to the wildtype, as demonstrated by the increases of dry weight after submergence ( Figure 4C). On the other hand, transgenic lines overexpressing MA-JcERFVII2 (MA-Line1 and MA-Line7) showed reduced plant growth when grown under aerobic conditions ( Figure 4A,B) and decreased submergence tolerance when compared with the wildtype ( Figure 4A,C).
Plants 2020, 9, x FOR PEER REVIEW 6 of 18 were selected for functional analysis. Semi-quantitative RT-PCR analysis confirmed the expression of JcERFVII-2 in the transgenic lines (Supplementary Materials Figure S2A). For submergence stress, the four transgenic lines and the wildtype A. thaliana Col-0 were grown until reaching the 10 leaf-stage and subjected to submergence stress for 3 d ( Figure 4A). While overexpression of MC-JcERFVII2 did not show any effect on the phenotype of the transgenic lines (MC-Line3 and MC-Line5) ( Figure 4A,B), it considerably improved submergence tolerance with respect to the wildtype, as demonstrated by the increases of dry weight after submergence ( Figure  4C). On the other hand, transgenic lines overexpressing MA-JcERFVII2 (MA-Line1 and MA-Line7) showed reduced plant growth when grown under aerobic conditions ( Figure 4A,B) and decreased submergence tolerance when compared with the wildtype ( Figure 4A,C).   For low oxygen survival assay, after 3 d of 2% oxygen and 3 d of recovery under aerobic condition, MC-Line3 and MC-Line5 showed significantly higher survival rate (63% and 60%, respectively) than that of the wildtype (38%) ( Figure 4D,E). However, MA-Line1 and MA-Line7 displayed a significantly lower survival rate (24% and 13%, respectively) ( Figure 4D,E). Together, these results clearly demonstrated that the constitutive overexpression of MC-JcERFVII2 could enhance growth and survival under low oxygen in transgenic Arabidopsis, while that of MA-JcERFVII2 resulted in growth reduction under aerobic conditions and poorly performed under low-oxygen stress.

Transcriptome Profiling of Transgenic Arabidopsis Overexpressing JcERFVII2
To analysis the impact of the N-terminal modification on the molecular function of the JcERFVII2 gene, we profiled the transcriptome of transgenic Arabidopsis overexpressing MAand MC-JcERFVII2 (MA-Line1 and MC-Line3, respectively) using Col-0 as a control genotype. Two biological replicates of total RNAs from 7 d.o. seedlings grown in aerobic conditions were isolated and subjected to RNA-seq. RNA-seq reads were mapped to the A. thaliana TAIR10 genome. The number of reads aligned back to each gene was obtained for differential gene expression analysis. Transcriptome analysis identified 344 and 282 differentially expressed genes (DEGs) with significant changes in gene expression as evaluated by false discovery rate (FDR) < 0.05 from MAor MC-JcERFVII2 overexpressing lines, respectively ( Figure 5A; Supplementary Materials Table S1). Of 282 DEGs from the MC-Line3, 29 DEGs (10%) were upregulated, and 253 DEGs (90%) were downregulated (Supplementary Materials Table S1), while, of 344 DEGs in the MA-Line1, 122 DEGs (35%) were upregulated, and 222 DEGs (65%) were downregulated (Supplementary Materials Table S1). Venn's diagram analysis revealed that 112 DEGs were commonly found in both MAand MC-JcERFVII2 transgenic lines, while 232 DEGs and 170 DEGs were exclusively found in MAand MC-JcERFVII2 transgenic lines, respectively ( Figure 5A). It should be noted that the endogenous ERFVIIs were not differentially expressed in transgenic lines overexpressing both MAand MC-JcERFVII2 (Supplementary Materials Table S1). To confirm that, we obtained the CPM (count per million) expression values from our RNA-seq data. Mostly, the expression of the endogenous ERFVII genes in transgenic lines is similar to the Col-0 (Supplementary Materials Figure S2B).
Since we observed more DEGs being upregulated in the MA-JcERFVII2 overexpressing line than that of the MC-JcERFVII2, we carefully examined the expression of the 122 upregulated DEGs from the MA-JcERFVII2 transgenic line ( Figure 5C). Of the 122 upregulated MA-JcERFVII2 DEGs, 22 of these were also upregulated in the MC-JcERFVII2 overexpressing line (Supplementary Materials Table S1). The rest of them (100 genes) were not differentially expressed in the MC-JcERFVII2 overexpressing line ( Figure 5C; Supplementary Materials Table S1). A possible explanation for these results is that the increase in JcERFVII2 protein abundance could elevate the expression of these 100 genes. GO analysis of the upregulated DEGs from the MA-JcERFVII2 transgenic line revealed their roles in response to multiple stresses, including hypoxia (FDR: 2.60 × 10 −7 ), oxidative stress (FDR: 6.60 × 10 −7 ), and other organisms (1.10 × 10 −6 ) ( Figure 5D; Supplementary Materials Table S2).  Based on GO enrichment results, DEGs in some specific classes demonstrated co-expression patterns ( Figure 6). Several Plant defensin (PDF) genes were upregulated in both MAand MC-JcERFVII2 lines ( Figure 6A, Supplementary Materials Table S1). Glutathione transferase and peroxidase genes were upregulated mainly in the MA-JcERFVII2 line ( Figure 6A, Supplementary Materials Table S1). In contrast, specific genes that function in ABA and JA responses were downregulated in both MAand MC-JcERFVII2 transgenic lines ( Figure 6B; Supplementary Materials Table S1). These results altogether indicate that post-translational modification of JcERFVII2 protein under aerobic conditions can affect its regulative function.

Validation of JcERFVII2 Target Genes
For verification of the RNA-seq results, quantitative reverse-transcription polymerase chain reaction (RT-PCR) was used to quantify 6 representative transcripts. The selected mRNAs included

Validation of JcERFVII2 Target Genes
For verification of the RNA-seq results, quantitative reverse-transcription polymerase chain reaction (RT-PCR) was used to quantify 6 representative transcripts. The selected mRNAs included three core hypoxia genes (HB1, PDC1, and PCO2), two plant defense responsive genes (PDF1.2 and PDF1.3), and Alternative oxidase 1D (AOX1D). The analysis confirmed that levels of these mRNAs are more induced in the MA-JcERFVII2 overexpressing line than those of the MC-JcERFVII2 and Col-0 grown in aerobic conditions (Figure 7). Furthermore, low oxygen-induced the accumulation of these mRNAs in all genotypes; however, the mRNA accumulation in some of these genes is slightly higher in the MA or MC-JcERFVII2 overexpressing lines (Supplementary Materials Table S2).
Plants 2020, 9, x FOR PEER REVIEW 11 of 18 three core hypoxia genes (HB1, PDC1, and PCO2), two plant defense responsive genes (PDF1.2 and PDF1. 3), and Alternative oxidase 1D (AOX1D). The analysis confirmed that levels of these mRNAs are more induced in the MA-JcERFVII2 overexpressing line than those of the MC-JcERFVII2 and Col-0 grown in aerobic conditions (Figure 7). Furthermore, low oxygen-induced the accumulation of these mRNAs in all genotypes; however, the mRNA accumulation in some of these genes is slightly higher in the MA or MC-JcERFVII2 overexpressing lines (Supplementary Materials Table S2).

Discussion
This study focuses on elucidating the roles of JcERFVIIs towards waterlogging and low oxygen response. Phylogenetic and domain architecture analyses reveal that JcERFVII1 and JcERFVII2 are orthologs of constitutively expressed Arabidopsis ERFVII genes, RAP2.2 and RAP2.12 and RAP2.3, respectively ( Figure 1). The last member of this JcERFVII family, JcERFVII3, is an ortholog of lowoxygen induced Arabidopsis HRE2 (Figure 1). This study reveals that the expression of JcERFVII1 is highly constitutive and remains unaffected following waterlogging, while JcERFVII2 and JcERFVII3 are upregulated by waterlogging (Figure 2 and Supplementary Materials Figure S1). Analysis of RAP2.3 in flooding tolerant Brassica species, Rorippa sylvestris and Rorippa amphibia, demonstrated that under flooding, no induction of RAP2.3 was observed [12]. Altogether, these data indicate that

Discussion
This study focuses on elucidating the roles of JcERFVIIs towards waterlogging and low oxygen response. Phylogenetic and domain architecture analyses reveal that JcERFVII1 and JcERFVII2 are orthologs of constitutively expressed Arabidopsis ERFVII genes, RAP2.2 and RAP2.12 and RAP2.3, respectively ( Figure 1). The last member of this JcERFVII family, JcERFVII3, is an ortholog of low-oxygen induced Arabidopsis HRE2 (Figure 1). This study reveals that the expression of JcERFVII1 is highly constitutive and remains unaffected following waterlogging, while JcERFVII2 and JcERFVII3 are upregulated by waterlogging (Figure 2 and Supplementary Materials Figure S1). Analysis of RAP2.3 in flooding tolerant Brassica species, Rorippa sylvestris and Rorippa amphibia, demonstrated that under flooding, no induction of RAP2.3 was observed [12]. Altogether, these data indicate that JcERFVII2 from waterlogging sensitive Jatropha and RAP2.3 from Brassica plants might undergo divergent evolution in gene expression.
In the dicot model Arabidopsis, all five ERFVIIs possess conserved motif function as N-degron that promotes the degradation of ERFVIIs via oxygen-and nitric oxide (NO) dependent N-end rule pathway of targeted proteolysis [8,10,13,14]. Overexpression of all five Arabidopsis ERFVIIs drastically improves low oxygen tolerance by promoting the expression of the genes involved in low oxygen adaptation [13,14,16,20]. Intriguingly, overexpression of stable version of HRE1 and HRE2 further improved low oxygen tolerance in Arabidopsis [14], while overexpression of stable version of RAP2.12 resulted in a reduction of plant growth in air and also decreasing submergence tolerance in Arabidopsis [13,22]. In this study, we demonstrated that the JcERFVIIs 1-3 are targeted at the N-end rule pathway in vitro (Figure 3), leading to a question of whether modulation of the JcERFVII2 stability can further improve low oxygen tolerance. Transgenic Arabidopsis lines overexpressing MC-JcERFVII2 are highly tolerant of both flooding and low oxygen stress, suggesting that JcERFVII2 could function as a low-oxygen determinant (Figure 4). In contrast, transgenic Arabidopsis lines overexpressing MA-JcERFVII2 are highly sensitive to low oxygen stresses ( Figure 4). Moreover, overexpression of MA-JcERFVII2 yields a decrease in rosette size and dry-weight when grown in air ( Figure 4A,B), demonstrating that modulation of the JcERFVII2 stability interferes with growth and development.
In this study, transcriptome profiling reveals that modification of JcERFVII2 stability affects transcript accumulation of multiple genes controlling cellular metabolism and stress responses ( Figure 5B). Previously, Bui et al. [32] demonstrated transcriptional activity of constitutively expressed RAP2.2, RAP2.3 or, RAP2.12 on a set of hypoxia-responsive promoters. Papdi et al. [20] showed that all three RAP2 genes, when overexpressed, can transactivate ADH (alcohol dehydrogenase) promoter. In addition, Gasch et al. [33] demonstrated that overexpression of all three RAP2 genes induced expression of ADH in transgenic Arabidopsis. Similarly, our study found that overexpression of MAand MC-JcERFVII2 upregulated the expression of 8 out of 49 core hypoxia-responsive genes (Supplementary Materials Table S1). Previous studies demonstrated that ectopic expression of ERFVIIs in transgenic plants increased tolerance to multiple abiotic stresses [8]. Some evidence suggested that RAP2.3 functions in pathogen response and ROS detoxification. Ogawa et al. [34] showed that tobacco BRIGHT YELLOW-2 cells overexpressing Arabidopsis RAP2.3 were more tolerant of H 2 O 2 and heat stress. Moreover, the expression of PDF1.2 and GST6 was enhanced in the transgenic Arabidopsis lines overexpressing RAP2.3 [34]. Furthermore, overexpression of the RAP2.3 ortholog, CaPF1 (Capsicum annuum pathogen and freezing tolerance-related protein 1), in Virginia pine upregulated several antioxidant enzymes including ascorbate peroxidase, glutathione reductase and superoxide dismutase [35]. In this study, we found that overexpression of both MAand MC-JcERFVII2 induced the expression of several PDF genes (PDFs 1.2, 1.2b, 1.2C, and 1.3; Figure 6A; Supplementary Materials Table S1). We also observed the upregulation of several GST and peroxidase genes in the transgenic line overexpressing MA-JcERFVII2 ( Figure 6A; Supplementary Materials Table S1). Altogether, these results demonstrate that JcERFVII2 may involve in pathogen response and reducing ROS accumulation in plant cells.
Our study demonstrated that overexpression of MA-JcERFVII2 interferes with growth and development ( Figure 4). Paul et al. [22] compared transgenic Arabidopsis lines overexpressing wildtype and stabilized forms of RAP2.12 under aerobic conditions and found that the stabilized RAP2.12 affected central metabolic processes by increasing activities of fermentative enzymes and accumulation of fermentative products including ethanol, lactate, alanine and γ-amino butyrate (GABA), therefore resulted in decreased ATP and starch levels. In this study, GO enrichment analysis revealed that the alpha-amino acid metabolic process was enriched in the upregulated DEGs from MA-ERFVII2 ( Figure 5D). This GO category includes genes encoding for several enzymes responsible for glutamate and GABA synthesis (AT2G02010: glutamate decarboxylase 4 (GAD4), AT5G37600: glutamine synthase (GSR1); AT5G38200: Class I glutamine amidotransferase-like superfamily protein, and AT4G35630: phosphoserine aminotransferase; Supplementary Materials Table S1), implying the possibility that the transgenic Arabidopsis lines overexpressing MA-JcERFVII2 could face carbohydrate starvation that leads to reduced growth and development.
In Arabidopsis, transcriptional activation of RAP2.12 can be counterbalanced by a trihelix transcriptional factor, namely HYPOXIA RESPONSE ATTENUATOR1 (HRA1) [36]. Giuntoli et al. [36] demonstrated that the interaction between RAP2.12 and HRA1 could enable an adaptive response to low oxygen, required for stress survival. Interestingly, transgenic wheat constitutively expressed the stabilized TaERFVII.1 showed no growth defect phenotype, which resulted from the upregulation of TaSAB18.1, an ortholog of HRA1, under aerobic condition [25]. In this study, we did not observe the upregulation of HRA1 from the transgenic Arabidopsis overexpressing both MAand MC-JcERFVII2 grown under aerobic conditions (Supplementary Materials Table S1).
Leon et al. [37] recently showed that enhanced RAP2.3 expression reduced NO-triggered transcriptome adjustment, and thus it functions as a brake for NO-triggered responses that included the activation of JA and ABA signaling in Arabidopsis. In addition, Vincente et al. [38] found that RAP2.3 enhanced abiotic stress responses by interacting with BRM, a chromatin-remodeling ATPase, that repressed ABA responses. Gibb et al. [39] demonstrated that RAP2.3 regulated the expression of ABSICISIC ACID INSENSITIVE5 (ABI5), a major negative regulator of germination in seed endosperm. Interestingly, our study found that the NO-scavenger gene, HB1, was upregulated (Supplementary Materials Table S1).
Additionally, genes involved in JA and ABA-activated signaling and responses were mostly down-regulated in transgenic Arabidopsis overexpressing MA and MC-JcERFVII2 ( Figure 6B), suggesting JcERFVII2 could modulate NO accumulation and hormonal response.
In summary, our study demonstrated that JcERFVII2 is an N-end rule regulated waterloggingresponsive transcription factor that functions by modulating gene expression of cellular metabolic and multiple stress-responsive genes, including low-oxygen, oxidative, and pathogen response. Constitutive upregulation of fermentative and stress-responsive genes could compromise growth and development in the transgenic Arabidopsis overexpressing the stabilized JcERFVII2. This study highlights several possibilities for future investigation, including genetic manipulation of the JcERFVII2 gene in Jatropha to determine whether it can improve waterlogging tolerance and elucidation of the JcERFVII2 roles in controlling physiological responses to multiple abiotic stresses in Jatropha and other crop plants.

Multiple Sequence Alignment and Motif Identification
Full-length amino acid coding regions of ERFVIIs were downloaded from the Jatropha genome database (https://www.kazusa.or.jp/jatropha/) and the Arabidopsis information resource (http://www. arabidopsis.org/). Multiple sequence alignment was performed using CLUSTALW, and then a phylogenetic tree was built by the neighbor-joining method (Poisson correction, pairwise deletion of gaps) using the MEGA10 software [40]. Domain analysis was performed using MEME [41] following the models published for Arabidopsis [5].

Genetic Materials
J. curcas (cv. "Chai Nat"-a local Thai variety) and A. thaliana genotypes including the Col-0 accession and 35S:MC-JcERFVII2 and 35S:MA-JcERFVII2 (ectopic expression) transgenic lines were used in this study. The genome of Col-0 has already been sequenced.

Plant Growth and Stress Condition
J. curcas seedlings were grown and waterlogged, as described in Juntawong et al. [29].
For growth in soil, A. thaliana plants were grown in soil containing 50% (v/v) peat, 25% (v/v) perlite, and 25% (v/v) coconut fiber with regular irrigation in a growth room at 120 µmol photon m −2 s −1 16 h light/8 h dark, at 23 • C. Submergence stress was performed using 10 leaf-stage plants grown in 5-cm 2 pots by placing them in a plastic container completely filled with water for 3 d.
For growth in sterile culture, A. thaliana seeds were surface sterilized, stratified at 4 • C for 48 h and plated on 0.5× solid Murashige and Skoog (MS) medium (0.215% (w/v) MS salts containing 1% (w/v) agar, pH 5.7) in 20-mm 2 dishes. Growth was in a vertical orientation in a growth room. Hypoxia stress was performed under dim light at the end of the 16-h light cycle in a sealed argon chamber. For hypoxia stress, 98% argon and 2% oxygen mixture was passed through water and into the chamber while ambient air was pushed out by positive pressure. Control was placed in an open chamber side by side.

Quantitative Reverse Transcription PCR
Total RNA samples were extracted using TRIzol reagents (Thermo Fisher Scientific, Waltham, MA, USA), subjected to DNase treatment, and RNA cleanup using an RNA-mini kit (Qiagen, Hilden, Germany). Three replicates of total RNA samples were used. One microgram of total RNAs was used to construct cDNA using MMuLv reverse transcriptase (Biotechrabbit, Berlin, Germany) in a final volume of 20 µL. The cDNA was diluted five times. Quantitative-realtime PCR (qPCR) reaction was performed according to Butsayawarapat et al. [42] using QPCR Green Master Mix (Biotechrabbit, Berlin, Germany) on a MasterCycler RealPlex4 (Eppendorf, Hamburg, Germany). For each sample, the PCR reaction was performed in triplicate. Each reaction contained 1 µL of diluted cDNA, 0.5 µM of each primer and 10 µL of QPCR Green Master Mix in a final volume of 20 µL. The PCR cycle was 95 • C for 2 min, followed by 45 cycles of 95 • C for 15 s and 60 • C for 30 s. Amplification specificity was validated by melt-curve analysis at the end of each PCR experiment. Relative gene expression was calculated using the 2 −∆∆CT method. Primers used to study Jatropha's gene expression were previously reported by Juntawong et al. [29]. The genes and primers used in Arabidopsis are shown in Supplementary Materials Table S3.

Analysis of Protein Stability
To construct the plasmids used for in vitro protein stability assay, cDNAs were amplified from J. curcas total cDNA using gene-specific primers (Supplementary Materials Table S3). The PCR products were ligated into a modified version of the pTNT (Invitrogen, Carlsbad, CA, USA) expression vector (pTNT-3xHA) [14]. N-terminal mutations were incorporated by modifying the forward primer sequences accordingly (Supplementary Materials Table S3).
For in vitro protein expression, TNT T7 Coupled Reticulocyte Lysate System (Promega, Madison, WI, USA) and 2 µg plasmid template was used according to manufacturer's guidelines. Where appropriate, 100 mM MG132 (Sigma, St. Louis, MO, USA) was added. Reactions were incubated at 30 • C. Samples were taken at indicated time points before mixing with protein loading dye to terminate the reaction. Equal amounts of each reaction were subjected to anti-HA immunoblot analysis. All blots were checked for equal loading by Coomassie Brilliant Blue staining.

Generation of Transgenic Lines
To construct Ti binary plasmids for plant transformation, JcERFVII2 open reading frame was amplified by RT-PCR from RNA extracted from roots of J. curcas using gene-specific primers (Supplementary Materials Table S3). The PCR product was inserted into the pCXSN binary plasmid [43], transformed into E. coli DH5α, and selected with 50 µg mL −1 kanamycin. N-terminal mutations were incorporated by changing the forward primer sequences accordingly (Supplementary Materials Table S3). The pCXSN, a plant overexpression vector, provides a CaMV 35S promoter and nopaline synthase terminator sequence in a Ti binary plasmid with a hygromycin-resistant gene. After sequence confirmation, the plasmid was electroporated into Agrobacterium tumefaciens GV3101 and colonies selected with 50 µg mL −1 kanamycin. Col-0 transformation was performed according to Clough and Bent [44]). T1 seeds were collected, seedlings resistant to 35 µg mL −1 hygromycin were propagated, and homozygous single insertion events were established.

RNA-Seq, Differential Gene Expression Analysis, and Gene Ontology Enrichment
Total RNA samples were extracted using TRIzol reagents (Thermo Fisher Scientific, Waltham, MA, USA), subjected to DNase treatment, and RNA cleanup using an RNA-mini kit (Qiagen, Hilden, Germany). Two replicates of total RNA samples were used for transcriptome analysis according to the ENCODE recommended RNA-seq standards (https://genome.ucsc.edu/ENCODE/protocols/ dataStandards/ENCODE_RNAseq_Standards_V1.0.pdf). The integrity of the RNA samples (RIN) was evaluated on an RNA 6000 Nano LapChiprun on Agilent2100 Bioanalyzer (Agilent Technologies, Waldbronn, Germany). Samples with a RIN > 7 were used in RNA-seq library preparation.
For each sample, 3 µg of total RNAs were used to generate a sequencing library using an Illumina ® TruSeq TM RNA Sample Preparation Kit v2 (Illumina, San Diego, CA, USA). Paired-end, 100 bp RNA-seq was performed on a NovaSeq6000 platform. FASTQ files were generated with the base caller provided by the instrument. Quality control filtering and 3 end trimming were analyzed using the FASTX-toolkit (http://hannonlab.cshl.edu/fastx_toolkit/index.html) and Trimmomatic software [45], respectively. The raw read files were deposited in the NCBI GEO database under the accession numbers GSE154601.
Differential gene expression analysis was performed according to Juntawong et al. [29]. The FASTQ files were aligned to the reference transcriptome using TopHat2 software (v2.0.13) [46]. A binary format of sequence alignment files (BAM) was generated and used to create read count tables by the HTseq python library (citation). Differentially-expressed genes were calculated using the edgeR program [47] with an FDR cutoff of <0.05.
Author Contributions: Conceived, designed, and supervised research, P.J.; performed research, P.J., P.B., P.S., R.P., and S.V.; performed in-silico analysis, interpreted the results and wrote the manuscript, P.J. All authors have read and agreed to the published version of the manuscript.