Group VII ethylene response factors forming distinct regulatory loops mediate submergence responses

Abstract

Group VII ethylene response factors (ERFVIIs), whose stability is oxygen concentration-dependent, play key roles in regulating hypoxia response genes in Arabidopsis (Arabidopsis thaliana) and rice (Oryza sativa) during submergence. To understand the evolution of flooding tolerance in cereal crops, we evaluated whether Brachypodium distachyon ERFVII genes (BdERFVIIs) are related to submergence tolerance. We found that three BdERFVIIs, BdERF108, BdERF018, and BdERF961, form a feedback regulatory loop to mediate downstream responses. BdERF108 and BdERF018 activated the expression of BdERF961 and PHYTOGLOBIN 1 (PGB1), which promoted nitric oxide turnover and preserved ERFVII protein stability. The activation of PGB1 was subsequently counteracted by increased BdERF961 accumulation through negative feedback regulation. Interestingly, we found that OsERF67, the orthologue of BdERF961 in rice, activated PHYTOGLOBIN (OsHB2) expression and formed distinct regulatory loops during submergence. Overall, the divergent regulatory mechanisms exhibited by orthologs collectively offer perspectives for the development of submergence-tolerant crops.

Three Group VII ethylene response factors form a refined regulatory loop involving transcriptional and protein stability controls in the submergence stress response in Brachypodium.

Introduction

The majority of cereal crops cannot survive long periods of hypoxia in flood-prone areas (Fukao and Bailey-Serres 2004). An exception is rice (Oryza sativa), which can tolerate hypoxic stress in roots for a prolonged period of time (Fukao et al. 2006). However, even in rice, submergence of the whole plant under water will cause drastic damage to roots and shoots (Ismail et al. 2009). Different rice cultivars, including both domesticated and wild species, display great variation in submergence tolerance. The SUBMERGENCE 1A (SUB1A) gene locus, which encodes a group VII ethylene response factor (ERFVII), is present in indica cultivars but absent in japonica cultivars (Fukao et al. 2006). SUB1A consists of two alleles, SUB1A-1 and SUB1A-2. Cultivars that possess SUB1A-1, including FR13A, are highly tolerant to submergence, whereas cultivars with SUB1A-2 or without SUB1A are sensitive to submergence (Xu et al. 2006). Overexpression of SUB1A-1 in a flood-sensitive japonica cultivar results in a submergence-tolerant phenotype, indicating that SUB1A-1 is the main determinant of submergence tolerance (Xu et al. 2006; Fukao and Bailey-Serres 2008).

Arabidopsis (Arabidopsis thaliana) RELATED TO APETALA2.12 (RAP2.12), RAP2.2, and RAP2.3 are APETALA2/ERF-type transcription factors that belong to the same ERFVII subfamily as rice SUB1A-1. Under hypoxic conditions, overexpression and repression of these transcription factors in Arabidopsis result in higher and lower survival rates, respectively (Papdi et al. 2015). In contrast with SUB1A-1, which is induced by submergence (Xu et al. 2006), Arabidopsis RAP2.12, RAP2.2, and RAP2.3 are constitutively expressed under normoxia and activate hypoxic responses under oxygen-limited conditions (Papdi et al. 2015). Fine-tuning of hypoxia-triggered anaerobic metabolism is achieved by HYPOXIA RESPONSE ATTENUATOR1 (HRA1), which is highly upregulated by hypoxia and counteracts the phenotypic and transcriptional effects of RAP2.12 in Arabidopsis. HRA1 induction is necessary for negative feedback regulation of RAP2.12 to modulate the extent of anaerobic response (Giuntoli et al. 2014, 2017).

Arabidopsis ERFVIIs contain the N-terminal amino acids “Met-Cys” (MC) and are degraded rapidly by N-degron-targeted proteolysis when exposed to O₂ and nitric oxide (NO) (Gibbs et al. 2011, 2014). Decreasing NO levels in Arabidopsis through the use of chemical NO scavengers, genetically reducing nitrate reductase (NR) activity, or enhancing PHYTOGLOBIN 1 (PGB1) scavenging results in increased stability of ERFVIIs (Gibbs et al. 2014, 2018; Hartman et al. 2019). Previous studies have also reported that the elimination of NO through lower levels of NR and higher levels of PGB is a substantial determinant of tolerance in Brachypodium distachyon ecotype Bd2-3 compared to Bd21 (Rivera-Contreras et al. 2016). Additionally, increased expression of PGB has been shown to enhance survival in Arabidopsis under hypoxic conditions (Hunt et al. 2002) and improve the mitigation of reactive oxygen species during hypoxia and reoxygenation (Liu et al. 2022). Therefore, increasing the level of PGB gene expression to reduce NO levels and stabilize ERFVII proteins is crucial for survival under hypoxia (Hartman et al. 2019). In the PROTEOLYSIS 6 (PRT6) N-degron pathway in plants, the first methionine of ERFVII is removed by a methionine aminopeptidase, and then cysteine oxidases oxidize the exposed Cys to Cys-sulfinic acid (CysO₂) in a reaction that requires O₂. CysO₂ is then arginylated to Arg-CysO₂ by ARGINYL tRNA TRANSFERASE 1, which may depend on NO (Hu et al. 2005), and finally Arg-CysO₂ is recognized by E3 ligase PRT6 and degraded via the ubiquitin-proteasomal (26S proteasome) pathway (Graciet and Wellmer 2010). However, ERFVIIs are stabilized and activate hypoxia response genes (HRGs) under O₂-limited conditions (Licausi et al. 2011; Bailey-Serres et al. 2012). Members of the ERFVII family have been reported to trigger adaptative responses to hypoxic stress (Xu et al. 2006); therefore, understanding how the ERFVII family is regulated is essential to improve the flooding tolerance of cereal crops to minimize economic losses.

Similar to Arabidopsis ERFVIIs, rice ERFVIIs play essential roles in regulating the expression of HRGs under submergence (Fukao et al. 2006; Gibbs et al. 2011; Licausi et al. 2011). There are eighteen ERFVIIs in the rice genome and some of them are cultivar-specific, including SUB1A. The expression levels and alleles in different cultivars have been shown to be associated with submergence tolerance. SUB1A-1 from the tolerant cultivar FR13A is highly induced during submergence, whereas SUB1A-2 from the sensitive cultivar IR29 has a lower expression level (Lin et al. 2019). SUB1A-1 can directly regulate the expression of two ERFVIIs, OsERF66 and OsERF67, which in turn activate downstream HRGs. Overexpression of SUB1A-1, OsERF66 or OsERF67 in the submergence-sensitive cultivar TNG67 improves submergence tolerance (Lin et al. 2019). OsERF66 and OsERF67 are also induced in IR29 during submergence, but at much lower levels (Lin et al. 2019). This raises the possibility that OsERF66/OsERF67 are the central switches that relay upstream signals to turn on downstream HRGs during submergence.

In this study, we used Brachypodium distachyon, which was proposed as a model species for genetics and molecular genomics in cereals, to investigate whether similar regulatory cascades exist in other grass species. We identified three upstream regulators of HRGs in B. distachyon: BdERF108, BdERF018, and BdERF961. BdERF108 and BdERF018 acted as positive regulators, enhancing the expression of HRGs, including PGB1, during submergence. However, BdERF961 acted as a negative regulator, inhibiting the activation of PGB1 mediated by BdERF108 and BdERF018. Expression of BdERF961 was also activated by BdERF108 and BdERF018. In rice, OsERF67, the ortholog of BdERF961, was also found to suppress the activation of OsHB2 mediated by OsERF70 and OsERF71. However, OsERF67 exhibited an activation of OsHB2 expression, leading to its positive role in submergence stress (Lin et al. 2019). These findings indicate that ERFVIIs form distinct regulatory loops during submergence in rice and Brachypodium.

Results

Stability of BdERFVII proteins under aerobic conditions

In Arabidopsis and rice, ERFVIIs are regulated by PRT6 N-degron pathway-mediated proteolysis under aerobic conditions (Gibbs et al. 2011; Licausi et al. 2011). Since PRT6 N-degron-recognizing sites (MCGG) were found in all eight of the BdERFVII proteins (Graciet and Wellmer 2010) (Fig. 1A), we investigated whether they were also regulated by the PRT6 N-degron pathway. The half-lives of the BdERFVII proteins produced by a TnT T7 quick coupled transcription/translation system (Lee et al. 2005) were examined by in vitro analysis of the PRT6 N-degron pathway after application of the protein synthesis inhibitor cycloheximide (CHX) in the presence of O₂. All BdERFVII proteins, except BdERF053 and BdERF961, were degraded rapidly within 60 min after application of CHX (Fig. 1B). The half-lives of BdERF053 and BdERF961 were longer than 12 h (Fig. 1C), suggesting that they were tolerant to the degradation mediated by PRT6 N-degron pathway. When the second cysteine of PRT6 N-degron-recognizing sites (MC) was changed to alanine (MA), all MA form BdERFVII proteins exhibited increased stability by resisting Cys oxidation mediated by plant cysteine oxidases (Fig. 1B). To further investigate the protein stability of BdERFVIIs ex vivo, a pFRETgc-2in1-CC (Hecker et al. 2015) vector containing ERFVII:EGFP and luciferase, both driven by the CaMV 35S promoter (pro35S), was generated for transfection into B. distachyon (Bd21-3) protoplasts. As shown in Fig. 1D, there was no detectable BdERF108::GFP and BdERF018::GFP at 1 h, indicating they are substrates of the PRT6 N-degron pathway. In contrast, the protein level of BdERF961::GFP at 1 h was substantially higher with a half-life longer than 3 h, even higher than that of Sub1A-1::GFP, suggesting that BdERF961 is stable under aerobic conditions.

Figure 1.

Stability of BdERFVII proteins. A) Multiple sequence alignment of N-terminal BdERFVII amino acids by Clustal Omega, yellow color represents conserved sequences. B and C) In vitro analysis of BdERFVII protein degradation via the PRT6 N-degron pathway after application of CHX. Proteins were translated using the TNT T7 quick coupled transcription/translation system. Proteins stained with amido black in each were used as a protein loading control. D) Ex vivo analysis of BdERFVII protein degradation via the PRT6 N-degron pathway in Bd21-3 protoplast cells after application of CHX. Luciferase protein levels were used as protein expression and loading controls. CHX, cycloheximide; MC, Met-Cys for PRT6 N-degron-recognizing sites; MA, Met-ala, resistant to PRT6 N-degron pathway.

BdERF961 plays a negative regulatory role during submergence

It has been reported that ERFVII genes activate HRGs in response to hypoxic stress (Licausi et al. 2011; Gasch et al. 2016), which leads to reduced damage and adaptation to low O₂ conditions (Xu et al. 2006). To determine which BdERFVIIs respond to submergence stress, B. distachyon was subjected to submergence treatment and assayed for ERFVII transcript levels. Based on transcriptional profiles, these genes were divided into two groups: BdERF018/031/108/113 were constitutively expressed with or without submergence treatment (Fig. 2A), and BdERF032/044/053/961 were highly induced by submergence (Fig. 2B).

Figure 2.

Expression profiles of BdERFVII in Bd21-3 during submergence. B. distachyon seedlings were subjected to submergence for 1 to 6 h. Constitutive (A) and submergence-induced (B)BdERFVII transcript levels in shoots and roots were determined by RT-qPCR during submergence. UBIQUITIN (Yacoubi et al. 2022) was used as an endogenous control to normalize transcript levels. Average values (2^−ΔCq) were relative to Ct of UBI ± SE Asterisks indicate significant differences compared with the zero time point of submergence (*, P < 0.05; **, P < 0.01; ***, P < 0.001, T.TEST, n = 3). Transcript levels in shoot and root are indicated on top of each panel. x axes represent time under submergence.

Since OsERF66 and OsERF67 are the central transcription factors necessary for activation of HRGs and submergence tolerance in rice (Lin et al. 2019), we investigated whether any of the BdERFVIIs plays a similar role. Phylogenetic analysis showed that among eight BdERFVIIs, BdERF961 is most closely related to OsERF66 and OsERF67 (Fig. 3A). Furthermore, through synteny analysis, it was confirmed that BdERF961 and OsERF67 are orthologs (Fig. 3B). Additionally, sequence alignment unveiled several conserved motifs shared between these two proteins (Fig. 3C) and, similar to OsERF66 and OsERF67, BdERF961 was dramatically induced by submergence. Therefore, we decided to examine the function of BdERF961 in B. distachyon during submergence. The submergence tolerances of a T-DNA insertion mutant of BdERF961 (erf961) and the wild-type line Bd21-3 were compared. Phenotypic analysis and quantification of survival rate revealed that erf961 was more tolerant to submergence than Bd21-3 (Fig. 4, A and B). In addition, erf961 also had better tolerance to anoxia than Bd21-3 (Fig. 4, C and D). These results suggest that BdERF961 plays a negative regulatory role during low O₂ stress. Although erf961 had greater submergence tolerance, it produced significantly fewer lateral roots compared with Bd21-3 (Supplemental Fig. S1, A and B). The reduction of lateral root production in erf961 might be responsible for the observed lower seedling fresh weight and seed dry weight (Supplemental Fig. S1, C to E).

Figure 3.

Unrooted phylogenetic tree, synteny analysis and protein motifs of the ERFVII gene family. A) An unrooted phylogenetic tree was constructed using NGPhylogeny.fr to illustrate the genetic relationships of ERFVII proteins between Oryza sativa and Brachypodium distachyon. OsERF and BdERF denote ERFVII proteins from O. sativa and B. distachyon, respectively. The scale bar indicates 5% nucleotide sequence divergence. B) Synteny analysis of ERFVIIs genes, where orthologus genes are connected by arrows. Conserved genes are connected by lines. Bd21-3 represents Brachypodium distachyon. C) The comprehensive motif discovery and searching of BdERFVIIs and OsEFVIIs were conducted by the MEME SUITE. The presence of the same motif symbol among ERVIIs indicates their conservation within this group.

Figure 4.

BdERF961 plays a negative regulatory role in submergence stress. Analysis of the phenotypes (A) and (C) and survival rates (B) and (D) of Bd21-3 and erf961 Brachypodium distachyon plants subjected to submergence for 65 h or anoxia treatment for 9 h, followed by 7 to 10 days of recovery, white arrows indicate the survivals. Asterisks indicate significant differences between Bd21-3 and erf961 (*, P < 0.05; **, P < 0.01, T.TEST, n ≥ 3, error bars represent the standard error). E) The transcript levels of fermentative and PGB genes in Bd21-3 and erf961 during submergence determined using RT-qPCR. Average values relative to air-treated Bd21-3 ± SE are shown. x axes represent time under submergence, y axis represents the fold changes compared to Bd21-3 at 0 time point. Asterisks indicate significant differences compared with Bd21-3 at the same time point (*, P < 0.05; **, P < 0.01; ***, P < 0.001, T.TEST, n = 3). F) The transcript levels of fermentative and PGB genes in Bd21-3 protoplasts that were transfected with BdERF961MA-pGWB417 (BdERF961MA-OE) or empty pGWB417 (vector control). x axes represent time point of submergence, y axis represents the fold changes compared to vector control at 0 time point. Average values relative to vector control ± SE are shown. Asterisks indicate significant differences compared with vector control (*, P < 0.05; **, P < 0.01; ***, P < 0.001, T.TEST, n = 3).

To investigate how erf961 increased submergence tolerance, the transcript levels of HRGs were examined in erf961 and Bd21-3 during submergence. The transcript levels of ADH1 and PDC2, which are involved in fermentation necessary for energy production under hypoxic conditions (Ismond et al. 2003), were higher in erf961 than in Bd21-3 (Fig. 4E). Phytoglobin genes PGB1 and PGB2, which are involved in scavenging NO (Dordas et al. 2003), were dramatically more highly expressed in erf961 (Fig. 4E). As overexpression of the PGB gene in Arabidopsis was found to increase survival under hypoxic stress (Hunt et al. 2002), the enhanced PGB levels in erf961 might contribute to improved survival of Bd21-3 under submergence stress. Furthermore, Bd21-3 protoplasts transiently overexpressing BdERF961MA showed lower levels of PDC2, PGB1, and PGB2 expression than in those expressing an empty vector control (Fig. 4F). Taken together, these results suggest that BdERF961 functions as a negative regulator that inhibits the induction of HRGs during submergence.

BdERF108 and BdERF018 play positive roles during submergence

To search for BdERFVIIs that play positive roles during submergence, we examined the phenotypes of the B. distachyon ecotypes Bd2-3 and Bd1-1 and found that both were more submergence tolerant and had higher transcript levels of ADH, PDC, and PGB under submergence than Bd21-3 (Supplemental Fig. S2, A to C). The transcript levels of BdERF108 and BdERF018 in Bd2-3 and Bd1-1 under submergence were higher than those in Bd21-3, raising the possibility that they might function as positive regulators (Supplemental Fig. S2D). We generated homozygous/biallelic double mutants of BdERF108 and BdERF018 (erf108/018) using the CRISPR-Cas9-mediated-genome editing technique to further investigate their functions during submergence (Fig. 5A). The erf108/018 mutants were more sensitive to submergence stress (Fig. 5, B and C) and had lower transcript levels of PGB1 than wild-type Bd21-3 during submergence (Fig. 5D). To elucidate the regulatory mechanisms involved in BdERF108 and BdERF018 overexpression, we performed transient overexpression experiments in protoplasts. The activation levels of ADH1, PDC2, and PGB1 were also dramatically higher in Bd21-3 protoplasts transiently overexpressing BdERF108MA or BdERF018MA than in those expressing an empty vector control (Fig. 5E). However, it was noted that the expression of ADH1 and PDC2 remained unchanged in Bd21-3 and erf108/018 stable mutants during submergence (Fig. 5D), indicating the presence of an alternative regulation mechanism for these hypoxia-responsive genes. Taken together, these results indicated that BdERF108 and BdERF018 are positive regulators that promote HRG expression during submergence.

Figure 5.

BdERF108 and BdERF018 play a positive role in submergence stress. A) CRISPR/Cas9-mediated-genome editing of the BdERF108 and BdERF018 coding sequences. The minus sign (−) indicates nucleotide deletion and the plus sign (+) indicates nucleotide insertion. Red letters and dash symbol (−) indicate insertion and deletion, respectively. PAM is protospacer adjacent motif. The phenotypes (B) and survival rates (C) of Brachypodium distachyon Bd21-3 and erf108/018 homozygous-biallelic double mutants subjected to submergence for 50 h, followed by 7 days of recovery. Asterisks indicate significant differences between Bd21-3 and the erf108/018 mutants (**, P < 0.01, T.TEST, n ≥ 3, error bars represent the standard error). D) Relative expression levels of ADH1, PDC2, and PGB1 in Bd21-3 and erf108/018 during submergence examined by RT-qPCR. Average values relative to air-treated Bd21-3 ± SE are shown. Asterisks indicate significant differences compared with Bd21-3 at the same time point (*, P < 0.05; **, P < 0.01; ***, P < 0.001, T.TEST, n = 3). E) The transcript levels of ADH1, PDC2, PGB1, BdERF108, and BdERF018 in Bd21-3 protoplasts transfected with BdERF108MA-pGWB417 (108-OE), BdERF018MA-pGWB417 (018-OE), or empty pGWB417 (control). Average values relative to control ± SE are shown. Asterisks indicate significant differences compared with the control (*, P < 0.05; **, P < 0.01; ***, P < 0.001, T.TEST, n = 3). y axis scales are non-linear for ADH1, PDC2, and PGB1.

Interplay between BdERF961 and BdERF108/BdERF018

Since PGB is crucial for survival under hypoxia (Hunt et al. 2002), and PGB1 expression was oppositely regulated by BdERF961 and BdERF108/BdERF018 (Figs. 4, E and F and 5, D and E), a dual-luciferase reporter assay based on PGB1 promoter-driven firefly luciferase was used to further investigate whether the other BdERFVIIs regulate PGB1 (Fig. 6A). With the exception of BdERF961 and BdERF053, all other BdERFVIIs could activate PGB1 expression, with BdERF018 and BdERF108 having the highest activation activities (Fig. 6B). To investigate if PGB1 expression was coregulated by BdERF961, BdERF018, and BdERF108, a PGB1 reporter assay was conducted by cotransfecting with respective BdERF961 and BdERF108, or BdERF961 and BdERF018. The activation activities of BdERF018 and BdERF108 were significantly reduced and nearly abolished when they were coexpressed with BdERF961 (Fig. 6C), suggesting that BdERF961 competes with BdERF018 and BdERF108 for promoter binding. To test this hypothesis, chromatin immunoprecipitation (ChIP)-qPCR assays were performed to identify their respective binding sites within the PGB1 promoter in vivo. BdERF108, BdERF018, and BdERF961 mainly bound to the P3 fragment, which contains a GCC 4 box near the ATG start codon (Fig. 6D). When a modified PGB1 promoter containing a mutated GGC 4 box (from GCCGCC to GCCTCC) in the P3 region was used in the dual-luciferase reporter assay, BdERF108- and BdERF018-mediated activation of PGB1 was dramatically decreased (Fig. 6E). However, BdERF108- and BdERF018-mediated activation was not significantly reduced by the addition of the BdERF961 effector in the presence of the modified PGB1 promoter (Fig. 6E). These results demonstrated that BdERF018, BdERF108, and BdERF961 directly regulate the expression of PGB1 by competing for the same promoter region.

Figure 6.

BdERF108 and BdERF018 activate proPGB1::luc, while BdERF961 represses it. A) Diagram of constructs used for the dual-luciferase reporter assay system: pGreenII 0800-LUC reporter containing proPGB1; effectors with/without pro35S::BdERFVIIs-MA:myc. B and C) The reporter along with the indicated effectors were cotransfected into Bd21-3 protoplasts for reporter assays. Average values of relative luciferase activity ± SE are shown; activity of a reporter cotransfected with empty effector was set to 1.0 (control). Asterisks indicate significant differences compared with the control (B) or with/without coexpression of BdERF961MA (C) (*, P < 0.05; **, P < 0.01; ***, P < 0.001, T.TEST, n ≥ 3). D) ChIP-qPCR analysis revealed the ability of BdERF108MA/018MA/961MA/053A to bind to the PGB1 promoter. Relative enrichment of PGB1 promoter fragments was determined by ChIP-qPCR and normalized against the value of normal rabbit IgG. Average values ± SE are shown; letters indicate significant differences (P < 0.05, one-way ANOVA, Tukey B, n ≥ 3). E) Reporter containing proPGB1 with or without deletion of the 4th GCC box together with indicated effectors used for reporter assays. Average values of relative luciferase activity ± SE are shown; a reporter cotransfected with the empty effector was set to 1.0 (control). Letters indicate significant differences (P < 0.05, one-way ANOVA, Tukey B, n ≥ 3).

Increasing the level of PGB in erf961 decreases NO content and inhibits PRT6 N-degron-targeted proteolysis

Class 1 PGB genes are induced by hypoxic stress and modulate NO concentration in plants (Hartman et al. 2019; Becana et al. 2020; Manrique-Gil et al. 2021). Moreover, augmenting the scavenging activity of PGB results in heightened stability of ERFVIIs, thereby enhancing the plant's ability to tolerate hypoxic conditions (Hartman et al. 2019). Hence, we examined whether the increased PGB1 can influence the NO level in erf961. We compared the cellular NO levels in Bd21-3 and erf961 and found that under normoxia NO was detected in Bd21-3 but barely detectable in erf961 (Fig. 7A). Upon submergence treatment, the NO level increased significantly in Bd21-3 but only slightly in erf961. As a result, the NO level in Bd21-3 was substantially higher than that in erf961 during submergence (Fig. 7A). The lower NO content in erf961 might be caused by the higher level of induction of PGB1 (Fig. 4E), since NO is oxidized to nitrate (NO₃⁻) by PGB in the PGB/NO cycle (Igamberdiev and Hill 2004).

Figure 7.

Nitric oxide content decreased in erf961 but increased in erf108/018 compared with Bd21-3. A and C) Nitric oxide (NO) content was visualized and quantified by fluorescence microscopy with a fluorescent probe, DAF-FM DA, in Bd21-3 and erf961(A) and erf108/018(C) seedling root tips treated with/without submergence. Letters indicate significant differences (P < 0.05, one-way ANOVA, Duncan, n ≥ 9, error bars represent the standard error). B and D) Ex vivo analysis of PRT6 N-degron pathway-mediated degradation of the substrate MCGG:GFP in Bd21-3 and erf961(B) and erf018/108(D) protoplasts, the last four lanes were coexpressed with BdERF961MA:myc(B) and PGB1:myc(D), respectively. MCGG:GFP protein levels were detected by immunoblot analysis with anti-GFP antibody after cycloheximide (CHX) treatment. E) Ex vivo analysis of PRT6 N-degron pathway-mediated degradation of the substrate MCGG:GFP in erf018/108 protoplasts coexpressed with PGB1:myc. Anti-myc antibody was used to detect expression of pro35S::BdERF961MA:myc(B) or pro35S::PGB1:myc(D and E). Luciferase protein levels were used as protein expression and loading controls.

Several studies have shown that NO and O₂ are essential for PRT6 N-degron-mediated proteolysis of ERFVII and that increased levels of PGB enhance the stability of ERFVII during normoxia (Gibbs et al. 2014; Hartman et al. 2019). Therefore, a reduction of NO content in erf961 might lead to inhibition of the PRT6 N-degron-targeted proteolysis of BdERFVII. To test this hypothesis and exclude the effect of O₂ supply, a protein degradation assay using the Nt-Cys-containing substrate MCGG:GFP was performed using erf961 and Bd21-3 protoplasts during normoxia. Immunoblotting revealed that MCGG:GFP proteins were more abundant and stable in erf961 than in Bd21-3 protoplasts after the addition of CHX. Furthermore, the level of MCGG:GFP protein was reduced by the expression of the BdERF961MA:myc in erf961 (Fig. 7B). This suggests that BdERF961 represses the expression of PGB1, leading to an increased NO content and promotion of the PRT6 N-degron pathway-mediated degradation of BdERFVII. Consistent with this observation, erf961 exhibited a higher level of PGB1 expression (Fig. 4E) and lower NO content compared to Bd21-3 (Fig. 7A). In contrast, the transcript levels of PGB1 were reduced in erf108/018 mutants (Fig. 5D), resulting in higher NO content than that in Bd21-3 under normoxia, but no significant difference was observed in NO content under submergence (Fig. 7C). It is noteworthy that an increase in NO content of erf108/018 corresponded with higher proteolysis of MCGG:GFP compared to Bd21-3 during normoxia, whereas coexpression with PGB1:myc partially restored the protein levels and stability of MCGG:GFP (Fig. 7, D and E). This indicated that the increasing abundance of PGB1 protein is crucial for reducing the levels of NO and consequently stabilizing MCGG:GFP.

BdERF108 and BdERF018 activate the expression of other BdERFVIIs

In addition to the HRGs ADH1 and PGB1, two BdERFVIIs, BdERF961 and BdERF053, were also upregulated by BdERF108 and BdERF018. Transcript levels of BdERF961 and BdERF053 in the erf108/018 mutants were significantly reduced during submergence treatment (Fig. 8A), and expression of these genes was dramatically higher in Bd21-3 protoplasts expressing BdERF108MA or BdERF018MA (Fig. 8B). Moreover, ChIP-qPCR assays revealed that BdERF108 and BdERF018 mainly bound to the Pa fragment of the BdERF961 promoter (Fig. 8C). Interestingly, BdERF961 also bound to the Pa fragment of its own promoter, implying that BdERF961 may regulate its own expression. To further confirm ChIP-qPCR results, the GCC 2 box of the Pa fragment in the BdERF961 promoter was mutated and used for transient reporter assays. The results revealed that the GCC 2 box of BdERF961 promoter was crucial for activation and repression by BdERF108 and BdERF961, respectively; however, partial activation of BdERF961 by BdERF018 was observed even when the GCC 2 box was mutated (Fig. 8D). These results revealed that BdERF108, BdERF018, and BdERF961 directly regulate the expression of BdERF961 by directly binding to the promoter region.

Figure 8.

BdERF961 is directly regulated by BdERF108 and BdERF018. A) Relative transcript levels of BdERF961 and BdERF053 in Bd21-3 and erf108/018 during submergence determined by RT-qPCR. x axes represent time point of submergence, y axis represents the fold changes compared to Bd21-3 at 0 time point. Average values relative to air-treated Bd21-3 ± SE are shown. Asterisks indicate significant differences compared with Bd21-3 at the same time point (*, P < 0.05; **, P < 0.01; ***, P < 0.001, T.TEST, n = 3). B) The transcript levels of BdERF961 and BdERF053 in Bd21-3 protoplasts transfected with BdERF108MA-pGWB417 (108-OE), BdERF018MA-pGWB417 (018-OE), or empty pGWB417 (control). x axes represent time under submergence, y axis represents the fold changes compared to vector control at 0 time point. Average values relative to control ± SE are shown. Asterisks indicate significant differences compared with the control (*, P < 0.05; ***, P < 0.001, T.TEST, n = 3). C) ChIP-qPCR analysis of the binding of BdERF108MA/018MA/961MA/053MA to the BdERF961 promoter. Relative enrichment of BdERF961 promoter fragments was determined by ChIP-qPCR and normalized against the value of normal rabbit IgG. Average values ± SE are shown. Letters indicate significant differences (P < 0.05, one-way ANOVA, Tukey B, n ≥ 3). D) pGreenII 0800-LUC reporter containing proBdERF961 with or without the 2nd GCC box was cotransfected into protoplasts with the indicated effectors for the reporter assay. Average values of relative luciferase activity ± SE are shown. A reporter cotransfected with empty effector was set to 1.0 (control). Letters indicate significant differences (P < 0.05, one-way ANOVA, Tukey B, n ≥ 3).

The transient reporter assays revealed that BdERF053 was upregulated by BdERF108 and BdERF018 but downregulated by BdERF961 (Supplemental Fig. S3A). In addition, BdERF053 also acted as a repressor to inhibit BdERF108/018-mediated activation of PGB1 (Supplemental Fig. S3B). Furthermore, a BdERF053 T-DNA mutant (erf053, JJ13579) and Bd21-3 protoplasts transiently overexpressing BdERF053MA had respectively higher and lower levels of PGB1 expression during hypoxia (Supplemental Fig. S4, A and B). ChIP-qPCR and transient reporter assays also demonstrated that BdERF053 can directly bind to the P2 fragment of the PGB1 promoter (Fig. 6D and Supplemental Fig. S3B). Similar to erf961, erf053 exhibited greater tolerance to submergence than Bd21-3 (Supplemental Fig. S4C), and BdERF053 was also found to be essential for Brachypodium root and seed development (Supplemental Fig. S4, D and E). Interestingly, BdERF053MA bound to the Pa fragment of the BdERF961 promoter and repressed its transcription as well (Fig. 8C, Supplemental Figs. S3C and S4B). These results suggested that BdERF053 played a similar role with BdERF961.

Overall, our results suggested that BdERF961 and BdERF108/BdERF018 play negative and positive roles in submergence stress, respectively. BdERF108/BdERF018 activate the expression of BdERF961, BdERF053, and PGB1 under hypoxic conditions. An increased level of PGB1 facilitates the turnover of NO, thereby delaying protein degradation of BdERFVII; however, this activation of PGB1 is decreased by a gradual increase in BdERF961 and BdERF053 soon after (Supplemental Fig. S5).

Conserved ERFVIIs forming distinct regulatory loops mediate submergence responses in Brachypodium and rice

To explore the conservation of the regulatory loop mediated by BdERF961, BdEF108, and BdERF018 in rice, we conducted an examination of this regulatory loop in the rice orthologs. Phylogenetic and motif analyses showed that rice OsERF70 and OsERF71 are closely related to BdERF108 and BdERF018 (Fig. 3). Similar to their counterparts in Brachypodium, OsERF70 and OsERF71 were constitutively expressed under normoxia and during submergence treatment, and the PGB gene OsHB2 was highly induced in TNG67 under submergence (Fig. 9A). OsERF70/71 also strongly activated the expression of OsHB2 in transient reporter assays using TNG67 protoplasts (Fig. 9B), and this OsERF70/71-mediated activation of OsHB2 was decreased by coexpression with OsERF67 and OsERF61, the respective orthologs of BdERF961 and BdERF053 (Fig. 9, C and D). Furthermore, reporter assays revealed that the expression of OsERF67 and OsERF61 was also upregulated by OsERF70 and OsERF71 (Supplemental Fig. S6A). Surprisingly, the expression of OsERF61 was downregulated by OsERF67 (Supplemental Fig. S6A); moreover, OsERF70/71-mediated activation of OsERF61 was inhibited by coexpression with OsERF67 (Supplemental Fig. S6B). Based on these findings, we observed that the regulatory loop mediated by OsERF67, OsERF70, and OsERF71 is conserved between rice and Brachypodium. However, it is noteworthy that OsERF67 and OsERF61 serve as positive regulators for OsHB2.

Figure 9.

OsERF61/67 inhibit OsERF70- and OsERF70-enhanced activation of proOsHB2::Luc. A) Rice TNG67 seedlings treated with submergence for 1 to 6 h; the gene expression profiles in roots during submergence were determined by RT-qPCR. TUBULIN (TUB) was used as an endogenous control to normalize transcripts. x axes represent time point of submergence. Average values (2^−ΔCq) are relative to Ct of TUBULIN ± SE. Asterisks indicate significant differences compared with the zero time point of submergence (*, P < 0.05; **, P < 0.01, T.TEST, n = 3). B to D) A reporter containing proOsHB2 and the indicated effectors were cotransfected into TNG67 protoplasts for reporter assays. Average values of relative luciferase activity ± SE are shown. Activity of a reporter cotransfected with empty effector was set to 1.0 (control). Asterisks indicate significant differences compared with control (B) or with/without coexpression of OsERF67MA(C) or OsERF61MA(D) (*, P < 0.05; **, P < 0.01; ***, P < 0.001, T.TEST, n ≥ 3).

Discussion

Under limited O₂ conditions most ERFVIIs activate HRGs to trigger downstream responses; however, this transcriptional activation needs to be fine-tuned to save carbohydrate reserves. For example, Arabidopsis requires the trihelix transcription factor HRA1 to counteract the induction of RAP2.12-mediated hypoxia responses (Giuntoli et al. 2014). Here, we demonstrated that Brachypodium also has an excellent ability to prevent the overproduction of submergence-induced genes via feedback inhibition. BdERF108 and BdERF018 activated the expression of ADH1, PDC2, and PGB1 during submergence (Fig. 5, D and E). However, BdERF108- and BdERF018-upregulated BdERF961 decreased the expression of ADH1, PDC2, and PGB1 (Fig. 4, E and F). As BdERF961 lacks any known repression domain (Licausi et al. 2013), it exerts repression on PGB1 activation mediated by BdERF108 and BdERF018 through binding to the same promoter region (Fig. 6, C to E). We hypothesize that BdERF961 acts as a passive repressor by competing with BdERF108 and BdERF018 to bind to the promoter of HRGs, functioning to maintain dynamic homeostasis of hypoxia-induced responses during submergence. Although the erf961 T-DNA mutant was more tolerant to submergence than wild-type Bd21-3 (Fig. 4, A and B), it had fewer lateral roots, a lower seedling fresh weight, and a lower seed dry weight than Bd21-3 (Supplemental Fig. S1), implying that BdERF961 is essential for normal growth. In sum, these three BdERFVIIs form an antagonistic regulatory loop to maintain the balance between submergence adaption and normal growth.

Similar to BdERF961, BdERF053 is also involved in reducing PGB1 expression and submergence tolerance (Supplemental Fig. S4, A to C). Furthermore, the effect of BdERF053 on root and seed development was similar to that of BdERF961 (Supplemental Fig. S4, D and E), and there was mutual repression between BdERF053 and BdERF961 when they were overexpressed (Fig. 4F and Supplemental Fig. S4B). These results suggest that BdERF053 and BdERF961 share redundant functions. The regulatory network involving PGB in rice was conserved with that in Brachypodium; OsERF70/71-mediated activation of the OsHB2 promoter was decreased by coexpression with OsERF61 or OsERF67 (Fig. 9, C and D). OsERF61 and OsERF67 themselves even slightly increased activation of the OsHB2 promoter compared with the control (Fig. 9B). This implies that OsERF61 and OsERF67 are minor activators competing with the strong activators OsERF70 and OsERF71 for the same promoter region of OsHB2, resulting in a lower level of activation of OsHB2. In a similar manner to the inhibition of BdERF053 by BdERF961 (Supplemental Fig. S3A), OsERF67 also repressed OsERF61 expression (Supplemental Fig. S6), suggesting that OsERF67 has as a dual role as a transcriptional regulator, inhibiting OsERF61 and activating OsHB2. Even though OsERF67 positively regulated OsHB2 (Fig. 9B), its ortholog in Brachypodium, BdERF961, negatively regulated this PGB gene (Figs. 6B and 10). Rice usually grows in semi-aquatic conditions, whereas Brachypodium was originally grown in a dry environment (Scholthof et al. 2018). Thus, these opposing regulatory mechanisms between OsERF67 and BdERF961 may be caused by adaption to a variety of environmental conditions during evolution (Anderson et al. 2011).

Figure 10.

Proposed model for the regulation of BdERFVIIs and OsERFVIIs under submergence. Distinct regulatory loops mediated by ERFVIIs have been conserved throughout evolution and facilitate submergence responses in both Brachypodium and rice. Sub1A-1 has been reported to activate OsERF67 and led to increased HRG expression in FR13A. In our study, PGB expression was dramatically activated by BdERF108, BdERF018, OsERF70, and OsERF71; however, this activation was repressed by BdERF961, BdERF053, OsERF67, and OsERF61 soon after. In addition, the expression of OsERF61 and BdERF053 was repressed by OsERF67 and BdERF961, respectively. Interestingly, OsERF67 and OsERF61 alone can slightly activate PGB expression. Furthermore, an increased amount of PGB converts NO into nitrate (NO₃⁻) through the PGB-NO cycle, leading to the inhibition of PRT6 N-degron-mediated proteolysis and an increase in ERFVII stability in Brachypodium. Black cap lines and red arrows represent inhibition and activation, respectively.

With the exception of SUB1A-1, all ERFVIIs have been considered to be PRT6 N-degron pathway substrates up until now, and their proteins are degraded rapidly when exposed to oxygen (Lin et al. 2019). Moreover, they mostly function as positive regulators to trigger adaptive responses under hypoxia (Gibbs et al. 2011; Licausi et al. 2011; Fukao et al. 2019; Lin et al. 2019; Perata 2020). Here, we found that BdERF961 had a prolonged half-life even in the presence of O₂ (Fig. 1), and it acted as a negative regulator to inhibit the induction of HRGs during submergence (Fig. 4, E and F). This suggests BdERF961 is an PRT6 N-degron-resistant ERFVII protein and may play an important role in the prevention of excessive hypoxic response following the onset of hypoxia.

Continuously expressed ERFVIIs have been thought to be crucial factors in hypoxia tolerance, since their proteins are stabilized immediately when hypoxia occurs to trigger the expression of downstream HRGs as soon as possible to minimize damage (Gibbs et al. 2011; Licausi et al. 2011; Leon et al. 2020). Here, we discovered that BdERF018 and BdERF108 had high expression levels and that their expression persisted after submergence (Fig. 2A); however, their proteins were degraded rapidly in the presence of O₂, which prevented their overaccumulation (Fig. 1, B and D). In terms of function, BdERF108 and BdERF018 were shown to be essential for PGB1 expression (Fig. 5, D and E). As a result of the BdERF108- and BdERF018-mediated higher expression levels of PGB1, a positive regulatory loop might be formed to scavenge NO and lead to stabilize the BdERF108 and BdERF018 proteins further. However, this loop was interrupted by BdERF961, which acts as a repressor competing for the same GCC box of the PGB1 promoter with BdERF108 and BdERF018 (Fig. 6, D and E). Noticeably, BdERF108 activated higher levels of expression of PGB1 and BdERF961 compared with BdERF018 (Figs. 6E and 8D). Moreover, the protein stability of BdERF108 was also slightly higher than that of BdERF018 (Fig. 1, B and D). These data imply that BdERF108 is the main regulator of the expression of PGB1 and BdERF961. In addition, PRT6 N-degron non-degradable MA forms of BdERF018 and BdERF108 were also unstable during CHX treatment (Fig. 1, B and D), suggesting that there is another process involved in the degradation of these proteins. In TNG67, the orthologs of BdERF108 and BdERF018, OsERF70 and OsERF71, respectively, were continuously expressed during hypoxia (Fig. 9A), and the encoded proteins shared conserved motifs with BdERF108 and BdERF018 (Fig. 3B). Hence, OsERF70 and OsERF71 may also act as positive regulators during submergence.

Although the PGBs play a crucial role in rapidly reducing the level of NO during early hypoxia and ensuring the stability of ERFVII proteins to activate HRGs (Gupta et al. 2005; Planchet et al. 2005; Hartman et al. 2019), the upstream regulatory mechanism of PGBs in cereals is still unclear. Our study showed that the expression of PGB1 was directly upregulated and downregulated by BdERF108/018 and BdERF961, respectively, to adjust the level of NO (Figs. 4F, 5E, 6D, and 7, A and C) in Brachypodiun during submergence. Importantly, the complementary expression of PGB1 protein in erf108/018 showed a positive association with the protein levels and stability of the PRT6 N-degron pathway target protein (MCGG:GFP) during normoxia (Fig. 7, D and E). These findings strongly suggest that BdERF108- and BdERF018-mediated activation of PGB1 indeed contributed to NO scavenging and repressed the N-degron pathway-mediated degradation of BdERFVII. Furthermore, it was demonstrated that the submergence-tolerant ecotype Bd2-3 had lower transcript levels of NR under submergence compared to Bd21 (Rivera-Contreras et al. 2016). Additionally, Bd2-3 exhibited higher transcript levels of HEMOGLOBIN (PHYTOGLOBIN) in both normoxia and submergence conditions. These findings further support the notion that the efficient removal of NO plays a pivotal role as a tolerance determinant. The rice putative ortholog of PGB1, OsHB2, is activated by OsERF70, OsERF71, OsERF67, OsERF66, as well as submergence treatment (Lin et al. 2019), suggesting that OsHB2 may have a similar role to Brachypodium PGB1. Arabidopsis putative orthologs of PGB1 and PGB2, namely AtPGB1 (AT2G16060) and AtPGB3 (AT4G32690), respectively, are activated by ethylene and NO (Mukhi et al. 2017; Hartman et al. 2019). Remarkably, elevated levels of AtPGB1 have been shown to enhance ERFVII stability and promote hypoxia survival (Hartman et al. 2019), whereas AtPGB3 plays a crucial role in defense response against Sclerotinia sclerotiorum (Mukhi et al. 2017), indicating distinct biological functions for these PGBs. In addition, an increase in NO content in rice was found to promote the growth of lateral roots (Sun et al. 2018), which may explain why erf961, which has a low NO content, had fewer lateral roots than Bd21-3 (Fig. 7A, Supplemental Fig. S1, A and B). As a result of the low NO content in erf961, the continuously expressed BdERF108 and BdERF018 proteins may become stable and activate PGB1 expression, forming a positive regulatory loop to further reduce NO content. Taken together, our findings demonstrate that Brachypodium has an excellent ability to fine-tune the degradation of ERFVII proteins through ERFVII-targeted PGB1.

The divergent regulatory mechanisms exhibited between orthologs, OsERF67 and BdERF961, can be attributed to their evolutionary adaptation to diverse environmental conditions (Anderson et al. 2011). This revelation of orthologous ERFVIIs serving as positive regulator in rice and negative regulator in Brachypodium offers important insights into the enhancement of submergence tolerance in crop breeding. Consequently, it presents valuable perspectives for the development of submergence-tolerant crops.

Materials and methods

Plant materials

Seeds of Brachypodium distachyon ecotypes Bd21-3, Bd2-3, Bd-1, and T-DNA insertion lines JJ22103 (erf961), and JJ13579 (erf053) were obtained from DOE Joint Genome Institute (JGI), Lawrence Berkeley National Laboratory (Bragg et al. 2012). Brachypodium seeds were immersed in 75% (v/v) ethanol for 1 min and rinsed once with sterile water. The lamella and palea were carefully removed with forceps, followed by sterilization with 10 mL of 1% sodium hypochlorite and 0.1% Tween 20 for 20 min and five washes with 10 mL of sterile water. The sterilized seeds were cultured horizontally in a 9 cm petri dish with Whatman qualitative filter paper No. 1 containing 3 mL sterile water or were cultured vertically on Murashige and Skoog (Catalan et al. 1997) medium (Duchefa Biochemie, M0231) containing 1% sucrose, 5 g/L phytagel, pH 5.7, and then incubated in dark at 4 °C for 7 days before germination and growth under long-day conditions at 23 °C (16 h light/8 h dark, photon flux density = 150–180 μmol m⁻² s⁻¹, RH = 50%). Three-day-old seedlings were transferred from the petri dish to soil (Jiffy, packaging [EN12580] plus 1:1:6 v/v/v vermiculite: perlite: substrate) for survival assays under submergence. Seven-day-old seedlings cultured vertically on MS medium were transferred to new MS medium for another 3 days, followed by submerging in 1/2 MS medium to analyze transcriptional profiles of BdERFVIIs and HRGs.

Rice (O. sativa subsp. japonica) cultivar TNG67 was used in this study, and the growth conditions were as described in Lin et al. (2019).

Plasmid constructs

To generate constructs for the TNT T7 quick coupled transcription/translation system (Promega, L1170), BdERFVII coding sequences (CDSs) were cloned into pTNT_4xMYC containing the T7 and SP6 promoters and C-terminal 4x myc tag (Lin et al. 2019). For protein degradation assays in Brachypodium protoplasts, the CDSs of BdERFVIIs, SUB1A-1, and Nt-Cys-mediated proteolysis substrate (MCGGAIIPDCIPEH) were cloned into the P3 and P2 sites of pFRETgc-2in1-CC in frame with EGFP; the firefly luciferase CDS from pGreenII 0800-LUC was cloned into the P1 and P4 sites of pFRETgc-2in1-CC for use as a protein expression and loading control. The constructs above were generated using Invitrogen Gateway recombination cloning technology. For the dual-luciferase reporter assay system constructs, MA forms of BdERFVII CDSs were individually cloned into pGWB417 for use as effectors using Gateway recombination cloning technology; the upstream sequences preceding the translational start site (1.5 kb for proPGB1, proBdERF053, proOsERF61, proOsERF67, and proOsHB2; 0.6 kb for proBdERF961) were cloned into pGreenII 0800-LUC with specific restriction enzyme sites for use as reporters, the restriction enzyme used is labeled on the primer name (Hellens et al. 2005). The primers used for construction are presented in Supplemental Table S3.

Analysis of the PRT6 N-degron pathway

In vitro analysis of the PRT6 N-degron pathway was conducted using substrates obtained using the TnT T7 quick coupled transcription/translation system (Lee et al. 2005) (Promega, L1170). pTNT_4xMYC containing BdERFVII coding sequences were used to perform in vitro transcription and translation at 30 °C for 30 min, followed by application of 100 μM CHX to stop the translation. Samples were then incubated at room temperature for 30–60 min in the presence of O₂. An equal volume of 2X SDS-PAGE sample buffer was added to the translated protein samples, and then proteins were denatured at 70 °C for 10 min before immunoblot analysis.

For ex vivo analysis of the PRT6 N-degron pathway, transfected protoplasts were centrifuged at 100 rcf for 7 min, and W5 solution was removed and replaced with 100 μL fresh W5 solution containing 100 μM CHX followed by incubation at room temperature for 1 to 4 h in presence of O₂. Protoplasts were harvested by centrifuging at 100 rcf for 3 min, and the W5 solution was replaced with 40 μL 2X SDS-PAGE sample buffer. Proteins were denatured at 70 °C for 10 min before immunoblot analysis.

Experimental setup and treatments

For transcriptional profiling of BdERFVIIs during submergence, 10-day-old Brachypodium seedlings were cultured vertically on MS medium submerged in 20-cm deep 1/2 MS medium (pH 5.7) supplied with 3% O₂ in the dark for 1 to 6 h. Whole seedlings were harvested for further RT-qPCR analysis (Cho et al. 2019).

For submergence tolerance, 3-week-old Brachypodium plants grown in a 3-inch round plastic pot containing soil were submerged in 25-cm deep water in the dark for 50 to 80 h, and then recovered under normal growth conditions for 7 to 10 days. Plants treated in the same tank were used to calculate the survival rate. Brachypodium plants grown under normal conditions were used as a control.

For anoxia tolerance, 6-day-old Brachypodium seedlings cultured vertically on MS medium were treated with anoxia in air-tight jars containing a pack of O₂ absorbent (GasPak EZ Anaerobe Container System Sachet, Becton, Dickinson and Company, 260678) and filled with nitrogen gas for 5 min (Hsu et al. 2013). After 9 h of anoxia treatment in the dark, treated plants were allowed to recover under normal growth conditions for 1 week, and the survival rate was calculated from treated plants in the same jar. Brachypodium seedlings grown under normal conditions were used as a control.

Protoplast isolation and transformation

The protocol for Brachypodium protoplast isolation was based on procedures described in Jung et al. (2014) and Hong et al. (2012). Healthy 1st and 2nd true leaves (0.2 g) were excised from 3-week-old soil-grown Brachypodium plants and finely chopped using a sharp razor blade in 15 mL of TVL solution (0.3 m mannitol, 50 mm CaCl₂). Leaf pieces were transferred to a 9 cm petri dish containing 20 mL of filter-sterilized enzyme digestion solution (0.6 m mannitol, 10 mm MES [pH 5.7], 1.5% cellulose RS (Yakult Pharmaceutical), 0.75% macerozyme R-10 (Yakult Pharmaceutical), 0.1% bovine serum albumin, 1 mm CaCl₂) through a 0.45 µm syringe filter (Pall Corporation, 4614). The enzyme digestion solution was shaken at 40 rpm for 3 h, followed by shaking at 60 rpm for another 30 min to increase protoplast release. Released protoplasts were collected by passing through a 40 µm cell strainer (Falcon, 352340); one volume of W5 solution (154 mm NaCl, 125 mm CaCl₂, 5 mm KCl, 2 mm MES, pH 5.7) was added with gentle mixing, and then the mixture was centrifuged at 100 rcf for 7 min to harvest protoplasts. Protoplasts were resuspended in 10 mL W5 solution, and cell number was quantified using a hemocytometer with a bright-field microscope. Protoplasts were incubated on ice for 30 min, followed by centrifuging at 100 rcf for 7 min. The pelleted protoplasts were resuspended in MMG solution (0.4 m mannitol, 15 mm MgCl₂, 4 mm MES-KOH, pH 5.7) adjusted to 2.5 * 10⁶ cells/mL and then incubated with effector or reporter plasmids on ice for 10 min (200 μL volume of MMG containing 5 to 10 μg plasmid for each reaction). One volume of PEG solution (0.2 m mannitol, 100 mm CaCl₂, 40% PEG-4000) was added to the mix by gently inverting the tube. The transfection mixture was incubated at room temperature for 20 min, and then the reaction was terminated by adding 2 volumes of W5 solution. Transfected protoplasts were centrifuged at 100 rcf for 7 min and resuspended in 1 mL fresh W5 solution containing 5 mm glucose and transferred to a 12 well-tissue culture plate (JET Biofil, TCP011012), which was pretreated with 1% BSA, and incubated in the dark at 23 °C for 17 h.

The rice protoplast isolation and transformation were conducted following the procedures described by Lin et al. (2018).

RNA isolation and RT-qPCR

RNA was extracted from seedlings and protoplasts (2.5 * 10⁶ cells) using TRIzol reagent (Invitrogen, 15596026) and digested with TURBO DNase (Invitrogen, AM2238) at 37 °C for 30 min. Total RNA (2 μg each) was used to synthesize cDNA using M-MLV reverse transcriptase (Invitrogen, 18057018). RT-qPCR was performed in a C1000 touch thermal cycler (BIO-RAD) using SYBR green PCR master mix (Applied Biosystems, 4309155) and 5 μL of a 1:20 cDNA dilution (15 μL total volume). The relative gene expression (2^(ΔCq)) per sample was normalized to the Cq value of UBI; the relative fold change (2^(ΔΔCq)) per sample was compared with the ΔCq value of the control sample.

Immunoblot analysis

The proteins from in vitro and ex vivo translation were adjusted to the same concentration as SDS-PAGE sample buffer for immunoblot analysis as described by Hsiao et al. (2017).

The dual-luciferase reporter assay system

pGreenII 0800-LUC vectors containing the indicated promoters were cotransfected into Bd21-3 protoplasts with effectors, and reporter activity assays were performed by following the procedures in the Dual-Luciferase Reporter Assay System instructions (Promega, E1960). The value of firefly luciferase (Fir) in each sample was normalized to the value of Renilla luciferase (Ren), and the relative luciferase activity of each sample was normalized to Fir/Ren ratio of the control sample expressing the empty effector vector.

ChIP-qPCR assay

ChIP-qPCR was conducted following the procedures in the Pierce Magnetic ChIP Kit (Thermo Scientific). The BdERF108MA-, BdERF018MA-, BdERF961MA-, and BdERF053MA-pGWB417 constructs were transfected into Bd21-3 protoplasts. DNA/protein complexes were isolated using the anti-Myc antibody. Relative enrichment of PGB1 or BdERF961 promoter fragments was determined by ChIP-qPCR and normalized against the value of normal rabbit IgG.

NO quantification in roots

Quantification of NO was conducted following the procedures described in Hartman et al. (Sun et al. 2018; Hartman et al. 2019). Ten-day-old seedlings cultured vertically on Murashige and Skoog (Catalan et al. 1997) medium were treated with/without submergence in 1/2 MS medium for 1 h. Roots were excised immediately and subsequently incubated in 20 mm HEPES-NaOH buffer (pH 7.5) containing 10 μM diaminofluorescein-FM diacetate (DAF-FM DA, Sigma-Aldrich, D1946) in the dark for 30 min. After incubation, roots were washed three times for 5 min with fresh 20 mm HEPES-NaOH buffer and visualized by a fluorescence stereomicroscope (Zeiss Stereo Lumar V12) with excitation at 488 nm and emission at 490 to 555 nm using 0.8 objective magnification, −1,938 focus position, 59 ms exposure time. NO was quantified using ImageJ software, and relative fluorescence was normalized to the normoxia control of Bd21-3.

Unrooted phylogenetic tree, synteny analysis, and protein motifs of the ERFVII gene family

The complete amino acids sequences of ERFVIIs obtained from the Rice Genome Annotation Project and JGI Phytozome websites were utilized to construct an unrooted phylogenetic tree. Sequences with or without the AP2 domain were utilized to construct a tree using the PhyML/OneClick tool available on the NGPhylogeny.fr website. The workflow involved the following steps: (i) Input Fasta format data, (ii) multiple alignment of nucleotide sequences using MAFFT, (iii) cleaning of aligned sequences using BMGE, (iv) phylogenetic analysis based on maximum-likelihood using PhyML, and (v) displaying the resulting phylogenetic tree in SVG format using Newick Display.

Synteny analysis of ERFVIIs genes was conducted, with the flanking genes of ERFVIIs retrieved from jBrowse through JGI Phytozome. Next, their conserved genes were identified using JGI Phytozome's Protein Homologs analysis tools.

The complete amino acids sequences of ERFVIIs underwent comprehensive motif discovery and searching using the MEME SUITE, a widely acclaimed tool for identifying conserved motifs in biological sequences. During the analysis, we configured the MEME SUITE to search for 20 distinct motifs. The identified candidate motifs were then aligned with several known motifs analyzed by the Tomtom Motif Comparison Tool.

Statistical analyses

IBM SPSS was employed to conduct a one-way ANOVA analysis. If significant differences among groups were detected (F = 0.05), a Duncan and Tukey post hoc test was applied to make specific group mean comparisons.

Accession numbers

Sequence data from this article can be found in the JGI Phytozome under gene identifier: BdERF018 (BdiBd21-3.1G0613800), BdERF108 (BdiBd21-3.3G0791600), BdERF961 (BdiBd21-3.1G0235000), and PGB1 (BdiBd21-3.1G0934500).

Author contributions

P.Y.H. conducted most of the experiments and wrote the manuscript; C.Y.Z. assisted with the OsERFVII reporter assays. M.C.S. conceived and supervised the project, contributed to the writing and was involved in data interpretation.

Supplemental data

The following materials are available in the online version of this article.

Supplemental Figure S1. BdERF961 expression is necessary for the normal development of lateral roots and seeds.

Supplemental Figure S2. Submergence-tolerant ecotypes of B. distachyon, Bd2-3 and Bd1-1, exhibited higher expression of HRGs than Bd21-3.

Supplemental Figure S3. BdERF053 gene expression is regulated by BdERF108, BdERF018, and BdERF961, and BdERF053 also inhibits the expression of PGB1 and BdERF961.

Supplemental Figure S4. BdERF053 plays a negative role under submergence stress.

Supplemental Figure S5. Proposed model for the regulation of BdERFVIIs under submergence.

Supplemental Figure S6. OsERF61 gene expression is regulated by OsERF70, OsERF71, and OsERF67.

Supplemental Table S1. List of RT-qPCR and genotyping primers used in this study.

Supplemental Table S2. List of ChIP-qPCR primers used in this study.

Supplemental Table S3. List of primers used for vector construction.

Supplemental Table S4. Genes used in this study.

Funding

This study was supported by funding from Agricultural Biotechnology Research Center, Academia Sinica, Taiwan.

Data availability

Materials and data are available for scientific research. Biological materials can be requested from the corresponding author. All data and materials used in the analysis in the main text or the supplementary information are available. Gene accession numbers of Brachypodium genes used in this study are listed in Supplemental Table S4. The transcriptomic data were deposited in the BioProject database ID PRJNA942969.

Dive Curated Terms

The following phenotypic, genotypic, and functional terms are of significance to the work described in this paper:

• PRT6 Gramene: At5g02310

• PRT6 Araport: At5g02310

• adh1 Gramene: AT1G77120

• adh1 Araport: AT1G77120

• PDC2 Gramene: AT5G54960

• PDC2 Araport: AT5G54960

• MES CHEBI: CHEBI:39010

• SUB1A Gramene: DQ011598

• SUB1A Araport: DQ011598