Discussion
Transcription factors of the ERF-VII family have been shown to be key regulators in plant responses to low oxygen stress (Gibbs et al. , 2011; Licausi et al. , 2011), and there is evidence that hypoxic niches are central in specific developmental programs (Shukla et al., 2019; Weits et al., 2021).
We identified four genes in the M. truncatula genome that belong to the ERF-VII TF family. Using a genome-wide analysis of the M. truncatula genome, Shu et al. (2016) identified 123 putative AP2/ERF genes, which were designated MtERF1-123 . Unfortunately,MtERF73 (Medtr1g087920) was missing from this analysis, which led us to propose a different nomenclature based on phylogenetic analysis with well-characterized orthologs in Arabidopsis (Table S3). A phylogenetic analysis showed a clear separation between rice ERF-VII involved in submergence tolerance (SUB1 and SK orthologs) and ERF-VII proteins from Glycine max . Moreover, the separation between the different clades preceded the separation between Fabales andBrassicales which would be in favor of clade-specific specialization. Transcriptional analysis revealed that onlyMtERF73 is up-regulated by hypoxia stress, such as HREs ortholog in Arabidopsis (Licausi et al., 2010), and during nodule development.MtERF73 was found to be expressed mainly in the nodule zone III, which is microoxic (Appleby, 1992). The hypoxia-inducible ERF-VIIsHRE1 and HRE2 in Arabidopsis likely maintain the transcription of HRGs, while the ERF-VII RAPs initiate their up-regulation (Bui et al., 2015; Licausi et al., 2010), the presence of MtERF73 in microaerophilic nodule tissues suggests a role in long-term adaptation to microoxic conditions. Future work could clarify the precise role of MtERF73 in metabolic changes and developmental reprogramming under the conditions observed in microoxic nodules.
It should be emphasized that, constitutively expressed MtERF74and MtERF75 , derived from recent duplications like their ortholog in different plant species analysed (van VEEN et al., 2014). Furthermore, their similarity to RAP2.2 and RAP2.12 in Arabidopsis suggests a possible similar function as major activators of hypoxia-responsive genes. Using MtERF74/75 RNAi transgenic roots, we confirmed that TF are involved in expression of several HRGs(Mustroph et al., 2009), namely ADH1 , PDC1 , Pgb1.1 ,AlaAT , and ERF73 (Fig. 3). However, expression ofLDH and Susy genes, belonging to the core anaerobic genes, was not affected in transgenic roots. Among the deregulated HRGgenes, we identified MtPgb1.1 , a gene coding for an enzyme involved in NO-scavenging (Berger et al., 2020b), as one of the genes mostly affected by silencing of MtERF74/75 . Hartman et al. (2019) have recently shown that AtPGB1 from A. thaliana (ortholog ofMtPgb1.1 ) plays a key role in low O2 responses and links ethylene signalling to hypoxia tolerance through its ability to scavenge NO. In M. truncatula , the silencing ofPhytogb1.1 was found to conduct the increase in NO concentration, negatively impact the nodulation and mimics a situation of hypoxia by the overexpression of ADH and PDC (Berger et al., 2020b). Our results suggest that ERF74 and ERF75 control the expression of MtPgb1.1during hypoxic stress in M. truncatula .
Silencing of MtERF74/75 demonstrates the importance of these TFs during the initial phase of nodule development, with a 30% reduction in nodulation efficiency (Fig. 4). In addition, we observed a phenotype, albeit very weak (18% reduction), in the number of lateral roots inMtERF74/75 RNAi (Fig. S4). Previous reports described the presence of an hypoxic niche in LRM during lateral root formation that promote the stability of ERF-VII and thereby regulate the transition from LRP to LR (Shukla et al., 2019). It should be emphasized that LRP and nodule primordia (NP) share some common features. Schiessl et al. (2019) showed that there is overlap in the mechanisms of organogenesis and transcriptional regulation between nodule and lateral root initiation. The same authors showed that auxin and some auxin-responsive regulators are accumulated similarly in LRP and NP. In addition, NO is detected in dividing cortical cells of these two different developmental stages (Correa-Aragunde et al., 2004; del Giudice et al., 2011). Recent work showed that the production of NO, observed in nodule primordia 4 days post-inoculation in the M. truncatula- S. meliloti symbiosis is involved in the onset of nodule organogenesis (Berger et al., 2020b). NO scavenging by pharmacological (treatment with NO scavenger) or genetic approaches (overexpression of Pgb1.1 ) resulted in down-regulation of many genes related to cell division and growth (Berger et al., 2020b; Boscari et al., 2013; del Giudice et al., 2011). Thus, the 30% reduction in nodulation efficiency in MtERF74/75RNAi transgenic roots could be related to deregulation of HRGgenes involved in this chronic hypoxia tissue at the NP stage.
In addition to the phenotypes in the early stages of nodule development, knock-down of MtERF74/75 revealed an important role of these genes in mature nodule functioning with a marked decreased N2 fixation activity (Fig. 4). Interestingly, although nodule development was impaired (Fig. 4c), the analysis of nodule ultrastructure showed no change (data not shown) and the unchanged level of Lb4 expression (gene marker for the N2fixation zone) in the transgenic knock-down MtERF74/75 and control root nodules (Fig. 5) suggests that the reduction in N2 fixation was not related to a failure in the development of the nodule N2 fixation zone. Conclusion also confirmed by the absence of induction of CP6 and VPE(Fig. 5), two markers of nodule senescence (Pierre et al., 2014).
N2 fixation in the nodule is O2sensitive since nitrogenase activity is irreversibly inhibited by O2 (Appleby, 1992). Consequently, legume nodules have evolved mechanisms to reduce the level of O2 by the presence of an O2 diffusion barrier and by expressing leghemoglobin proteins (Appleby, 1992). Legume nodules work efficiently even if internal nodule O2 concentration is between 10 and 50 nM of O2 (Kuzma et al. , 1993), suggesting a peculiar and organ specific adaptation to innate low O2 concentration. In normoxia, M. truncatulanodules exhibit ATP/ADP ratios of 6-8 (Horchani et al. , 2011), demonstrating a high energy state. It is generally assumed that in the O2-limited environment of the nodule, glycolysis is shifted to malic acid synthesis, with further reductive synthesis to fumarate and succinate to feed the bacteroids (Vance and Gantt, 1992). Impairment of N2 fixation activity in the transgenic nodules knockdown of MtERF74/75 could be the consequence of the down-regulation of HRG genes necessary for maintaining optimal function in the microoxia prevailing in nodules. An observed exception, the AlaAT expression was not significantly affected in knockdownMtERF74/75 nodules. As alanine has been involved in the adaptation to microoxia in roots (Ricoult et al., 2005) and nodules (Berger et al., 2020a), this suggests that AlaAT is regulated via an ERF74/75-independent pathway.
Interestingly, NR1 and Pgb1.1 , two important actors of the Pgb-NO respiration (Gupta and Igamberdiev, 2011), were severely impaired in the transgenic nodules (Fig. 8). NR enzymes are described as the major source of NO in nodules, enabling the reduction of NO3- to NO2-, which is subsequently reduced to NO via the mitochondrial electron transfer chain (Berger et al., 2020a; Hichri et al., 2015; Horchani et al., 2011). Recently, Berger et al. (2020a) observed that the decrease in NR activity in nodules either by a double RNAi::NR1-2 or by the use of tungstate (NR inhibitor) was accompanied by a decrease in N2 fixation activity. It was concluded that NRs in cooperation with Phytogb1.1 enables the maintenance of cell energy status in nodules and N2-fixing metabolism through the functioning of the Phytogb-NO respiration. Therefore, it seems that one of the functions of MtERF74 and MtERF75 in long-term adaptation of nodule to microoxia is the induction of MtNR1 and Pgb1.1 expression for efficient Phytogb-NO respiration.
Furthermore, we investigated whether the stability of the ERF75 protein is controlled via the N-degron proteolytic pathway. We observed the highest stability of the chimeric protein after substitution of the terminal Cys by Ala (Fig. 6a) and in the prt6 mutant in Arabidopsis protoplasts (Fig. 6b), indicating that the Cys residue at the N-terminal part of MtERF75 is involved in the stability of the protein. A result that was also observed with the GFP-fused MtERF75 protein (pMA-ERF75::GFP) in M. truncatula protoplasts. These results suggest that the turnover of the different members of the MtERF-VII family in M. truncatula may be regulated by the N-degron proteolysis pathway, as shown for RAP2.12 in Arabidopsis (Gibbs et al., 2011; Licausi et al., 2011). The investigation of the role of N-terminal amino acid residues in the subcellular localization of MtERF75 fused to GFP in M. truncatula protoplasts revealed that under aerobic conditions, the fusion protein may be localized in the nucleus (Licausi et al., 2011).
Both O2 and NO are required to destabilise ERF-VII (Gibbs et al., 2015, 2014b). Regarding the involvement of NO, in animal cells it has been proposed that Cys nitrosylation precedes Cys oxidation (Hu et al., 2005). The high concentration of NO in nodules prompted us to investigate the role of NO in the targeted proteolysis of MtERF75. The internal O2 concentration in nodules (10-50 nM (Kuzma et al., 1993)) should lead to stabilisation of ERF-VII. This implies that the O2-sensitive regulation of the transcription factor MtERF-VII might be different in nodules than in root. An interesting hypothesis to test would be that high content of NO in nodules favour Cys oxidation at a low O2concentration, degrading MtERF-VII under normal conditions. It has recently been demonstrated that enzymes from Plant Cystein Oxidase family (PCOs 1 and 4) are dioxygenases that catalyse the direct incorporation of O2 into RAP2.2/RAP2.12 peptides to form Cys-sulfinic acid (White et al., 2017). However, a role for NO in the formation of a Cys-sulfonic acid product that is also a substrate for ATE1 has not been ruled out (White et al., 2017). In our work (Fig. 9), we show that S-nitrosylation, NO-dependent posttranslational modification (PTM) targeting specific Cys residues can be detected for NO concentrations on the order of a few micromoles (in the range of physiological values) in peptides CR20 lacking the first methionine. Using this original and innovative approach, we demonstrated that S-nitrosylation, the first step in the oxidation of the Cys residue, could eventually occur. Whether this is altering the intervention of PCO1 and PCO2 enzymes, which catalyse the oxidation of the N-terminal Cys of the substrate with molecular O2 as a co-substrate (Weits et al., 2014) is still to be defined.
In conclusion, MtERF74 and MtERF75 plays a key role in activating the anaerobic response in M. truncatula . Their impairment conduct to a significant reduction in nodulation capacity and nitrogen fixation activity in mature nodules that could be explain by a reduce efficiency of Pgb-NO respiration in nodules knock-down for both TF. Our results suggest that MtERF75 is target of the N-degron pathway and stable forms of the protein are localised in the nucleus. Moreover, S-nitrosylation of the Cys2 residue is promoted when the first Met is removed can occur. This evidence opens new questions regarding the existence of a specific function of NO in nodules related to the presence of an hypoxic niche. The NO regulation of MtERF-VII, and consequently of yet unknown target genes, for efficient N2 fixation and nodule organogenesis need to be explored in more depth.