The submergence-induced hypoxic condition negatively affects the plant growth and development, and causes early onset of senescence. Hypoxia alters the expression of a number of microRNAs (miRNAs). However, the molecular function of submergence stress-induced miRNAs in physiological or developmental changes and recovery remains poorly understood. Here, we show that miR775 is an Arabidopsis thaliana-specific young and unique miRNA that possibly evolved non-canonically. miR775 post-transcriptionally regulates GALACTOSYLTRANSFERASE 9 (GALT9) and their expression is inversely affected at 24 h of complete submergence stress. The overexpression of miR775 (miR775-Oe) confers enhanced recovery from submergence stress and reduced accumulation of RBOHD and ROS, in contrast to wild-type and MIM775 Arabidopsis shoot. A similar recovery phenotype in the galt9 mutant indicates the role of the miR775-GALT9 module in post-submergence recovery. We predicted that Golgi-localized GALT9 is potentially involved in protein glycosylation. The altered expression of senescence-associated genes (SAG12, SAG29 and ORE1), ethylene signalling (EIN2 and EIN3) and abscisic acid (ABA) biosynthesis (NCED3) pathway genes occurs in miR775-Oe, galt9 and MIM775 plants. Thus, our results indicate the role for the miR775-GALT9 module in post-submergence recovery through a crosstalk between the ethylene signalling and ABA biosynthesis pathways.
Abiotic factors, such as light, temperature, water, oxygen, nutrition, etc., act as external cues to modulate plant growth and development (Kraehmer, 2016). Globally, flood is one of the major stresses that lead to severe loss of crop yield and productivity (Ismail, 2018; Kumar and Dash, 2019; Bui et al., 2020). Flooding stress is divided into two categories depending upon the water exposure: complete submergence and partial submergence (or waterlogging). Under complete submergence, the whole plant is fully immersed in water; under partial submergence, the shoot terminal is maintained above the water surface (Fukao et al., 2019). Premature senescence, necrosis, chlorosis and cessation of growth are the major consequences of submergence stress (Zhang et al., 2000; Visser et al., 2003). Submergence initiates various molecular cascades that adversely affect plant growth and development. Submergence also leads to excessive reactive oxygen species (ROS) production and cell death, due to the reduced availability of oxygen (hypoxic conditions). Prolonged submergence gradually affects gaseous exchange, which leads to cell damage and chlorophyll breakdown causing early onset of senescence (Fukao et al., 2019). RESPIRATORY BURST OXIDASE HOMOLOGUE D (RBOHD) is one of the NAPDH oxidases involved in the production of ROS and is widely used as a marker gene to assess ROS accumulation during submergence stress. RBOHD promotes the expression of alcohol dehydrogenase (ADH), pyruvate decarboxylase 1 (PDC), lactate dehydrogenase (LDH), calcium (Ca2+) levels and various hypoxia-responsive genes, induced during hypoxia stress. An abundance of RBOHD also changes under submergence due to oxidative stress (Yamauchi et al., 2013; Yeung et al., 2018). Senescence-associated genes (SAGs), such as SAG12, SAG29 and ORESARA1 (ORE1/NAC6) are mainly expressed in senescent tissue, which ultimately leads to chlorophyll breakdown and leaf senescence. SAG12 encodes a cysteine protease (vacuolar protein) in Arabidopsis thaliana (Arabidopsis) and is associated with oxidative stress (H2O2) and senescence (Weaver et al., 1998). Increased SAG12 transcript levels lead to early leaf senescence in Arabidopsis (Ding et al., 2016; Ueda et al., 2020). Similarly, SAG29 is also prominently expressed in senescing plant tissue and its expression is increased by osmotic stresses through an abscisic acid (ABA)-dependent pathway (Seo et al., 2011). ORE1 has been identified as an accession-specific regulatory gene that is expressed at high levels in the Bay-0 ecotype (an Arabidopsis accession) and has a predominant role in chlorophyll breakdown. The ore1 mutant showed intermediary submergence tolerance (Yeung et al., 2018). Both ORE1 and SAG29 accelerate leaf senescence in Arabidopsis (Kim et al., 2009; Qiu et al., 2015; Zhang et al., 2018).
Besides genes and transcription factors (TFs), a large number of miRNAs have been also reported to be dynamically regulated under submergence stress (Zhang et al., 2008; Moldovan et al., 2010; Licausi et al., 2011; Liu et al., 2012; Jeong et al., 2013; Zhai et al., 2013; Jin et al., 2017; Li et al., 2017; Franke et al., 2018; Fukao et al., 2019). MicroRNAs (miRNAs) belong to a class of endogenous small non-coding RNAs that negatively regulate their target gene expression at the post-transcriptional level through complementary pairing with their specific target mRNAs in most of the eukaryotes. Several miRNAs have recently been implicated in various developmental processes, including shoot and root development. Many miRNAs have been shown to be differentially expressed under various abiotic stress conditions (Singh et al., 2020a; Wang et al., 2020).
Previous reports showed the involvement of some miRNAs in the regulation of SAG genes, e.g. miR164-mediated regulation of ORE1 triggers early senescence by regulating various SAG genes (Balazadeh et al., 2010; Glazińska et al., 2014; Yeung et al., 2018). miR159 was found to be upregulated in maize root during waterlogging or flood conditions and its targets, gibberellin-mediated expression of myeloblastosis genes (GAMYBs; e.g. MYB33 and MYB101) were downregulated (Liu et al., 2012). miR166 is shown to have an important role in response to flood stress through regulating calcium spikes and accumulation of ROS during root growth and development (Fukao et al., 2019). miR167 was found to regulate short-term waterlogging or submergence in maize root by targeting auxin response factors (ARFs) (Zhang et al., 2008; Liu et al., 2012). Recently, a report showed that miR167 was differentially upregulated in Alternanthera philoxeroides and Populus tomentosa plants during flood response (Li et al., 2017). Moreover, miR156 have been indicated for its potential role in submergence and hypoxia by targeting squamosa promoter binding protein-likes genes (SPLs) in lotus and Arabidopsis (Moldovan et al., 2010; Jin et al., 2017; Franke et al., 2018). Long-term waterlogging downregulated the expression of miR172, leading to accumulation of target AP2/ERF mRNAs and thereby promoting crown root development in maize (Zhai et al., 2013). The expression of miR775, which targets GALT9, was reported to be induced by hypoxic conditions caused by flood or high altitude in Arabidopsis (Moldovan et al., 2010; Liu et al., 2012; Jin et al., 2017; Tripathi et al., 2019). However, the functional role of miR775 in hypoxia remains to be understood. In the present study, we have addressed the potential role of the miR775-GALT9 module in submergence stress-induced hypoxia and post-submergence recovery in Arabidopsis. Our study shows that miR775-GALT9 module plays an important role in post-submergence recovery and senescence in Arabidopsis, by modulating the expression of SAG genes, RBOHD, ethylene and ABA pathway genes.
Gene structure of MIR775A and its possible origin through promoter acquisition
MIR775A (AT1G78206) and its mature miRNA miR775, are uniquely present in Arabidopsis, at chromosome 1 (Araport11; with gene coordinate Chr1:29422452-29422574). Interestingly, we did not find any promoter elements in the upstream [transcription starts site (TSS) and TATA-box] of MIR775A up to the next gene (snoRNA). Hence, we searched for the existence of a canonical promoter element upstream of the MIR775A and snoRNA genes. We retrieved an intergenic sequence upstream of MIR775A to identify or predict the promoter through software ‘TSSPlant’ (http://www.softberry.com/cgi-bin/programs/promoter/tssplant.pl). We found an adjacent gene, snoRNA (AT1G09787.1), situated very closely (92 bp apart) in the upstream of MIR775A, having no predicted TSS or TATA-box (Fig. 1A). Therefore, we extended our search for the intergenic sequence between the next upstream gene, AT1G78200 (upstream of MIR775A). We retrieved an intergenic sequence of 587 bp, including the 3′ UTR of the AT1G78200 gene, which also overlapped with the snoRNA sequence, and then predicted the TSS and TATA-box elements. (Fig. 1A). TSSPlant predicted two TSS: one without a TATA box and another with a TATA box (Fig. S1A). The predicted TSS with a higher score (1.9844), which also possessed a TATA box, was considered to be the most likely candidate (Fig. S1A). This result suggests that MIR775A putatively acquires the promoter elements to become functional, which correlates with earlier findings that promoter acquisition is also one of the modes of evolution of miRNAs (Lu, 2019). However, we also checked other canonical modes of miRNA evolution to understand the origin of MIR775A in Arabidopsis. To address this, we first investigated whether MIR775A originated canonically through inverted duplication of its target (Allen et al., 2004). We observed that MIR775A and its validated target GALT9 genes are not situated in tandem, as they are 9.55 Mbp apart from each other in chromosome 1 of the Arabidopsis genome (Fig. 1B), and possess no duplicated fragment, which rules out their possible origin through an inverted duplication mode, unlike common miRNAs. Furthermore, we also checked the sequence variability at the miR775-binding site in GALT9 of Arabidopsis against GALT9 orthologs (Fig. S1B). We found a highly conserved miR775-binding site in GALT9 orthologs (Fig. S1B), which excludes the possibility that MIR775A originated from GALT9. Therefore, our analysis indicates that the MIR775A possibly originated specifically in the genome of Arabidopsis non-canonically through acquisition of a TSS and TATA box in its upstream region (Fig. 1A), which is a unique feature of miR775.
Non-conserved miR775 targets conserved GALT9
During our ongoing work, other studies recently validated GALT9 as a target of ath-miR775 in Arabidopsis (Fahlgren et al., 2007; Mishra et al., 2021; Zhang et al., 2021). Some more targets, DCL1 (AT1G01040), WRKY19 (AT4G12020) and KFB (KELCH DOMAIN-CONTAINING F-BOX PROTEIN, AT1G23390) were also predicted but not validated (Zhang et al., 2011). GALT9 achieved the status of a bona fide target of miR775; surprisingly, however, studies differed in terms of its cleavage site(s) (Fig. 1C). Our degradome analysis of the most recent TAIR database (AraPort11) using the CleaveLand tool confirmed GALT9 as the strongest target of miR775 in Arabidopsis. Variable cumulative scores for a perfect cleavage site (10th) were obtained during degradome analysis in 11-day-old seedling and leaf (Fig. 1C, Tables S1 and S2). In seedling tissue, we observed the perfect cleavage site of miR775 in GALT9 mRNA (position 10th), where all reads fell to CleaveLand analysis category (CAC) zero (0), the best score. However, in leaf, the miR775-mediated cleavage of GALT9 appeared in CAC (2), suggesting moderate deviation from the perfect cleavage site in some aligned reads (Fig. 1C). Our results indicate tissue-specific variation of miR775-mediated cleavage of GALT9 (Fig. 1C, Tables S1 and S2). Thus non-conserved miR775 is uniquely present specifically in Arabidopsis; however, its target GALT9 and GALT9 orthologues are widespread and conserved across the plant species (Fig. S1B).
The expression pattern in the co-infiltration experiment and in mutants or transgenics revealed negative regulation of GALT9 by miR775
To further validate the transcriptional cleavage of GALT9 by miR775, 4-week-old Nicotiana benthamiana (tobacco) leaves were co-infiltrated with the construct 35S:GALT9-GFP (control), with empty vector (EV)+35S:GALT9-GFP, and with 35S:MIR775A+35S:GALT9-GFP (Fig. S2A). If cleaving GALT9, co-infiltrated miR775 should cleave the GALT9 transcripts and reduce the expression of GFP. In parallel, we used another miRNA, miR839, as an additional control to validate that the GALT9 is specifically targeted by miR775 and co-infiltrated the tobacco leaves with 35S:MIR839A+35S:GALT9-GFP. Furthermore, we quantified the level of GFP by quantitative RT-PCR (qRT-PCR) (Fig. S2B) and observed significant reduction in co-infiltrated (35S:MIR775A and 35S:GALT9:GFP) tobacco leaves in comparison with tobacco leaf co-infiltrated with 35S:GALT9-GFP, EV+35S:GALT9-GFP and 35S:MIR839A+35S:GALT9-GFP. Reduced expression of GFP confirmed the post-transcriptional cleavage of GALT9 transcripts by miR775 (Fig. S2B). To further confirm the miR775-mediated downregulation of GALT9 in Arabidopsis, we analysed the expression of GALT9 in two independent miR775-overexpression lines (miR775-Oe1 and miR775-Oe2) (Fig. 2A) and observed its reduced expression (by 48% and 63%, respectively) (Fig. 2B). Furthermore, we observed significant upregulation of GALT9 transcript level in two independent lines with a target mimic of miR775 (MIM775-1 and MIM775-2) by 3.98- and 2.55-fold, respectively (Fig. 2C). These results confirmed the post-transcriptional regulation of GALT9 by miR775, which is consistent with a recent study (Zhang et al., 2021) published during the processing of this article.
Functional annotation by GALT9 protein structure analysis
The 3D structure of GALT9 was predicted using I-TASSER, which further deduces protein functions and biological annotations. We analysed the enzyme commission (EC) numbers and gene ontology (GO) terms, and annotated their putative molecular function based on modelled 3D structure of GALT9. The best consensus putative function of GALT9 is that of an enzyme, β-1,3-galactosyl-O-glycosyl-glycoprotein β-1,6-N-acetylglucosaminyltransferase (EC number: 184.108.40.206). GALT9 has a role in protein glycosylation (GO:0006486), where the ligand is uridine-diphosphate (UDP), which is involved in UDP-glycosyltransferase activity (GO:0008194), and is localized in Golgi membrane (GO:0000139) (Figs S4 and S5). These results indicate that GALT9 is involved in the transfer of glycosyl groups (GO:0016757) to protein moieties (e.g. the addition of glycan chains to proteins) through peptidyl-asparagine modification (GO:0018196). The peptidyl-asparagine modification helps in N-glycosylation, where glycans are attached to the side-chain nitrogen atoms of asparagine (Asn) residues in a conserved consensus sequence asparagine-Xaa-serine/threonine (Asn-Xaa-Ser/Thr), where Xaa could be any amino acid (Schwarz and Aebi, 2011).
Arabidopsis GALT9 is a member of the Carbohydrate Active Enzyme (CAZy) family and the protein localizes in Golgi apparatus
The coding sequence of GALT9 consists of 1038 bp, including seven exons that are interrupted by six introns while GALT9 proteins consist of 345 amino acid residues (https://www.arabidopsis.org/servlets/TairObject?name=AT1G53290&type=locus). GALT9 protein catalyzes the transfer of a glycosyl group from a UDP-glycosyl donor molecule to protein moieties (Qu et al., 2008; Gille et al., 2013; Qin et al., 2013). Arabidopsis GALT9 shows strong homology with Gossypium hirsutum (cotton) GhGALT1 and belongs to the Carbohydrate Active Enzyme glycosyltransferases (CAZy GT) family (Qin et al., 2013). A total 31 members of the CAZy GT-family are present in Arabidopsis and, out of 31, 20 members have a β-(1-3)-GT motif, similar to mammalian systems (Qu et al., 2008). As GALT9 is potentially involved in post-translational modification of proteins, we investigated its subcellular localization. Using the Arabidopsis Subcellular Localization Prediction Server (AtSubP, http://bioinfo3.noble.org/AtSubP/index.php), we predicted GALT9 protein was localized in the Golgi apparatus, at a subcellular level. To validate this, we generated a translation fusion reporter (35S:GALT9-GFP) construct and co-infiltrated it with a GmManI-pBIN2 Golgi mCherry marker for transient subcellular localization in Nicotiana benthamiana leaves. Using differential interference contrast (DIC) microscopy analysis, we found fluorescence from 35S:GALT9-GFP to be primarily localized in the Golgi apparatus which is consistent with a recently published study (Zhang et al., 2021), further validating the in silico prediction (Fig. S5).
MIR775A and target GALT9 showed complementary and dynamic expression pattern predominantly in shoot and root tissues
To investigate the spatiotemporal expression pattern of MIR775A and its target GALT9, we performed the histochemical GUS assay in 5 days after germination (dag) seedlings and in 35-day-old pMIR775A:GUS and pGALT9:GUS plants. In 5 dag seedlings, MIR775A showed expression in the shoot, hypocotyl, cotyledons, leaves, root-shoot junction (RSJ), stomata, trichomes and rosette leaves of 35-day-old plants (Fig. 3A-E). GALT9 was also expressed in shoot, root-shoot junction, hypocotyl, leaves, stomata and trichomes in 5 dag seedlings and in rosette leaves of 35-day-old plants (Fig. 3F-J). The GUS expression of pMIR775A:GUS and pGALT9:GUS was moderately complementary in shoot, shoot apex, root-shoot junction, hypocotyl, stomata, trichome and rosette leaves.
Furthermore, we quantified and compared the expression level among 5 dag whole seedlings, shoot and roots (Fig. 4A), which showed higher expression of MIR775A in shoot when compared with root (Fig. 4A). MIR775A and GALT9 showed complementary expression in young rosette, cauline leaves and flower (Fig. 4B). The availability of mature miRNA depends on many factors, such as the activity of miRNA biosynthesis pathway genes, miRNA processing machinery, etc. (Chen, 2009). Therefore, we also validated the accumulation of functional mature miRNA in different tissues of 35-day-old Col-0 plants by stem-loop qRT-PCR. The mature miR775 was expressed abundantly in a young rosette, young cauline, floral bud and flowers (Fig. 4B). These results suggest a dynamic spatiotemporal expression pattern of miR775 and its target GALT9 in Arabidopsis, which is likely to have an impact on its physiological or developmental role.
Submergence stress affects the expression of miR775 and GALT9 in Arabidopsis seedlings
Hypoxia stress conditions, caused by flooding or high altitude, are known to induce the expression of miR775 in Arabidopsis (Moldovan et al., 2010; Liu et al., 2012; Jin et al., 2017; Tripathi et al., 2019). To understand the molecular basis of this, we first studied the effect of submergence-induced hypoxia stress on the expression of miR775 and its target gene GALT9 in Arabidopsis shoot. To analyse the submergence-induced expression of MIR775A, we treated 7 dag seedlings with complete submergence and analysed the expression of miR775 by stem-loop qRT-PCR and of GALT9 transcripts by qRT-PCR at four different time points (4 h, 8 h, 12 h and 24 h) (Fig. 5A). We found that the expression of miR775 was upregulated by fivefold; however, the expression of its target GALT9 was downregulated by 62% after 24 h of submergence. However, the expression of miR775 was significantly upregulated at 4 h but the expression of its target GALT9 was not significantly changed. Furthermore, to analyse the tissue-specific expression, 7 dag shoot seedlings of pMIR775A:GUS and pGALT9:GUS transgenic plants were analysed at two time points, 4 h and 24 h, during submergence stress (Fig. 5B-G). Interestingly, we found similar results to those obtained using stem-loop qRT-PCR: under normoxic (normal level of oxygen) conditions, pMIR775A:GUS and pGALT9:GUS seedlings showed a basal level of expression in untreated seedlings (0 h). At 4 h after submergence, Arabidopsis seedlings had a mild elevation of pMIR775A:GUS expression; however, 24 h after submergence, pMIR775A:GUS expression was upregulated (Fig. 5B-D), whereas expression of the pGALT9:GUS expression was significantly downregulated in shoot tissue (Fig. 5G). This result highlights the potential role of miR775 and its target GALT9 during submergence stress in Arabidopsis.
Submergence recovery in miR775-Oe, MIM775 and galt9 plants
Complete submergence leads to the overproduction of different reactive oxygen species (ROS) due to the oxygen-deficient conditions (Bailey-Serres and Voesenek, 2008). Previously, we have shown that the expression of miR775 and GALT9 was dynamically affected in a complementary way during submergence stress (Fig. 5). Therefore, we were interested to estimate the survival ability of different miR775 transgenic lines and of galt9. To estimate the survival rate, we treated different transgenic lines at the same stage (3 weeks old) (miR775-Oe1, miR775-Oe2, galt9 and MIM775-1, MIM775-2) in complete submergence for 5 days, and estimated their survival rate after 5 days of desubmergence (5 DADS). We observed that nearly ∼75% of the wild-type (Col-0) plants survived during submergence recovery; however, the survivability rate was higher in miR775-Oe lines (∼85-90%) and galt9 (∼90%). The survival rate of MIM775 was reduced to ∼50%. (Fig. 6M). These results indicate that miR775-mediated regulation of its target gene GALT9 contributes to the plant survival during post-submergence recovery.
Submergence stress altered the ROS levels in miR775-Oe, MIM775 and galt9 plants
To further understand the molecular function of the miR775-GALT9 module during submergence stress in Arabidopsis, we analysed reactive oxygen species (ROS) accumulation. Hydrogen peroxide (H2O2) is one of the major ROS that accumulates during various abiotic and biotic stresses (Liu et al., 2010). We have estimated the ROS (here H2O2) level in different miR775 transgenic lines (miR775-Oe and MIM775) and in the galt9 mutant in 12 dag seedlings after 5 days of submergence through DAB staining. The accumulation of H2O2 was highest in MIM775-1 and MIM775-2 (Fig. 7A-F). The level of H2O2 reduced in miR775-Oe1, miR775-Oe2 and galt9, when compared with Col-0 control (Fig. 7A-F). Furthermore, we quantified the ROS intensity by using ImageJ (https://imagej.nih.gov/ij/download.html) (Fig. 7G). Next, we quantified the accumulation of RBOHD, a core hypoxia marker gene, after the 5 days of desubmergence and we found increased expression of RBOHD in MIM775-1 (2.19) but decreased expression of RBOHD in miR775-Oe1 (by 99%) and galt9 (by 97%) (Fig. 7H). Increased expression of RBOHD in MIM775 was persistent even after 5 DADS, suggesting ROS accumulation. This might have led to impaired growth during submergence due to the higher cell death. These higher levels of functional miR775, in miR775-Oe lines, promoted the plant survival ability during recovery from submergence stress, which might have reduced accumulation of ROS. Altered ROS levels in different miR775 transgenic and galt9 lines suggest the differential ability of cells to survive.
Expression of senescence-associated genes (SAGs) and chlorophyll content were altered by post-submergence recovery
Our results showed the poor recovery in MIM775 lines after submergence, in contrast to miR775-Oe and galt9 lines (Fig. 6M). The MIM775 transgenic lines have small rosette leaves; however, the miR775-Oe and galt9 lines were phenotypically similar to the Col-0 control under normal growth and developmental conditions. MIM775 transgenic lines exhibited enhanced senescence and a high degree of chlorosis after 5 DADS (Fig. 6). We observed a reduction in senescence and chlorosis in miR775-Oe and galt9 transgenic lines, when compared with MIM775. To further quantify chlorophyll breakdown, we estimated the level of chlorophyll A, chlorophyll B, total chlorophyll, carotenoids and xanthophylls at 5 DADS (Fig. 8A-D). We observed a reduction in total chlorophyll levels in the MIM775 transgenic lines, in contrast to miR775-Oe and galt9 transgenic lines. The level of chlorophyll A, chlorophyll B and xanthophyll was also reduced in MIM775 lines (Fig. 8A-D).
An increase in ROS level and reduction in chlorophyll content after submergence stress is related to precocious senescence. Our results indicate that miR775-Oe and galt9 plants survived through the submergence stress by reducing ROS levels and chlorosis, and, therefore, by delaying the senescence. On the other hand, the early senescence observed in MIM775 plants (post-submergence) could be due to excessive ROS levels and reduced chlorophyll content (Fig. 7A-F,G; Fig. 8A-D).
Different abiotic and biotic stresses affect the expression of SAG genes such as SAG12, SAG29 and ORE1, which further govern leaf senescence (Seo et al., 2011; Yeung et al., 2018; Bengoa Luoni et al., 2019; Ueda et al., 2020). We sought to determine whether miR775-GALT9-mediated post-submergence recovery and senescence, as evidenced by the phenotype in miR775 transgenic and galt9 lines, involved SAGs. To address this, we compared the expression levels of SAG12, SAG29 and ORE1 during post-submergence recovery. The expression levels of SAG12, SAG29 and ORE1 was downregulated in miR775-Oe1 and galt9 transgenic lines, whereas expression was upregulated in MIM775-1 (Fig. 8E). These data suggest that the miR775-GALT9 module regulates submergence tolerance and post-submergence recovery by modulating the expression of SAGs (SAG12, SAG29 and ORE1) in Arabidopsis.
Putative molecular mechanisms and hormonal crosstalk during post-submergence recovery
Phytohormones, such as ethylene and ABA, promote leaf senescence (Wang et al., 2021). In Arabidopsis, ethylene promotes leaf senescence through a signalling cascade mediated by ETHYLENE INSENSITIVE 2 (EIN2), EIN3 and ORE1 (Kim et al., 2009, 2014). EIN3, a pivotal TF of ethylene signalling directly activates the expression of ORE1 and SAG29 to promote leaf senescence (Kim et al., 2014; Qiu et al., 2015; Zhang et al., 2018). To confirm this further, we quantified the expression of EIN2 and EIN3, which act upstream of ORE1 and SAG29, and we found EIN2 and EIN3 upregulated by 2.65- and 22.39-fold, respectively, in MIM775 (Fig. 8F). However, EIN2 and EIN3 were downregulated in miR775-Oe1 (by 92% and 79%) and in galt9 (by 55% and 66%). It has been reported that senescence-associated NAM/ATAF1,2/CUC2 (NAC) TFs, such as ANAC019, ANAC047, ANAC055, ORE1 SISTER1 (ORS1) and ANAC102, and ANAC087, act downstream of EIN2 and EIN3, respectively, to accelerate leaf senescence, chlorophyll breakdown and cell death, which ultimately inhibit post submergence recovery (Kim et al., 2014; Yeung et al., 2018).
It has been shown previously in the current study that the expression of RBOHD, a core hypoxia marker gene was increased in MIM775-1 (2.19-fold) when compared with miR775-Oe1 and galt9 lines, after 5 DADS (Fig. 7H). Increased expression of RBOHD in MIM775 was persistent even after 5 DADS, suggesting ROS accumulation. It has been shown that osmotic stress causes oxidative stress, leading to oxidative cell damage, and expression of the ABA biosynthesis gene NINE-CIS-EPOXYCAROTENOID DIOXYGENASE 3 (NCED3) is drastically upregulated by osmotic stress (Xiong et al., 2002; Tamang and Fukao, 2015). Furthermore, we quantified NCED3 which has been shown to be upregulated by osmotic/oxidative stress and we found the expression of NCED3 was upregulated by 5.84-fold in MIM775-1 (Fig. 8F) compared with miR775-Oe1 and galt9 (reduced by 89% and 90%, respectively). This increased NCED3 leads to accumulation of ABA and thus promotes stomatal closure, which inhibits dehydration and ultimately affects post-submergence recovery. These data suggest that the miR775-GALT9-ET-ABA module regulates post-submergence recovery by modulating ethylene signalling and ABA biosynthesis in Arabidopsis.
The miRNA-mediated gene regulation is crucial for the plant growth, development and physiological responses, which facilitated the origin and evolution of miRNA through positive selection. Many of the miRNAs are evolutionarily categorized as ancient or young, based on their origin in plant species (Zhang et al., 2006; Voinnet, 2009). Similarly, miR775 appears to be a so-called young miRNA, which is due to its species-specific presence in Arabidopsis (Lu et al., 2006; Rajagopalan et al., 2006). Earlier studies reported that plant species from Poaceae and Brassicaceae possessed many species-specific non-conserved miRNAs that originated through rapid spontaneous evolution (Zhang et al., 2006; Cui et al., 2017). Our study shows the de novo origin of MIR775A in Arabidopsis. The genomic location of MIR775A and its validated target GALT9, and higher conservation of miR775-binding site at GALT9 does not support the canonical modes of origin and evolution of miR775 (Fig. 1). Furthermore, we found its gene structure acquired through the acquisition of TSS and TATA-box upstream of two consecutive genes, snoRNA and MIR775A dicistronic in nature (Qu et al., 2015) (Fig. 1A, Fig. S1A). The continuous preferential selection pressure might be acted upon by the Arabidopsis genome for the acquisition of promoter elements to regulate target gene GALT9. However, genetic drift, due to lack of selection, might have resulted in the loss of these genes in subsequent species divergence (Voinnet, 2009).
Identification of miRNA, target genes and their conservation across the species reveals the role of miRNA-target gene modulation in various aspects of plant growth and their physiology. Techniques such as degradome PARE analysis often enables the identification of conserved and novel targets. Our degradome analysis suggested GALT9 was a strong target of miR775 in Arabidopsis (Fig. 1C). Consistent with other observations, GALT9 is validated and established as a bona fide target of miR775 (Fig. 1C) (Fahlgren et al., 2007; Mishra et al., 2021; Zhang et al., 2021). Interestingly, however, variations in the cleavage site of miR775 were obvious (Fig. 1C), which might depend upon the stage of growth and growth condition, and be due to the biogenesis of miR775 from the 3′ end of its stem-loop precursor (Fig. 1B). Additionally, in the present study, we validated our in silico-based prediction using I-TASSER and AtSubP to show that GALT9 protein localizes in the Golgi body (Figs S4-S5). The role of GALT9 was deduced from the predicted protein 3D structure through I-TASSER, which shows its protein glycosylation activity, as a part of post-translational modifications (PTM) required for folding and stabilizing the translated protein structures.
Environmental factors or abiotic stresses are known to influence the expression of many miRNAs that are involved during the regulation of physiology, as well as during growth and development of plants. It has been reported previously that the hypoxic condition, caused by flood or high altitude, leads to the induction of miR775, indicating its potential role in hypoxia-inducing stresses (Moldovan et al., 2010; Liu et al., 2012; Tripathi et al., 2019).
Submergence severely affects plant growth, development, survivability and yield by reducing light availability and stomatal opening, and therefore gaseous exchanges, which ultimately affects photosynthesis and respiration. So efficient recovery from the submergence stress is vital for plant growth and survival. In the current study, we show the complementary regulation of the miR775-GALT9 module during the submergence stress-induced senescence and post-submergence recovery.
Our results indicate the dynamic spatio-temporal expression pattern of miR775 and its target GALT9 through reporter lines of pMIR775A:GUS and pGALT9:GUS under normal and submergence stress (Figs 3 and 5). Under normal conditions (without submergence), all the miR775 transgenic lines and galt9 mutants were green and healthy; however, leaves of MIM775 lines were small in size (Fig. 6A-F). Mutation in the light-response/signalling gene HY5 is known to affect plant organ or leaf size (Zhang et al., 2021). We showed that the expression of miR775 was increased in the shoot of the hy5-1 mutant (Fig. S6B). Furthermore, HY5 protein is directly bound to the promoter of MIR775A, as shown by the yeast-one-hybrid experiment (Fig. S6A). Therefore, HY5-mediated regulation of miR775 contributes to the altered leaf shape (organ size) (Fig. S7). These results implicate HY5 in miR775-GALT9-mediated organ size regulation (Zhang et al., 2021). Many GALT9 colocalized genes that encode Golgi body localized proteins (such as ACLA1, CGR2 and CGR3; Table S4) have previously been shown to play a key role in plant development, such as ATP-CITRATE LYASE A-1 (acla1) and COTTON GOLGI-RELATED 2 (cgr2-1 and cgr3-1), the mutants of which have relatively smaller organ size (Fatland et al., 2005; Kim et al., 2015b).
We have characterized the miR775-GALT9 module for its role in submergence stress recovery. Submergence stress causes reduction in oxygen level (hypoxia) (Nishiuchi et al., 2012; Ahmed et al., 2013; Chen et al., 2015; Yeung et al., 2018; Loreti and Striker, 2020; Nakamura and Noguchi, 2020). In our results, transgenics [miR775 overexpression (miR775-Oe1) and target mimic of miR775 (MIM775)] and obtained galt9 mutant lines revealed significant differences in their submergence stress tolerance and post-submergence recovery. Among these miR775 transgenic and mutant lines, miR775-Oe1 and galt9 exhibited the reduced accumulation of ROS (Fig. 7A-F,G). However, the MIM775-1 line displayed an increased accumulation of ROS. It has been previously reported that after desubmergence, the reillumination conditions lead to the production of ROS in recovering tissues (Elstner and Osswald, 1994; Smirnoff, 1995). ROS production differed between the miR775-Oe1, galt9 and MIM775 lines, which corresponded to higher RBOHD accumulation during recovery in MIM775. RBOHD, a key hypoxia gene, and the RBOHD-mediated ROS burst are crucial for submergence tolerance and recovery (Yeung et al., 2018). Balanced ROS production is crucial and needs to be countered by an effective antioxidant mechanism that can control excessive ROS production and associated damage in Arabidopsis. However, the recovery signals regulating RBOHD remain to be determined. A recent report shows that ABA and ethylene responses in Arabidopsis are crucial for submergence tolerance and recovery (Yeung et al., 2018). Increased cell death in MIM775-1 during submergence stress is also evident in the shoot after desubmergence (Fig. 7A-F,G). Upregulation of SAG genes, including ORE1, SAG12 and SAG29, is marked during senescence (Weaver et al., 1998; Seo et al., 2011; Kim et al., 2014; Qiu et al., 2015; Ding et al., 2016; Yeung et al., 2018; Ueda et al., 2020). Our results show higher upregulation of EIN2, EIN3, ORE1, SAG12 and SAG29 in MIM775 lines, in comparison with miR775-Oe1, galt9 and Col-0 (Fig. 8E).
The miR775-Oe1 and galt9 transgenics lines were healthy in comparison with Col-0 and MIM775-1 at 5 DADS (Fig. 6A-L). MIM775 lines were poorly affected by the submergence-induced stress and showed early senescence during the submergence (Fig. 6M) (Qiu et al., 2015; Yeung et al., 2018). Submergence stress led to the yellowing of leaves in MIM775-1 lines where chlorophyll content was lowest, in comparison with miR775-Oe1 and galt9 (Fig. 8A-D), suggesting that the miR775- GALT9 module regulates senescence and post-submergence recovery (Fig. 8A-D).
Ethylene is known to be a senescence-accelerating hormone and its signalling involves the ETHYLENE–INSENSITIVE 2 (EIN2), EIN3 and NAC TFs to regulate leaf senescence and chlorophyll degradation in Arabidopsis (Neljubow, 1901; Crocker, 1932; Kim et al., 2015a; Iqbal et al., 2017). EIN3 directly activates the master regulators of SAG genes, ORE1 and SAG29, to promote leaf senescence (Wang et al., 2021).
EIN2 mediates leaf senescence by transducing ANAC019, ANAC047, ANAC055 and ORS1 TFs via an EIN3-independent pathway. However, ANAC087 and ANAC102 TFs were preferentially activated by ORE1 to promote leaf senescence, chlorophyll degradation and cell death (Kim et al., 2014). Phytohormone ABA and ethylene regulate dehydration and senescence during submergence recovery by modulating SAG genes and ORE1 (Yeung et al., 2018). Enhanced expression of RBOHD promotes ROS accumulation, which ultimately accelerates the expression of NCED3 by oxidative stress (Xiong et al., 2002). The ABA accumulation inhibits dehydration by stomatal closure and hampers post-submergence recovery (Xiong et al., 2002; Yeung et al., 2018).
The role of GALT9 is deduced from the modelled protein structure using I-TASSER, which reveals its protein glycosylation activity, which is a part of the post-translational modifications (PTM) required for folding and stabilizing the translated proteins [e.g. many membrane proteins, secreted proteins and vacuolar proteins (Schwarz and Aebi, 2011; von Schaewen et al., 2008)]. Earlier studies have shown that the hampered maturation of N-glycan in Golgi complexes reduces the stress tolerance ability (Kang et al., 2008; Nagashima et al., 2018). Some of the proteins (EIN2 and RBOHD, etc.) involved in the post-submergence-directed senescence network, were membrane bounded and acted as signal receptors. However, any perturbation in N-glycosylation disrupts the activity of membrane-bounded signal receptors (Nagashima et al., 2018). Furthermore, we observed that SAG12, a senescence-associated vacuolar protein, undergoes N-glycosylation (https://www.uniprot.org/uniprot/Q9FJ47#ptm_processing) and might be processed by GALT9 (Figs S4 and S5).The connection between the miR775-GALT9 module and ethylene and ABA pathways is likely to be indirect. Given that GALT9 is not a transcription factor, it is likely to regulate the downstream genes through post-translational modification. N-linked glycosylation of phytohormonal pathway components (proteins) is crucial for hormonal homeostasis, and is required for physiological and developmental responses (Jiao et al., 2020; Li and Lan, 2017; Ostrowski and Jakubowska, 2014). Recently, it was predicted that ethylene biosynthetic pathway genes, such as S-adenosyl-l-methionine (SAM), aminocyclopropane-1-carboxylic acid synthase (ACS) and ACC oxidase (ACO) proteins have N-linked glycosylation sites (Ahmadizadeh et al., 2020). Ethylene is known to promote the accumulation of EIN2 and EIN3 (Qiao et al., 2009), which showed enhanced expression in MIM775 lines (Fig. 8F), which have higher GALT9 levels. Therefore, we hypothesized that GALT9 mediated possible N-glycosylation of ethylene biosynthetic component (e.g. SAM, ACS and ACO) might have led to the altered accumulation of EIN2 and EIN3, which have affected post-submergence recovery in mutants/transgenics of miR775/GALT9 (Figs 8F and 9). On the other hand, the accumulation and homeostasis of endogenous ABA levels is regulated by AtBG1 (an efficient β-glucosidase), which is involved in the hydrolysis of glucose-conjugated ABA (Jin et al., 2011; Lee et al., 2006). As the glycoprotein AtBG1 possessed N-glycosylation sites, it is possible that GALT9-mediated glycosylation helps in the activation of AtBG1, which in turn hydrolyses the conjugated form of ABA (glucose-ABA) to produce active ABA. ABA increases the expression of NCED3 (an ABA biosynthetic gene) in a positive-feedback mechanism (Barrero et al., 2006). We show the increased expression of NCED3 in MIM775 lines during post submergence recovery, which might be due to GALT9-mediated glycosylation of AtBG1 (Figs 8F and 9).
In summary, we illustrate the role of the miR775-GALT9 module during submergence-induced recovery response. Our results suggest that miR775 promotes recovery after submergence by downregulating the expression of target GALT9. GALT9 overexpression, produced in the MIM775 line, led to the severe senescence of plants (Fig. 6). The miR775-GALT9 module regulates the post-submergence recovery through the regulation of EIN2, EIN3 and SAGs either directly or indirectly. Increased expression of EIN2, EIN3, NCED3 and SAGs in the MIM775 lines promotes cell death and chlorosis, which may be the consequence of increased ROS due to altered GALT9 abundance and changes in the ethylene and ABA biosynthesis pathway genes (Fig. 8E). Based on our findings, we proposed a hypothetical model showing the importance of miR775-GALT9 module in regulating the post-submergence recovery process in Arabidopsis (Fig. 9).
MATERIALS AND METHODS
Evolutionary analysis of miR775 in Arabidopsis
We evaluated the mode of origin of miR775 in Arabidopsis. We calculated the distance between MIR775A (Chr1:29422452) and GALT9 (Chr1:19873727) genes and found to be 9.55 Mbp apart on chromosome 1 (Araport11). Furthermore, sequence alignment between GALT9 and MIR775A did not show any duplication of GALT9 in MIR775A. Consequently, we also searched for conservation or variation of complementary miR775-binding sites in GALT9 homologs through multiple sequence alignment using ClustalX (Larkin et al., 2007). Multiple sequence alignment of the top 19 hits of the GALT9 homologs have shown conserved complementary miR775-binding sites (Fig. S1b). Furthermore, we predicted the promoter of MIR775A using the tool TSSPlant (http://www.softberry.com/cgi-bin/programs/promoter/tssplant.pl). Upstream of MIR775A, we found snoRNA (AT1G09787.1) at a distance of 92 bp, but no transcription start site (TSS) or TATA box. Therefore, we took the upstream sequence extended from MIR775A to AT1G78200, and predicted the TATA box and TSS using TSSPlant at distance of 387 bp and 356 bp, respectively, upstream of MIR775A.
Validation of the miR775 target using degradome data
We analysed the target of miR775 through the analysis of Arabidopsis degradome PARE data available at Sequence Read Archive (SRA) in NCBI (SRR3143654–11-day-old seedling, SRR7093799 – leaf sample of stage 5) using the tool CleaveLand v4.5 (used to find evidence of sliced targets of small RNAs from degradome data) (Addo-Quaye et al., 2009). We retrieved SRA datasets SRR3143654 and SRR7093799, and then further converted these into fastq and, subsequently, into fasta file formats using locally installed tools fastq-dump from SRA Toolkit (https://www.ncbi.nlm.nih.gov/books/NBK158900/) and FASTX-Toolkit (http://hannonlab.cshl.edu/fastx_toolkit/), respectively. The tool fastx_trimmer from FASTX-Toolkit was used for trimming the adapter sequences from these datasets. Furthermore, the mature miR775 sequence was used as a query to search against the whole-genome cDNA sequences of Arabidopsis (https://www.arabidopsis.org/download/index-auto.jsp?dir=%2Fdownload_files%2FSequences%2FAraport11_blastsets). The SRA datasets, mature miR775 sequences and cDNA reference sequences of Arabidopsis were used for the analysis of the miR775 target through the CleaveLand4 pipeline using default settings. The aligned reads of SRA datasets were considered as targets that were being cut at the 10th position in miR775. The miRNA-target cleavage site that has CleaveLand analysis category (0-4) ≤2, lowest Allen et al. (2004) score (0-∞), and highest MFE ratio (0-1) were considered as the best target.
Construction of transgenic lines
For miR775 overexpression, the DNA fragments corresponding to the precursors of 238 bp were cloned, fused to the cauliflower mosaic virus 2X35S promoter gateway cloning vector pMDC32 and transformed into ecotype Columbia (Col-0). For the expression pattern of pGALT9:GUS, the 1256 bp length of promoter sequence of GALT9 was cloned into pCAMBIA1301, fused to the glucuronidase (GUS) reporter gene as pGALT9:GUS and transformed into Col-0 via Agrobacterium-mediated transformation. For localization and target validation, the full-length CDS of GALT9 (1038 bp) was cloned in the pCAMBIA1304 vector frame in the same way as 35S:GALT9:GFP, by removing the stop codon. The target mimic line MIM775 was generated by modifying the IPS1 gene (Todesco et al., 2010). MIM775 target mimic constructs were placed in pGREEN vectors under the constitutive CaMV 35S promoter, which is resistant against BASTA. Col-0 ecotype was used as a control throughout the experiment. Seeds of galt9 (AT1G53290) SALK_015338 were obtained from Arabidopsis Biological Resource Center (ABRC) (Ohio State University). The two independent lines showing the maximum upregulation of miR775, miR775-Oe1 (14.19-fold) and miR775-Oe2 (5.46-fold) were selected on hygromycin B for further analysis in theT3 generation. For miR775 mimic (MIM775) lines, the two lines exhibiting the maximum upregulation of GALT9 in MIM775-1 (3.98-fold) and MIM775-2 (2.55-fold) were selected on BASTA for the further analysis in the T3 generation. Additionally, we generated GALT9-Oe1 (2.44-fold) and GALT9-Oe2 (3.04-fold), and selected on hygromycin B for further analysis in the T3 generation. Primer details are provided in Table S3.
Plant growth and submergence treatment
For all the experiments, the Arabidopsis seeds were first sterilized by seed wash buffer [70% ethanol and 0.1% (v/v) Triton X-100] and then germinated on half-strength Murashige and Skoog (½MS) medium (HiMedia) supplemented with 1% sucrose and 0.8% agar (Murashige and Skoog, 1962). The plants were grown vertically in a controlled environment at 21-22°C, under the 16 h light:8 h dark cycle of white light intensity at 120 µmol m−2 s−1. The above experiments were repeated in triplicates to ensure precision and reproducibility.
For submergence treatment, we used a 3-week-old Col-0 plant. The disinfected containers were filled with Milli-Q water overnight before the treatment to maintain temperature equilibrium (21-22°C), as previously described with some modifications (Yeung et al., 2018), and submerged (8 h after the start of the photoperiod) in ∼6 cm water depth in a dark, humidity-controlled room (Yeung et al., 2018). After 5 days of submergence, de-submerged plants were returned to normal growth conditions for 5 days to follow the post-submergence recovery. Submergence-related experiments were performed at 2:00 PM (8 h after the start of the photoperiod).
Conditions for the expression analysis of MIR775A and its target GALT9 during submergence stress in Arabidopsis seedlings at different time points
For submergence treatment, the disinfected containers were filled with Milli-Q water overnight before the treatment to maintain the temperature equilibrium (21-22°C). The 7 dag seedlings of pMIR7775A: GUS, pGALT9:GUS and Col-0 plants were grown in square Petri plates containing ½ MS media (120×120×17 mm; Praveen Scientific Corporation, India). The plants were submerged (8 h after the start of the photoperiod) at 2:00 PM in ∼6 cm water depth in a dark, humidity-controlled room. The samples were harvested for GUS analysis at early and late time points of 4 h and 24 h. For control plants, the samples were kept in the dark for the same period without being submerged to rule out the effect of the dark on the results. Samples were then harvested for GUS analysis. For GUS analysis, samples were incubated at 37°C for 14.5 h.
Agrobacterium infiltration with transgenic constructs for validation of miRNA target
Four-week-old Nicotiana benthamiana leaves were used for target validation. miR775-Oe and miR839-Oe cloned in the pSITE-4NB vector and sensitive 35S:GALT9:GFP constructs were used for transformations in the Agrobacterium tumefaciens GV3101 strain. For infiltrations, the overnight cultures of individual constructs were harvested and then suspended in an infiltration buffer [pH 5.7, 0.5% glucose 10 mM MgCl2, 150 μM acetosyringone and 10 mM 2-(N-morpholino) ethanesulfonic acid (MES)] and incubated at room temperature for 6 h. For target validation, Nicotiana leaves were infiltrated by 1 ml syringe with the target constructs (sensitive 35S:GALT9:GFP) alone or with EV+35S:GALT9:GFP, 35S:MIR775A+35S:GALT9:GFP and 35S:MIR839A+35S:GALT9:GFP in a 1:1 ratio. The plants were kept in a growth chamber maintained at 26°C (±2) and light intensity of 250 μmol m−2 s−1 and harvested after 48 h for RNA extraction. For target validation, the HPTII gene in the vector was used to normalize target abundance in qRT-PCR experiments. For precursor efficiency in qRT-PCR assays, the Ct value of precursor expression was checked to confirm the synthesis of precursor miRNA in the overexpression construct.
A homology search for miR775 and target GALT9
We performed BLAST for the identification of precursor MIR775A and GALT9 homologs in other plants. We did not find any homolog of miR775 in plants, even by changing the search parameters of BLAST. However, GALT9 shows homologs in other plants. A phylogenetic tree was reconstructed using the BLAST hits of GALT9 through NCBI with the identity and query coverage (https://blast.ncbi.nlm.nih.gov/Blast.cgi).
3D structure prediction and analysis
The functional annotation of GALT9 protein was accomplished by analysing its 3D structure. However, a 3D solved structure of GALT9 is not present in the Protein Data Bank archive (PDB; https://www.rcsb.org/). Therefore, we used I-TASSER (Iterative Threading ASSEmbly Refinement), an online server (https://zhanglab.dcmb.med.umich.edu/I-TASSER/), for protein 3D structure prediction and structure-based function annotation, which use a hierarchical and multiple threading approaches. The 3D structures of the best predicted model of GALT9 were visualized using Chimera v1.6.2 (Fig. S3A). The stereochemical quality check through Ramachandran plot of the predicted GALT9 structure confirmed its stable structure (Fig. S3B). Simultaneously, the functional annotation was also predicted based on assigned enzyme commission (EC) numbers, gene ontology (GO) terms and ligand-binding sites with binding affinity (Zhang, 2008; Roy et al., 2010; Yang et al., 2015).
GALT9 protein and its subcellular localization in planta
To explore subcellular localization of GALT9 protein, we used an Arabidopsis Subcellular Localization Prediction Server (AtSubP, http://bioinfo3.noble.org/AtSubP/index.php). For the prediction of GALT9 subcellular localization, the amino acid composition-based Support Vector Machine (SVM) found that GALT9 is a Golgi apparatus protein. To validate the subcellular localization, the coding sequence of the GALT9 gene (without the stop codon) was cloned into the pCAMBIA1304. The 35S:GALT9:GFP construct was then transformed into the Agrobacterium tumefaciens GV3101 strain. Agrobacterium containing the 35S:GALT9:GFP or GmManI-pBIN2 Golgi apparatus mCherry marker constructs were grown to saturation in Luria-Bertani (LB) medium. Cultures were centrifuged and resuspended in 10 mM MgCl2, 10 mM MES and 150 mM acetosyringone, and kept at room temperature for 2 h. The cultures were then diluted to one OD600 unit and co-infiltrated into the abaxial side of young tobacco (Nicotiana benthamiana) leaf epidermis (4-week-old seedlings grown at 22°C) using a 1 ml syringe without the needle. Transformed leaves were analysed 72 h after infection of the lower epidermis. Subsequently, fluorescence microscopy was performed on a Nikon 80i to record and process the digital images. At least three independently transformed leaves were analysed.
Histochemical detection of GUS assay and microscopy
Histochemical GUS analysis was carried out by putting the samples into appropriate amounts of GUS histochemical buffer [50 mM sodium phosphate (pH 5.7), 50 mM EDTA (pH 8.0), 0.1% Triton X-100, 2 mM potassium ferrocyanide, 2 mM potassium ferricyanide and 1 mM 5-bromo-4-chloro-3-indolyl-β-D-glucuronic acid (X-Gluc)] and incubated at 37°C for 14.5 h. Stained samples were washed with a de-staining solution of ethanol:acetone:glycerol (3:1:1, v/v/v) to remove chlorophyll and then, after incubating the samples in chloral hydrate (TCI Chemicals) for 1 h, microscopy was performed using a Nikon80i and an Olympus SZX16 to record and process the digital images.
Total RNA extraction and quantitative real-time PCR
Gene-specific primers were designed using Integrated DNA Technologies software and custom synthesized by Sigma Aldrich. The total RNA was extracted by using TRIzol (TRI reagent). Purified RNA used for single-stranded cDNA was synthesized from 2.5 μg RNA using an oligo (dT) primer using a high-capacity cDNA reverse transcription kit (Thermo Fisher Scientific). The RT reaction consisted of total RNA, 0.8 µl of 100 mM dNTP mix, 4 µl of 5× reaction buffer, 1 µl of random hexamer primer, 1 µl of oligodT and 1 µl of Revert aid RT enzyme in a final volume of 20 µl. The reaction was carried out at 25°C for 10 min and 37°C for 2 h followed by denaturation at 85°C for 5 min. For performing qRT-PCR, the cDNAs were diluted to 20 ng with sterile MilliQ water. For each tissue type, separate PCR amplification reactions were set up for detecting different genes. The qRT-PCR reaction was set up by mixing 5 μl of 2× PowerUp SYBR Green PCR Master Mix (Applied Biosystems), 0.5 μl of 10 µM each of forward and reverse primers, 2 μl (40 ng) of cDNA and sterile MilliQ water to adjust the reaction volume to 10 μl. qRT-PCR was carried out in an Applied Biosystems ViiA 7 Real-Time PCR System with PowerUp SYBR Green Master Mix. The relative transcript level was calculated by using the 2−ΔΔCT method, which was normalized to ACTIN7 as previously described (Wang et al., 2014; Singh et al., 2020b).
For stem-loop cDNA synthesis, a total of 200 ng of purified RNA was taken then mixed with 0.5 μl 10 mM dNTP and nuclease-free water. The mixture was incubated at 65°C for 5 min, kept on ice for 2 min, briefly centrifuged and added with 4 μl of 5× Reaction buffer (Thermo Fisher Scientific), 2 μl of 0.1 M DTT, 0.25 μl of RiboLock RNase Inhibitor (stock: 20 units/μl), 0.25 μl of RevertAid H Minus M-MuLV Reverse Transcriptase (stock: 200 units/μl), and 1 μl of stem-loop primer (stock: 1 μM) to make a final reaction volume of 20 μl. After gentle mixing, a centrifuge was used to bring the solution to the bottom of the tube and cDNA synthesis was performed as previously described (Varkonyi-Gasic et al., 2007). The cDNA was diluted up to two times before performing the real-time PCR. Fold change was calculated using the formula FC=2−ΔΔCt as previously described (Singh et al., 2017, 2020b; Gautam et al., 2020).
Histochemical detection of H2O2
The H2O2 staining agent 3,3′diaminobenzidine (DAB) (SRL), was dissolved in H2O by adjusting the pH to 3.8 with KOH. Freshly prepared DAB solution was used to avoid auto-oxidation. The 7 dag seedlings were transferred for 5 days for submergence; after 5 days of submergence the seedlings were exposed to 1 h of normal conditions and then transferred for treatment by immersion and infiltration under vacuum with 1.25 mg ml−1 DAB staining solution for 15 min followed by incubation at room temperature for 6 h. The stained seedlings were then bleached out in ethanol:acetic acid:glycerol (3:1:1, v/v/v) solution for 30 min and then images were captured using an Olympus SZX16 microscope. ImageJ system software (https://imagej.nih.gov/ij/download.html) was used for the quantification of ROS intensity. The brown colour visualization of H2O2 was due to DAB polymerization.
Chlorophylls and xanthophyll's estimation
Yeast one hybrid assays
Yeast one-hybrid assays (Y1Hs) were performed to verify the gene-gene interactions, using the Matchmaker Gold Y1H Library Screening System. The full-length CDS of HY5 was subcloned into the pGADT7 AD vector and the promoter of pMIR775A (∼152 bp) was constructed into the public vector according to the ClonExpress II One-Step Cloning Kit. Auto-activation and then interaction analyses were performed.
Determination of auto-activation concentration of AbA
A healthy colony was picked from the bait strains. The colony was resuspended in SD-Ura broth. The dilution was adjusted to 0.1, 0.01, 0.001 and 0.0001, and 10 μl of the culture was patched on the following media: (1) SD/-Ura with AbA (150 ng/ml), (2) SD/-Ura with AbA (250 ng/ml) and (3) SD/-Ura with AbA (500 ng/ml). Vector and construct details are as follows: (1) pGADT7 (for cloning of prey), (2) pAbAi (for cloning of bait), (3) pGADT7-Rec-p53/p53-AbAi (positive control) and (4) pGADT7 transformed in Y1H gold cells (negative control). Colonies were grown for 2-3 days at 28°C on SD/-Ura plates.
All data in this study were obtained from three independent experiments. Error bars indicate standard error of the mean (±s.e.m.). The data were analysed using a two-tailed Student's t-test with GraphPad Prism 9.0.0 software (https://www.graphpad.com/quickcalcs/ttest1/?format=SEM). Asterisks indicate significant statistical differences: ***P≤0.001, **P≤0.01 and *P≤0.05.
Arabidopsis Genome Initiative (AGI) locus identifiers for the genes mentioned in this article are listed as follows: MIR775A (AT1G78206), GALT9 (AT1G53290), SAG12 (AT5G45890), SAG29 (AT5G13170), ORE1 (AT5G39610), RBOHBD (AT5G47910), HY5 (AT5G11260), EIN2 (AT5G03280), EIN3 (AT3G20770) and NCED3 (AT3G14440).
We acknowledge the National Institute of Plant Genome Research for providing necessary research facilities (plant growth facility, confocal/other microscopic facility and other central instrument facility) and internal grants. We also acknowledge the DBT-eLibrary Consortium (DeLCON) for providing access to e-resources.
Conceptualization: A.K.S.; Methodology: V.M.; Software: A.K., V.M.; Validation: V.M., A.K.S.; Formal analysis: V.M., A.S., A.K.S.; Investigation: V.M., A.S., A.K.S.; Resources: A.K.S.; Data curation: V.M.; Writing - original draft: V.M., A.S.; Writing - review & editing: V.M., A.S., N.G., S.S.D., S.Y., A.K., A.K.S.; Visualization: V.M., A.S., N.G., S.S.D., S.Y., A.K., A.K.S.; Supervision: A.K.S., A.S.; Funding acquisition: A.K.S.
V.M. thanks the Department of Biotechnology (DBT), Ministry of Science and Technology, India (fellowship DBT/JRF/15/AL/223), A.S. thanks the Council of Scientific and Industrial Research, India (CSIR) [CSIR-SRA fellowship 13(9166-A)/2021-Pool], N.G. thanks the University Grants Commission (UGC) (fellowship 939/CSIR-UGC NET JUNE 2017), S.S.D. acknowledges the DBT (DBT-RA fellowship DBT/July/2020/22), A.K. thanks the CSIR [CSIR-SRA fellowship 13(9100-A)/2020-Pool). A.K.S. acknowledges the DBT for financial support (BT/PR12766/BPA/188/63/2015) and also support from National Institute of Plant Genome Research (NIPGR) and School of Life Sciences at Jawaharlal Nehru University (JNU), New Delhi.
The authors declare no competing or financial interests.