Abstract
Glutathione (GSH) is a ubiquitous molecule known to regulate various physiological and developmental phenomena in plants. Recently, its involvement in regulating iron (Fe) deficiency response was established in Arabidopsis. However, the role of GSH in modulating subcellular Fe homeostasis remained elusive. In this study, we dissected the role of GSH in regulating Fe homeostasis in Arabidopsis shoots under Fe limited conditions. The two GSH depleted mutants, cad2-1 and pad2-1 displayed increased sensitivity to Fe deficiency with smaller rosette diameter and higher chlorosis level compared with the Col-0 plants. Interestingly, the expression of the vacuolar Fe exporters, AtNRAMP3 and AtNRAMP4, chloroplast Fe importer, AtPIC1, along with AtFer1 and AtIRT1 were significantly down-regulated in these mutants. The expression of these genes were up-regulated in response to exogenous GSH treatment while treatment with BSO, a GSH inhibitor, down-regulated their expression. Moreover, the mutants accumulated higher Fe content in the vacuole and lower in the chloroplast compared with Col-0 under Fe limited condition suggesting a role of GSH in modulating subcellular Fe homeostasis. This regulation was, further, found to involve a GSNO-dependent pathway. Promoter analysis revealed that GSH induced the transcription of these genes presumably via S-nitrosylation of different Fe responsive bHLH transcription factors.
Highlight Glutathione regulates subcellular iron homeostasis under iron deficiency via GSNO dependent transcriptional activation of AtNRAMP3, AtNRAMP4, AtPIC1, AtFer1 and AtIRT1 genes presumably by S-nitrosylation of different iron responsive bHLH factors.
Introduction
Iron (Fe) is an essential micronutrient for plant growth and development. It plays an important role in regulating numerous cellular responses because of its physicochemical properties. This micronutrient is known to coordinate metalloprotein active sites and regulate many important enzymatic reactions required for nitrogen fixation, DNA synthesis and biosynthesis of various phytohormones (Briat et al., 2015). Although Fe is found abundantly in the earth’s crust, it is usually present in an oxidized form and its availability to plants is limited. The deficiency of Fe in plants causes chlorosis and perturbs photosynthesis and oxidative phosphorylation by hampering different electron carriers of photosynthetic machineries as well as mitochondrial electron transport system. Therefore, Fe deficiency is a challenging issue for plant growth, development and productivity.
Plants have evolved highly sophisticated mechanisms for Fe uptake and transport. Two basic strategies for Fe uptake are described in the graminecious and non-graminecious plants (Morrissey and Guerinot, 2009). Among them, the strategy I or reduction based strategy is found in the non-graminecious plants like Arabidopsis where the plasmamembrane H+-ATPases (AHAs) release protons to increase rhizosphere acidification. This promotes the reduction of Fe3+ to the more soluble Fe2+ by ferric reductase oxidase 2 (FRO2) (Robinson et al., 1999). Iron regulated transporter 1 (IRT1), a member of the ZIP metal transporter family, imports the reduced Fe2+ into the root cells (Vert et al., 2002). Also, shoot specific FRO7 plays an important role in Fe delivery to chloroplasts. Besides, FRO3 and FRO8, present in the mitochondrial membrane, also regulate Fe homeostasis in mitochondria (Jeong and Connolly, 2009).
Modulating intracellular Fe homeostasis is crucial particularly under altered Fe conditions and involves several Fe transporters. A high-affinity Fe transporter, natural resistance against macrophage protein (NRAMP), having significant similarity with its mammalian counterpart, was identified in A. thaliana (Curie et al., 2000). Although their functional role in plants was not entirely revealed, complementation assays in yeast showed that these proteins were induced under Fe deficient condition. NRAMP 1, 3 and 4 represent multi-specific metal transporters localized in roots and leaves (Curie et al. 2000; Thomine et al., 2003; Lanquar et al., 2005). In Arabidopsis, NRAMP3 and NRAMP4 were identified to retrieve and export Fe from vacuoles to chloroplast during the Fe deficient condition (Bastow et al., 2018). In fact, the vacuolar Fe released by these two transporters was found to be the primary source of Fe in germinating seeds (Bastow et al., 2018). On the other hand, vacuolar iron transporter 1 (VIT1) was identified as one of the key transporters involved in Fe influx into the vacuole. It thus functions in the vacuolar sequestration process essential for detoxification under the excess Fe condition (Kim et al., 2006). In plants, maintaining the plastidal Fe homeostasis is crucial for survival. Several Fe transporters like permease in chloroplasts 1 (PIC1), non-intrinsic ABC protein 11 (NAP11) and NAP14 were identified as Fe importers that transport Fe across the chloroplast envelope (Duy et al., 2007; Shimoni-Shor et al., 2010). On the other hand, several transporters like yellow stripe like 1 (YSL1) and YSL3 help in maintaining Fe homeostasis by regulating Fe efflux from the chloroplast (Waters et al., 2006; Chu et al., 2010).
Glutathione (GSH) is a multifunctional metabolite that has drawn extensive attention due to its unique structural properties, abundance, broad redox potential, and wide distribution in most living organisms. Along with ascorbate, GSH is considered as one of the most abundant redox couples in plant cells (Foyer and Halliwell, 1976). In plants, GSH is known to play a pivotal role in regulating stress responses as well as growth and development. Phenotypic analysis of GSH deficient Arabidopsis mutants demonstrated that GSH was essentially required for plant development, particularly embryo and meristem development (Vernoux et al., 2000; Cairns et al., 2006; Reichheld et al., 2007; Frottin et al., 2009; Bashandy et al., 2010). The rml1 mutant, which was severely deficient in GSH, developed non-functional root meristem while the shoot meristem remained largely unaffected (Vernoux et al., 2000). In addition to serving as a source of reduced sulfur during secondary metabolite biosynthesis, GSH was widely reported to play crucial role in plant defense signalling network. Ball et al. (2004) reported that several stress responsive genes were altered due to changed GSH metabolism in A. thaliana rax1-1 and cad2-1 mutants of GSH biosynthesis enzyme. Another GSH deficient mutant, pad2-1, was demonstrated to be susceptible to Pseudomonas syringae as well as P. brassicae infections (Glazebrook and Ausubel, 1994; Glazebrook et al., 1997; Parisy et al., 2007; Datta and Chattopadhyay, 2015). Besides, GSH was shown to be involved in the modulation of NPR1-dependent and independent salicylic acid signalling pathways (Ghanta et al., 2011; Han et al., 2013). Subsequently it was reported that GSH induces ethylene biosynthetic pathway via transcriptional as well as post transcriptional regulation of the key enzymes (Datta et al., 2015).
Again, GSH exhibits a wide range of metal chelating activities and plays important role to reduce metal toxicity in plants. Detoxification of heavy metals within plant cell occurs via phytochelatins (PCs) that are synthesized from GSH by the enzyme phytochelatin synthase (Grill et al., 1989; Clemens et al., 1999; Vatamaniuk et al., 1999). On cadmium or copper exposure A. thaliana plants responded by increasing the transcription of glutathione synthetase and glutathione reductase (GR) genes which were involved in GSH synthesis and reduction respectively (Queval et al., 2009). Further, the involvement of GSH was reported in combating metal toxicity against arsenic stress in maize (Requejo and Tena, 2012), chromium toxicity in rice (Zeng et al., 2012; Qiu et al., 2013) and cadmium stress in Pinus and A. thaliana (Schützendübel et al., 2001; Jobe et al., 2012). Cross-talk of GSH with zinc and Fe homeostasis was also reported (Shanmugam et al., 2012). In another study, GSH was reported to trigger the upregulation of genes related to Fe uptake and transport and to increase the Fe concentration in A. thaliana seedlings under Fe deficiency (Koen et al., 2012). GSH-ascorbate redox cycle was studied against Fe deficiency as well (Ramírez et al., 2013). Previous studies demonstrated the association of GSH in maturation of Fe-S molecule and transport of dinitrosyl-Fe complexes in plants (Hider and Kong, 2011; Kumar et al., 2011). GSH also served as the reservoir of nitric acid (NO) by formation of S-nitrosoglutathione (GSNO) complex which helped in NO mediated signalling including Fe deficiency responses (Chen et al., 2010; Ramirez et al., 2011). Subsequently, Shanmugam et al. (2015) revealed the role of GSH in enhancement of Fe deficiency tolerance in plants. The activities of several Fe related transporters like AtIRT1, AtFRO2 and AtFIT was reduced in zir1, a GSH depleted mutant of Arabidopsis, compared to the wild-type (WT). This observation attributed the role of GSH in regulation of Fe transport under Fe-limiting condition. However, the involvement of GSH in modulating subcellular Fe homeostasis in shoot has not been elucidated so far.
In this study, we report that the two GSH depleted mutants, cad2-1 and pad2-1 displayed increased sensitivity towards Fe deficiency. Several organellar Fe transporters, viz. AtNRAMP3, AtNRAMP4, AtPIC1 as well as AtFer1 were found to be down-regulated in both the mutants in addition to AtIRT1. This GSH mediated regulation involved transcriptional activation of the identified genes presumably via S-nitrosylation of different Fe responsive bHLH transcription factors.
Materials and methods
Plant growth, stress treatment and morphological analysis
The seeds of A. thaliana Columbia ecotype (Col-0) along with 2 GSH depleted mutants, cad2-1 and pad2-1, were procured from Nottingham Arabidopsis Stock Centre, UK. The seeds were inoculated in MS medium (Murashige and Skoog, 1962) after surface sterilization with 4 % sodium hypochlorite and Tween 20. The plants were maintained at 21°C with 60 % relative humidity and a photoperiod of 16 h light/8 h dark cycles (Datta et al., 2015).
The 7 d old seedlings were transferred to either minimal iron (MI) or depleted iron (DI) medium and maintained for 7 d. The MI medium contained 10 μM Fe in contrast to 100 μM Fe in the MS medium. For preparation of DI medium, the Fe salt was omitted from MS medium keeping all other components unaltered. In addition, 300 μM ferrozine [3-(2-pyridyl)-5,6-diphenyl-1,2,4-triazine sulfonate] was also added to remove any trace amount of Fe from the medium (Eroglu et al., 2016). For control set, seedlings were maintained in MS medium for the entire period. After 7 d of stress treatment, the morphological parameters including primary root length, lateral root density and rosette diameter were measured using Image J software.
RNA extraction and quantitative-RT PCR analysis
Total RNA was isolated from the tissue samples by Trizol method. Complementary DNA (cDNA) was prepared subsequently using iScript™ cDNA Synthesis Kit (Bio-rad) following manufacturer’s protocol. Quantitative PCR amplification was carried out in CFX96 Touch™ Real-Time PCR Detection System (Bio-rad) using iTaq™ Universal SYBR® Green Supermix (Bio-rad) and gene specific primers (Supplementary Table S1). AtActin2 was used as a reference gene to normalize the relative expression and AtIRT1 was used as Fe-responsive marker gene to confirm the Fe deficient conditions.
Chemical treatment of seedlings
15 d old seedlings were used for chemical treatments. For GSH feeding, freshly prepared 100 μM GSH was used while for BSO treatment, 1 mM BSO was used for 72 h as standardized before (Datta et al., 2015). For DTT treatment, a 5 mM DTT solution was used for 24 h treatment. In case of GSNO feeding, the seedlings were treated with 250 μM GSNO for 72 h (Kailasham et al., 2018). For tungstate treatment, seedlings were treated with 1 mM sodium tungstate for 72 h (Chen et al., 2010). Control seedlings were maintained in half strength MS medium.
Estimation of Fe content
Fe content from different tissues was measured through dry ash digestion method with slight modifications (Jiang et al., 2007). Briefly, tissues were harvested from 15 d old seedlings, washed with HPLC grade water, pat dried, weighed and used for ash preparation. The ash was digested with 0.5 M HNO3 and filtered through Whatman no. 42 filters. The solution was again filtered through Millex® GV 0.22 μm PVDF membrane filter (Merck Millipore). The Fe content was analysed using ICP-OES (iCAP 6300 Duo ICP-OES, Thermo Scientific).
Estimation of Chlorophyll content
The chlorophyll content was estimated following Lichtenthaler (1987). Briefly, 200 mg tissue was homogenized in 80 % acetone, followed by centrifugation at 5000 g for 5 min. The supernatant was used for the estimation of total chlorophyll.
Estimation of total GSH content and GSH:GSSG ratio
GSH estimation was performed following Anderson (1985). Briefly, 200 mg of tissue was homogenized in 5 % sulphosalicylic acid followed by centrifugation at 12000 g for 20 min. GSH was estimated from the supernatant by 5,5’-dithiobis-(2-nitrobenzoic acid) (DTNB) method. For estimation of total GSH, 0.4 mM NADPH and GR were added to the reaction buffer. Optical density was measured at 412 nm using a Double Beam UV-Vis Spectrophotometer (U2900, Hitachi). Amount of GSSG was calculated by subtracting GSH amount from total GSH.
Isolation of chloroplast and vacuole
Chloroplast was isolated from leaves as standardized before (Kleffmann et al., 2004; Ghanta et al., 2014). Briefly, leaves were finely chopped in chloroplast extraction buffer (0.3 M sorbitol, 1 mM MgCl2, 50 mM HEPES/KOH, 2 mM EDTA, 0.04 % β-marcaptoethanol, 0.1 % PVPP, pH 7.8), filtered and centrifuged at 4 °C for 10 min at 1300 g. The pellet was suspended in isolation buffer (0.3 M sorbitol, 1 mM MgCl2, 50 mM HEPES/KOH, 2 mM EDTA, pH 7.8). The solution was loaded on the top of percoll gradient and centrifuged at 4 °C and 8000 g for 20 min. Intact chloroplasts were isolated and washed twice with the isolation buffer.
Vacuoles were isolated from leaves following Zouhar (2016). Briefly, leaves were collected, weighed, cut with razor blade into 2 mm strips and immersed in the protoplast isolation buffer [(1 % (w/v) Cellulase R10, 1 % (w/v) Macerozyme R10, 0.4 M mannitol, 25 mM CaCl2, 5 mM β-mercaptoethanol, 10 mM 2-[N-morpholino]ethanesulfonic acid (MES)-KOH, (pH 5.7)]. The solution was then infiltrated and incubated for 4 h at room temperature with continuous shaking. The released protoplasts were filtered followed by centrifugation at 80 g at 20 °C for 15 min. 10 % Ficoll buffer was added to the pellet for protoplast disruption. The lysed protoplast solution was then treated with 4 % Ficoll buffer followed by vacuole isolation buffer (0.45 M mannitol, 5 mM sodium phosphate, 2 mM EDTA, pH 7.5). The solution was then ultracentrifuged at 50,000 g for 50 min at 10 °C. Vacuoles were collected from the 4 % Ficoll buffer/vacuole buffer interface.
The isolated chloroplast and vacuole fractions were used for Fe estimation as described above.
Fe localization by Perls prussian blue and DAB staining
The Fe localization using Perls prussian blue staining technique was carried out following Rochhartdz et al. (2009). Briefly, the plant samples were fixed in a 6:3:1 methanol:chloroform:acetic acid (v/v) solution and washed twice with ultrapure water. The plants were then transferred to the pre-warmed staining solution (4 % (w/v) K4Fe(CN)6 and 4 % (v/v) HCl) and incubated at room temperature for 1 h. After removing the staining solution, the samples were washed twice with ultrapure water. For DAB intensification, the samples were incubated in preparation solution (0.01 M Na2CO3 and 0.3 % H2O2 in methanol) for 1 h. After washing with the 0.1 M phosphate buffer (pH 7) the samples were incubated in intensification solution (0.025 % (v/v) H2O2 and 0.005 % (w/v) CaCl2) and kept at room temperature for 30 min. The samples were then washed with ultrapure water and photographed.
Vector construction and raising of transformed Arabidopsis plants
Genomic DNA was extracted from shoots following CTAB method. To clone the promoter regions of AtIRT1, AtFer1, AtPIC1, AtNRAMP3 and AtNRAMP4 genes, approximately 1.5 kb of intergenic region upstream of the transcription start site was amplified by PCR using gene specific primers (Supplementary Table S1). The amplified promoter regions were cloned into pCAMBIA1304 between BamHI and BglII restriction enzyme sites to generate AtIRT1pro::GUS, AtNRAMP3pro::GUS and AtNRAMP4pro::GUS constructs and between BamHI and SpeI to produce AtPIC1pro::GUS and AtFer1pro::GUS constructs. These constructs were then transformed into Col-0 plants through Agrobacterium mediated floral dip transformation method (Clough and Bent, 1998). A vector control line harbouring the CaMV35S::GUS construct was generated as well. The transformed lines harbouring the recombinant constructs were selected and maintained up to T2 generations in a growth chamber as described above.
Histochemical GUS assay
The seedlings from the transgenic lines harbouring the AtIRT1pro::GUS, AtFer1pro::GUS, AtPIC1pro::GUS, AtNRAMP3pro::GUS, AtNRAMP4pro::GUS and CaMV35S::GUS constructs were incubated under DI condition or fed with GSH and GSNO as described above. The samples were then infiltrated with GUS staining solution (0.5 mg mL-1 5-bromo-4-chloro-3-indolyl-β-D-glucuronic acid, 0.5 mM potassium ferrocyanide, 0.5 mM potassium ferricyanide, 0.1 % (v/v) Triton X-100, 100 mM phosphate buffer, pH 7.0, and 10 mM EDTA) following Jefferson et al. (1987). The stained samples were washed with 70 % ethanol and photographed.
In silico prediction of S-nitrosylation sites
The 12 basic helix loop helix (bHLH)) transcription factors (TFs) like AtbHLH6, AtbHLH11, AtbHLH18, AtbHLH19, AtbHLH20, AtbHLH25, AtbHLH34, AtbHLH47 (POPEYE), AtbHLH104, AtbHLH105, AtbHLH115, and AtbHLH121 were considered in this study and their protein sequences were retrieved from UniProt database (http://www.uniprot.org/uniprot/). These sequences were submitted to GPS-SNO 1.0 software to predict the presence of S-nitrosylated cysteine residues using threshold level as high (Xue et al., 2010). The cysteine (cys/c) residues which undergo S-nitrosylation were considered as positive hits (+) while the non-nitrosylated cysteine residues were considered as negative hits (-). The enrichment for S-nitrosylated cysteine residues was depicted using CentriMo 5.3.3 web server (https://meme-suite.org/meme/tools/centrimo; Bailey and Machanick, 2012). In addition, the presence of bHLH TF binding motifs were identified in the promoter regions of AtNRAMP3, AtNRAMP4, AtPIC1, AtIRT1 and AtFER1 using PlantPan 3.0 (plantpan.itps.ncku.edu.tw/promoter.php) and MEME suite 5.3.3 (https://meme-suite.org/; Bailey et al., 2009).
Statistical analysis
Statistical analysis of the data was performed using GraphPad Prism version 8.3.0 software (GraphPad Software, San Diego, California USA). The variation of morphological, and biochemical parameters as well as the relative transcript abundance among different samples were analyzed following two-way ANOVA followed by Sidak’s multiple comparison tests. The statistical significance at P ≤ 0.05 was considered to determine the difference between two sets of data. Details of replicates, sample size, and the significance levels of P-values were indicated in respective figure legends. The data were represented as mean ± standard error of mean (SEM).
Results
GSH depleted mutants displayed increased sensitivity to Fe deficient conditions
To explore if GSH plays a role in regulating Fe homeostasis, two GSH depleted mutants, cad2-1 and pad2-1, containing only 40 % and 22 % GSH respectively, were selected for this study. The MS medium containing 100 μM Fe was used as a control medium while the MI medium containing 10 μM Fe and the DI medium lacking any Fe source and containing the Fe-chelator, ferrozine were used as Fe deficient media. Since, the expression of AtIRT1 gene was found to be the highest after 7 d treatment, this time point was used for all downstream analyses (Supplementary Fig. S1). The mutant lines when grown under MI and DI conditions displayed increased sensitivity to Fe deficiency as compared with the Col-0 plants (Fig. 1). The rosette diameter was found to be significantly reduced in the mutants compared with the Col-0 plants after 7 d of treatment. In addition, the root morphology was also altered in response to Fe deficiency. The primary root length decreased in all 3 lines with the shortest root length in the pad2-1 plants under both MI and DI conditions. Lateral root density was found to be considerably increased in all the 3 lines. Since Fe deficiency is widely known to affect the photosynthetic machinery, we estimated the total chlorophyll content. While all the 3 lines displayed chlorosis in response to Fe deficiency, the mutant lines showed significantly lower total chlorophyll content under both MI and DI conditions as compared with the Col-0 plants.
To determine any correlation between GSH and Fe contents, we also analyzed the total GSH content, GSH:GSSG ratio as well as the Fe content under control and Fe deficiency conditions. Interestingly, the mutant lines exhibited significantly lower Fe content in the shoot as well as the root as compared with Col-0 plants (Fig. 2A, D). The lower Fe content in the mutant lines was also visualized by Perls-DAB staining (Supplementary Fig. S2). On the other hand, the total GSH content increased under MI and DI conditions while the GSH:GSSG ratio decreased. As expected, the mutant lines displayed lower GSH levels in both the shoot and the roots as compared with the Col-0 plants (Fig. 2B, C, E, F).
GSH played a crucial role in regulating Fe transporters and ferritin genes
Since the GSH depleted mutants were sensitive to Fe deficiency, the next pertinent point was to identify the Fe transporters that might be regulated by GSH. To this end, the expression of the Fe transporter genes were analyzed in the mutant lines and compared with the Col-0 plants. Interestingly, it was found that the expression of AtNRAMP3, AtNRAMP4, AtPIC1 and AtIRT1 genes were significantly down-regulated in the mutant lines as compared with the Col-0 plants (Fig. 3). On the other hand, the expression of AtYSL1 and AtYSL3 genes were found to be up-regulated in the mutant lines. Besides, several genes involved in regulating Fe homeostasis were also analyzed. Among them, AtFer1 gene was found to be significantly down-regulated in the mutant lines as compared with the Col-0 (Fig. 4). This observation strongly suggested the involvement of GSH in modulating Fe homeostasis in plants. Since the AtNRAMP3, AtNRAMP4, AtPIC1, AtFer1 and AtIRT1 genes are known to be induced by Fe deficiency, we checked their expression under DI condition. It was found that under DI condition as well, the relative transcript abundance for all these genes were significantly lower in the mutant lines further supporting the previous observations (Fig. 5).
Exogenously altered GSH level was sufficient to regulate Fe responsive gene expression
To further confirm the GSH-mediated regulation of the identified Fe responsive genes, Col-0 seedlings were exogenously treated with GSH or an inhibitor of GSH biosynthesis, BSO. In addition, a non-specific reducing agent DTT was also used for treatment. A half strength MS medium was used as a control set. To confirm the efficiency of the feeding treatments, total GSH content was estimated from each set. The total GSH level was found to be significantly increased in the GSH-fed plants, decreased in the BSO-fed plants while no significant alteration was observed in the DTT-fed plants (Fig. 6A). Next, the relative transcript abundance of the identified genes was analyzed. It was observed that the expression of the AtNRAMP3, AtNRAMP4, AtPIC1 and AtIRT1 transporters along with the AtFer1 gene were induced in response to GSH feeding while the expression was decreased in response to BSO treatment (Fig. 6B-F). The DTT feeding, however, failed to significantly alter the expression levels thus indicating that non-specific reducing condition was not sufficient for this regulation.
Since the pad2-1 mutant is severely deficient in GSH content, we supplemented these mutant seedlings with exogenous GSH or DTT. It was observed that only GSH supplementation, and not DTT feeding, could compensate for its deficiency and resulted in an increased expression of the identified Fe responsive genes (Supplementary Fig. S3).
GSH modulated subcellular Fe homeostasis under Fe depleted condition
The Fe transporters AtNRAMP3 and AtNRAMP4 are vacuolar exporters for Fe while the AtPIC1 is responsible for Fe influx into the chloroplast. In addition, AtFer1 acts as a major Fe accumulator in the chloroplast. Since GSH was found to regulate the expression of all of these genes, it was hypothesized that GSH might be involved in modulating the subcellular Fe homeostasis under Fe deficient condition. To dissect this, the organellar Fe content was estimated from the cad2-1 and pad2-1 mutants along with the Col-0 plants under both control and DI conditions. It was observed that the chloroplast Fe content was significantly lower in the mutant lines as compared with the Col-0 plants under control as well as DI conditions (Fig. 7A). On the other hand, the vacuolar Fe content was lower in the mutants under control condition but higher under DI condition as compared with the Col-0 plants (Fig. 7B). This observation suggested that the mutant lines with depleted GSH levels failed to efficiently export the Fe from vacuoles and channelize them into the chloroplast under DI condition. The lower vacuolar Fe content in the mutants under control condition could be attributed to their lower AtIRT1 expression and subsequent impaired Fe uptake.
GSH mediated regulation of Fe homeostasis involved GSNO
Since the association of GSH with GSNO is known in regulating Fe deficiency response, we were curious if this GSH mediated modulation of subcellular Fe homeostasis also involved GSNO. Therefore, the Col-0 and pad2-1 plants were treated with GSH and GSNO in combination with their inhibitors and DI condition. The expression of AtNRAMP3, AtNRAMP4, AtPIC1, AtIRT1 and AtFER1 genes were then analyzed. When Col-0 plants were treated with GSH or GSNO under DI condition, stronger induction of the genes was observed as compared with DI condition alone (Fig. 8). Treating the plants with GSNO along with the GSH inhibitor, BSO under DI condition did not display this effect. Yet again, plants treated with GSH in combination with the nitrate reductase inhibitor, tungstate failed to induce the gene expression even under DI condition. These observations suggested that the GSH mediated regulation of the identified Fe responsive genes occurred in a GSNO-dependent manner. Similarly, in case of pad2-1 mutant, the treatment with GSH or GSH-tungstate combination under DI condition showed a similar trend like that of Col-0 plants. On contrary, GSNO treatment under DI condition failed to augment gene expression indicating that GSNO alone was not sufficient to trigger the expression of these genes under the GSH depleted condition.
GSH triggered promoter activation of the Fe responsive genes
Since GSH modulated the relative transcript abundance for the identified Fe responsive genes, it was assumed that a probable transcriptional regulation might be involved. To identify the mechanism of this regulation, transgenic lines harboring the AtNRAMP3pro::GUS, AtNRAMP4pro::GUS, AtPIC1pro::GUS, AtFer1pro::GUS and AtIRT1pro::GUS constructs were generated (Supplementary Fig. S4). A transgenic line harboring the CaMV35S::GUS construct was used as a negative control. The promoter activities in response to GSH and GSNO treatment were analyzed by histochemical GUS assay. The DI condition was used as a positive control for these treatments. The GUS activity in all the transgenic lines was found to be induced in response to both GSH and GSNO treatments which was also supported by the relative transcript abundance of the GUS gene (Fig. 9). On contrary, the GUS activity as well as its expression was unaltered in the CaMV35S::GUS containing negative control line. This observation strongly suggested that the GSH mediated regulation of these Fe responsive genes occurred via transcriptional activation.
GSH mediated transcriptional induction presumably involved S-nitrosylation of Fe-responsive TFs
Seventeen bHLH TFs are known to be involved in the transcriptional regulation of various Fe responsive genes (Gao et al., 2019; 2020). Among them, the AtFIT, AtbHLH38, AtbHLH39, AtbHLH100, and AtbHLH101, were reported to be regulated directly or indirectly by GSNO (Meiser at al., 2011; Kailasham et al., 2018). This made us curious to check if the other known Fe responsive TFs were also regulated in a GSNO-dependent fashion. Therefore, the 12 known Fe responsive bHLH TFs, viz. AtbHLH6, AtbHLH11, AtbHLH18, AtbHLH19, AtbHLH20, AtbHLH25, AtbHLH34, AtbHLH47 (POPEYE), AtbHLH104, AtbHLH105, AtbHLH115, and AtbHLH121 were selected for this study and screened for probable S-nitrosylation sites using the GPS-SNO 1.0 software. Among them, AtbHLH11, and AtbHLH34 were found to contain 2 putative S-nitrosylation sites while AtbHLH6, AtbHLH18, AtbHLH19, AtbHLH105, and AtbHLH115 contained 1 putative S-nitrosylation site each (Supplementary Fig. S5; Supplementary Table S2). No sites were predicted in rest of the TFs. The enrichment analysis for the presence of S-nitrosylated cysteine residues in these TFs was performed as well (Supplementary Fig. S6). The bHLH TFs were reported to bind at the conserved E-box DNA binding (CANNTG) motif (Toledo-Ortiz et al., 2003). Therefore, the promoter regions of AtNRAMP3, AtNRAMP4, AtPIC1, AtIRT1 and AtFer1 genes were analyzed for the presence of this conserved CANNTG motif. Interestingly, the promoters of all these identified Fe responsive genes contained the conserved E-box motif indicating their probable interaction with the bHLH TFs (Supplementary Fig. S7). Together, these observations suggested that the GSH mediated regulation of the identified Fe responsive genes presumably involved transcriptional modulation via these S-nitrosylated bHLH TFs.
Discussion
In the last few decades the role of GSH in regulating multiple stress responses in plants was elaborately studied. Recently, emerging evidences suggested a positive role of GSH in mitigating Fe deficiency responses in plants (Koen et al., 2012; Ramirez et al., 2013). It was demonstrated that during Fe deficiency in Arabidopsis, the elevated level of GSH could significantly regulate different Fe responsive genes like FRO2, IRT1, NAS4, and FIT1 to enhance Fe uptake and transport through roots (Koen et al., 2012). Subsequently, this GSH mediated regulation was found to occur in a GSNO dependent manner (Garcia et al., 2010, Shanmugam et al., 2015; Kailasam et al., 2018). However, the role of GSH in regulating subcellular Fe homeostasis in plants was not elucidated so far. In our previous study, comparative proteomic analysis of pad2-1, a GSH depleted mutant, revealed the down-accumulation of AtFer1 protein in shoot as compared with the Col-0 plants (Datta et al., 2015). This clue prompted us to investigate the role of GSH in regulating the expression of different Fe responsive genes in Arabidopsis shoot. Since ferritin is abundantly found in the chloroplast, we were also curious to identify any role of GSH in maintaining chloroplast Fe homeostasis. To begin with, we selected two GSH depleted mutants cad2-1 and pad2-1, and analyzed their response under Fe limited conditions. Both the mutants displayed severe sensitivity to Fe deficiency as compared with the Col-0 plants with smaller rosette diameter and higher chlorosis level (Fig. 1). This observation was in line with the earlier study where Fe deficiency responses were found to be aggravated by reduced GSH content in Arabidopsis (Ramirez et al., 2013; Shanmugam et al. 2015). The next approach was to identify the Fe transporters and homeostasis related genes that might be responsible for this observation. Out of the 32 Fe transporters analyzed, 2 vacuole Fe exporters, AtNRAMP3 and AtNRAMP4, and a chloroplast Fe importer, AtPIC1 were found to be significantly down-regulated in the mutants, while 2 chloroplast Fe exporters, AtYSL1 and AtYSL3 were up-regulated (Fig.3–5). Several Fe homeostasis related genes were analyzed as well among which AtFer1 was found to be down-regulated in the mutants. Exogenously altered GSH levels also supported this observation (Fig 6).
The altered expression of these organellar Fe transporters suggested a possible role of GSH in modulating the subcellular Fe homeostasis under Fe limited condition. The vacuole, during Fe sufficient condition stores the excess amount of Fe to rescue the cell from Fe mediated oxidative damage. This stored Fe is released under Fe limited condition to mitigate Fe deficiency responses in plants (Morrissey and Guerinot, 2009). On contrary, AtPIC1, a chloroplast Fe importer, was found to be up-regulated during Fe deficiency to maintain the Fe homeostasis in chloroplast (Duy et al., 2005). In addition, AtFer1 was reported to act as Fe reservoir in chloroplast and to diminish the Fe deficiency responses (Briat et al., 2010). In a previous study, it was reported that AtNRAMP3 and AtNRAMP4 function as major vacuolar Fe exporters in germinating seeds where vacuolar reserve is the primary source of Fe (Bastow et al., 2018). Further, it was demonstrated that the Fe mobilization into chloroplast was limited when the vacuolar Fe could not be retrieved. The expression of AtFer1 was also reported to be significantly reduced in the nramp3nramp4 mutant with perturbed vacuolar Fe efflux. In this study, GSH was found to positively regulate the expression of AtNRAMP3, AtNRAMP4, AtPIC1 and AtFer1 genes. This made us inquisitive if GSH can help in channelizing the Fe retrieved from vacuoles into the chloroplast in shoot under Fe deficient conditions. We, therefore, analyzed the Fe content in the vacuoles and the chloroplasts in the GSH depleted mutants. Surprisingly, both the vacuolar and chloroplast Fe contents were found to be lower in the mutants under Fe sufficient condition (Fig. 7). This lower Fe accumulation in the mutants can be attributed to their impaired Fe uptake due to reduced AtIRT1 expression in the roots. On contrary, the vacuolar Fe content was found to be higher in the mutants while the chloroplast Fe content remained lower under DI condition. This observation strongly suggested that the GSH depleted mutants failed to efficiently retrieve the vacuolar Fe under DI condition and channelize them into the chloroplast.
GSH combines with NO leading to the formation of GSNO which acts as an important signaling intermediate. Earlier, GSNO was reported to act as a key modulator of Fe responsive genes in Arabidopsis (Koen et al., 2012; Shanmugam et al. 2015; Kailasam et al., 2018). This made us interested to dissect the role of GSNO in regulating these Fe responsive genes. The Col-0 and pad2-1 plants, when fed with GSH or GSNO under DI condition showed marked increment in the expression of all 5 genes (Fig 8). On contrary, feeding the plants with GSNO/BSO combination under DI condition, did not display this augmented expression pattern as compared with the DI condition alone. Again, no induction in gene expression was observed in case of combined GSH/tungstate treatment even under DI condition. These observations suggested that GSH regulated the subcellular Fe homeostasis via a GSNO mediated pathway.
The next pertinent question was how this GSH-GSNO module induced the expression of these genes. To dissect this mechanism, 5 different transgenic lines harbouring AtNRAMP3pro::GUS, AtNRAMP4pro::GUS, AtPIC1pro::GUS, AtFer1pro::GUS, and AtIRT1pro::GUS constructs were generated and the transgenic plants were used for promoter analysis. Interestingly, the histochemical GUS assay confirmed that all the 5 promoters were activated in response to GSH or GSNO treatments. This strongly indicated that the GSH-GSNO module regulated the Fe responsive genes via transcriptional activation. Since neither GSH nor GSNO can directly bind to these promoters, this transcriptional activation presumably involved one or more TFs. Earlier reports have identified 17 bHLH TFs that are involved in regulating various Fe responsive genes in Arabidopsis (Yuan et al., 2008; Gao et al., 2019; 2020). Among them, AtFIT1 (bHLH29, clade IIIa) and the clade Ib bHLH factors (bHLH38, bHLH39, bHLH100, bHLH101) were reported to regulate two crucial Fe responsive genes, AtIRT1 and AtFRO2 (Colangelo and Guerinot, 2004; Yuan et al., 2008). Further, these TFs were found to be regulated directly or indirectly by GSNO (Darbani et al., 2013; Kailasam et al., 2018). This GSNO mediated regulation often involved S-nitrosylation of the TFs (Darbani et al., 2013). We were curious if the rest of the 12 bHLH TFs could be regulated by GSNO as well and analyzed for the presence of putative S-nitrosylation sites. Seven out of these 12 TFs were found to contain at least one putative S-nitrosylation sites supporting the hypothesis for a GSNO-mediated regulation (Supplementary Fig. S5-6). Moreover, the identified Fe responsive genes, viz. AtNRAMP3, AtNRAMP4, AtPIC1, AtIRT1 and AtFer1 were found to contain conserved E-box motifs in their promoter regions indicating the probable transcriptional regulation via the S-nitrosylated bHLH TFs (Supplementary Fig. S7). This remains an open area for in depth analysis in the future.
In summary, it can be postulated that during Fe deficiency the accumulation of GSH in cells can activate the vacuolar Fe exporters like AtNRAMP3 and AtNRAMP4 to facilitate Fe export from the vacuolar reserve. On the other hand, GSH was found to trigger the expression of chloroplast Fe importer, AtPIC1 and Fe responsive gene AtFer1 to maintain chloroplast Fe content during Fe deficiency in plants (Fig. 10). This regulation involved GSH-GSNO mediated transcriptional activation of these genes presumably via S-nitrosylation of different Fe responsive bHLH TFs.
Supplementary data
Fig. S1. Expression of AtIRT1 gene in Col-0 shoot under DI condition. 7 d old MS grown seedlings were transferred to DI medium and samples were collected after 0 d, 2 d, 5 d, 7 d, and 10 d of treatment.
Fig. S2. Perls-DAB staining of Col-0, cad2-1 and pad2-1 plants grown under MS condition.
Fig. S3. Response of pad2-1 plants to exogenous GSH and DTT treatments. (A) Total GSH content. (B-F) Relative transcript abundance of the identified Fe responsive genes.
Fig. S4. Screening of transgenic lines harbouring (A) CaMV35S::GUS, (B) AtPIC1pro::GUS, (C) AtNRAMP3pro::GUS, (D) AtIRT1pro::GUS, (E) AtFer1pro::GUS, and (F) AtNRAMP4pro::GUS constructs.
Fig. S5. Prediction of putative S-nitrosylation sites in different Fe responsive bHLH TFs using GPS-SNO 1.0 Software and Centrimo 5.3.3 web server.
Fig. S6. Enrichment analysis for the presence of putative S-nitrosylation cysteine residues in different Fe responsive bHLH TFs.
Fig. S7. In silico promoter analysis of the identified Fe responsive genes for the presence of conserved E-box motif.
Table S1. List of primers used.
Table S2. Prediction of putative S-nitrosylation sites using GPS-SNO 1.0 software.
Author Contributions
RD and SP conceived and designed the research plan; RS raised the transgenic lines, performed expression analysis, feeding experiments and GUS assay; SG standardized the plant growth and stress treatments, PK performed biochemical experiments, SS prepared the constructs; CR helped in maintaining the transgenic lines, DS performed in silico analysis; RD and SP analyzed the data and prepared the manuscript.
Acknowledgement
This work has been supported by the Science and Engineering Research Board, Government of India [ECR/2017/000231] and Start up Grant, University Grant Commission, Government of India, India [F30-363/2017(BSR)]. We thank the central instrumentation facility of the Department of Botany, University of Calcutta and Department of Botany, Dr A. P. J. Abdul Kalam Government College. We thank Prof. Manoj Prasad, Staff Scientist VII, NIPGR, New Delhi for providing us the pCAMBIA1304 vector.
Footnotes
Email addresses: Ranjana Shee: ranjana.shee22{at}gmail.com, Soumi Ghosh: soumimpplab{at}gmail.com, Pinki Khan: pinkimpplab{at}gmail.com, Salman Sahid: salmansahid1991{at}gmail.com, Chandan Roy: suchandan.roy35{at}gmail.com, Dibyendu Shee: dibyendushee01{at}gmail.com, Soumitra Paul: psoumitra{at}ymail.com, Riddhi Datta: riddhi.bot{at}gmail.com