Nucleoredoxin 1 in Wheat: Its Genes and Role in Maintaining Balance and Stress Resistance

BLASTP analysis using the A. thaliana NRX1 protein as a query identified three putative homologous genes in the wheat (T. aestivum) genome, located on chromosomes 2A, 2B, and 2D (Table 1). These genes encode five distinct protein isoforms: NRX1.1a, NRX1.1b, NRX1.2a, NRX1.3a, and NRX1.13b. All predicted wheat NRX1 proteins exhibited high sequence similarity to the Arabidopsis NRX1 and contained conserved thioredoxin domains, indicating a potential redox‐regulatory role. Domain validation through SMART and PFAM confirmed the presence of three redoxin domains and a C1‐like domain in each isoform, consistent with their classification within the thioredoxin superfamily (Figure S1). Physicochemical properties calculated using the Expasy ProtParam tool (Swiss Institute of Bioinformatics, n.d.) revealed that the molecular weights of wheat NRX1 proteins ranged from approximately 63.37 to 65.12 kDa. Their theoretical isoelectric points (pI) were between 4.78 and 4.91, indicating a slightly acidic nature (Table 1). These values suggest potential solubility and stability in diverse cellular environments. Subcellular localization predictions indicated that these proteins predominantly localize in the cytoplasm. However, additional signals suggest distribution to the chloroplast, nucleus, endoplasmic reticulum, vacuole, cytoskeleton, and extracellular space (Table 1), pointing toward multifunctional roles in various cellular compartments.

Phylogenetic analysis was conducted on genes for NRX1 proteins from T. aestivum and 29 other plant species. The analysis revealed clustering into seven distinct clades, reflecting evolutionary relationships among the 30 species (Figure 1). Wheat NRX1s grouped closely with orthologs from Hordeum vulgare, Triticum turgidum, Aegilops tauschii, Thinopyrum intermedium, and Triticum urartu, indicating strong conservation within the Triticeae lineage. Among all the 30 species, Saccharum officinarum had the highest number of genes for NRX1 proteins (17), followed by Andropogon gerardi (9), Panicum virgatum (6), and T. aestivum (5). Other species, such as Setaria viridis (4), Miscanthus sinensis, Oryza sativa, and Sorghum bicolor (3 each), displayed variable NRX1 gene counts, suggesting lineage‐specific gene expansion (Figure 1A). MEME analysis identified 10 conserved motifs across all wheat NRX1 protein genes, including characteristic redoxin catalytic sites, supporting the potential functional roles of the proteins in redox regulation (Figure 1B). Additionally, gene structure analysis using GSDS revealed conserved exon–intron patterns among wheat NRX1 genes, highlighting the evolutionary stability of their gene architecture (Figure 2).

Subcellular localization prediction using WoLF PSORT indicated that all five wheat NRX1 proteins are primarily localized in the cytosol, aligning with their proposed roles in redox regulation. NRX1.1a and NRX1.2a also showed potential extracellular localisation, while NRX1.1b, NRX1.2a, and NRX1.3a exhibited the possibility of chloroplast localization, suggesting additional roles in plastid redox processes (Table 1). Promoter analysis using PlantCARE revealed a total of 79 cis‐regulatory elements (CREs) across the NRX1 gene promoters, encompassing both hormone‐responsive and stress‐related elements (Figure 3A,B). Hormonal regulatory motifs included MeJA‐responsive elements, ABRE (abscisic acid), ERE (ethylene), GARE (gibberellin), SARE (salicylic acid), and ARE (auxin), suggesting complex hormonal crosstalk in NRX1 regulation. Stress‐responsive elements included ASIE (anoxic stress), D&SRE (defense and stress), and LTRE (low‐temperature responsiveness), highlighting their potential roles under environmental challenges. Notably, multiple MYB binding sites (MYBBS) involved in flavonoid biosynthesis, along with five MYBHv1 binding sites (MYBHv1BS), were also detected, indicating transcriptional regulation via MYB transcription factors. Comparative promoter analysis of related antioxidant genes (catalases, SODs, APXs) and HDR genes revealed that while most showed a diverse regulatory element profile, HDR had fewer cis‐elements, suggesting a more constitutive or housekeeping expression pattern (Figure 3B).

PTM prediction using NetPhos (DTU Health Tech Bioinformatic Services, n.d.‐a) identified multiple serine and threonine phosphorylation sites across all NRX1 proteins, suggesting involvement in kinase‐mediated signaling (Figure 4B). NetNGlyc analysis (DTU Health Tech Bioinformatic Services, n.d.‐b) predicted one N‐glycosylation site in each of NRX1s (Figure 4A), indicating potential for post‐translational modulation of NRX1 stability and function. The secondary structure of all these proteins was predicted using the SOPMA webserver, and α–helices, β–sheets, extended strands, and random coils were predicted between 35.42%–36.71%, 10.66%–11.03%, 22.41%–23.78%, and 29.37%–31.03%, respectively (Figure S2). Disordered regions were analyzed using the Mobi DB server (https://mobidb.bio.unipd.it↗). All three proteins showed no disordered regions and contained the functional domain (Figure S3).

Protein–protein interaction analysis using the STRING database predicted that wheat NRX1 proteins are highly connected with other redox‐related proteins, including thioredoxins, peroxiredoxins, and redoxins, consistent with them playing a central role in redox regulation networks (Figure 5). A particularly strong interaction was predicted with 4‐hydroxy‐3‐methylbut‐2‐enyl diphosphate reductase (HDR), a key enzyme in the methylerythritol phosphate (MEP) pathway, which links NRX1 to isoprenoid biosynthesis. Table 2 below summarizes functional annotations and subcellular localization of NRX1‐interacting proteins. Many of these proteins localize to chloroplasts, mitochondria, cytoplasm, and nuclei—compartments critical for redox homeostasis (Table 2). For instance, several thioredoxins and thioredoxin‐like proteins (e.g., Traes_2BL_9C3ACD499.1, Traes_2DL_397237F3E.1, and Traes_5BL_369BF7271.1) co‐localize in the chloroplast and mitochondria, indicating potential coordination of redox signaling across organelles. KEGG pathway enrichment further indicated NRX1's association with antioxidant defense and metabolic pathways, especially isoprenoid biosynthesis via the MEP pathway. These findings collectively suggest that wheat NRX1 proteins may act as critical regulators of redox‐dependent signaling and secondary metabolism.

Expression data retrieved from WheatExp (www.wheat‐expression.com↗) revealed that most wheat NRX1 genes are substantially upregulated in response to both abiotic and biotic stresses (Figure 6). Under drought conditions, NRX1 expression showed a transient induction followed by suppression, consistent with a time‐dependent regulatory role during stress adaptation (Figure 6A). Similarly, biotic stresses such as powdery mildew and stripe rust also triggered NRX1 expression, suggesting its involvement in plant defense signaling (Figure 6A). In parallel, key antioxidant defense genes—including superoxide dismutase (SOD), catalase (CAT), and ascorbate peroxidase (APX)—also displayed dynamic expression patterns under the same stress conditions. SOD genes were notably responsive to drought and stripe rust (Figure 6D), while CAT genes showed variable upregulation under both heat and biotic stress conditions (Figure 6C). APX genes exhibited strong stress‐inducible expression under all tested conditions, particularly under heat stress (Figure 6B). These coordinated expression trends suggest that NRX1 may function in a broader redox regulatory network, interacting with core antioxidant systems to mitigate oxidative damage during environmental stress.

Exonic regions of NRX1 were targeted in Cas9‐wheat plants using BSMV‐delivered gRNAs (Figure 7). Targeted wheat plants exhibited mutations confirmed by PCR amplification and sequencing of the target regions (Figure 8E,F). Genome‐specific primers were designed for the amplification of target regions of NRX1 alleles from A (Figure 8E‐a), B (Figure 8E‐b), and D (Figure 8E‐c) genomes of inoculated plants to evaluate the mutation rates induced by different BSMV vectors. Target regions were amplified using the genomic DNA from the M0 and M1 generations. We identified three plants with deletions in the B and D subgenomes. Seeds from the confirmed mutants in the M1 generation were used for downstream phenotypic and biochemical assays in the M2 generation.

To evaluate the physiological impact of CRISPR‐Cas9‐mediated mutations in NRX1, two mutant lines nrx1‐bd and nrx1‐b were used (labeled as XD44 and XD13 in Figure 8F and Table 3). Nrx1‐bd and nrx1‐b contain small deletions in the conserved active sites, as shown in Figures 8F and S4 and Table S1. They result in frameshift mutations in the modified NRX1 protein produced, rather than knockout of the complete NRX1 gene, which would be a desirable additional experiment. Phenotypic traits were assessed in mutants (nrx1‐bd and nrx1‐b) and wild‐type BW208 (Bobwhite 208) wheat lines under stripe rust and salinity stress (Figure 9). Under control conditions, shoot lengths ranged between 42–44 cm and root lengths between 18 and 21 cm across all lines. Under salinity stress, nrx1‐bd and nrx1‐b exhibited reduced shoot and root lengths (43.5% and 35% in nrx1‐bd and 47.4% and 36% in nrx1‐b, respectively), compared to BW208 (24.5% shoot; 27% root). Fresh biomass of shoots and roots also declined sharply in mutants, with a reduction of 55.6% and 49% in shoot weight and 38.6% and 39% in root weight in nrx1‐bd and nrx1‐b, respectively, under salinity, compared to BW208's 33% and 31% reductions, respectively. These results highlight compromised stress tolerance in mutant lines. Stripe rust infection caused major reductions in shoot weight. There was a 22% and 10.6% reduction in shoot weight in nrx1‐bd and nrx1‐b, respectively, compared to an 8.4% reduction in the shoot weight of BW208.

Stripe rust infection was evaluated based on infection type and severity of symptoms. Wild type BW208 exhibited the lowest infection type, with an average score of 2 on a 0–9 scale, which indicates a relatively resistant response, while nrx1‐b and nrx1‐bd plants showed higher infection types of 4 and 5, respectively. Disease severity, based on the percentage of damaged leaf area, showed a similar trend. BW208 showed low average severity at 16%, while the mutants nrx1‐bd and nrx1‐b exhibited increased severities of 26% and 29%, respectively. The results suggest the susceptibility of BW208 toward stripe rust increased after mutation in the NRX1 genes.

The altered sensitivity observed in nrx1‐bd and nrx1‐b mutants is consistent with a functional role of NRX1 in redox regulation, stress resilience, and potentially isoprenoid‐mediated antioxidant defense pathways. The summarized phenotypic data are presented in Figure 9.

Total chlorophyll content was measured to assess the impact of salinity and stripe rust stress on the photosynthetic efficiency of CRISPR‐Cas9‐induced mutant lines (nrx1‐bd and nrx1‐b) compared to the wild‐type (BW208). Chlorophyll was extracted using the dimethyl sulfoxide (DMSO) method, and absorbance was recorded at 470 nm, 645 nm, and 665 nm for total chlorophyll calculation. Under control conditions, all lines showed high chlorophyll content, with BW208 and nrx1‐b exhibiting slightly higher levels, while nrx1‐bd had slightly lower content (Figure 10). Chlorophyll contents were significantly affected by salinity in all the plants (p < 0.05). Stripe rust also had a significant effect on normal plants, but the chlorophyll contents of the mutants were affected more (p < 0.05). The highest reduction in chlorophyll contents was observed in nrx1‐bd mutants under stress conditions. A reduction of up to 35% in chlorophyll a, 24% in chlorophyll b, and 28% in total chlorophyll content was observed in the mutants under stripe rust stress compared to 20%, 13%, and 17% reduction in the normal BW208 plants under stripe rust. Salinity caused significantly more reduction in chlorophyll contents with 49% in chlorophyll a, 45% in chlorophyll b, and 47% reduction in total chlorophyll, while the reduction was 42%, 32%, and 37% respectively, in normal BW208 plants under salinity stress (p < 0.05). In summary, there was a major impact on the carotenoid contents of the mutants under stress conditions. Overall, the photosynthetic pigments in NXR1 knockout plants were substantially more impacted by both kinds of stresses compared to wild‐type plants (Figure 10).

To further investigate the physiological effects of the CRISPR‐Cas9‐induced mutations in NRX1, oxidative stress levels and antioxidant enzyme activities were evaluated under control, salinity, and stripe rust conditions. To investigate the redox‐regulatory role of NRX1 under stress conditions, lipid peroxidation and antioxidant enzyme activity in wild‐type (BW208) and CRISPR‐Cas9 mutant wheat lines (nrx1‐bd and nrx1‐b) exposed to salinity and stripe rust were evaluated. The results highlight clear differences in oxidative stress indicators between wild‐type and mutant genotypes (Figure 11). Malondialdehyde (MDA) levels, a marker of lipid peroxidation and membrane damage, remained low under control conditions across all genotypes.

An increase in the MDA content was observed in plants under stress conditions, and MDA contents of mutants were affected significantly more (p < 0.05, Tukey's HSD test), indicating heightened membrane damage due to insufficient ROS detoxification. The increase in MDA content was higher in the case of salinity compared to that of stripe rust. The increase in MDA content with the onset of stripe rust stress was up to 43% in the mutants compared to 41% wild‐type plants. Similarly, the mutants showed an increase of MDA up to 66% under salinity stress compared to a 61% increase in normal plants under salinity stress. This shows that mutants had more ROS, and they had lost their ability to deal with their lipid peroxidation under stress conditions. Although the MDA contents were higher, there was not much difference in MDA contents of nrx1‐bd and nrx1‐b type plants under stripe rust.

The activity of key antioxidant enzymes followed a similar trend. Superoxide dismutase (SOD) activity in BW208 remained consistently high under both stress conditions, effectively dismutating superoxide radicals. However, SOD activity declined in nrx1‐bd and nrx1‐b, suggesting a reduced capacity to counteract oxidative stress at the primary level. The increase in SOD activity under stripe rust stress was up to 12% in the mutants when compared to 17% in BW208 plants. Similarly, the mutants showed an increase of almost 18% under salinity stress compared to a 25% increase in normal plants under salinity stress.

Similarly, an increase in peroxidase (POD) and catalase (CAT) and ascorbate peroxidase (APX) activities was significantly lower in the mutant lines compared to wild type plants. The increase in POD activity under stripe rust stress was up to 23% in the mutants when compared to 33% in wild type plants, which is almost 10% less increase. Similarly, the mutants showed an increase of up to 34% under salinity stress compared to a 45% increase in normal plants under salinity stress.

A little less increase in catalase activity under stripe rust stress was observed in the mutants, compared to the increase of 35% in BW208 plants, while APX activity also showed a similar kind of change. Similarly, the mutants showed up to 30% increase in catalase activity under salinity stress compared to a 43% increase in activity in normal plants under salinity stress and APX activity increased up to 40% in mutants compared to a 42% increase in wild‐type plants under salinity stress, reinforcing the overall trend of weakened antioxidant defense in the edited lines. Collectively, these biochemical findings provide strong evidence that NRX1 disruption impairs oxidative stress tolerance, as evidenced by higher MDA accumulation and reduced antioxidant enzyme activity in mutant plants. The wild‐type BW208 line's superior performance under stress conditions further validates NRX1's essential role in redox regulation and stress resilience in wheat. In summary, antioxidant enzyme activities were reduced in NRX1 knockout mutants with the largest effect on catalase while MDA levels were increased (Figure 11).

qPCR expression analysis showed a reduced expression of NRX1 genes in the mutants nrx1‐bd and nrx1‐b, compared to wild‐type plants under unstressed conditions as well as following abiotic salinity stress and biotic stress in the form of stripe rust disease (Figure 12). An increase in expression of NRX1 was observed under both types of stress. The expression was higher under salinity compared to stripe rust infection. Reduced expression of NRX1 genes in the mutants was associated with the increased susceptibility of wheat plants to the effects of both salinity and stripe rust. We observed an increase in expression of NRX1 genes under stress in mutants also, but the increase was lower than the increase in NRX1 expression in normal plants under stress. Expression was more affected in nrx1‐bd than in nrx1‐b.

To identify NRX1 homologs in wheat (T. aestivum), the protein sequence of A. thaliana NRX1 was retrieved from The Arabidopsis Information Resource (TAIR), and the domain sequences (three thioredoxins and one C1‐like domain) were used as a query for a BLASTP search against the wheat genome (IWGSC RefSeq v2.2) through the database Phytozome (https://phytozome‐next.jgi.doe.gov/↗). A stringent e‐value threshold of < 1e−5 was used to ensure specificity. The candidate wheat NRX1 sequences were screened for the presence of thioredoxin‐like domains using the PFAM and SMART databases, confirming domain conservation and structure. Basic physicochemical properties such as isoelectric point (pI), molecular weight, and protein length were determined using the ProtParam tool available on the Expasy server (Swiss Institute of Bioinformatics n.d.).

Multiple sequence alignment of NRX1 protein sequences from wheat and 29 other plant species (S. officinarum, A. gerardi, P. virgatum, T. aestivum, S. viridis, Brachypodium hybridum, Miscanthus sinensis, Paspalum vaginatum, Sorghum bicolor, Oryza sativa, T. intermedium, T. turgidum, Brachypodium distachyon, Brachypodium mexicanum, Eleusine coracana, Gossypium barbadense, Panicum hallii, Pharus latifolius, Brachypodium arbuscula, Brachypodium stacei, Brachypodium sylvaticum, Chasmanthium laxum, Dioscorea alata, H. vulgare, Joinvillea ascendens, Oropetium thomaeum, Setaria italica, Urochloa fusca, A. tauschii, T. urartu) was conducted using MEGA‐X software (Kumar et al. 2018). The phylogenetic relationships were inferred using the Neighbor–Joining (NJ) method with 1000 bootstrap replications to assess the reliability of the clades. Gene structure analysis was performed using the Gene Structure Display Server (GSDS 2.0) (Hu et al. 2015), where exon‐intron arrangements were visualized by comparing genomic DNA and coding DNA sequences (CDS). Conserved motifs among the NRX1 proteins were identified using the MEME Suite (v5.4.1), with parameters set to detect up to 10 motifs ranging from 6 to 50 amino acids in width (Bailey et al. 2009).

The subcellular localization of wheat NRX1 proteins was predicted using the WoLF PSORT web server (Horton et al. 2007) to infer the likely compartmentalzation within the cell, such as cytosol, chloroplast, nucleus, or mitochondria. To understand the regulatory potential of NRX1 genes, HDR genes, and genes for antioxidant enzymes, promoter regions extending 1.5 kb upstream of the start codon were obtained from the wheat genome using the database Phytozome (https://phytozome‐next.jgi.doe.gov/↗). These promoter sequences were submitted to the PlantCARE database (Lescot et al. 2002) for the identification of CREs.

Post‐translational modifications play a crucial role in regulating protein function, stability, and interactions. Phosphorylation sites on serine, threonine, and tyrosine residues were predicted using the NetPhos 3.1 server (DTU Health Tech Bioinformatic Services, n.d.‐a), providing both generic and kinase‐specific predictions. These phosphorylation events are essential for activating signal transduction pathways in response to stress. Glycosylation potential was evaluated using NetNGlyc 1.0 (DTU Health Tech Bioinformatic Services, n.d.‐b), where sequences were analyzed for the presence of N‐linked glycosylation motifs. A threshold score of 0.5 or higher, with a jury agreement of 9/9, was used to confirm high‐confidence glycosylation sites. These PTM predictions helped infer the potential post‐translational regulation of NRX1 proteins during stress responses. Secondary structures were predicted using the SOPMA web server (Geourjon and Deléage 1995). Disordered regions were analyzed using the Mobi DB server (Piovesan et al. 2021).

To explore the functional interactions of NRX1, protein–protein interaction networks were constructed using the STRING database (www.string‐db.org↗). Amino acid sequences of three wheat NRX1 proteins were analyzed via the STRING database, and the network was filtered using an enrichment p value threshold to include the most significant interactors. Special attention was given to proteins co‐expressed and co‐localized with NRX1, such as HMBPP reductase (HDR) and thioredoxins, suggesting links with redox regulation and the methylerythritol phosphate (MEP) pathway. Additionally, in silico expression profiles of NRX1 and major antioxidant genes (e.g., SOD, CAT, and APX) were analyzed using the WheatExp database (Borrill et al. 2016). Genes of different antioxidant enzymes were identified using already published data, specifically for SOD genes (Jiang et al. 2019), catalases (Zhang et al. 2022), and APXs (Tyagi et al. 2020).

To functionally validate the role of NRX1, CRISPR‐Cas9‐mediated genome editing was employed. Guide RNAs (gRNAs) (Table 4) were designed to target conserved exonic regions of NRX1s using the CRISPR‐direct online tool (www.crispr.dbcls.jp↗). Three gRNAs were designed from the conserved regions in NRX1 homologs. Each gRNA was synthesized with overhangs suitable for Golden Gate cloning into a modified Barley Stripe Mosaic Virus (BSMV)‐based vector, pGY036, which contains the gamma part of the BSMV genome. Golden Gate cloning was performed using standard protocols (Marillonnet and Grützner 2020), followed by transformation into Escherichia coli and Agrobacterium tumefaciens for vector propagation and Agro‐infiltration (Yuan et al. 2011). Agrobacteria were also transformed separately with plasmid constructs having the alpha and beta genomes of the BSMV virus. At this stage, we had three kinds of Agrobacterium strains, each containing one of the three parts (alpha, beta, and gamma with gRNA) of the BSMV genome. An equal concentration of all three strains was mixed and infiltrated into the underside of approximately the 5th true leaf of Nicotiana benthamiana plants using a 1‐mL needleless syringe (Hu et al. 2019). For the mechanical inoculation of wheat plants, we harvested A. tumefaciens‐infiltrated leaves of Nicotiana benthamiana after 10 days of infiltration. The leaves were finely ground in 20 mM Na‐phosphate buffer (pH 7.2) containing 1% silicon carbide powder, and the sap was rub‐inoculated onto the plants of "Bobwhite 208" BW208 wheat line already expressing the Cas9 gene for the delivery of gRNA via viral expression (Li et al. 2021; Smedley et al. 2021; Wang et al. 2022). Editing efficiency was evaluated using PCR amplification and Sanger sequencing of target regions from both systemically infected tissues and progeny plants. The Voytas laboratory, University of Minnesota, provided plant materials and vectors for these experiments, and the mutants were developed there (www.voytaslab.umn.edu↗).

To assess the impact of NRX1 gene editing, mutant and control wheat plants were grown under normal and stress conditions induced by salinity and stripe rust infection. For stripe rust infection, the urediniospore mixture was applied to wheat leaves with a paintbrush. Inoculated seedlings were placed under high humidity for 24 h at 10°C in the dark. After 24 h, inoculated leaves were placed under normal conditions with 75% relative humidity. Salinity stress was applied by irrigating plants with a 150 mM solution of NaCl. Morphological parameters, such as root and shoot lengths and fresh biomass, were recorded 10 days after the application of both stresses. Photosynthetic pigments, including chlorophyll a, b, and total chlorophyll, were measured from leaf tissues using the method described by Ronen and Galun (1984). Lipid peroxidation was assessed using the method described by Heath and Packer (1968). Superoxide dismutase (SOD), catalase (CAT), and peroxidase (POD), ascorbate peroxidase (APX) activities were measured as described by Demirezen Yilmaz and Uruç Parlak 2011; Taggar et al. 2012; Alavifard et al. 2012, and Altaf et al. 2022, respectively. The differences in physiological and biochemical responses between wild‐type and mutant lines helped validate the predicted role of NRX1 in modulating redox balance and stress tolerance.

qPCR expression analysis was done by using the conserved primers for all three homeologs of NRX1 under stripe rust stress and salinity stress, 10 days after the application of stress. RNA extraction was performed using the Trizol method. RNA was resuspended in 30 μL of RNase‐free water. The RNA concentration was quantified using a NanoDrop spectrophotometer at 260 nm under RNase‐free conditions, and the RNA was stored at −80°C.

First‐strand cDNA synthesis was done by using the RevertAid First Strand cDNA Synthesis Kit (Thermo Fisher Scientific, Catalog#K1622). Maxima SYBR Green qPCR Master Mix (ThermoFisher Scientific) was used for real‐time expression analysis. The 12.5‐μL Maxima SYBR Green Master Mix (2×), 0.3‐μL forward primer, 0.3‐μL reverse primer, ≤ 500‐ng cDNA template, and nuclease‐free water were mixed to raise the volume to 25 μL. Reaction tubes were inserted into the cycler, and then the program for quantitative expression analysis on Mic qPCR (Bio‐Molecular System) was run. CT values were obtained for expression analysis.

Experiments were conducted in a factorial design to assess the effects of both the stresses and mutations and their interactions on mutant plants compared to wild‐type plants. All experiments were performed with three replications, and data were collected and tabulated in Microsoft Excel. Means and standard deviations were calculated in Microsoft Excel. Statistical analyses were done by two‐way ANOVA followed by Tukey's HSD test at a 95% level of significance, and all pairwise means were compared.