Surface dielectric barrier discharge plasma priming enhances salt-tolerant wheat germination through predicted AMY1.1 redox stabilization Authors: Aeshah A. Awaji, Khaled Lotfy, Essam F. El-Hashash, Salem Mesfir Al-Qahtani, Karima Mohamed El-Absy, Eid Abd-Elfattah Eid, Sayed A. Abo-Marzoka Published: 2026 ABSTRACT Soil salinity severely constrains wheat (Triticum aestivum) germination, the developmental stage most dependent on redox-sensitive starch hydrolysis. Surface dielectric barrier discharge (SDBD) cold atmospheric plasma is a chemical-free seed-priming technology, yet the molecular basis of its transition from beneficial low-dose priming to inhibitory overexposure is unresolved. We combined a systematic SDBD dose-response (ten exposure durations, 0–180 s) under 4000 ppm NaCl with comprehensive agrophysiological profiling — germination, photosynthetic pigments, antioxidant enzymes, osmolytes, gas exchange, membrane stability, mineral nutrition and multivariate analysis — and with in silico molecular docking and protein-flexibility analysis of the high-pI germination α-amylase (AMY1.1) against three plasma-generated redox species (GSH, GSNO, ONOO⁻). Germination and salt tolerance peaked at 140 s (66.67% germination; salt-tolerance index 88%) within a statistically indistinguishable 120–160 s optimum, declining sharply at 180 s. We propose an integrated hormetic model in which moderate doses act through three hierarchical tiers — seed-surface remodelling, antioxidant and osmolyte priming, and predicted AMY1.1 stabilization by GSH and GSNO — whereas excessive doses shift the chemistry toward ONOO⁻-associated distress. The docking analyses generate testable molecular hypotheses rather than proof. SDBD priming at ~ 140 s is a scalable, chemical-free strategy for salt-tolerant wheat establishment.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1038/s41598-026-70789-8. Introduction Global food security is increasingly challenged by changing climatic patterns and the progressive degradation of arable land quality, necessitating sustainable agricultural interventions to stabilize cereal crop productivity1. Wheat (Triticum aestivum L.), the world’s most extensively cultivated cereal crop and the dietary staple of approximately 2.5 billion people, contributing roughly 20% of global caloric and protein intake, is a glycophyte with moderate-to-high salinity sensitivity, making it acutely vulnerable to the dual ionic and osmotic dimensions of NaCl stress2,3. Salinity exerts its phytotoxic effects through three interacting mechanisms: (i) osmotic stress, which reduces soil water potential below the threshold for seed imbibition and radicle emergence; (ii) ionic toxicity, mediated by accumulation of Na⁺ and Cl⁻ in embryonic and seedling tissues to concentrations that displace essential cations (K⁺, Ca²⁺, Mg²⁺) from membrane transport sites and enzymatic cofactor positions; and (iii) secondary oxidative stress, in which perturbation of electron transport and NADPH-oxidase activation drives overproduction of reactive oxygen species (ROS) — superoxide (O₂˙⁻), hydrogen peroxide (H₂O₂) and hydroxyl radical (˙OH) — that damage proteins, membrane lipids and nucleic acids essential for germination metabolic competence4,5. Recent advances in crop physiology emphasize that overcoming salinity-induced seed dormancy requires holistic mitigation strategies that concurrently alleviate physiological drought, reduce ion cytotoxicity, and reactivate embryonic metabolic pathways6,7.The germination stage, governed by mobilization of seed reserves through starch hydrolysis, water uptake and radicle cell-wall loosening, is particularly susceptible to salinity, as these transitions depend on enzyme systems — most critically α-amylase — whose catalytic integrity is exquisitely sensitive to the cellular redox environment8,9. Conventional mitigation strategies — salt-tolerant cultivar development through breeding and marker-assisted selection, soil amendment (gypsum, biochar, organic matter), and chemical seed priming with osmoprotectants, polyamines or hormonal elicitors — have achieved incremental but collectively insufficient progress, constrained by lengthy cultivar timelines, high cost, genotype-environment specificity, and the environmental liability of synthetic inputs10,11. This gap has motivated investigation of physical seed-priming technologies that deliver broad-spectrum stress-tolerance induction without chemical inputs, residual contamination or genotype-specific optimization. Non-thermal (cold) atmospheric plasma has emerged as one of the most promising such technologies, exploiting the compositionally complex mixture of reactive oxygen and nitrogen species (RONS), UV photons, charged particles and electric fields generated in a dielectric barrier discharge to remodel seed-surface physicochemistry and trigger adaptive redox-sensitive signalling in germinating embryos12–14. Meta-analyses document improvements in germination rate, seedling vigour, biomass, chlorophyll content and antioxidant enzyme activity across crop species, competitive with or superior to chemical priming under stress15,16. By regulating antioxidant activities, boosting osmolyte accumulation and reducing oxidative damage, plasma exposure increases seedling growth and stress tolerance17. Among reactor configurations, the Surface Dielectric Barrier Discharge (SDBD) system offers advantages for scale-up: a spatially uniform, large-area plasma sheet at atmospheric pressure and ambient temperature, with no vacuum or noble-gas requirement and precise independent control of exposure, power density and geometry18,19. Despite this phenomenological evidence, the mechanistic basis of plasma priming in wheat under salinity — in particular the transition from beneficial low-dose priming to inhibitory high-dose overexposure — remains poorly resolved. The hormetic principle, in which sub-cytotoxic stressor doses elicit adaptive benefit while supra-threshold doses are inhibitory, is the operative paradigm for RONS-mediated plasma bioactivity in plants20. However, the molecular targets whose integrity is preserved or compromised as a function of plasma dose have not been characterized at the enzyme structural level for wheat germination. A mechanistically critical gap concerns the high-pI α-amylase (AMY1.1; UniProtKB P08117; legacy Swiss-Prot entry AMY3_WHEAT), the principal isoform responsible for starch-granule hydrolysis and soluble-sugar mobilization during early germination21,22. Under the modern unified cereal α-amylase nomenclature, this high-pI, germination-associated isoform corresponds to AMY1.1, and is distinct from the grain-development isoform TaAMY3, whose over-expression delays germination23,24. The structural and catalytic integrity of this isoform under the redox microenvironments generated by different plasma doses is a key, yet unexplored, determinant of dose-dependent germinability. Three plasma-generated RONS are of particular relevance: glutathione (GSH), the primary low-molecular-weight antioxidant buffer that accumulates at sub-saturation power densities; S-nitrosoglutathione (GSNO), a physiological NO reservoir and mediator of reversible S-nitrosylation signalling; and peroxynitrite (ONOO⁻), the highly reactive oxidant formed by the diffusion-limited combination of superoxide and nitric oxide that dominates at excessive doses and mediates irreversible protein nitration25–28. The present study addresses this gap through an integrative, multi-scale investigation combining systematic SDBD dose-response experiments across ten exposure durations (0–180 s) with comprehensive agrophysiological characterization — germination kinetics, photosynthetic pigment dynamics, antioxidant enzyme profiling, osmolyte quantification, gas exchange, cell-membrane stability, mineral nutrition and multivariate PCA — and with in silico molecular docking of GSH, GSNO and ONOO⁻ against the AlphaFold2-predicted AMY1.1 (high-pI α-amylase) structure (UniProtKB P08117), followed by flexibility analysis using CABS-flex 2.0 and ProDy Anisotropic Network Modelling (ANM). We hypothesize that the hormetic optimum at ~ 140 s — the representative peak of a statistically indistinguishable 120–160 s optimal cluster — favours a GSH/GSNO-dominated redox microenvironment that stabilizes the AMY1.1 catalytic architecture and sustains starch-hydrolytic capacity, whereas exposure at ~ 180 s shifts the balance toward ONOO⁻-mediated distress chemistry that destabilizes AMY1.1 and underlies the germination decline. Because molecular docking provides binding hypotheses rather than biological proof, this in silico layer is intended to generate testable, molecular-level hypotheses to guide subsequent evidence-based optimization of plasma priming for wheat under saline field conditions, rather than to establish causation on its own. Results Germination response to SDBD plasma and salinity Salinity reduced mean germination percentage by 22% under 4000 ppm NaCl (72.14 ± 3.05% control vs. 46.19 ± 3.92%; Fig. 1). Main effects of salinity and plasma exposure were individually non-significant for germination, but their interaction was highly significant (P < 0.01; Fig. 1). Plasma at all durations (20–180 s) gave germination percentages superior to the untreated control under both conditions (Fig. 2). A hormetic dose-response was observed, with maxima at 120 s (85.71 ± 0.01% control; 57.14 ± 8.25% salinity) and 140 s (76.19 ± 4.76% control; 66.67 ± 4.76% salinity — the absolute maximum under salinity), declining at 160 and 180 s (Fig. 2). The 140 s exposure was the representative optimum for salinity-stressed wheat — within a statistically indistinguishable 120–160 s cluster (interaction LSD = 30.96) — achieving a 133% improvement over untreated salt-stressed seeds (66.67% vs. 28.57%) and a salt-tolerance index in the “high” class (Sect."Salt tolerance index (STI)"). Key agrophysiological values across representative doses are consolidated in Table 1; full ten-dose datasets appear in Supplementary Tables S1–S6. Table 1Consolidated agrophysiological response of Triticum aestivum cv. Giza 171 to SDBD plasma dose under 4000 ppm NaCl salinity.Parameter0 s(control)120 s140 s(optimum)160 s180 s(distress)Germination & salt toleranceGermination (%)28.5757.14 66.67 57.1447.62Salt-tolerance index, STI (%)5767 88 7556Cell-membrane stability, CMS (%)41.1655.3445.2452.2758.20Photosynthetic pigmentsChlorophyll a (mg g⁻¹ FW)2.203.763.203.173.77Chlorophyll b (mg g⁻¹ FW)1.121.381.391.401.42Total chlorophyll (mg g⁻¹ FW)4.056.055.535.566.22Carotenoids (mg g⁻¹ FW)0.730.910.931.001.03Chl a/Chl b ratio1.972.732.312.272.21Oxidative status & antioxidant enzymesH₂O₂ (µmol g⁻¹ FW)53.1230.9535.2541.3355.90MDA (µmol g⁻¹ FW)0.550.860.790.800.80SOD (U g⁻¹ FW) †15.428.496.968.0612.94CAT (U mg⁻¹ protein)23.1512.599.7512.1817.99APX (µmol min⁻¹ mg⁻¹ protein)9.004.924.636.079.98OsmolyteProline (µmol g⁻¹ FW)1.142.362.192.282.26Chlorophyll fluorescenceFv/Fm0.250.220.150.150.26 Table 1Consolidated agrophysiological response of Triticum aestivum cv. Giza 171 to SDBD plasma dose under 4000 ppm NaCl salinity.Parameter0 s(control)120 s140 s(optimum)160 s180 s(distress)Germination & salt toleranceGermination (%)28.5757.14 66.67 57.1447.62Salt-tolerance index, STI (%)5767 88 7556Cell-membrane stability, CMS (%)41.1655.3445.2452.2758.20Photosynthetic pigmentsChlorophyll a (mg g⁻¹ FW)2.203.763.203.173.77Chlorophyll b (mg g⁻¹ FW)1.121.381.391.401.42Total chlorophyll (mg g⁻¹ FW)4.056.055.535.566.22Carotenoids (mg g⁻¹ FW)0.730.910.931.001.03Chl a/Chl b ratio1.972.732.312.272.21Oxidative status & antioxidant enzymesH₂O₂ (µmol g⁻¹ FW)53.1230.9535.2541.3355.90MDA (µmol g⁻¹ FW)0.550.860.790.800.80SOD (U g⁻¹ FW) †15.428.496.968.0612.94CAT (U mg⁻¹ protein)23.1512.599.7512.1817.99APX (µmol min⁻¹ mg⁻¹ protein)9.004.924.636.079.98OsmolyteProline (µmol g⁻¹ FW)1.142.362.192.282.26Chlorophyll fluorescenceFv/Fm0.250.220.150.150.26 Consolidated agrophysiological response of Triticum aestivum cv. Giza 171 to SDBD plasma dose under 4000 ppm NaCl salinity. Fig. 1 Main effects of SDBD plasma exposure duration (0–180 s) and salinity (control vs. 4000 ppm NaCl) on germination percentage of Triticum aestivum cv. Giza 171. Values represent means ± SD (n = 3). Germination was scored as the proportion of seeds producing a radicle ≥ 2 mm within 7 days of imbibition. Treatment differences were assessed by two-way ANOVA with LSD post-hoc test at P ≤ 0.05 and P ≤ 0.01. Main effects of SDBD plasma exposure duration (0–180 s) and salinity (control vs. 4000 ppm NaCl) on germination percentage of Triticum aestivum cv. Giza 171. Values represent means ± SD (n = 3). Germination was scored as the proportion of seeds producing a radicle ≥ 2 mm within 7 days of imbibition. Treatment differences were assessed by two-way ANOVA with LSD post-hoc test at P ≤ 0.05 and P ≤ 0.01. Fig. 2Interactive effects of SDBD plasma exposure duration (0–180 s) and salinity level (control; 4000 ppm NaCl) on germination percentage of Triticum aestivum cv. Giza 171. Values represent means ± SD (n = 3). LSD values at P ≤ 0.05 = 30.96; at P ≤ 0.01 = 41.43. Bold values indicate the absolute maximum germination percentage achieved under each salinity condition. The interaction was highly significant (P < 0.01), indicating salinity-dependent modulation of the plasma dose-response relationship. Interactive effects of SDBD plasma exposure duration (0–180 s) and salinity level (control; 4000 ppm NaCl) on germination percentage of Triticum aestivum cv. Giza 171. Values represent means ± SD (n = 3). LSD values at P ≤ 0.05 = 30.96; at P ≤ 0.01 = 41.43. Bold values indicate the absolute maximum germination percentage achieved under each salinity condition. The interaction was highly significant (P < 0.01), indicating salinity-dependent modulation of the plasma dose-response relationship. Photosynthetic pigments and chlorophyll fluorescence Both salinity (P < 0.01) and plasma (P < 0.05–0.01) significantly affected pigment contents, with a significant interaction for Chl a, Chl b, CAR, Tchl, the Chl a/Chl b ratio and FM (P < 0.05 for FV; NS for Fv/Fm; Tables S1 and S2). Under salinity, Chl a (3.18 vs. 2.04 mg g⁻¹), Tchl (5.39 vs. 4.68) and the Chl a/Chl b ratio (2.40 vs. 1.93) rose, whereas Chl b, CAR and FV/FM/Fv/Fm fell (Table S1). Across the pooled plasma effect, Chl a and Tchl peaked at 180 s, Chl b/CAR/FM at 160 s, and Fv/Fm (0.45) and FV at 20 s. Under the interaction, plasma raised Chl a and Tchl above the salt control (max at 180 s), while under control conditions 120–160 s raised Chl b and CAR with the Chl a/Chl b ratio falling to 0.39–0.42 (Table S2). Plasma-treated salt-stressed plants showed higher FV and FM than untreated salt-stressed plants (Table S2). Oxidative stress biomarkers and antioxidant defense Salinity was highly significant (P < 0.01) for proline, phenolics, MDA, APX and H₂O₂ (NS for CAT, SOD); plasma was highly significant (P < 0.01) for all; and the interaction was highly significant (P < 0.01) for all (Tables S3 and S4). Under salinity, proline (1.90 vs. 1.16), MDA (0.76 vs. 0.30) and H₂O₂ (57.61 vs. 42.29) rose, while phenolics (23.01 vs. 48.45) and APX (6.60 vs. 7.54) fell (Table S3). Under the pooled plasma effect, proline peaked at 100/180 s (min 20 s); phenolics at 20/100 s (min 120 s); MDA was lowest untreated and plateaued at 100–180 s; APX peaked at 180 s (min 140 s); CAT and SOD were highest untreated and lowest at 140 s; H₂O₂ was highest at 20 s and lowest at 120 s (Table S3). Gas exchange, cell-membrane stability and mineral nutrition Salinity was highly significant (P < 0.01) for Ci, Gs, N, P and K; plasma for Ci, CMS and N; and the interaction (P < 0.01) for Ci and CMS (Tables S5 and S6). Salinity suppressed all parameters: Ci − 60.5% (417.33 vs. 1056.38), Gs − 23.3%, N − 52.7%, P − 57.4% and K − 58.0%; CMS was non-significantly reduced (Table S5). Under the interaction, 180 s plasma under salinity gave the highest CMS, N, P and K, with minima at 60 s (Table S6). Salt tolerance index (STI) STI for germination ranged from 41% to 88% (Fig. 3): 140 s (88%) was the sole very-high-tolerance treatment; 40, 80, 100, 120 and 160 s were high tolerance; and 0, 20 and 180 s were medium tolerance. The STI dose-response was an inverted U, rising from 57% (untreated) to 88% at 140 s and declining to 56% at 180 s (Fig. 3). Fig. 3Salt Tolerance Index (STI) for germination percentage of Triticum aestivum cv. Giza 171 as a function of SDBD plasma exposure duration (0–180 s) under 4000 ppm NaCl salinity stress. STI was calculated as \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:(G{P}_{\mathrm{salinity}}\times\:G{P}_{\mathrm{control}})/{\stackrel{\prime }{GP}}_{\mathrm{control}}^{2}\times\:100$$\end{document}, where GP denotes germination percentage and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\stackrel{\prime }{GP}}_{\mathrm{control}}$$\end{document} is the grand mean under control conditions (n = 3). Horizontal dashed threshold lines (not shown; implied by classification brackets) delineate tolerance categories: very high tolerance (STI ≥ 85%, green); high tolerance (60–84%); and medium tolerance (< 60%).The 140 s treatment (STI = 88%) was the sole very-highly-tolerant classification, while 40, 80, 100, 120, and 160 s achieved high tolerance; untreated (0 s), 20 s, and 180 s were classified as medium tolerant. Values represent means (n = 3). Salt Tolerance Index (STI) for germination percentage of Triticum aestivum cv. Giza 171 as a function of SDBD plasma exposure duration (0–180 s) under 4000 ppm NaCl salinity stress. STI was calculated as \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:(G{P}_{\mathrm{salinity}}\times\:G{P}_{\mathrm{control}})/{\stackrel{\prime }{GP}}_{\mathrm{control}}^{2}\times\:100$$\end{document}, where GP denotes germination percentage and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\stackrel{\prime }{GP}}_{\mathrm{control}}$$\end{document} is the grand mean under control conditions (n = 3). Horizontal dashed threshold lines (not shown; implied by classification brackets) delineate tolerance categories: very high tolerance (STI ≥ 85%, green); high tolerance (60–84%); and medium tolerance (< 60%).The 140 s treatment (STI = 88%) was the sole very-highly-tolerant classification, while 40, 80, 100, 120, and 160 s achieved high tolerance; untreated (0 s), 20 s, and 180 s were classified as medium tolerant. Values represent means (n = 3). Molecular docking of plasma-derived redox ligands GSH and GSNO achieved the most favourable AutoDock Vina scores against AMY1.1 (best − 5.4 kcal mol⁻¹ each), whereas ONOO⁻ scored − 3.7 (Table 2). LigPlot+ analysis (Table 2; Fig. 4): GSH formed eight hydrogen bonds across Glu74, Glu228, Asp203 and Asp296 and seven hydrophobic contacts; GSNO formed ten hydrogen bonds and additionally contacted Gln301 and Leu303 in the ~ 295–310 C-terminal loop; ONOO⁻ contacted only Glu228, Asn294 and His295 at the active-site rim, without engaging the nucleophile Asp203. The active-site machinery is the GH13 catalytic triad (nucleophile Asp203, acid/base Glu228, transition-state-stabilizing Asp296) together with the substrate-binding residue Glu74. Because AutoDock Vina scores are not size-normalized, the GSH/GSNO-over-ONOO⁻ ranking is partly a molecular-size effect rather than redox-specific affinity (developed in the Discussion); the contact maps are reported as hypothesis-generating. Table 2AutoDock Vina binding energies and LigPlot+ v2.2-derived molecular interaction profiles for plasma-derived redox ligands docked against Triticum aestivum α-amylase 3 (AMY1.1).LigandBest Score (kcal·mol⁻¹)H-bond ResiduesBond Lengths (Å)Hydrophobic ContactsGSH −5.4 Glu74, Asp203, Glu228, Asp2962.95–3.09Ser72, Pro73, Tyr76, Leu77, Trp34, Phe167, Phe204GSNO −5.4 Glu74, Asp203, Glu228, Asp296, Ser722.82–3.14Tyr76, Trp34, Phe204, Gln301, Leu303ONOO ⁻ −3.7 Glu228, Asn294, His295—Trp63, Phe251, Phe253 Table 2AutoDock Vina binding energies and LigPlot+ v2.2-derived molecular interaction profiles for plasma-derived redox ligands docked against Triticum aestivum α-amylase 3 (AMY1.1).LigandBest Score (kcal·mol⁻¹)H-bond ResiduesBond Lengths (Å)Hydrophobic ContactsGSH −5.4 Glu74, Asp203, Glu228, Asp2962.95–3.09Ser72, Pro73, Tyr76, Leu77, Trp34, Phe167, Phe204GSNO −5.4 Glu74, Asp203, Glu228, Asp296, Ser722.82–3.14Tyr76, Trp34, Phe204, Gln301, Leu303ONOO ⁻ −3.7 Glu228, Asn294, His295—Trp63, Phe251, Phe253 AutoDock Vina binding energies and LigPlot+ v2.2-derived molecular interaction profiles for plasma-derived redox ligands docked against Triticum aestivum α-amylase 3 (AMY1.1). Fig. 4LigPlot+ v2.2 two-dimensional interaction diagrams for plasma-derived redox ligands docked against the active site of Triticum aestivum α-amylase 3 (AMY1.1); AutoDock Vina v1.2.3). (A) AMY1.1–GSH complex (B) AMY1.1–GSNO complex (C) AMY1.1–ONOO⁻ complex. The 1.7 kcal mol⁻¹ binding energy deficit of ONOO⁻ relative to GSH/GSNO, and the associated shift from deep catalytic-core engagement to peripheral rim contacts, constitutes the molecular basis for plasma dose-dependent AMY1.1 functional impairment at supra-optimal SDBD exposure times. Residue colors: blue = hydrogen-bond acceptors; orange/brown = hydrophobic contacts; red dashed lines = non-bonded contacts; bond distances in Å. LigPlot+ v2.2 two-dimensional interaction diagrams for plasma-derived redox ligands docked against the active site of Triticum aestivum α-amylase 3 (AMY1.1); AutoDock Vina v1.2.3). (A) AMY1.1–GSH complex (B) AMY1.1–GSNO complex (C) AMY1.1–ONOO⁻ complex. The 1.7 kcal mol⁻¹ binding energy deficit of ONOO⁻ relative to GSH/GSNO, and the associated shift from deep catalytic-core engagement to peripheral rim contacts, constitutes the molecular basis for plasma dose-dependent AMY1.1 functional impairment at supra-optimal SDBD exposure times. Residue colors: blue = hydrogen-bond acceptors; orange/brown = hydrophobic contacts; red dashed lines = non-bonded contacts; bond distances in Å. Backbone flexibility (CABS-flex 2.0/ProDy ANM) Three dynamical phenotypes were resolved for AMY1.1 (Figure S1; Table S7): GSH produced broad rigidification (mean ΔRMSF − 0.126 Å; ΔSqFluct − 0.405 Ų); GSNO produced a net-flexibilizing, region-selective profile (mean ΔRMSF + 0.054 Å) with a pronounced + 4.02 Ų excursion at Node 2 (~ 295–310, peak ~ 303); and ONOO⁻ produced a near-neutral global profile (mean ΔRMSF − 0.069 Å) with a localized spike exceeding 5.0 Å at residue ~ 298. Node 1 (residues 150–170) corresponds to the substrate-binding cleft (Table S7). Principal component analysis PCA of 22 variables across 22 treatment combinations yielded ten PCs, of which five had eigenvalue > 1, together explaining 92.33% of variance (Table S8). PC1 (38.19%) and PC2 (22.13%) captured 60.32% jointly. PC1 separated a mineral-nutrition/photosynthetic cluster (N, P, K, Phe, Ci, FM, FV, Fv/Fm, CMS; positive) from a stress-biomarker cluster (proline, MDA, Chl a, H₂O₂, CAT, SOD; negative); PC2 resolved a germination–pigment optimization axis (Fig. 5). PCA was performed in SPSS v26. Fig. 5Principal Component Analysis (PCA) biplot of PC1 (38.19% of total variance) versus PC2 (22.13% of total variance) for germination percentage and 21 chemical composition variables measured in Triticum aestivum cv. Giza 171 across SDBD plasma exposure treatments (0–180 s) under two salinity conditions (control; 4000 ppm NaCl). Blue vectors: variable loadings (arrows); direction and magnitude indicate correlation strength and orientation with respect to principal axes. Red points: treatment score projections; green points: salinity condition scores. Treatment clusters: Quadrants I/IV (positive PC1) — control-condition treatments with high photosynthetic-nutritional scores (untreated, 20 s, 40 s); Quadrants II/III (positive PC2) — moderate-dose plasma treatments with high germination-pigment scores (120 s, 140 s, 160 s under control). Variables abbreviations: G% = germination percentage; Chl a/b = chlorophyll a/b; CAR = carotenoids; Tchl = total chlorophyll; Chl a/Chl b = chlorophyll a: b ratio; FM/FV/FV·FM⁻¹ = chlorophyll fluorescence parameters; CMS = cell membrane stability; Ci = intercellular CO₂; Gs = stomatal conductance; N/P/K = leaf macronutrients; Phe = total phenolics; Pro = proline; MDA = malondialdehyde; H₂O₂ = hydrogen peroxide; APX/CAT/SOD = antioxidant enzyme activities. PCA performed in SPSS v26 (IBM) using correlation matrix standardization; eigenvalues and cumulative variance for all 10 PCs are detailed in Table S8. Principal Component Analysis (PCA) biplot of PC1 (38.19% of total variance) versus PC2 (22.13% of total variance) for germination percentage and 21 chemical composition variables measured in Triticum aestivum cv. Giza 171 across SDBD plasma exposure treatments (0–180 s) under two salinity conditions (control; 4000 ppm NaCl). Blue vectors: variable loadings (arrows); direction and magnitude indicate correlation strength and orientation with respect to principal axes. Red points: treatment score projections; green points: salinity condition scores. Treatment clusters: Quadrants I/IV (positive PC1) — control-condition treatments with high photosynthetic-nutritional scores (untreated, 20 s, 40 s); Quadrants II/III (positive PC2) — moderate-dose plasma treatments with high germination-pigment scores (120 s, 140 s, 160 s under control). Variables abbreviations: G% = germination percentage; Chl a/b = chlorophyll a/b; CAR = carotenoids; Tchl = total chlorophyll; Chl a/Chl b = chlorophyll a: b ratio; FM/FV/FV·FM⁻¹ = chlorophyll fluorescence parameters; CMS = cell membrane stability; Ci = intercellular CO₂; Gs = stomatal conductance; N/P/K = leaf macronutrients; Phe = total phenolics; Pro = proline; MDA = malondialdehyde; H₂O₂ = hydrogen peroxide; APX/CAT/SOD = antioxidant enzyme activities. PCA performed in SPSS v26 (IBM) using correlation matrix standardization; eigenvalues and cumulative variance for all 10 PCs are detailed in Table S8. Discussion An integrated multi-tier hormetic framework Synthesizing the experimental physiological dataset — germination kinetics, photosynthetic apparatus integrity, antioxidant enzyme dynamics, mineral nutrition, gas exchange, cell-membrane stability and multivariate PCA — with the in silico findings from AutoDock Vina docking and CABS-flex 2.0/ProDy ANM profiling, we advance a unified multi-tier hormetic model for SDBD plasma action in salinity-challenged T. aestivum cv. Giza 171 that reconciles the germination dose-response, antioxidant kinetics, photosynthetic pigment dynamics, gas exchange and AMY1.1 conformational behaviour within a single framework (Fig. 6). The hormetic principle — low-to-moderate stressor doses elicit beneficial adaptation while supra-threshold doses are inhibitory — is the operative paradigm for RONS-mediated plasma bioactivity in plants20,29. The model has two temporal phases — priming eustress (~ 120–140 s) and distress (~ 180 s) — separated by a critical RONS chemistry transition, accommodating three hierarchical tiers that operate synergistically during eustress but are dismantled during distress. A fourth dimension, developmental stage-specific dose dissociation, resolves the discordance between the germination and pigment optima. Fig. 6Integrated hormetic model for SDBD plasma action in salinity-stressed Triticum aestivum cv. Giza 171 (4000 ppm NaCl). (A) Hormetic dose–response: eustress optimum at 140 s (STI = 88%) and distress zone at 180 s (STI = 56%), separated by a GSH/GSNO→ONOO⁻ chemistry transition. (B) Dose-dependent RONS speciation. (C) Three tiers — seed-surface remodelling (seconds), antioxidant/osmolyte priming (hours), AMY1.1 stabilization (days). (D) Developmental stage–dose dissociation (embryonic AMY1.1 at 140 s; chloroplast biogenesis at 160–180 s). Integrated hormetic model for SDBD plasma action in salinity-stressed Triticum aestivum cv. Giza 171 (4000 ppm NaCl). (A) Hormetic dose–response: eustress optimum at 140 s (STI = 88%) and distress zone at 180 s (STI = 56%), separated by a GSH/GSNO→ONOO⁻ chemistry transition. (B) Dose-dependent RONS speciation. (C) Three tiers — seed-surface remodelling (seconds), antioxidant/osmolyte priming (hours), AMY1.1 stabilization (days). (D) Developmental stage–dose dissociation (embryonic AMY1.1 at 140 s; chloroplast biogenesis at 160–180 s). The priming eustress phase (~ 120–140 s) At moderate exposure, the discharge generates a controlled, sub-cytotoxic RONS flux balanced between short-lived surface oxidants (O₃, ¹O₂, ˙OH) and stable diffusible species (H₂O₂, NO₂⁻, NO₃⁻, GSH, GSNO) penetrating embryonic tissues17,30. Three tiers collectively produce the germination phenotype of Sect. "Germination response to SDBD plasma and salinity" and "Salt tolerance index (STI)". Tier 1 — Seed-surface physicochemical remodelling. Moderate exposure remodels the seed coat through etching and oxidative functionalization: new polar groups (-OH, -COOH, -C = O) reduce the water contact angle, raise hydrophilicity and porosity, and lower diffusional resistance to water uptake31,32. Under the 4000 ppm NaCl gradient, this counteracts a primary mechanism of salt-induced germination inhibition9,19. The CMS data (84.85% at 180 s control; 58.20% at 180 s salinity; Sect. "Gas exchange, cell-membrane stability and mineral nutrition") are consistent with membrane lipid remodelling that increases bilayer rigidity and electrolyte retention33. Tier 2 — Preemptive antioxidant and osmolyte priming. Moderate RONS at 120–140 s activate antioxidant networks (SOD/CAT/APX; Sect. "Oxidative stress biomarkers and antioxidant defense") and stimulate osmolyte biosynthesis, particularly proline, before full ionic-stress responses34–37. This preconditioning — analogous to H₂O₂/NO hormetic stress memory via WRKY/MYB/ERF4,33,37— explains the lower H₂O₂ and MDA under salinity at these doses34,38. Proline, positively correlated with germination by PCA (Sect. "Principal component analysis"), serves as both osmotic solute and chaperone34,39. Suppression of the Chl a/Chl b ratio at 140 s extends Tier 2 protection to photochemical machinery38. Tier 3 — Molecular stabilization of AMY1.1. At the molecular level, moderate doses favour GSH and GSNO over ONOO⁻, consistent with the kinetic hierarchy of RONS generation5,26,28. The favourable, equivalent docking scores of GSH (− 5.4 kcal mol⁻¹) and GSNO (− 5.4) against AMY1.1, with complementary dynamical consequences (GSH global dampening, mean ΔRMSF − 0.126 Å; GSNO Node-2 ΔSqFluct + 4.02 Ų), are consistent with a catalytically competent conformation5,9,40. As in Sect. "Molecular docking of plasma-derived redox ligands", however, the GSH/GSNO-over-ONOO⁻ preference is partly attributable to ligand molecular size rather than redox-specific affinity; the Tier-3 mechanism is therefore advanced as a structurally motivated hypothesis. Sustained AMY1.1 activity would secure soluble-sugar supply to the embryo, addressing the starch-hydrolysis bottleneck that limits germinability8,21,22. GSNO’s remodelling of the ~ 295–310 regulatory loop may facilitate catalytic loop-gating via reversible S-nitrosylation28,41,42. These tiers are temporally and spatially hierarchical: Tier 1 within seconds; Tier 2 over hours; Tier 3 AMY1.1 stabilization over days — a coordinated cascade whose output is the salt-tolerant germination phenotype (STI = 88%) at 140 s8,9,39, confirmed by the unique PC2 clustering of this treatment (Sect. "Principal component analysis")4,20. The distress phase (~ 180 s) and the critical RONS chemistry transition At ~ 180 s, RONS accumulation shifts from a priming signal to an oxidative distress stressor, driven by GSH/GSNO depletion and ONOO⁻ dominance25–27. At the seed-coat level, etching exceeds the pericarp integrity threshold, consistent with the germination decline from 66.67% (140 s) to 47.62% (180 s) and the STI reduction from 88% to 56%43. At the biochemical level, 180 s under salinity produced the highest distress markers (H₂O₂ 55.90; MDA 0.80; APX/CAT/SOD hyperinduction; Table S6); simultaneous elevation of lipid-peroxidation markers and antioxidant activities is the signature of damage-reactive compensation rather than preconditioning3,4,36. At the molecular level, ONOO⁻ dominance is mirrored in the in silico profile — weaker AMY1.1 docking score (− 3.7), peripheral rim-only contacts, absence of protective ΔSqFluct, and a localized destabilization spike at residue ~ 298 consistent with His295 nitration25,27,44,45; these are hypotheses consistent with — not proof of — the proposed mechanism (Sect. "Limitations and future directions"). The proposed consequence, impaired starch hydrolysis and restricted sugar supply, is a candidate molecular correlate of the germination decline8,9,21. Developmental stage–dose dissociation The dissociation between the germination optimum (140 s) and the pigment optimum (160–180 s) is coherent under developmental stage-specific RONS threshold architecture — optimal dose is set by the rate-limiting target in each tissue at each stage29–31,46. Germination is governed by AMY1.1-dependent starch hydrolysis, optimally primed at 140 s and damaged at 160–180 s8,9,42. Pigment accumulation depends on chloroplast biogenesis machinery with higher RONS activation thresholds, so higher doses enhance pigments while germination declines33,38,47,48. This has direct agronomic implications: no single exposure optimizes all endpoints, so treatment duration must follow the primary objective30,46,49. The model provides the mechanistic foundation for such optimization — constituting, to our knowledge, among the first comprehensive multi-tier models for SDBD plasma action in salinity-stressed wheat integrating experimental agrophysiology with molecular docking and protein-dynamics evidence. Limitations and future directions The computational analysis uses an AlphaFold v2.0 ab initio prediction (AF-P08117-F1) rather than an experimental structure; while AlphaFold approaches experimental accuracy for folded domains61, docking in loop regions (pLDDT < 70) warrants caution. A central limitation concerns the in silico molecular tier (Sect. "Molecular docking of plasma-derived redox ligands"–"Backbone flexibility (CABS-flex 2.0 / ProDy ANM)"): molecular docking identifies candidate binding geometries and contact residues but does not measure binding affinity or establish biological causation, and AutoDock Vina scores are not size-normalized — so the GSH/GSNO-over-ONOO⁻ ranking is partly a molecular-size artifact rather than evidence of redox-specific selectivity. No in vitro or in vivo assay in this study measured AMY1.1 activity or stability in the presence of GSH, GSNO or ONOO⁻; the Tier-3 mechanism is presented as a testable hypothesis, not a demonstrated result, and the experimentally grounded Tier-1 and Tier-2 findings stand independently of it. Several quantitative parameters in the physiological dataset show internal unit or value inconsistencies (Sect. "Germination response to SDBD plasma and salinity"–"Gas exchange, cell-membrane stability and mineral nutrition") being reconciled against primary records; the qualitative dose-response patterns and the tier structure of the model are unaffected. Future studies should extend this approach to: (i) multiple cultivars and salinity levels; (ii) field-scale validation of germination and yield; (iii) direct quantification of intracellular RONS at each dose; (iv) site-directed mutagenesis of the catalytic triad (Asp203/Glu228/Asp296, plus substrate-binding Glu74) with in vitro AMY1.1 activity assays under GSH/GSNO/ONOO⁻ — the direct biological test of the Tier-3 hypothesis; and (v) atomistic molecular-dynamics simulations to validate the CABS-flex findings and characterize allosteric communication. Methods Plant material Giza 171 (Triticum aestivum L.; bread wheat), moderately tolerant to soil salinity50, was used. Seeds were surface-sterilized (1% NaOCl, 5 min), rinsed three times in sterile distilled water, and air-dried. SDBD plasma system and seed treatment Treatment used a Surface Dielectric Barrier Discharge device of four plasma-generating plates (9 × 4 cm² alumina dielectric, 1 mm; hexagonal aluminium electrodes, 7 mm circumscribed diameter) operated under ambient conditions18,29. An AC HV supply operated at 8–30 W, ~ 10 kHz and 5–10 kVpp32,51. Seeds were arranged in a non-overlapping monolayer on the plate (20 seeds/plate, 80/experimental unit)31,49 and exposed for 0, 20, 40, 60, 80, 100, 120, 140, 160–180 s; treated seeds were transferred immediately to germination assays to minimize decay of short-lived species18. Each condition was replicated three times (n = 3). Salinity stress and germination assay A two-factor factorial design crossed two salinity levels (0; 4000 ppm NaCl ≈ 6.25 dS m⁻¹, a moderate-to-high salinity level selected from preliminary range-finding trials) with the ten exposure durations. Twenty-one seeds per dish were placed on Whatman No. 1 paper moistened with 5 mL distilled water or 4000 ppm NaCl, sealed, and incubated at 25 ± 1 °C under 16/8 h light/dark; germination (radicle ≥ 2 mm) was scored daily for seven days. Germination percentage (GP) and the salt-tolerance index (STI = GPstress/GPcontrol × 100) followed Ali et al.52, with tolerance classified as > 80 (very high), 60–80 (high), 40–60 (medium), 20–40 (low) and < 20% (very low)52. Photosynthetic pigments and fluorescence Pigments were extracted from 0.2 g leaf tissue in 80% acetone, centrifuged (5,000 × g, 10 min, 4 °C), and read at 663, 645 and 470 nm; Chl a, Chl b, Tchl and carotenoids were computed per Lichtenthaler53 and expressed as mg g⁻¹ FW. Chlorophyll fluorescence (FV, FM, Fv/Fm) was measured on dark-adapted (20 min) leaves with a portable PAM fluorometer54. Oxidative stress biomarkers Hydrogen peroxide and lipid peroxidation H₂O₂ was determined by the titanium-sulphate method55 at 415 nm and expressed as µmol g⁻¹ FW. Lipid peroxidation (MDA) was measured by the TBARS assay56 at 532 nm (corrected at 600 nm) and expressed as µmol g⁻¹ FW† using ε = 155 mM⁻¹ cm⁻¹. Compatible solutes and phenolics Free proline was determined by the acid-ninhydrin method57 at 520 nm and expressed as µmol g⁻¹ FW†. Total phenolics were quantified by Folin–Ciocalteu58 and expressed as mg gallic-acid equivalents g⁻¹. Antioxidant enzyme assays Tissue (0.5 g) was homogenized in ice-cold 50 mM potassium phosphate (pH 7.0) with 1 mM EDTA and 1% PVP, centrifuged (12,000 × g, 20 min, 4 °C); protein was determined by Bradford59. APX60 was followed at 290 nm (ε = 2.8 mM⁻¹ cm⁻¹); CAT61 at 240 nm (ε = 39.4 mM⁻¹ cm⁻¹); and SOD62 by 50% inhibition of NBT photoreduction at 560 nm. Computational methods Protein structure The structure of Triticum aestivum high-pI α-amylase (AMY1.1; UniProtKB P08117; legacy Swiss-Prot entry AMY3_WHEAT) was retrieved from the AlphaFold database (AF-P08117-F1) (https://alphafold.ebi.ac.uk)63,64. Under the modern unified cereal α-amylase nomenclature, the high-pI germination isoform encoded by P08117 is designated AMY1.1 and must not be conflated with the grain-development isoform TaAMY3 (over-expression of which delays germination)20,21. As no experimental structure is deposited in the PDB65, the AlphaFold model (pLDDT ≥ 70) was used; the structure was prepared in PyMOL v2.5 (https://pymol.org) with polar hydrogens and Gasteiger charges assigned in AutoDockTools version 1.5.7 (https://autodock.scripps.edu)66. The signal peptide (residues 1–23) was excluded. Ligand preparation GSH (CID 124886), GSNO (CID 119032) and ONOO⁻ (CID 25200) were retrieved from PubChem67 and energy-minimized (MMFF94) in Open Babel v3.1.1(http://openbabel.org)68. Molecular docking Docking used AutoDock Vina v1.2 (https://vina.scripps.edu)69,70 with a 25 × 25 × 25 ų grid centered on the active site (the conserved GH13 catalytic triad: nucleophile Asp, acid/base Glu, transition-state-stabilizing Asp). Exhaustiveness was 8 with 10 independent runs per ligand; poses were ranked by ΔG and analyzed in LigPlot+ v2.2 (https://www.ebi.ac.uk/thornton-srv/software/LigPlotPlus/)71. Protein flexibility Backbone flexibility was assessed with CABS-flex 3.0 (https://biocomp.chem.uw.edu.pl/CABSflex3/)72 for apo-AMY1.1 and each holo-complex (Chain A, residues 11–403); ΔRMSF = RMSFHolo − RMSFApo. Atomistic ΔSqFluct was computed with the ANM in ProDy v2.5 (http://prody.csb.pitt.edu)73; critical nodes used |ΔSqFluct| > 0.5 Ų. Statistical analysis A completely randomized two-factor factorial design was used; measurements were performed in triplicate (n = 3) and expressed as mean ± SD. Normality was confirmed by the Kolmogorov–Smirnov test. Two-way ANOVA was performed in SPSS Statistics v26.0 (IBM Corp. https://www.ibm.com/products/spss-statistics), with means compared by the LSD test at P ≤ 0.05 and P ≤ 0.0174. Principal component analysis was performed in SPSS Statistics v26. Conclusions This study establishes an integrated, multi-tier hormetic model for SDBD cold atmospheric plasma priming in salinity-stressed bread wheat (Triticum aestivum cv. Giza 171). Exposure for ~ 140 s represents the physiological eustress optimum under 4000 ppm NaCl, maximizing germination capacity (66.67%) and salt tolerance (STI = 88%), whereas prolonged exposure (180 s) triggers oxidative distress. Mechanistically, optimal priming synergizes seed-surface etching with preemptive antioxidant/osmolyte activation and predicted in silico AMY1.1 active-site stabilization by GSH and GSNO against peroxynitrite distress. Furthermore, a developmental stage-dose dissociation separates the embryonic germination optimum (140 s) from the subsequent foliar pigment peak (160–180 s). SDBD plasma priming at ~ 140 s provides an effective, chemical-free pre-sowing intervention for wheat cultivation under saline environments, presenting clearly delineated molecular hypotheses for future vitro enzymatic verification. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1