Untargeted metabolomics reveals key metabolites and genes underlying salinity tolerance mechanisms in maize

Abstract

Understanding the physiological, metabolic, and genetic mechanisms underlying salt tolerance is essential for improving crop resilience and productivity, yet their complex interactions remain poorly defined. We compared physiological and metabolic responses to salinity between two contrasting maize (Zea mays L.) inbred lines: the salt‐sensitive C68 and the salt‐tolerant NC326. The sensitivity of C68 was characterized by reduced shoot and root dry weights and plant height, high tissue accumulation of Na and Cl but low K, and lower leaf proline accumulation compared to the salt‐tolerant NC326. Untargeted metabolomics identified 56 metabolites categorized as constitutively upregulated or salt‐responsive. In NC326, constitutive accumulation of flavonoids, including schaftoside, tricin, and kaempferol‐related compounds in leaves, suggests adaptive priming against oxidative stress, while constitutively higher lipids and fatty acids in roots may enhance membrane stability. Salt‐responsive metabolites, notably antioxidants and lanosterol, highlighted inducible oxidative‐stress mitigation and membrane‐stabilization strategies. By integrating metabolomic and genetic analyses, we identified 10 candidate genes involved in the biosynthesis of key metabolites. These findings establish a comprehensive platform for functional validation of metabolites and candidate genes for developing maize varieties with improved resilience to soil salinity through targeted breeding or biotechnological strategies.

Core Ideas

• Maize genotypes differ in growth, ion balance, and proline under salt stress.

• Metabolomics reveals pre‐stress buildup of protective flavonoids and fatty acids.

• Salt triggers metabolic changes like sterol increase to protect membranes.

• Genetic analysis links genes to key metabolites controlling salt tolerance.

• Metabolite‐gene interactions may advance breeding of maize for salt resilience.

Plain Language Summary

Salinity, or high salt content in soil, is a major challenge to crop growth worldwide, reducing food production. Maize, an important crop globally, struggles to grow under salty conditions. This study compared two maize types—one that grows well in salty soil and one that struggles—to understand how maize plants adapt to salt stress. We measured hundreds of compounds (metabolites) in plant tissues to identify protective substances. We discovered that plants resistant to salt stress naturally produce higher amounts of protective substances, helping them avoid damage from salt. Additionally, we identified salt‐induced metabolites that likely contribute to cellular protection by stabilizing membranes and mitigating osmotic and oxidative stress. We also identified genes that control the production of these important metabolites. This research provides new insights into how maize plants manage salt stress and highlights potential targets for developing salt‐resilient crops.

Abbreviations

EC

electrical conductivity

ICP‐OES

inductively coupled plasma optical emission spectrometry

LC‐MS

liquid chromatography‐mass spectrometry

LPC

lysophosphatidylcholine

mQTL

metabolite quantitative trait locus

PCA

principal component analysis

QC

quality control

ROS

reactive oxygen species

UPLC

ultra performance liquid chromatography

1. INTRODUCTION

Salinity is a major abiotic stress that significantly reduces crop yields, affecting approximately 800 million ha (6%) of global agricultural land and contributing to annual agricultural losses estimated between $12 and $27.3 billion (Butcher et al., 2016; Munns & Tester, 2008; Qadir et al., 2014; X. Zhou et al., 2022). The impact of salinity on crops, water, and agricultural soils is expected to worsen due to climate change, unsustainable irrigation practices, soil degradation, and the expansion of irrigated farming. Rising temperatures and increased drought conditions accelerate evaporation, resulting in higher salt concentrations in the soils, while sea‐level rise causes saltwater intrusion into coastal aquifers used to irrigate farmland. Additionally, poor irrigation management, excessive fertilizer use, inadequate drainage, urbanization, and deforestation contribute to soil salinization by altering natural water flows and exposing underground salts, further reducing agricultural productivity (Tarolli et al., 2024). Saline soil typically has a higher accumulation of sodium, potassium, calcium, magnesium, and chloride ions. While salinization and sodification affect all climatic regions, arid and semi‐arid climate zones are particularly affected due to limited natural leaching (Gheyi et al., 2023; Rengasamy, 2006). Salt stress disrupts essential physiological processes, including photosynthesis, respiration, and water uptake, leading to significant losses in crop productivity (Ferreira et al., 2024). Given the increasing severity of soil salinity and its impact on agriculture, improving salt tolerance in crop plants is crucial for developing climate‐resilient varieties and mitigating yield losses.

Maize (Zea mays L.), one of the most widely grown cereals in the world, is a multi‐purpose crop that provides food, livestock feed, and various industrial products, including biofuels, alcohol, and vegetable oil. Abiotic stresses, including drought and salinity, significantly reduce maize productivity and threaten global food security, feed availability, and industrial supply chains (Farooq et al., 2015; Nepolean et al., 2018; Sandhu et al., 2020). High soil salinity disrupts essential physiological processes as excess sodium interferes with potassium uptake, impairs stomatal regulation, and increases evapotranspiration, ultimately leading to leaf necrosis and stunted growth (Fortmeier & Schubert, 1995; Mass & Hoffman, 1977; Menezes‐Benavente et al., 2004). Moreover, salt stress results in higher production of reactive oxygen species (ROS) that cause oxidative damage to the cells (de Azevedo Neto et al., 2006; Hasanuzzaman et al., 2021; Methela et al., 2024). Given the expanding extent of saline soils, there is an urgent need to develop maize cultivars capable of withstanding salt stress. Yield losses in maize under salinity are often severe, particularly during early developmental stages, making it critical to dissect the physiological and genetic mechanisms that confer tolerance. Breeding and genetic engineering approaches, including marker‐assisted breeding and transgenic modifications, offer promising strategies for developing salt‐tolerant maize genotypes (M. Zhang, Li, et al., 2023; X. Zhou et al., 2022).

Metabolites represent the functional readout of cellular biochemistry and, therefore, underlie external plant phenotypes. Plants exposed to salinity undergo dramatic metabolic shifts to maintain basic metabolism and adapt to the stressful environment (Arbona et al., 2013; Bundy et al., 2009). The accumulation of certain primary and secondary metabolites enhances salt tolerance by alleviating oxidative stress and regulating osmotic balance. Since the metabolome is closely linked to the phenotype, characterizing global metabolic changes is crucial for identifying novel metabolites and mechanisms underlying salt tolerance. However, there are critical knowledge gaps in the salinity‐metabolome nexus, including variation in metabolic profiles across different tissues in maize and the identity of metabolites associated with the salinity response and their genetic regulation. Recent metabolomic studies have provided insights into metabolic variation and the underlying molecular responses to salinity (Han et al., 2023; P.‐C. Li et al., 2021; Liang et al., 2021; Richter et al., 2015; Tareq et al., 2024; D. Wang et al., 2023; Widodo et al., 2009; Wu et al., 2024). However, a comprehensive understanding of metabolic changes across different tissues during salt stress and their contributions to salt tolerance in maize remains largely unexplored.

Our recent study on the characterization of natural variation for salt tolerance in maize germplasm revealed tremendous variation and identified salt‐tolerant and salt‐sensitive inbred lines (Sandhu et al., 2020). The current study investigates the physiological, ionic, and metabolic responses of NC326 and C68 inbred lines to salt stress to uncover the mechanisms underlying salt tolerance. Through a time‐course analysis of leaf and root metabolome, this study provides mechanistic insights into tissue‐specific metabolic perturbations and identifies putative pathways and genes underlying salt tolerance. These findings enhance our understanding of the biological processes governing salt tolerance and provide a platform for development of salt tolerance maize and related grasses.

2. MATERIALS AND METHODS

2.1. Plant materials and growth conditions

The maize inbred lines used in this study, C68 (salt‐sensitive) and NC326 (salt‐tolerant), were selected based on the phenotypic evaluation of a maize diversity panel for salinity response (Sandhu et al., 2020). These inbred lines were evaluated under controlled greenhouse conditions using lysimeters of 120 cm in length, 60 cm in width, and 50 cm in depth at the US Salinity Laboratory, Riverside, CA. A randomized complete block design was employed, with 12 seeds sown per inbred line and subsequently thinned to nine plants each. Three biological replicates were used in the experiment. Plants were grown under natural light conditions from March 15 to April 26, 2023. The greenhouse was maintained at a day/night temperature of 27°C/16°C with a relative humidity of approximately 65%. Initially, plants were cultivated in half‐strength Hoagland's solution with an electrical conductivity (EC_iw) of 1.46 dS m⁻¹ for 2 weeks (Table 1). Subsequently, the solution for the plants designated for the salinity treatment was gradually increased to 16 dS m⁻¹ over three days to avoid osmotic shock (Table 1). The moment when the target salinity concentration (16 dS m⁻¹) was reached was designated as 0 h. The pH of the irrigation water was monitored weekly and adjusted to 7.0 using sulfuric acid as needed. The control plants continued to grow in half‐strength Hoagland's solution (EC_iw = 1.46 dS m⁻¹). To simulate field‐relevant salinity conditions, careful attention was given to maintaining ionic balance in the irrigation water. A mixture of salts was used to ensure appropriate ratios among sodium (Na⁺), calcium (Ca²⁺), and magnesium (Mg²⁺), which is essential for preventing nutrient imbalances and preserving membrane stability. Additionally, cations and anions were balanced to maintain electroneutrality and prevent pH shifts or salt precipitation. This strategy ensured that plants experienced realistic osmotic and ionic stress conditions typical of naturally saline environments. Given that plants under salt stress first undergo osmotic stress before acclimating to saline conditions, tissue samples were collected at two additional time points (24 and 48 h after reaching the final salinity level) to effectively capture physiological and metabolic adjustments during acclimation. These samples were immediately frozen in liquid nitrogen and stored at −80°C for metabolic analysis. The remaining plants were grown under their respective treatments for an additional 3 weeks.

TABLE 1.

Composition of irrigation water.

Treatment	ECiw (dS m−1)	Ion concentration (mmolc L−1)
		Cl−	SO4 2−	NO3 −	PO4 3−	HCO3 −	Na+	K+	Ca2+	Mg2+
Control	1.46	1.52	2.02	5.39	1.5	3.07	1.69	6.58	3.2	2.05
Saline	16.0	128.52	27.27	5.39	1.5	3.07	106.69	6.58	29.45	23.05

Abbreviation: EC, electrical conductivity.

Core Ideas

• Maize genotypes differ in growth, ion balance, and proline under salt stress.

• Metabolomics reveals pre‐stress buildup of protective flavonoids and fatty acids.

• Salt triggers metabolic changes like sterol increase to protect membranes.

• Genetic analysis links genes to key metabolites controlling salt tolerance.

• Metabolite‐gene interactions may advance breeding of maize for salt resilience.

2.2. Data collection and sampling

Root and shoot tissue samples, obtained by separating plants at the scutellar node, were dried at 70°C for 96 h to determine the dry root and shoot weight, respectively. Plant height was measured in centimeters (cm) from the tip of the primary root to the tip of the topmost visible leaf in the whorl.

2.3. Mineral analyses

Oven‐dried leaf and root tissues were ground into a fine powder prior to mineral analyses. Chloride concentrations were determined using a colorimetric assay involving mercuric thiocyanate in the presence of ferric nitrate in an AQ300 discrete analyzer. Concentrations of other ions were measured after nitric acid digestion using inductively coupled plasma optical emission spectrometry (ICP‐OES; 3300 DV, Perkin‐Elmer Corp.). ICP‐OES calibration was performed using certified standards. For potassium and sodium, SPEX CertiPrep Assurance Grade standards (10,000 µg/mL in 5% HNO₃) were used, with calibration curves ranging from 0.05 to 100 mg/L (0, 0.5, 1, 10, 50, and 100 ppm). For chloride, a 1000 ppm stock solution was prepared from analytical reagent‐grade NaCl crystals and diluted to working standards of 0, 1, 2, 5, 15, 25, 50, 75, and 100 ppm. All ion measurements fell within the calibration range, and detection limits were verified through blank and standard checks during analysis.

2.4. Proline analysis

Proline extraction was carried out using 250 mg of finely ground, oven‐dried leaf samples, each combined with 25 mL of deionized water. Samples were incubated in a water bath at 45°C for 1 h, with vigorous vortex mixing every 15 min (four times total). Following incubation, tubes were centrifuged at 2520 g, and the supernatant was filtered and used for proline determination. For each assay, 1 mL of leaf extract was combined with 1 mL of acid ninhydrin reagent and 1 mL of glacial acetic acid. The mixture was shaken vigorously at 375 rpm for 12 min on a platform shaker (Innova 2000, New Brunswick Scientific), followed by heating in a boiling water bath (100°C) for 1 h in a closed system to form the proline‐ninhydrin chromophore complex. L‐Proline (Sigma‐Aldrich) was used as the standard. After cooling, the proline‐ninhydrin complex was extracted by vigorous shaking with 2 mL toluene. The upper toluene layer containing the chromophore was transferred into a glass cuvette, and absorbance was measured at 520 nm using a spectrophotometer (DU 7500, Beckman Coulter), with pure toluene serving as a blank. Proline concentrations were quantified using a standard curve and expressed as µmol g⁻¹ dry weight. Proline analysis was performed using three biological replicates, each with one technical replicate.

2.5. Metabolite extraction and metabolomic analyses

Freeze‐dried leaf and root samples were ground to a fine powder using Geno/Grinder 2010 (SPEX SamplePrep) at a frequency of 1450 rpm for 2 min with two grinding cycles. For each sample, the powder was resuspended in a monophasic extraction solvent (30 acetonitrile:30 methanol:20 isopropanol:20 water, all liquid chromatography‐mass spectrometry [LC‐MS] grade solvents) using a ratio of 100 µL/mg of powder (9.5–10 mg used in the extractions). The samples underwent sonication on ice for 5 min, followed by vortex‐agitation for 90 min at 4°C, centrifugation at 4°C for 10 min at 21,255 G, and collection of supernatants into LC‐MS vials. A pooled quality control (pooled QC) sample, comprising equal volumes of all samples, was used to monitor system stability and performance. LC‐MS metabolomics analysis was performed at the UC Riverside Metabolomics Core Facility as described (Rothman et al., 2019), with minor modifications. Briefly, a Synapt G2‐Si quadrupole time‐of‐flight mass spectrometer (Waters Corporation) coupled to an Acquity I‐class ultra performance liquid chromatography system (Waters Corporation) was used, employing an Acquity CSH phenyl‐hexyl column (2.1 × 100 mm, 1.7 µM) (Waters Corporation) for separations. The mobile phases consisted of (A) water with 0.1% formic acid and (B) acetonitrile with 0.1% formic acid. The column was maintained at 40°C with a flow rate of 250 µL/min. Injection volume was 1 µL. The gradient elution was as follows: 0 min, 1% B; 1 min, 1% B; 8 min, 40% B; 24–26.5 min, 100% B; 27 min, 1% B; 30 min, 1% B. MS data were acquired over a scan range of 50–1200 m/z with a 100 ms scan time, and MS/MS spectra were obtained in a data‐dependent manner. The source temperature was set at 150°C, and the desolvation temperature was set at 600°C, with desolvation, nebulizer, and cone gas at 600 L/h, 6.5 bar, and 0 L/h, respectively. Nitrogen served as the desolvation and nebulizer gas, while argon was used as collision gas. The capillary voltage was maintained at 1 kV in positive‐ion mode, and leucine enkephalin was infused for mass correction. The pooled QC sample was acquired in the beginning of the run, in the end, and after every six injections.

2.6. Data processing

Untargeted LC‐MS data processing, including peak picking, alignment, deconvolution, integration, normalization, and spectral matching, was performed in Progenesis QI software (Nonlinear Dynamics). Post‑acquisition mass correction was performed in Progenesis QI. No batch correction was applied in subsequent statistical analyses. Peak intensity data were normalized using total ion abundance, excluding features with variation >30% across QC injections (Barupal et al., 2018; Dunn et al., 2011). To account for insource fragments, features were assigned to a cluster‐ID group using RAMClust (Broeckling et al., 2014). Annotation level confidence was assigned based on an extension of the Metabolomics Standard Initiative guidelines (Schymanski et al., 2014; Sumner et al., 2007), with level 1a indicating an MS, MS/MS, and retention time match to an in‐house database; level 1b indicating an MS and MS/MS match to an in‐house database; level 2a representing an MS and MS/MS match to an external database; and level 2b denoting an MS and MS/MS match to the LipidBlast and in silico database (Kind et al., 2013). Several mass spectral metabolite libraries were searched, including Mass Bank of North America libraries, Metlin (Smith et al., 2005), and an in‐house library, to enhance metabolite identification.

2.7. Identification of genes regulating metabolites

Genes encoding enzymes catalyzing the biosynthesis of kaempferol and amyrin were obtained from CornCyc 10.0.1 (https://www.plantcyc.org) online database. Genes related to schaftoside, tricin, isovitexin, coumaroyl tyramine, and lysophosphatidylcholine (LPC) 18:3 were found searching the online literature (Niu et al., 2023; Schmidt et al., 1999; Z.‐L. Wang et al., 2020; Zhou et al., 2008). Gene for lanosterol biosynthesis was found by blasting the protein sequence of house mouse lanosterol synthase (BC029082.1) against maize protein database.

2.8. Statistical analysis

The prcomp R package was used to perform principal component analysis (PCA). For PCA, the data were mean‐centered (by subtracting the mean from each variable, producing a data set with zero as mean) and scaled (by dividing each variable by its standard deviation) to unit variance with no additional transformation. To identify differently abundant metabolites (false discovery rate adjusted p ≤ 0.05), we performed 3‐way analysis of variance with genotype, treatment, and time points as independent variables, followed by a two‐tailed t‐test for pairwise comparisons by calculating estimated mean measures using emmeans function in R. To control for multiple testing, p‐values were adjusted using Benjamini and Hochberg correction to reduce false discovery rate.

3. RESULTS

3.1. Distinct physiological responses of two inbred lines to salt stress

Exposure of two maize inbred lines to salt stress revealed that C68 (salt‐sensitive) exhibited a significant decrease in shoot and root dry weight, whereas NC326 (salt‐tolerant) had no significant change in these traits, as evident from comparison with the respective control plants and the salt tolerance index (Figure 1A,B). Furthermore, while both inbred lines experienced a reduction in plant height under saline conditions, the decline was greater in C68 (68.5%) compared to NC326 (43.1%) (Figure 1C). These observations align with our previous findings that NC326 and C68 represent salt‐tolerant and salt‐sensitive inbred lines, respectively (Sandhu et al., 2020). To investigate if these phenotypic responses were associated with changes in ion concentrations, we analyzed sodium (Na), chloride (Cl), and potassium (K) concentrations in the leaves and roots of both inbred lines. Salt stress led to a significant increase in Na and Cl concentrations and a significant decrease in K concentration in both tissues (Figure 1D–I). Notably, the relative increase in Na concentration (3.34‐fold in leaf and 1.93‐fold in roots) was substantially higher in C68 across both tissues, whereas the relative increase in Cl concentration (3.13‐fold in leaf and 0.94‐fold in roots) was greater in the leaves but comparable in roots between C68 and NC326. These results indicate that salinity perturbs the ion homeostasis to a greater extent in the C68 inbred line. Consistent with the positive association of proline accumulation and salt tolerance (El Moukhtari et al., 2020), NC326 accumulated significantly more leaf proline than C68 under salt stress, indicating that genetic differences in the synthesis of this amino acid contribute to salinity stress response in maize (Figure 1J). Collectively, these results demonstrate that NC326 and C68 exhibit distinct responses in tissue accumulation, physiological changes, and growth under salt stress, with NC326 showing greater resilience through better ion homeostasis and higher proline accumulation.

FIGURE 1.

Salt stress impacts plant growth and physiology in maize. (A–C), Effect of salinity on growth parameters, including shoot dry weight (A), root dry weight (B), and plant height (C), with salt tolerance index (STI) shown by red circles. (D–I), effect of salinity on ion accumulation in the leaves (D–F) and roots (G–I). (J) Effect of salinity on proline accumulation in leaves. Relative change (RC) is marked by red circles in (D–J). Asterisks indicate significant differences between control (gray) and salinity (black) treatments (p ≤ 0.05, n = 3). Error bars represent the standard error.

3.2. Global change in leaf and root metabolome in response to salt stress

To capture the metabolome variation underlying salt tolerance, we examined metabolite changes in the leaves and roots of NC326 and C68 inbred lines under control and salinity conditions. Untargeted metabolomic analysis detected 3498 and 3044 mass features in the leaf and root tissue, respectively. PCA based on leaf mass features revealed that PC1 and PC2 explained 31.6% and 24.1% variation in the metabolome, respectively (Figure 2A). Similarly, PCA based on root mass features showed that PC1 and PC2 explained 29.3% and 19.7% variation in the root metabolome, respectively (Figure 2B). Remarkably, separate clustering of C68 and NC326 inbreds under control conditions in leaf and roots indicated that these inbred lines possess inherently distinct metabolic profiles. Furthermore, the leaf and root metabolome of the salt‐sensitive inbred C68 had a clear distinction between salt stress and control conditions, indicating large‐scale changes in response to salinity. However, in the salt‐tolerant inbred NC326, we observed a less pronounced global distinction in leaf and root metabolomes between salt stress and control conditions, suggesting that targeted changes in specific metabolites, rather than broad metabolic shifts, are key to salinity tolerance.

FIGURE 2.

Salt stress drives temporal shifts in leaf and root metabolomes. Principal component analysis (PCA) illustrates temporal changes in the metabolome of leaves (A) and roots (B) during salt stress in salt‐sensitive C68 (squares) and salt‐tolerant NC326 (triangles) inbred lines. Each shape represents the mean of three biological replications. Different colors represent distinct time points: 0 h (sky blue), 24 h (light purple), and 48 h (light orange). Solid and pattern‐filled shapes represent control and salinity treatment, respectively, and PC1 and PC2 denote the first and second principal components, respectively.

The lack of annotation of most identified mass features in metabolomic studies remains a major constraint due to limited availability of reference spectra, preventing their designation as metabolites (Alseekh et al., 2021; da Silva et al., 2015). Of 3498 mass features identified in leaves, we annotated 177 (5.06%) mass features as specialized metabolites belonging to 43 classes of organic compounds (Supporting Information Dataset S1). Similarly, 143 (4.70%) of 3044 mass features identified in the roots were annotated as specialized metabolites belonging to 51 classes of organic compounds (Supporting Information Dataset S2). Phosphatidylcholines was the most abundant compound class, accounting for 17 metabolites each in leaf and root.

3.3. Constitutive accumulation of certain metabolites suggests adaptive priming

Since the global metabolome suggested inherent differences in metabolic profiles of the two inbred lines, we first compared the metabolomes of both inbred lines under control conditions to identify the baseline differences in metabolite accumulation. In leaves, 19 (10.7%) of the annotated specialized metabolites showed differential abundance between the two inbred lines (Supporting Information Dataset S1) (Figures S1 and S2), with 12 metabolites classified as flavonoids and the remaining seven representing unique compound classes. Of these, 11 flavonoids had significantly higher accumulation in the leaves of NC326 compared to C68 at all three timepoints (Figure 3A–K) (Supporting Information Dataset S1). These include antioxidant flavonoids such as schaftoside, tricin, and kaempferol‐related compounds that are expected to impart protection against excessive oxidative stress triggered by biotic or abiotic stresses. In roots, 11 (7.7%) of the annotated specialized metabolites differentially accumulated between the two inbred lines, and, unlike leaves, all these metabolites belonged to unique classes of compounds (Supporting Information Dataset S2) (Figures S1 and S3). Roots of NC326 had constitutively significantly higher levels of a polyunsaturated fatty acid 28:6 at all three timepoints (Figure 3L) (Supporting Information Dataset S2). Constitutively higher levels of several specialized metabolites, each significantly enriched in NC326 leaves and roots, indicate that this subset of metabolites may prime the genotype for adaptive salinity tolerance. To ask how exposure to salt stress impacted these constitutive metabolites, we compared leaf and root tissues from the control and saline treatment. Because plants were gradually exposed to salt until reaching the desired water salinity, achieved at 0 h stage, we observed small differences in metabolite abundance between the C68 and NC326 inbred lines under both control and salinity conditions at 0 h in both tissues (Figures 3). All the compounds maintained significantly higher levels in NC326 compared to C68 under salt stress at all stages.

FIGURE 3.

Adaptive priming of the metabolome underlies salt tolerance in maize. Metabolite levels (normalized peak intensity in thousands) in salt‐tolerant NC326 (triangles) and salt‐sensitive C68 (squares) are shown under control (gray) and salinity (black) conditions at different time points. (A‐K), changes in various flavonoids in these inbred lines under control and salinity conditions. (L), levels of fatty acid (28:6) in roots. All shown metabolites were significantly different with a significance threshold of false discovery rate adjusted p ≤ 0.05 for pairwise comparisons discussed in the text.

3.4. Identification of salt‐responsive metabolites

To identify metabolites associated with salt stress, we compared the metabolome of both inbreds under control and salinity conditions for all three time points. In leaves, 11 metabolites showed differential abundance in C68, while only three were differentially abundant in NC326 (Supporting Information Dataset S1) (Figures S1 and S2), suggesting that susceptible genotypes undergo broader metabolic shifts in response to salinity, whereas tolerant genotypes maintain a more stable metabolic profile. The majority (eight) of the differentially accumulated metabolites in C68 were upregulated. While most differentially abundant metabolites in leaves were specific to each inbred line, a few—including a galactolipid (DGDG;18:3/20:3) and a sterol (lanosterol)—were shared between both inbreds and exhibited similar changes in response to salt stress (Figure 4A,B) (Supporting Information Dataset S1). Consistent with the role of sterols in enhancing membrane stability (Du et al., 2022), lanosterol levels significantly increased in both inbred lines at all timepoints under salt stress. In contrast, galactolipid levels significantly decreased in both inbreds at 0 h, suggesting the reprogramming of plastid membranes in response to salt stress.

FIGURE 4.

Metabolic changes associated with differential response to salinity. Levels (normalized peak intensity in thousands) of key metabolites in salt‐tolerant NC326 (triangles) and salt‐sensitive C68 (squares) are shown under control (gray) and salinity (black) conditions at different time points in leaves (A–B) and roots (C–E). All shown metabolites were significantly different with a significance threshold of false discovery rate adjusted p ≤ 0.05 for pairwise comparisons discussed in the text.

Similarly, in roots, seven metabolites were differentially accumulated in C68, while only three showed differential accumulation in NC326 inbred lines (Supporting Information Dataset S2) (Figures S1 and S3). About two‐thirds (71%) of differentially abundant metabolites were phospholipids and downregulated in C68 roots, and two of these phospholipids were downregulated in NC326 roots. Notably, two common lysophosphatidylcholines (LPC 18:1 and LPC 18:2), a subclass of phospholipids, declined significantly at all timepoints under salt stress in C68 roots, whereas in NC326, the levels of these metabolites decreased at 0 h but later stabilized to levels comparable to control plants (Figure 4C,D) (Supporting Information Dataset S2). This pattern of root metabolites maintaining normal levels after fluctuations under salinity in NC326 is consistent with observations of global metabolome and suggests that NC326 roots adapt more rapidly to salt stress (Figure 2B). While the flux of phospholipids changed similarly in these inbreds, enhanced accumulation of metabolites with antioxidant properties was unique to C68 under salt stress. For example, moupinamide (also known as feruloyltyramine) increased significantly in C68 at 24 h under salt stress, likely as a stress‐mitigation response to oxidative damage (Figure 4E) (Supporting Information Dataset S2). These findings highlight distinct metabolic strategies between the inbreds, with NC326 showing early metabolic adjustments that may contribute to its salt tolerance.

3.5. Identification of candidate genes underlying salt tolerance

Metabolites serve as a functional readout of cellular biology and, therefore, provide a direct link between the genome and the external plant phenotypes. To understand the genetic architecture of salt tolerance, we identified the candidate genes involved in the biosynthesis of metabolites associated with response to salt stress. Our analyses revealed 10 genes associated with eight metabolites, including eight genes linked to leaf metabolites and two to root metabolites (Table 2). Among these, UDP‐glycosyltransferase, O‐methyltransferase, and flavonol synthase were associated with the synthesis of antioxidant flavonoids such as schaftoside, tricin, and kaempferol‐related compounds, likely conferring stress protection through antioxidant activity or anti‐inflammatory capacities. Genes encoding patatin‐like phospholipase A, associated with phospholipid metabolism, likely mediate lipid remodeling to stabilize membranes under salinity. Additionally, N‐hydroxycinnamoyl transferase, implicated in coumaroyl tyramine biosynthesis, might contribute to stress responses through phenolic metabolism. Upregulation of genes encoding patatin‐like phospholipase A and N‐hydroxycinnamoyl transferase in maize seedlings under salt stress (Zhang et al., 2021) further supports the role of these metabolites in salt tolerance. Besides the two flavanol synthases and an O‐methyltransferase, the role of the remaining eight genes in salt tolerance remains unexplored, suggesting that these are novel targets for improving in salt tolerance. Collectively, these candidate genes provide potential targets for improving maize salt tolerance through breeding or genetic engineering.

TABLE 2.

Putative genes encoding enzymes upstream of metabolites.

Metabolites	Enzymes	Putative genes
Schaftoside	UDP‐glycosyltransferase 708A6	Zm00001d037382
Tricin	O‐methyltransferase	Zm00001d049541
Similar to Kaempferol	Flavonol synthase	Zm00001d018187 and Zm00001d018181
Lanosterol	Lanosterol synthase	Zm00001d008674
Amyrin	Terpene cyclase	Zm00001d052971, Zm00001d014695, and Zm00001d035389
Isovitexin	C‐glucosyl transferase	Zm00001d037382
Coumaroyl tyramine	N‐hydroxycinnamoyl transferase	Zm00001d048407
LPC 18:3	Patatin‐like phospholipase A	Zm00001d031811

Abbreviation: LPC, lysophosphatidylcholine.

4. DISCUSSION

Salinity negatively impacts plant growth and vigor across all species, although salt‐sensitive genotypes exhibit a more pronounced decline, suggesting the development of adaptive mechanisms to cope with salt stress. In this study, the salt‐sensitive C68 inbred line showed a pronounced decrease in shoot dry weight, root dry weight, and plant height under salt stress compared to salt‐tolerant NC326 inbred. While the concentration of Na and Cl increased significantly in the leaves of both inbred lines under salt stress, their relative accumulation of these salt ions and the resulting ion imbalance was significantly higher in C68. In contrast, both inbred lines had higher accumulation of these two ions in roots. These observations indicate that, compared to roots, ion imbalance in the leaves is more detrimental to plant growth and vigor, likely due to disruption of photosynthesis, water balance, and other physiological functions. Additionally, NC326 maintained better Na⁺ homeostasis, thus reducing the negative effects of excessive Na⁺ accumulation. Significantly higher proline in the leaves of NC326 indicates that the genetic differences for the synthesis of this osmoprotectant play a critical role in salt tolerance by stabilizing cellular functions, mitigating oxidative stress, and aiding in osmotic adjustment. Consistent with our findings, a recent study demonstrated that application of the short peptide VTID significantly enhanced proline accumulation in maize seedlings under salt stress, contributing to improved growth and stress tolerance (Wu et al., 2024). Taken together, these results reinforce the findings from previous studies (El Moukhtari et al., 2020; Gupta & Huang, 2014; Sandhu et al., 2021) and that underscore ion homeostasis and metabolic adjustments represent key physiological mechanisms conferring salt tolerance in maize.

Identification of specific metabolites differentiating salt‐sensitive and salt‐tolerant inbred lines is crucial for determining the metabolic pathways contributing to salt tolerance in maize. While the untargeted approach identified over 3000 mass features each in root and leaves, only fewer than 5% of these features could be annotated to specific metabolites due to various reasons, including the limited availability of reference spectra and insufficient published information (Alseekh et al., 2021). This significant gap in specialized metabolites annotation, consistent with our previous observations (Brar et al., 2025), poses a major challenge for uncovering mechanisms and genes underlying salinity tolerance. However, this limitation presents exciting opportunities to characterize these unknown metabolites, which are significantly and differentially regulated between inbred lines and by salt stress. In the current study, we highlighted the top five unknown metabolites from each of the root and leaf tissues that showed the highest fold‐change in the salt‐tolerant NC326 inbred under salinity stress (Table S1). Identification of these unknown metabolites will enhance our understanding of salt tolerance mechanisms in maize and provide valuable targets for future functional studies and metabolic engineering efforts. Future research, including improved analytical techniques, metabolite‐wide association studies for comprehensive linking of metabolites and unknown features to salt tolerance, and metabolite quantitative trait locus (mQTL) mapping to identify genes and enzymes responsible for synthesizing key metabolites and unknown features associated with salt tolerance, will propel our knowledge of salt tolerance in maize and facilitate crop improvement strategies.

Metabolomics analysis revealed constitutively higher flavonoids in the leaves of NC326 compared to C68 under control conditions, suggesting that adaptive priming through enhanced accumulation of antioxidants, including schaftoside, tricin, and kaempferol‐related compounds, could be a key mechanism underlying salt tolerance. Similar constitutive patterns were also observed in roots, although involving fewer metabolites primarily representing fatty acids, indicating that maintaining membrane fluidity and permeability may be the principal adaptive strategy employed by roots. Schaftoside, a major bioactive compound in medicinal plants like Clinacanthus nutans, exhibits anti‐inflammatory, antioxidant, and antiviral properties and was shown to attenuate acute liver injury by inhibiting ferroptosis through activation of the Nrf2/GPX4 pathway (Lu et al., 2024). However, ROS were not directly measured in this study to validate these mechanisms. Overexpression of flavonoid 3‐hydroxylase in rice (Oryza sativa L.) enhanced kaempferol accumulation and salt tolerance (Jan et al., 2021), while exogenous application of kaempferol similarly alleviated salinity stress in potato (Ramzan et al., 2024). Tricin, a lignin monomer especially important in monocots, contributes to cell wall rigidity and plays roles in growth, stress tolerance, and pest defense (M. Li et al., 2016). Overexpression of CYP75B4, which catalyzes tricin biosynthesis, increased tricin levels in rice, enhancing salt tolerance likely by reinforcing cell walls via lignin production and mitigating oxidative stress (Lam et al., 2015; Ruan et al., 2025). Our tricin data at 0 h and 48 h, despite no significant change at 24 h, align with previous findings showing a general decrease in tricin levels under salt and drought stress compared to controls at the same developmental stage (Moheb et al., 2013). Several flavonoids identified in this study also exhibit protective functions in human cells by stabilizing membranes and reducing ROS‐induced damage, underscoring parallels between plant and animal antioxidant mechanisms. The constitutively elevated flavonoids observed in the salt‐tolerant NC326 inbred are likely key contributors to its salinity tolerance. This is consistent with reports showing that apigenin‐related flavonoids enhance salt tolerance in maize (Liang et al., 2021) and that isovitexin is significantly upregulated under salinity, potentially mitigating oxidative stress (E. Zhang, Zhu, et al., 2023). Furthermore, high baseline levels of galactaric acid, cellobiose, and 2,4‐dihydroxybutanoic acid have also been associated with salt tolerance in maize hybrids (Richter et al., 2015).

This constitutive metabolite upregulation could reflect inherent genetic differences arising from natural selection, or a correlated response to artificial selection for other stresses, given the limited targeted breeding for salinity stress in maize. Additionally, such adaptive priming might result from transgenerational epigenetic memory generated by past exposures of these genotypes to salt stress. Future systems genetics studies integrating genomic and epigenomic analyses will be valuable for disentangling the genetic and epigenetic contributions to salt tolerance. Furthermore, these studies may provide insights into potential resource allocation trade‐offs and reveal the energetic costs associated with constitutive defense mechanisms and their impact on crop yields under non‐stress conditions.

Compared to the constitutive metabolome, relatively fewer metabolites showed differential abundance in response to salinity, suggesting that inducible changes in metabolome are more targeted. A higher number of metabolites responded to salt in C68 leaves and roots than in NC326, suggesting that a large proportion of the inducible metabolome is involved in mitigating rather than preventing stress. For example, the metabolite moupinamide (feruloyltyramine) specifically increased in C68 at 24 h under salt stress (Figure 4E), suggesting a potential role in mitigating oxidative damage. Consistent with such a role, feruloyltyramine has been shown to accumulate by up to 10‐fold in tomato (Solanum lycopersicum) leaves following wounding (Pearce et al., 1998). The overall increase in antioxidant metabolites in C68 underscores the importance of oxidative stress management as a reactive defense. Besides genotype‐specific responses, several metabolites were commonly responsive in both lines, reflecting shared mechanisms against salinity stress. Notably, lanosterol accumulation increased in leaves of both lines, suggesting a role in membrane stabilization and reduced ion permeability (Chalbi et al., 2015; Kerkeb et al., 2001; Salama & Mansour, 2015), possibly through conversion to phytosterols (Ohyama et al., 2009). Such a role is furthers supported by upregulation of a lanosterol derivative (14‐demethyllanosterol) in response to salinity in maize (E. Zhang, Zhu, et al., 2023). While sterols are broadly linked to abiotic stress resilience (Du et al., 2022), the precise role of lanosterol in stress tolerance remains to be clarified.

We also observed a decline in galactolipid levels in leaves, possibly due to their conversion into sulfolipids that enhance chloroplast membrane stability under saline conditions (Guo et al., 2019). Similar responses have been documented in barley (Hordeum vulgare) and other plants, where galactolipid depletion has been proposed to protect chloroplast membranes and limit photoinhibition to improve salt tolerance (Bejaoui et al., 2016; Müller & Santarius, 1978; Sui & Han, 2014). In roots, the reduction in lysophosphatidylcholines content under salt stress suggests active remodeling of membrane lipids, potentially to enhance membrane integrity and reduce ion leakage. As the primary interface with saline soil, roots rely heavily on such lipid adjustments as a critical adaptive mechanism for maintaining membrane stability and function under stress (Han & Yang, 2021; M. Li et al., 2023; Xu et al., 2021). Supporting this argument, exogenous application of four amino acids has been shown to improve salt tolerance in maize by altering lipid and lipid‐like molecules (Wu et al., 2024). Collectively, targeted adjustments in antioxidant production and membrane lipid composition emerge as central components of inducible metabolic strategies contributing to salt tolerance in maize. However, determining whether these metabolites act as causal drivers of tolerance or are downstream responses to stress will require further investigation, including exogenous metabolite applications and targeted genetic studies.

The identification of 10 unique genes associated with the biosynthesis of eight specific metabolites provides valuable targets for improving salt tolerance in maize. The role of these candidate genes in conferring salt tolerance can be further validated through reverse genetic approaches. Seven genes involved in flavonoid biosynthesis are immediate candidates for crop improvement strategies. The identification of genes involved in altered metabolite regulation remain a significant limitation, as only a few candidates could be identified in this study. To fully understand the genetic basis of complex traits, such as salt tolerance, future genomic approaches that include mQTL analyses and genome‐wide association studies are necessary.

In conclusion, this study identified 56 metabolites underlying differential salt‐stress responses in two maize inbred lines. Metabolomics revealed two key adaptive strategies, including constitutive accumulation of flavonoids and lipids for antioxidant defense and membrane stabilization, and inducible metabolic adjustments targeting oxidative stress and membrane integrity. Increased salt induced lanosterol levels in both inbreds emphasized possible roles in maintaining membrane integrity, enhanced phytosterol synthesis and ion transport. Although 10 candidate genes regulating eight metabolites were identified as candidate targets for crop improvement, limited annotation restricted further genetic insights. Future studies employing mQTL mapping, metabolite‐wide association analyses, and epigenomic approaches will enhance the genetic understanding of salt tolerance, informing targeted maize breeding strategies.

AUTHOR CONTRIBUTIONS

Manwinder S. Brar: Data curation; formal analysis; investigation; methodology; validation; visualization; writing—original draft; writing—review and editing. Amancio De Souza: Data curation; formal analysis; investigation; methodology; writing—original draft; writing—review and editing. Avineet Ghai: Investigation; methodology; writing—review and editing. Jorge F.S. Ferreira: Formal analysis; writing—original draft; writing—review and editing. Devinder Sandhu: Conceptualization; data curation; funding acquisition; investigation; methodology; project administration; resources; supervision; writing—original draft; writing—review and editing. Rajandeep S. Sekhon: Conceptualization; data curation; funding acquisition; investigation; methodology; project administration; resources; supervision; writing—original draft; writing—review and editing.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts interests.

Supporting information

Supplementary Figure S1. Differentially abundant metabolites in different treatments.

Supplementary Figure S2. Differentially abundant metabolites in leaves of salt‐sensitive C68 and salt‐tolerant NC326 inbred lines.

Supplementary Figure S3. Differentially abundant metabolites in roots of salt‐sensitive C68 and salt‐tolerant NC326 inbred lines.

Supplementary Table S1. List of top five unknown mass features in NC326 inbred lines that were differentially abundant in leaf and roots in response to salinity.

Supplementary Dataset S1. List of identified mass features and differentially abundant metabolites in leaves of NC326 and C68 inbred lines under control and salinity conditions.

Supplementary Dataset S2. List of identified mass features and differentially abundant metabolites in roots of NC326 and C68 inbred lines under control and salinity conditions.

Brar, M. S. , De Souza, A. , Ghai, A. , Ferreira, J. F. S. , Sandhu, D. , & Sekhon, R. S. (2025). Untargeted metabolomics reveals key metabolites and genes underlying salinity tolerance mechanisms in maize. The Plant Genome, 18, e70102. 10.1002/tpg2.70102

DATA AVAILABILITY STATEMENT

Data could be found in Supporting Information Datasets S1 (leaf metabolome) and S2 (root metabolome).