A total of 0.1 mL of ZnO nanoparticles (ZnO NPs) suspension was dropped onto copper grids for transmission electron microscopy (TEM) observation (FEI Talos F200X, Thermo Fisher Scientific, USA). After air-drying, images were captured and particle size distribution was analyzed using ImageJ software to generate a particle size distribution histogram, and ZnO NPs are in a stable state in solution (Supplementary Figure 1).

Seeds of Glycyrrhiza uralensis Fisch. were collected from Yanchi County, Ningxia, China. To remove surface contaminants, the seeds were thoroughly rinsed with ultrapure water three times, immersed in 75% (v/v) ethanol for 30 s, and subsequently treated with 5.0% sodium hypochlorite solution for 10 min. After sterilization, seeds were washed three additional times with ultrapure water. The disinfected seeds were placed in Petri dishes for germination. When the seedlings reached approximately 2 cm in height, uniform individuals were transferred into plastic pots filled with farmland soil at a density of 20 plants per pot. Plants were maintained under controlled environmental conditions with a 14 h light (24 °C)/10 h dark (16 °C) cycle and relative humidity maintained at 60%.

Based on preliminary experiments, 160 mM NaCl was selected as the salinity level and 40 mg·L^-¹ as the ZnO NPs concentration. Four treatments were established: (1) CK (control, no NaCl and ZnO NPs); (2) S (160 mM NaCl); (3) CK + ZnO NPs (40 mg·L^-¹ ZnO NPs); and (4) S + ZnO NPs (160 mM NaCl + 40 mg·L^-¹ ZnO NPs). The first treatment was applied when the first true leaf emerged, and 200 mL of the corresponding solution was evenly applied to each pot. Treatments were repeated every three days for a total of ten applications. Control plants were irrigated with the same volume of ultrapure water. The experiment was conducted using a completely randomized design. Each treatment consisted of three independent biological replicates, with each replicate comprising 20 seedlings grown in a single pot. All physiological and biochemical measurements were performed using independent biological samples, and technical replicates were averaged prior to statistical analysis.

After one month of treatment, plant samples were collected. Different tissues were harvested to evaluate the effects of salinity and ZnO NPs. Root samples were immediately frozen in liquid nitrogen and stored at −80 °C for transcriptomic and metabolomic analyses.

Physiological traits were determined using five fully expanded healthy leaves per treatment. Net photosynthetic rate (Pn), transpiration rate (Tr), stomatal conductance (Gs), and intercellular CO₂ concentration (Ci) were measured with a portable photosynthesis system (Yaxin-1105, Beijing, China) following previously reported protocols (Tirani et al., 2013; Fan et al., 2021). Measurements were performed between 09:00 and 11:00 under a controlled CO₂ concentration of 400 μmol·mol^-¹ and an airflow rate of 0.6 m·min^-¹. Malondialdehyde (MDA) levels were quantified using the thiobarbituric acid (TBA) method to evaluate lipid peroxidation. Enzymatic activities of superoxide dismutase (SOD), and peroxidase (POD) in leaf tissues were determined using commercial assay kits (Jiangsu Edison Biotechnology Co., Ltd., China) in accordance with the manufacturer’s instructions (Kubis, 2005; Jambunathan, 2010; Kim et al., 2012). The concentrations of sodium (Na⁺) and zinc (Zn²⁺) in plant tissues were measured using inductively coupled plasma mass spectrometry (ICP–MS). For elemental analysis, 0.1 g of dried leaf and root samples was digested in 2.5 mol·L^-¹ nitric acid, and the resulting solutions were subjected to analysis after appropriate dilution.

Fresh root samples were freeze-dried, finely ground into powder, and extracted using 70% (v/v) methanol solution. The extracts were passed through 0.22 μm membrane filters and transferred into autosampler vials for analysis. Metabolite profiling was conducted using an ultra-performance liquid chromatography system coupled with tandem mass spectrometry (UPLC–ESI–MS/MS; ExionLC™ AD, Shanghai, China). Chromatographic separation was carried out on a Kinetex C18 column (2.1 mm × 100 mm, 2.6 μm). The mobile phase consisted of solvent A (0.01% acetic acid in water) and solvent B (50% acetonitrile/isopropanol). The column temperature was maintained at 25 °C, while the autosampler temperature was set at 4 °C. A sample volume of 2 μL was injected at a flow rate of 0.3 mL·min^-¹. Mass spectrometric detection was performed on a QTrap 6500+ system (Sciex Technologies). The electrospray ionization parameters were as follows: ion spray voltage of 5500 V (positive mode) and −4500 V (negative mode), ion source temperature of 400 °C, curtain gas at 35 psi, and ion source gas 1 at 50 psi.

LC–MS data acquisition and analysis were performed by Baipu Biotechnology Co., Ltd. (Shanghai, China). Differential metabolites were screened based on the criteria of fold change ≤ 0.5 or ≥ 2 and adjusted p-value < 0.05. Identified differential metabolites were mapped to the KEGG database for pathway enrichment analysis.

Total RNA was isolated from licorice root tissues using the RNeasy Pure Plant Kit (DP432, Tiangen, China). RNA integrity was evaluated by 1% agarose gel electrophoresis, and purity and concentration were assessed using a NanoPhotometer spectrophotometer (IMPLEN, USA).

Double-stranded cDNA libraries were prepared using the RNA Library Prep Kit (Tiangen, Beijing, China), and high-throughput sequencing was performed on the Illumina HiSeq platform by Baiqu Biotechnology Co., Ltd. (Shanghai, China). Raw sequencing reads were quality-filtered to eliminate low-quality reads, and GC content and sequencing error rate were analyzed to ensure data reliability. Clean reads were assembled de novo using Trinity software to generate reference transcripts. Unigenes were obtained through hierarchical clustering using Corset. Differentially expressed genes (DEGs) were identified using the DESeq2 package with the thresholds of fold change ≤ 0.5 or ≥ 2 and adjusted p-value < 0.05. Functional annotation was performed based on Gene Ontology (GO), and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses were conducted. Weighted gene co-expression network analysis (WGCNA) was carried out using the WGCNA package implemented in R software (v3.3.0) to identify gene modules associated with different treatments.

All statistical analyses were conducted using SPSS version 22.0 (IBM Corp., USA). Differences among treatment groups were assessed by one-way analysis of variance (ANOVA) followed by Tukey’s multiple comparison test. A significance level of p < 0.05 was applied. Visualization of transcriptomic and metabolomic datasets, including Venn diagrams, heatmaps, and principal component analysis (PCA) plots, was generated using the OmicShare online platform (www.omicshare.com/tools). Additional figures were produced using GraphPad Prism 10.0 (GraphPad Software, USA) and Adobe Illustrator 2024 (Adobe Inc., USA). The transcriptomic data have been deposited in the NCBI Sequence Read Archive (SRA) under accession number PRJNA1421034.

As demonstrated by the transmission electron microscopy scanning results, the ZnO NPs exhibited a predominant spherical morphology under electron microscope observation (Figures 1A–C). The size distribution of the particles ranged from 2–12 nm, with an average and median size of 6.5 nm (Figure 1D).

To assess the effects of salinity and zinc oxide nanoparticles (ZnO NPs) on G. uralensis, leaf morphology was examined. Salt treatment caused visible chlorosis at the leaf tips, whereas this symptom was markedly alleviated under combined salt and ZnO NPs exposure (Figure 2A), indicating that ZnO NPs promoted plant growth under salt stress. Photosynthetic performance and leaf physiological status were further evaluated using multiple gas-exchange parameters. Exogenous application of ZnO NPs significantly mitigated the salt-induced decreases in stomatal conductance (Gs), net photosynthetic rate (Pn), transpiration rate (Tr), and intercellular CO₂ concentration (Ci); however, these values remained lower than those observed under control (CK) conditions (Figure 2B). In addition, ZnO NPs treatment resulted in increased chlorophyll and carotenoid contents in leaves (Supplementary Figure 2), suggesting improved photosynthetic pigment accumulation. Among the antioxidant-related indicators analyzed, ZnO NPs treatment was associated with a significant increase in SOD activity compared with salt stress alone., whereas no significant effects were observed for malondialdehyde (MDA) content or peroxidase (POD) activity. Nevertheless, in the S + ZnO NPs treatment group, SOD activity, as well as MDA and POD levels, were higher than those in the control group (Figure 2C). Elemental analysis showed that Na⁺ accumulation in roots was highest under salt stress alone. In contrast, under combined S and ZnO NPs treatment, Na⁺ contents in both leaves and roots declined, while Zn²⁺ contents increased (Figure 2D), indicating that ZnO NPs altered ion uptake and distribution in G. uralensis seedlings.

Metabolomic profiling was performed to elucidate the global metabolic responses of G. uralensis roots to salt stress and ZnO NPs. Across all four treatments, a total of 580 metabolites were identified in G. uralensis roots. Principal component analysis (PCA) showed that the first two principal components (PC1 and PC2) explained 33.13% and 23.16% of the total variance, respectively. Clear separations were observed between CK samples and those subjected to salt stress and combined salt and ZnO NPs treatments (Figure 3A).

A total of 257 differentially accumulated metabolites (DAMs) were identified across the four pairwise comparisons. Based on relative abundance, 124 DAMs were detected in CK vs S, 71 in CK vs CK + ZnO NPs, and 109 in S vs S + ZnO NPs. In addition, 137 DAMs differed significantly between CK and S + ZnO NPs samples (Figure 3B). Among these DAMs, 134 were shikimate- and phenylpropanoid-related compounds, 35 were alkaloids, 25 were terpenoids, 19 were amino acids and short peptides, 14 were carbohydrates, 14 were polyketides, 12 were fatty acids, and 4 belonged to other categories (Supplementary Table 1).

Comparative analysis revealed pronounced differences in metabolic profiles among CK, S, CK + ZnO NPs, and S + ZnO NPs treatments. In CK vs CK + ZnO NPs, differential accumulation was observed in shikimate and phenylpropanoid-related compounds, alkaloids, carbohydrates, terpenoids, amino acids and short peptides, fatty acids, and polyketides, with only three shikimate-derived metabolites and one carbohydrate being downregulated. Under salt stress, shikimate and phenylpropanoid-related compounds constituted the most enriched class, followed by alkaloids, amino acids and short peptides, terpenoids, and fatty acids. In contrast, the concentrations of 31 shikimate and phenylpropanoid-related metabolites, 6 alkaloids, 5 amino acids and short peptides, 3 terpenoids, 2 fatty acids, and 1 polyketide were reduced under salt exposure. These findings indicate that ZnO NPs and salt treatments exerted distinct effects on the abundance and composition of root metabolites.

In the CK vs S + ZnO NPs comparison, shikimate and phenylpropanoid-related metabolites were the most enriched category, followed by terpenoids, alkaloids, and fatty acids. Among the DAMs, 47 shikimate and phenylpropanoid-related metabolites were upregulated, whereas the downregulated metabolites were mainly associated with shikimate and phenylpropanoid metabolism, terpenoids, alkaloids, carbohydrates, and amino acids and short peptides (Supplementary Table 1). These results suggest that metabolites related to the shikimate and phenylpropanoid pathways may play important roles in the beneficial effects of ZnO NPs on G. uralensis under salt stress.

To identify key metabolic pathways associated with S, ZnO NPs, and combined treatments, KEGG pathway enrichment analysis was conducted for the top 20 significantly enriched pathways. In the CK vs S comparison, 124 DAMs were significantly enriched in pathways including phenylpropanoid biosynthesis, biosynthesis of secondary metabolites, flavonoid biosynthesis, and biosynthesis of various plant secondary metabolites (Figure 3C). In CK vs CK + ZnO NPs, 16 enriched pathways were identified, including biosynthesis of secondary metabolites, alkaloid biosynthesis, plant secondary metabolite biosynthesis, phenylpropanoid biosynthesis, and flavonoid biosynthesis (Figure 3D). Notably, phenylpropanoid and flavonoid biosynthesis pathways were significantly enriched in both CK vs S and CK vs CK + ZnO NPs comparisons.

In CK vs S + ZnO NPs, enriched pathways included isoflavonoid biosynthesis, valine, leucine and isoleucine biosynthesis, phenylpropanoid biosynthesis, amino acid biosynthesis, biosynthesis of various plant secondary metabolites, flavonoid biosynthesis, and biosynthesis of secondary metabolites (Figure 3E), with most DAMs exhibiting significantly higher concentrations than those in the control group. Furthermore, in the S vs S + ZnO NPs comparison, DAMs were primarily enriched in nicotinate and nicotinamide metabolism, isoflavonoid biosynthesis, phenylalanine, tyrosine and tryptophan biosynthesis, amino acid biosynthesis, plant secondary metabolite biosynthesis, flavonoid biosynthesis, and secondary metabolite biosynthesis (Figure 3F).

Collectively, these results indicate that DAMs involved in secondary metabolism, phenylpropanoid biosynthesis, amino acid biosynthesis, and flavonoid biosynthesis may play critical roles in the mitigation of salt stress in G. uralensis roots by ZnO NPs.

Transcriptomic analysis was conducted to investigate global gene expression profiles of G. uralensis roots in response to ZnO NPs and salt stress. A total of twelve root cDNA libraries were constructed and subjected to RNA sequencing. All libraries exhibited Q30 scores exceeding 98.09%, indicating high sequencing quality and reliability (Supplementary Table 2). Principal component analysis (PCA) revealed detectable differences in gene expression patterns among treatments (Figure 4A), with the twelve samples clustering into four distinct groups. In addition, the distribution of gene expression levels and fragments per kilobase of transcript per million mapped reads (FPKM) values further confirmed the distinct transcriptional responses across the four treatments (Figures 4B, C), highlighting treatment-specific transcriptomic profiles.

Across CK vs CK + ZnO NPs, CK vs S, CK vs S + ZnO NPs, and S vs S + ZnO NPs comparisons, a total of 23,697 differentially expressed genes (DEGs) were identified (Figure 4D). Venn diagrams illustrated the overlap and uniqueness of DEGs among the four comparison groups (Figure 4E). The CK vs S + ZnO NPs comparison yielded the highest number of DEGs, suggesting that combined salt and ZnO NPs treatment exerted the strongest impact on the G. uralensis root transcriptome. In contrast, the fewest DEGs were detected in the CK vs CK + ZnO NPs comparison. A total of 4,941 DEGs were identified in CK vs S, reflecting transcriptional adjustments in response to salt stress alone. Moreover, 5,637 DEGs were found between S and S + ZnO NPs, indicating that ZnO NPs substantially altered gene expression under saline conditions. Notably, 418 DEGs were shared across all comparisons and were considered associated candidate genes responsive to ZnO NPs exposure under salt stress (Figure 4E).

Gene Ontology (GO) enrichment analysis revealed that DEGs in CK vs S were significantly associated with biological processes and molecular functions such as plant-type secondary cell wall biogenesis, response to hydrogen peroxide, glycolytic process, phenylpropanoid biosynthetic process, dioxygenase activity, and peroxidase activity, which are closely related to salt stress responses and cellular homeostasis (Figure 4F). Specifically, upregulated DEGs were mainly enriched in response to heat, response to hydrogen peroxide, dioxygenase activity, response to chitin, and response to caffeine, whereas downregulated DEGs were primarily associated with plant-type secondary cell wall biogenesis, extracellular region, fatty acid biosynthetic process, and glycolytic process (Figure 4F). GO enrichment analysis of DEGs in CK vs S + ZnO NPs showed that genes associated with plastid chromosome, anaerobic respiration, photosynthesis, and trehalose biosynthetic process were significantly upregulated. In contrast, DEGs related to cell wall, plant-type secondary cell wall biogenesis, structural constituent of ribosome, and glycolytic process were downregulated (Figure 5A), which may contribute to enhanced salt tolerance in G. uralensis.

KEGG pathway enrichment analysis indicated that DEGs in all comparison groups were significantly enriched in the sesquiterpenoid and triterpenoid biosynthesis pathway. In addition, DEGs in CK vs S were enriched in multiple pathways, including fatty acid biosynthesis, nitrogen metabolism, phenylpropanoid biosynthesis, and flavonoid biosynthesis (Figure 5B). In the CK vs S + ZnO NPs comparison, enriched pathways included glycine, serine and threonine metabolism, isoflavonoid biosynthesis, citrate cycle (TCA cycle), and starch and sucrose metabolism (Figure 5C).

Weighted gene co-expression network analysis (WGCNA) identified nine distinct gene modules (Supplementary Figure 3). Strong correlations were observed between specific modules and treatments. As shown in Figures 5D, the blue module was positively correlated with the S + ZnO NPs treatment, the green and yellow modules showed positive associations with salt-treated samples, and the red module was positively correlated with ZnO NPs treatment alone. KEGG analysis of genes in the blue module revealed significant enrichment in plant hormone signal transduction, glycolysis/gluconeogenesis, amino acid metabolism, flavonoid biosynthesis, phenylpropanoid biosynthesis, and carbohydrate metabolism (Supplementary Table 3). The upregulation of these metabolism-related genes is consistent with the metabolomic findings and suggests that ZnO NPs alleviate salt stress in G. uralensis roots by modulating associated metabolic pathways.

To further narrow down the associated pathways involved in the response to salt stress and to comprehensively elucidate the mitigating effects induced by ZnO NPs, an integrated analysis of transcriptomic and metabolomic data was performed. The results showed that significantly enriched pathways at both the transcript and metabolite levels were consistently detected only in the CK vs S, CK vs S + ZnO NPs, and S vs S + ZnO NPs comparisons, whereas relatively few shared pathways were identified in the CK vs CK + ZnO NPs comparison (Supplementary Figure 3). These findings indicate that transcripts and their associated metabolites exhibited similar functional responses and expression trends under different treatments. Ultimately, the phenylpropanoid biosynthesis and flavonoid biosynthesis pathways were identified as the major common pathways associated with both differentially expressed genes (DEGs) and differentially accumulated metabolites (DAMs).

As shown in Figure 6, a total of 16 DEGs and 8 DAMs involved in the phenylpropanoid and flavonoid biosynthesis pathways were identified. The concentrations of five DAMs—4’,7-dihydroxyflavone, sinapic acid, kaempferol, eleutheroside, and 5-caffeoylshikimic acid—were increased in the S, CK + ZnO NPs, and S + ZnO NPs groups. In contrast, the levels of kaempferol, 5,7-dihydroxyflavanone, and isoscutellarein were higher in the CK group, whereas trihydroxychalcone and 7-dihydroxyflavanone were downregulated under salt treatment but significantly upregulated in the S + ZnO NPs group.

At the transcriptional level, the expression levels of PAL, HCT, SCE, CCR, and E5.5.1.6 were downregulated in the S + ZnO NPs group compared with the control, whereas 4CL, F5H, UGT72E, CYP75A, CYP75B, and FLS were significantly upregulated, indicating that ZnO NPs treatment was associated with transcriptional reprogramming consistent with enhanced flavonoid biosynthesis under salt stress.