Transcriptome-based identification of the WRKY gene family in Polygala tenuifolia Willd. and their response to cold stress

Abstract

Background

Polygala tenuifolia Willd. is an important traditional medicinal plant widely used in prescriptions for cognitive enhancement. Cold stress severely affects seedling growth and medicinal quality in P. tenuifolia; however, the molecular mechanisms underlying cold tolerance remain poorly understood. WRKY transcription factors (TFs) are key regulators of abiotic stress responses in plants, but their role in cold stress adaptation of P. tenuifolia has not been systematically explored.

Results

Cold stress caused pronounced but reversible growth inhibition in P. tenuifolia seedlings and was accompanied by dynamic physiological changes. During early cold exposure, proline content, chlorophyll a and b levels, and antioxidant enzyme activities increased, while membrane lipid peroxidation was effectively suppressed, indicating an active stress mitigation response. Transcriptome analysis revealed significant enrichment of genes involved in photosynthesis, oxidative stress response, secondary metabolite biosynthesis, and plant hormone signaling pathways. A total of 104 cold-responsive WRKY TF genes (PtWRKYs) were identified from transcriptome data. Promoter analysis showed that many PtWRKY genes contained low-temperature and abscisic acid-responsive cis-acting elements. Expression profiling further demonstrated marked diversity in cold responsiveness among PtWRKYs. Notably, Pt13G01037 and Pt17G01891 showed strong induction under cold stress and distinct tissue-specific expression patterns, suggesting their potential roles in regulating cold tolerance.

Conclusions

This study reveals coordinated physiological adaptation and WRKY-mediated transcriptional regulation in P. tenuifolia under cold stress. The identification of cold-responsive PtWRKY genes, particularly Pt13G01037 and Pt17G01891, provides promising candidate regulators for future functional studies aimed at improving cold tolerance and stabilizing medicinal traits in P. tenuifolia.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12870-026-08377-1.

Background

Polygala tenuifolia Willd. is a perennial herb that possesses high medicinal value and was first documented in Shen Nong’s Herbal Classic, the earliest pharmaceutical monograph in China. This species ranks among the top three single herbs used clinically in traditional Chinese nootropic prescriptions [1], and is known for its therapeutic effects, including calming the mind, enhancing cognitive function, resolving phlegm, and reducing swelling. Clinically, it is used to treat insomnia, excessive dreaming, forgetfulness, anxiety, trance-like states, ineffective coughing with retained sputum, ulcers, and swellings, as well as breast distension associated with heart-kidney disharmony [1]. Phytochemical studies have shown that the major active constituents of P. tenuifolia include saponins, xanthones, and oligosaccharide esters. Among these, saponins exhibit notable protective effects on the central nervous system, such as anti-neuroinflammatory activity, inhibition of neuronal apoptosis, and promotion of neuronal proliferation [2–4]. Due to the sharp decline, wild P. tenuifolia has been designated as a third-class protected medicinal species in China. Increasing clinical demand has therefore driven large-scale cultivation of artificial materials, particularly in Shanxi and Shaanxi provinces [4].

However, cold stress, especially freezing injury is a major abiotic constraint in these primary production regions of P. tenuifolia, with low-temperature conditions prevailing at the seedling stage. This abiotic stress not only severely impairs P. tenuifolia’s growth, development, and survival, but also significantly reduces the yield and quality of medicinally active components such as saponins and flavonoids. Cold stress exerts distinct physiological effects on plants: exposure to low temperatures (0–12℃) suppresses growth and development, while prolonged freezing below 0℃ disrupts membrane integrity and causes cellular damage or death [5]. To survive under such conditions, P. tenuifolia undergoes multiple physiological and biochemical adaptations, which are largely driven by altered transcription of various genes, with physiological indicators serving as direct metrics for its stress responses. To simulate field-relevant cold stress conditions, seedlings were exposed to a constant 4℃, representing sustained low-temperature stress rather than extreme freezing. In major cultivation regions of P. tenuifolia, such as northern Shaanxi, seedlings frequently encounter prolonged low temperatures during early autumn prior to subzero freezing events. Climatic background data from the main cultivation region are provided in Table S1. Preliminary experiments indicated that 4℃ effectively induced stable physiological and transcriptional responses without causing rapid seedling mortality, allowing assessment of both stress response and acclimation processes. Throughout the cold stress treatment, environmental conditions other than temperature were strictly controlled. Photoperiod, light intensity, and relative humidity were kept constant to minimize the confounding effects, ensuring that temperature was the primary variable influencing stress responses.

WRKY proteins constitute one of the largest transcription factor (TF) families in plants. They are characterized by a highly conserved WRKYGQK motif of approximately 60 amino acids at the N-terminus and a zinc finger motif at the C-terminus [6]. In addition to the canonical WRKYGQK sequence, several motif variants, including WRRY, WSKY, WKRY, WVKY, and WKKY, have been reported. Based on domain architecture, WRKY proteins are classified into three groups: Group I contains two WRKY domains, whereas Groups II and III each possess a single WRKY domain. The zinc finger motifs are further divided into two types [7, 8]: C2-H2 (C-X_4–5-C-X_22–23-H–X-H) and C2-HC (C-X₇-C-X₂₃-H- X-C). Members of Groups I and II possess the C2-H2 motif, while Group III proteins contain the C2-HC motif. A small number of atypical WRKY proteins are occasionally categorized as Group IV [7]. The structural diversity of zinc finger motifs in WRKY proteins is considered to play an important role in plant evolution. This classification can be further refined into subgroups based on phylogenetic relationships, conserved domain features, and intron distribution patterns [8]. WRKY TFs are key regulators of plant growth and development, participating in processes such as seed germination, flowering, fruit ripening, and coloration [9], glucose metabolism and photosynthesis [10], secondary metabolites biosynthesis [11], and responses to biotic and abiotic stresses [12], including drought, temperature extremes, salinity, and mechanical injury. Genes involved in sugar and proline biosynthesis are commonly upregulated during cold stress, whereas genes related to photosynthesis are often downregulated [13, 14]. Many WRKY genes facilitate stress adaptation by regulating these physiological processes; similarly, under other stresses such as cold, WRKY and related TFs modulate plant stress resistance via distinct, specific mechanisms.

WRKY TFs regulate gene expression by specifically recognizing and binding to W-box elements (TTGACC/T) [11, 15] within the promoter of target genes. In cold stress responses, a typical example is the cucumber WRKY gene CsWRKY46, whose expression is induced by cold stress and exogenous abscisic acid (ABA). CsWRKY46 protein localizes to the nucleus and binds to the W-box element in the promoter of the ABI5 gene. Transgenic Arabidopsis thaliana overexpressing CsWRKY46 exhibit enhanced freezing tolerance, characterized by increased seedling survival rates, elevated proline accumulation, reduced electrolyte leakage, and decreased malondialdehyde (MDA) levels, as well as hypersensitivity to ABA during seed germination [16]. In terms of glucose metabolism regulation, heterologous overexpression of Camellia sinensis L. CsWRKY29 in A. thaliana increases soluble sugar content under both normal and cold conditions. It induces glucose-sensitive phenotypes, such as shortened primary roots, on glucose-supplemented media [17]. In the regulation of photosynthetic adaptation, RNA interference (RNAi)-mediated silencing of PsnWRKY70 enhances the expression of core photosystem genes under salt stress in Populus simonii × P. nigra [18].

In addition to the aforementioned regulatory functions in cold stress responses, glucose metabolism, and photosynthesis, WRKY and related TFs also play important roles in adapting to other stresses. MusaWRKY18 expression is upregulated in leaves of different banana cultivars under various stress conditions, indicating its role in regulating stress responses. Meanwhile, MusaSNAC1 can trigger stomatal closure by increasing H₂O₂ levels in guard cells. Overexpression of MusaSNAC1 in banana plants increases the proportion of closed stomata and reduces water loss, thereby enhancing drought resistance. The significant upregulation of β-glucuronidase (GUS) activity driven by the MusaSNAC1 promoter in guard cells under drought stress further confirms this function [19, 20].

Furthermore, WRKY genes can act as either positive or negative regulators of pathways [5]. For example, AtWRKY38 and AtWRKY62 negatively regulate defense responses against pathogenic bacteria [21], while AtWRKY53 negatively affects drought tolerance by regulating stomatal movements [22]. In A. thaliana, mitogen-activated protein kinase 4 (MPK4) interacts with MKS1 transition zone complex subunit 1 (MKS1) and phosphorylates WRKY25 and WRKY33 in vitro. The WRKY33 mutant exhibits increased PR1 expression under short-day conditions, heightened susceptibility to necrotrophic fungal pathogens, and impaired jasmonic acid (JA)/ethylene (ET)-regulated defense gene expression, demonstrating that WRKY33 functions as a positive regulator in the JA/ET signaling pathways essential for resistance to necrotrophic fungi [21].

Most previous studies on P. tenuifolia have focused on its phytochemical composition, pharmacological activities, and cultivation techniques, whereas few have investigated its adaptation to abiotic stresses. Although functional and physiological studies of WRKY TFs have been reported in multiple medicinal plants, including Forsythia [10], Platycodon grandiflorus [12], Dendrobium officinale [23], and Andrographis paniculata [24]. To address this knowledge gap, the present study leveraged RNA-sequencing (RNA-seq) transcriptome data of P. tenuifolia seedlings under cold stress to systematically characterize 104 WRKY TFs (PtWRKYs) with respect to gene structure, cis-acting regulatory elements, chromosomal localization, and subcellular distribution. Furthermore, we validated the expression profiles of PtWRKY genes exhibiting distinct cold-responsive patterns via quantitative real-time PCR (qRT-PCR), and identified two key cold-responsive candidates (Pt13G01037 and Pt17G01891). Collectively, a systematic identification and functional characterization of WRKY TFs under cold stress is necessary to clarify their role in cold adaptation and to provide a molecular basis for improving stress resistance and medicinal trait stability in this species.

Materials and methods

Plant materials and cold stress treatments

Polygala tenuifolia seeds were obtained from Shanxi Deji Pharmaceutical Co., Ltd. and taxonomically identified by Professor Hong Zhang. Seeds were surface-sterilized, thoroughly rinsed with sterile water, and sown in a soil–perlite–vermiculite mixture (1:1:1, v/v/v). Seedlings were grown at 25℃ under a 16 h light/8 h dark photoperiod with light intensity of 12,000 lx and relative humidity of approximately 65%. For cold stress treatment, 40-day-old seedlings were transferred to a growth chamber maintained at 4℃, while photoperiod, light intensity, and relative humidity were maintained at levels identical to those in normal growth conditions. Above-ground tissue was harvested at 0, 1, 3, 5, and 7 days after cold treatment, designated CS0, CS1, CS3, CS5, and CS7, respectively. Seedlings maintained at 25℃ under the same environmental conditions served as controls, and samples were collected at the corresponding time points (CK0, CK1, CK3, CK5, and CK7). All harvested tissue was immediately frozen in liquid nitrogen and stored at −80℃ for subsequent physiological and molecular analyses.

Phenotypic and physiological measurements

Fresh weight of above-ground seedling tissues was recorded, and seedling height was randomly measured using ImageJ software (version 2.5.5) to calculate average values. To evaluate the effects of cold stress on P.tenuifolia seedlings, six physiological parameters were analyzed. Proline content was determined using the acidic ninhydrin colorimetric method. Chlorophyll a and chlorophyll b contents were measured following the Plant Physiology Experiment Guide edited by Li et al. (5th Edition, 2016, pp. 50–52).

Antioxidant-related parameters were quantified using commercial assay kits, including superoxide dismutase (SOD; Total SOD Assay Kit with NBT method, Beyotime, S0109), peroxidase (POD; POD Assay Kit, Nanjing Jiancheng Bioengineering Institute, A084-3–1), catalase (CAT; Catalase Assay Kit, Beyotime, S0051), and MDA (MDA Assay Kit with TBA method, Nanjing Jiancheng Bioengineering Institute, A003-1). All physiological measurements were performed using three independent biological replicates, with each biological replicate analyzed in technical triplicate. Statistical analysis and graphical visualization were performed using Origin 2021.

RNA extraction, reverse transcription, and quantitative real-time PCR

Above-ground seedling tissues were ground into a fine powder in liquid nitrogen. Total RNA was extracted using RNAiso Plus (TaKaRa, 9109), according to the manufacturer’s instructions. RNA quality and concentration were assessed by 1.2% agarose gel electrophoresis and a Multiskan SkyHigh Full-Wavelength Microplate Reader (Thermo Fisher Scientific). Only RNA samples with an A260/A280 ratio between 1.7 and 2.1 were used for downstream analysis.

Reverse transcription and genomic DNA removal were performed using the TransScript® All-in-One First-Strand cDNA Synthesis SuperMix for qPCR (One-Step gDNA Removal, AT341, TRAN, China). qRT-PCR was carried out on a Bio-Rad CFX96 Real-Time PCR system (Bio-Rad, USA) following the manufacturer's protocol. The GAPDH gene was used as an internal reference [25]. Each sample was analyzed using three biological replicates, with each biological replicate run in technical triplicate. Relative gene expression was calculated by the 2^−ΔΔCT method [26]. Primers used for qRT-PCR analysis are listed in Table S2.

Transcript abundance estimation and differential expression analysis

For each treatment and control group, three biological replicates were prepared for RNA-sequencing. Sequencing libraries were constructed using a standard Illumina-compatible RNA library preparation kit according to the manufacturer’s protocol, including mRNA enrichment, fragmentation, first- and second-strand cDNA synthesis, end repair, adaptor ligation, and PCR amplification. Library quality and insert size distribution were evaluated prior to sequencing. The qualified libraries were sequenced on an Illumina sequencing platform to generate paired-end reads. Raw sequencing reads were processed to remove adaptor sequences, poly-N reads, and low-quality reads using standard quality control procedures. Clean reads were used for downstream transcriptome analysis. Gene expression levels were quantified as fragments per kilobase of transcript per million mapped reads (FPKM).

Principal component analysis (PCA) was performed to evaluate correlations and overall variation among samples subjected to cold stress. Differentially expressed genes (DEGs) were identified using the DESseq2 package. Genes with an absolute |log₂ (fold change)|> 1 and an adjusted p-value (padj) < 0.05 were considered significantly differentially expressed. The padj values were obtained after multiple testing correction using the Benjamini–Hochberg false discovery rate (FDR) method.

Gene Ontology (GO) enrichment analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis were conducted to identify significantly enriched functional categories and metabolic pathways associated with the DEGs.

Identification and structural analysis of the PtWRKY gene family

PtWRKY genes were identified using a combination of domain-based and sequence-similarity approaches. Hidden Markov models (HMM) profiles of the WRKY domain (PF03106) were obtained from the InterPro database (http://www.ebi.ac.uk/interpro/) [27, 28] and were used for initial screening. TBtools software (v2.149) [29] was used to search for candidate PtWRKY genes with an e-value threshold of 1 × 10⁻⁵ and to extract corresponding sequences. All candidates were further validated using the NCBI Conserved Domain Database (CDD, https://www.ncbi.nlm.nih.gov/cdd/) was used for further validation.

In total, 104 PtWRKY genes were identified using transcriptomic data, and detailed information is provided in Table S3. Based on genomic annotation and characteristic sequences features from the P. tenuifolia genome, gene IDs for 104 PtWRKYs were confirmed, and their protein sequences were downloaded from the NCBI database. WRKY protein sequences from A. thaliana (AtWRKYs) were obtained from TAIR (https://www.arabidopsis.org/) for comparative analysis.

Physiochemical properties of PtWRKY proteins, including amino acid number, molecular weight, theoretical pI, aliphatic index, grand average of hydropathicity (GRAVY), and instability index, were analyzed using TBtools. Subcellular localization was predicted using WoLF PSORT (https://wolfpsort.hgc.jp/). Based on phylogenetic relationships with AtWRKY proteins, conserved motif composition, and gene structure characteristics, each PtWRKY gene was assigned to a specific group and subgroups. Subgroup classification information was included in Table S4 to facilitate data interpretation.

Evolutionary relationships, motif analysis, domain identification, and gene structure analysis of PtWRKYs

To elucidate the evolutionary relationships between WRKY proteins of P. tenuifolia and A. thaliana, a phylogenetic tree was constructed using TBtools software based on the maximum likelihood (ML) method. The bootstrap value was set to 5000 replicates, and all other parameters were kept at their default settings.

WRKY protein sequences from A. thaliana were included as references. Conserved motifs within the PtWRKY protein sequences were identified using the online MEME tool (https://meme-suite.org/meme/doc/meme.html) [30]. A maximum of ten motifs was predicted using default parameters. On NCBI (https://www.ncbi.nlm.nih.gov/Structure/bwrpsb/bwrpsb.cgi), access “Batch CD-Search” under “Conserved Domain Database (CDD)” in “Domains & Structures”, upload PtWRKYs protein sequences with default parameters to obtain the family's domain and structure file. Finally, the phylogenetic tree, the distributions of conserved motifs, WRKY domains, and gene structures were integrated and visualized using TBtools to facilitate comparative analysis.

Chromosomal localization and duplication event analysis

The chromosomal distribution of PtWRKY genes was visualized by the MG2C online tool (http://mg2c.iask.in/mg2c%5Fv2.1/) [31]. The physical positions of PtWRKY genes on the P. tenuifolia chromosome were mapped, and genes were named sequentially according to their chromosomal locations. Intra-species collinearity analyses among PtWRKY genes and inter-species collinearity analyses between PtWRKYs and AtWRKYs genes were performed to identify gene duplication events contributing to the expansion of the PtWRKY gene family.

Gene duplication events within the PtWRKY gene family were classified as tandem or segmental duplications based on chromosomal location, sequence similarity, and collinearity relationships. Tandem duplication events were defined as homologous gene pairs located on the same chromosome, separated by ≤ 100 kb and with no intervening functional genes. These gene pairs were further required to exhibit a CDS sequence identity of at least 80% and sequence coverage ≥ 80%. Segmental duplication events were defined as homologous gene pairs located on different chromosomes with 5 BLAST hits and confirmation via collinearity block analysis using MCScanX software to exclude other duplication mechanisms.

Analysis of cis-elements in the PtWRKYs promoter

For PtWRKYs identified based on transcriptome data, promoter sequences (2,000 bp upstream of the translation start codon ATG) of 104 PtWRKY genes were retrieved from the P. tenuifolia genome database [2]. These sequences were then analyzed using the PlantCARE database (https://bioinformatics.psb.ugent.be/webtools/plantcare/html/) [32] to predict cis-acting regulatory elements, followed by classification and visualization of the identified elements via TBtools. Then, cis-acting elements were classified according to the functional annotation file, followed by the generation of bar charts and bubble charts using WPS Office (12.1.0.22215) and the heatmap module integrated in TBtools, respectively.

Tissue-specific expression analysis of Pt13G01037 and Pt17G01891

To investigate the tissue-specific expression patterns of Pt13G01037 and Pt17G01891 under normal growth conditions, samples were collected at two distinct developmental stages. Seedlings at the vegetative growth stage (approximately 40 days old) were cultivated under controlled conditions (25℃, 16 h light/8h dark photoperiod, light intensity of 12,000 lx, and approximately 65% relative humidity) without cold stress. Root, stem, and leaf tissues were harvested. Mature plants at the harvest stage (August 2024) were grown under natural field conditions, during which the average maximum temperature and mean light duration were 27.4℃ and 10.4h (Table S1), respectively. Roots, stems, leaves, flower buds, early flowering tissue, and full flowering tissue were collected.

In total, nine tissue types were sampled, immediately frozen in liquid nitrogen, and used for qRT-PCR analysis, each sample was analyzed using three biological replicates. Relative expression levels were calculated using the 2^−ΔΔCT method.

Results

Phenotypic and physiological responses to cold stress

Compared with the control group, P. tenuifolia seedlings exhibited pronounced phenotypic changes after CS1, including visible wilting of the two cotyledons and a significant reduction in the fresh weight of the above-ground tissue. With prolonged cold exposure, seedlings gradually recovered from wilting, and normal growth was largely restored after CS7, indicating the presence of an adaptive response to cold stress. Physiological measurements showed that chlorophyll a, chlorophyll b, and proline contents increased significantly during the early phase of the cold stress (0–1 days, Fig. 1). In contrast, other physiological indicators displayed distinct response patterns (Figure S1). MDA content increased significantly under cold stress, reflecting enhanced membrane lipid peroxidation and indicating oxidative damage to the cellular membrane, while MDA levels were moderated at later stages, coinciding with substantial proline accumulation. As a known osmoprotectant and membrane stabilizer, proline likely alleviates membrane damage by maintaining cellular osmotic balance and protecting membrane integrity.

Fig. 1.

Phenotypic and physiological responses of Polygala tenuifolia seedlings under cold stress (4℃). A Representative phenotypes of control and cold-stressed seedlings. Close-up views of cotyledons are shown to highlight cold-induced wilting symptoms. Scale bar = 1 cm. B-E Changes in chlorophyll a (B), chlorophyll b (C), proline (D), and malondialdehyde (MDA) content (E) under cold stress. Data are presented as mean ± SE (n = 3). Different lowercase letters indicate significant differences among treatments at P < 0.05

In addition to proline accumulation, antioxidant-related parameters showed coordinated, biologically meaningful responses. SOD activity showed a transient increase during the early stage of cold stress, indicating rapid activation of the antioxidant defense system to scavenge excessive reactive oxygen species (ROS). However, prolonged cold exposure led to a gradual decline in SOD activity, possibly due to stress-induced impairment of enzyme structure or catalytic efficiency. Notably, this decline was accompanied by relatively stable or slightly enhanced activity of POD and CAT, whose activities did not show significant changes compared with the controls, suggesting a compensatory interaction among these antioxidant enzymes. The coordinated action of POD and CAT may partially offset the reduced SOD activity and contribute to the maintenance of ROS homeostasis under sustained cold stress.

Collectively, these findings indicate that the period from 0 to 1 day represents the initial stress response phase, while the period from 3 to 7 days corresponds to a physiological adaptation phase. During this process, P. tenuifolia seedlings primarily mitigate cold stress by enhancing proline accumulation. These results further demonstrate that P. tenuifolia seedlings respond to cold stress through an integrated physiological strategy involving coordinated antioxidant enzyme activities, and the dynamic balance among ROS scavenging, membrane protection, and osmolyte accumulation underlies cold adaptation in this species.

Transcriptome analysis of PtWRKYs under cold stress

The reliability of the transcriptome sequencing data of P. tenuifolia under cold stress was first evaluated (Table S4), confirming high sequencing quality and strong data reliability. PCA based relationships among samples. Pearson correlation coefficients were calculated to evaluate reproducibility and exclude outliers. As shown in Fig. 2A, the retained samples displayed distinct expression patterns and were clearly separated, with PC1 and PC2 explaining 64.22% and 11.78% of the total variation, respectively. Consistently, the correlation heat map (Figure S2) showed strong correlations among samples, with a correlation coefficient exceeding 0.7, indicating excellent reproducibility.

Fig. 2.

Analysis of the Transcriptome of P. tenuifolia seedlings under cold stress. A PCA score plot of PtWRKY gene expression profiles in P. tenuifolia seedlings under cold stress. PC1 and PC2 accounted for 64.22% and 11.78% of the total variance. Each data point represents three biological replicate. Cold-treated samples were distinctly separated from the control group along the PC1 axis. B Venn diagram of DEGs between cold-treated (CS) and control (CK) groups in P. tenuifolia seedlings. A total of 2,932 unique DEGs were identified across all comparisons. C Volcano diagram of DEGs between cold-treated (CS1) and control (CK1) groups in P. tenuifolia seedlings. A total of 4,081 genes were upregulated and 3,558 genes were downregulated. D Volcano diagram of DEGs between cold-treated (CS3) and control (CK3) groups in P. tenuifolia seedlings. A total of 5,213 genes were upregulated and 6,382 genes were downregulated. E Volcano diagram of DEGs between cold-treated (CS5) and control (CK5) groups, 5,387 genes were upregulated and 6,953 genes were downregulated. F Volcano diagram of DEGs between cold-treated (CS7) and control (CK7) groups, 4,868 genes were upregulated and 6,586 genes were downregulated

DEGs were identified through pairwise comparisons between cold stress treatments and their corresponding control. As illustrated in the Venn diagram (Fig. 2B), the number of DEGs was 1,766 for CS1 vs. CK1, 1,570 for CS3 vs. CK3, 1,611 for CS5 vs. CK5, and 1,846 for CS7 vs. CK7, with a total of 2,932 unique DEGs across all comparisons. Further analyses revealed 7,639 DEGs (4,081 upregulated and 3,558 downregulated) in CS1 vs. CK1, 11,595 DEGs (5,213 upregulated and 6,382 downregulated) in CS3 vs. CK3, 12,790 DEGs (5,837 upregulated and 6,953 downregulated) in CS5 vs. CK5, and 11,454 DEGs (4,868 upregulated and 6,586 downregulated) in CS7 vs. CK7 (Fig. 2C-F; Table S5). These results indicate a progressive and dynamic transcriptional response to prolonged cold stress.

To explore the biological functions of DEGs, GO enrichment analysis was conducted. Enriched GO terms were classified into biological processes, cellular components, and molecular function categories, with ten significant enriched terms identified in each category. Prominent molecular functions included tetrapyrrole binding (GO:0046906), heme binding (GO:0020037), iron ion binding (GO:0005506), oxidoreductase activity (GO:0016705), and hydrolase activity hydrolyzing O-glycosyl compounds (GO:0004553). Primary cellular components included thylakoids (GO:0009579, GO:0044436), photosystems (GO:0034357, GO:0009521), and cell walls (GO:0005618). Notable biological processes were photosynthesis (GO:0015979), response to oxidative stress (GO:0006979), and protein ubiquitination (GO:0016567). Notably, DEGs in CS1 vs. CK1 were mainly enriched in molecular functions, whereas those in CS3 vs. CK3 were predominantly associated with cellular components. In contrast, CS5 vs. CK5 and CS7 vs. CK7 showed enrichment in both cellular components and molecular functions categories (Fig. 3; Table S6).

Fig. 3.

GO enrichment analysis of P. tenuifolia seedlings under cold stress. The top five significantly enriched GO terms in each comparison further highlighted stage- specific responses to cold stress. A Early cold exposure (CS1 vs. CK1) was characterized by enrichment of binding and redo related functions, such as tetrapyrrole binding (GO:0046906, padj = 3.7873e⁻¹¹), heme binding (GO:0020037, padj = 3.7873e⁻¹¹), iron ion binding (GO:0005506, padj = 1.3532e⁻⁹), oxidoreductase activity, acting on paired donors, with incorporation (GO:0016705, padj = 5.8173e⁻⁹), and transferase activity, transferring glycosyl groups (GO:0016757, padj = 4.8790e⁻⁷); B For the CS3 vs. CK3 group, they were thylakoid (GO:0009579, padj = 1.9724e⁻¹⁰), photosynthetic membrane (GO:0034357, padj = 1.9724e⁻¹⁰), thylakoid part (GO:0044436, padj = 1.9724e⁻¹⁰), photosystem (GO:0009521, padj = 9.5943e⁻¹⁰), and photosynthesis (GO:0015979, padj = 4.5727e⁻⁶). C For the CS5 vs. CK5 group, they were photosystem (GO:0009521, padj = 7.0217e⁻⁶), thylakoid (GO:0009579, padj = 8.6148e⁻⁶), photosynthetic membrane (GO:0034357, padj = 8.6148e⁻⁶), thylakoid part (GO:0044436, padj = 8.6148e⁻⁶), and extracellular region (GO:0005576, padj = 4.5200e⁻⁵). D For the CS7 vs. CK7 group, they were oxidoreductase activity, acting on paired donors, with incorporation (GO:0016705, padj = 4.3918e⁻⁶), hydrolase activity, hydrolyzing O-glycosyl compounds (GO:0004553, padj = 2.3699e⁻⁵), heme binding (GO:0020037, padj = 4.2283e⁻⁵), tetrapyrrole binding (GO:0046906, padj = 4.2283e⁻⁵), and hydrolase activity, acting on glycosyl bonds (GO:0016798, padj = 4.2822e.⁻⁵)

KEGG pathway enrichment analysis identified the top enriched pathways for each comparison group (Fig. 4). The major pathways included biosynthesis of secondary metabolites, plant hormone signal transduction, starch and sucrose metabolism, and glutathione metabolism. Notably, phenylpropanoid biosynthesis was enriched in all comparisons except CS3 vs. CK3, whereas carbon metabolism showed high enrichment in CS3 vs. CK3 and CS7 vs. CK7. Furthermore, the MAPK signaling pathway-plant was significantly enriched only in CS1 vs. CK1. In summary, numerous genes associated with these pathways contributed substantially to cold resistance in P. tenuifolia. The data used are presented in Table S7.

Fig. 4.

KEGG enrichment analysis of P. tenuifolia seedlings under cold stress. The top 5 significantly enriched KEGG pathways in each comparison group as follows. A For the CS1 vs. CK1 group, these were Phenylpropanoid biosynthesis (ath00940, padj = 3.5314e⁻⁸), MAPK signaling pathway—plant (ath04016, padj = 8.5462e⁻⁶), Biosynthesis of various plant secondary metabolites (ath00999, padj = 4.2753e⁻⁵), Plant hormone signal transduction (ath04075, padj = 0.0005), and Plant-pathogen interaction (ath04626, padj = 0.0015). B For the CS3 vs. CK3 group, they were Photosynthesis (ath00195,padj = 7.7352e⁻⁵), Starch and sucrose metabolism (ath00500, 5.7068e⁻⁵), Photosynthesis-antenna proteins(ath00196, padj = 0.0002), Biosynthesis of various plant secondary metabolites (ath00999, padj = 0.0011), Plant hormone signal transduction (ath04075, padj = 0.0038). C For the CS5 vs. CK5 group, they were Phenylpropanoid biosynthesis (ath00940, padj = 0.0001), Starch and sucrose metabolism (ath00500, padj = 0.0003), Plant hormone signal transduction (ath04075, padj = 0.0003), Photosynthesis-antenna proteins (ath00196, padj = 0.0007) and Tryptophan metabolism (ath00380, padj = 0.0038). D For the CS7 vs. CK7 group, they were Starch and sucrose metabolism (ath00500, padj = 1.2718e.⁻⁵), Linoleic acid metabolism (ath00591, padj = 0.0004), Phenylpropanoid biosynthesis (ath00940, padj = 0.0004), Biosynthesis of various plant secondary metabolites (ath00999, padj = 0.0004) and Steroid biosynthesis (ath00100, padj = 0.0004)

For a more in-depth dissection of the molecular regulatory mechanisms underlying the cold stress-responsive pathways described above, we performed a systematic annotation analysis of DEGs derived from transcriptome data under cold stress. The results showed that members of transcription factor families associated with abiotic stress tolerance were significantly enriched among the DEGs; among these, the PtWRKY gene family ranked among the top five in terms of the number of DEGs and accounted for a relatively high proportion of stress-responsive DEGs. As well-documented master regulators of plant cold stress responses, PtWRKY proteins mediate cold stress adaptation by modulating the expression of downstream target genes involved in pathways including plant hormone signal transduction, phenylpropanoid biosynthesis, and the MAPK signaling pathway-all of which were highly enriched in our KEGG enrichment analysis (Fig. 4). On this basis, we focused subsequent analyses on the PtWRKY gene family to elucidate its specific regulatory role in the cold stress response of P. tenuifolia seedlings.

Identification and characterization of the PtWRKY gene family

Based on transcriptome data of P. tenuifolia seedlings exposed to cold stress, 104 highly expressed PtWRKY genes were identified and mapped to their respective chromosomal locations (Fig. 5A). These genes were unevenly distributed across 19 chromosomes, with no PtWRKY genes detected on chromosome 12. Chromosomes 1, 13, and 18 harbored the largest numbers of PtWRKY genes, containing 15, 11, and 10 genes, respectively. The two target genes, Pt13G01037 and Pt17G01891, were located on chromosomes 13 and 17, respectively. In parallel with the above analyses, eight pairs of homologous genes were also identified, namely Pt1G00645/Pt1G00648, Pt1G01386/Pt1G01388, Pt6G01096/Pt6G01097, Pt6G01097/Pt6G01102, Pt9G00390/Pt9G00392, Pt17G00574/Pt17G00575, Pt18G00931/Pt18G00932, and Pt19G00319/Pt19G00322.

Fig. 5.

Identification and characterization of the PtWRKY gene family. A Chromosome distribution of PtWRKY genes in P. tenuifolia. The scale on the left is indicated in mega-bases (Mb). A total of 104 PtWRKY genes were unevenly mapped onto 19 chromosomes, with no genes detected on chromosome 12; additionally, eight pairs of homologous PtWRKY genes were identified. B Interspecific homology analysis of of WRKY genes between P. tenuifolia and A. thaliana. A genome-wide comparison identified 84 homologous genomic regions. The highest number of homologous gene pairs was detected between chromosomes 1/13 of P. tenuifolia and chromosome 4 of A. thaliana, indicating partial conservation of genomic structure between the two species. C Phylogenetic tree of WRKY genes from P. tenuifolia and A. thaliana. Based on conserved N-terminal WRKYGQK motifs, C-terminal zinc-finger motifs, evolutionary relationships with AtWRKY proteins, conserved motif composition and gene structural characteristics, PtWRKY genes were classified into four major groups and further assigned to specific subgroups. D Conserved motifs, domains and gene structure diagram of PtWRKY genes. PtWRKY proteins contained 1–8 conserved motifs (motifs 1/2 most prevalent); 100 of 104 proteins had WRKY domains, while 4 showed superfamily characteristics. Intron numbers varied 0–5 in subgroup Ia-1; Group II had two intron patterns (4 in IIa-IId, 2 in IIe-IIf); Group III genes all had 2 introns

The predicted amino acid length, molecular weights, theoretical isoelectric points (pI), physicochemical properties, and subcellular localizations of each PtWRKY proteins are summarized in Table S8. The lengths of the PtWRKY proteins ranged from 97 amino acids (Pt10G00092) to 733 amino acids (Pt13G01826), with corresponding molecular weights ranging from 10,946.73 Da (Pt10G00092) to 79,051.13 Da (Pt13G01826). The theoretical pI values varied from 4.72 (Pt9G00392) to 10.69 (Pt1G02795). The aliphatic indices ranged from 40.71 (Pt13G02091) to 80.31 (Pt10G00092), while the grand average of hydropathicity ranged from −1.094 (Pt13G02091) to −0.378 (Pt15G01282), indicate that all PtWRKY proteins are hydrophilic. The instability indices ranged from 33.50 to 72.40, with seven PtWRKY proteins showing favorable stability: Pt1G02821 (33.50), Pt13G01372 (37.09), Pt3G02486 (38.28), Pt7G01399 (38.77), Pt11G01665 (38.97), Pt17G01907 (39.18), and Pt5G00252 (39.24). Subcellular localization predictions revealed that most PtWRKY proteins were localized to the nucleus (90 proteins), while others were predicted to localize to chloroplasts (5), cytoplasm (3), peroxisomes (3), plasma membrane (1), extracellular space (1), and vacuoles (1).

To investigate evolutionary relationships, an unrooted phylogenetic tree was constructed using the ML method, incorporating 104 PtWRKY proteins and AtWRKY proteins. Based on the conserved N-terminal WRKYGQK motifs and C-terminal zinc-finger motif, the PtWRKY genes were classified into four major groups [9, 33, 34]. Group I contains two WRKY domains with a C2-H2 motif, whereas Groups II and III each harbor a single WRKY domain, featuring C2-H2 and C2HC motifs, respectively. Group I was further subdivided into two subgroups: Ia, which contains a C2-H2 zinc finger motif (C-X₄-C-X_22–23-H-X₁-H), and Ib, which features a C2HC zinc finger motif (C-X₇-C-X₂₃-H-X₁-C). Group II was divided into six subgroups, with each defined by a specific conserved motif: IIa (TVA(A/P)(G/E)), IIb (TM(A/G)(V/A/T)(G/A)), IIc (TIA(S/P)S), IId (S(F/Y)AP (S/D/N)), IIe (S(S/T)(A/M/V)(R)/K(G/R)), and IIf (S(S/T)(S/A/M)/R(K)G). Group III was classified into two subgroups (IIIa and IIIb), while sequences lacking either the WRKY domains or zinc-finger motifs were assigned to Group IV (Fig. 5C). Evolutionary analyses revealed that some PtWRKY genes had undergone structural variations during evolution. For instance, Pt18G00931 likely diverged from Group Ia following the deletion of the WRKYGQK motif, while other genes underwent evolutionary divergence due to zinc-finger motif loss. Specifically, Pt15G01242 (Ia), Pt18G01652 (IIb), Pt6G00075 (IId), Pt10G00092 (IIe), and Pt10G01763 (IIe) lack the zinc-finger motifs, whereas Pt6G01097 exhibited mutations in both the WRKY and zinc-finger motifs.

Each PtWRKY member was assigned to a specific group and subgroup based on its evolutionary relationships with AtWRKY proteins, conserved motif composition, and gene structural characteristics. Although a small number of PtWRKY genes (such as Pt6G01097) lacked a complete WRKY domain or a typical zinc-finger motif, they were still retained as members of the PtWRKY gene family on the basis of multiple lines of evidence. First, phylogenetic analysis clustered these genes within well-supported WRKY clades rather than in distinct, independent clades, indicating that they share a close evolutionary relationship with bona fide WRKY members. Second, conserved motif analysis revealed the presence of partial WRKY-associated conserved motifs, suggesting that these genes may represent divergent or truncated WRKY variants generated during gene duplication and evolutionary divergence. Similar WRKY-like genes lacking complete domains have also been identified in other plant species and are generally recognized as evolutionarily derived members of the WRKY family [34]. The classification of the aforementioned genes was inferred based on their evolutionary relationships combined with the established classifications reported in previous literature [15, 34]. To facilitate data interpretation, the subgroup classification of each PtWRKY gene, as determined by phylogenetic analysis, was included in Table S3.

Intraspecific collinearity analysis was performed on the 104 PtWRKY genes (Figure S3A), with a total of 572 homologous regions detected, which laid a solid foundation for identifying the conserved core members of the WRKY gene family in P. tenuifolia. Among the 8 homologous gene pairs, only one pair of tandemly duplicated genes (Pt6G01096 and Pt6G01097) exhibited 99.63% sequence homology. Combined with the tandem duplication events identified, these results further indicated that segmental duplication and tandem duplication contributed significantly to the evolutionary expansion of the PtWRKY gene family, with segmental duplication acting as the predominant mechanism. For interspecific homology analysis, a genome-wide comparison was conducted between P. tenuifolia and A. thaliana, which identified 84 homologous regions (Fig. 5B). Among these regions, the highest number of homologous gene pairs was detected between chromosomes 1 and 13 of P. tenuifolia and chromosome 4 of A. thaliana. This observation demonstrated that the genomic structure of P. tenuifolia exhibits partial conservation relative to that of A. thaliana.

Analysis of conserved motifs, domains, gene structures, and promoter cis-acting elements showed strong consistency with phylogenetic classification. Most PtWRKY proteins contained between one and eight conserved motifs, with 1 and 2 being the most prevalent. Among the 104 proteins, 100 possessed conserved WRKY domains, whereas four proteins (Pt5G01242, Pt18G01652, Pt17G00574, Pt1G02525) exhibited characteristic features of the WRKY superfamily. Genes with similar motif compositions clustered together in the phylogenetic tree (Fig. 5D). A significant variation in intron number (0–5) was detected among PtWRKYs in subgroup Ia-1. Genes in Group II displayed two intron distribution patterns: most members of subgroups IIa-IId contained four introns, while subgroups IIe and IIf generally harbored two introns. All six PtWRKY genes in Group III contained two introns. These findings demonstrate that PtWRKYs with close evolutionary relationships typically share similar motif profiles and gene structures, implying evolutionary divergence during species evolution and contributing to functional diversity within the gene family.

Promoter analysis identified 62 cis-acting regulatory elements (Figure S3B), which were classified into four categories: hormone-responsive elements, growth-and development-related elements, stress and defense-related elements, and light-responsive elements. The abundance and diversity of these elements suggest that PtWRKY genes are subject to complex transcriptional regulation in response to developmental cues and environmental stresses.

To clarify the role of cis-acting elements in PtWRKY-mediated cold stress responses, promoter regions were systematically analyzed for low-temperature-related regulatory elements. These included low-temperature responsive elements (LTR), MYB-binding sites involved in stress regulation (MBS), and hormone-related elements. The number, type, and positional distribution of these cold-responsive cis-elements within the 2-kb promoter regions upstream of each PtWRKY gene were visualized in Figure S3B. Quantitative analysis revealed substantial variation in the abundance of cold-responsive cis-elements among PtWRKY genes. Some genes contained multiple LTR elements (≥ 2 copies) at different positions relative to the transcription start site, whereas others harbored only a single cold-related element or none at all. This heterogeneity suggests differential sensitivity of PtWRKY genes to low-temperature signals at the transcriptional level.

To further explore the relationship between promoter architecture and gene expression, cis-element data integrated with RNA-seq expression profiles. PtWRKY genes containing at least two LTR elements in their promoter regions, such as Pt11G01665 (highly expressed in CS1), Pt18G011921 (highly expressed in CS3), and Pt14G00989 (highly expressed in CS5), showed significant upregulation under cold stress, with FPKM values of 322.6907, 72.7039, and 130.2598 (Table S3), respectively. These results suggest a positive association between LTR copy number and cold-induced transcriptional activation of PtWRKY genes.

Genes harboring multiple types of cold-responsive and hormone-related cis-elements displayed faster and more sustained transcriptional induction than genes containing only a single cold-responsive element. For example, Pt13G01037, which contains abscisic acid-responsive elements (ABREs), CGTCA-motif, and gibberellin-responsive elements (GARE) in its promoter, showed high expression at the CS5 time point. Similarly, Pt17G01891, which contains MBS and TGACG-motif elements, also showed strong upregulation at CS5. These findings indicate that the synergistic action of multiple cis-acting elements enhances the responsiveness of PtWRKY genes to cold stress signals.

In contrast, PtWRKY genes lacking identifiable cold-stress-related cis-elements showed no significant transcriptional changes under cold stress. The clear correspondence between promoter cis-element and gene expression behavior supports a central role for cold-responsive cis-elements, particularly LTR, in mediating cold-induced transcriptional activation of PtWRKY genes. Collectively, these results establish a regulatory link between promoter architecture, signal perception, and transcriptional responsiveness, providing mechanistic evidence for the involvement of PtWRKY genes in cold stress adaptation in P. tenuifolia.

Expression pattern of PtWRKY genes under cold stresses

WRKY TFs are plant-specific regulators that play key roles in plant stress responses. Based on transcriptome profiling, 57 PtWRKY genes were identified as significantly differentially expressed under cold stress. Among these, 17 PtWRKY genes showed high expression in CK0 and were continuously downregulated under cold stress. Five genes were highly expressed specifically at CS1; 14 genes showed high expression in both CK0 and CS1; 9 genes were highly expressed at CS3; 7 at CS5; and 5 at CS7 (Table S3).

Genes that were significantly upregulated at each cold stress stage were selected for qRT-PCR validation. The expression trends of 15 differentially expressed PtWRKY genes were largely consistent between qRT-PCR and transcriptome expression profiles (Fig. 6, Figure S4). To further verify the reliability of the transcriptome data, correlation analysis was performed between the RNA-seq and qRT-PCR results for the selected PtWRKY genes. Gene FPKM values obtained from RNA-seq and the relative expression levels measured by qRT-PCR across all cold stress time points were visualized and compared using line charts and bar charts, respectively. As shown in Fig. 6, RNA-seq and qRT-PCR data showed strong concordance, confirming the reliability and accuracy of the transcriptome-based expression analysis. Based on qRT-PCR data, two genes (Pt13G01037 and Pt17G01891) with the highest expression among verified differential PtWRKYs were selected for further experiments. Both genes were significantly upregulated with prolonged cold stress, peaking at 5 days, and subsequently downregulated after 7 days. In addition, several candidate reference genes including Elongation1 [35], GAPDH [25], β-actin1 [36], β-actin2 [37], and Cdc42 [38] were evaluated (Figure S4B-F). Among them, GAPDH exhibited the lowest Ct value and the most stable expression and was therefore selected as the internal reference gene. Primers used in qRT-PCR analysis are listed in Table S2.

Fig. 6.

Identification and validation of cold stress-responsive PtWRKY genes via transcriptome profiling and qRT-PCR. Fifty-seven DEGs were classified into six expression patterns: highly expressed in CK0 and continuously downregulated under cold stress; specific to CS1; abundant in both CK0 and CS1; highly expressed in CS3, CS5, and CS7, respectively. Fifteen highly expressed DEGs were selected for qRT-PCR validation. (Left Y-axis: qRT-PCR relative expression; Right Y-axis: gene FPKM values)

Tissue-specific expression patterns of Pt13G01037 and Pt17G01891

To further investigate the potential roles of PtWRKY genes in response to cold stress, the expression of two candidate genes, Pt13G01037 and Pt17G01891, was analyzed by qRT-PCR in different tissues of seedlings (roots, stems, leaves) and mature plants (flower buds, early flowering tissue, full flowering tissue, root phloem, stems, leaves). The two genes showed distinct tissue-specific expression patterns (Fig. 7), with overall higher transcript levels in seedlings than in mature plant tissues. In seedlings, Pt17G01891 exhibited the highest expression in stems, whereas Pt13G01037 showed relatively lower but stable expression across tissues. In mature plants, both genes displayed strong root-preferential expression. Pt13G01037 reached its highest expression level in mature roots, while Pt17G01891 also showed peak expression in roots, with markedly lower expression in stems, leaves, and floral tissues. In silico subcellular localization analysis predicted that both proteins are localized to the nucleus (Table S8), consistent with their roles as transcription factors. The consistent enrichment of Pt13G01037 and Pt17G01891 transcripts in root tissues across developmental stages suggests that these genes may be involved in root-associated physiological processes and could play important regulatory roles in mediating cold stress adaptation in P. tenuifolia.

Fig. 7.

Tissue-specific expression analysis of PtWRKY genes in P. tenuifolia via qRT-PCR. A Seedlings (roots, stems, leaves) and mature plants (flower buds, early-flowering tissues, full-flowering tissues, root phloem, stems, leaves). B–C Expression patterns of Pt17G01891 and Pt13G01037: in seedlings, Pt17G01891 peaked in stems, while Pt13G01037 showed low but stable expression across tissues; in mature plants, both genes were root-preferential, with peak expression in roots and markedly lower levels in stems, leaves, and floral tissues

Discussion

Physiological responses of seedlings to cold stress and potential signaling pathways

Physiological indicators provide direct evidence of how plants respond to environmental stress. MDA is a terminal product of membrane lipid peroxidation and is widely used as an indicator of oxidative damage. Excessive MDA accumulation can accelerate lipid peroxidation, disrupt membrane integrity, and alter membrane permeability [39, 40], ultimately leading to cellular damage [41, 42]. Therefore, higher MDA levels generally reflect more severe stress-induced injury. Osmotic adjustment is another key mechanism underlying plant stress tolerance. Proline, a representative osmoprotectant, plays an essential role in maintaining cellular osmotic balance, stabilizing proteins and membranes, and scavenging ROS [43]. Under normal conditions, proline concentration remains low, but its rapid accumulation is a common response to abiotic stress and is positively associated with enhanced stress tolerance [44]. Previous studies have shown that transgenic A. thaliana overexpressing MaDREB14/22/51 exhibited enhanced tolerance to drought and cold stress by reducing MDA accumulation while increasing proline and soluble sugar content [45]. In addition, chlorophyll content is highly sensitive to stress. Damage to chloroplast structure under stress conditions often leads to chlorophyll degradation and impaired photosynthesis [46].

In the present study, P. tenuifolia seedlings exhibited clear physiological adjustment in response to cold stress. MDA content significantly increased on day 5 of cold treatment and declined by day 7, suggesting that membrane damage initially intensified but was partially alleviated during prolonged exposure. In contrast, proline and chlorophyll (a and b) contents increased from days 1 to 5, followed by a decreased at day 7. This coordinated pattern indicates a dynamic balance among membrane damage, osmotic regulation, and photosynthetic adjustment. Together,these results suggest that P. tenuifolia seedlings undergo a reversible stress response, maintaining viability through the activatinon of multiple protective mechanisms.

Physiological responses and transcriptional regulation are closely interconnected during plant adaptation to cold stress. Such as MusaNAC29 enhance stress resistance by modulating hormone balance, secondary metabolism, and stress-responsive gene expression, although often at the cost of reduced growth and yield [47]. By integrating physiological transcriptome profiling and pathway enrichment analyses, we further explored the molecular mechanisms lunderlying cold tolerance in P. tenuifolia. Correlation analysis revealed that the expression levels of the two candidate genes, Pt13G01037 and Pt17G01891, were strongly associated with key physiological indicators. Both genes showed significant positive correlations with proline accumulation and SOD activity, and significant negative correlations with MDA content. These relationships suggest that PtWRKY genes may contribute to cold tolerance by enhancing osmotic adjustment and antioxidant capacity while limiting membrane lipid peroxidation.

Further insight into downstream regulatory networks was obtained by examining cold-responsive pathways. Genes involved in tetrapyrrole and heme binding, and in photosystem processes, were significantly enriched under cold stress. Promoter analysis revealed that many of these genes contained conserved W-box elements (TTGACC/T), a typical binding site for WRKY TFs. These findings suggests that PtWRKY proteins may directly regulate photosynthesis-related genes during cold stress. In addition, differentially expressed genes were significantly enriched in plant hormone signal transduction and phenylpropanoid biosynthesis pathways, both of which are closely associated with abiotic stress responses.

Notably, the promoters of Pt13G01037 and Pt17G01891 harbor ABREs, indicating potential involvement in ABA-dependent cold signaling pathways. It is therefore plausible that cold stress activates ABA signaling, which in turn induces PtWRKY expression. The activated PtWRKY proteins may then regulate downstream genes associated with proline biosynthesis, antioxidant defense, and adjustments in photosynthesis, thereby reducing cellular damage under low-temperature condition Overall, our results support a PtWRKY-centered regulatory framework in which cold stress-induced WRKY TFs coordinate physiological adaptation by integrating osmotic regulation, ROS homeostasis, and hormone-mediated signaling pathways in P. tenuifolia.

The WRKY gene family is highly conserved in P. tenuifolia, and its expansion is primarily driven by fragment duplication

This study represents the first systematic identification of the WRKY gene family in P. tenuifolia. WRKY TFs have been extensively characterized in many plant species, including economic crops such as rice [48], soybean [49], and maize [50], fruits and vegetables like Vitis vinifera [44] and Saccharum spontaneum [7], medicinal plants such as Forsythia suspensa [10], Platycodon grandiflorus [12], Dendrobium officinale [23], and Andrographis paniculata [24]; as well as the model plant A. thaliana [33].

In the present study, 104 PtWRKY genes were identified from P. tenuifolia. Phylogenetic analysisclassified these genes into 12 subgroups: Ⅰa1, Ⅰa2, Ⅰb, Ⅱa, Ⅱb, Ⅱc, Ⅱd, Ⅱe, Ⅱf, Ⅲa, Ⅲb, and Ⅵ [9]. Although variations in the conserved WRKY heptapeptide motif have been reported in other species, such as cucumber (WRKYGKK), grapes (WRKYGKK, WKKYGQK), poplars (WRKYGKK, FWRKYGQK, WKKYGQK, WRKYGRK, WRKYGEK), and eggplant (WIKYGQK, WRKYGEK, WRKYGKK, WKKYEIG, WRKYGON, WHKFGQK, WDKFGQK) [51], most PtWRKY proteins contain a conserved WRKYGQK sequence. Several PtWRKY proteins exhibited heptapeptide variations, including WCKYGQK in Ia1, WRKYGKK in two genes in Ia2, FKVYCPK in subgroup IIe, and YPNYGQK in subgroup IIf. In addition, five PtWRKY proteins lacked a conserved and identifiable heptapeptide motif: Pt6G01097 and Pt18G00932 (Group VI), Pt17G00574 (Group VI), Pt18G00931 (Group Ia1), and Pt8G00938 (Group Ib). Mutations or losses of the conserved motif in these genes may lead to functional divergence. Moreover, WRKY function is also influenced by altered protein–protein interactions or variations in cis-regulatory elements beyond the W-box [51].

Several distinct hypotheses have been proposed to explain the evolutionary origin and diversification of WRKY TFs [8, 34]. One model emphasizes protein structural features, including the number of WRKY domains and zinc-finger motifs [7, 8]. while another focuses phylogenetic relationships based on conserved domains and intron locations [7, 8]. A third hypothesis, known as the “IIa + IIb” model, highlights the importance of whole-genome duplication, tandem duplication, and fragment duplication in the evolution of complex plant genomes [34]. In the present study, the 104 PtWRKY genes showed strong conservation in motif composition, domain organization, and intron number, indicating that the PtWRKY gene family in P. tenuifolia has remained relatively conserved during evolution. Gene family expansions can result from chromosome or segmental duplication, tandem duplication, transposition, and homing [12]. Compared to A. thaliana, the WRKY gene family in P. tenuifolia has undergone significant expansion, with 572 fragment duplication events (Figure S3A) and 1 tandem duplication events (Fig. 5A: Pt6G01096/Pt6G01097). These results suggest that fragment duplication contributes significantly to the expansion of PtWRKY gene family.

Expression of PtWRKY genes is regulated by multiple factors

Promoter analysis showed that 103 out of the 104 PtWRKY genes contain cis-acting elements related to dehydration-responsive element (DRE), LTR, and salt stress-responsive elements, with Pt17G01095 being the only exception. Most PtWRKY promoters contained multiple cis-acting elements, which were classified into four groups: hormone response elements (12 types), growth and development-related elements (16 types), defense and stress-responsive elements (8 types), and light-responsive elements (26 types). Similar regulatory complexity has been seen in other medicinal plants. For example, three DoWRKY genes in Dendrobium officinale are significantly induced by low-temperature and methyl jasmonate (MeJA) treatments [23], while PgWRKY genes in Platycodon grandiflorus contain numerous stress and hormone-related cis-acting elements, indicating potential roles in abiotic stress regulation [12].

In the present study, 54 PtWRKY genes were associated with stress homeostasis-related elements, particularly LTR elements, highlighting their potential involvement in cold stress adaptation. Previous studies have reported that bananas rewire WRKY proteins from immune functions to respond to cold stress responses, disrupting intracellular homeostasis and eventually causing browning [52]. In tomatoes, loss of cold tolerance results from evolutionary deletion in the WRKY34 promoter, which disrupts the CBF-mediated cold-response pathway at both transcriptional and protein levels [53].

These findings indicate that the expression of the PtWRKY gene family is regulated by integrating hormonal, light, and stress-related signaling pathways. This characteristic is also manifested in Pt13G01037 and Pt17G01891, providing multi-dimensional evidence for inferring their functions in cold stress responses. Although our initial functional inference of these two genes was based on their similarity to homologous genes in A. thaliana, subsequent sequence conservation analysis, promoter cis-acting element identification, and tissue-specific expression profiling further confirmed their potential roles in cold stress responses and highlighted their typical feature of being regulated by multiple factors. Among these, tissue-specific expression characteristics intuitively reflect their adaptive regulation in response to environmental stresses: the preferential expression of Pt13G01037 and Pt17G01891 in roots important functional implications for cold stress adaptation-roots are directly exposed to lows oil temperatures and are critical for water uptake and osmotic balance under cold conditions. High basal expression of PtWRKY genes in roots may enable rapid transcriptional reprogramming in response to temperature fluctuations, thereby enhancing root cold tolerance. Notably, Pt17G01891 showed particularly high expression in seedling stems and mature roots suggesting a potential role coordinating root-shoot responses during cold stress. Although tissue-specific expression was analyzed conducted under normal growth conditions, the observed spatial expression patterns, together with cold-induced transcriptional activation, suggest that these PtWRKY genes may contribute to cold resistance by reinforcing root-associated stress adaptation mechanisms.

For the two candidate genes analyzed in detail (Pt13G01037 and Pt17G01891), their promoter regions were notably enriched in cis-acting elements related to MeJA and light responsiveness, and both contained ABA-responsive elements. Given that WRKY TFs serve as core regulators in ABA signaling networks, particularly under low-temperature conditions [54], these findings suggest that Pt13G01037 and Pt17G01891 expression may integrate hormone (ABA, MeJA) and light signal pathways to establish a multi-signal synergistic response mechanism. Further promoter analysis provided additional functional insights: Pt13G01037 contained multiple ABREs, implying its potential involvement in ABA-dependent cold signaling; Pt17G01891 harbored both ABRE and MBS elements (typically linked to abiotic stress regulation), and combined with its strong root-specific expression under cold stress, this promoter architecture indicates a role for Pt17G01891 in mediating root-specific cold tolerance.

In summary, Pt13G01037 and Pt17G01891 possess the typical characteristic of WRKY genes, being regulated by multiple factors including hormones, light signals, and stress signals: the high degree of sequence and domain conservation ensures the conservation of their regulatory functions, while the enrichment of various response elements in the promoter region provides the molecular basis for their multi-signal integration and regulation. Combined with their tissue-specific expression patterns adapted to cold stress, these findings fully confirm that Pt13G01037 and Pt17G01891 are important regulators in the cold stress response of P. tenuifolia, and they participate in the improvement of plant cold tolerance through a multi-factor synergistic regulatory mechanism. Taken together, the high level of sequence conservation, preservation of functional WRKY domains, and enrichment of stress- and hormone-responsive cis-elements strongly support the conclusion that Pt13G01037 and Pt17G01891 function as important regulators in the cold stress response of P. tenuifolia.

Pt13G01037 and Pt17G01891 play important roles under cold stress

Cold stress is a major factor limiting the overwintering survival of P. tenuifolia seedlings in northwest China. In plants, three core components of cold signaling have been identified: Inducer of CBF Expression (ICE), C-repeat Binding Factors (CBFs), and Cold-Regulated genes (CORs). The ICE-CBF-COR signaling cascade plays a central role in alleviating cold stress and has been well documented in rice [55], wheat, tea [55], and P. mume ‘Huahudie’ [56]. Typically, P. tenuifolia seedlings must withstand subzero temperatures for nearly one month to successfully overwinter. Based on correlation analysis of DEGs between control (CK) and cold-stressed (CS) treatments and their expression patterns, Pt13G01037 and Pt17G01891 were identified as two key candidate genes involved in cold stress responses. Gene cloning and qRT-PCR verification further confirmed that the expression of both genes was significantly upregulated in the CS5, indicating their crucial roles in cold stress responses of P. tenuifolia.

Phylogenetic classification showed that Pt13G01037 belongs to Subgroup Ia, while Pt17G01891 belongs to Group IIf. To characterize the functions of Pt13G01037 and Pt17G01891, we performed protein sequence alignment in the NCBI database and identified the top five genes with the highest homology for each of the two target genes, respectively. The results showed that the homology of Pt13G01037 with AtWRKY75 and AtWRKY45 from A. thaliana, as well as WRKY75 from Ziziphus jujuba, Cajanus cajan and Theobroma cacao, ranged from 63.5% to 84.8% (Figure S5). Among these homologs, the highest sequence identity was observed with AtWRKY75 (84.8%) and AtWRKY45 (78.3%). Concurrently, Pt17G01891 shared the highest homology of 54.5% with AtWRKY27 from A. thaliana, and also exhibited relatively high homology with WRKY29 from other plant species including Gastrolobium bilobum, Rosa rugosa, Vigna radiata var. radiata and Durio zibethinus (Figure S6). Based on the evolutionary relationships and a comprehensive review of relevant literature, existing studies on these homologous genes provide important clues for inferring the functions of Pt13G01037 and Pt17G01891. Specifically, AtWRKY45 responds to salt and osmotic stress by inducing transcriptional activation, binding to the RD29A promoter, and regulating stress-related genes, which ultimately increases proline and chlorophyll content while reduced MDA levels and reactive ROS levels [57]. AtWRKY75, the closest homolog of AtWRKY45 in A. thaliana with 75% amino acid similarity, positively regulates salt stress tolerance by directly binding the SOS1 promoter to modulate its expression-evidenced by salt sensitivity in AtWRKY75 T-DNA mutants and enhanced salt tolerance in AtWRKY75-OX plants. Unlike AtWRKY75, however, AtWRKY45 targets the promoters of RD29A and RD29B instead of SOS pathway genes; following salt or sorbitol treatment, AtWRKY45-OX plants exhibit elevated expression of ABA-related genes without significant changes in SOS1, SOS2, and SOS3 expression, suggesting an ABA-dependent stress tolerance mechanism mediated by AtWRKY45 [58]. For the homologs of Pt17G01891, AtWRKY27 is associated with plant development, breeding, and immune responses [59], while AtWRKY29 enhances stress tolerance by binding to the AtBAG7 promoter and activating transcription [60]. Given these functional characteristics of their homologs, we speculate that Pt13G01037 and Pt17G01891 may similarly regulate abiotic stress responses to temperature changes in P. tenuifolia.

Mechanistically, the regulatory role of Pt13G01037 and Pt17G01891 in cold stress responses is supported by multiple lines of evidence. First, as WRKY TFs, they are consistent with the well-documented role of WRKY TFs in positively regulating stress resistance by influencing downstream genes. For example, overexpression of PmWRKY57 increases the expression of cold-response genes (AtCOR6.6, AtCOR47, AtKIN1, and AtRCI2A) under low temperatures [56]. Similarly, over-expressing TaWRKY19 improves tolerance to salt, drought, and freezing stress by activating genes such as DREB2A, RD29A, RD29B, and Cor6.6 [61]. In Verbena bonariensis, VbWRKY32 enhances cold tolerance through upregulation of cold-responsive genes [40], while in bermudagrass, CdWRKY2 directly activates sucrose biosynthesis and the CBF-signaling pathway by directly binding to the promoters of genes like CdSPS1 and CdCBF1  [62]. These findings suggest that Pt13G01037 and Pt17G01891 may confer cold stress tolerance in P. tenuifolia by regulating genes in the ICE-CBF-COR pathway, a hypothesis warranting further experimental validation.

Subsequently, the presence of multiple cold and hormone-responsive cis-acting elements (such as ABRE) in the promoter regions of Pt13G01037 and Pt17G01891 indicates that these genes may serve as effective regulatory nodes in cold stress signaling networks. This inference is further supported by the enrichment of plant hormone signal transduction pathways (especially ABA-related pathways) under cold stress. ABA signaling is well-documented to modulate osmotic adjustment, antioxidant defense, and secondary metabolism-processes critical for stress tolerance in medicinal plants-suggesting that Pt13G01037 and Pt17G01891 mediate hormonal regulation to coordinate cold stress responses.

Notably, the functional characteristics of Pt13G01037 and Pt17G01891 also imply their potential role in coordinating cold stress responses with medicinal component biosynthesis in P. tenuifolia. On one hand, WRKY TFs are known to regulate genes involved in phenylpropanoid biosynthesis and antioxidant defense. Under cold stress, the phenylpropanoid biosynthesis pathway is strongly enriched in P. tenuifolia, this pathway is a central metabolic route for synthesizing phenolic compounds, flavonoids, and other bioactive metabolites that contribute to the medicinal properties of P. tenuifolia (such as, neuroprotective and antioxidant activities). On the other hand, cold stress induces metabolic reprogramming of carbohydrate metabolism pathways to balance energy supply, stress protection, and secondary metabolite biosynthesis. Collectively, these findings suggest that Pt13G01037 and Pt17G01891 may modulate the phenylpropanoid biosynthesis pathway and carbohydrate metabolism pathways, acting as core regulators in the interconnected networks of stress adaptation and medicinal component accumulation.

In terms of practical application value, Pt13G01037 and Pt17G01891 exhibit distinct advantages as candidate genes for downstream genetic experiments: they show strong and reproducible induction under cold stress, clear tissue-specific expression patterns, and high conservation of the WRKY DNA-binding domain-features that ensure reliable stress responsiveness and functional stability, which are critical for candidate genes targeted in genetic engineering.

Collectively, this study identifies Pt13G01037 and Pt17G01891 as key cold-responsive WRKY genes in P. tenuifolia, revealing that cold stress responses in this species are closely linked to pathways governing both stress adaptation and medicinal compound biosynthesis. These findings provide a mechanistic basis for understanding how environmental stress shapes the quality and pharmacological potential of P. tenuifolia, and offer insights into the potential mechanisms underlying WRKY-mediated abiotic stress regulation in P. tenuifolia seedlings. Consequently, Pt13G01037 and Pt17G01891 represent promising targets for future functional studies (such as, transgenic overexpression, gene silencing, or CRISPR/Cas-mediated genome editing) aimed at enhancing cold tolerance and optimizing medicinal component accumulation in P. tenuifolia and related species.

Conclusions

In this study, the PtWRKY gene was comprehensively identified and characterized for the first time based on transcriptome data under cold stress of P. tenuifolia. A total of 104 PtWRKY genes were identified and classified into 12 phylogenetic subgroups. Their chromosomal distribution, collinearity relationships, gene structures, conserved motifs, and promoter cis-acting elements were systemically analyzed. Notably, the promoter regions of most PtWRKY genes contained multiple stress-responsive elements, indicating potential roles in abiotic stress regulation. To investigate the involvement of PtWRKY genes in cold stress responses, 40-day-old P. tenuifolia seedlings were subjected to low-temperature treatment at 4℃, followed by physiological measurements and transcriptome sequencing. Integrative analysis identified two key genes, Pt13G01037 and Pt17G01891, as strong candidate regulators of abiotic stress responses. The full-length transcripts of these genes were cloned, and their expression patterns were examined across different tissues under cold stress. Both genes showed strong and consistent induction, supporting their involvement in the cold-responsive regulatory process. Overall, the results indicate that P. tenuifolia seedlings exhibit an adaptive response to cold stress, reflecting an inherent level of cold tolerance.

Despite these findings, several limitations should be acknowledged. First, the functional role of PtWRKY genes were inferred primarily from transcriptomic profiling and qRT-PCR validation. Direct functional evidence from genetic manipulation experiments, such as gene overexpression, knockout, or gene silencing, was not performed. As a result, the causal roles of Pt13G01037 and Pt17G01891 in cold tolerance remain to be confirmed. Second, although promoter analysis and expression patterns suggest that PtWRKY genes may participate in classical cold-responsive regulatory pathways, the interactions between PtWRKY TFs and their downstream target genes were not experimentally validated. In particular, whether Pt13G01037 and Pt17G01891 directly regulate key cold-responsive genes, such as CBF and COR family members, remains unclear. Third, cold stress was applied at a single temperature (4℃), which may not fully capture the range of cold conditions experienced by P. tenuifolia in natural environments.

Future research will aim to address these limitations. The functional roles of Pt13G01037 and Pt17G01891 will be validated through genetic manipulation approaches, including transgenic or heterologous expression systems, to assess their effects on cold tolerance-related physiological traits. Regulatory relationships between PtWRKY proteins and downstream cold-responsive genes will be elucidated using molecular interaction assays, including yeast one-hybrid (Y1H), chromatin immunoprecipitation (ChIP-qPCR), and bimolecular fluorescence complementation (BiFC). In addition, the expression dynamics of PtWRKY genes will be investigated under multiple cold stress gradients and freezing conditions to reflect field environments. Finally, integrative multi-omics approaches will be used to investigate potential links between PtWRKY-mediated cold stress responses and the biosynthesis of medicinally important secondary metabolites in P. tenuifolia.

Authors’ contributions

**Jing Liu** was responsible for conducting the overall experiments and writing the original manuscript; **Shuai Liu** designed the experiments and revised the manuscript; **Hong Zhang** reviewed and edited the manuscript; Qianqian Wang and Zenghui Wang performed the gene family analysis and collected the sample; Shanrong Han and Dilare Tuohetuohuli performed the qRT-PCR.

Funding

This research was supported by key project at central government level: Special Scientific Research Program of Shaanxi Provincial Department of Education for Serving Local Development (25JC017), The ability establishment of sustainable use for valuable Chinese medicine resources (2060302), Key Research and Development Program of Shaanxi (2024SF-YBXM-470, 2023-YBSF-491), Science and Technology Special Project of the State Administration of Traditional Chinese Medicine (GZY-KJS-2023–036), Shaanxi Innovative Healthcare Team for Qin Medicine Quality Control and Product Research and Development, Institutional Research Project of Shaanxi Academy of Traditional Chinese Medicine (2025–6).

Data availability

*Polygala tenuifolia* genome sequence information was obtained from the Genome Sequence Archive website under the accession numbers CRA009096 (http://bigd.big.ac.cn/gsa). The annotations of the *P. tenuifolia* genome can be downloaded from the TCM Plant Genome Database (TCMPG: http://cbcb.cdutcm.edu.cn/TCMPG/resource/genomes/details/?id=TCMPG20196). Transcriptome data of the above-ground parts of *P. tenuifolia* seedlings from different treatment groups under cold stress have been deposited in the National Center for Biotechnology Information (NCBI), with the accession number PRJNA1308798. Data supporting the findings of this study are available within the article and its supplementary data.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.