Lilium ‘Siberia’ bulbs used in this study were purchased from Hongyue Horticultural Corporation (Zhejiang, China; official website: https://www.hongyue.com). Lilium ‘Siberia’ plants were cultivated at the Horticultural Experiment Station of Jinling Institute of Technology. When the flower buds reached a length of approximately 8 cm, the root, stem, leaf, perianth, anther, pollen, and pistil tissues were harvested, immediately frozen in liquid nitrogen, and stored for subsequent analyses. Three biological replicates were prepared for each sample.

Arabidopsis thaliana seeds (ecotype Columbia) were surface-sterilized by soaking in 1% sodium hypochlorite for 10 min, sown on Murashige and Skoog (MS) medium supplemented with 3% sucrose and 0.8% agar (pH 5.8). After stratification at 4°C in darkness for 3 days, the seedlings were then cultured in a light incubator at 22°C under a 16 h/8 h light/dark photoperiod.

Total RNA was extracted from the leaves of Lilium ‘Siberia’ Plants using the TRIzol method [3], and first-strand cDNA was synthesized using the HiScript II kit (Vazyme, Nanjing, China). The open reading frame (ORF) of LoWRKY22 was subjected to polymerase chain reaction (PCR) amplification using cDNA extracted from the mature leaves of potted Lilium ‘Siberia’ plants as the template and LoWRKY22-F/R as gene-specific primers (Table S1). The PCR reaction procedure was as follows: initial denaturation at 98°C for 30 s; 35 cycles of 98°C for 10 s, 60°C for 30 s and 72°C for 60 s; followed by a final extension at 72°C for 10 min. The amplified products were purified and ligated into the pMD18-T vector (Takara, Japan), and the recombinant plasmids were transformed into Escherichia coli DH5α competent cells. Positive clones were screened and verified by sequencing.

Multiple sequence alignment of the LoWRKY22 protein was performed using Clustal W (version 1.81) and BioEdit software (version 7.0) with default parameters. The alignment results were imported into MEGA (version 7), and a phylogenetic tree was constructed using the neighbor-joining (NJ) method with 1000 bootstrap replicates for statistical validation, and other parameters were kept as default.

pGBKT7 (BD) vector (CLONTECH Laboratories, USA) was used in the experiment. The BD-LoWRKY22 recombinant expression vector was constructed by homologous recombination using the ClonExpress II One Step Cloning Kit (Vazyme, Nanjing, China). The sequences of the primers used for vector construction are provided in Table S1. A GAL4 vector (positive control), an empty BD vector (negative control), and the recombinant BD-LoWRKY22 vector were independently transformed into Saccharomyces cerevisiae AH109 competent cells. The transformed yeast cells were plated on SD/-Trp solid medium and incubated in an inverted manner at 30°C for 2–3 days. Single colonies were selected and verified by PCR and sequencing. The confirmed positive yeast strains were resuspended in 100 μL of sterile water. Then, 5 μL aliquots of the yeast (Saccharomyces cerevisiae) suspension were spotted onto solid culture plates including SD/-Trp, SD/-Trp-His, SD/-Trp-His containing 5 mM 3-AT and SD/-Trp-His with 15 mM 3-AT, followed by incubation at 30°C for 2–3 days. After incubation at 30°C for 2–3 days, the growth status of yeast colonies was observed and photographed to identify the self-transactivation activity of BD-LoWRKY22 fusion protein.

The growth of the yeast colonies was observed and documented photographically. The β-galactosidase chromogenic assay was performed as follows: a sterile filter paper was placed flat on SD/-Trp plates containing fresh yeast colonies, and the colonies were transferred onto the filter paper. The filter paper was then clamped with tweezers, rapidly frozen in liquid nitrogen for 1 min, thawed at 15–30°C, and this freeze-thaw cycle was repeated 3–5 times. The filter paper was subsequently placed in a clean Petri dish, with the yeast colonies facing upward, and 1 mL of Z buffer/X-Gal solution was evenly applied. The plate was then incubated in the dark at 30°C, following which color development was monitored and β-galactosidase activity was recorded photographically.

The full-length coding sequence of LoWRKY22 without stop codon was inserted into the expression vector by homologous recombination. Competent cells of Agrobacterium tumefaciens strain GV3101 were transformed with the the constructed fusion plasmid. The transformed bacterial cells were cultured in liquid Luria-Bertani (LB) medium supplemented with kanamycin and rifampicin, with shaking at 200 rpm and 30°C for 14 to 16 hours, until the OD₆₀₀ absorbance value reached 1.5 to 2.0. Bacterial cells were gathered by centrifugal treatment and re-dissolved in infiltration buffer consisting of 10 mM MgCl₂ and 10 mM 2-(N-morpholino)ethanesulfonic acid, 200 μM acetosyringone, and pH = 5.6. The bacterial suspension was adjusted to an OD₆₀₀ value of 1.0, followed by dark static incubation at 15–30°C for 4 h. A needleless 1 mL syringe was used to inject the agrobacterial suspension into tobacco leaf mesophyll tissues from the abaxial side. The infiltrated tobacco plants were cultured in darkness for 1 day and then under normal light conditions for another day. Leaf tissues around the injection sites were sampled, and the fluorescence signals of target protein were observed and photographed using a laser scanning confocal microscope (Zeiss LSM780, Germany).

The LoWRKY22 promoter was cloned using the FPNI promoter cloning method [16]. Three reverse primers (R1, R2, and R3) were designed to amplify the 5′ region located within 500 bp upstream of the start codon of the LoWRKY22 ORF in the Lilium ‘Siberia’ cultivar. The first primer (R1) was positioned farthest from the start codon. Each primer was 20–25 bp in length, spaced 60–100 bp apart, and exhibited a GC content of 45–55%. The sequences of the three reverse primers are provided in Table S1. Following confirmation of the PCR-positive amplicons by sequencing, new specific primers were designed, and the LoWRKY22 promoter was re-amplified using high-fidelity PCR with 2× Super Pfx Master Mix (CWBIO, Beijing, China). The re-sequencing data were compared with the primary cloning sequences and calibrated to ensure accuracy.

The structural organization of the LoWRKY22 promoter was visualized using TBtools [17]. The LoWRKY22 promoter sequence was submitted to the PlantCARE database (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/) for predicting the cis-acting regulatory elements, and the results were visualized and annotated using TBtools.

The ORF sequence of LoWRKY22 excluding the stop codon was inserted into the expression vector via homologous recombination, and the resultant recombinant plasmid was transformed into A. tumefacien strain GV3101 competent cells. The bacterial cells were harvested and resuspended in 10 mM MgCl₂ solution to adjust the OD₆₀₀ value to 0.6, and wild-type A. thaliana was transformed using the floral dip technique [18]. After obtaining transgenic Arabidopsis lines, seedlings of wild-type and transgenic plants were transplanted into potted soil. The bolting time of each seedling was recorded 14 days after transplantation, and six uniformly growing plants from each line were selected for statistical investigation. At 60 days post-transplantation, five representative plants per line were chosen to measure plant height and count the number of dehiscent siliques. All statistical analyses were performed using SPSS 17.0 software, and Duncan’s new multiple range test was applied for significance difference analysis. GraphPad Prism software was used for graph plotting. All experimental data were presented as mean ± standard deviation (SD).

For quantitative real-time PCR (qRT-PCR) assays, seeds of wild-type and transgenic A. thaliana were surface-sterilized with 1% sodium hypochlorite and sown on solid MS medium containing 3% sucrose and 0.8% agar. After stratification treatment, the seeds were transferred to a plant growth chamber at 25°C with a 16 h light/8 h dark photoperiod. Whole seedlings were harvested 10 days after sowing for total RNA extraction and subsequent cDNA synthesis. Specific qRT-PCR primers were designed as shown in Table S1. Actin was used as the internal reference gene for A. thaliana, and 18S rRNA was selected as the reference gene for Lilium ‘Siberia’. The qRT-PCR reactions were carried out using ChamQ Universal SYBR qPCR Master Mix (Vazyme, Nanjing, China) with the following program: pre-denaturation at 95°C for 10 min; 40 cycles of 95°C for 15 s, 55°C for 15 s and 72°C for 20 s. The relative gene expression levels were calculated according to the 2^−ΔΔCt method. Each sample was prepared with three independent biological replicates, and all PCR primer sequences are shown in Table S1.

The TRINITY_DN192_c0_g1_i1_3_1 gene was isolated from the Lilium ‘Siberia’ cultivar and encodes a Trinity transcript derived from the RNA-seq dataset associated with NCBI BioProject PRJNA767274. It contains a full-length cDNA sequence of 834 bp (Table S1) and an ORF encoding 277 amino acids. Phylogenetic analysis revealed that this gene formed a cluster with the AtWRKY22, AtWRKY29, and AtWRKY27 genes of Arabidopsis thaliana (Fig. 1). Amino acid sequence alignment of LoWRKY22 with AtWRKY22, AtWRKY27 and AtWRKY29 revealed that it shares the closest evolutionary relationship with AtWRKY22 (Fig. S1). Further sequence alignment analysis revealed that the protein encoded by LoWRKY22 contains a canonical WRKY repeat domain at the N-terminus, which mediates DNA binding, as well as a conserved C2H2-type zinc finger domain at the C-terminus (Fig. 2).

As depicted in Fig. 3, the transcriptional activation assays revealed that yeast colonies transformed with the GAL4 vector (positive control), the empty BD vector (negative control), and the recombinant BD-LoWRKY22 vector grew normally on—Trp medium. Yeast cells transformed with GAL4 or BD-LoWRKY22 exhibited normal growth on—Trp-His medium, whereas those transformed with the empty BD vector (negative control) failed to grow. The addition of 5 mM or 15 mM 3-AT to the—Trp-His medium did not inhibit the growth of yeast colonies transformed with the GAL4 or BD-LoWRKY22 vectors. β-galactosidase staining assays further confirmed that both GAL4 and BD-LoWRKY22 exhibit transactivation activity.

To determine the subcellular localization of LoWRKY22 fused with green fluorescent protein (GFP)—specifically, its distribution within the cytoplasm, nucleus, or other cellular compartments—Agrobacterium tumefaciens harboring the recombinant plasmid was infiltrated into tobacco leaf epidermal cells using a needleless syringe. The fluorescence signals were visualized using a Zeiss LSM800 laser scanning confocal microscope following 48 h of incubation. Bright-field images confirmed that the cells remained spherical and structurally intact, indicating that the observed green fluorescence originated from intracellular protein expression rather than cellular damage or extraneous contaminants. These findings collectively demonstrate that LoWRKY22 localizes to the nucleus (Fig. 4).

To clarify the specific regulatory basis governing LoWRKY22 expression, we conducted an in silico prediction of cis-regulatory motifs in the promoter region upstream of its transcription start site (Fig. 5). The analysis revealed that, besides canonical core promoter features such as the TATA box (positioned approximately 30 bp upstream of the transcription start site) and the CAAT box (a ubiquitous cis-regulatory motif in promoters and enhancers), the LoWRKY22 promoter also contains two light-responsive motifs (GT1-motif and Box 4), one low-temperature-responsive motif (LTR), and one gibberellin-responsive motif (P-box) (Table 1).

Analysis of the tissue-specific expression of LoWRKY22 (Fig. 6) revealed that its relative expression levels were highest in leaves (63.36). LoWRKY22 expression did not differ significantly between the roots and perianths, which exhibited the second-highest expression levels (13.13 and 14.61, respectively; mean = 13.87). The relative expression of LoWRKY22 was lowest in the stems, anthers, pollen, and pistils, with no significant differences in expression among these tissues (2.67, 2.40, 0.93, and 2.21, respectively; mean = 2.05). The relative expression level of LoWRKY22 in the leaves was significantly higher than that in the roots and perianths, representing a 356.81% increase. Furthermore, the mean relative expression level of LoWRKY22 in the roots and perianths was significantly higher than that in the stems, anthers, pollen, and pistils, with a 576.59% increase. The relative expression level of LoWRKY22 in the leaves was 6.77-fold higher than its mean expression in stems, anthers, pollen, and pistils.

For functional characterization, LoWRKY22 was heterologously expressed in Arabidopsis thaliana. Three independent transgenic lines (OE-4, OE-5, and OE-8) were confirmed by RT-PCR and selected for subsequent functional analyses. Under normal growth conditions, the flowering phenotype of seedlings expressing LoWRKY22 was recorded two weeks following transplantation from MS medium to soil. The transgenic plants exhibited significantly earlier flowering than the WT Arabidopsis plants (p < 0.05) (Fig. 7A). At 60 days post-transplantation, the transgenic lines expressing LoWRKY22 displayed a dwarf phenotype, characterized by reduced plant height (Fig. 7B) and premature silique maturation compared to that observed in WT plants.

To determine the effects of LoWRKY22 overexpression on the expression of genes involved in plant growth and development, 10-day-old whole seedlings of transgenic Arabidopsis thaliana were subjected to qRT-PCR analysis. As depicted in Fig. 8, LoWRKY22 overexpression in Arabidopsis thaliana significantly upregulated the genes associated with plant senescence (p < 0.05). Specifically, the expression levels of AtWRKY53, AtSAG12, and AtSAG13 increased by 4.24-, 2.75-, and 6.47-fold, respectively, in transgenic lines, compared to those in WT plants. Furthermore, analysis of flowering-related genes revealed that LoWRKY22 overexpression significantly altered their expression. Compared to those in WT plants, the expression levels of AtFT and AtAP1 increased by 2.75- and 2.74-fold, respectively, in transgenic lines, whereas AtFLC levels decreased by 0.71-fold.