Genome-wide identification and expression analysis of the bHLH gene family in Lilium davidii var. unicolor and its response to abiotic stress Authors: Yue Yang, Mengyue Jing, Yao Qu, Jiaxing Li, Youguo Wang, Tao Yin, Xinyu Yang Published: 2026 ABSTRACT BackgroundBasic helix-loop-helix (bHLH) transcription factors regulate plant growth, stress responses, and hormone signaling. Advances in bioinformatics, the bHLH transcription factor family has been comprehensively identified and characterized across a wide range of organisms. However, a genome-wide analysis of this family has not yet been performed in Lilium davidii var. unicolor, an economically important species in China valued for its nutritional and medicinal properties.ResultsUsing whole-genome sequencing data, we identified 214 LdbHLH genes from the L. davidii var. unicolor genome and classified them into 23 subfamilies based on phylogenetic analysis. Gene structure was highly conserved within subfamilies. Collinearity analysis identified five syntenic gene pairs (LdbHLH3-LdbHLH9, LdbHLH17-LdbHLH61, LdbHLH176-LdbHLH125, LdbHLH181-LdbHLH197, and LdbHLH210-LdbHLH200), all derived from segmental duplication, indicating limited large-scale expansion. Promoter analysis showed that LdbHLH genes commonly contain cis-regulatory elements associated with development, abiotic stress, and hormone responses, suggesting their potential roles in signal crosstalk. Expression profiling demonstrated clear tissue specificity, with certain genes highly expressed in leaves, scales, stems, or tepals. Under hormone and stress treatments, most genes exhibited transient expression changes, while some genes, such as LdbHLH76 under salt stress and LdbHLH7 under ABA treatment, were strongly upregulated. Functional assays further confirmed that LdbHLH12, LdbHLH46, LdbHLH76, and LdbHLH189 positively contribute to salt tolerance.ConclusionsThis study provides the first comprehensive functional and genomic characterization of the bHLH family in L. davidii var. unicolor, offering insights into their roles in stress adaptation and tissue development, and identifying candidate targets for molecular breeding aimed at enhancing stress resilience.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12870-026-08961-5. Background Transcription factors (TFs) are DNA-binding proteins, also known as trans-acting factors, which recognize specific cis-regulatory elements (CREs) in gene promoters to activate or repress downstream gene expression in response to environmental signals [1, 2]. In higher plants, numerous TFs involved in biotic and abiotic stress responses have been identified, where they regulate genes associated with drought, salinity, temperature extremes, phytohormones, and pathogen defense [3]. Plant TFs usually contain four functional domains: a DNA-binding domain (DBD), a transcriptional regulatory domain, a nuclear localization signal, and an oligomerization site [4]. Based on differences in their DBDs, TFs are classified into several families, such as bZIP, Zinc finger, basic helix-loop-helix (bHLH), and MYB, with bHLH representing the second largest family in plants after MYB [5]. The bHLH family is widely distributed in plants and is characterized by a conserved bHLH domain of approximately 60 amino acids. This domain contains an N-terminal basic region (15 − 20 amino acids, rich in basic residues) that specifically recognizes and binds CREs [6, 7], and a C-terminal HLH region (40 − 50 amino acids) that facilitates homo- and heterodimer formation through two amphipathic α-helices connected by a loop [8]. Therefore, bHLH TFs predominantly function as dimers. They play important roles in regulating plant development and growth, including seed germination and the formation of stomata, flowers, and guard cells, as well as responses to abiotic stresses [9, 10]. In addition, bHLH TFs interact with other TFs to regulate the biosynthesis of secondary metabolites, such as terpenoids, alkaloids, phenylpropanoids, and anthocyanins, which are critical for plant-environment interactions [11, 12]. With the rapid progress of genome sequencing and the increasing availability of plant genomes, bHLH gene families have been comprehensively identified in many species. For instance, 162 bHLH members have been identified in A. thaliana [13], 167 in rice (Oryza sativa) [14], and 208 in maize (Zea mays) [15]. Functional studies have shown that these proteins participate in diverse biological processes, including stress responses, light signaling, phytohormone signaling, iron uptake, anthocyanin and flavonoid biosynthesis, as well as the development of stomata, roots, and petals. [16]. For example, MYC2 integrates JA signaling with other phytohormone pathways, including abscisic acid (ABA), salicylic acid (SA), gibberellins (GAs), and auxin (IAA) [17]. In Rosa persica, several RbebHLHs regulate low-temperature tolerance through the JA pathway. In rice, OsbHLH148 interacts with OsJAZ proteins to modulate the jasmonate signaling and enhance drought tolerance [18]. Similarly, MdbHLH093 interacts with MdMYB116 and activates the SA signaling pathway by increasing hydrogen peroxide accumulation, thereby improving resistance to powdery mildew in apple [19]. Recent studies have also highlighted the role of bHLH TFs in salt stress responses. For example, bHLH123 enhances salt tolerance in tobacco by activating the expression of the NADPH oxidase gene NtRbohE; bHLH142 improves salt-alkali tolerance in winter rapeseed; and bHLH319 is involved in salt tolerance regulation in wheat. However, most of these studies focus on crops or model species such as rice, wheat, and Arabidopsis, with relatively little is known about bHLH genes in Lilium davidii var. unicolor. Therefore, their functions and molecular regulatory mechanisms, particularly in salt stress responses and hormonal signaling, remain largely unclear. L. davidii var. unicolor is an economically important crop endemic to China, valued for its ornamental, edible, and medicinal properties [20]. Its bulbs are large, with a delicate texture, sweet flavor, and white color, and are rich in amino acids, trace elements, vitamins, total phenols, flavonoids, flavanols, and polysaccharides. These bioactive components contribute to its high nutritional value, supporting its reputation of “vegetable ginseng,” as well as medicinal benefits such as moistening the lungs, relieving coughs, and regulating nerve function, supporting its role as a dual-purpose resource for food and medicine [21]. The species is mainly cultivated in the semi-arid regions of Northwest China, particularly in Gansu Province, where it contributes significantly to the local economy. Its natural tolerance to high salinity, drought, and poor soils also makes it a useful system for studying abiotic stress adaptation in Lilium. Understanding the physiological and molecular regulatory networks underlying its stress responses will elucidate stress adaptation mechanisms in Lilium, support the discovery of stress-related genes, advance molecular breeding, and promote sustainable cultivation in ecologically vulnerable regions [22]. The availability of the L. davidii var. unicolor genome enables in-depth study of its evolution, development, and gene functions [23]. In this study, we identified bHLH genes at the whole-genome level and analyzed their physicochemical properties, chromosomal distribution, cis-elements, gene structures, and phylogeny. We further examined their tissue-specific expression and transcriptional responses to phytohormones, drought, and salt stress. Preliminary functional validation was performed for selected LdbHLH genes potentially involved in salt tolerance. These results provide candidate genes for studies of stress responses and phytohormone signaling and establish a basis for understanding the biological functions and evolutionary mechanisms of the bHLH gene family. Results Whole-genome characterization of bHLH TFs in L. davidii var. unicolor Putative bHLH gene family members were identified from the genome of L. davidii var. unicolor using hidden Markov models (HMMs) [24]. The presence of conserved bHLH domains was subsequently verified using SMART and Pfam databases. After removal of redundant sequences and transcripts lacking complete bHLH domains, a total of 214 non-redundant LdbHLH transcripts were retained. These genes were further characterized in terms of amino acid length, molecular weight, isoelectric point (pI), and genomic localization (Supplementary Table S1). Protein lengths ranged from 84 to 971 amino acids. The pI values varied from 4.43 to 11.02, and molecular weights ranged from 9,494.7 Da to 106,950.2 Da. The observed diversity in physicochemical properties suggests functional heterogeneity among LdbHLH proteins, with most exhibiting an acidic bias (pI < 7). Subcellular localization prediction using WoLF PSORT indicated that 173 proteins were localized to the nucleus, 24 to the cytoplasm, 11 to the chloroplast, three to the Golgi apparatus, and one protein each to the peroxisome, mitochondrion, and vacuole. To characterize the sequence features of bHLH TFs, multiple sequence alignment was performed utilizing all 214 protein sequences. Analysis of the bHLH domains revealed that these proteins contain the typical conserved bHLH motif, comprising four regions: the second helix, loop, first helix, and basic region (Fig. 1A, B). A total of 25 amino acid residues showed a conservation frequency exceeding 50%, among which nine residues (Leu-61, Tyr-57, Lys-36, Pro-26, Leu-21, Arg-11, Arg-10, Arg-8, and Glu-7) exhibited conservation greater than 75% (Fig. 1C, D). The basic region, essential for DNA-binding activity, contained four highly conserved residues (Glu-7, Arg-8, Arg-10, and Arg-11). The first helix region included eight conserved residues (Lys-13, Leu-14, Asn-15, Leu-21, Ser-23, Leu-24, Val-25, and Pro-26), while the loop region showed conservation at Lys-36 and Asp-41. The second helix region comprised Lys-42, Ala-43, Ser-44, Leu-51, Ala-54, Ile-55, Tyr-57, Lys-59, Leu-61, Gln-62, and Gln-64. Notably, Arg-11 and Leu-21 were highly conserved across all 214 LdbHLH proteins, suggesting their critical roles in bHLH dimer formation.Fig. 1Characterization and distribution of bHLH domains. A Three-dimensional structure of the bHLH domain in LdbHLH13 (Lily02G06180). B Two-dimensional representation of the bHLH domain: the basic region (residues 125–135, pink), first helix (residues 136–150, green), loop region (residues 151–156, purple), and second helix (residues 157–186, yellow). C Sequence logo of the bHLH domain derived from alignment of 214 LdbHLH proteins. D Amino acid distribution in the conserved bHLH motif of bHLH TFs. Numbers below indicate residue positions in the alignment Characterization and distribution of bHLH domains. A Three-dimensional structure of the bHLH domain in LdbHLH13 (Lily02G06180). B Two-dimensional representation of the bHLH domain: the basic region (residues 125–135, pink), first helix (residues 136–150, green), loop region (residues 151–156, purple), and second helix (residues 157–186, yellow). C Sequence logo of the bHLH domain derived from alignment of 214 LdbHLH proteins. D Amino acid distribution in the conserved bHLH motif of bHLH TFs. Numbers below indicate residue positions in the alignment Classification and phylogenetic relations of LdbHLHs We constructed a phylogenetic tree to investigate the grouping and evolutionary trends of LdbHLHs, incorporating bHLHs from L. davidii var. unicolor (214 LdbHLHs) and A, thaliana (162 AtbHLHs). Based on the established classification system of A, thaliana bHLHs [25], the 372 proteins were grouped into 26 subfamilies. The 214 LdbHLHs were unevenly distributed across 23 subfamilies (Fig. 2). The number of members per subfamily ranged from one to 24, with subfamily IVa containing the largest number of LdbHLH (24 members), whereas subfamilies III(d + e) and IVc each contained only a single LdbHLH. Certain clusters displayed pronounced species-specific biases, suggesting that specific gene duplications occurred after the divergence of these species [26]. Furthermore, three subfamilies (Ib(1), III(a + c), and X) were absent in L. davidii var. unicolor. These differences in subfamily composition and gene abundance indicate substantial interspecific divergence of the bHLH gene family between the two species.Fig. 2Phylogenetic tree of bHLH proteins in L. davidii var. unicolor and A. thaliana. Subfamilies are labeled as XV, XIV, XIII, XII, XI, Ⅹ, Ⅸ, Ⅷc, Ⅷb, Ⅷa, Ⅶ(a + b), Ⅵ, Ⅴb, Ⅴa, Ⅳd, Ⅳc, Ⅳb, Ⅳa, Ⅲf, Ⅲ(d + e), Ⅲb, Ⅲ(a + c), Ⅱ, Ⅰb(2), Ⅰb(1), and Ⅰa Phylogenetic tree of bHLH proteins in L. davidii var. unicolor and A. thaliana. Subfamilies are labeled as XV, XIV, XIII, XII, XI, Ⅹ, Ⅸ, Ⅷc, Ⅷb, Ⅷa, Ⅶ(a + b), Ⅵ, Ⅴb, Ⅴa, Ⅳd, Ⅳc, Ⅳb, Ⅳa, Ⅲf, Ⅲ(d + e), Ⅲb, Ⅲ(a + c), Ⅱ, Ⅰb(2), Ⅰb(1), and Ⅰa Gene structure, conserved motifs, and domains analysis of LdbHLHs To gain a deeper understanding of the structural characteristics of LdbHLH proteins, we analyzed their conserved motif composition, domain distribution, and gene structure. MEME analysis identified ten conserved motifs in LdbHLH proteins, all of which were associated with the bHLH domain (Fig. 3B). Among these motifs, Motif 2 and Motif 1 were present across nearly all subfamilies, whereas Motif 6 was specifically in detected in subfamilies IVa and Va. The highly conserved motif distribution within each subfamily suggests potential functional similarity among members. Conserved domain analysis identified 15 types of bHLH domains across LdbHLH proteins (Fig. 3C). Consistent with the motif distribution, all subfamilies contained both shared and subfamily-specific conserved domains, reflecting structural conservation alongside functional diversification.Fig. 3Phylogenetic tree, gene structure, conserved motifs, and conserved domains of LdbHLHs. A Phylogenetic tree of LdbHLHs. B Gene structure and motif distribution of LdbHLHs. C Conserved domains distribution of LdbHLHs Phylogenetic tree, gene structure, conserved motifs, and conserved domains of LdbHLHs. A Phylogenetic tree of LdbHLHs. B Gene structure and motif distribution of LdbHLHs. C Conserved domains distribution of LdbHLHs Chromosomal localization of LdbHLHs Chromosomal localization analysis identified a total of 214 LdbHLH genes, of which 212 were distributed across 12 chromosomes, while the remaining two were located on two contig regions (Fig. 4). The distribution of these genes was heterogeneous: LG04 harbored the largest number of LdbHLH genes, whereas LG06 contained the fewest. Notably, although Chromosome 1 was the longest, it did not contain the highest number of LdbHLH genes, indicating that chromosome length was not significantly linked with the distribution of the LdbHLH gene family.Fig. 4Chromosomal localization of LdbHLH genes. LG represents the chromosomes of L. davidii var. unicolor; chromosome numbers are shown on the left, and the scale on the left represents chromosome length Chromosomal localization of LdbHLH genes. LG represents the chromosomes of L. davidii var. unicolor; chromosome numbers are shown on the left, and the scale on the left represents chromosome length Intra-and Inter-specific synteny analysis of bHLH genes in L. davidii var. unicolor MCScanX was conducted to investigate potential relationships and gene duplication events among LdbHLH genes (Fig. 5A). Five syntenic gene pairs were identified: LdbHLH3-LdbHLH9, LdbHLH17-LdbHLH61, LdbHLH176-LdbHLH125, LdbHLH181-LdbHLH197, and LdbHLH210-LdbHLH200, all of which originated from segmental duplication events. These results suggest that the LdbHLH gene family has been relatively conserved during evolution, with few chromosomal rearrangements or large-scale duplication events.Fig. 5Synteny analysis of LdbHLH genes. A Collinearity analysis of LdbHLHs in the genome of L. davidii var. unicolor. B Syntenic analysis of bHLH genes in L. davidii var. unicolor versus Lilium regale, Pyrus spp., A. thaliana, and Zea mays Synteny analysis of LdbHLH genes. A Collinearity analysis of LdbHLHs in the genome of L. davidii var. unicolor. B Syntenic analysis of bHLH genes in L. davidii var. unicolor versus Lilium regale, Pyrus spp., A. thaliana, and Zea mays Interspecific synteny analysis provides valuable insights into the evolution and phylogenetic relations of genes across species. To further investigate the evolutionary mechanisms of LdbHLH genes, comparative synteny analyses were performed with four species: two dicotyledons (Pyrus spp. and A. thaliana) and two monocotyledons (Lilium regale and Z. mays). A total of 103 LdbHLH genes showed syntenic relationships with these species, including nine with Z. mays, 70 with L. regale, 23 with Pyrus spp., and one with A. thaliana. These results indicate that L. davidii var. unicolor exhibits the strongest syntenic conservation with the congeneric species Lilium regale, whereas lower levels of synteny were observed with dicotyledonous genomes (Pyrus spp. and A. thaliana) and the monocotyledonous genome (Z. mays) (Fig. 5B). Cis-acting element analysis TFs play a key role in coordinating plant responses to abiotic and biotic stresses, with CREs in promoter regions serving as key determinants of gene expression [27]. To investigate the transcriptional regulatory mechanisms of the LdbHLH gene family, 2,000 bp upstream promoter sequences were extracted and their CREs were analyzed utilizing the PlantCARE database, in conjunction with phylogenetic classification. A total of 3,135 CREs were identified in the promoter regions of LdbHLHs (Table S2), with 15 representative CREs revealed in Fig. 6A. Among the LdbHLH genes, LdbHLH117, LdbHLH163, LdbHLH213, and LdbHLH212 contained the highest numbers of CREs, with 67, 40, 40, and 37 elements, respectively. The number of CREs varied across diverse subfamilies (Fig. 6B), and all CREs could be grouped into three major categories (Fig. 6C).Fig. 6Cis-acting elements in LdbHLH Promoters. Different colors denote distinct types of cis-acting elements. A Distribution of cis-acting elements in the 2,000 bp upstream promoter regions of LdbHLHs. A total of 3,135 cis-acting elements were identified, with 15 representative elements displayed. Genes such as LdbHLH117, LdbHLH163, LdbHLH213, and LdbHLH212 contained the highest numbers of elements (67, 40, 40, and 37, respectively). B Number of cis-acting elements on the putative promoters of LdbHLHs across subfamilies. MeJA- and ABA-responsive elements were enriched in most subfamilies, while certain elements (e.g., auxin-, GA-, and stress-responsive elements) showed subfamily-specific enrichment. C Venn diagram showing the three major categories of cis-acting elements: hormone-responsive, development- and growth-related, and stress-responsive elements. Overlapping regions indicate promoters containing multiple categories of regulatory elements Cis-acting elements in LdbHLH Promoters. Different colors denote distinct types of cis-acting elements. A Distribution of cis-acting elements in the 2,000 bp upstream promoter regions of LdbHLHs. A total of 3,135 cis-acting elements were identified, with 15 representative elements displayed. Genes such as LdbHLH117, LdbHLH163, LdbHLH213, and LdbHLH212 contained the highest numbers of elements (67, 40, 40, and 37, respectively). B Number of cis-acting elements on the putative promoters of LdbHLHs across subfamilies. MeJA- and ABA-responsive elements were enriched in most subfamilies, while certain elements (e.g., auxin-, GA-, and stress-responsive elements) showed subfamily-specific enrichment. C Venn diagram showing the three major categories of cis-acting elements: hormone-responsive, development- and growth-related, and stress-responsive elements. Overlapping regions indicate promoters containing multiple categories of regulatory elements The first category comprised hormone-responsive CREs, totaling 1,503 elements. These included ABA-responsive elements (476, 15.18%), SA-responsive elements (121, 3.9%), auxin-responsive elements (43, 1.37%), methyl jasmonate (MeJA)-responsive elements (676, 21.56%), and GA-responsive elements (187, 5.96%). Promoter regions across most LdbHLH subfamilies were enriched in MeJA-responsive elements (CGTCA-motif and TGACG-motif) and ABA-responsive elements (ABRE). SA-responsive elements were detected in 90 of the 214 LdbHLHs and were primarily distributed in subfamilies IVb, IVa, and XII, but absent in III(d + e) and IVc. Auxin-responsive elements (e.g., TGA-box, TGA-element, AuxRR-core, and AuxRE), which were associated with plant growth and development, were identified in specific subfamilies. Approximately one quarter of all LdbHLHs (59 of 214), including members of subfamilies IVb, IX, XII, Ia, Ib, and II, contained auxin-responsive elements in their promoters. Similarly, 95 LdbHLHs harbored GA-responsive elements (P-box, TATC-box, and GARE-motif) predominantly in subfamilies IVa, IVb, Ib, and XIV. Notably, some genes, such as LdbHLH4, LdbHLH46, LdbHLH76, LdbHLH118, LdbHLH170, and LdbHLH175, contained multiple copies of identical hormone-responsive elements, suggesting potential for strong or rapid responses to specific hormonal signals. Conversely, other genes, including LdbHLH7, LdbHLH12, LdbHLH142, LdbHLH152, LdbHLH183, and LdbHLH186, harbored diverse hormone-responsive elements, implying possible involvement in multiple hormone signaling pathways. The second category was linked with development and growth, comprising 608 elements. These included circadian control elements (71, 2.26%), seed-specific elements (17, 0.54%), endosperm-specific negative regulatory elements (6, 0.19%), meristem expression elements (117, 3.73%), and light-responsive elements (397, 12.66%). Light-responsive elements, such as Box I, ACE, G-box, GATA-motif, GT1-motif, and CAG-motif, were widely dispersed across nearly all promoters of LdbHLH genes, with the exception of four genes in subfamily IVc (Fig. 6B). In contrast, other CREs related to development and growth were relatively scarce, with several motifs absent in many subfamilies, such as NON-box and CAT-box (meristem expression), GCN4_motif (endosperm regulation), and RY-element (seed-specific regulation). The third category comprised stress-responsive elements, totaling 765. These included MYB-binding sites involved in drought inducibility (MBS, 158, 5.04%), low-temperature responsiveness elements (126, 4.02%), MYB-binding sites regulating flavonoid biosynthesis (9, 0.28%), defense- and stress-responsive elements (91, 2.9%), and anaerobic induction elements (ARE, 381, 12.15%). Among these, AREs and MBS elements were abundant and broadly distributed across all subfamilies, whereas other stress-related elements were restricted to specific subfamilies. For instance, MYB-binding sites related to flavonoid biosynthesis were enriched in subfamilies IIIb, IIIf, IVa, IVb, IX, Va, and VIIIa, but absent in 14 other subfamilies. Overall, both the number and composition of CREs in the promoters of LdbHLH genes vary substantially within and between subfamilies, implying that bHLH gene expression in L. davidii var. unicolor is regulated by diverse CREs associated with hormone signaling, developmental and growth processes, and stress responses. Homology modeling analysis and protein interactions of the LdbHLH protein Based on syntenic relationships, bit scores, and phylogenetic analyses, orthologous bHLH proteins from A. thaliana were used to predict the protein–protein interaction (PPI) network of LdbHLHs (Fig. 7A). The resulting network comprised 69 nodes, with each node interacting with multiple other nodes. Some proteins exhibited direct interactions, such as LdbHLH7 and LdbHLH76, whereas others displayed complex polygenic interactions. For example, LdbHLH4, LdbHLH12, LdbHLH39, LdbHLH46, LdbHLH77, LdbHLH152, LdbHLH170, LdbHLH175, and LdbHLH183 formed extensive connectivity, suggesting functional diversity of LdbHLH proteins. Protein secondary structure plays critical roles in determining protein sequence, activity, stability, and abundance, and three-dimensional (3D) structures provide deeper insights into protein localization, function, and interaction networks. Using the AlphaFold3 database [28], homology modeling was performed for LdbHLH proteins, and 3D structural models were predicted for members of 12 subfamilies. For each protein, the model with the highest pTM score was selected as the representative structure (Fig. 7B). LdbHLH proteins exhibited diverse 3D structural features across subfamilies, whereas proteins within closely related subfamilies showed relatively similar structures. Many models demonstrated central axis symmetry, as observed in LdbHLH46 and LdbHLH118 (subfamilies XIV and IIIf), as well as LdbHLH175 and LdbHLH183 (subfamilies II and IIIf). Overall, these homology models offer a preliminary foundation for understanding the molecular functions and interaction mechanisms of LdbHLH proteins.Fig. 7A Predicted PPI of LdbHLHs based on orthologs in Arabidopsis thaliana; B Predicted 3D structures of representative LdbHLH proteins from L. davidii var. unicolor A Predicted PPI of LdbHLHs based on orthologs in Arabidopsis thaliana; B Predicted 3D structures of representative LdbHLH proteins from L. davidii var. unicolor Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis of LdbHLHs GO enrichment analysis revealed that LdbHLH genes are significantly enriched in multiple biological processes, cellular components, and molecular functions (Fig. 8B). Within the biological process category, the most enriched term was “response to oxidative stress,” followed by “ABA-activated signaling pathway” and “response to water deprivation.” In the cellular component category, LdbHLH genes were predominantly associated with the nucleus, followed by plasmodesmata and the cytosol. In terms of molecular function, “TF binding” was the most enriched term, followed by “protein kinase activity” and “sequence-specific DNA binding.” KEGG pathway enrichment analysis classified LdbHLH genes into five major functional categories (Fig. 8A). Within cellular processes, “plant hormone signal transduction” was the most prominent pathway, followed by the “MAPK signaling pathway-plant,” highlighting their key roles in hormone-mediated signaling cascades. In the environmental information processing category, “ubiquitin-mediated proteolysis” was prominently enriched, suggesting involvement in protein degradation and regulatory mechanisms. Among pathways associated with genetic information processing, “RNA polymerase” was the most enriched, followed by “spliceosome” and “protein processing in the endoplasmic reticulum.” In the metabolism category, the top enriched pathways included “biosynthesis of amino acids,” “starch and sucrose metabolism,” and “phenylpropanoid biosynthesis.” Finally, within organismal systems, only the “plant-pathogen interaction” pathway was enriched.Fig. 8KEGG (A) and GO (B) enrichment analyses of the LdbHLH gene family KEGG (A) and GO (B) enrichment analyses of the LdbHLH gene family Expression patterns of LdbHLHs in different tissues of L. davidii var. unicolor To further explore the potential functions of LdbHLH genes in the growth and development of L. davidii var. unicolor, we selected candidate genes based on transcriptome data, focusing on those enriched in the “plant hormone signal transduction” and “response to oxidative stress” pathways, as well as those harboring abundance hormone- and stress-responsive elements in their promoter regions. Considering that gene functions often vary across tissues, 12 LdbHLH genes from different subfamilies were selected for expression analysis in four tissues (leaves, tepals, stems, and scales) using qRT-PCR. The results revealed distinct tissue-specific expression patterns (Fig. 9). Specifically, LdbHLH7, LdbHLH46, LdbHLH142, and LdbHLH152 were predominantly expressed in leaves; LdbHLH170 and LdbHLH175 showed elevated expression in scales; LdbHLH4, LdbHLH12, LdbHLH118, and LdbHLH183 were highly expressed in stems; and LdbHLH76 and LdbHLH189 exhibited strong expression in tepals. These findings suggest that tissue-specific expression patterns of LdbHLH genes are closely associated with their regulatory roles in the development and growth of different tissues in L. davidii var. unicolor.Fig. 9qRT-PCR analysis of 12 LdbHLH genes across four tissues. Data are presented as mean ± SD (n = 3 biological replicates). Different lowercase letters indicate significant differences among tissues (P < 0.05), as determined by one‑way ANOVA followed by Duncan's multiple‑range test qRT-PCR analysis of 12 LdbHLH genes across four tissues. Data are presented as mean ± SD (n = 3 biological replicates). Different lowercase letters indicate significant differences among tissues (P < 0.05), as determined by one‑way ANOVA followed by Duncan's multiple‑range test Expression patterns of LdbHLHs under abiotic stress and hormone treatments Cis-acting element analysis of 214 LdbHLH genes revealed the presence of hormone-responsive elements for ABA, MeJA, GA, SA, auxin, and ethephon. To further investigate the involvement of LdbHLH genes in hormonal responses, L. davidii var. unicolor seedlings were treated with these hormones, and samples were collected at 0, 6, 12, 24, and 48 h post-treatment (with 0 h serving as the control). The expression patterns of 12 representative LdbHLH genes from different subfamilies were analyzed utilizing qRT-PCR (Fig. 10). The results revealed substantial variation in relative expression levels under different hormone treatments, indicating that these LdbHLHs respond to all six phytohormones. Despite expression patterns varying among genes, a common trend was observed in which most genes exhibited an initial upregulation followed by a decline. Representative genes displaying this pattern under each treatment are as follows:ABA: LdbHLH142, LdbHLH152, LdbHLH175, LdbHLH189Auxin: LdbHLH4, LdbHLH7, LdbHLH12, LdbHLH46, LdbHLH118, LdbHLH142GA: LdbHLH46, LdbHLH76, LdbHLH118, LdbHLH142, LdbHLH152, LdbHLH170MeJA: LdbHLH7, LdbHLH46, LdbHLH142, LdbHLH152, LdbHLH170, LdbHLH175, LdbHLH189SA: LdbHLH7, LdbHLH46, LdbHLH118, LdbHLH142, LdbHLH152, LdbHLH170, LdbHLH175, LdbHLH183, LdbHLH189Ethephon: LdbHLH4, LdbHLH7, LdbHLH12, LdbHLH46, LdbHLH76, LdbHLH118, LdbHLH152, LdbHLH170, LdbHLH175, LdbHLH183, LdbHLH189Fig. 10qRT-PCR analysis of 12 LdbHLH genes under ABA, auxin, GA, and MeJA treatments. All experiments were replicated independently three times. Error bars represent the standard deviation of three replicates. Different lowercase letters indicate significant differences (P < 0.05), as determined by one‑way ANOVA followed by Duncan's multiple‑range test ABA: LdbHLH142, LdbHLH152, LdbHLH175, LdbHLH189 ABA: LdbHLH142, LdbHLH152, LdbHLH175, LdbHLH189 Auxin: LdbHLH4, LdbHLH7, LdbHLH12, LdbHLH46, LdbHLH118, LdbHLH142 Auxin: LdbHLH4, LdbHLH7, LdbHLH12, LdbHLH46, LdbHLH118, LdbHLH142 GA: LdbHLH46, LdbHLH76, LdbHLH118, LdbHLH142, LdbHLH152, LdbHLH170 GA: LdbHLH46, LdbHLH76, LdbHLH118, LdbHLH142, LdbHLH152, LdbHLH170 MeJA: LdbHLH7, LdbHLH46, LdbHLH142, LdbHLH152, LdbHLH170, LdbHLH175, LdbHLH189 MeJA: LdbHLH7, LdbHLH46, LdbHLH142, LdbHLH152, LdbHLH170, LdbHLH175, LdbHLH189 SA: LdbHLH7, LdbHLH46, LdbHLH118, LdbHLH142, LdbHLH152, LdbHLH170, LdbHLH175, LdbHLH183, LdbHLH189 SA: LdbHLH7, LdbHLH46, LdbHLH118, LdbHLH142, LdbHLH152, LdbHLH170, LdbHLH175, LdbHLH183, LdbHLH189 Ethephon: LdbHLH4, LdbHLH7, LdbHLH12, LdbHLH46, LdbHLH76, LdbHLH118, LdbHLH152, LdbHLH170, LdbHLH175, LdbHLH183, LdbHLH189 Ethephon: LdbHLH4, LdbHLH7, LdbHLH12, LdbHLH46, LdbHLH76, LdbHLH118, LdbHLH152, LdbHLH170, LdbHLH175, LdbHLH183, LdbHLH189 qRT-PCR analysis of 12 LdbHLH genes under ABA, auxin, GA, and MeJA treatments. All experiments were replicated independently three times. Error bars represent the standard deviation of three replicates. Different lowercase letters indicate significant differences (P < 0.05), as determined by one‑way ANOVA followed by Duncan's multiple‑range test Certain genes exhibited sustained upregulation under specific treatments, including LdbHLH7 under ABA, LdbHLH175 under auxin, LdbHLH189 under GA, and LdbHLH4, LdbHLH12, and LdbHLH76 under SA treatment. To further evaluate responses to abiotic stress, qRT-PCR analysis was conducted under NaCl and drought treatments (Fig. 11). Under NaCl treatment (samples collected at days 0, 2, 4, 6, and 8 of continuous treatment), LdbHLH76 expression was significantly induced, showing a steady increase over time, peaking at day 8. In contrast, the expression of LdbHLH12, LdbHLH46, and LdbHLH189 initially increased but then decreased. Under drought stress (samples collected at days 0, 5, 10, 15, and 20 of water withdrawal), the expression of LdbHLH4, LdbHLH76, and LdbHLH152 was significantly downregulated. In contrast, LdbHLH7, LdbHLH46, LdbHLH142, and LdbHLH170 showed transient upregulation, with the first two peaking on day 10 and the latter two on day 15, followed by a decline.Fig. 11qRT-PCR analysis of 12 LdbHLH genes under SA, ethephon, NaCl, and drought treatments. All experiments were replicated independently three times. Error bars represent the standard deviation of three replicates. Different lowercase letters indicate significant differences among tissues (P < 0.05), as determined by one‑way ANOVA followed by Duncan's multiple‑range test qRT-PCR analysis of 12 LdbHLH genes under SA, ethephon, NaCl, and drought treatments. All experiments were replicated independently three times. Error bars represent the standard deviation of three replicates. Different lowercase letters indicate significant differences among tissues (P < 0.05), as determined by one‑way ANOVA followed by Duncan's multiple‑range test Subcellular localization of LdbHLH12, LdbHLH46, LdbHLH76, and LdbHLH189 To further investigate the functions of LdbHLHs, we focused on LdbHLH12, LdbHLH46, LdbHLH76, and LdbHLH189, which exhibited relatively high expression levels across various tissues and stress conditions, particularly under salt stress, where their expression initially increased and then declined. These four genes were cloned from L. davidii var. unicolor (Figure S1 and Figure S2) and fused with GFP for subcellular localization analysis. The results identified that the green fluorescence signals of LdbHLH12-GFP, LdbHLH46-GFP, LdbHLH76-GFP, and LdbHLH189-GFP were exclusively observed in the nucleus, indicating nuclear localization of these proteins (Fig. 12), consistent with previous predictions based on subcellular localization analyses.Fig. 12Subcellular localization of LdbHLH12, LdbHLH46, LdbHLH76, and LdbHLH189 in tobacco leaf epidermal cells. Each LdbHLH protein was fused to GFP (LdbHLH-GFP), and a nuclear marker (Super1300‑35S‑H2B‑mCherry) was co-expressed as a reference. Scale bar: 25 μm Subcellular localization of LdbHLH12, LdbHLH46, LdbHLH76, and LdbHLH189 in tobacco leaf epidermal cells. Each LdbHLH protein was fused to GFP (LdbHLH-GFP), and a nuclear marker (Super1300‑35S‑H2B‑mCherry) was co-expressed as a reference. Scale bar: 25 μm Overexpression of LdbHLH12, LdbHLH46, LdbHLH76, and LdbHLH189 enhances salt tolerance in transgenic Arabidopsis To further characterize the roles of LdbHLH12, LdbHLH46, LdbHLH76, and LdbHLH189 in salt stress responses, the recombinant plasmids Super1300-LdbHLH12, Super1300-LdbHLH46, Super1300-LdbHLH76, and Super1300-LdbHLH189 were introduced into Agrobacterium tumefaciens strain GV3101 and then transformed into A. thaliana via the floral dip method [29]. Homozygous T3 generation lines were obtained, and successful transgene insertion into A. thaliana was confirmed by PCR (Figure S3). For each construct, at least four independent homozygous transgenic lines were analyzed. The expression levels of these lines were examined by qRT‑PCR, and all four lines showed moderate and comparable expression levels. Therefore, all four lines were used for subsequent phenotypic analyses. The data presented in Fig. 13 represent the means ± SE of four independent lines for each construct, all of which exhibited consistent responses under salt stress. Wild‑type (WT) plants and empty vector (EV) transformants served as controls to exclude any effects due to the vector itself. Root length and fresh weight were compared among transgenic lines (OX-LdbHLH12, OX-LdbHLH46, OX-LdbHLH76, and OX-LdbHLH189), WT plants, and EV controls under varying NaCl concentrations. As NaCl concentration increased, WT and EV plants showed substantially reductions in root lengths and fresh weights compared to the transgenic lines. Notably, under 150 mM NaCl treatment, WT and EV plants exhibited severe wilting, whereas the transgenic plants maintained higher growth vigor. These results indicate that overexpression of LdbHLH12, LdbHLH46, LdbHLH76, and LdbHLH189 enhances salt tolerance in Arabidopsis seedlings (Fig. 13).Fig. 13Overexpression of LdbHLH12, LdbHLH46, LdbHLH76, and LdbHLH189 enhances salt tolerance in transgenic Arabidopsis. Data are presented as mean ± SD (n = 3 independent biological replicates). Different lowercase letters indicate significant differences among treatments (P < 0.05), as determined by one‑way ANOVA followed by Duncan's multiple‑range test Overexpression of LdbHLH12, LdbHLH46, LdbHLH76, and LdbHLH189 enhances salt tolerance in transgenic Arabidopsis. Data are presented as mean ± SD (n = 3 independent biological replicates). Different lowercase letters indicate significant differences among treatments (P < 0.05), as determined by one‑way ANOVA followed by Duncan's multiple‑range test Discussion Increasing evidence has illustrated that plant bHLH genes play central regulatory roles in diverse biological processes, including cell differentiation and proliferation, organ development, light signal transduction, hormone responses, and secondary metabolism, thereby influencing plant life growth throughout the life cycle [30–33]. L. davidii var. unicolor is an important edible lily with high nutritional and medicinal value owing to its rich content of amino acids, proteins, starch, polysaccharides, and bioactive compounds such as lily saponins, flavonoids, and colchicine [34]. Studies have demonstrated that it possesses multiple pharmacological activities, including blood glucose-lowering effects, anti-aging, antioxidant, anti-tumor, and anti-inflammatory, making it an economically important crop with both nutritional and medicinal value that is highly favored by consumers [35]. The development and growth of L. davidii var. unicolor are strongly influenced by environmental conditions. Despite extensive studies of bHLH genes in many plant species, knowledge of this gene family in L. davidii var. unicolor remains limited [36–38]. In this study, 214 bHLH TF-encoding genes were identified in the genome of L. davidii var. unicolor and categorized into 23 subfamilies based on phylogenetic analysis. Each subfamily exhibited highly conserved motif composition and gene structure, consistent with the established classifications in A. thaliana, and comparable to observations in other species such as Passiflora edulis [7]. Motif and gene structure analyses revealed that closely related members shared similar gene architectures, conserved domains, and motifs, indicating that bHLH TFs display substantial structural and functional diversity across plant species, as reported in rice and Arabidopsis [39]. Notably, Motif 2 and Motif 1 were highly conserved across most proteins, forming core components of the bHLH domain and likely contributing to functional specificity [15]. Furthermore, genes in L. davidii var. unicolor were found to contain prominent ultra-long introns, together with a large genome size and exceptionally high content of repetitive sequences, particularly transposable elements (TEs). A similar genomic architecture has been reported in the congeneric species L. regale, with both species sharing features such as extremely large genomes, a high proportion of repetitive sequences, and ultra-long introns [40]. These genomic characteristics are thought to increase the complexity and plasticity of gene regulation, provide abundant material for long-term evolution, and thereby enhance evolutionary potential and environmental adaptability [41]. Homology-based protein modeling further demonstrated that closely clustered gene members share similar secondary and three-dimensional (3D) structural features, whereas LdbHLH proteins from different subfamilies exhibit diverse 3D conformations. In summary, these results indicate that the structural characteristics of LdbHLH proteins within each subfamily are relatively conserved, supporting the phylogenetic classification, while structural diversity among subfamilies likely underlies the functional diversification of the bHLH gene family [42]. Tissue-specific gene expression is a key regulatory mechanism in plant development, enabling precise control of gene activity across different tissues to establish their unique structures and functions, thereby supporting coordinated growth and environmental adaptation [43]. This pattern is also observed in L. davidii var. unicolor, where tissue differentiation is closely associated with the tissue-specific expression of bHLH TFs. For example, LdbHLH170 and LdbHLH175 are highly expressed in bulb scales, which are rich in polysaccharides, amino acids, and flavonoids. Given the involvement of LdbHLH genes in flavonoid biosynthesis, as indicated by GO enrichment analysis, these genes may promote nutrient accumulation and contribute to the formation of medicinally important compounds by regulating flavonoid and polysaccharide biosynthesis. Flower development is regulated by environmental factors, such as temperature and light, as well as internal signaling pathways [44], and bHLH genes are known to play important roles in this process. Thus, LdbHLH76 and LdbHLH189, which show tepal-specific expression, may function as regulators of flower development in this species [45, 46]. Lignin, which is essential for stem strength and lodging resistance, is synthesized under the regulation of bHLH TFs, and GO analysis further suggests the involvement of LdbHLH genes in lignin biosynthesis. Therefore, stem-enriched genes such as LdbHLH4, LdbHLH12, LdbHLH118, and LdbHLH183 may participate in stem development and potentially contribute to the mechanical support required for the tall growth habit of L. davidii var. unicolor [47]. Phytohormones act as key signaling molecules under abiotic stress, integrating external cues, activating stress-responsive genes, and coordinating physiological adjustments to enhance tolerance to salinity and drought [48, 49]. In this study, cis-element analysis revealed that LdbHLH promoters commonly contain both abiotic stress-responsive and hormone-responsive elements, providing a molecular basis for signaling crosstalk. Expression data further support this regulatory interaction. For instance, LdbHLH7 contains ABA- and drought-responsive elements and exhibits sustained upregulation under ABA treatment, along with a transient increase under drought conditions, suggesting its potential role as a downstream mediator of ABA-dependent drought defense responses. LdbHLH76, which contains auxin- and salt-responsive elements, is strongly induced by NaCl treatment and shows transient response to auxin, indicating its potential function in balancing salt tolerance and growth regulation. Similarly, LdbHLH46, which contains GA- and drought-responsive elements, displays transient upregulation under both treatments (peaking at 12 − 24 h under drought), reflecting its potential role in integrating GA and drought signals to modulate growth under water deficit. Taken together, these findings indicate that hormone-stress crosstalk in L. davidii var. unicolor represents an important molecular adaptation, enabling the species to thrive in the arid, semi-arid, and saline environments of Northwest China. Finally, we performed preliminary functional validation of four genes: LdbHLH12, LdbHLH46, LdbHLH76, and LdbHLH189. Subcellular localization analysis confirmed that all four proteins are localized in the nucleus, consistent with prior predictions. To further assess their functions, transgenic Arabidopsis lines overexpressing each gene were generated, and root length and fresh weight were evaluated under media supplemented with different NaCl concentrations. The results demonstrated that overexpression of LdbHLH12, LdbHLH46, LdbHLH76, and LdbHLH189 significantly improved salt tolerance in transgenic seedlings compared with WT and EV controls. These findings indicate that all four genes act as positive regulators of salt stress responses. Taken together, this study provides a foundation for further functional analyses and highlights these genes as promising candidates for molecular breeding and germplasm improvement. Conclusion In this study, 214 bHLH gene family members were identified in the genome of L. davidii var. unicolor and categorized into 23 subfamilies based on phylogenetic relationships with Arabidopsis bHLH proteins. Structural analyses, including gene architecture, homologous protein modeling, and motif composition, revealed that LdbHLH proteins are relatively conserved within each subfamily. Collinearity analysis identified five syntenic gene pairs (LdbHLH3-LdbHLH9, LdbHLH17-LdbHLH61, LdbHLH176-LdbHLH125, LdbHLH181-LdbHLH197, and LdbHLH210-LdbHLH200), all derived from segmental duplication events, indicating evolutionary conservation of this gene family with limited large-scale chromosomal rearrangements or gene duplications. Many LdbHLH genes were predicted to be involved in plant development and to respond to various abiotic stresses, highlighting their potential for coordinated regulatory roles. Promoter analysis further indicated that these genes commonly contain cis-acting elements associated with stress responses, development, and hormone signaling, which may underlie their functional diversity. In addition, interactions among LdbHLH proteins and with other TFs, such as MYBs, suggest the formation of a complex and dynamic bHLH regulatory network. Preliminary functional validation demonstrated that LdbHLH12, LdbHLH46, LdbHLH76, and LdbHLH189 act as positive regulators under salt stress. These findings offer a foundation for further investigation of the regulatory roles of bHLH genes in L. davidii var. unicolor under different environmental conditions. Materials and methods Materials The plant material used in this study was Lilium davidii var. unicolor (Lanzhou lily). Lanzhou lily is a sweet edible lily endemic to China. Bulb materials were obtained from the National Flower Germplasm Resource Repository-National Lily Germplasm Resource Repository of the Beijing Academy of Agriculture and Forestry Sciences, which is consistent with the source used for genome sequencing. The genome sequence and corresponding GFF annotation files of L. davidii var. unicolor were downloaded from the CNCB database (https://ngdc.cncb.ac.cn) [23], and coding sequences were extracted using Gffreads. For comparative analyses, bHLH gene data from the model plant A. thaliana were sourced from the AID database (https://www.arabidopsis.org). Identification, physicochemical property analysis, localization, and synteny analysis of LdbHLH gene family members Candidate bHLH proteins in lily were identified using Arabidopsis bHLH protein sequences in BLASTp alignment with an E-value of 1e−5. The HMM profile of bHLH proteins (PF00010) was downloaded from the Pfam database (http://pfam.xfam.org/) and used for further screening with HMMER, using an E-value threshold of 0.001 [50]. To confirm the presence of the bHLH domain in all candidate genes, their protein sequences were further analyzed using the Pfam and SMART (http://smart.embl.de/) databases with default parameters. The chromosomal locations of the identified LdbHLH genes were visualized using Map Gene to Chromosome (MG2C, http://mg2c.iask.in/mg2c_v2.1). Synteny analysis was conducted using BLASTp and MCScanX [51], and results were visualized with TBtools [52]. The physicochemical properties of LdbHLH proteins, such as isoelectric point, instability index, amino acid length, molecular weight, hydrophilicity, and aliphatic index, were analyzed using the ProtParam tool (https://web.expasy.org/protparam/). Subcellular localization predictions were carried out using WoLF PSORT (https://www.genscript.com/wolf-psort.html). Multiple sequence alignment, subfamily classification, and phylogenetic analysis of LdbHLH genes The amino acid sequences of all LdbHLH and AtbHLH proteins were aligned using MUSCLE [53]. A phylogenetic tree was then constructed using the maximum likelihood (ML) method in IQ-TREE [54], with node support assessed by 1,000 bootstrap replicates, while other parameters set to default. The resulting tree was visualized using ChiPlot (https://www.chiplot.online/), and LdbHLH family members were classified into subfamilies based on the established classification of the AtbHLH gene family. Analysis of LdbHLH gene structure, conserved domains, and motifs The conserved motifs in LdbHLH proteins were identified using MEME, with the number of motifs set to 10 and the motif length constrained to 6 − 100 amino acids, while other parameters were set to default. Exon and intron positions of LdbHLH genes were extracted from GFF annotation files. Amino acid sequences were submitted to NCBI for conserved domain identification, and the results were visualized using TBtools. Conserved domain sequences were further extracted based on a second round of domain annotation, followed by multiple sequence alignment using MUSCLE. Finally, WebLogo (https://weblogo.berkeley.edu/logo.cgi) and GeneDoc software were employed to analyze and visualize the conserved domains. Prediction of cis-acting elements in the promoters of LdbHLH genes The 2,000 bp upstream sequences of the start codon for each LdbHLH gene were extracted and analyzed for cis-acting elements using PlantCARE (https://bioinformatics.psb.ugent.be/webtools/plantcare/html). The types, numbers, and putative functions of the identified elements were statistically analyzed and visualized using TBtools. GO and KEGG analyses of LdbHLH genes The RNA-seq data used in this study were obtained from Li et al. (2023) and downloaded from the NCBI SRA database under accession number PRJNA721267 [55]. The samples consisted of middle leaves of Lilium davidii var. unicolor subjected to three water regimes: control (C, watered every 5 days), moderate drought (MS, watered every 15 days), and severe drought (S, watered every 25 days). Each treatment included three biological replicates, with five pots per replicate. Samples were collected after drought treatment and immediately stored at − 80 °C. Total RNA was extracted using the RNAprep Pure Plant Kit (Tiangen, Beijing), and RNA purity, concentration, and integrity were assessed using a NanoDrop spectrophotometer and an Agilent 2100 Bioanalyzer. mRNA was enriched using oligo(dT) magnetic beads, fragmented, and reverse-transcribed into cDNA. An Illumina sequencing library was constructed, and de novo transcriptome assembly was performed using Trinity software to obtain unigene sequences. The raw sequencing data were deposited in the NCBI SRA database. Functional annotation was performed using the eggNOG database [56]. GO and KEGG enrichment analyses of LdbHLH genes were then conducted with TBtools, and the results were visualized using ggplot2. PPI prediction and protein structure analysis PPI and structural analyses of LdbHLH proteins were conducted as follows. Amino acid sequences were submitted to STRING (http://string-db.org) using A. thaliana orthologs as references. After comparison, Table S4 lists the top 30 results of homology similarity with Arabidopsis. Interaction networks were constructed based on the highest-scoring BLAST hits (bit score), and non-interacting proteins were excluded. Tertiary (3D) structures of LdbHLH proteins were predicted using AlphaFold3, and structural conservation was analyzed using ConSurf (https://consurf.tau.ac.il/). All structural models were visualized with PyMOL. Stress treatments, RNA extraction, and qRT‑PCR To examine the tissue-specific expression of LdbHLH genes, plant materials were obtained from plants cultivated in the greenhouse of Yunnan Agricultural University. The seedlings were cultivated under controlled conditions: a constant temperature of 22 °C, a 16‑h light/8‑h dark photoperiod, and 60–70% relative humidity. Plants were grown in a soil mixture of peat moss, perlite, and vermiculite in a 3:1:1 ratio (v/v/v). All experiments were conducted using four‑week‑old plants at a uniform developmental stage. Healthy, disease-free L. davidii var. unicolor seedlings with well-developed roots and shoots were subjected to abiotic stress and hormone treatments, with three biological replicates per treatment. For drought stress, watering was withheld for 20 days, and samples were collected on days 0, 5, 10, 15, and 20 after the cessation of watering. According to a previous study, 15 days of water withholding induces severe drought stress [55]. For salt stress, seedlings were treated with 125 mmol·L−1 NaCl solution, a concentration previously shown to effectively induce salt stress-related gene expression in lily [57]. The treatment lasted for 8 days, and samples were collected on days 0, 2, 4, 6, and 8 after treatment. For hormone treatments, seedlings were sprayed with 100 μM solutions of ABA, auxin, GA, MeJA, and SA, respectively, prepared in deionized water containing 0.1% (v/v) ethanol or DMSO as the solvent [58]. For ethephon treatment, a 200 mg·L−1 solution was applied as a foliar spray based on previous methods [59], until both adaxial and abaxial leaf surfaces were fully wetted and solution runoff occurred. Leaf samples (the third fully expanded leaf from the apex) were collected at 0, 6, 12, 24, and 48 h post-treatment, rinsed with deionized water, blotted dry, immediately frozen in liquid nitrogen, and stored at − 80 °C until further analysis. Total RNA was extracted using TRIzol reagent (Invitrogen, USA) following the manufacturer’s instructions. First-strand cDNA was synthesized using the ThermoScript RT-PCR kit (Thermo Fisher Scientific, USA). qRT-PCR was performed on a Bio-Rad system using SYBR Premix Ex Taq II (TaKaRa, Japan). The LdFP gene was used as an internal reference for normalization [60]. Each reaction included three biological replicates and three technical replicates. Relative gene expression levels were calculated using the 2−ΔΔCT method. All qRT-PCR primer sequences are listed in Table S3. Subcellular localization The coding sequences of candidate TFs, excluding the stop codon (TGA), were cloned into the Super1300‑eGFP vector (primer sequences listed in Table S3), with the empty Super1300‑eGFP vector serving as a control. Nicotiana benthamiana leaves were co-infiltrated with Agrobacterium tumefaciens strain GV3101 carrying the recombinant plasmid and a nuclear marker (Super1300‑35S‑H2B‑mCherry) at a 1:1 ratio. GFP fluorescence was observed using a confocal laser scanning microscope (LSM800, Zeiss, Germany) with excitation at 488 nm and emission at 500–550 nm. Subcellular localization was determined based on co‑localization of GFP signals with the nuclear marker. Arabidopsis transformation The recombinant plasmids (candidate TF-Super1300) were introduced into Agrobacterium tumefaciens strain GV3101 and used to transform A. thaliana via the floral-dip method [29]. T3 seeds were collected for subsequent analyses. Transgenic plants were selected on ½ MS medium containing 30 mg L−1 kanamycin and further confirmed by colony PCR using the gene-specific primers (Table S3). For stress assays, verified 5-day-old seedlings were transferred to ½ MS medium containing different NaCl concentrations (0, 50, 100, and 150 mM) and grown under controlled conditions (21–23 °C, 150 μmol·m−2·s−1, 8 h/16 h light/dark photoperiod) for 12 days, with fresh weight and root length measured. Control treatments were also performed, and control Arabidopsis plants were used for comparison. All confirmed transgenic lines were subsequently grown in a controlled growth chamber at 24 °C under a 12 h light/12 h dark photoperiod. Statistical analysis Each assay was replicated at least three times. The data are presented as the mean ± standard deviation (SD). After checking normality and homogeneity of variance, the data were analyzed by one‑way ANOVA followed by Duncan’s multiple range test using SPSS (v 18.0; IBM Co., Endicott, NY, USA). Visualization was performed using GraphPad Prism 8.0 (GraphPad Software Inc., USA; http://www.graphpad.com/). Differences were considered statistically significant at P < 0.05. Supplementary Information Supplementary Material 1. Supplementary Material 2. Supplementary Material 1. Supplementary Material 2.