For drought treatment, two-year-old S. baicalensis plants were cultivated in 7.5 × 45 cm nutrient pots under controlled greenhouse conditions. The culture parameters were set to a 24-hour cycle with the following settings: from 0 to 8 hours, a temperature of 15°C, light intensity of 0 lux, and 50% humidity; from 8 to 24 hours, a temperature of 22°C, light intensity of 30,000 lux, and 50% humidity. We determined the most suitable time and the most suitable concentration of polyethylene glycol (PEG) solution through the experiments of different drought stress time and different drought stress degree in the early stage.The treatment group was irrigated with 120 mL of 20% PEG solution, while the control group received an equal volume of distilled water. Plants were harvested 24 hours after treatment. A total of nine samples were collected from each group, with every three samples pooled together. The pooled samples were wrapped in aluminum foil, snap-frozen in liquid nitrogen, and stored at -80°C. Each treatment was conducted with three biological replicates.

S. baicalensis genome data (GWHAOTO00000000) were obtained from CNCB (https://www.cncb.ac.cn/, accessed on 8 April 2024) (Xu et al., 2020). To identify potential bHLH genes, two complementary approaches were employed. First, the Hidden Markov Model (HMM) for the bHLH domain (PF00010) was retrieved from the InterPro database (https://www.ebi.ac.uk/interpro/, accessed on 10 April 2024), and the S. baicalensis protein sequence database was scanned using HMMER 3.0 (Huang et al., 2024). Second, A. thaliana bHLH protein data were downloaded from the Plant Transcription Factor Database (PlantTFDB, https://planttfdb.gao-lab.org/, accessed on 10 April 2024) (Mohammad et al., 2023). BLAST 2.15.0 was performed using A. thaliana bHLH protein sequences as queries, with an e-value <1^e−5, to identify potential bHLH protein sequences in S. baicalensis. Redundant protein IDs and sequences were removed, and all identified candidates were further verified using the Pfam (https://www.ebi.ac.uk/interpro/, accessed on 15 April 2024) (Mohammad et al., 2023), SMART (https://smart.embl.de/, accessed on 16 April 2024) (Huang et al., 2024), and Batch CD-search (https://www.ncbi.nlm.nih.gov/cdd/, accessed on 15 April 2024) databases. The physicochemical properties of SbbHLH proteins were analyzed using the ExPASy (https://web.expasy.org/protparam/, accessed on 6 May 2024) (Zhang et al., 2015). Subcellular localization of the SbbHLH proteins was predicted using multiple tools, including WoLF-PSORT (https://wolfpsort.hgc.jp/, accessed on 6 May 2024) (Zhou et al., 2020), CELLO v.2.5 (http://cello.life.nctu.edu.tw/, accessed on 18 November 2024) (Huang et al., 2024), and BUCSA (https://busca.biocomp.unibo.it/, accessed on 18 November 2024) (Liang et al., 2023).

The SbbHLH and AtbHLH protein sequences were aligned using MAFFT v7.520. A phylogenetic tree was then constructed using the Maximum Likelihood (ML) method with 1000 bootstrap replicates and visualized using iTOL (https://itol.embl.de/, accessed on 20 May 2024). The ka/ks ratios were calculated with ka/ks Calculator 2.0 to assess natural purifying selection among target gene pairs. The SbbHLH genes of S. baicalensis were mapped to nine chromosomes using the genome data. Additionally, the collinear relationships between chromosome pairs were analyzed using TBtools v2.025 (Chen et al., 2023).

Gene structure annotation was obtained using the GFF3 file of the S. baicalensis genome, and visualized using TBtools v2.025 (Chen et al., 2023). Conserved motifs within the SbbHLH proteins were identified using the MEME (https://meme-suite.org/meme/ accessed on 8 June 2024). The 2kb region upstream of the start codon of the SbbHLH gene in the S. baicalensis genome was used as the promoter sequence and subjected to analysis in the PlantCARE database (https://bioinformatics.psb.ugent.be/webtools/plantcare/html/ accessed on 10 June 2024) to predict potential cis-elements.

The transcriptome data of S. baicalensis in different tissues were downloaded from public resources (https://ngdc.cncb.ac.cn/bioproject/browse/PRJCA009556 accessed on 30 April 2024) (Hu et al., 2023). The sequence reads were processed using FastQC 0.11.8 to remove adapters and low-quality sequences, then assembled using HISAT2 2.1.0. The expression patterns of SbbHLH genes were analyzed using normalized values expressed as Fragments Per Kilobase Million (FPKM). A heatmap was generated to visualize these expression patterns, with Z-score normalization applied to the data.

Accurately weigh 0.5 g of fresh S. baicalensis root samples. Following the instructions provided in the assay kits from the Nanjing Jiancheng Bioengineering Institute, measure the activities of superoxide dismutase (SOD, A001-1-1) and peroxidase (POD, A084-3-1), as well as the concentrations of proline (A107-1-1) and soluble protein (A045-2-2). To determine the baicalin content, The root tissues of S. baicalensisto a constant weight. A 0.3 g sample of the dried powder was then combined with 10 mL of 70% ethanol. Microwave extraction was performed at 80°C for 6 minutes using the MARS6 system (CEM, USA). Following extraction, filter the solution and dilute it to a final volume of 50 mL with 70% ethanol. Pass 2 mL of the diluted solution through a 0.22 μm filter membrane to prepare the sample for analysis. The baicalin content was determined using high-performance liquid chromatography (HPLC, Agilent 1100, USA) according to the elution protocol described by Han et al (Zhong-Ming et al., 2011). The concentration was calculated based on the standard curve equation: y = 2626.76x − 96.06 (R² = 0.9997). Each sample was measured in triplicate, and the average value was reported as the final result, with the standard deviation represented as the error bar.

RNA was extracted from the roots of both control and treatment groups of S. baicalensis. To ensure RNA quality, the total RNA was assessed and quantified using the Qubit fluorescence quantifier and the Qsep400 high-throughput biological fragment analyzer. Breaks intact RNA into short RNA fragments. Then reverse transcribed to form cDNA. The cDNA was purified and Polymerase Chain Reaction (PCR) amplification to complete the construction of cDNA library. After sequencing, the raw data were obtained. The raw data were filtered using FastQC 0.23.2 and Trimmomatic 0.36.5, and assembled using HISAT2 2.1.0. Normalization of the number of mapped reads against transcript length facilitated the calculation of gene alignment using FeatureCounts 1.6.2. Changes in gene expression were determined based on FPKM values.

The protein interaction network analysis of the differentially expressed genes was performed using the STRING database (https://cn.string-db.org/, accessed on July 27, 2024). Homologous genes from A. thaliana were selected to construct the interaction network, with a prediction score > 0.7 and a p-value < 1^e-10. To validate the transcriptome data, 10 SbbHLH genes were randomly selected from the transcriptome data of S. baicalensis. Primers for each SbbHLH gene were designed using Primer Premier 5.0 ( Supplementary Table S1 ). Quantitative real-time PCR (qRT-PCR) was performed to analyze these genes, with 18S rRNA serving as the reference gene. The qRT-PCR reaction mixture consisted of 10 µL of SYBR Premix Ex Taq (RR820 Takara, China), 1 µL of each primer, 1 µL of cDNA template, and 7 µL of ddH₂O, making a total volume of 20 µL. The PCR protocol started with a pre-denaturation step at 95°C for 30 s, followed by 40 cycles of denaturation at 95°C for 5 s, annealing at 56°C for 30 s, and extension at 72°C for 15 s. Finally, the 2^-ΔΔCt method was used to normalize the experimental results, with 18S rRNA as the internal reference gene. To ensure the reliability of the data, each sample was repeated three times. Statistical analysis was performed using one-way analysis of variance (ANOVA), and different lowercase letters indicate significant differences (p < 0.05).

To identify bHLH genes on a genome-wide scale in S. baicalensis, we used the AtbHLH proteins as queries and applied the HMM model containing the bHLH domain to search the complete amino acid database of S. baicalensis. After removing redundant sequences, the candidate SbbHLH genes were screened using the SMART, Batch CD-search, and Pfam databases, resulting in the identification of 142 SbbHLH genes ( Supplementary Table S2 ). Based on their chromosomal locations, the 142 SbbHLH genes were renamed sequentially (SbbHLH1-SbbHLH142) ( Supplementary Table S3 ). Analysis of their physicochemical properties revealed that the amino acid (aa) lengths of the SbbHLH proteins ranged from 91 (SbbHLH97) to 912 (SbbHLH56), with molecular weights (MW) varying from 10,431.7 Da (SbbHLH97) to 99,345.31 Da (SbbHLH56). The isoelectric points (pI) of these proteins ranged from 4.76 (SbbHLH104) to 11.54 (SbbHLH134), with an average pI of 6.94. Notably, 90 SbbHLH proteins had pI values below 7, classifying them as acidic, while 52 proteins exhibited pI values above 7, indicating they are basic. Subcellular localization predictions showed that 134 SbbHLH genes were localized to the nucleus, while the remaining proteins were distributed across other cellular compartments: chloroplasts (SbbHLH42, SbbHLH43, SbbHLH58, SbbHLH66, SbbHLH69, SbbHLH83), extracellular space (SbbHLH95), and mitochondria (SbbHLH52).

A phylogenetic tree was constructed using the protein sequences of 142 SbbHLH genes and 161 AtbHLH genes. The resulting phylogenetic tree was divided into 16 branches based on the homology of bHLH proteins ( Figure 1 ). Significant variation was observed in the number of genes within each branch; branch 4 contained the highest number of members, with 25 SbbHLH genes (17.6%), while branch 7 had the fewest, comprising only one SbbHLH gene. Notably, each branch included bHLH genes from both S. baicalensis and A. thaliana, highlighting the evolutionary conservation of bHLH genes across species. Since genes within the same branch are likely to share similar functions, this phylogenetic analysis provides a foundational framework for predicting the potential functions of SbbHLH genes.

We constructed a phylogenetic tree to elucidate the relationships among all SbbHLH genes ( Figure 2A ). Based on the topology of the tree, we further analyzed conserved motifs and gene structures. A total of 10 conserved motifs were identified ( Figure 2B ). Notably, motifs 1 and 2 were present in all SbbHLH genes, with these two motifs constituting the complete bHLH domain. Specifically, motif 1 corresponds to the partial loop and helix 2 region of the bHLH domain, while motif 2 includes the basic region, helix 1, and part of the loop region. Similar motifs were typically distributed among genes within the same branch. For example, the branch containing SbbHLH112, SbbHLH113, SbbHLH111, and SbbHLH142 displayed motifs 1, 2, 6, 7, and 10. Whereas the branch comprising SbbHLH5, SbbHLH1, SbbHLH70, SbbHLH97, and SbbHLH138 only contained motifs 1 and 2. This distinct motif distribution suggests that these branches may have divergent functions, providing insights into the conservation and diversity of SbbHLH genes while supporting the phylogenetic tree classification. Gene structure analysis provides valuable insights into the phylogenetic relationships among SbbHLH genes. Among the 142 SbbHLH genes analyzed, the number of exons ranged from 1 to 14 ( Figure 2C ), with the majority containing between 2 and 8 exons. Notably, SbbHLH51 exhibited the highest exon count at 14, while several genes, including SbbHLH11, SbbHLH12, SbbHLH42, SbbHLH43, SbbHLH83, SbbHLH86, SbbHLH101, and SbbHLH123, each contained only a single exon. Such variations likely reflect the unique functional roles and evolutionary trajectories of these genes, suggesting diverse regulatory mechanisms within the SbbHLH gene family. Moreover, the similarity in the number and length of exons within the same branch further supports the phylogenetic tree classification.

Chromosome localization analysis revealed that the 142 SbbHLH genes are distributed across all nine chromosomes of S. baicalensis ( Figure 3A ). Chromosome 1 contains the highest number of SbbHLH genes (26, 18.30%), while chromosome 6 has the fewest (7, 0.05%). Gene duplication events are known to facilitate the evolution and expansion of gene families in plants. To investigate the gene duplication events within the SbbHLH gene family, we analyzed a total of 142 SbbHLH genes ( Figure 3A ). Our analysis identified 26 segmental duplication events and 3 tandem duplication events. These findings were further supported by phylogenetic analysis, indicating the critical role of segmental duplication in the evolution of the SbbHLH gene family. Additionally, a ka/ks analysis was performed on 29 gene pairs ( Supplementary Table S4 ). The results showed that, except for the SbbHLH48/49 gene pair, which is under positive selection, the remaining 28 gene pairs underwent purifying selection. This further supports the notion that the SbbHLH gene family is crucial for environmental adaptation and the emergence of new functions. To clarify the phylogenetic relationships among bHLH gene families across different species, we constructed a collinearity map comparing the bHLH genes of A. thaliana, S. barbata, and S. baicalensis ( Figure 3B ). The analysis identified 122 homologous gene pairs between S. baicalensis and A. thaliana and 172 homologous gene pairs between the two Scutellaria species. The higher number of homologous gene pairs within Scutellaria indicates a closer evolutionary relationship between these species. Further analysis revealed that S. baicalensis, A. thaliana, and S. barbata had 121, 115, and 78 homologous genes, respectively. Notably, 70 homologous bHLH genes are shared among all three species, accounting for 49.30% of the total SbbHLH genes in S. baicalensis.

A total of 3,521 cis-elements were predicted across the 142 SbbHLH genes ( Figure 4 ). These cis-elements were grouped into 21 types, which were further categorized into three main functional groups: growth and development, hormone response, and stress response. Among them, 2,081 cis-elements (59.10%) were associated with growth and development, followed by 910 (25.85%) linked to hormone responses and 530 (15.05%) related to stress responses. Notably, 138 of the 142 SbbHLH genes contained hormone-responsive elements, including those responsive to salicylic acid, abscisic acid, and gibberellins. This suggests that hormonal regulation plays a significant role in the expression of SbbHLH genes. Notably, although stress response elements constituted a smaller proportion overall, 139 SbbHLH genes were found to contain such elements, including those responsive to anaerobic conditions, low temperatures, and MYB-drought inducibility. Among them, SbbHLH128, SbbHLH71, and SbbHLH135 had the largest number of stress response elements, with 30, 18, and 17 elements, respectively. Furthermore, seven SbbHLH genes were identified to contain MYB regulatory elements associated with flavonoid biosynthesis.

Baicalin is the primary medicinal component of S. baicalensis and classified as an RSF, and is predominantly synthesized in the root tissue. To explore the expression patterns of SbbHLH genes and identify potential genes involved in baicalin biosynthesis, we analyzed transcriptomic data from different tissues of S. baicalensis. Principal component analysis (PCA) revealed that the expression profiles of SbbHLH genes formed distinct clusters corresponding to the four tissue types ( Supplementary Figure S1 ). This clustering was primarily driven by specific SbbHLH genes with high contributions to the principal components, indicating their tissue-specific expression patterns. Almost all SbbHLH genes were expressed in the four tissue types, with the exception of six genes (SbbHLH19, SbbHLH22, SbbHLH84, SbbHLH103, SbbHLH104, and SbbHLH120). We generated a heatmap using the FPKM values of the remaining 136 SbbHLH genes, which were not uniformly expressed across different tissues ( Figure 5 ). Notably, the number of differentially expressed genes (DEGs) was highest in the flowers compared to the root, stem, and leaf tissues. In flower tissue, we identified 13 significant DEGs, of which 9 were upregulated and 4 were downregulated. In root tissue, we found 10 upregulated and 2 downregulated genes, totaling 12 DEGs. In stem tissue, 8 DEGs were observed, while only 2 DEGs were found in leaves. To validate our findings, we measured the expression levels of 10 SbbHLH genes in different tissues by qRT-PCR ( Supplementary Table S1 ), which further confirmed the stability of the transcriptome data ( Supplementary Figure S2 ).

To explore the effects of drought stress on the physiological and pharmacological components of S. baicalensis. we measured the activities of SOD and POD, as well as the contents of proline and Soluble Protein ( Figure 6A, B ). The results indicated that the activities of SOD and POD increased significantly under drought conditions, rising by 159.61% and 29.76%, respectively, compared to the control group. Similarly, the contents of proline and Soluble Protein also exhibited significant increases, with proline rising by 0.9-fold and Soluble Protein by 0.6-fold relative to the control group. Additionally, we assessed the changes in baicalin content under drought treatment. The results demonstrated a significant increase in baicalin content as well ( Figure 6C ). This increase in baicalin content is particularly beneficial for enhancing the quality of S. baicalensis as a medicinal material. We further analyzed the expression patterns of SbbHLH genes and baicalin biosynthesis genes under drought stress using transcriptomic data. A total of 82 SbbHLH genes and 22 baicalin biosynthesis genes were identified in both the control and drought-treated groups. After normalizing the FPKM values for these genes, a clustered heatmap was generated ( Figure 7A ). The results showed that the CYP82D-1 and CCL-7 genes were upregulated under drought stress, along with 30 other SbbHLH genes exhibiting similar expression trends. This suggests that CYP82D-1 and CCL-7, together with the 30 SbbHLH genes grouped into one class, may be involved in the biosynthesis of baicalin under drought stress. We divided the 15,179 genes into 11 distinct modules using weighted correlation network analysis (WGCNA) ( Figure 7B ). The turquoise module, containing 8,839 genes, is the largest, followed by the blue module with 4,981 genes. The purple module is the smallest, with only 75 genes. Notably, the expression pattern of genes in the blue module was strongly positively correlated with baicalin content, with correlations were both greater than 0.85 and p-values<0.01 ( Figure 7B , Supplementary Figure S3 ). Therefore, we analyzed the genes in the blue module and searched for matches based on the original IDs of the 142 SbbHLH genes we identified. Among the blue module genes, we found 32 SbbHLH genes. Additionally, sequence alignment revealed three genes involved in baicalin biosynthesis ( Supplementary Table S5 ). To validate the expression patterns of these genes, we performed qRT-PCR on 10 SbbHLH genes, which confirmed the reliability of the transcriptome data ( Supplementary Figure S4 ).

Baicalin is one of the primary medicinal components of S. baicalensis. Our previous research demonstrated that moderate drought stress activates the stress response in S. baicalensis, inducing physiological changes that enhance baicalin content (Cheng et al., 2018; Cheng et al., 2023). We analyzed the promoter regions of CYP82D-2, CYP82D-1, and CCL-7, which are involved in the biosynthesis of the major secondary metabolite baicalin within the blue module. Our analysis revealed that the promoter regions of these genes in the baicalin biosynthetic pathway contain cis-elements associated with bHLH transcription factors ( Figure 8A ), suggesting that CYP82D-2, CYP82D-1, and CCL-7 may be regulated by SbbHLH genes. We merged the 30 SbbHLH genes with expression patterns similar to CYP82D-1 and CCL-7 obtained from the clustered heatmap with the 32 SbbHLH genes from the blue module. After removing duplicates, a total of 36 unique SbbHLH genes were identified. Subsequently, we constructed a protein interaction network using the 36 SbbHLH genes and 3 baicalin biosynthesis pathway genes ( Supplementary Table S6 ; Figure 8B ; Supplementary Figure S5 ). The results showed that the CYP82D-1 interacts with SbbHLH142 and SbbHLH98, while CCL-7 interacts with SbbHLH74. Notably, SbbHLH53 emerged as a central node in the protein interaction network, interacting with multiple proteins, including SbbHLH142, SbbHLH140, SbbHLH20, SbbHLH85, SbbHLH40, SbbHLH68, SbbHLH35, and SbbHLH125. Phylogenetic analysis revealed that SbbHLH53 in S. baicalensis shares a high degree of homology with AtbHLH112 in A. thaliana. Previous studies have shown that, under drought conditions, AtbHLH112 in A. thaliana is induced by drought stress and then regulates the expression of genes related to SOD, POD and proline metabolism. Based on previous studies and the findings of this study, we propose that the SbbHLH gene regulates the baicalin biosynthetic pathway under drought stress. Specifically, when S. baicalensis is subjected to drought stress, its physiological state undergoes significant changes, including alterations in SOD activity, POD activity, and proline content. At the genetic level, the expression of SbbHLH53 is upregulated under drought stress, activating a regulatory network centered on this transcription factor and ultimately promoting baicalin biosynthesis. Within this complex regulatory framework, SbbHLH74 may positively regulate the expression of CCL-7. Meanwhile, SbbHLH98 and SbbHLH142 are likely to enhance the expression of CYP82D-1 (F6H), which accelerates the enzymatic conversion of chrysin to baicalein through the action of F6H. Ultimately, these interactions and regulatory mechanisms contribute to an increased biosynthesis rate of baicalin under drought stress conditions.

The bHLH gene family play crucial roles in signal transduction networks (Seo et al., 2011; Zhang et al., 2020; Yin et al., 2023). The identification of transcription factor families has been facilitated by the rapid increase in plant genomic data. The bHLH genes had been comprehensively analyzed in multiple species, including Salvia miltiorrhiza ( Zhang et al., 2015), Ginkgo biloba ( Zhou et al., 2020), A. thaliana ( Li et al., 2006), Juglans mandshurica ( Li et al., 2024), Passiflora edulis ( Liang et al., 2023) and Citrus sinensis ( Huang et al., 2024), underscoring the significance of this transcription factor family. However, comprehensive information regarding bHLH gene family in S. baicalensis remains scarce. Therefore, we identified and analyzed 142 SbbHLH genes in S. baicalensis. This number is comparable to that reported in P. edulis (138) (Liang et al., 2023) and C. sinensis (135) (Huang et al., 2024), fewer than in A. thaliana (161), and more than in G. biloba ( Zhou et al., 2020) and S. miltiorrhiza ( Zhang et al., 2015). The genome size and chromosome number of different species may lead to differences in the number of bHLH genes (Flagel and Wendel, 2009; Kamran et al., 2023).

The phylogenetic tree, comprising the 142 SbbHLH genes and 161 AtbHLH genes, showed that each SbbHLH gene clustered with at least one AtbHLH gene, forming sister branches ( Figure 1 ). This clustering suggests that bHLH genes diversified prior to species evolution, thereby enhancing the reliability of classification and functional predictions for SbbHLH genes in S. baicalensis ( Yao et al., 2024). The analysis of the conserved motifs in the 142 SbbHLH genes revealed that genes with similar conserved motifs are generally grouped within the same branch of the phylogenetic tree. This suggests that members within the same branch may share a common ancestor and perform similar functions, further supporting the phylogenetic relationships within S. baicalensis ( Figure 2A ). All SbbHLH genes contained motifs 1 and 2, indicating that these motifs represent the conserved domains of bHLH genes, forming the DNA-binding domain and the protein dimerization region (Kamran et al., 2023). In contrast, However, the distribution of motifs 2-10 was random but conserved within individual branches, reflecting the specific functions of genes within those branches. Among the 142 SbbHLH genes, the highest number of exons observed was 14, while 134 genes have more than two exons ( Figure 2B ). These results suggest that exon insertions or deletions may have occurred during evolution, contributing to the expansion of the SbbHLH gene family (Li et al., 2024).

The 142 SbbHLH genes are unevenly distributed across the nine chromosomes of S. baicalensis ( Figure 3A ), suggesting their involvement in evolutionary development (Reams and Neidle, 2004). Gene duplication events are common in transcription factor families, enabling rapid expansion and contraction in response to environmental changes. This process contributes to the quantitative variation of transcription factor families and promotes genetic diversity (Guo et al., 2024). Gene duplications are typically categorized into segmental and tandem duplications, with segmental duplications being particularly prevalent in plant transcription factor families (Yao et al., 2024). As anticipated, we identified three tandemly duplicated and 26 segmentally duplicated gene pairs in S. baicalensis, suggesting that segmental duplication likely plays a pivotal role in the evolution of bHLH genes and the development of novel functions for environmental adaptation ( Figure 3A ). Similar conclusions in other species, such as J. mandshurica and C. praecox, further support the idea that segmental duplication is a common mechanism driving the expansion of bHLH gene families (Kamran et al., 2023; Li et al., 2024). Collinearity analysis provided further insights into the origin and evolution of the bHLH gene family ( Figure 3B ). We identified 122 gene pairs between S. baicalensis and A. thaliana, and 172 gene pairs between Scutellaria species. This indicates a closer phylogenetic relationship within Scutellaria, while still showing 48.45% similarity with A. thaliana. This suggests that bHLH gene diversity may have emerged early in terrestrial plants. The ka/ks analysis revealed that SbbHLH genes primarily underwent purifying selection, reflecting highly conserved evolution ( Supplementary Table S4 ) (Zhan et al., 2023). Additionally, we identified 70 collinear genes among S. baicalensis, S. barbata, and A. thaliana, which are crucial for understanding evolutionary dynamics.

The growth and development of plant are complex processes influenced by both external environmental factors and internal hormone levels (Li et al., 2024). To further clarify the role of SbbHLH genes in S. baicalensis, we identified the cis-elements of 142 SbbHLH genes. The results showed that SbbHLH genes play significant roles in stress tolerance, hormone responses, and growth and development in S. baicalensis ( Figure 4 ). Changes in transcription levels are a primary mode of regulation in organisms. We further examined the expression patterns of these 142 SbbHLH genes in different tissues ( Figure 5 ). Our findings revealed that 136 SbbHLH genes were expressed in at least one tissue, with varying expression levels among different tissues, indicating tissue-specific expression of SbbHLH genes. Notably, 12 genes were highly expressed in the roots, suggesting their potential roles in root growth and development (Zhao et al., 2016; Cheng et al., 2018).

Given the importance of S. baicalensis for its antibacterial and anti-inflammatory properties (Wang et al., 2022). Moreover, the yield of S. baicalensis and the content of baicalin are closely linked to its market price, making the understanding of baicalin biosynthesis mechanisms a current research hotspot. Our study demonstrated that appropriate drought treatment could stimulate the production of secondary metabolites in S. baicalensis, including baicalin and baicalein (Cheng et al., 2018; Cheng et al., 2023). Under drought treatment, the physiological state of S. baicalensis undergoes significant changes, including the active removal of harmful substances through increased activity of protective enzyme, and osmotic adjustment via elevated proline and soluble protein content ( Figure 6B ). The content of the secondary metabolite baicalin also significantly increased, indicating that secondary metabolites help resist drought stress ( Figure 6C ). A substantial body of literature shows that these processes involve complex signal transduction between the environment, hormones, and transcription factors, ultimately altering the activity of downstream key enzymes and thereby affecting the biosynthesis of plant secondary metabolites (Guo et al., 2024). The bHLH gene family plays a crucial role in plant stress resistance. For example, overexpression of MdCIB1 (cryptochrome-interacting bHLH1) in Malus domestica callus enhanced drought tolerance to PEG6000, while MdCIB1 transgenic A. thaliana exhibited improved root development under drought stress (Ren et al., 2021; Li et al., 2024).

To investigate the potential roles of SbbHLH genes in regulating baicalin biosynthesis under drought treatment, we performed transcriptome data and qRT-PCR analysis on root tissues of S. baicalensis subjected to drought treatment and control treatments. To investigate the potential role of SbbHLH genes in regulating baicalin biosynthesis under drought stress, we performed transcriptome and qRT-PCR analyses on root tissues of S. baicalensis subjected to drought stress and control treatments. A total of 36 SbbHLH genes exhibited expression patterns consistent with the trend in baicalin content ( Figure 7A, B ). Additionally, we identified cis-elements associated with drought stress and hormone responses in the promoter regions of SbbHLH genes, further supporting their role in regulating the stress response in S. baicalensis ( Figure 4 ). Protein interaction analysis suggested that SbbHLH53 may act as a central component of the regulatory network under drought stress, modulating the expression of downstream genes, including CCL7 and CYP82D-1, in response to drought conditions and hormonal changes ( Figure 8B , Supplementary Figure S5 ). This suggests that these bHLH genes are part of a complex cross-regulatory network involving both drought stress and hormone induction. For instance, ScbHLH5 and ScbHLH65 exhibited significant positive correlations under various hormonal inductions in Secale cereale, indicating a synergistic regulatory effect between different endogenous hormonal metabolic pathways (Chen et al., 2024). Given the conservation of the bHLH gene family, we can predict the functions of SbbHLH genes through homology analysis (Feller et al., 2011b). Many SbbHLH genes have been functionally validated in A. thaliana, allowing us to infer their roles in S. baicalensis. The homologous gene of SbbHLH53 in S. baicalensis is AtbHLH112, which has been shown to mediate physiological responses at the transcriptional level, including proline biosynthesis and ROS scavenging, thereby enhancing the salt and drought tolerance of A. thaliana ( Liu et al., 2015).

The bHLH genes may exhibit complex regulatory mechanisms in response to abiotic stress and function by forming both homologous and heterologous complexes (Guo et al., 2024). Previous studies have indicated that the HLH domain facilitates dimer formation, with specific amino acid residues in the helix regions being essential for this process (Carretero-Paulet et al., 2010). Since the biosynthesis pathways of anthocyanins and baicalin share similarities, it is plausible that the SbbHLH gene may also form complexes through protein interactions, playing a crucial role in regulating downstream gene expression in response to drought stress signals. Our study revealed that 36 SbbHLH genes exhibited elevated expression under drought stress. Notably, we identified binding sites for bHLH genes in the promoter regions of key genes involved in baicalin biosynthesis ( Figure 8A ). The biosynthesis of baicalin may be influenced by drought stress, which can increase the expression of SbbHLH genes and consequently the content of baicalin (Yang et al., 2022; Cheng et al., 2023). Based on our analysis and previous studies, we hypothesize that SbbHLH74, SbbHLH98, and SbbHLH142 play crucial roles in regulating baicalin biosynthesis. Specifically, under drought stress, the regulatory network centered on the SbbHLH53 gene is activated, influencing the expression of SbbHLH74, SbbHLH98, and SbbHLH142, which in turn enhances the activities of CYP82D-1 and CCL-5 and promotes baicalin biosynthesis. Our study aims to further investigate the target genes that regulate baicalin biosynthesis to clarify the role of SbbHLH genes in this process. In future studies, we will validate the functions of SbbHLH genes and their regulatory mechanisms under drought stress.