Abstract Aims Scutellaria baicalensis Georgi is a commonly utilized bulk Chinese medicinal material in China. In clinical practice, it is often divided into ZiQin and KuQin according to their growth years and different medicinal purposes. Extensive herbal research and pharmacological studies have demonstrated distinct effects of ZiQin and KuQin. Therefore, we collected HuangQin along with its rhizosphere soils at different growth stages to investigate the impact of soil microorganisms on flavonoid synthesis in HuangQin. Methods In this study, high-throughput sequencing and UPLC-MS/MS-based metabolomics method were employed for the analysis of 16S rDNA sequences in HuangQin rhizosphere soil samples and metabolic profiling of HuangQin, respectively. Results The results revealed that the number of operational taxonomic units (OTUs) for the four years were 7594, 10,227, 10,280, and 9796, respectively. Furthermore, an extended cultivation period led to a gradual decrease in the abundance of Pseudarthrobacter, Achromobacter and other beneficial bacteria. In addition to this, a total of 552 secondary metabolites were identified in the metabolome. The analysis revealed that 56 components could be successfully mapped to the KEGG database, encompassing five distinct metabolic pathways: isoflavonoid biosynthesis, metabolic pathways, biosynthesis of secondary metabolites, flavonoid biosynthesis and flavone and flavanol biosynthesis. Correlation analysis between soil physicochemical properties and differential microorganisms demonstrated significant associations: Pseudarthrobacter exhibited strong correlations with TN (total nitrogen), AN (ammonium nitrogen), AK (available potassium), and NH[4]^+-N; Nocardioides displayed notable correlations with TN and AK as well as a significant association with AN and pH; Haliangium showed a significant correlation with AP (available phosphorus). Conclusion The results suggest that flavonoid accumulation in S. baicalensis may be influenced by soil physicochemical properties and rhizosphere microorganisms, providing a valuable basis for future research on optimizing cultivation practices. Supplementary Information The online version contains supplementary material available at 10.1186/s12866-025-04229-4. Keywords: Scutellaria baicalensis georgi, Rhizosphere microbiome, 16S rDNA sequencing, Soil microbial diversity, Targeted metabolomics, Flavonoid dynamics Introduction Scutellaria baicalensis Georgi (S. baicalensis) is a perennial herb commonly referred to as ‘HuangQin’, highly esteemed for its desiccated roots according to the Chinese Pharmacopoeia Commission [[42]1]. Modern pharmacological investigations have unveiled a wide range of therapeutic properties associated with HuangQin, encompassing its anti-inflammatory, anti-tumor, antiviral, and antimicrobial effects [[43]2–[44]4]. Within traditional Chinese medicine, HuangQin roots aged two to three years are referred to as ‘ZiQin’, while those surpassing three years are termed ‘KuQin’ [[45]5, [46]6]. Interestingly, classical medical texts have reported notable differences between ZiQin and KuQin in both physical characteristics and therapeutic efficacy [[47]7, [48]8]. ZiQin typically exhibits no hollow decay, whereas KuQin progressively develops a decaying area and that darkens over time [[49]9]. The principal bioactive constituents of S. baicalensis are flavones, including baicalin, wogonoside, baicalein, and wogonin [[50]10–[51]12]. Variations in the types and ratios of these constituents, primarily flavones, between ZiQin and KuQin in vivo suggest potential disparities in therapeutic efficacy [[52]13]. Considering that divergent pharmacological effects of HuangQin stem solely from differing growth periods, we conducted a comprehensive collection of S. baicalensis roots and their rhizosphere soil across various growth years to investigate the influence of soil microorganisms on the synthesis of medicinal compounds. The biosynthetic pathway of flavones in HuangQin has been extensively elucidated, primarily through the root-specific flavone pathway [[53]14]. The root specific flavone synthesis pathway originates from cinnamic acid (Fig. [54]1), which is derived from the amino acid phenylalanine as a biosynthetic precursor under the catalysis of phenylalanine ammonia lyase (SbPAL). Subsequently, cinnamic acid undergoes enzymatic conversion to from cinnamoyl -coenzyme A (cinnamoyl CoA) through the action of cinnamate-CoA ligase (SbCLL-7). The pinocembrin-chalcone synthetase (SbCHS-2) facilitates the conversion of cinnamoyl CoA into pinocembrin chalcone, which subsequently undergoes isomerization mediated by chalcone isomerase (SbCHI) to yield pinocembrin. And then, pinocembrin is converted to chrysin by flavone synthetase (SbFNSII-2), followed by hydroxylated of chrysin to baicalein by flavone 6-hydroxylase (SbF6H) [[55]15]. Baicalein is glucuronidated to baicalin by baicalin-7-O-glucuronosyltransferase (UBGAT), while chrysin is converted to norwogonin by flavone 8-hydroxylase (SbF8H). Norwogonin undergoes O-methylation at position 8 catalyze by O-methyl transferase (OMT) to yield wogonin, which ultimately undergoes glucuronidation to form wogonoside mediated by baicalin-7-O-glucuronosyltransferase (UGAT) [[56]16]. Fig. 1. [57]Fig. 1 [58]Open in a new tab Root specific flavone synthesis pathway in HuangQin Increasingly, investigations have revealed the pivotal role of soil factors in shaping the growth and quality attributes of medicinal plants [[59]17]. Notably, distinct nitrogen utilization patterns have been associated with the modulation of secondary metabolism networks, particularly flavonoid biosynthesis in Glycyrrhiza uralensis Fisch [[60]18]. Additionally, the microbial diversity in the rhizosphere has been demonstrated to confer advantageous effects on plant growth, development, and the accumulation of bioactive compounds [[61]19, [62]20]. The emission and concentration of terpenes in Artemisia annua L are significantly influenced by the interaction between soil microorganisms and roots [[63]21]. Tomato root metabolites have been found to recruit beneficial rhizosphere components, thus improving plant resilience [[64]22]. Concurrently, in Angelica sinensis, the composition of rhizosphere microorganisms exhibits a robust correlation with the levels of active constituents such as organic acids, flavonoids, and fatty acids over multiple years of cultivation [[65]23]. Furthermore, root exudates play a pivotal role in governing both the yield and quality of Panax ginseng by exerting influence on rhizome biomass [[66]24]. Isolates of Bacillus derived from the rhizosphere soil of A. sinensis have been demonstrated to promote plant growth and biomass accumulation, while modulating the accumulation of specific compounds such as butylphthalide and ligustolide [[67]25]. Additionally, rhizosphere microorganisms in citrus soils have been found to activate terpene synthesis and enhance monoterpene accumulation through interactions with the host immune system [[68]26]. Significantly enriched genera, such as Rhizobacter, Variovorax, Polaromonas, and Mycobacterium, have been associated with variations in anisodines contents, aboveground biomass, and nitrogen levels in the rhizospheres of Anisodus tanguticus [[69]27]. The integration of microbiome and metabolome analyses enables the identification of beneficial bacteria highly correlated with essential active ingredients, thereby enhancing the quality of medicinal plant. For instance, in G. uralensis Fisch, joint transcriptome, microbiome, and metabolome analyses have revealed crucial environmental factors influencing the accumulation of liquiritin and glycyrrhizic acid. Consequently, in this investigation, rhizosphere soil and corresponding roots were collected from HuangQin plants of different ages (one-year, two-year-old, three-year-old, and four-year-old) for 16S rDNA gene sequencing and metabolomic profiling analysis, respectively. Materials and methods Plant roots and rhizosphere soil sample collection Rhizosphere soil and roots of HuangQin were collected from Fengxiang District, Baoji City, Shaanxi Province, utilizing a five-point sampling method [[70]28]. First, rhizosphere soil and root samples are collected within 5 cm of the plant roots. After removing the topsoil with a shovel, the roots were collected at a depth of 15–30 cm. Then, gently remove the soil that is not tightly attached to the roots, collect the soil that is tightly attached to the roots into a 50 ml sterile tube, place the roots in a collection bag, and send them back to the laboratory. The collected soil samples were sieved with a 20-mesh sieve, snap-frozen with liquid nitrogen and stored at −80 °C. Carefully remove fibrous roots and residual shoots after root sample sampling and rinse 3 times with PBS buffer to eliminate surface soil residue. Subsequently, the root samples are dried, sliced into 2 cm lengths, snap-frozen with liquid nitrogen and stored at −80 °C. Rhizosphere soil samples from one to four years old HuangQin were designated as Soil-HQ_1Y, Soil-HQ_2Y, Soil-HQ_3Y, and Soil-HQ_4Y, respectively. Similarly, roots from one to four years old HuangQin were denoted as HQ_1Y, HQ_2Y, HQ_3Y, and HQ_4Y, correspondingly. Each sample was represented by three biological replicates. To determine the growth years of HuangQin, we employed phloroglucinol-HCl staining, a method that selectively stains wooden walls red [[71]29]. The specific operation is as follows: a healthy 1-4-year-old taproot of HuangQin was selected and separated from the rhizome. Transverse Sect. (3-mm-thick) were sequentially obtained from the apical to basal root regions using a sterile blade. The fresh sections were mounted on glass slides and sequentially stained with phloroglucinol-CaCl[2] solution (1% w/v phloroglucinol in 95% ethanol) for 10 min, followed by acidification with concentrated HCl (36% v/v) at a 1:1 volumetric ratio. After 5 min of color development at room temperature, the lignin-stained samples were immediately observed under a stereomicroscope (SteREO Discovery 2.0) and digitally captured with an integrated camera system. The meteorological data of sample collection region were obtained from the National Meteorological Information Centre of China meteorological data service centre ([72]http://data.cma.cn/en). Soil physicochemical properties testing The soil physical and chemical properties were analyzed by Yangling Xinhua Ecological Technology Co., Ltd. Soil pH was measured using a Sardolus PB-10 pH meter, while the total Nitrogen (TN) content in the soil was determined employing the sulfuric acid-catalyst digestion method. The flame photometric method was employed to assessed the total Potassium (TK) content in the soil. Total Phosphorus (TP) and Available Phosphorus (AP) in the soil was determined using the NaOH melting method coupled with the molybdenum-antimony anti-colorimetric method. Organic carbon and organic matter were quantified using the potassium dichromate-concentrated sulfuric acid external heating method. Ammonium nitrogen (NH[4]^+-N), Alkali-hydrolyzable nitrogen (AN) and Nitrate nitrogen (NO[3]-N) was analyzed utilizing the potassium chloride extraction method. Flame photometry was used to determine Available Potassium (AK) and slowly available potassium (SAK). The drying method was used to determine the Soil Moisture Content (SMC) [[73]30]. Microbial DNA extracted from soil samples and sequencing Soil samples were collected from each of the four different years of HuangQin, and genomic DNA was extracted using the CTAB method. An appropriate amount of sample DNA was collected in a centrifuge tube, and the purity and concentration of DNA were evaluated using spectrophotometry. Subsequently, samples were diluted to a concentration of 1 ng/µL using sterile water. Specific primers targeting the 16S rDNA gene sequence of microorganisms were utilized for PCR amplification with 15 µL of Phusion High-Fidelity PCR Master Mix (New England Biolabs). The total DNA extracted was utilized as a template for PCR using primers targeting the V4 region of the 16S gene: forward primer 515 F (5’-GTGCCAGCMGCCGCGGTAA-3’) and reverse primer 806R (5’-GGACTACHVGGGTWTCTAAT-3’). The PCR reaction mixture consisted of Phusion Master Mix (2×) 15 µL, forward primer (1 µM) 0.2 µL, reverse primer (1 µM) 0.2 µL, genomic DNA (1 ng/µL) 10 µL (5 ~ 10 ng), and ddH[2]O to make up a final volume of 30 µL. The PCR conditions included an initial denaturation step at 98 °C for 1 min, followed by denaturation at 98 °C for 10 s, annealing at 50 °C for 30 s, extension at 72 °C for 30 s with a total of 30 cycles, and final extension step at 72 °C for 5 min. The resulting PCR products were pooled, purified, and subjected to library preparation. PCR products were detected by electrophoresis using a 2% agarose gel; The PCR products that passed the test were purified by magnetic beads, quantified by enzyme labels, mixed in equal amounts according to the concentration of PCR products, fully mixed and detected by 2% agarose gel electrophoresis, and recovered the products by gel recovery kit provided by qiagen for the target bands. Subsequently, libraries were constructed using the TruSeq^® DNA PCR-Free Sample Preparation Kit, Qubit and Q-PCR quantitated, and the libraries were qualified for on-machine sequencing using the NovaSeq6 × 000. Rhizosphere microbiome analysis The 16S sequencing data obtained from Illumina NovaSeq sequencing underwent quality control and were subsequently trimmed to generate Clean Tags. Chimera filtering was implemented to retain valid data suitable for subsequent analysis. Each sample’s valid data were clustered into Operational Taxonomic Units (OTUs) based on the 97% sequence similarity principle, followed by species annotation of the OTU sequences. Metabolite profiling analysis In this study, a total of 12 samples were selected and categorized into 4 groups for metabolic analysis. Sample preparation and extraction procedures were conducted as follows: Biological samples were subjected to vacuum freeze-drying using a lyophilizer (Scientz-100 F) until completely dehydrated. Subsequently, the dried samples were ground to a fine powder using a grinder (MM 400, Retsch) operating at 30 Hz for 1.5 min. A precise amount of 50 mg of sample powder was weighed using an electronic balance (MS105DΜ). To this, 70% methanol water internal standard extract pre-cooled to −20 °C at a volume of 1200 µL per 50 mg of sample was added (less than 50 mg of sample required a proportionate adjustment of the extraction agent volume to maintain the 1200 µL ratio). The mixture was vortexed once every 30 min for a total of 6 vortex cycles, each lasting 30 s. Following centrifugation at 12,000 rpm for 3 min, the supernatant was aspirated, and the sample was filtered through a microporous membrane with a pore size of 0.22 μm. The filtrate was then transferred to a vial for subsequent analysis using UPLC-MS/MS. The data acquisition instrument system mainly includes UPLC (ExionLC™ AD, [74]https://sciex.com.cn/) and Tandem mass spectrometry. The liquid phase conditions mainly include: Column: Agilent SB-C18 1.8 μm, 2.1 mm × 100 mm; Mobile phase: ultrapure water (0.1% formic acid) and acetonitrile (0.1% formic acid). Elution gradient: the proportion of phase B was 5% at 0.00 min, the proportion of phase B increased linearly to 95% within 9.00 min, and remained at 95% for 1 min, the proportion of phase B decreased to 5% at 10.00–11.10 min, and equilibrated at 5% to 14 min. Flow rate 0.35 mL/min; Column temperature 40 °C; The injection volume was 2 µL. The mass spectrometry conditions were configured as follows: Electrospray ionization (ESI) temperature was maintained at 550 °C. The ion spray voltage (IS) was set to 5500 V for positive ion mode and − 4500 V for negative ion mode. Gas settings included ion source gas I (GSI), gas II (GSII), and curtain gas (CUR), which were adjusted to 50, 60, and 25 psi, respectively. Collision-induced ionization parameters were set to high, and the collision gas (nitrogen) was maintained at medium pressure during QQQ scan operation. Further optimization involved the adjustment of declustering potential (DP) and collision energy (CE) for each multiple reaction monitoring (MRM) ion pair. Specific MRM transitions were monitored in each epoch based on the elution profile of metabolites within that epoch. The data collected by the software Analyst 1.6.3 was matched with the public database of metabolite information and the reference material database of the metabolome platform to identify the metabolites in the sample. According to the MRM monitoring mode, the peak area data of metabolites in the same metabolites in different samples were corrected, and the relative content of metabolites was finally obtained. PCA (Principal Component Analysis) was conducted using R ([75]www.r-project.org) to analyze the metabolomic data. For two-group comparisons, differential metabolites were identified based on Variable Importance in Projection (VIP) scores (VIP > 1) and absolute Log[2] Fold Change (|Log[2]FC| ≥ 1.0). In multi-group analyses, differential metabolites were determined using VIP scores (VIP > 1) and P-values (P-value < 0.05). The identified metabolites were annotated using the KEGG compound database ([76]http://www.kegg.jp/kegg/compound/), and subsequently mapped to the KEGG pathway database ([77]http://www.kegg.jp/kegg/pathway.html). Results Determining the growth years of HuangQin The root’s thickness visibly increased with advancing growth years (Fig. [78]2). Observation of HuangQin slices revealed that from the third year onwards, the central hollow expanded with increasing growth years. Staining results indicated that one-year-old HuangQin slices displayed a single circle of red xylem, whereas two-year-old slices exhibited two distinct rounds of red xylem. Additionally, the central regions of one- and two-year-old HuangQin slices remained structurally intact. However, starting from the third year, a conspicuous black cavity emerged at the core of the slice, accompanied by the presence of three concentric rings of red xylem. Slices from the third and fourth years exhibited three and four rounds of red xylem, respectively. Remarkably, there was a progressively increase in the proportion of decay from the third to fourth year, accompanied by a color transition from brown to brownish-black. Overall, the phloroglucinol staining results provided unequivocal confirmation of HuangQin’s growth age. Fig. 2. [79]Fig. 2 [80]Open in a new tab Photos of HuangQin samples and corresponding staining photos. a1, b1: HQ_1Y; a2, b2: HQ_2Y; a3, b3: HQ_3Y; a4, b4: HQ_4Y The geographical locations of the FengXiang exhibits resulted in slight variations in precipitation and annual average air temperature over the course of four growth years (Stable 1). The third year shows the highest air temperature and land surface temperature, accompanied by the longest sunshine duration and ultraviolet radiation. The first year, however, exhibits the highest amount of precipitation (Stable 1). Soil physical and chemical properties in different years The physicochemical of the rhizosphere soil were analyzed to obtain a comprehensive understanding. The results revealed that the pH remained slightly alkaline and exhibited a slightly increased with increasing planting duration. The levels of SAK exhibited a gradual upward trend from the first year to the fourth year, while TN, TK, and AK displayed an initial increase over the first two years followed by a subsequent decline from the second to fourth years (Table [81]1). Conversely, TP, NH[4]^+-N, AP, and SMC exhibited a declining trend from the first to second years, followed by an increase from the second to third years, and subsequently a decrease from the third to fourth years. On the other hand, AN, NO[3]-N, and acid-soluble potassium content increased from the first to second year, decreased from the second to third year, and then increased again from the third to fourth year. Overall, Soil-HQ_2Y exhibited the highest levels of TN, TK, AN, NO[3]-N and AK. Soil-HQ_3Y recorded the highest levels of TP and AP. In contrast, Soil-HQ_4Y displayed the highest levels of acid-soluble potassium, SAK, and pH value. Only NH[4]^+-N content was found to be highest in Soil-HQ_1Y (SFig [82]1). Table 1. Mineral element content in rhizosphere soil of HuangQin at different years HQ_1Y HQ_2Y HQ_3Y HQ_4Y Total Nitrogen (g/kg) 0.7767 ± 0.0153 0.9700 ± 0.0200 0.8700 ± 0.0100 0.7833 ± 0.0231 Total Phosphorus (g/kg) 0.5767 ± 0.0153 0.5100 ± 0.0100 0.6500 ± 0.0100 0.4633 ± 0.0058 Total Potassium (g/kg) 20.2633 ± 0.1650 21.1600 ± 0.1758 21.1533 ± 0.4292 20.7600 ± 0.2722 Acid soluble potassium (g/kg) 645.3333 ± 4.6188 888.0000 ± 24.0000 866.6667 ± 4.6188 925.3333 ± 23.0940 slowly available potassium (g/kg) 548.0000 ± 4.3589 748.0000 ± 24.5153 750.0000 ± 3.6056 823.0000 ± 22.5389 Alkali-hydrolyzable nitrogen (mg/kg) 35.9100 ± 0.6750 44.7200 ± 1.1778 37.9467 ± 1.1720 39.0767 ± 1.0333 Ammonium nitrogen (mg/kg) 0.7767 ± 0.3215 0.5600 ± 0.0100 0.6633 ± 0.0252 0.6233 ± 0.0058 Nitrate nitrogen (mg/kg) 4.3233 ± 0.1274 4.1667 ± 0.0058 1.8533 ± 0.1234 2.9667 ± 0.1102 Available Phosphorus (mg/kg) 8.9133 ± 0.0513 5.8100 ± 0.1852 10.5633 ± 0.1528 3.2100 ± 0.2307 Available Potassium (mg/kg) 97.3333 ± 0.5774 140.0000 ± 1.0000 116.6667 ± 1.5275 102.3333 ± 1.1547 pH 8.0267 ± 0.03215 8.0400 ± 0.0361 8.0567 ± 0.0252 8.1500 ± 0.0200 Soil Moisture Content (%) 16.1067 ± 0.0777 14.5067 ± 0.1629 16.0800 ± 0.1353 14.3033 ± 0.2540 [83]Open in a new tab Microbial community structure of HuangQin planted in different years In rhizosphere soil samples from the first year to the fourth year, the number of operational taxonomic units (OTUs) observed were 7594, 10,227, 10,280, and 9796, respectively (Fig. [84]3). Notably, Soil-HQ_3Y exhibited the highest number of OTUs, whereas Soil-HQ_1Y displayed the lowest count. The total OTU count increased from Soil-HQ_1Y to Soil-HQ_3Y, followed by a decrease from Soil-HQ_3Y to Soil-HQ_4Y. This pattern suggests an initial increase followed by a decrease in OTU count with increasing planting years of HuangQin, with a peak observed at 3 years. To delineate common and unique characteristics of soil microorganisms across different years, a Venn plot analysis was performed. The total number of common OTUs across all four groups was 2432, 32.03% (Soil-HQ_1Y), 23.78% (Soil-HQ_2Y), 23.66% (Soil-HQ_3Y), and 24.83% (Soil-HQ_4Y) of the total OTU in each group, respectively. And a total of 4636 OTUs were shared between Soil-HQ_2Y and Soil-HQ_3Y, whereas Soil-HQ_1Y and Soil-HQ_2Y exhibited the fewest shared OTUs, totaling 3430. Additionally, Unique OTU counts for Soil-HQ_1Y to Soil-HQ_4Y were 2835, 4474, 4275, and 4269, respectively. The results showed that the total number of bacterial OTU and the number of endemic bacteria in HuangQin soil at different growth years decreased with the increase of years. The soil bacterial species of Soil-HQ_1Y and Soil-HQ_2Y were the most different, and the soil bacterial species of Soil-HQ_2Y and Soil-HQ_3Y were the smallest, and nearly half of the bacterial species were the same. The above results indicated that planting years had a great effect on the amount of OTU in HuangQin soil. Fig. 3. Fig. 3 [85]Open in a new tab Temporal variations in soil microorganism’s populations as depicted by Histograms across multiple years. Abscissa, different groupings. Ordinate, number of OTUs A total of 41 phyla, 78 classes, 159 orders, 298 families, 619 genera, and 380 species were detected in 4 groups of soil samples. Analysis of soil microorganism composition and structural changes across different years revealed the top 10 microflora at the phylum level (Fig. [86]4a), with varying abundance percentages. These included Myxococcota (2.20%−3.15%), Bacteroidota (1.75%−4.06%), Planctomycetes (3.91%−4.95%), Gemmatimonadota (3.74%−5.93%), unidentified_Bacteria (5.26%−7.36%), Crenarchaeota (3.13%−8.59%), Actinobacteria (7.45%−21.35%), Acidobacteriota (8.75%−15.14%), Actinobacteriota (11.47%−14.33%), and Proteobacteria (15.57%−32.13%). Notably, Proteobacteria, Actinobacteriota, Acidobacteriota, Actinobacteria and Crenarchaeota emerged as dominant bacterial phyla. Proteobacteria was the most dominant microbial group in the soil, with a relative abundance of 20.83%, proteobacteria is one of the most abundant bacterial groups in soil, and it participates in soil nutrient cycling through key links such as nitrogen cycle, sulfur cycle and phosphorus cycle [[87]31], followed by Actinobacteriota, Acidobacteriota, Actinobacteria and Crenarchaeota, all of which accounted for about 12%. These five dominant groups together accounted for 58.07% of the total bacterial abundance. Actinobacteriota plays a variety of roles in soils, including participating in the decomposition of organic matter, participating in soil carbon, nitrogen, and sulfur cycling, and possibly promoting plant growth. These functions are essential for maintaining soil health and fertility [[88]32]. Actinobacteria are abundant aerobic microorganisms in the soil, which can decompose organic matter, inhibit pathogenic bacteria, participate in nitrogen fixation and degradation of pollution, improve nutrient and mineral availability, promote plant growth, and do not pollute the environment, and maintain soil biological balance [[89]33]. Moreover, the abundance of the five phyla Myxococcota, Planctomycetes, Crenarchaeota and Acidobacteriota increased year by year, reaching the highest value in the fourth year. The abundance of Bacteroidota, unidentified_Bacteria and Proteobacteria showed a decreasing and then increasing trend, among them, Bacteroidota and Proteobacteria both had only small increases, and they were still the most abundant in the first year. Finally, the abundance of Actinobacteria and Actinobacteriota showed a trend of first increasing and then decreasing. It can be seen that with the increase of planting years, the abundance of some beneficial bacteria such as Actinobacteria, Acidobacteriota and Proteobacteria in the soil decreases. Fig. 4. [90]Fig. 4 [91]Open in a new tab Species composition and abundance of bacterial communities in the rhizosphere soil of Skullcap with varying cultivation durations (a: door level, b: genus level) c: ASV-based histogram of LDA value distribution. The displayed species exhibit an LDA score exceeding 4 At the genus level, Achromobacter was the dominant genus in the Soil-HQ_1Y bacterial community composition, with a relative abundance of 12.30% (Fig. [92]4b), followed by unidentified_Burkholderiaceae (4.08%), Pseudarthrobacter (2.68%), Solirubrobacter (0.86%), Gaiella (1.68%), Blastococcus (0.82%), Sphingomonas (0.70%), Nocardioides (0.66%), Haliangium (0.63%) and unidentified_Methylomirabilota (0.54%). Among the Soil-HQ_2Y bacterial community composition, Pseudarthrobacter was the dominant genus with a relative abundance of 13.25%, followed by Gaiella (1.81%), Blastococcus (1.5%), Nocardioides (1.31%), Haliangium (0.84%), Solirubrobacter (0.83%), unidentified_Methylomirabilota (0.67%), Sphingomonas (0.56%) and Achromobacter (0.01%). Among the Soil-HQ_3Y bacterial community composition, Pseudarthrobacter was the dominant genus with a relative abundance of 2.61%, followed by Gaiella (2.03%), Blastococcus (1.44%), Solirubrobacter (0.95%), unidentified_Methylomirabilota (0.92%), Nocardioides (0.84%), Haliangium (0.81%), Sphingomonas (0.67%), Achromobacter (0.02%) and unidentified_Burkholderiaceae (0.01%). Lastly, in the Soil-HQ_4Y of the bacterial community composition, Pseudarthrobacter was the dominant genus with a relative abundance of 2.34%, followed by Gaiella (1.79%), unidentified_Methylomirabilota (1.15%), Haliangium (1.03%), Solirubrobacter (0.93%), Blastococcus (0.71%), Sphingomonas (0.68%), Nocardioides (0.56%), Achromobacter (0.04%) and unidentified_Burkholderiaceae (0.02%). Among them, the abundance of Pseudarthrobacter in the four groups of samples has been greater than 2%, and the overall trend is to increase first and then decrease, reaching a peak in the second year. Blastococcus and Nocardioides are in perfect tune with its trend. Gaiella has a similar trend, peaking only in its third year. The abundance of Sphingomonas, unidentified_Burkholderiaceae and Achromobacter decreased first and then increased, while the abundance of unidentified_Methylomirabilota and Haliangium increased overall. The abundance of Solirubrobacter changes in a trend similar to N. Based on the linear discriminant analysis effect size (LEfSe), a linear discriminant analysis (LDA) distribution histogram with a score cutoff of ± 4 was employed to illustrate taxa with significant differences in abundance between different groups (Fig. [93]4c). LEfSe analysis showed that the bulk soil indicator bacteria in Soil-HQ_1Y were Burkholderia_pseudomallei (5.04), Achromobacter (5.02) and Proteobacteria (4.98), among the Soil-HQ_2Y were Pseudarthrobacter (5.14), Micrococcaceae (4.92) and unidentified_Actinobacteria (4.89), and in the Soil-HQ_3Y were Nitrososphaerales (4.56), Nitrososphaeria (4.51) and Crenarchaeota (4.46), and finally in the Soil-HQ_4Y are Nitrosomonadaceae (4.31), Metagenome (4.16) and Pyrinomonadaceae (4.15). These results also showed that there were significant differences in bacterial abundance in rhizosphere soils of HuangQin in different planting years. Flavonoid metabolomics analysis and differential metabolite screening To gain insights into the changes occurring in the main active ingredients of HuangQin during the transition from ZiQin to KuQin, flavonoids in the four groups of samples were identified using UPLC-MS platform with broadly targeted metabolome technology. For root samples from HQ_1Y to HQ_4Y, we detected 550, 547, 547, and 547 metabolites, respectively. Flavonoids constituted 97.28% of the metabolite composition, while tannins accounted for 2.7%. Among the identified flavonoids, a total of 537 were classified into secondary structures (Fig. [94]5a), comprising 282 flavones (52.5%), 105 flavonols (19.6%), 49 flavonones (9.1%), 37 isoflavones (6.9%), 25 other flavonoids (4.7%), 18 chalcones (3.4%), 15 flavanonols (2.8%), and 6 flavanols (1.0%). Fig. 5. [95]Fig. 5 [96]Open in a new tab Composition of metabolites categories in HuangQin. a: Histogram depicting the secondary classification of all identified metabolites. b: PCA grouping analysis. c: OPLS-DA verification Diagram. d: OPLS-DA score graph The principal component analysis (PCA) results demonstrated good repeatability of biological samples, with distinct separation among different samples (Fig. [97]5b). The contribution rates of the first two principal components were 32.42% and 21.68%, respectively. The cumulative rate reached 54.10%, which effectively discriminates HuangQin samples for different years, indicating that there were certain differences in the chemical composition of HuangQin across different years. Orthogonal partial least squares discriminant analysis (OPLS-DA), a supervised pattern recognition method, effectively excluded non-study-related effects to identify differential metabolites. The model’s Q2 value of 0.937 indicated its appropriateness (Fig. [98]5c). The OPLS-DA score plot (Fig. [99]5d) revealed significant segregation among the different comparison groups. Pairwise comparisons were conducted among the radix samples of the four HuangQin accessions to identify metabolites responsible for observed differences. Through correlation analysis, PCA, cluster heat map analysis, and OPLS-DA model analysis, high biological sample repeatability was confirmed. Differential metabolite analysis 552 compounds were screened for differential metabolites, resulting in the identification of 222 differential metabolites (121 upregulated and 101 downregulated) in HQ_2Y compared to HQ_1Y (Fig. [100]6a), 140 (106 upregulated and 34 downregulated) in HQ_3Y compared to HQ_2Y, 178 (61 upregulated and 117 downregulated) in HQ_4Y compared to HQ_3Y, 206 (142 upregulated and 64 downregulated) in HQ_3Y compared to HQ_1Y, 210 (109 upregulated and 101 downregulated) in HQ_4Y compared to HQ_1Y, and 151 (75 upregulated and 76 downregulated) in HQ_4Y compared to HQ_2Y (SFig. [101]2). The types of differential metabolites in HQ_1Y and HQ_2Y encompass 45.9% flavones (Fig. [102]6b), 20.3% flavonols, 9.9% flavanones, 7.7% isoflavones, 5.9% chalcones, 3.2% flavanonols, 3.6% other flavonoids, 3.2% tannin, and 0.5% flavanols. In HQ_2Y and HQ_3Y, the differential metabolite types included 53.6% flavones, 18.6% flavonols, 9.3% flavanones, 8.6% isoflavones, 4.3% other flavonoids, 3.6% chalcones, and 2.1% flavanonols. The increased proportion of flavones by 8% was the main reason for the change, while the proportions of other types did not show significant changes, and no significant differences were observed in tannin and flavanol metabolites. Over the course of 2 to 3 years, the contents of chrysin, baicalein, norwogonin, and baicalin increased, albeit not significantly, indicating ongoing accumulation of active ingredients in HuangQin during this period. In HQ_3Y and HQ_4Y, the differential metabolite types comprised 48.3% flavones, 21.3% flavanols, 9.6% flavanones, 8.4% isoflavones, 4.5% tannin, 3.9% other flavonoids, 1.7% flavanonols, 1.1% chalcones, and 1.1% flavanols. Notably, baicalin, a principal component in HuangQin, exhibited a significant down-regulation by 0.44 times over the course of three to four years, with its content halved, while chrysin, baicalein, and norwogonin all showed up-regulation, although not significantly. Furthermore, the content of chrysin in HQ_4Y reached its highest value, and oroxylin A content in HuangQin was significantly up-regulated at each stage, reaching its peak in the fourth year. In HQ_2Y and HQ_4Y, the differential metabolite types comprised 48.3% flavones, 19.9% flavonols, 7.9% flavanones, 6.6% isoflavones, 6.0% other flavonoids, 4.0% chalcones, 3.3% tannin, 3.3% flavanonols, and 0.7% flavonols. At this stage, significant changes occurred in the main components of HuangQin, with chrysin, baicalein, norwogonin, and oroxylin A being significantly up-regulated, while baicalin was significantly down-regulated. This suggests continuous accumulation of norwogonin, chrysin, oroxylin A, and baicalein over the 2 to 4-year period, while baicalin underwent significant reduction. Therefore, it is speculated that the 3-year growth period of HuangQin may be the time node to distinguish ZiQin and KuQin. Fig. 6. [103]Fig. 6 [104]Open in a new tab Metabolic profiling of differential metabolites. a: Differential metabolite volcano plot. The horizontal axis numbers represent different difference groups, I: HQ_1Y vs. HQ_2Y; II: HQ_2Y vs. HQ_3Y; III: HQ_3Y vs. HQ_4Y; IV: HQ_2Y vs. HQ_4Y. Vertical axis, log[2]FC value of difference substances in each difference group. Red dot, difference species (p value < 0.05). Black color, difference species (p value > = 0.05). b: Differential metabolite composition map. Horizontal axis, substances classified by the secondary classification of HuangQin. Vertical axis, number of metabolites All the detected differential metabolites were compared with the KEGG database, and the metabolic pathways closely related to different differential expressions were identified through pathway enrichment analysis, which provided a basis for gene function mining in the later stage. It is included isoflavonoid biosynthesis, metabolic pathways, biosynthesis of secondary metabolites, flavonoid biosynthesis and flavone and flavonol biosynthesis. Effects of soil physicochemical properties on bacteria at different planting years of HuangQin By integrating the prominent microorganisms in soil physicochemical properties and the abundance at the genus level, a notable correlation emerged with Pseudarthrobacter exhibiting a significant association with TN, AN, AK, and NH[4]^+-N (Fig. [105]7a). Specifically, as TN, AN, and AK increased, the abundance of Pseudarthrobacter demonstrated a corresponding rise, whereas with a decrease in NH[4]^+-N, the abundance of Pseudarthrobacter increased, peaking at two years. Pseudarthrobacter has been recognized for its role in promoting plant growth, supported by studies revealing that Pseudarthrobacter sp. NIBRBAC000502770 can enhance the growth of Geum aleppicum and facilitate the accumulation of flavonoids [[106]34]. Fig. 7. [107]Fig. 7 [108]Open in a new tab Integrated analysis of soil physicochemistry and soil microbial communities. a: Edge width corresponds to Mantel’s R statistic for the corresponding distance correlations, Blue and red for gradients with negative to positive correlations, respectively. Orange, very significant line (Mantel’s p < 0.01). Green, prominent line (0.01 ≤ Mantel’s p < 0.05). Gray, the inconspicuous line color (Mantel’s p > 0.05). b: RDA analysis. Black, different microorganisms. Blue labels and arrows, different physicochemical factors. Colored dots, samples belonging to different groups. Length of the arrow, intensity of the influence of environmental factors on microbial changes A decrease in NH[4]^+-N decreases during the 1–2 year period is accompanied by a corresponding reduction in the abundance of Sphingomonas. The NH[4]^+-N concentration reaches its peak at 3 years, coinciding with the highest abundance of Sphingomonas during the same period, while AN exhibit an inverse trend. Sphingomonas species have exhibited significant agricultural applications, particularly in promoting plant growth and enhancing plant stress resistance Certain strains of Sphingomonas have been identified as potential enhancers of crop growth, particularly under stress conditions such as drought, salinity, and heavy metal exposure in agricultural soils. This is attributed to their ability to synthesize plant growth hormones like gibberellin and indoleacetic acid [[109]35]. Furthermore, a significant correlation was observed between Haliangium and AP, with both exhibiting a decline in abundance from 3 to 4 years, coinciding with the lowest AP levels at year 4. Haliangium species may play a crucial role in the degradation of polycyclic aromatic hydrocarbons (PAHs) in biochar-enhanced rhizosphere soils [[110]36]. Nocardioides exhibited a notable correlation with TN and AK, as well as a significant association with AN and pH. The abundance of Nocardioides increased during the 1–2 year period but decreased from 2 to 4 years, mirroring the trends observed in TN and AK. AN followed a similar trajectory as Nocardioides, with a slight increase observed during 3–4 years, whereas pH remained relatively stable during 1–3 years before increasing during 3–4 years. Achromobacter demonstrated a significant correlation with TP, SK, SAP, and NH[4]^+-N, with its highest abundance observed at 1 year and gradually decreasing over subsequent years. The general trend for TP, SK, SAP, and NH[4]^+-N was an increase during 1–3 years followed by a decrease in 3–4 years. Moreover, a significant correlation was noted between Blastococcus and AP, with Blastococcus being most abundant at 3 years, consistent with the trend observed in AP. RDA analysis was used to further explore the relationship between rhizosphere soil microbial community and soil physicochemical factors in HuangQin. The results showed that the contents of pH, TN, TK, NO[3]-N and AP had an effect on the distribution of soil bacterial communities (Fig. [111]7b). The influence of pH, TN and TK was greater, and the influence of NO[3]-N and AP was relatively small. Pseudarthrobacter was positively correlated with TN, TK and NO[3]-N in soil. and it was negatively correlated with the content of AP and pH. Achromobacter and unidentified_Burkholderiaceae were positively correlated with the content of AP in soil. There was a negative correlation with the contents of TN, TK, NO[3]-N and pH, and the rest of the bacteria were all gathered at the origin, which was not significantly correlated with soil physical and chemical factors. Correlation analysis of differential metabolites with differential microorganisms Spearman correlation hierarchical cluster analysis (Fig. [112]8a) revealed a significant decrease in baicalin content among the differential microorganisms and metabolites over the 3 to 4 year period. The data of differential metabolites and the spearman correlation analysis of differential microorganisms were listed (Stable2). Rhizosphere bacteria were found to exhibit both negative and positive correlations with baicalin levels. On the one hand, rhizosphere bacteria that are negatively correlated with baicalin include 13 species of bacteria, such as Polycyclovorans, Pseudomonas and Aquicella, etc. Conversely, the rhizosphere bacteria exhibiting positive correlation with baicalin encompass a total of 17 bacterial species, including Nocardioides, Iamia and Gemmatimonas, etc. Notably, the majority of these positively associated bacteria comprise beneficial strains that actively contribute to soil fertility and plant growth through their involvement in crucial processes such as organic matter decomposition, carbon fixation, and nitrogen metabolism. Therefore, we hypothesize that certain bacterial species, such as Polyceclovolens, Psadormonas, and Aquitella, may exert inhibitory effects on baicalin synthesis, while others like Nocatiodes, Yamia, and Germatimonas could potentially enhance the accumulation of baicalin. Fig. 8. [113]Fig. 8 [114]Open in a new tab Hierarchical cluster heat map of spearman correlation between differential microorganisms and differential metabolites. a: HQ_4Y_vs_HQ_3Y; b: HQ_4Y_vs_HQ_2Y Between two to four years of age (Fig. [115]8b), baicalin content exhibited a significant increase and was found to be significantly negatively correlated with 12 bacteria species, including Pirellula, Polycyclovorans and Nitrosospira, while positively correlated with 23 bacteria species such as Blastococcus, Skermanella and unidentified_Acidimicrobiia. The content of chrysin exhibited a significant increase and demonstrated a negative correlation with 20 bacteria species, including Blastococcus, Skermanella, and unidentified_Acidimicrobiia. Conversely, it displayed a positive correlation with 9 bacteria such as Pirellula, unidentified_Chloroflexi and Polycyclovorans. The content of norwogonin exhibited a significant increase and displayed negative correlations with 23 bacteria taxa, including Pseudarthrobacter, Cellulomonas and Microbacterium. Conversely, it showed a positive correlation with 15 bacteria including Pirellula, Polycyclovorans and Nitrosospira. The content of baicalein increased significantly and was negatively correlated with 13 bacteria such as unidentified_Acidimicrobiia, Cellulomonas and Actinomycetospora, and positively correlated with 12 bacteria such as Pirellula, unidentified_Chloroflexi and Polycyclovorans. While the Mantel tests and redundancy analysis (RDA) indicated significant correlations (p < 0.05) between specific microbial taxa and flavonoid accumulation, these findings reflect statistical associations rather than established causal relationships. Additional mechanistic investigations such as targeted inoculation or genetic manipulation, are necessary to confirm whether these microbes directly regulate flavonoid biosynthesis in HuangQin. Discussion Soil bacteria exhibit robust metabolic activity, rapid proliferation, diverse species composition, and substantial population size, thereby constituting a pivotal component of the soil microecological environment. Concurrently, they exert a significant influence on the growth dynamics of medicinal plants [[116]37, [117]38]. The present investigation revealed a close association between alterations in bacterial communities and the duration of plant cultivation. Specifically, the diversity of soil bacterial communities exhibited an initial increase followed by a subsequent decline after four years of continuous HuangQin cultivation. Subsequent analysis of bacterial abundance at the microbial genus level identified Sphingomonas, unidentified_Burkholderiaceae, and Achromobacter as the predominant taxa in the first year, while Nocardioides, Blastococcus, and Pseudarthrobacter dominated in the second year. Solirubrobacter and Gaiella were prevalent in the third year, with Haliangium being predominant in the fourth year. Notably, Achromobacter exhibited a growth-promoting effect on plants, with its abundance peaking in the initial two years of HuangQin cultivation before declining. The gram-negative bacillus Achromobacter sp. 5B1, recognized as a probiotic, modulates plant physiological traits, including root development via auxin signaling pathways [[118]39]. Similarly, unidentified_Burkholderiaceae, known for its beneficial plant functions such as colonization, nitrogen fixation, and pollutant degradation, displayed a substantial impact on plant growth, exemplified by Burkholderia phytofirmans PsJN. The multifaceted functions of Sphingomonas genus species encompass the production of beneficial plant hormones, enhancement of plant growth under stress conditions, and remediation of environmental pollutants [[119]40]. Consequently, the dominance of Achromobacter during the initial HuangQin cultivation period implies its pivotal role in root development, while the modest abundance of Burkholderiaceae may contribute to the accumulation of active ingredients in HuangQin. Pseudarthrobacter, exhibiting effects on growth enhancement and flavonoid content augmentation in Geum aleppicum, evidenced its peak abundance in the second year, coinciding with the rapid growth phase of HuangQin and the accumulation of active ingredients [[120]41, [121]42]. However, with prolonged HuangQin cultivation, beneficial bacteria abundance declined gradually, as evidenced by the reduction in Pseudarthrobacter and unidentified_Burkholderiaceae from the second to fourth years. Soil microorganisms, serving as drivers of organic matter transformation and nutrient cycling, significantly influence soil nutrient availability. Alterations in soil microorganism diversity and structure may consequently exert an influence on plant composition. However, this study primarily elucidates microbiome-metabolite covariation patterns, but causation cannot be inferred from observational data alone. For instance, microbial community shifts in the microbial community might either drive flavonoid production or indirectly reflect plant-microbe feedback mediated by root exudates. To address this limitaion, future studies should integrate synthetic microbial community (SynCom) experimetns with transcriptomic profiling of flavonoid pathway genes under controlled microbial colonization conditions. The physicochemical properties of soil play a crucial role in shaping the composition of soil microbial communities [[122]43]. Soil pH is an important chemical property of soil, exerting significant influence on the activity of soil microorganisms, the decomposition of minerals and organic matter, as well as the release, fixation and migration of soil nutrients, thus affecting the bioavailability and toxicity of the elements. Furthermore, it has been established that optimal pH level plays a pivotal role in the accumulation of organic matter, and alteration in soil pH will also exert an influence on the carbon cycling process of microorganisms. For example, soil pH is a major factor influencing the microbial community of the massive soil and rhizosphere soil of Ageratina adenophora [[123]44]. Several studies have demonstrated that alterations in soil pH exert significant influence on the composition and structure of soil microbial communities, thereby giving rise to persistent challenges in continuous cropping obstacles [[124]45, [125]46]. The soil pH consistently exhibited weak alkalinity throughout the duration of this study, with only a slightly observed over the years of cultivation. This finding aligns with the optimal growth conditions for HuangQin, which thrive under neutral to weakly alkaline environments. The soil microbial community structure can be directly or indirectly influenced by factors such as nitrogen and phosphorus content in the soil, as well as soil moisture levels, which ultimately impact crop yields [[126]47]. Nitrogen is a large nutrient that is necessary for plant growth and development. The element actively participates in numerous vital physiological and biochemical reactions as well as material metabolism processes, thereby serving as the primary limiting factor in plant growth and yield. Phosphorus and potassium are also indispensable nutrients for plant growth and development [[127]48, [128]49]. The three elements of fertilizer, namely nitrogen, phosphorus and potassium, collectively play an indispensable role in plant growth. Their effects on plants are not isolated but rather exhibit a coordinated influence on each other [[129]50]. The latest findings indicate that the judicious application of N, P and K fertilizers can increase the yield of yam, establishing a significantly correlation between yam yield and the appropriate utilization of N, P and K fertilizers [[130]51]. The composition of the microbiome in the core area of Citrus reticulata ‘Chachi’ was related to factors such as pH, organic matter, AN, K, and P. The diversity of bacterial communities in Mulberry and Alfalfa intercropping soils was correlated with variations in soil total carbon, available phosphate, and available potassium [[131]52]. Therefore, we hypothesized that pH, TN, TK, NO[3]-N and AP might exert significant influence on the composition and diversity of microbial communities in the rhizosphere soil of HuangQin. In this investigation, 12 samples were systematically selected and grouped for metabolic profiling. Leveraging the UPLC-MS/MS detection platform and a bespoke database, a comprehensive analysis identified a total of 552 metabolites, predominantly flavonoids (537) with a minor presence of tannins (15). The subsequent analysis revealed a consistent metabolite composition across the four vintages of HuangQin, indicating minimal variation likely attributed to their shared origin. Employing OPLS-DA analysis with stringent criteria, 222, 140, and 178 differential metabolites were delineated for the transitions from the first to second year, second to third year, and third to fourth year, respectively. Integrating this with the biosynthesis pathway of main flavonoids in HuangQin revealed stable levels of chrysin, baicalin, baicalein, and norwogonin during the initial 1 to 2 years, with chrysin and baicalin showing accumulation trends while baicalein and norwogonin exhibited decreased levels (SFig [132]3). Notably, the subsequent 2 to 3-year period witnessed a notable accumulation phase, particularly marked by a significant surge in chrysin content, reaching a four-fold increase compared to the previous biennial period. Additionally, although the levels of baicalin, baicalein, and norwogonin displayed modest changes, all demonstrated a consistent trend towards accumulation. This phase coincided with a speculated enhancement in SbFNSII-2 activity, consequently elevating the catalytic formation of these compounds. In the subsequent 3 to 4-year period, while baicalin levels exhibited a substantial decline, the levels of chrysin, baicalein, and norwogonin continued to rise steadily, indicating persistent accumulation. This trend could be attributed to the enhanced activities of SbFNSII-2, F6H, and F8H enzymes, facilitating increased catalysis. Remarkably, over the extended 2 to 4-year duration, there were significant elevations in the levels of chrysin, baicalein, and norwogonin, highlighting an almost six-fold increase in chrysin content and a two-fold rise in both baicalein and norwogonin levels. Conversely, baicalin levels exhibited a pronounced reduction by half. These findings underscore the distinct developmental stages of HuangQin and its active constituents, indicating that the third year as the optimal harvest period due to the peak baicalin content. This is similar to previous studies, which is a suitable harvest period in the third year [[133]53, [134]54]. However, the fourth year represents a favorable harvesting window for elevated levels of chrysin, baicalein, and norwogonin. Notably, differences in baicalin, baicalein, and norwogonin content between ZiQin and KuQin variants indicate potential clinical implications, necessitating further investigation for tailored therapeutic applications. While our study has certain limitations, such as data being collected from a single geographic region, which may restrict the generalizability of the findings to more heterogeneous environments, future studies should aim to replicate this work across regions with diverse soil conditions. Since we utilized only non-cultivation method for rhizosphere microorganism characterization, future research could incorporate cultivation-based approaches to identify additional key strains for biological role confirmation. Additionally, metatranscriptomic and metabolomic analyses could be applied to elucidate microbial functional activity and construct strain-host gene co-expression networks. Furthermore, targeted SynComs (Synthetic Communities) could be using high-performance strains identified in this study, and their efficacy in promoting plant growth or enhancing stress resistance could be evaluated through rhizosphere colonization experiments. Supplementary Information Below is the link to the electronic supplementary material. [135]Supplementary Material 1^ (427.3KB, zip) Acknowledgements