Aspergillus niger ZJ-17 enhances the yield and quality of Angelica dahurica var. formosana by modulating beneficial rhizobacteria: a sustainable strategy for plant production Authors: Meiyan Jiang, Renlang Liu, Zhengchuan Li, Hailing Xu, Jiuyu Zeng, Wenjie Qu, Zhaoyong Hu, Yinyin Chen, Dongju Feng, Wei Wu Published: 2026 ABSTRACT AbstractBackgroundAngelica dahurica var. formosana is a medicinal and edible plant. Higher abundance of Proteobacteria is an excellent characteristic of its rhizosphere bacterial community. Aspergillus niger ZJ-17 (AN17) significantly improved plant yield and quality while reducing fertilizer input application in both pot and field experiments. However, the impact of inoculants on resident soil bacteria directly determines their field application. Therefore, to systematically analyze how AN17 remodels host rhizosphere bacterial communities, we inoculated AN17 into the roots of A. dahurica var. formosana. The rhizosphere bacterial communities and root exudates were investigated at the harvest stage. Rhizosphere bacteria were isolated for in vitro experiments to elucidate the reasons for the changes in the rhizosphere bacterial community.ResultsAN17 promoted nutrient utilization and absorption in rhizosphere soil and increased the accumulation of IAA and JA in plant roots. In the microbiome, the relative abundance of rhizosphere Proteobacteria increased after inoculation. A total of 832 bacterial strains were isolated from the rhizosphere of the host for in vitro experiments. In in vitro experiments, AN17 induced the enrichment of Proteobacteria through microbial interactions and the modulation of host root exudates. These root exudates promoted the proliferation of bacterial genera with higher abundance and diversity. In the metabolome, host root activity increased following AN17 inoculation. According to the KEGG analysis of root exudates, AN17 upregulated microbial metabolism in diverse environments, the biosynthesis of alkaloids derived from the shikimate pathway and tryptophan metabolism. Correlation analysis and in vitro tests revealed that AN17 regulated the secretion of phenolic acids (4-chlorophenol, 2-oxoadipic acid, pyrogallol, and vanillic acid) from roots, which serve as crucial components driving the enrichment of Proteobacteria. In both plate and pot experiments, these bacteria promoted the growth of A. dahurica var. formosana and activated nutrient availability.ConclusionsWe supplemented multiple growth-promoting strategies by which AN17 improves rhizosphere bacterial communities. Phenolic acids in root exudates were recognized during the stimulation of rhizosphere bacterial proliferation by AN17. The interaction relationships among plants, beneficial fungi and rhizobacteria were explored and revealed. This result provides a sustainable approach for rhizosphere bacterial optimization and chemical fertilizer reduction in agricultural production.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1186/s12870-026-09252-9.Keywords: Plant-microbial interaction, Rhizosphere bacteria, Root exudates, Growth promotion, Phenolic acids Introduction As a perennial medicinal plant, Angelica dahurica var. formosana is produced mainly in Suining city, Sichuan Province, China. The root is the main medicinal component and contains many coumarin and volatile oil components. Coumarins (imperatorin and isoimperatorin) are regarded as the major active ingredients in the root [1]. These active ingredients can have anti-inflammatory, analgesic, antihypertensive, anticancer, and whitening effects [2]. With the wide application of A. dahurica var. formosana in food, spices, skin care and beauty, its market demand is expanding. In crop production, fertilizer is the material basis for increasing production. However, excessive reliance on fertilizers and agrochemicals to ensure productivity adversely affects environmental quality, microbial diversity and sustainability, ultimately leading to lower crop yields. To combat environmental impacts from agricultural runoff, various effective solutions have been put into practice. To combat the environmental impacts of agricultural runoff, various effective solutions have been put into practice [3]. Among them, the use of rhizosphere microorganisms to develop microbial fertilizer instead of chemical fertilizer has become a research hotspot [4]. Compared with traditional fertilizers, microbial fertilizer is low cost and environmentally friendly. It not only increases production but also improves the soil environment and microbiome [5]. The utilization of microbial resources provides novel insights into the sustainable production of medicinal plants. We obtained the growth-promoting fungus Aspergillus niger ZJ-17 (referred to as AN17) from the rhizosphere of A. dahurica var. formosana. It possesses strong nitrogen-fixing, phosphorus-solubilizing, and potassium-solubilizing activities but exhibits negligible endogenous auxin synthesis ability. Our previous pot and field experiments have confirmed that this microbial inoculant significantly enhances plant biomass and the accumulation active component of A. dahurica var. formosana through pot tests and field experiments [6, 7]. However, the physiological mechanism of AN17 and its impact on the rhizosphere microbiome remain unknown. This knowledge gap largely limits the field application of AN17. The biotechnological application of A. niger is extensive. These fungi produce a variety of organic acids and degrading enzymes and is therefore often used in the fermentation industry [8]. In agricultural cultivation, A. niger positively influences crop nutrition, yield, soil fertility and pathogen resistance. Multiple degrading enzymes secreted by A. niger promote organic matter degradation and available phosphorus mobilization [9]. It is applied to promote corn germination and growth [10]. Dephosphorylating A. niger PH1 improves maize growth and mitigates lead stress by improving photosynthetic and antioxidant responses [11]. In addition, remediation of Pb-polluted soils involves increasing phosphate release and decreasing Pb phytoavailability [12]. Moreover, citric acid and oxalic acid secreted by A. niger exhibit nematocidal activity [13]. However, the mechanisms for the improvement of host rhizosphere bacteria by A. niger remain poorly understood. The rhizosphere region surrounding roots harbors numerous microbes in the soil. Many bacteria benefit plants by secreting growth-regulating substances, inhibiting pathogen invasion, and facilitating nutrient acquisition from rhizosphere soil [14]. In actual production, artificially added microbial inoculants also recruit more growth-promoting bacteria by interacting with the indigenous soil bacteria [15, 16]. This optimized soil microbiome promotes the health and stability of the soil ecosystem to increase nutrient use efficiency and productivity [17, 18]. Microbial inoculants increase plant growth and development in multiple ways [19, 20]. In previous studies, increased Proteobacteria populations were positively correlated with increased yield and improved phytochemical characteristics in A. dahurica var. formosana [1, 21]. Therefore, when evaluating the growth-promoting mechanism of microbial inoculants, monitoring changes in the indigenous bacterial community in the rhizosphere is particularly important. In the rhizosphere, besides microbial interactions, root exudates represent another critical pathway for rhizosphere microbiome regulation. These exudates serve as primary mediators of plant‒microbe communication and function as essential energy substrates for soil microorganisms [22]. Plants modulate microbial composition via root exudate secretion [23–25]. Therefore, root exudates should be highlighted when exploring how fungal inoculant reshapes the rhizosphere microbiome. Numerous rhizosphere bacteria prefer to consume aromatic organic acids (salicylic, nicotinic, cinnamic, shikimic and indole-3-acetic acids) secreted by roots [26]. A. niger promotes ferulic acid production of plants and its conversion into organic acids such as vanillic acid [27]. This fungus degrades oat bran to elevate the release of phenolic acids [28]. These phenolics and organic acids are available as carbon sources for rhizosphere microbes [29]. Accordingly, we infer that AN17 likely stimulates phenolic acid exudation by A. dahurica var. formosana, which in turn promotes rhizosphere bacterial growth. The crucial functions of rhizosphere bacteri a and root exudates in soil ecosystems are as follows. Therefore, we investigated the changes in NPK nutrients, root exudates and the rhizosphere bacterial composition of A. dahurica var. formosana after the addition of AN17. We aimed to explore: (i) whether AN17 exerts positive effects on the host rhizosphere bacterial community, thereby improving its yield and quality, (ii) how AN17 regulates the rhizosphere bacterial community, and (iii) whether root exudates act as a key factor regulating rhizosphere bacteria. These studies will elucidate the unique growth-promoting pathway of AN17, which specifically regulates rhizosphere Proteobacteria by modulating root exudates. These studies complement the mechanism of AN17 for crop promotion and broaden the application of this microbial inoculant. Materials and methods Potting test design and soil, plant and root exudate sampling The growth-promoting performance of AN17 was evaluated via a pot experiment. AN17 was previously isolated from the rhizosphere of A. dahurica var. formosana. This fungus was preserved at the China General Microbiological Culture Collection Center (CGMCC) under accession number CGMCC N0.3852. The fungus exhibited growth-promoting properties, including nitrogen fixation and phosphorus and potassium solubilization. However, its capacity for iron carrier and auxin production is deficient. The fermentation process of AN17 was conducted via PDB. Thereafter, the mycelial suspension of AN17 was prepared through filtration and resuspension in sterile water, yielding a concentration of 10 g/L [6]. The growth-promoting effect of AN17 was determined for two cultivars in two different soils. These soils were obtained from the genuine production area of A. dahurica var. formosana (Suining city, Sichuan Province, China) and the nongenuine production area (Chongzhou city, Sichuan Province, China). The basic physicochemical properties of the soil collected from Suining were as follows: pH 7.50, organic matter 39.67 g/kg, alkali-hydrolyzable nitrogen 31.80 mg/kg, available phosphorus 17.35 mg/kg, available potassium 70.12 mg/kg, total nitrogen 0.58 g/kg, total phosphorus 0.68 g/kg, and total potassium 19.33 g/kg. Chongzhou: pH 6.37, organic matter 49.41 g/kg, alkali-hydrolyzable N 100.86 mg/kg, available P 15.42 mg/kg, available K 114.45 mg/kg, total N 1.64 g/kg, total P 0.56 g/kg, total K 19.44 g/kg. The two cultivars used were BZA001 and BZB003. They were provided by the research team of Prof. Wu Wei from Sichuan Agricultural University, originating from Suining City, Sichuan Province, China [1]. These two materials are elite cultivars developed by our team through multi-year selective breeding from various local varieties. They are now the primary recommended cultivars in Suining. The pot experiment was carried out in an open field of Sichuan Agricultural University. All pots were placed uniformly. Temperature and light were maintained under natural conditions, with uniform irrigation and routine management. Seedlings with uniform growth before the rapid growth period were selected. The mycelial suspension was watered 3 cm from the roots at 100 mL per plant. The control was watered with the same amount of sterile water. The samples were watered once a month for a total of 3 times. Some plants were dug at the harvesting stage, and the fresh weights of the aboveground parts and roots were recorded. The levels of Imperatorin and isoimperatorin were determined via high-performance liquid chromatography (HPLC) [21]. Each treatment includes six replicates, one pot per replicate and five plants per pot. Following the evaluation of the yield- and quality-enhancing effects on the host of AN17, BZA001 planted in Suining soil was selected as the main target for the subsequent study of the involvement of AN17 in plant‒bacteria interactions. The treatments were named SA (plants with sterile water), SA + AN17 (plants with AN17), and IV (initial values before treatment). The SA control group in this experiment was the same as the SA treatment group in a previous study on rhizosphere bacteria of A. dahurica var. formosana in different soils [1]. Therefore, the original yield and quality, soil nutrient, microbiome and metabolome data of the SA control group in this study were obtained from a previous study. A portion of the roots was removed at 0, 1, 5, 10, 15, 30 and 45 d after the first treatment with AN17 for phytohormone determination. The dynamic changes in the fresh weight of the roots in the two treatment groups were recorded simultaneously. Each treatment includes three replicates, one pot per replicate and five plants per pot. At harvest, following three fungal applications, the remaining seedlings were sampled. The intact plants were uprooted, and the rhizosphere soil adhering to the roots was collected after the loose soil was removed. Five plants from each replicate were combined into one sample. After washing, the roots were soaked in 250 mL of 0.05 mol/L CaCl2 solution for 4 h to collect the root exudates. This root exudate mixture was collected. A portion of the solution was freeze-dried for LC–MS analysis. Additionally, the aboveground and root fresh weights of the plants were recorded. Six independent replicates per treatment group were used. Colonization of AN17 in rhizosphere soils qRT-PCR was employed to assess AN17 colonization [30]. DNA was extracted from rhizosphere soil samples via a TIANamp Soil DNA for Soil Kit (TIANGEN, China). The specific primers forward Pf (TTGTCACATGCTCTGTAGGATT) and reverse Pr (GGAGCTTTACACGGGCA) were used for amplification. The enzyme used for the reaction was TB Green® Premix Ex Taq™ II (Tli RNaseH Plus) (Takara, Japan). The PCR products of the A. niger target gene were subsequently cloned via TA cloning. The plasmids were extracted, and a standard curve was generated via tenfold serial dilutions and corresponding Ct values. The number of A. niger colonies was quantified on the basis of the standard curve. Physiological indices of roots at the seedling stage and plant and soil nutrients at the harvesting stage The roots 0–45 days after the first AN17 inoculation were used to measure the changes in related phytohormones. A plant ELISA kit (mlbio, China) was used to determine the levels of auxin (IAA), jasmonic acid (JA) and reactive oxygen species (ROS) via an enzyme-linked immunosorbent assay [31]. The plant samples at harvest were killed at 105 °C and then dried at 55 °C to a constant weight. The sulfuric acid and hydrogen peroxide digestion method was used for the determination of NPK accumulation in plants [32]. The soil samples at harvest were shade-dried. The physicochemical properties and nutrient changes of the samples were analyzed according to previously described methods [1, 33]. A graphical representation of these data was generated with GraphPad Prism 8.3.0. Soil 16S rDNA amplicon sequencing DNA extraction and 16S rDNA amplicon sequencing were performed on rhizosphere soil samples collected at harvest. The methods used for genome extraction, library construction and raw data processing were consistent with those used in previous studies [1]. The raw SA and SA + AN17 data were stored in the NCBI Sequence Read Archive (SRA) database under accession numbers PRJNA1058974 and PRJNA1230362. LC‒MS detection of root exudates The freeze-dried sample processing, analysis system, database comparison and original data analysis were consistent with previous studies of LC-MS detection [1]. UHPLC-MS/MS analysis was conducted via a Vanquish UHPLC system (Thermo Fisher, Germany) integrated with an Orbitrap Q ExactiveTM HF mass spectrometer (Thermo Fisher, Germany) at Novogene Co., Ltd. (Beijing, China). The raw data files were imported into CD 3.1 database search software for processing. The data were compared against the mzCloud, mzVault, and Masslist databases. The raw data of nontargeted metabolomics techniques of SA and SA + AN17 treatments were deposited in the MetaboLights database with the numbers MTBLS9259 and MTBLS12285. Isolation and identification of recruited bacterial strains After fermentation for 2 d using 1/2 PDB, the culture was subjected to sterile filtration to prepare the fermentation broth of AN17. The bacterial strains were isolated via the dilution plating method and grown in TSB. From the rhizosphere soils of the two treatment groups, 338 and 439 single colonies were isolated, which were named BS (bacteria from SA) and BA (bacteria from SA + AN17), respectively. Four working solutions were used to assess bacterial recruitment in vitro: root exudates from the SA and SA + AN17 groups, AN17 sterile fermentation broth, and sterile water (CK) as a control. To avoid nutrient overload, 1/5 TSB was used as the medium. Each group received 10% working solution and a 1% bacterial suspension (1.0 × 108 CFU/mL), with three independent biological replicates. Growth curves were generated on the basis of OD600 measurements, and the area under the curve (AUC) was determined. The normalized growth ratio (GR) (AUC of SA + An/AUC of SA) was used to determine bacterial recruitment by root exudates. The GR of AN17 fermentation broth to sterile water was used to assess direct recruitment by AN17. A GR > 1 and P < 0.05 indicated promotion. GraphPad Prism 8.3.0 was used to generate a volcano map to visualize the recruitment of strains. To quickly screen for more recruited strains, strains exhibiting a GR greater than 1.25 were chosen for further identification. The selected bacteria were identified via molecular biology, as described by Jiang [1]. Effects of root exudates on recruited bacteria On the basis of the results of the root exudate analysis, four compounds (4-chlorophenol, 2-oxoadipic acid, pyrogallol, and vanillic acid) were selected for studying the recruited bacterial strains. Solutions of 1, 0.1, 0.01, 0.001, and 0.0001 mM were prepared. The aforementioned in vitro method was used with 10% compound solutions added to the culture. Sterile water served as the negative control, whereas SA and SA + AN17 root exudates were positive controls. Three independent biological replicates were used per group. The GR was calculated with sterile water as the reference. The GR significantly exceeded that of SA root exudates, indicating that GR promoted bacterial growth. Heatmaps were plotted with log2-transformed values. Growth-promoting potential of the screened strains The growth-promoting potential of the recruited bacteria was assessed via selective media [34]. 7 µL bacterial suspension were spotted onto plates and photographed after 3-5 days. N fixation activity was evaluated on the basis of bacterial growth. P solubilization was measured by the appearance of transparent zones. K solubilization was assessed by the degree of yellowing. Siderophore production is indicated by yellow transparent zones. IAA production was determined on the basis of pink coloration in fermentation broth and quantified by OD530 absorbance. Three independent biological replicates were used per group. A pot experiment with seedlings was used to quickly determine the growth-promoting manifestations of each bacterial genus. The recruited strains were fermented via TSB media. After filtration and resuspension in sterile water, bacterial suspensions with an OD600 of 1 were prepared. The bacterial suspensions of different strains from the same genus were mixed in equal amounts. The sources of the soil and A. dahurica var. formosana were the same as those used in the SA treatment (BZA001 in the Suining soil). Seedlings whose growth was consistent before the rapid growth period were selected. The bacterial mixture was pooled at a distance of 3 cm from the root, with 100 mL per plant. The control was watered with the same amount of sterile water. The plants were harvested 45 days later, and the fresh weights of the aboveground parts and roots were recorded. The contents of imperatorin and isoimperatorin were determined via high-performance liquid chromatography [21]. Each treatment includes six replicates, one pot per replicate and five plants per pot. Statistical analysis Bacterial community analysis utilized DADA2 or the deblurring module in QIIME1 for denoising. The final amplicon sequence variants (ASVs) were obtained by filtering out sequences whose abundances were less than 5. ASVs were taxonomically annotated via classify-sklearn against the SILVA database (Release 138.1). The observed ASVs and the Shannon index were calculated via QIIME (version 1.9.1). PCA was performed via the ade4 and ggplot2 packages of R (version 2.15.3). LEfSe software (LEfSe 1.0) was used to analyze the LDA results. Welch's t-test was used to identify differential taxa between groups, and the results were corrected for multiple comparisons using the Benjamini–Hochberg method. For CCA, the vegan package of R (version 3.4.1) was used. For the Spearman correlation analysis, the corr.test function in the psych package of R (version 3.4.1) was used to calculate the Spearman correlation values between species and environmental factors and to test their significance. The correlations were tested for significance using t-tests, with Benjamini–Hochberg correction for multiple comparisons. A heatmap function was subsequently used for visualization. Significantly differential metabolites are variable importance projection (VIP) > 1, P-value < 0.05 and FC ≥ 1.2 or FC ≤ 0.83. These metabolites were annotated via the HMDB, KEGG and LIPIDMaps databases. The PCA results were calculated via metaX. Spearman correlation analysis between metabolites and bacteria was performed via Python 3.7.9. We applied univariate analysis (t-test) to calculate the statistical significance (P-value). The bubble map was drawn via ggplot2 in the R package. A volcano plot was produced with GraphPad Prism 8.3.0. Results Promotion effect of AN17 on rhizosphere nutrient utilization A. dahurica var. formosana at harvest after AN17 inoculation was analyzed. Overall, AN17 inoculation promoted the utilization and accumulation of available nutrients in the rhizosphere. The qRT-PCR results revealed that AN17 significantly colonized the rhizosphere soil at harvest (Fig. 1B, C). Previously, we demonstrated that AN17 promoted the yield and quality of two cultivars (BAZ001 and BZB003) in different soils (Suining and Chongzhou) (Figs. 1A, S1). In this pot experiment, the accumulation of nitrogen (N), phosphorus (P) and potassium (K) in the roots of plants inoculated with AN17 increased significantly. In addition to elevated aboveground N accumulation, the contents of aboveground P and K were also significantly improved after AN17 inoculation (Fig. 1D-I). No significant differences in the pH, total nitrogen (TN), total phosphorus (TP) or total potassium (TK) of the rhizosphere soil (Fig. 1J-M). The growth of A. dahurica var. formosana depleted soil organic matter (OM) and fast-acting nutrients. Notably, AN17 inoculation reduced OM and increased the available phosphorus (AP) and available potassium (AK) in rhizosphere soil (Fig. 1N-Q). These findings suggested that AN17 indeed promoted soil OM mineralization and NPK nutrient mobilization, thereby enhancing root NPK accumulation and ultimately improving crop yield. Plant NPK accumulation and rhizosphere soil nutrient changes at the harvest stage after AN17 treatment. A Growth-promoting effect of AN17. B Standard curve of A. niger gene copies via qRT‒PCR. c Copies of A. niger in rhizosphere soil determined by qRT-PCR (P < 0.05). D-I NPK accumulation in aboveground parts and roots (P < 0.05). J-Q Changes in rhizosphere soil nutrients (P < 0.05). SA, sterile water treatment; SA + AN17, AN17 treatment; IV, initial value before planting Root hormone levels at the initial stage of AN17 inoculation The roots of the seedlings were continuously measured within 45 d of inoculation with AN17. Differences in seedling root fresh weight emerged at 5 days post-inoculation (Fig. 2A). IAA and JA were significantly upregulated from 1-10 d after treatment (Fig. 2B, C), while ROS levels were significantly decreased (Fig. 2D). Obvious primary root thickening and lateral root proliferation were observed at 45 d of AN17 treatment (Fig. 2E, F). Therefore, we hypothesized that, in addition to its nutrient strategy, AN17 promoted yield and quality by regulating the growth promotion pathway related to IAA and the resistance pathway related to JA and ROS. Root growth and hormone of the seedlings from 0 to 45 d after AN17 treatment. Root fresh weight (A), IAA (B), JA (C) and ROS changes (D) from 0-45 d after AN17 treatment. E, F Changes in primary roots and lateral roots after AN17 treatment for 45 d Analysis of rhizosphere bacterial community diversity in response to AN17 treatment The rhizosphere bacterial community was characterized through 16S rDNA amplicon sequencing. We obtained 8473 ASVs after quality filtering (minimum abundance threshold: 5). Individual samples presented between 1653 to 2247 ASVs. AN17 inoculation did not significantly affect the number of ASVs but markedly reduced the Shannon index (Fig. 3A, B). Principal component analysis (PCA) revealed clear separation between the AN17 treatment group and the control group. The application of AN17 significantly changed the rhizosphere bacterial community structure (Fig. 3C). Proteobacteria dominated the rhizosphere bacterial community (Fig. 3D), which aligns with prior findings (Jiang et al., 2023b; Jiang et al., 2024). Linear discriminant analysis (LDA) revealed that Proteobacteria were the most significantly enriched taxa following AN17 inoculation (adjusted P < 0.05) (Fig. 3E), with Pseudomonas being significantly enriched in the soil treated with AN17 (Fig. S2). In the canonical correlation analysis (CCA), AN17-inoculated soil samples were positively correlated with soil AP and AK contents and negatively correlated with OM (Fig. 3F). Moreover, in the Spearman correlation analysis, Proteobacteria showed an extremely significant negative correlation with OM (adjusted P < 0.05) (Fig. 3G). These results suggested that Proteobacteria are closely involved in rhizosphere OM utilization and NPK nutrient transformation. Diversity of the bacterial community in rhizosphere soil and the causes of bacterial differences. Changes in the observed ASVs (A) and Shannon indices (B) of the bacterial community in the rhizosphere. C PCA of the bacterial community. D Histogram of bacterial relative abundance at the phylum level. E Histogram of the LDA value distribution (LDA > 3.5). F Canonical correlation analysis (CCA) between soil nutrients and soil samples. G Spearman correlation analysis of bacteria at the phylum level (top 10) and soil nutrients. H Growth curves of Pseudomonas BS1–83 in each working fluid. I Volcanic map of strain growth preference for two root exudates (SA + AN17 vs SA) (P < 0.05). J Volcanic map of AN17-induced growth enhancement in bacterial strains. K Categories and quantities of bacteria recruited from the root exudates (SA + AN17/SA) or AN17 solution (AN17/CK) after treatment with AN17. BS: Bacterial strains isolated from the aseptic water treatment group (SA); BA: Bacterial strains isolated from the AN17 treatment group (SA + AN17) Factors influencing rhizosphere bacteria changes caused by AN17 treatment Microbial interrelationships and root exudate-mediated effects on rhizobacterial communities were investigated. We isolated 393 and 439 bacterial strains from soils treated with sterile water and AN17, respectively. To determine the main reason for the changes in rhizosphere bacteria, we collected root exudates from the two treatment groups and prepared fermentation solutions of AN17. The root exudates or fermentation solution of AN17 were added to cocultures with isolated rhizosphere bacterial strains in vitro. We found that both approaches promoted the growth of many bacterial strains (Fig. 3H-J). For rapid screening of recruited bacteria, bacteria with a growth ratio (GR) greater than 1.25 were identified. According to the results of the molecular identification, most of the recruited bacterial strains belonged to Proteobacteria (Table S1, S2). Both AN17 fermentation broth and host root exudates inoculated with AN17 promoted the growth of Pseudomonas and Enterobacter. AN17 not only directly enriched these rhizosphere bacteria but also indirectly regulated the bacteria by modifying host root exudates. In addition, the growth of Pantoea, Klebsiella, Variovorax and Stenotrophomonas was regulated only by the root exudates inoculated with AN17 (Fig. 3K). The root exudates after inoculation with AN17 promoted the growth of more bacteria and more species of bacteria than did the AN17 fermentation broth (Figs. 3K, S3). Therefore, the metabolism of the root exudates of A. dahurica var. formosana after AN17 treatment was highly important for altering the structure of the beneficial Proteobacteria in the rhizosphere. Impact of host root exudates on the chemotaxis and growth of Proteobacteria The root exudates of A. dahurica var. formosana caused by AN17 were analyzed via nontargeted metabolomics. There were significant differences in the root exudate patterns between the two treatment groups according to the PCA results (Fig. 4A). The root secretion activity improved after AN17 inoculation. By applying the selection criteria of P < 0.05 and FC > 1.2, we identified significant differences in 506 root exudate components. Of these, 379 components were significantly upregulated, whereas 127 were notably reduced (Fig. 4B). These upregulated components of the root exudates were involved in microbial metabolism in diverse environments, the biosynthesis of alkaloids derived from the shikimate pathway and tryptophan metabolism according to the KEGG enrichment analysis (Fig. 4C). A total of 24 components enriched in microbial metabolism in different environments included citric acid, benzoic acid and vanillic acid. These organic acids might be involved in the regulation of rhizosphere bacteria by AN17 (Fig. 4D, Table S3). Flavonoid biosynthesis was significantly upregulated in the negative ion mode according to the KEGG analysis (Fig. S4). A total of 6 upregulated compounds, including IAA synthetic precursors, were enriched in tryptophan metabolism (Table S4). The upregulation of this pathway was consistent with the determination of the IAA content in the roots (Fig. 2B). Thus, AN17 promoted host growth by increasing the synthesis of tryptophan and IAA in the roots of plants. Analysis of root exudates after AN17 inoculation and their relationships with Proteobacteria. A PCA of the root exudates in the two treatment groups. B Volcanic plot of differential root exudates (P < 0.05, FC > 1.2). C Enrichment analysis diagram of the KEGG pathways (top 10). D Correlation analysis of biomarkers significantly enriched in the rhizosphere and differential root exudate components in microbial metabolic pathways in different environments (P < 0.05) Four root exudate compounds that recruit bacteria of the phylum Proteobacteria The root exudate changes in the AN17 treatment caused changes in the rhizosphere bacterial community. We selected 24 compounds involved in the top pathway in the KEGG analysis for correlation analysis. Correlation analysis with the bacterial biomarkers in the LDA analysis, 4-chlorophenol, 2-oxoadipic acid, pyrogallol, and vanillic acid exhibited the strongest correlation with Proteobacteria (R2 > 0.9, adjusted P < 0.05) (Fig. 4D). To verify the regulatory effects of the 4 compounds on Proteobacteria, representative strains of different genera were selected from the recruited strains for subsequent validation experiments. The isolated representative strains were cultured in media supplemented with the four individual compounds, and their growth ratios (GR) were determined. The four compounds had different effects on promoting the growth of the strains. All 4 compounds promoted the growth of Pseudomonas and Klebsiella. and Stenotrophomonas significantly, particularly at 10–3 mmol/L and 10–4 mmol/L. Enterobacter proliferation was promoted by pyrogallol and vanillic acid. The growth of Pantoea was more affected by 2-oxoadipic acid, pyrogallol and vanillic acid. Although these 4 compounds promoted the growth of Variovorax, their effects were not significantly different from those of the root exudate treatment without AN17 (Fig. 5). Effects of 4 phenolic acid components on the growth of different strains from the rhizosphere. “*” indicates a significant difference from the SA group (P < 0.05) Growth-promoting effects of recruited rhizosphere Proteobacteria strains on A. dahurica var. formosana The growth-promoting potential of the recruited strains was measured via selective media and pot tests. Stenotrophomonas BS2-86 exhibited weak phosphorus solubilization. Pseudomonas BS1-83 and Variovorax BA1-8 showed no potassium-solubilizing activity. The remaining of the strains presented varying degrees of nitrogen fixation, phosphorus solubilization, potassium solubilization and iron carrier production activities (Fig. 6A). These enriched bacterial community indeed increased the availability of NPK nutrient. During a 45-day pot experiment, these bacterial genera exerted distinct growth-promoting effects on the seedlings. With the exception of Stenotrophomonas, all other bacteria increased the fresh weight of the seedlings (Fig. 6B, C). Moreover, Enterobacter, Klebsiella, Pantoea and Stenotrophomonas increased the contents of IMP and isoimperatorin. Ultimately, all six bacterial genera increased the accumulation of the main active components (Fig. 6C). Collectively, bacteria recruited by Proteobacteria in response to AN17 treatment increased NPK nutrient availability, facilitating host growth and nutrient absorption. Additionally, these strains exhibited low IAA production capacity (Fig. S5), indicating that the IAA increase was produced mainly by host metabolic changes induced by AN17. Growth-promoting capabilities of the recruited bacterial strains. A Nitrogen fixation capacity was evaluated via selective medium growth. Phosphate solubilization was determined by transparent zone formation. The abilities of potassium release and siderophore production were measured on the basis of the degree of yellowing on specific media. B The growth-promoting ability of recruited strains in pot experiments (45 d after treatment). C Promotion of the fresh weight and imperatorin and isoimperatorin levels of seedlings by different bacterial genera. Abbreviations: N, nitrogen fixation. P, phosphorus dissolution. K, potassium solubilization. Fe, Siderophore production Discussion The rhizosphere-promoting fungus AN17 significantly improved the yield and quality of A. dahurica var. formosana. However, the interactions between AN17 and the rhizosphere bacterial community are not clear. Here, we demonstrated the advantage of AN17 in coordinating and optimizing the rhizosphere bacterial community via host root exudates. This predominance significantly contributes to soil health enhancement and ecosystem stability. AN17 inoculation altered the bacterial composition of the host rhizosphere. Proteobacteria exhibited the most significant increase in abundance (Fig. 3). Proteobacteria are well-documented predominant growth-promoting bacteria in the rhizosphere [34, 35]. Higher Proteobacteria abundance is recognized as a key characteristic of high yield and quality in the A. dahurica var. formosana rhizosphere [21]. This medicinal plant secretes more flavonoids to recruit these bacteria, especially Pseudomonas [1]. Consequently, the increase of Proteobacteria is an improvement in rhizosphere bacteria caused by AN17, promoting a more beneficial bacterial composition. In addition, the abundance of Nitrososphaeraceae associated with nitrogen cycling also increased significantly (Fig. 3E), which might facilitate the transformation and utilization of rhizosphere nutrients [36]. It should be noted that this study did not investigate the regulatory effects of AN17 on rhizosphere pathogenic bacteria, neutral bacteria and other beneficial bacteria. This limitation wiould be addressed in future research. The changes in rhizosphere bacteria are primarily driven by host root exudates and microbial interspecific interactions. Both factors collectively facilitate the enrichment of diverse Proteobacteria species (Fig. 3I, J). Notably, root exudates recruited more bacteria, such as Pantoea, Klebsiella, Variovorax and Stenotrophomonas (Fig. 3K). Pseudomonas [37–39], Enterobacter [40] and Klebsiella [34, 41] are well-characterized plant growth-promoting bacteria that promote nutrient uptake, lateral root development and stress resistance [37]. Variovorax coordinates IAA secretion by other rhizosphere bacteria to maintain root growth and development [42]. Pantoea also promote NP availability, regulate phytohormone secretion, and participate in plant cell wall formation and root exudate metabolism [43, 44]. Stenotrophomonas promotes sunflower growth through the hormones associated with root growth and stress [45] and mitigates soil pesticide residue pollution [46]. In this study, these bacteria were inversely proportional to the rhizosphere OM and directly proportional to AP (Fig. 3F, 3G). After testing, these recruited strains were weakly IAA-producing, but they improved the availability of NPK nutrients and significantly enhanced the growth of seedlings and the accumulation of active components (Fig. 6). Therefore, we are even more convinced that the increase in Proteobacteria is one of the ways by which AN17 promotes plant growth. Moreover, host root exudates after AN17 inoculation are indispensable for recruiting beneficial taxa of Proteobacteria to the rhizosphere. Root exudates are essential for substance exchange and signal transmission between plants and soil, as well as for attracting beneficial bacteria to the rhizosphere [47]. Under nitrogen deficiency, maize roots promoted nitrogen uptake by secreting flavonoids to enrich rhizosphere Oxalobacter [48]. Citric acid in cucumber root exudates recruited Pseudomonas to inhibit Fusarium oxysporum infestation [49]. Similarly, root exudates of A. dahurica var. formosana have an ecological function. AN17 greatly enhanced root vigor, resulting in the upregulation of 379 root secretions. These compounds are involved mainly in microbial metabolism in diverse environments, the biosynthesis of alkaloids derived from the shikimate pathway, tryptophan metabolism, and flavonoid biosynthesis (Fig. 4). The shikimic acid metabolism pathway is the upstream pathway for coumarin and flavonoid synthesis [50]. Coumarin is the core active component of A. dahurica var. formosana [51]. This medicinal plant promotes nutrient uptake by exuding more flavonoids to recruit Pseudomonas [1]. AN17 promoted the secretion of flavonoid compounds. Thus, AN17 not only promoted the accumulation of coumarin but also reshaped root exudate profiles to drive the targeted the enrichment of Proteobacteria. Following AN17 inoculation, the secretion levels of 4-chlorophenol, 2-oxoadipic acid, pyrogallol and vanillic acid significantly increased. While correlation analysis suggested a link between these compounds and Proteobacteria abundance, we experimentally confirmed that each of the four compounds promoted the growth of representative strains in vitro (Fig. 5). The secretion of organic acids directly enhances the release of available nutrients from the soil [52]. and serves as a carbon source for rhizosphere microorganisms [29]. Phenolic acids are mostly associated with plant resistance. For example, tea plants secrete phenolic acids and flavonoids that regulate their rhizosphere microbial community to improve gray blight resistance [53]. We proposed that these phenolic acids are key components driving the proliferation of Proteobacteria. Future research should further investigate their roles in bacterial colonization and chemotaxis, to distinguish their function as either signaling molecules or energy substrates. Vanillic acid secreted by roots exerts positive microecological functions. It recruits beneficial endophytes of Abelmoschus esculentus that are involved in disease resistance [54]. The symbiosis between arbuscular mycorrhizal fungi and the host regulates the secretion of vanillic acid in roots [55]. In addition to vanillic acid, we discovered and demonstrated the role of another compound in promoting plant-microbial interactions for the first time. The other three compounds have been identified in root exudates of various plant species [56, 57], but there are few reports on their roles in plant–microbe interactions. Among these 3 compounds, 4-chlorophenol enhances plant peroxidase activity and thereby improves plant defense against pests and diseases [58]. Pseudomonas is capable of degrading and utilizing this compound [59]. This observation is in line with our result that low concentrations of 4-chlorophenol facilitated the growth of Pseudomonas. 2-Oxoadipic acid is a significant metabolite in the TCA cycle and in the biosynthesis of tryptophan and lysine. Pyrogallol possesses antioxidant properties and inhibits the growth of several phytopathogens [60]. We hypothesized that these phenolic compounds promote the proliferation of Proteobacteria while may inhibiting the growth of certain pathogens. Furthermore, AN17 increased tryptophan metabolism in the roots (Fig. 4). This finding was consistent with the observed increase in auxin content in roots (Fig. 2). Both AN17 and the representative strains exhibited low auxin synthesis capacity. Consequently, root auxin accumulation was mainly attributed to endogenous hormone synthesis in host plants induced by AN17. Auxin regulated root architecture, promoted root growth and expanded nutrient absorption area [61], matching the thickening of primary roots and increased lateral root formation (Fig. 2). AN17 inoculation elevated root JA content and reduced ROS accumulation. JA and ROS acted as core signals activating plant secondary metabolism and modulate the biosynthesis of alkaloids, phenolics, terpenoids and phenylpropanoids [62]. Moderately reduced ROS alleviated root oxidative damage and coordinated with JA to regulate plant growth [63]. Auxin, JA and ROS were key signaling molecules for microbe-mediated growth and metabolic reprogramming in hosts [64]. Collectively, we proposed that AN17 exerted multiple physiological effects via modulating endogenous hormones. On the hand, it induced endogenous auxin synthesis to improve root development and nutrient uptake. On the other hand, elevated JA and balanced ROS homeostasis might activate the shikimate pathway and flavonoid metabolism, facilitating the production and release of root exudates. The detailed regulatory mechanisms remained to be further verified. Future work would explore the coordinated synthesis of auxin and JA, and clarify the molecular basis underlying AN17-mediated regulation of root development and secondary metabolism via hormone signaling. Conclusion In summary, the growth-promoting mechanism of AN17 were added and outlined. AN17 colonizes the A. dahurica var. formosana rhizosphere. First, AN17 promoted the synthesis of growth-related hormone and the accumulation of active ingredients in host roots. Second, it improves nutrient uptake in the rhizosphere soil. Third, it directly or indirectly enriched various beneficial genera of Proteobacteria. Root exudates were important substances throughout this regulatory process. AN17 enhances the exudation of root phenolic acids, which further facilitated the recruitment of beneficial Proteobacteria. There are multiple ways to increase the abundance of Proteobacteria in the rhizosphere and restructure soil microbial communities (Fig. 7). Our study complements the potential of AN17 in improving rhizosphere bacterial communities. In previous research, inoculation with AN17 improved the yield of Angelica dahurica var. formosana by 37.84%-73.59% under different soils and cultivars [6]. This study only investigated the growth-promoting performance of the most effective treatment combination. Whether the growth-promoting effects of AN17 remain consistent across different cultivars and soil types required further exploration. Growth promotion strategies of AN17 by stimulating resident rhizobacteria in Angelica dahurica var. formosana Supplementary Information Supplementary Material 1. Acknowledgements This study was funded by grants from the Key R & D projects of the Sichuan Provincial Department of Science and Technology (Grant No. 2021YFYZ0012), the Sichuan Qihuang Scholar Capability Enhancement Project (Grant No. 2024-55), and the Key Discipline Construction Project of Traditional Chinese Medicine in Sichuan Province (Grant No. 2021-16-4). Authors’ contributions M. J. performed the experiments, prepared the draft manuscript, and was the major contributor. R. L., Z. L. and H. X. helped with the data analysis. J. Z., W. Q. and Z. H. helped with the preparation of the manuscript. Y. C., D. F. and W. W. designed and supervised the research and finalized the manuscript. All the authors contributed to the article and approved the submitted version. Funding This study was funded by grants from the Key R & D projects of the Sichuan Provincial Department of Science and Technology (Grant No. 2021YFYZ0012), the Sichuan Qihuang Scholar Capability Enhancement Project (Grant No. 2024–55), and the Key Discipline Construction Project of Traditional Chinese Medicine in Sichuan Province (Grant No. 2021–16-4), and the Double-Support Plan of Disciplinary Construction in Sichuan Agricultural University. Data availability The raw 16S rDNA amplicon sequencing data of the rhizosphere bacterial communities in SA and SA + AN17 are available in the Sequence Read Archive (SRA) database of NCBI with the accession codes PRJNA1058974 and PRJNA1230362. The raw nontargeted metabolomics data of the root exudates from the SA and SA + AN17 treatment groups are available in the MetaboLights database under the numbers MTBLS9259 and MTBLS12285. Declarations Ethics and consent to participate Not applicable. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Material 1. Data Availability Statement The raw 16S rDNA amplicon sequencing data of the rhizosphere bacterial communities in SA and SA + AN17 are available in the Sequence Read Archive (SRA) database of NCBI with the accession codes PRJNA1058974 and PRJNA1230362. The raw nontargeted metabolomics data of the root exudates from the SA and SA + AN17 treatment groups are available in the MetaboLights database under the numbers MTBLS9259 and MTBLS12285.