Metabolite-Mediated Alleviation of Iron Deficiency and Growth Promotion of Malus hupehensis by Bacillus licheniformis LCDD6 in Calcareous Soil

TL;DR

A soil bacterium helps apple seedlings grow better in iron-poor soil by producing specific metabolites that improve iron and phosphorus uptake.

Contribution

The study identifies bacillibactin and IAA as key metabolites produced by Bacillus licheniformis LCDD6 that alleviate iron deficiency and promote plant growth in calcareous soils.

Findings

Bacillibactin extract most strongly enhanced plant growth and iron accumulation.

IAA preferentially stimulated root development and phosphorus accumulation.

Bacillibactin enriched beneficial fungi like Coprinellus, which correlated with improved plant growth.

Abstract

Calcareous soils are typically deficient in essential nutrients such as iron, phosphorus, and potassium, which frequently results in nutrient deficiency in fruit trees. Bacillus licheniformis LCDD6 markedly enhanced Malus hupehensis seedling growth and plant iron nutrition in calcareous soil. This study aimed to elucidate the mechanism underlying these beneficial effects of strain LCDD6 under iron deficiency. Transcriptomic analysis revealed that iron deficiency induced metabolic reprogramming in strain LCDD6, characterized by a significant upregulation of genes involved in the biosynthesis of the siderophore bacillibactin and plant growth hormone indoleacetic acid (IAA). Consistently, metabolomic profiling identified bacillibactin and IAA as the dominant metabolites produced under iron-deficient conditions. A 60-day pot experiment further demonstrated that the cell-free fermentation broth of strain LCDD6 significantly enhanced plant growth and rhizosphere soil enzyme activities. The crude bacillibactin extract derived from the fermentation exerted the strongest effects on plant growth and iron accumulation, whereas IAA preferentially stimulated root development and promoted plant phosphorus accumulation. Additionally, different metabolites exerted distinct and selective effects on the rhizosphere microbial community, with fungi showing stronger and more metabolite-specific responses than bacteria. The crude bacillibactin extract enriched fungal taxa, particularly Coprinellus, which showed strong positive correlations with plant growth traits and iron accumulation, while Stachybotrys, enriched under IAA treatment, was positively correlated with plant phosphorus content. Overall, strain LCDD6 promotes plant growth under iron-deficient conditions through the coordinated action of multiple metabolites, with bacillibactin as the primary contributor and IAA providing complementary effects. These findings offer mechanistic insight and a scientific basis for developing Bacillus-based biofertilizers to improve nutrient acquisition in calcareous soils.

Figures

Full text

1. Introduction

Deficiencies in essential mineral elements, including iron, phosphorus, potassium, zinc, manganese, and copper, can severely limit fruit tree productivity [1]. Iron is an essential nutrient for many physiological processes in plants, including chlorophyll synthesis, photosynthesis, respiration, and various redox and enzymatic reactions [2]. Therefore, adequate iron nutrition is crucial for maintaining plant vitality and ensuring stable yield and quality [3,4]. However, iron deficiency is a widespread agricultural constraint, especially in calcareous soils, where iron mainly exists in the form of insoluble ferric oxides or hydroxides, resulting in extremely low bioavailability [5]. Calcareous soils are alkaline soils typically rich in calcium carbonate (CaCO_3_), which can significantly affect nutrient availability due to the high pH [6]. These soils are widely distributed in arid and semi-arid regions, where fruit tree cultivation is prevalent [7,8]. Therefore, improving the acquisition of essential nutrients by fruit trees has become an urgent challenge in orchard management and soil remediation.

Malus hupehensis Rehd. var. pingyiensis Jiang (Pingyi Tiancha, PYTC) is a widely used Chinese apple rootstock valued for its strong abiotic stress resistance, grafting compatibility, and ability to support high-quality apple production [9,10]. However, like many fruit trees grown in calcareous soils, M. hupehensis is highly susceptible to iron deficiency, resulting in chlorosis, reduced growth, and impaired root function [11]. Considering the agronomic importance of apple cultivation in arid and semi-arid regions and the sensitivity of M. hupehensis to iron availability, this species represents an ideal model plant for investigating microbial strategies that alleviate iron stress.

To cope with iron scarcity, both plants and soil microorganisms have evolved specialized iron-acquisition strategies [12]. One of the most effective microbial mechanisms is the secretion of siderophores, a class of low-molecular-weight, high-affinity iron-chelating compounds capable of solubilizing Fe^3+^ under alkaline and aerobic conditions [13,14]. By forming stable Fe^3+^-siderophore complexes and transporting iron through specific uptake systems, microbial siderophores can substantially enhance iron availability in the rhizosphere. Among the major siderophore classes, catecholate siderophores, such as bacillibactin produced by many Bacillus spp., exhibit particularly strong Fe^3+^ binding affinity and play a prominent role in improving plant iron uptake in alkaline soils [15].

Plant growth-promoting rhizobacteria (PGPR) have therefore gained considerable attention as sustainable tools for addressing iron deficiency in calcareous soils [16,17]. Bacillus licheniformis, a Gram-positive and environmentally resilient PGPR [18,19], is known to secrete a variety of bioactive metabolites including siderophores, plant hormones, enzymes, and antimicrobial compounds [20,21,22]. These traits contribute to pathogen suppression, improved nutrient acquisition, and enhanced plant growth. However, while the overall growth-promoting effects of B. licheniformis have been widely reported, the primary metabolites and their functions have rarely been reported under iron-deficient conditions. This knowledge gap limits the targeted development and application of Bacillus-based biofertilizers, particularly for calcareous soils in fruit production regions.

In this study, we focused on B. licheniformis LCDD6, a strain that has been shown to possess multiple plant growth-promoting characteristics, including phosphate and potassium solubilization, siderophore production, and IAA production [23]. By integrating transcriptomics and metabolomic analysis, and the pot experiments, we aimed to systematically elucidate: (i) the key growth-promoting metabolites produced by strain LCDD6 under iron-deficiency conditions and (ii) how these metabolites contribute to enhanced plant growth. This research not only advances fundamental understanding of PGPR–plant–soil interactions under iron-deficient conditions but also provides a scientific basis for developing siderophore-driven microbial fertilizers to enhance fruit tree production in calcareous soils.

2. Materials and Methods

2.1. Strain and Culture Conditions

B. licheniformis LCDD6 used in this study was isolated from a mixed fermentation pile of corn cobs and chicken manure, as reported in a previous study [23], and its complete genome sequence has been deposited in GenBank (accession number: CP065029.1). The strain was cultured in LB medium (yeast extract 5 g, peptone 10 g, NaCl 10 g, distilled water to 1 L) and SA iron-deficient medium (sucrose 20 g, L-asparagine 2 g, K_2_HPO_4_ 0.5 g, MgSO_4_·7H_2_O 0.5 g, distilled water to 1 L) [24]. Cultures were incubated at 37 °C with shaking at 180 rpm.

2.2. Plant Material and Soil Characteristics

Malus hupehensis Rehd. var. pingyiensis Jiang (Pingyi Tiancha, PYTC), a highly iron-deficiency sensitive apple rootstock, was used as the test plant. The soil was collected from Tai’an, Shandong Province, China (36°10′5.58″ N, 117°09′19.63″ E). The soil pH was 9.23, indicating strongly alkaline soil conditions. The calcium carbonate content of the soil was determined using a neutralization titration method according to standard procedures described by Bao (2000) [25], and the measured CaCO_3_ content was 47.1 g/kg (n = 3, oven-dried soil basis). Nutrient analysis showed that the soil contained 2.57 mg/kg available iron, 20.6 mg/kg available phosphorus, and 186 mg/kg available potassium. Based on the above indicators, the soil was classified as an iron-deficient calcareous soil [26].

2.3. Effect of LCDD6 on the Growth of M. hupehensis

2.3.1. Cultivation of M. hupehensis Seedlings

M. hupehensis seeds stored at −80 °C were sown in a substrate of peat soil, perlite, and vermiculite (3:1:1, v/v/v) and grown in a greenhouse under controlled conditions (25 ± 2 °C, 60–70% relative humidity). Seedlings were watered every three days. After 45 days, uniform three-leaf seedlings were transplanted into pots (23 cm × 18 cm, diameter × height) containing 2.5 kg of calcareous soil, with five biological replicates per treatment. Seedlings were acclimated for seven days prior to treatment.

2.3.2. Inoculation and Growth-Promotion Assay

Strain LCDD6 was cultured in LB medium at 37 °C and 180 rpm for 12 h. The culture, which included both bacterial cells and culture metabolites, was diluted 50-fold with sterile water and applied to the root zone of M. hupehensis seedlings (200 mL per pot), with sterile LB medium serving as the control. After 60 days, seedlings were removed intact, and plant tissues and rhizosphere soil were collected separately for the determination of agronomic traits and nutrient-related parameters.

2.4. Determination of LCDD6 Metabolites Under Iron-Deficient Conditions

2.4.1. Determination of Sampling Time Points

B. licheniformis LCDD6 was inoculated into SA liquid medium, and bacterial growth was monitored by measuring OD_600_. Spore formation was quantified by heating 1 mL of fermentation broth at 85 °C for 15 min, followed by plating on LB agar. In parallel, siderophore and IAA production in the culture supernatant was monitored using the Arnow and Salkowski assays, respectively [27,28]. These measurements were used to determine the optimal sampling time points for metabolite and cell collection.

2.4.2. Collection of Metabolite and Cell Samples

Strain LCDD6 was cultured in LB and SA media to the designated sampling time points. Cultures were centrifuged at 4000× g for 10 min at 4 °C to separate supernatants and bacterial pellets, with three biological replicates per group. Pellets were washed twice with pre-chilled PBS, resuspended, frozen in liquid nitrogen, and stored at −80 °C for subsequent analysis.

Siderophore and IAA levels in the supernatant were determined in the fermentation supernatant of LCDD6 cultured in both LB and SA media. The supernatant of the SA group was further centrifuged (16,000× g, 10 min) and extracted three times with ethyl acetate (2:1, v/v). The combined extracts were concentrated at 42 °C using a rotary evaporator, and the ethyl acetate was evaporated. The residue was then dissolved in methanol, centrifuged (16,000× g), and filtered through a 0.22 μm membrane. This step was performed with three biological replicates.

Liquid chromatography–mass spectrometry (LC-MS) analysis was performed using an UltiMate 3000 ultra-high-performance liquid chromatography (UPLC) system coupled to a high-resolution quadrupole time-of-flight mass spectrometer (5600 QTOF, AB SCIEX, Shanghai, China). Chromatographic separation was achieved on an Agilent Eclipse Plus C18 column (2.1 × 100 mm, 1.8 μm) using a binary mobile phase system consisting of ultrapure water containing 0.1% (v/v) formic acid (A) and acetonitrile (B). The flow rate was set at 0.20 mL·min^−1^, and the injection volume was 2 μL. A gradient elution program was applied, with mobile phase B increased from 10% to 95%. Metabolite identification was performed using MSDIAL software (v4.24), and metabolites detected in fewer than 50% of biological replicates were excluded from subsequent analyses.

2.4.3. Bacterial Transcriptome Sequencing Under Iron-Deficient Conditions

Bacterial transcriptome sequencing was performed using the PE150 sequencing method on the Illumina high-throughput sequencing platform by Guangdong Meige Gene Technology Co., Ltd. (Guangzhou, China). Total RNA was extracted from the microbial pellets collected in Section 2.4.2 using the Trizol method, following the manufacturer’s protocol. Quality control of the raw sequencing data was performed using fastp software (v0.23.0), removing adapter sequences and low-quality reads (Q ≤ 20). The ribosomal sequences were then removed, and the cleaned data were aligned with the reference genome using Bowtie2. To eliminate the effects of gene length and sequencing depth on gene expression calculation, FPKM (Fragments Per Kilobase of transcript per Million mapped reads) normalization was applied. Differential gene expression analysis was performed using DESeq2 (v1.30.0, Bioconductor release 3.14) or edgeR (v3.32.1, Bioconductor release 3.14) software, with significantly differentially expressed genes selected based on the criteria: p < 0.05 and |log2FC| ≥ 1. GO functional enrichment analysis was conducted using the clusterProfiler package.

2.4.4. Quantitative Real-Time PCR Validation

The qRT-PCR analysis was performed to validate transcriptomic results. Total RNA was extracted from qualified samples, assessed for quality, and reverse-transcribed into cDNA using a commercial kit. GAPDH (ICC24_RS15650) was used as the internal reference gene, and primers were designed using Beacon Designer 7 (Table S1). qPCR was conducted using SYBR Green chemistry under standard cycling conditions.

2.5. Effects of Metabolites on Plants and Rhizosphere Soil

2.5.1. Metabolite Preparation

Seedling cultivation and treatment were performed as described in Section 2.3.1. Strain LCDD6 was cultured in SA medium, and the supernatant was collected and filtered (0.22 μm) to obtain cell-free fermentation broth. The filtrate was divided into two portions: one stored at 4 °C, and the other (100 mL) placed into dialysis bags with a molecular weight cut-off (MWCO) of 1000 Da and dialyzed at 4 °C for 72 h, with sterile distilled water replaced every 12 h. The dialysis solution inside the bags was tested with Arnow reagent until no further color change was observed. The dialysate was concentrated to 50 mL and further dialyzed (MWCO 500–1000 Da) for 48 h to obtain the bacillibactin crude extract, which was stored at 4 °C.

2.5.2. Identification of Bacillibactin Crude Extract Components

Bacillibactin crude extract was analyzed using a Hola C18 column (2.1 × 100 mm, 1.9 μm) with 0.1% acetic acid aqueous solution (A) and 0.1% acetic acid acetonitrile solution (B) as the mobile phases under gradient elution. Metabolite identification was performed using MSDIAL software.

2.5.3. Preparation of IAA Solution

First, 1 mL of cell-free fermentation broth was mixed with an equal volume of Salkowski reagent, incubated in the dark for 30 min, and three biological replicates were set up. The OD_530_ value was measured using a UV spectrophotometer. IAA concentration was calculated using a standard curve (y = 107.61x − 0.1349, R^2^ = 0.9997). An IAA solution of the same concentration was prepared using analytical-grade IAA standard and stored at 4 °C in the dark.

2.5.4. Test Design and Determination and Analysis of Indicators for M. hupehensis and Soil

Through the above experiments, we obtained cell-free fermentation broth, bacillibactin crude extract, and IAA solution from the LCDD6 strain. To investigate the effects of these metabolites on M. hupehensis seedlings, four treatment groups were designed: SA medium as the control group (CK); the cell-free fermentation broth (F); a mixture of bacillibactin crude extract (S); and the IAA (I). Each treatment was diluted 50 times with sterile water before use. The above treatment solutions were used for root irrigation of M. hupehensis seedlings, with 200 mL applied to each pot, followed by regular cultivation under the same conditions. Three biological replicates were set up for each treatment.

For 60 days, the pots were gently inverted to remove the plants. Loose soil adhering to the roots was carefully shaken off, and rhizosphere soil samples were collected using a small brush. Subsequently, growth parameters were measured: plant height and root length were determined using a ruler, and stem diameter was measured with a vernier caliper. Fresh and dry weights were measured after cleaning and drying the plant samples. The chlorophyll content in the leaves was measured via acetone extraction. The soluble protein content in the leaves was determined using the Coomassie Brilliant Blue method [29], while the activities of superoxide dismutase (SOD), peroxidase (POD) and catalase (CAT) in the leaves were assessed using the Farhang–Abriz and Torabian method [30]. Total iron content in plants and soil was measured using atomic absorption spectrophotometry, and phosphorus and potassium contents were determined via molybdenum-antimony colorimetry and flame photometry [31,32]. Enzyme activities of sucrose hydrolysis, phosphatase, and peroxidase in rhizosphere soil were measured via colorimetry and titration [33]. Rhizosphere soil samples were collected and stored at −80 °C for further analysis.

2.5.5. Soil DNA Extraction and Sequencing

Total genomic DNA was extracted from rhizosphere soil samples obtained in Section 2.5.4 using the E.Z.N.A. Soil DNA Kit (Omega Bio-Tek, Norcross, GA, USA). For bacterial communities, the V3-V4 region of the 16S rRNA gene was amplified using specific primers B341F (5′-CCTACGGGNGGCWGCAG-3′) and B341R (5′-GACTACHVGGGGTATCTAATCC-3′) by PCR. For fungal communities, the ITS1 region of the ITS1 rDNA gene was amplified using specific primers F (5′-GGAAGTAAAAGTCGTAACAAGG-3′) and R (5′-GCTGCGTTCTTCATCGATGC-3′) by PCR. The PCR products were purified, and sequencing libraries were constructed. Paired-end sequencing was performed on the Illumina platform, generating paired-end short reads.

2.6. Data Analysis

Statistical analyses were conducted using SPSS 19.0. Differences among treatments were evaluated via one-way ANOVA followed by Tukey’s post hoc test (p < 0.05). PCA and APCS analyses were performed in R (v4.5.2). Microbial α-diversity (Shannon index) and β-diversity (NMDS based on Bray–Curtis distance) were analyzed using QIIME2 (2019.4), and community differences were assessed via ANOSIM. Taxonomic composition was visualized using MEGAN (v2.1) and GraPhlAn (v1.1.3). Co-occurrence networks were constructed based on Spearman correlations (R > 0.7, p < 0.01) using the “microeco” package and visualized with Gephi (v0.10.1).

3. Results

3.1. Strain LCDD6 Promotes Plant Iron Uptake and Growth in Calcareous Soil

The addition of the fermentation broth of strain LCDD6 significantly promoted the growth of M. hupehensis seedlings in calcareous soil (Figure 1A,B). Compared to the control group, strain LCDD6 treatment increased plant dry weight by 67.7%, enhanced total iron content in the plant by 4.62 mg/kg, and reduced soil available iron content by 20.8% (Figure 1C–E; Table S2; p < 0.05). These results indicate that under iron-deficient conditions, strain LCDD6 plays a beneficial role in promoting M. hupehensis growth and iron acquisition. To investigate the metabolic basis underlying the growth-promoting effects of strain LCDD6 under iron-deficient conditions, we first identified iron-responsive metabolites and subsequently evaluated their effects on plant growth.

3.2. Bacillibactin and IAA Are Major Differential Metabolites Under Iron Deficiency Condition

After 20 h of incubation in the SA medium, strain LCDD6 entered the stationary phase (Figure S1A), with no spore formation observed within 36 h. The content of the siderophore reached its peak at 28 h, then gradually decreased. Meanwhile, the content of the plant hormone IAA continuously increased over the 36 h period (Figure S1B). Cells and their supernatant were collected after 28 h of cultivation for transcriptomic and metabolomic analysis.

3.2.1. Comparative Transcriptomic Analyses of Strain LCDD6 Under Iron Deficiency

A total of 2641 differentially expressed genes (p < 0.05, |log2FC| ≥ 1) between the non-limiting iron and limiting iron conditions were identified. Among these, 1420 genes were significantly upregulated, 1221 genes were significantly downregulated (Figure S2A). GO functional annotation of the differentially expressed genes further revealed that strain LCDD6 exhibited elevated transcription levels in multiple biological processes, including biofilm formation, reproductive processes, signal transduction, carbohydrate utilization, carbon source utilization, and molecular sensor activity, with no downregulated genes observed in these processes (Figure S2B).

Compared with CK, genes involved in bacillibactin biosynthesis exhibited a remarkable overall upregulation under iron-deficiency conditions (Figure 2A,B). Key genes responsible for converting chorismate into 2,3-dihydroxybenzoate such as ICC24_RS16485, ICC24_RS19755, ICC24_RS19745, ICC24_RS19750, and ICC24_RS19765 showed sharply elevated read counts in the SA treatment. The dhbC gene, encoding an essential enzyme in bacillibactin assembly, also displayed a substantial increase. These results indicate that strain LCDD6 strongly activates the bacillibactin biosynthetic pathway in response to iron limitation.

Additionally, genes associated with IAA biosynthesis also demonstrated consistently higher expression under SA conditions (Figure 2C,D). Multiple genes in the tryptophan-dependent pathways, including trpB, trpC, ICC24_RS12345, and ICC24_RS12355 were significantly upregulated, enhancing the flux toward L-tryptophan synthesis. Downstream IAA-production modules, involving indolepyruvate, indole-3-acetaldehyde pathways, also showed increased transcription. Further qRT-PCR analysis confirmed the significant upregulation of the IAA biosynthesis genes trpC, trpB, the bacillibactin biosynthesis genes ICC24_RS06100, ICC24_RS19760, and the transport gene ICC24_RS12950 under SA treatment. The qRT-PCR results were highly consistent with the RNA-seq data, thereby validating the reliability of the transcriptomic analysis (Figure S2C).

3.2.2. Metabolomic Analysis of Culture Supernatants

Metabolomic analysis of the culture supernatants revealed the presence of several bioactive compounds. One prominent compound with a molecular weight (MW) of 883 was identified as bacillibactin, a catecholate-type iron chelator, as indicated by the [M+H]^+^ ion peak (Figure 3A,B). Additionally, a compound with an MW of 175 was detected and identified as IAA (Figure 3D,E). Further analysis also identified other metabolites (Table S5), including lactic acid, myo-inositol, and vitamin B12, which have been shown to play important roles in plant growth and stress resistance. Quantitative analysis further revealed that bacillibactin production was dramatically enhanced under the SA condition, showing a 970% increase relative to the CK treatment (Figure 3C), accompanied by an 84.7% elevation in IAA levels (Figure 3F).

Crude bacillibactin extract was obtained by dialyzing the culture supernatant of strain LCDD6 through a membrane with a molecular weight cutoff of 500–1000 Da. Metabolomic analysis showed that, in addition to bacillibactin, the resulting extract contained several co-enriched metabolites, including smenospongine, aconitic acid, and icos-19-ene-1,2,4-triol (Table S3), some of which are known to exhibit antimicrobial properties [34,35,36].

3.3. Effects of the Metabolites on Promoting M. hupehensis Seedlings Growth

Compared with the CK group, the S group, I group and F group all showed significant increases in plant height, aboveground fresh weight, underground fresh weight, aboveground dry weight, and underground dry weight, with increases ranging from 47.7–125%, 42.1–205%, 98.6–217%, 54.9–266%, and 65.2–80.0%, respectively. Among these treatments, the S group exhibited the strongest growth-promoting effect, with these parameters increasing by 125%, 205%, 217%, 266%, and 80.0%, significantly outperforming other treatments (Figure 4A,B and Table S4). Meanwhile, the I group exhibited the highest belowground dry weight, representing a 80.0% increase compared to the CK treatment. Moreover, all the other treatments significantly enhanced the measured physiological properties in plants. Specifically, chlorophyll content increased by 27.8–58.9%, soluble protein content by 46.5–64.7%, SOD activity by 23.6–38.3%, POD activity by 50.0–75.0%, and CAT activity by 48.9–115% compared to the CK group.

In addition, the treatments significantly improved rhizosphere soil properties, including pH and EC, as well as the activities of soil phosphatase, peroxidase, and sucrase (Figure 4B and Table S4). The F group exhibited the strongest soil enzyme activities. Compared with the control, soil sucrase, phosphatase, and peroxidase activities increased by 57.0%, 25.9%, and 14.6%, respectively. Notably, the S treatment exhibited the most pronounced effect on enhancing iron and potassium uptake (Figure 4B). In particular, plant iron content increased by 26.6% compared with the CK treatment (Figure 4C), accompanied by a 17.6% decrease in soil available iron content (Figure 4D). The I group had the most prominent effect on phosphorus absorption, with the phosphorus content in the plants increasing by 27.4% compared to CK treatment (Figure 4B and Table S4).

PCA analysis of all measured parameters showed a clear separation between the CK group and the other treatments, with the CK group positioned in the negative region (Figure 4E). The S group demonstrated strong growth-promoting ability in plants, primarily reflected in plant biomass and iron content. The I group also showed significant improvement, particularly in plant total phosphorus and root development. The F group exhibited a more balanced enhancement in both soil enzyme activities and plant physiological properties, such as leaf chlorophyll and soluble protein contents. APCS analysis further supported these results (Figure S3), showing that the F and S groups achieved very similar and the highest overall scores. These findings suggested that bacillibactin is the main bioactive compound responsible for growth promotion in the cell-free fermentation broth of strain LCDD6 under iron-deficient conditions, with IAA playing a supplementary role.

3.4. Effects of the Metabolites on the Rhizosphere Microbial Community

After 60 days of cultivation, the addition of different metabolites significantly altered the rhizosphere microbiome. Sequencing yielded 1,630,913 bacterial and 1,464,930 ITS reads, clustered into 55,586 bacterial and 2700 fungal OTUs (Figure 5A,B). While bacterial α-diversity remained unchanged (Figure 5C), fungal α-diversity increased significantly in the S group (Figure 5D). NMDS analysis further revealed that rhizosphere bacterial community composition in the F and S groups was clearly separated from the CK group (Figure 5E). In contrast, fungal communities exhibited pronounced separation among all four treatments (Figure 5F), indicating a stronger and more treatment-specific response of fungi to metabolite addition.

At the phylum level, the bacterial communities were dominated by Proteobacteria, Actinobacteria, Acidobacteria, Chloroflexi, and Bacteroidetes across all treatments, with no significant differences in their relative abundances (Figure 5G). At the genus level, several dominant taxa, including Pseudomonas, Subgroup_6, BIrii41, A4b, and SBR1031, were shared among treatments. Notably, Pseudomonas exhibited a metabolite-specific response, reaching its highest relative abundance in the F treatment (10.6%), compared with 4.80%, 4.31%, and 5.05% in the CK, S, and I treatments, respectively (Figure 5H).

Fungal communities displayed more pronounced and differential responses to metabolite addition. Ascomycota, Mortierellomycota, and Basidiomycota dominated all treatments, collectively accounting for 83.6–89.7% of the total fungal community. However, Basidiomycota was markedly enriched in the F (9.78%) and S (7.28%) treatments compared with the CK (1.53%) and I (2.24%) treatments (Figure 5I). At the genus level, clear treatment-specific patterns were observed: Botryotrichum was enriched in the F and I treatments but reduced in the S treatment, whereas Pseudogymnoascus increased exclusively in the S treatment. In addition, Arthrobotrys increased in both the S and I treatments, accompanied by a concomitant decline in Iodophanus. Coprinellus was preferentially enriched in the F and S treatments, while Stachybotrys showed significant enrichment in the F and I treatments (Figure 5J). Overall, these results demonstrate that different metabolites exert distinct and selective effects on the rhizosphere microbial community, with fungi showing stronger and more treatment-specific responses than bacteria.

Co-occurrence network analysis of the rhizosphere microbial communities showed that network size varied little among treatments, as indicated by comparable numbers of nodes (191–219). In contrast, the number of links differed markedly among treatments, ranging from 5528 in the F treatment to 7153 in the S treatment (Figure 6). Compared with the CK network (1039 fungal–fungal, 3294 bacterial–fungal, and 2552 bacterial–bacterial interaction edges), distinct metabolite-specific effects on interaction patterns among microbial groups were observed. In the S group, fungal–fungal interactions increased to 1367, and bacteria–fungi interactions increased to 3576, whereas bacteria–bacteria interactions decreased to 2210. In contrast, in the I group, fungal–fungal and bacteria–fungi interactions markedly declined to 787 and 2880, respectively, while bacteria–bacteria interactions remained comparable to those in the CK group. The F treatment exhibited a different pattern, characterized by pronounced reductions in bacteria–bacteria (1902) and bacteria–fungi interactions (2654), whereas fungal–fungal interactions remained comparable to those in the CK group. These results demonstrate that different metabolites reshaped rhizosphere microbial interaction networks in distinct ways: Bacillibactin enhanced fungal-associated and cross-kingdom interactions, IAA suppressed fungal and bacteria–fungi interactions, and the F treatment primarily weakened bacterial-associated interactions, potentially due to the presence of antibiotic-like compounds in the strain LCDD6 cell-free fermentation broth (Table S5).

3.5. Microbe–Plant Feedback in Response to the Metabolites

Correlation heatmap analysis was performed to examine the relationships between the top 10 most abundant bacterial and fungal genera and plant agronomic traits, plant physiological properties, plant nutrient contents, soil enzyme activities, and soil characteristics, based on which four microbial groups (I–IV) were identified (Figure 7A). Most microorganisms in Group I showed no significant correlations with plant growth parameters and soil properties. Group II included four fungal taxa, Coprinellus, Pseudogymnoascus, Arthrobotrys, and Chrysosporium. Among them, Coprinellus was enriched in the S and F groups, whereas the remaining taxa were predominantly enriched in the S group compared with the other treatments. Interestingly, Coprinellus was positively correlated with plant agronomic traits (aboveground dry weight, aboveground and underground fresh weight, and height), plant nutrient content (total iron, total potassium), and soil sucrase activity, while being negatively correlated with soil characteristics (available iron, phosphorus, and pH). Arthrobotrys and Chrysosporium showed a significant positive correlation with the underground dry weight of plants (Figure 7B).

In Group III, Stachybotrys, which was enriched in the I treatment, showed a significant positive correlation with plant total phosphorus content. Group IV consists of two bacterial genera, SBR1031 and BIrii41, and two fungal genera, Mortierella and Iodophanus. These taxa showed positive correlations with soil characteristics (except soil total iron), but negative correlations with plant nutrient content, plant agronomic traits, physiological properties, and soil enzyme activities.

4. Discussion

Calcareous soils, due to their high pH and poor physical structure, often lack an effective supply of essential elements such as nitrogen, phosphorus, iron, zinc, manganese, and copper, which severely affects the growth and yield of fruit trees. Under alkaline and nutrient-deficient stress conditions, certain microorganisms, such as B. licheniformis, can secrete various growth-promoting substances to enhance plant stress adaptation [37,38]. The production of these metabolites not only supports the growth of the microorganisms themselves but also promotes plant growth and health. This study analyzes the mechanism by which the metabolic products of B. licheniformis LCDD6 enhance the growth of M. hupehensis seedlings in calcareous soils. Under iron-deficient conditions, B. licheniformis LCDD6 is able to secrete various bioactive substances, including bacillibactin, IAA, lactic acid, chitobiose, myo-inositol, vitamin B12, scopoletin, etc. These substances play important roles in plant growth, stress resistance, and nutrient absorption. Many of these organic acids, such as lactic acid, are known to have metal-chelating properties, and their interactions with metals, including iron and other essential micronutrients, may further contribute to plant growth promotion.

Previous studies have shown that some metabolites of plant growth-promoting microorganisms can improve plant growth. For example, organic acids like lactic acid can help plants adapt to stress [39]; scopoletin, a plant growth factor, can inhibit acetylcholinesterase [40]; siderophores can effectively alleviate iron deficiency and enhance plant absorption of iron and phosphorus [14,41]; IAA, a common plant hormone, promotes root growth and improves water and nutrient absorption by plants [42,43]. Additionally, vitamins like vitamin B12 can promote plant growth [44]. This is consistent with our findings. We found that the metabolic products bacillibactin and IAA from strain LCDD6 exhibit complementary effects. Bacillibactin enhances the iron and potassium content in plants, promotes the growth of above-ground parts, and also increases the phosphorus content in plants, although its effect is not as strong as that of IAA. The secretion of IAA enhances phosphorus absorption and promotes root growth. Their synergistic effect significantly enhances the growth of M. hupehensis seedlings, especially under iron-deficient and alkaline soil stress conditions.

Transcriptome analysis indicated that under iron-deficient conditions, strain LCDD6 significantly upregulated key genes involved in the biosynthesis of bacillibactin and IAA. Metabolomics analysis further confirmed this finding, showing that the production of bacillibactin increased significantly by 970%, and the level of IAA increased by 84.7%. Through these metabolic regulation mechanisms, the microorganism not only met its own nutritional needs but also enhanced the bioavailability of nutrients in the soil, thereby promoting plant nutrient uptake. Overall, strain LCDD6 prioritized the biosynthesis of the iron chelator bacillibactin (increased by 970%) through the secretion of metabolic products to cope with iron deficiency and exert a pronounced plant growth-promoting effect.

In addition to directly chelating iron, bacillibactin also significantly altered the rhizosphere microbial community, particularly the fungal composition [45,46]. Bacillibactin treatment significantly enriched several beneficial fungal genera, including Coprinellus, Chrysosporium, Arthrobotrys, and Pseudogymnoascus. Among these, Coprinellus and Pseudogymnoascus are typical saprophytic fungi capable of efficiently decomposing lignocellulosic residues and plant litter in soil, thereby promoting nutrient mineralization and release and providing sustained nutritional support for plant growth [47,48]. In addition, Coprinellus, Chrysosporium, and Arthrobotrys exhibit notable biocontrol potential: the former two can suppress soil-borne pathogens through the production of bioactive compounds, whereas Arthrobotrys is well known for its ability to prey on or inhibit plant-parasitic nematodes [47,49,50]. Concurrently, bacillibactin treatment significantly reduced the relative abundance of fungi such as SBR1031, Mortierella, BIrii41, and Iodophanus. At present, no definitive taxonomic or functional information is available for SBR1031 and BIrii41, suggesting that they may represent unclassified OTUs or sequence identifiers generated through high-throughput sequencing analyses. Previous studies have reported that Mortierella is capable of synthesizing plant hormones such as IAA [51]; however, in the present study, both metabolite treatment and exogenous IAA application markedly decreased its relative abundance, indicating that elevated soil IAA levels may inhibit the growth of Mortierella. Notably, bacillibactin treatment also reduced the abundance of Mortierella, although to a lesser extent than metabolite and IAA treatments, implying that bacillibactin may influence this genus indirectly, and the underlying mechanisms warrant further investigation. With respect to Iodophanus, available studies are limited.

In addition to the direct effects of the metabolites, our study also emphasizes the feedback mechanism between microorganisms and plants. We found that fungi such as Coprinellus and Pseudogymnoascus showed significant positive correlations with plant growth traits, while Stachybotrys was positively correlated with plant phosphorus content. This finding further supports the importance of plant–microbe interactions in promoting plant growth. We speculate that bacillibactin and IAA, by altering the structure and interaction network of the microbial community, not only improve the nutrient supply in the soil but also enhance the plant’s adaptation to stress by activating metabolic activities in the root system. Compared to previous studies, this research provides a more detailed understanding of the dynamic interactions between microbial communities and plant growth.

Lastly, this study was conducted under controlled greenhouse conditions, and field trials are needed to assess the performance of strain LCDD6 in natural calcareous soils. In field environments, factors such as temperature, precipitation, and local microbial communities may influence the effectiveness of strain LCDD6 and its metabolites.

5. Conclusions

This study demonstrated that B. licheniformis LCDD6 effectively promotes the growth and improves the iron nutrition of M. hupehensis seedlings in calcareous soils. Integrated transcriptomic and metabolomic analyses identified bacillibactin and IAA as the dominant metabolites induced under iron deficiency, and their functions are complementary: bacillibactin enhances plant absorption of iron and potassium to promote above-ground growth, while IAA improves phosphorus absorption and facilitates root development, synergistically enhancing seedling stress resistance and growth. Additionally, different metabolites exerted distinct and selective effects on the rhizosphere microbial community, with fungi showing stronger and more metabolite-specific responses than bacteria. These findings reveal the plant growth-promoting mechanism mediated by the complementary functions of metabolites, providing a theoretical basis for the development of Bacillus-based biofertilizers to improve iron availability in calcareous soils.