Introduction

Pinus massoniana (P. massoniana) is a major pioneer tree species used for afforestation in Southern China (Quan and Ding 2017). Damping-off disease caused by Fusarium oxysporum can induce extensive death of Masson pine seedlings in a short time frame (Luo and Yu 2020), restricting the cultivation, afforestation, and resource utilization of P. massoniana. Chemical pesticides are widely used to control infectious pathogens because they are convenient, quick-acting, and effective (Sharma et al. 2020). However, excessive dependence on their use over a long period can worsen the physical and chemical properties of soil, destroy the soil microbial community, and increase the drug resistance of pathogens (Mehmood et al. 2021). Due to their safety, durability, and ecological nature, biological control agents (BCAs) are a promising alternative to chemical pesticides to suppress infectious diseases in Pine trees (Yu and Luo 2020). There is an urgent need for BCAs in addition to the commonly used agents, Bacillus subtilis, Beauveria bassiana, Trichoderma harzianum, and Paecilomyces lilacinus (Weerapol et al. 2019; Villa-Rodriguez et al. 2022), to be identified.

Plant growth-promoting fungi (PGPFs), such as Trichoderma species, are widely used in agriculture to control plant disease (Jogaiah et al. 2013). Trichoderma spp. use several mechanisms to effectively inhibit various pathogens in Pine trees. T. atroviride improves the systemic tolerance of Diplodia pinea-infected P. radiata and reduces 20% of the seedling dieback (Reglinski et al. 2012). Trichoderma spp. from P. sylvestris bark can produce volatile compounds with antimicrobial activity and have typical mycoparasitic manifestations against Botryosphaeriaceae (Karlicic et al. 2021). Trichoderma spp. decreases morbidity associated with Fusarium circinatum damping-off in P. radiata seedlings (Morales-Rodríguez et al. 2018). Recently, T. longibrachiatum was shown to play a vital role in helping crops resist diseases. This fungus accumulates key metabolites, which help onion plants (Allium cepa L.) to resist F. oxysporum (Abdelrahman et al. 2016). T. longibrachiatum TG1 controls wheat (Triticum aestivum L.) crown rot disease by activating the plant defense system and increasing the transcription of pathogenesis-related genes (Boamah et al. 2021). However, there are few reports on the effect of PGPFs on Fusarium damping-off disease in Pinus.

Several plant growth-promoting rhizobacteria (PGPR) are effective BCAs (Wang et al. 2021b), promoting plant growth by secreting indole acetic acid and dissolving minerals, and also protecting plants from infection by inducing systemic resistance and the production of antagonistic substances (Takishita et al. 2018). The genera Bacillus and Pseudomonas are predominant PGPRs that protect various crops against pathogens, such as V. dalhiae and F. oxysporum (Beneduzi et al. 2012; Essalimi et al. 2022). More recently, Burkholderia spp. (B. spp.) have also been identified as PGPRs, showing great potential against various soil-borne pathogens (Esmaeel et al. 2020). While B. contaminans, B. cepacia and B. stabilis are shown to produce various antimicrobial metabolites against wilt and root rot diseases (Jung et al. 2018; Kim et al. 2020; Heo et al. 2022), very little research has assessed the role of Burkholderia spp. in disease control in Pinus.

Recent studies have indicated that microbial combinations can more effectively inhibit plant pathogens than a single inoculate. Elshahawy and El-Mohamedy (2019) found that the combined use of five Trichoderma isolates was the most effective method of suppressing damping-off and root rot by activating defense enzymes. Few studies have assessed the combined effects of PGPF and PGPR in controlling fungal disease. Many BCAs can induce systemic acquired resistance (ISR) to combat plant pathogens (Heo et al. 2022). Plant rhizosphere microbial communities are the first line of defense for protecting plants against different biological and abiotic stresses (Dini-Andreote 2020). Thus, it is important to study the disease resistance mechanism of a PGPF and PGPR combination by analyzing its effect on the rhizosphere microbial community structure and plant physiological system.

PGPF and PGPR strains were recently isolated from P. massoniana to assess their ability to control Fusarium damping-off in seedlings. The current study sought to (1) evaluate the antifungal activity of PGPFs and PGPRs against F. oxysporum in vitro, (2) identify the growth promotion and biocontrol effects of a PGPF and PGPR combination in P. massoniana seedlings, and (3) analyze changes in the soil microbial community structure and physiological system after PGPF and PGPR co-inoculation.

Materials and methods

Fungal and bacterial strains

The fungal and bacterial strains used in this study were isolated from a 20-year-old healthy Masson pine in Guiyang City, Guizhou Province (26.44°N, 106.65°E), China, in June 2019. Three fungal strains, 3Y, 6Y, and 12Y were isolated from the root and determined to be IAA-producing strains, which identified them as PGPFs (Luo 2020). Four bacterial strains, K3, K15, K25, and K29 were isolated from rhizosphere soil using a method developed by Bagyalakshmi et al. (2017) and determined to be potassium solubilization and IAA-producing, which identified them as PGPRs. The F. oxysporum strain (GenBank accession no. MK356552), which causes damping-off disease in seedlings, was isolated from the roots of P. massoniana (Luo and Yu 2020).

Determination of antagonistic activity in vitro

The dual culture technique was used to measure the antagonistic effect of different PGPF and PGPR strains against the pathogenic F. oxysporum strain (Bell et al. 1982). A 5 mm diameter mycelial disc from an actively growing 7-day F. oxysporum culture was placed on the center of each PDA plate, and a 108 CFU/mL bacterial suspension from each PGPR strain (K3, K15, K25, and K29) was used to draw a line 10 mm from the disc between the left and right sides. Plates inoculated with F. oxysporum alone were used as a control. The plates were incubated in the dark at 28 °C for 5 days. The 5 mm diameter mycelial disc of F. oxysporum and each PGPF strain (3Y, 6Y, and 12Y) from the 7-day-old cultures were placed across from each other (about 60 mm apart) on a 90 mm diameter PDA plate. A PDA plate inoculated with an F. oxysporum disc served as the control. The plates were incubated in the dark at 28℃ for 7 days.

After dual culture, the inhibition rate was calculated using the following formula: inhibition rate (%) = ([R1 − R2] / R1) × 100 (Díaz-Gutierrez et al. 2021), where R1 is the radial growth of the pathogen on the control plate, and R2 is the radial growth of the pathogen on the dual culture plate. The hyphae at the contacting edge of the pathogen were collected to observe interactions using a light microscope (Olympus CX21, Tokyo, Japan). The experiment was performed with six replicates.

To measure extracellular enzyme activity (protease, cellulase, and chitinase), 2 µL suspension of the K25 and K29 strains (108 CFU/mL) were inoculated on detection medium containing 1.5% skim mile power (Heo et al. 2022), 1% carboxymethyl cellulose (Díaz-Gutierrez et al. 2021), and 1% colloidal chitin (Roberts and Selitrennikoff 1988). After culture at 30℃ for 5 days, the plates were assessed for the presence of a transparent halo. 3,5-Dinitrosalicylic acid (DNS) colorimetry and Folin-phenol methods was used to measure chitinase and protease activities.

Plant inoculation and treatment

The fungal strain 3Y (Trichoderma longibrachiatum, MT131278.1) and the bacterial strain K29 (Burkholderia stabilis, OR262944), identified based on their morphology and molecular sequences, were used for plant inoculation. The methods are described in detail in Method S1. A 1 × 108 CFU/mL bacterial suspension (of B. stabilis and 1 × 106 spores ml−1 spore suspensions of T. longibrachiatum and F. oxysporum were prepared. The phloem of fine roots from 3-month-old seedlings were wounded lightly with sterilized scalpels once and planted in a plastic pot (12.5 cm high, 14.5 cm upper diameter, 10.2 cm lower diameter) at three seedlings per pot. The suspension was then poured into the soil near the root using a sterilized syringe (Sofo et al. 2010). The following four treatment groups were used: (1) Fo, soil inoculated with a 5 mL F. oxysporum suspension (control), (2) Fo + 3Y, soil inoculated with a mixture of a 5 mL F. oxysporum and a 5 mL T. longibrachiatum suspension, (3) Fo + K29, soil inoculated with a 5 mL F. oxysporum and a 5 mL B. stabilis suspension, (4) Fo + 3Y + K29, soil inoculated with a 5 mL F. oxysporum, 5 mL T. longibrachiatum, and 5 mL B. stabilis suspension. Ten pots were used as a biological repeat, and each treatment included three biological repeats. After treatment, seedlings were grown at 25 ± 0.5℃ for 60 days, with 16/8 h day/night lighting and a relative humidity of 70%.

Disease and growth index measurements

After inoculation for 60 d, disease incidence (DI), disease severity (DS), and control index (CI) were evaluated. DI was assessed using the following formula: DI (%) = (I / R) × 100 (Toghueo et al. 2016, where I is the number of infected plants, and R is the total number of plants receiving treatment. DS was determined based on the severity of symptoms, which was divided into five incidence levels (0, 1, 2, 3, and 4) (Díaz-Gutierrez et al. 2021), and evaluated using the following formula: DS (%) = (∑(A × B) /(T×M)× 100, where A is the number of diseased plants at each level, B is the corresponding incidence level, T is the total number of plants, and M is the maximum incidence level. CI was calculated using the following formula: CI (%) = ((D1-D2) / D1) × 100, where D1 is the DS value in the control (Fo treatment) and D2 is the DS value in the corresponding treatment group. After the disease index assessment, seeding height, taproot length, and the number of lateral roots were determined. Meanwhile, above-ground shoots and roots were separately collected to assess their fresh and dry weight.

Assay of physiological indices

Physiological parameters were measured after 60 inoculations. The chlorophyll content was determined using the 80% acetone extraction method (Wood et al. 2020). The relative conductivity was determined using a conductometer (DDS-308+, INESA, Shanghai, China) as described by Fan et al. (1997). Malondialdehyde (MDA), soluble sugar, and proline content were detected using the thiobarbituric acid, anthrone-sulfuric, and acidic-ninhydrin methods, respectively (Draper et al. 1993). Peroxidase (POD), superoxide dismutase (SOD), and catalase (CAT) activities were determined using the nitroblue tetrazolium photoreduction, guaiacol colorimetry, and potassium permanganate titration methods, respectively (Lin et al. 2020).

Soil genomic DNA extraction, sequencing, and data processing

After 60 days of inoculation, the rhizosphere soil samples (Fo, Fo + 3Y, Fo + K29, Fo + 3Y + K29) were collected, homogenized, and sampled for DNA extraction. Ten pots were pooled as a replicate. Three biological replicates were used for each treatment. The 16 S rDNA gene V3-V4 region of the soil bacteria and the fungus ITS1-ITS2 region were amplified using the bacteria-specific primers, 341 F (5’- CTACGGGNGGWGCAG-3’) and 805R (5’- GACTACHVGGATCTAATCC-3’), and the fungal primers, ITS1F (5’- CTTGGTCATTTAGGAGATAAA-3’) and ITS2R (5’- CTGCGTTCTTCGCGATGC-3’), respectively. Purified PCR products were sequenced using the Illumina MiSeq sequencing system (Sangon Biotech, Shanghai, China). For raw sequencing data, operational taxonomic units (OTUs) were assigned based on a 97% sequence similarity threshold, and the representative OTU classification was obtained using the QIIME (v1.8.0) and R package (v3.2.0) (Edgar 2010). The relative abundance of each phylum or genus was calculated using the following formula: relative abundance (%) = ni/N, where ni is the number of sequences for each OTU, i represents an individual OTU, and N is the total number of sequences for all OTUs in the sample. Alpha diversity indices (Simpson, Shannon, ACE, and Chao1) were used to access the diversity and richness of the bacterial and fungal communities, and the permutational MANOVA (ADONIS) analysis and principal component analysis (PCA) was used to evaluate differences in the community structures associated with each treatment (Tuomisto 2010).

RNA extraction, cDNA library construction, and transcriptomics analysis

After 60 days of incubation, total RNA was extracted using whole Masson Pine seedlings from the Fo and Fo + 3Y + K29 treatment groups. Six seedlings was pooled as a replicate. Three biological replicates were used for each treatment. Six high-quality cDNA libraries were created as previously described by Liu et al. (2020) and sequenced using the Illumina HiSeq 2500 system (Sangon Biotech, Shanghai, China). Raw read data were stored in the NCBI database with accession number PRJNA997741. Clean reads were identified by removing the adapter, ambiguous, and low-quality reads from the raw data. Non-redundant unigenes were obtained using the Trinity method (Grabherr et al. 2011). The transcription level of each unigene was calculated and normalized using the fragments per kilobase of transcript per million fragments mapped reads (FPKM) method (Trapnell et al. 2010). The differentially expressed genes (DEGs) in the two different treatment groups were assigned with a p-value < 0.05 and |log2 fold change (FC)| >1 using the DEGseq package (Love et al. 2014). Significant enriched KEGG pathway was determined with a p adjusted using the Bonferroni correction (q values) ≤ 0.05.

Statistical analysis

Data were analyzed with SPSS software (IBM 21.0, New York, NY, USA) for one-way analysis of variance (ANOVA) using the Duncan test (P ≤ 0.05) to determine significant differences.

Results

Antagonistic activity of the PGPF and PGPR strains against the pathogen, F. oxysporum

Four PGPR strains, K3, K15, K25, and K29, were used for the antagonistic activity measurements. All four strains prevented the growth of F. oxysporum, with inhibition rates ranging from 35.00 to 55.67%. Of these, K25 and K29 had the strongest inhibition rates, reaching 50.33% and 55.67% respectively (Fig. 1A, B). In the dual culture conditions, K25 and K29 had significant inhibitory effects on the growth of F. oxysporum mycelium, causing breakage, bending, shrinking, and deformity (Fig. 1C). K29 strain solubilized skim milk and colloidal chitin by forming a halo zone on the agar, indicating that this strain had chitinase activity and was able to secrete proteases, while the K25 strain had protease activity (Fig. S1A). The proteases and chitinase activities of K29 strain were significantly higher than that of K25 strain (Fig. S1B). The growth of the three PGPF strains, 3Y, 6Y, and 12Y, against F. oxysporum are shown in Fig. 2. The 3Y strain strongly prevented F. oxysporum growth (Fig. 2A), with an inhibition rate of 75% (Fig. 2B). Microscopic examination showed that the mycelium of F. oxysporum became curled and deformed after co-cultivation with the 3Y strain (Fig. 2C). Based on these results, the K29 and 3Y strains were chosen for subsequent experiments.

Fig. 1
figure 1

Antagonistic activity of PGPR strains against F. oxysporum. A Dual cultures inoculated with F. oxysporum as a control (Fo), K3 strain against F.oxysporum (Fo + K3), K15 strain against F. oxysporum (Fo + K15), K25 strain against F.oxysporum (Fo + K25), K29 strain against F.oxysporum (Fo + K29). B Inhibition rate of the four PGPR strains against F. oxysporum in dual culture, different letters on the bars indicate significant differences at P ≤ 0.05 according to Duncan’s multiple range test. C Observation of F. oxysporum mycelium in dual culture after 5 days, Scale bars = 20 μm

Full size image
Fig. 2
figure 2

Antagonistic activity of PGPF strains against F. oxysporum. A Dual cultures inoculated with F. oxysporum as a control (i, Fo), 3Y strain against F. oxysporum (ii, Fo + 3Y), 6Y strain against F. oxysporum (iii, Fo + 6Y), 12Y strain against F. oxysporum (iv, Fo + 12Y). B Inhibition rate of the PGPF strains against F. oxysporum in dual culture. The different lower-case letters indicate significant difference at P ≤ 0.05 according to Duncan’s multiple range test. C Observation of mycelial interactions between 3Y strain and F. oxysporum after 7 days in dual culture, scale bars = 20 μm

Full size image

Characterization of the K29 and 3Y strains

The K29 strain was identified based on its cell and colony morphology, biochemical and physiological test results, and 16s rRNA gene sequence. Under a scanning electron microscope, cells of the K29 strain were short and rod-shaped, approximately 0.7–0.9 μm in length and 0.2–0.3 μm in width (Fig. 3A). The colonies had a milky white color with a moist surface, protrusions, no wrinkles, and fast growth (Table S1). The K29 strain was Gram-negative and oxidized bacteria. Amylolysis, contact enzyme, and gelatin hydrolysis test results were positive, the methyl red test (MR test) and Voges-Proskauer test (V-P test) results were negative, and the strain was able to grow at 25–35℃ with tolerance to 2% NaCl (Table S1). Based on its 16 S rDNA gene sequence, the K29 strain (OR262944) displayed the greatest sequence homology with Burkholderia stabilis (MG571686.1) (Fig. 3B). The upper side of the 3Y colony was yellowish green and the lower side was yellow, the colony grew rapidly and had an obvious whorl, the conidiophore had simple branches and a strong main shaft, and the conidia were oval, smooth, green, and 3.0–4.5 × 1.5–2.3 μm in size (Fig. 3C). Based on the ITS rDNA gene sequence, the 3Y strain (MT131278.1) had the highest sequence homology with T. longibrachiatum (MT102396.1) and formed a distinct clade from the other Trichoderma spp (Fig. 3D).

Fig. 3
figure 3

Morphological and molecular identification of K29 strain as Burkholderia stabilis and 3Y strain as Trichoderma longibrachiatum. A Morphology observation with a scanning electron microscope. B Phylogenetic tree based on 16 S rDNA sequence. These 16 S rDNA sequences of related strains were downloaded from NCBI GenBank database. The tree was structured using neighbor joining (NJ) method, with the bootstrap analyses of 1000 cycles. Bar represents sequence divergence of 0.02 nucleotides. C Morphology observation, colonial morphology on the upper side (i) and lower side (ii) on PDA medium; microscopic observation of conidiophore (iii) and conidia (iv). D Phylogenetic tree based on ITS sequence

Full size image

Biocontrol activity of combined T. longibrachiatum and B. stabilis treatment against Fusarium damping-off disease in P. massoniana

Fusarium oxysporum inoculation caused needle withering and yellowing and the symptoms were alleviated by K29 or 3Y treatment (Fig. 4A). Disease incidence and severity were significantly lower in response to Fo + K29, Fo + 3Y, and Fo + K29 + 3Y than to Fo treatment alone (P < 0.05, Fig. 4B, C), and K29 and 3Y together had the highest control index (Fig. 4D). The seedling height, number of lateral roots, and fresh and dry weight were significantly higher following Fo + K29 + 3Y than Fo treatment alone (P < 0.05, Fig. 4E-I). These results indicated that the combination of T. longibrachiatum and B. stabilis effectively reduced Fusarium infection and promoted seedling growth.

Fig. 4
figure 4

The effects of different treatment groups on disease incidence and seedlings growth. A The representative photographs, B Disease incidence, C Disease severity, D Control index, E Seedling height, F Taproot length, G Number of lateral roots, H Fresh weight, I Dry weight. Fo, plant inoculated with F. oxysporum; Fo + 3Y, plant inoculated with T. longibrachiatum and F. oxysporum; Fo + K29, plant inoculated with B. stabilis and F. oxysporum; Fo + 3Y + K29, plant inoculated with F. oxysporum, T. longibrachiatum and B. stabilis. Treatment with different strains was used as a variable, shoot or root was analyzed separately by one-way ANOVA. Bars indicate the mean ± SE for three independent biological replicates (n = 30). Different letters indicate significant differences at P ≤ 0.05

Full size image

Plant physiological characteristics under different treatment conditions

The seedlings in the Fo + K29 + 3Y treatment group had a higher chlorophyll content than those in the other treatment groups (Fig. 5A). The relative conductivity and MDA content were higher in plants receiving Fo treatment and lower in seedlings inoculated with the K29 or 3Y strain (Fig. 5B, C). The soluble sugar content of the shoots and proline content of the shoots and roots were higher in the Fo + K29 + 3Y treatment group than in the other treatment groups (Fig. 5D, E). Meanwhile, the activity of antioxidant enzymes, including SOD, POD, and CAT, was significantly higher in the Fo + K29 + 3Y treatment group (Fig. 5F-H).

Fig. 5
figure 5

The effects of different treatment groups on various physiological indicators. A Chlorophyll content, B Relative conductivity, C MDA content, D Soluble sugar content, E Proline content, F SOD activity, G POD activity, H CAT activity. Fo, plant inoculated with F. oxysporum; Fo + 3Y, plant inoculated with T. longibrachiatum and F. oxysporum; Fo + K29, plant inoculated with B. stabilis and F. oxysporum; Fo + 3Y + K29, plant inoculated with F. oxysporum, T. longibrachiatum and B. stabilis. Treatment with different strains was used as a variable, shoot or root was analyzed separately by one-way ANOVA. Different letters indicate significant differences at P ≤ 0.05

Full size image

Analysis of rhizosphere community structure

The permutational MANOVA (ADONIS) analysis and PCA analysis indicated that there was no significant difference between Fo and Fo + 3Y + K29 treatment for fungal community (P > 0.05, Fig. 6A, Table S2), but there was a marked difference in the bacterial community structure of the different treatment groups (P = 0.05, Fig. 6B, Table S2). The Shannon index of the fungal community was significantly higher in the Fo + 3Y treatment group than in the Fo treatment group (P < 0.05), but there was no significant difference in the total number of OTUs or in the ACE and Chao indexes of the different treatment groups (P > 0.05, Table S3). The number of OTUs and the Shannon, ACE, and Chao indexes of the bacterial community were significantly reduced in the Fo + K29, Fo + 3Y, and Fo + K29 + 3Y treatment groups than in the Fo treatment group, while the Simpson index was markedly increased (P < 0.05, Table S4). These results indicated that T. longibrachiatum inoculation improved fungal richness, while both T. longibrachiatum and B. stabilis inoculation reduced bacterial richness and diversity in F. oxysporum-infected seedlings.

Fig. 6
figure 6

Rhizosphere soil microbial community composition in different treatments. Principal component analysis (PCA) was used to study the effect of different treatments on the fungal (A) and bacterial (B) community structure. Relative abundances of the phyla in fungi (C) and bacteria (E). Heatmaps showing differences in the compositions of the top 30 genera of fungi (D) and bacteria (F), the heat map was generated from the relative abundance values, and the color change from blue to red suggests an increase in the relative abundance levels. Different letters in each row of the heatmap indicate significant differences at P ≤ 0.05 among Fo, Fo + K29, Fo + K29 + 3Y, and Fo + 3Y treatments according to Duncan’s multiple range test

Full size image

The rhizosphere soil microbial community composition is shown in Fig. 6. Ascomycota and Basidiomycota were the dominant fungi phyla, with relative abundances of 73–83% and 6–16%, respectively (Fig. 6C). At the top 30 fungal genera, the relative abundance of 19 genera had no significant difference among different treatments (Fig. 6D). Fo + 3Y treatment enhanced the relative abundance of some genera, including Trichoderma, Saitozyma, Penicillium, and Papiliotrema, K29 and 3Y together caused a marked improvement in the relative abundances of Trichoderma and unclassified Rozellomycota (P < 0.05, Fig. 6D). The dominant bacteria phyla included Proteobacteria, Acidobacteria, Bacteroidetes, and Actinobacteria; there was a significant increase in the relative abundance of Proteobacteria in response to Fo + K29, Fo + 3Y and Fo + K29 + 3Y compared to Fo treatment (P < 0.05, Fig. 6E). Most bacterial genera exhibited significant differences in the relative abundance, some key genera, including Buttiauxella, Sphingomonas, Pseudomonas, Comamonas, Cupriavidus as well as Flavobacterium, were markedly enriched in the Fo + K29 + 3Y treatment group than in the Fo treatment group (Fig. 6F).

Analysis of the KEGG pathway and key DEGs associated with growth and disease resistance

Compared to Fo treatment (CK) alone, there were a total of 8541 DEGs identified in the Fo + 3Y + K29 treatment group (W), of which 4549 were up-regulated and 3992 were down-regulated. These DEGs are enriched in 252 KEGG pathways, eight of which were significantly enriched (q value < 0.05), including ribosome, photosynthesis, photosynthesis-antenna proteins, starch and sucrose metabolism, pentose and glucuronate interconversions, ascorbate and aldarate metabolism, plant hormone signal transduction, and carbon fixation in photosynthetic organisms (Fig. 7A).

Fig. 7
figure 7

KEGG analysis and the expression profiles of DEGs involved in key biological pathways associated with plant growth and disease resistance. A KEGG enrichment analysis of the DEGs in W vs. CK treatment, the y-axis shows the pathway name, and the x-axis indicates the enrichment factor corresponding to the pathway; the color of the dot represents the q-value, and the size of the dot represents the number of DEGs. B DEGs involved in sugar metabolism, C DEGs involved in plant hormone signal transduction, D DEGs involved in photosynthesis pathway, E DEGs involved in antioxidant enzyme activity and proline metabolism; the bubble color was generated from the Log2 (Fold Change (FC) values)(FC is FPKM [W]/ FPKM [CK]), the color changing from red to blue represents an decrease in the expression levels, and red and blue represents significant up-regulated and down-regulated genes (p Value ≤ 0.05), respectively. The bubble size was generated from -log10 (pValue), the larger the bubble, indicating a higher significant level. CK, seedlings inoculated with F. oxysporum; W, seedlings inoculated with F. oxysporum, T. longibrachiatum and B. stabilis

Full size image

The expression profile of DEGs involved in important biological pathways is shown in Fig. 7B-E. Most DEGs involved in sugar metabolism, photosynthesis pathway, peroxidase, superoxide dismutase and proline metabolism were significantly up-regulated (p value < 0.05, Fig. 7B, D, E). A total of 30 DEGs were shown to participate in starch and sucrose metabolism, and those involved in sucrose, cellobiose, amylose, maltose, and trehalose synthesis were all up-regulated (Fig. 7B). A total of 33 DEGs mapped to the plant hormone signal transduction pathway and shown to be primarily involved in auxin, cytokinin, gibberellin, abscisic acid, ethylene, brassinosteroid, and jasmonic acid metabolism (Fig. 7C). A total of 24 DEGs participated in photosynthesis, 21 of which were up-regulated (Fig. 7D). There were 5 up-regulated DEGs involved in proline metabolism (Fig. 7E); a total of 14 DEGs involved in encoding antioxidant enzymes, including peroxidase (POD), superoxide dismutase (SOD) and catalase (CAT) (Fig. 7E), most DEGs encoding POD and SOD were up-regulated, while two DEGs encoding CAT were all down-regulated, indicating that those DEGs involved in POD and SOD enzyme activities may mainly contribute to regulating the ROS scavenging.

Discussion

The combination of T. longibrachiatum and B. stabilis effectively controlled Fusarium damping-off disease in P. massoniana

Damping-off disease caused by Fusarium spp. has been widely reported in crops (Scott and Punja 2023) and more recently in pine seedlings, causing serious economic losses (Luo and Yu 2020; Tahat et al. 2021). Biological control products are being extensively developed and applied as an alternative to chemical fungicides ( Kim et al. 2021). Scott and Punja (2023) found that five biological-control agents, including Trichoderma sp. and Bacillus sp., effectively reduced Fusarium infection in cannabis (Cannabis sativa L.) plants. Our previous study found that T. koningiopsis can control F. oxysporum damping-off disease in P. massoniana seedlings (Yu and Luo 2020). The current study isolated and identified a T. longibrachiatum 3Y strain and a B. stabilis K29 strain that exhibited plant growth-promoting traits, including potassium dissolution and IAA formation (Luo 2020) and had strong biocontrol activity. To our knowledge, this is the first study to use PGPR (B. stabilis) and PGPF (T. longibrachiatum) in combination to control damping-off disease in P. massoniana seedlings. The combined treatment reduced disease severity by 74%, which was higher than the response elicited by inoculation with a single biological control strain.

Both T. longibrachiatum 3Y and B. stabilis K29 inhibited the growth of F. oxysporum, causing dissolution and deformity of Fusarium hyphae. B. stabilis K29 has chitinase activity and can secrete proteases that degrade or hydrolyze the cell walls of fungal and fungal-like pathogens (Abdallah et al. 2015). Several beneficial bacteria isolated from various crops are also shown to suppress Fusarium growth by secreting cell wall degrading enzymes (Sriwati et al. 2023). Meanwhile, our previous study found that T. longibrachiatum 3Y produces various nonvolatile metabolites, including cyclohexanone, alcohols, and organic acids (Luo 2020), which may play key roles in inhibiting Fusarium growth. The current study showed that T. longibrachiatum and B. stabilis can synergize and have a stronger biological effect when used in combination.

T. longibrachiatum and B. stabilis impacted the rhizosphere microbial communities of P. massoniana against Fusarium disease and promoted seedling growth

Rhizosphere microorganisms use several mechanisms to protect plants from pathogen attack (Lee et al. 2021). For instance, the rhizosphere microbial consortium weakens the growth of pathogenic bacteria (Jain and Das 2020), beneficial microorganisms in the rhizosphere can improve disease prevention by enhancing the defense responses of the host (Jain et al. 2015; Jain and Das 2020), and rhizosphere beneficial microorganisms can affect the expression of genes related to stress and disease resistance in plants, dissolve phosphates in soil, and produce substances such as iron carriers, IAA, SA, and extracellular polysaccharides (Wang et al. 2021a). The composition of rhizosphere microbial communities is dependent on the plant species, soil type, and pathogens (Schreiter et al. 2014). Trichoderma and Burkholderia are well known for effectively colonizing soil or roots to control the invasion of pathogens into plants, which can impact the rhizosphere microbial community and improve plant disease resistance (He et al. 2018; Arici and Demirtas 2019; Yu and Luo 2020). The current study found that the composition of bacterial communities significantly changed when F. oxysporum-infected P. massoniana seedlings were inoculated with both T. longibrachiatum and B. stabilis.

In the process of controlling plant diseases, biocontrol strains may directly or indirectly affect the composition of rhizosphere microbial communities. Biocontrol strains may directly impact the growth of rhizosphere soil microorganisms through metabolites, active enzymes, nutrition and spatial competition (Whipps., 2001; Song et al. 2023), or indirectly regulate the growth of them through activating soil nutrients or affecting plant root exudates (Saeed et al. 2021). In this study, Ascomycota and Proteobacteria were the dominant phyla in all treatment groups, broadly corresponding to previously published surveys of soil microbial communities (Mendes et al. 2011). The relative abundance of Ascomycota was lower following T. longibrachiatum or B. stabilis treatment than F. oxysporum treatment, which supports previous findings that Ascomycota is less enriched in Fusarium wilt disease-free soil (Zhou et al. 2019). Interestingly, the decreased levels of Proteobacteria were associated with some fungal disease suppression (Shen et al. 2015; He et al. 2018), which contrasts with current findings that combined T. longibrachiatum and B. stabilis treatment significantly increased the abundance of Proteobacteria. These results suggest that Proteobacteria may help to protect P. massoniana seedlings against Fusarium infection. Meanwhile, we found that there were substantial differences in the relative abundances of most dominant bacterial genera among the four samples, but most fungal genera showed no significant differences. The use of T. longibrachiatum or B. stabilis increased the abundances of some beneficial microorganisms, including Sphingomonas, Pseudomonas, Comamonas and Cupriavidus, most of which are known to improve plant growth during drought, salinity, and oxidative stress (Asaf et al. 2020; Yasmin et al. 2022). However, further experiments are needed to confirm the direct or indirect effects of T. longibrachiatum or B. stabilis treatment on the relative abundances of those rhizosphere soil microorganisms when P. massoniana seedlings against Fusarium infection.

T. longibrachiatum and B. stabilis treatment changed key physiological pathways of P. massoniana seedlings against F. oxysporum infection

Plant cells can accumulate reactive oxygen species (ROS) during interactions with potential pathogens, causing oxidative damage, which results in lipid peroxidation and macromolecule break down (Mandal et al. 2008). Malondialdehyde (MDA) is a widely used marker of oxidative lipid injury caused by environmental stress (Kong et al. 2016). The current study found that F. oxysporum infection significantly enhanced MDA levels in P. massoniana seedlings, while combined T. longibrachiatum and B. stabilis treatment reduced MDA levels, possibly due to the significant up-regulation of DEGs (SODF, PER4, PER1, etc.) associated with ROS-scavenging enzymes and reduced cell damage (Mandal et al. 2008). These results suggest that the combined use of T. longibrachiatum and B. stabilis can increase antioxidant enzyme activity (mainly including peroxidase and superoxide dismutase) to help mitigate cell membrane damage caused by F. oxysporum.

Sugars are the primary substrates that provide energy and structural material for plant defense responses. They can also act as signaling molecules to activate immune responses against pathogens (Morkunas and Ratajczak 2014). Trichoderma-plant-pathogen interactions are associated with an increase in arabinose, xylose, and carbohydrate metabolism, and the higher sugar content in plant tissue improves resistance against F. oxysporum (Abdelrahman et al. 2016). The current study identified several up-regulated DEGs associated with sucrose, cellobiose, amylose, maltose, and trehalose metabolism induced by combined T. longibrachiatum and B. stabilis treatment. The sugars accumulated over time, indicating that their metabolism may play a pivotal role in protecting P. massoniana against F. oxysporum infection; however, further study is required to verify the mechanism.

Plant hormones modulate the expression of genetic networks involved in defense reactions, of which JA and SA constitute the hormonal backbone of plant immunity (Shigenaga et al. 2017). Trichoderma activates ISR through signal transduction pathways activated by JA/ET, but also includes crosstalk with SA and phytohormones associated with plant development (Hermosa et al. 2012). Trichoderma asperellum induces systemic resistance in Arabidopsis thaliana by activating the expression of genes related to ET, SA, and JA (Huang et al. 2015). Auxin can regulate pathogen resistance (Kazan and Manners 2009) as well as plant lateral root, leaf, flower, and vasculature development. In the present study, many DEGs associated with JA/auxin signal transduction were activated after inoculation with T. longibrachiatum and B. stabilis, indicating that JA/auxin pathways may play key roles in reducing F. oxysporum infection and promoting P. massoniana seedling growth.

Biotic attack from fungal, bacterial, and viral pathogens can decrease the photosynthetic rate and down-regulate photosynthesis-related gene expression (Kangasjärvi et al. 2012). Chlorophyll (Chl) is a vital photosynthetic pigment in plants that greatly influences photosynthetic capacity and plant growth and is easily degraded in response to biotic attack (Bilgin et al. 2010). While chlorophyll content is reduced in F. oxysporum-infected tomatoes, levels are significantly increased in the presence of Arbuscular mycorrhiza or effective microorganisms (Alshammari et al. 2022). Results from the current study indicate that co-inoculation of F. oxysporum-infected P. massoniana with T. longibrachiatum and B. stabilis up-regulates the expression of abundant photosynthetic genes and increases photosynthesis and chlorophyll content, illustrating that the photosynthetic system pathway was activated.

Conclusion

T. longibrachiatum and B. stabilis showed antagonistic activity against F. oxysporum, effectively controlling Fusarium damping-off disease in P. massoniana and promoting seedling growth. Combined use of the two biocontrol strains significantly changed the soil bacterial microbial community composition, and affecting the relative abundances of some beneficial microorganisms in the rhizosphere. Meanwhile both biocontrol strains treatment altered plant physiological pathway, and some DEGs associated with growth and disease resistance were up-regulated in response to Fusarium infection.