Field experiment and plant disease determination
From April 2012, a field experiment was conducted at the field station of the Chinese Academy of Sciences (Jiangxi Province, China). The soil at the study site is classified as Udic Ferrosol (FAO classification), covering an area of approximately 1×108 ha in southern China. Because of low soil organic matter content and fertility, and climatic suitability, peanut (Arachis hypogaea L.) is a particularly popular crop in the region. The field experiment included two cropping regimes (treatments): (1) monocropping with peanut; and (2) rotation, with a 2-year rotation of peanut alternated with other crops. For monocropping, the same peanut cultivar (Ganhua-5) was consecutively planted for the growing season (April–August) of each year. For crop rotation, peanut was grown first (2012), and then maize (Zea mays L.), potato (Solanum tuberosum), and soybean (Glycine max) were ordinally planted in every other peanut planting year (Fig. 1a). Three plots (replicates) of the two cropping regimes were laid out in a randomized block design, and lay fallow after harvest until the following sowing period. The management practices and fertilizer application are described in detail elsewhere5.
In the 2018 growing season, all plots were planted with peanut. The incidence of diseases caused by soil-borne fungi on each plot, a limitation of peanut productivity under consecutive monoculture, was continuously evaluated across the entire growing season (Fig. 1a). For each examination, 30 plants from each plot were carefully removed from the soil, and peanut root rot was evaluated using a five-class rating scale52. The examination of peanut root rot covered the peanut seedling, flowering, pod-bearing, and maturity stages. Overall, 720 plants (6 plots, 4 time points) were removed from the plots for examination. Before peanut planting, the soil was collected as the bulk soil, and the soil tightly attached to the roots was collected as the rhizosphere soil. Pathogens were analyzed from the roots of mature monocropped peanuts. Soil samples from each plot were pooled as one replicate, with three replicates for each treatment. Soil DNA was extracted for microbial community analysis. Pathogen density was determined in the rhizosphere soil by quantitative Real-time PCR (qPCR). Following a gentle wash with tap water, the plant height, shoot and root fresh weight, root length, pod weight, and nodule number were determined at each sampling.
Isolation And Identification Of Potential Pathogens
Potential fungal pathogens were isolated from mature peanuts from the monocropped plots that displayed symptoms of root rot, as described by Schuck et al. Briefly, a clean knife was used to cut the pathogen-infected roots into sections. The root sections were surface-sterilized (submerged in 4v/v sodium hypochlorite for 5 min), washed (two times, in sterile distilled water), and placed on a PDA plate containing streptomycin and penicillin (20 µg/mL) to obtain fungal isolates54,55. DNA was extracted from each fungal culture by using the FastDNA Spin Kit (MP Bio, Solon, OH), according to the manufacturer’s instructions. DNA sequences from the fungal 18S rRNA region were amplified using primers NS1(5′- GTAGTCATATGCTTGTCTC − 3′) and NS8(5′- TCCG-CAGGTTCACCTACGGA-3′). The PCR mixture contained (per 50 µL) 1.5 U of Taq polymerase (Red Taq, Sigma Chemical Co.) and the following reagents: 1⋅ Sigma PCR buffer, 0.20 mM PCR nucleotide mix (Promega), 4.0 mM MgCl2, 6.25 mg bovine serum albumin (Roche Diagnostics), and 25 pmol of each primer. For the amplification reaction, the DNA samples were initially denatured for 3 min at 95°C. This was followed by 35 cycles of denaturation (94°C, 30 s), annealing (57°C, 30s), and elongation (72°C, 105 s). The PCR program ended with a 2-min incubation at 72°C. Each fragment was compared phylogenetically to sequences of known species in the GenBank database of the National Center for Biotechnology Information (NCBI) by using BLAST. Phylogenetic trees were analyzed using MEGA v5. Phylogenetic trees were constructed by using the neighbor-joining (NJ) method. To confirm pathogen, cultures of isolated fungi increased in diseased roots were inoculated to 14 day-old peanuts rhizosphere (30°C, 70% relative humidity, light intensity 500 µM m− 2 s− 1), and the incidence of disease was recorded 30 d after inoculation.
Quantitative analysis of F. oxysporum
The changes of F. oxysporum abundance in the peanut rhizosphere from the seedling to maturity stages were quantitatively analyzed by qPCR. The following primer pair was used: Fa, 5′-TCGTCATCGGCCACGTCGACTCT-3′, and Ra, 5′-CAATGACGGTGACATAGTAGCG-3′. The reaction contained 7 µL of double-distilled H2O, 10 µL of 2× SYBR® Green master-mix, 0.5 µL of each primer (10µM), and 2 µL of template DNA. The PCR procedure was as follows: 95°C for 30 s; and 40 cycles of 95°C for 5 s, 57°C for 60 s, and 72°C for 60 s. Each sample was analyzed in three replicates.
Bioinformatic Analysis Of Root Endosphere Fungi And Rhizosphere Bacteria
Roots from healthy, mild or severe disease peanuts at maturity were surface-sterilized in 3% hydrogen peroxide, and washed with sterile water and 70% ethanol. Excess fluid on the sterilized root surface was wiped off using sterilized filter papers and root samples were cut into pieces. Then, root samples were ground in liquid nitrogen and DNA was extraction using the MiniBEST Plant Genomic DNA Extraction Kit (Takara). DNA was also extracted from the rhizosphere soil using the FastDNA® SPIN Kit for the Soil (MP Biomedicals, Santa Ana, CA). Specific sequences were amplified using the primer pairs ITS1F(5’-CTTGGTCATTTAGAGGAAGTAA-3’)/ITS2 (5’-GCTGCGTTCTTCATCGATGC-3’), specific for the fungal ITS1 region, and 338F (5’‑ ACTCCTACGGGAGGCAGCAG‑3’)/806R(5’-GGACTACHVGGGTWTCTAAT‑3’), specific for the V4 hypervariable region of the bacterial 16S rRNA gene. The samples were initially denatured for 3 min at 95°C; this was followed by 27 cycles of denaturation (95°C, 30 s), annealing (55°C, 30 s), and elongation (72°C, 45 s). The PCR program ended with a 10-min incubation at 72°C. The PCR products were separated by electrophoresis on 1% agarose gel. Amplicons were sequenced at Majorbio Bio-Pharm Technology Co. Ltd. (Shanghai, China) using the Illumina MiSeq platform, according to the manufacturer’s instructions. Raw fastq files were quality-filtered using QIIME. Low-quality sequences (< 150 bp long, with an average quality score < 25) were removed. The reads were trimmed and assigned based on unique 7-base barcodes. The barcode and primer sequences were then removed. The forward and reverse reads were incorporated into full-length sequences based on the thresholds: overlap length > 10 bp and mismatch ratio < 0.2. After discarding unqualified reads, the OTUs were assigned at 97% identity similarity level using UPARSE. Chimeric sequences were identified and removed using UCHIME.
Peanut Seedling Cultivation In Pot Experiment
In the current study, all plots were planted with peanuts for the 2018 planting season. Hence, to isolate the peanut rhizosphere microbes and assess their ability to suppress fungal pathogen invasion, soil samples were collected for pot cultivation experiments in the greenhouse after the 2018 planting season, which prevented disorganizing the field plot experiment (Fig. 1b). Peanut seeds were surface-sterilized with 0.1% mercury chloride for 5 min, and then rinsed five times with deionized distilled water. For each plot, 20 kg of the soil (0–20 cm layer) was randomly collected into 10 pots, and one surface-disinfected peanut seed (Ganhua-5) was sown in each pot. Overall, there were 10 biological replicate pots for each plot, with 60 experimental units (10 pots × 3 field plots × 2 crop treatments). After cultivation for 28 d (30°C, 70% relative humidity, light intensity 500 µM m− 2 s− 1), the plants were carefully removed from the pots and rhizosphere samples were collected by brushing off the soil adhering to the roots. The rhizosphere soil from 10 pots per plot was pooled. These six independent replicates from monocropping and rotation regimes were used for subsequent isolation of the rhizosphere bacterial colonies, the assessment of fungal pathogen suppression, and metatranscriptome analysis. Growth status of all plants, i.e., plant height, fresh weight, and root length and weight were determined, and the roots were scored for disease symptoms as described in the subsection on field experiment.
Suppression Of Fungal Pathogen Development By Peanut Rhizosphere Community
The suppression of a fungal pathogen by the rhizosphere soil was determined by in vitro microcosm antagonism test (Fig. 1b). F. oxysporum isolated from the peanut root rot was used. Before the cultivation experiment, a fungal plug (6-mm diameter) was transferred to PDA medium and placed in a biochemical incubator at 28°C for 3 d in the dark. To prepare soil bacterial suspensions, 1 g (dry weight equivalent) of fresh rhizosphere soil was placed in 9 mL of phosphate buffer (KH2PO4, 1 g/L, pH = 6.5), and mixed on a rotary shaker at 4°C and 150 rpm for 1.5 h. The suspension was then sonicated for 1 min at 47 kHz, twice, and mixed again for 0.5 h40. The suspension was filtered through a 5-µm filter to remove most fungal propagules before testing in vitro.
For the microcosm separation, an inverted assay was performed using sterile Petri dishes (9-cm diameter), as described in56. Briefly, a plug containing pathogenic fungal hyphae was inoculated at the bottom of a Petri dish containing PDA, and incubated at 28°C for 24 h. A layer of nutrient agar (TSA) (1.5 g/L tryptone, 0.5 g/L soytone, 0.5 g/L NaCl, and 15 g/L agar, pH 7.0) was poured into the lid of the Petri dish. After solidifying, 200 µL aliquots of bacterial suspension were evenly spread on the TSA. As a control, an equal amount of sterile water instead of the bacterial suspension was used. Then, the top of the dish was placed over the matching bottom part, sealed with Parafilm, and incubated at 28°C until the control PDA plates were filled with mycelia. In this way, the fungi to be tested were exposed to any antifungal substances produced by bacteria in the upper compartment. Each treatment was repeated three times. Percentage inhibition was calculated as [(fungal mycelium diameter in the control)−(fungal mycelium diameter in the treatment)]⋅[100/(fungal mycelium diameter in the control)].
For the direct microcosm experiments, 60 µL of bacterial suspension was inoculated evenly on a piece of sterile filter paper (1.8 cm × 4.8 cm) placed on one side of NA medium. After 3 d, a PDA plug (8-mm diameter) containing fungal hyphae was placed 2 cm away from the filter paper. The plates were sealed with Parafilm and incubated at 28°C. The fungal pathogens were exposed to the secretions produced by the bacterial communities on the filter paper. The extension of fungal hyphae was measured after 72 h and compared with that on the control plates inoculated with sterile water instead of the bacterial suspension. Each treatment was repeated three times.
Determination Of Fungal Suppressive Compounds In Rhizosphere Community
Since our previous studies indicated high suppressive activity of VOCs produced by the soil bacterial community against plant pathogenic fungi57, the VOCs produced by bacterial communities from the peanut rhizosphere were determined, using thermal desorption coupled with GC-MS. For VOC collection, 200 µL of the rhizosphere bacterial suspension was inoculated on TSA medium, sealed with Parafilm, and cultured in the dark in a 25°C incubator for 3 d. Then, the septum was pierced with a solid phase microextraction needle (Supelco, Bellefonte, PA) and the fibers were exposed to the sample headspace for 60 min at 30°C. After the extraction, the fibers were retracted into the needle, transferred immediately to the injection port of a GC (GC-CP3800, Agilent Technologies), and desorbed at 250°C for 3 min. The compounds collected in the headspace were analyzed by the GC connected to a mass spectrometer (Varian Saturn 2200). The column used was 30 m × 0.25 mm i.d., with film thickness of 0.25 µm (CP-8, Agilent Technologies). Helium was the carrier gas, at a constant linear velocity of 1 mL/min. The following temperature program was used: 50°C for 2 min, then ramping up to 260°C at 5°C/min, and holding for 2 min. The mass spectra were scanned at 70 eV over a mass range from m/z 35 to 600. VOCs were identified based on their individual mass spectrum, retention index (RI), and retention time (RT) by comparing with those in reference databases (NIST Mass Spectral Data 08′ edition). Peak areas of all components were calculated by using Xcalibur 2.0, and relative amounts (RAs) were calculated based on the peak-area ratios of all volatiles.
Isolation And Identification Of Cultivable Bacterial Strains
For the experiment, 1 g of rhizosphere sample was mixed with 9 mL of MS buffer solution [50 mM Tris-HCl (pH 7.5), 100 mM NaCl, 10 mM MgSO4, and 0.01% gelatin] on a rotary shaker at 170 rpm for 30 min at 30°C. Then, 100-µL volumes of serially diluted soil suspensions were plated on TSA, and incubated for 48 h at 30°C in the dark. All colonies were picked from plates and further re-streaked on fresh TSA plates for purification. According to the morphology (color and size) of the bacterial colonies, 200 isolates were selected for 16S rRNA gene sequencing for identification. DNA was extracted using the FastDNA® SPIN Kit for the Soil (MP Biomedicals, Santa Ana, CA), according to the manufacturer’s instructions. The region of interest was amplified using PCR primers 27F, 5′-AGAGTTTGATCCTGGCTCAG-3′, and 1492R, 5′-GGTTACCTTGTTACGACTT-3′. The PCR cycle consisted of an initial denaturation at 94°C for 5 min; followed by 35 cycles of denaturation at 94°C for 1 min, annealing at 54°C for 1 min, and extension at 72°C for 1 min; and a final extension for 5 min at 72°C. PCR products were verified on 1% agarose gel and sequenced by Sangon Biotech Co., Ltd. (Shanghai, China). The BLAST search program was used for sequence identification. Highly homologous sequences were aligned, and NJ trees were generated using MEGA v5. To confirm the existence of these bacterial isolates in the peanut rhizosphere, phylogenetic analysis was performed based on the 16S rRNA sequence of the bacterial isolates and gene sequences of OTUs from the same genus.
The V4 region of the 16 S rRNA gene sequences were aligned with the consensus sequences of the depleted/enriched OTUs. The isolates and OTUs with the closest evolutionary distance were considered to be the same strain.
Characteristics Of Bacterial Strains Required For Rhizosphere Colonization
To investigate why certain strains isolated from the rotation peanut rhizosphere were depleted from the monocropped peanut rhizosphere, their ability to form biofilm and respond to peanut root exudates were tested. The former was determined using a microtiter plate assay with crystal violet staining, following the protocol described by Robledo et al. Briefly, bacterial suspension were prepared by culturing in NB medium (5 g/L beef extract, 10 g/L soytone, and 5 g/L NaCl, pH 7.2) and adjusting density to OD600 = 1.0 with sterile water. Wells of a 96-well plate containing 180 µL of sterile NB medium per well were each inoculated with 20 µL of the bacterial suspension. Negative control wells contained sterile NB medium but were not inoculated. The plate was sealed with Parafilm and incubated at 25°C for 48 h. Three replicates for each strain were tested. The cultures were then aspirated, and each well was washed three times with 200 µL of sterile PBS. The wells were stained with 100 µL of 1% crystal violet and incubated at 28°C for 5 min. After incubation, the crystal violet solution was removed, and the wells were washed six times with 150 µL of sterile water. The remaining crystal violet was solubilized with 100 µL of 33% acetic acid and sample OD590 was recorded.
To determine the response of certain strains to peanut root exudates, peanut seeds (Ganhua-5 variety) were surface-disinfected with 2% NaClO solution for 10 min, and rinsed four times in sterilized distilled water. The sterilized seeds were transferred to a box containing vermiculite and incubated in a growth chamber at 28°C, with a 16-h light/8-h dark photoperiod. After 15 d, the seedlings were uprooted from the substrate, and the roots were gently washed to remove the adhered vermiculite. The individual seedlings were transplanted into 50-mL flasks containing 50 mL of sterile liquid 1/2 sucrose-free Hoagland medium at 28°C. After incubation for 30 d, root exudates were collected from 20 seedlings. To do that, plant roots were washed four times with sterile double-distilled water to remove the nutrient solution. Then, each plant was placed into a 50-mL flask and the roots were submerged in 50 mL of sterile double-distilled water. All plants were placed in a growth chamber for 24 h (16-h light/8-h dark photoperiod) at 28°C with gentle shaking (50 rpm). Next, 20 µL of bacterial suspension and 180 µL of root exudates were placed in a sterile 96-well plate. Each bacterial solution was tested in triplicate, and a bacterial suspension mixed with sterile water was used as a control. Sample absorbance was measured using a spectrophotometer at OD600 every 12 h, and the response of different bacteria to root exudates was determined based on changes in bacterial density. To understand the functional properties of the monocropping-depleted and enriched bacteria, phosphate solubilization, and siderophore and IAA production capacity of the bacteria were further tested, as well as their inhibitory effect on F. oxysporum59.
Metatranscriptome Sequencing Of Culturable Rhizosphere Microbiome
Following the inverted antagonism assay described in “suppression of fungal pathogen development by peanut rhizosphere community”, 4 mL of sterile deionized water were added to the plate containing bacterial colonies, and suspensions of rhizosphere microbiome were obtained by mixing with a sterile inoculating loop. Total RNA was extracted from the bacterial suspension using the UltraClean® Microbial RNA Isolation Kit (Mobio), according to the manufacturer’s instructions. Then, mRNA was enriched using the MICROBExpress™ Bacterial mRNA Enrichment Kit (Ambion). The enriched mRNA sample was used as a template in a reverse-transcription reaction to synthesize cDNA. The constructed metatranscriptome library was then sequenced using the Illumina HiSeq 4000 platform at Shanghai Personal Biotechnology Co., Ltd (Shanghai, China). Based on the FastQC quality report, filtered mRNA reads were cleaned using Cutadapt v1.2.160. Specifically, reads < 50 bp and 10 bp minimum overlap with maximum 20% mismatch were removed.
For functional analysis, the quality-filtered reads were sorted using SortMeRNA61, which separated rRNA from non-rRNA and thus, potential mRNA. Trimmed high-quality sequences were assembled de novo using Trinity v2.2.062. Sequences were not assembled into contigs to avoid the problem of chimera formation arising from highly diverse communities, as previously recommended63. Cluster Database at High Identity with Tolerance (CD-HIT) was used for sample collation after assembly and splicing. Transcripts were merged to eliminate redundancy at a similarity of 0.95 and a minimum coverage of 0.9, and the longest sequence was used as the representative sequence in UniGene. Bowtie2 v2.2.9 and RSEM v1.3.0 were used to functionally annotate Trinity contigs, and to align reads and quantify transcripts. Five samples out of six passed quality control after sequencing; sample “C1” yielded very low sequence counts and was not included in the analysis. The resulting alignments were annotated using KEGG database64 with MEGAN65.
Suppression of F. oxysporum by monocropping-depleted strains
To determine whether the monocropping-depleted strains increased pathogen resistance of the rhizosphere, we first determined the inhibitory effect of the addition of depleted strains on F. oxysporum. Briefly, suspension of depleted stains were prepared by culturing in NB medium and adjusted density to OD600 = 1.0 with sterile water. Equal depleted stains suspension volumes were mixed to create a SynCom. Following the assay described in “metatranscriptome sequencing of culturable rhizosphere microbiome”, suspensions of rhizosphere microbiome were obtained and adjusted density to OD600 = 1.0. The SynCom (5 mL) was added to a suspension (10 mL) of rhizosphere microbiome from monocropped peanut. Inhibition of F. oxysporum by the rhizosphere microbiome with or without SynCom supplementation was examined. Each treatment was tested in quadruplicate.
Next, the protective effect of depleted strain addition on the root and plant growth was assessed. Peanut seedlings were cultivated under sterile conditions, as described by Li et al66, with slight modification. Surface-disinfected peanut seed were placed on sterile moist filter paper for five days to germinate. Well-grown and uncontaminated seedlings were planted in 200-mL beakers (1 seedling per beaker) containing sterile vermiculite and 50 mL of sterile Hoagland’s nutrient solution (1/4 strength). Six 200-mL beakers were then placed in a 5-L beaker, covered with four layers of sterile gauze to prevent microbial contamination, and incubated in a plant growth chamber (30°C, 70% relative humidity, light intensity 500 µM/m2s1). After 3 d of cultivation, 10 mL of SynCom and 10 mL of monocropping microbiome suspension, or 10 mL of sterile water and 10 mL of monocropping microbiome suspension, were added to the vermiculite in 200-mL beakers. Three days later, the vermiculite was inoculated using 10 mL of F. oxysporum spore suspension (105 CFU/mL) or sterile water. After 30 d of incubation, the disease incidence and plant growth status, i.e., height, fresh weight, and root length and weight, were determined.
Indoor And Field Analysis Of Syncom Biocontrol Effect On Fungal Pathogen
To investigate whether the monocropping-depleted strains have an additive effect on plant disease defenses, three different SynCom types, composed of 7, 4, or 2 depleted strains, were created. Only one strain combination was used for the SynCom composed of 7 depleted strains, while 35 and 21 combinations of strains were tested in the SynComs composed of 4 and 2 strains, respectively. SynComs were prepared as described in the preceding subsection. Determination of the inhibitory effect of different SynComs on fungal spore germination was done by co-cultivation. First, to wells of a 96-well plate containing 100 µL of F. oxysporum conidial suspension (105/mL) in each well, 100 µL of the SynCom suspension was added. The control group was mixed with an equal volume of sterile water instead of a SymCom. After 24 h of co-cultivation at 28°C, spore germination was evaluated under a microscope. Next, 60 µL of SynCom suspension was inoculated evenly on a piece of sterile filter paper placed on one side of NA medium. After 2 d of co-cultivation at 28°C, a PDA plug (8-mm diameter) containing fungal hyphae was placed 2 cm away from the filter paper. The growth of fungal hyphae was assessed after 72 h and compared with that on control plates, where sterile water was used instead of the bacterial suspension. Each SynCom combination was tested in six replicates.
Finally, the effect of different SynComs on plant health and growth was verified in a monocropped peanut field. The field experiment was set up at the Red Soil Ecological Experiment Station of the Chinese Academy of Sciences in Yingtan (Jiangxi Province, China). The control and 7-strain SynCom treatments were arranged in 4 plots (2 m ⋅ 2 m). For 4- and 2-strain SynComs, each combination was arranged in one plot, for a total of 64 plots. Before sowing, peanut seeds were soaked in a SynCom suspension for 12 h. Forty-five days after planting, all peanut plants were harvested from each plot to determine their health and growth status.
Statistical analysis
All statistical analyses were performed in the R environment (v4.1.0, http://www.r-project.org/). When the data satisfy the assumptions of normality and homoscedasticity, we performed Student’s t tests (two-sided) to compare the statistical significance between pairs of samples and analysis of variance (ANOVA) to determine the statistical significance of multiple comparisons. Otherwise, Mann-Whitney U test (pairs samples) and Kruskal–Wallis tests (multiple comparisons) and were performed. Alpha-diversity indices (including the ACE, Chao1, Simpson, and Shannon indices) were analyzed using the “vegan” package in R. The results were visualized using the “ggplot2” package. Alpha-diversity indices of different samples were analyzed using the analysis of variance. PCoA based on Bray–Curtis dissimilarity matrix was performed and plotted using the “vegan” package to describe the bacterial community under different treatments and at different sampling times. Differences in bacterial family abundance were examined using STAMP test. Based on the culturable rhizosphere microbial community, monocropping-depleted OTUs were defined as OTUs not detected under the monocropping regime, and monocropping-enriched OTUs were defined as OTUs not detected under the rotation regime. Phylogenetic trees of bacterial rhizosphere isolates, and of the enriched and depleted OTUs were constructed using MEGA v5, based on 16S rRNA sequences. Phylogenetic trees were annotated and visualized using iTOL software67. Spearman correlation scores were calculated, and only robust (Spearman’s r > 0.5 or r < m–0.5) and statistically significant (P < 0.01) correlations were retained. KEGG enrichment scatter plot analysis was performed using OmicStudio tools available at https://www.omicstudio.cn/tool/11.