Host or pathogen primed bacterial consortia alters microbial community structures in Arabidopsis roots and suppresses Fusarium oxysporum

doi:10.21203/rs.3.rs-1990907/v1

Download PDF

Research Article

Host or pathogen primed bacterial consortia alters microbial community structures in Arabidopsis roots and suppresses Fusarium oxysporum

https://doi.org/10.21203/rs.3.rs-1990907/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background: The plant-associated microbiota confers beneficial traits to the plant host that promote growth and prevents disease. It is, however, not fully understood how the host, and pathogens, modulate the assembly of the associated microbiota. In this work, we investigated the effects of inoculation with bacterial consortia (BCs) that had been primed with either the host plant Arabidopsis thaliana alone or by a combination of the host and the pathogen Fusarium oxysporum f. sp. mathioli (FOM), on the root-associated microbiota and FOM. We used 16S rRNA and ITS amplicon sequencing for profiling root bacterial and fungal communities, respectively. qPCR was employed for the assessment of FOM quantities in the shoots of Arabidopsis.

Results: Our results showed distinct effects of host or pathogen primed BCs on microbial community structures and networks. The host-primed BCs inhibited FOM proliferation in shoots, distinctively affected rhizo-microbial community structures and further enhanced the connectivity of the microbial networks. Beneficial bacterial genera including Flavobacterium, Azospirillum, Massilia and Pedobacter were significant biomarkers in the samples inoculated with host-primed BCs and could potentially be involved in FOM antagonism. Likewise, FOM-primed BCs reduced FOM in the shoots of Arabidopsis and altered microbial networks. Bacterial taxa Bacteriodetes, Flavobacterium and Chthoniabacteraceae and the fungal, Exophiala and Phaeosphaeriaceae were the most enriched biomarker taxa in the samples treated with FOM-primed BCs. Moreover, the host genotype effect was significant in overall microbial community network structuring and could contribute to resistance to FOM pathogens.

Conclusions: Altogether, these findings deepen our understanding of both host- and pathogen-priming on plant-associated microbiomes, thus, revealing pivotal engineering routes in exploring microbiomes to manage plant diseases.

rhizosphere engineering

microbiome inoculation

disease suppression

fungal pathogens

invasion networks.

The plant as a meta-organism derives enormous functionality from its associated microorganisms. This includes beneficial microbial processes involved in mechanisms that promote plant growth such as nutrient acquisition, disease suppression including induction of systemic resistance, tolerance to abiotic stresses, and phytostimulation [1–3]. Below-ground, plant-microbe interactions mostly occur in the rhizosphere, the root zone actively influenced by root secretions that are serving as a rich carbon source that attracts diverse soil microbes [4, 5], as well as providing chemical signals that mediate symbiotic interactions [6].

Plants are often challenged by pathogens that are highly dependent on inoculum levels, host genotype and environmental conditions as well as management practices for successful infection [7], factors that also affect overall microbial community networks [8–11]. Plant fitness and performance is dependent on a resilient rhizosphere, characterized with positive feedback repertoires that revivify plant nutrition and persistently suppress pathogen attack in varying growing conditions [12]. Distinct microbiomes in terms of composition and abundance have been observed in the rhizosphere of healthy and diseased plants [13–15] supporting the role of the plant-associated microbiota as determinants of plant health. Beneficial microbial taxa characterize suppressive soils and are enriched in response to disease, counteracting pathogen progression [16, 17]. A classic example is the suppression of the take-all pathogen, Gaeumannomyces graminis var. tritici resulting from the enrichment of 2,4-diacetylphloroglucinol (2,4-DAPG)-producing Pseudomonas fluorescens in continuous wheat or barley monocultures [18]. Application of Bacillus amyloliquefaciens resulted in Fusarium wilt suppression that correlated with an increase of Pseudomonas species [19], while the inoculation of bacterial core microbiota members enhanced Fusarium wilt resistance in Arabidopsis ([20].

Although the rhizosphere microbiota has a huge potential in sustainable cropping systems, its complexity, in terms of its physicochemical and biological dynamics have limited its exploitation [21, 22]. Recently however, novel techniques focusing on the plant and the plant-associated microbiota have become targets for optimizing the rhizosphere. Rhizosphere engineering approaches utilizing plant microbiomes have drawn research interest primarily due to the overwhelming positive impact it confers to plants. While single or few beneficial microbial strains such as plant growth promoting rhizobacteria (PGPR) have been utilized to modify the rhizosphere, several setbacks including a surcease due to competition in the long term by resident microbial populations have been reported [23]. Hence, microbial activity is dependent on other microbial members within the microbiota [24]. Inoculum diversification via utilization of sub or whole microbial communities is suggested to facilitate microbial establishment and promoting its overall activity [12, 23, 25].

Microbiome inoculation has been shown to enhance growth by altering physiological traits such as root elongation and increased shoot biomass and flowering time. For example, Panke-Buisse et al., [26] observed shifts in flowering time of plants inoculated with microbiota harvested from early- or late- flowering plants. Moreover, microbial consortia from host plants may have a high abundance of beneficial microbes with a high functional redundancy, including antagonists that suppress pathogens [26–28]. Jiang et al., [29] reported the protective potential of rhizosphere microbiome consortia against the bacterial wilt pathogen Ralstonia solanacearum. Overall, a successful transfer of “primed” or induced plant-associated microbial consortia could confer both growth stimulating traits as well as enduring resistance against pathogens [28, 30, 31].

Our aim in the present work was to study the effects of either host- or pathogen bacterial consortia (BCs) priming on the robustness of the rhizosphere microbiota and on the root invading Fusarium oxysporum f. sp. mathioli (FOM). We hypothesized that pre-inoculation of either host- or pathogen primed bacterial microbiota would differentially affect the resident plant root microbiomes and their response to FOM infection. The main objectives were thus to compare the effects on host and pathogen primed bacterial consortia on i) microbial community structures and ii) FOM disease progression. For our experiments we used Arabidopsis grown in natural soils across five growth generations including a sub-community approach where rhizosphere bacterial consortia isolated from the preceding Arabidopsis generations were inoculated into soils of the succeeding seedlings.

Plant materials and fungal strains

Two Arabidopsis wildtypes Col-0 and Ler-0 obtained from the Nottingham Arabidopsis Stock Center (NASC), Nottingham, United Kingdom, were used in this study. Prior to the start of our experiments, Arabidopsis thaliana seeds were multiplied to obtain uniform seed lots for each planting generation. Col-0 and Ler-0 exhibit distinct root morphologies [32, 33] chemical profiles [34] and disease resistances against F. oxysporum [35].

The Fusarium oxysporum f.sp. mathioli culture, isolate 726 [36] was kindly provided by Dr. Corby Kistler, USDA, St. Paul, Minnesota and was maintained on potato dextrose agar (PDA).

Experimental design

Arabidopsis seeds were sown in pots (8 cm x 8 cm x 6 cm) containing field soil (fine sand 32.2%, coarse sand 52.8%, humus 4.7%, clay 3%, silt 7.3%) [37], pH 5.95, collected from Jyndevad Research station (N 46° 28' 20.716''/E 9° 28' 45.347''), Denmark, in 2016. We replicated each genotype five times with each pot (containing 15–20 seeds) representing a replicate. For stratification, pots with seeds were arranged in trays and loosely covered with clear plastic wraps in the dark for 3 days at 4ºC. The pots were later transferred, completely randomized and maintained under greenhouse conditions (16h light, 8h dark at 18–23 ºC). Seedlings were watered (100 ml/pot) two times per week and weeds removed regularly upon emergence. After germination, thinning was done to allow 10 seedlings to grow in each pot.

The experiment consisted of three manipulated setups (Fig. 1A). Setup one (S1) comprised five generations; initial host-primed BC generations (G1and G2) followed by FOM-primed BC generations (G3-G5). S2 consisted of an unprimed generation (G3) followed by FOM-primed generations (G4-G5). S3 (control) consisted of an unprimed generation (G3) followed by G4 and G5 that were inoculated with host-primed BC. The BCs were applied to the soils containing 10 days old seedlings of the following planting generation. In S1 and S2, FOM was inoculated three days after BC treatment. Complete randomization was done in all setups and seedlings were maintained as previously stated. To confirm successful FOM infection of Arabidopsis seedlings, shoots (six weeks old) were sampled for FOM biomass quantification using qPCR.

For collection of BCs and characterization of the microbiota, roots of six weeks old Arabidopsis plants were harvested by gently pressing the sides of planting pots to loosen soils around the plant and then carefully pulling out the plant and shaking gently to remove loosely attached soils. Roots and the remaining attached soils (collectively referred to as root samples) from the 10 plants in each pot were pooled into 15 ml sterile collection tubes and homogenized before subsampling. From this pool, samples (ca. 3g) were subsampled into 2ml Eppendorf tubes and used for metabarcoding while the remaining samples were pooled into 50 ml falcon tubes and placed on ice for isolation of BCs. All samples for metabarcoding and qPCR were kept in liquid nitrogen until storage at -20 ºC.

Purified bacteria consortia preparation and application

To purify BC, 45ml sterile deionized water was added to the root samples in falcon tubes (Supplementary Figure S1A). The content was shaken vigorously for 5 minutes and centrifuged at 4000 rpm for 5 minutes and the supernatant transferred into new sterile falcon tubes. The collected supernatant was filtered using a 5-µm sterile Sterlitech filters that retain eukaryotes while most bacterial taxa pass through. To confirm the presence of bacteria and absence of fungi in the filtrate, we prepared cultures from pre- and post- filtered supernatants by streaking onto LB media. (Supplementary Figure S1A). 40 mls of this filtrate was applied to soils of 10 days old plants in each planting generation.

FOM culture preparation, quantification, and inoculation

FOM was cultured on synthetic nutrient agar (SNA) media. Mycerial plugs (3 plugs/100ml) from 7 days old cultures were transferred into a 1L sterile liquid carboxymethylcellulose (CMC) (high viscosity sodium salt, #C5013: Sigma Aldrich, St. Louis, MO, USA) broth containing 1.5% CMC-Na, 0.1% NH4NO3, 0.1% KH2PO4, 0.1% yeast extract (YE) and 0.05% MgSO4. 7H2O) and cultured for 3 days at 22 ˚C in the dark with gyratory shaking (125 rpm). A total of 5L of CMC was prepared for each generation of FOM culture. Fungal spores were harvested by filtering through sterile Miracloth to remove mycelia followed by centrifugation (4500 g) of the filtrate containing FOM spores for 15 minutes at room temperature to pellet spores. The spores were then washed twice with sterile deionized distilled H₂O, centrifuged at 7500g for 5 min at room temperature and the supernatant discarded. The spore pellet was resuspended in sterile water and spore concentrations estimated using a hemacytometer. Inoculation was done by carefully pipetting the spore suspension adjacent to the roots of two weeks old Arabidopsis seedlings. 300 uL of the FOM spore suspension with a concentration of 1.1 x10⁶ spores per mL was inoculated into soils of G3, G4 and G5 plants. The non-FOM infected setup (S3) was inoculated with distilled water.

DNA extraction

Root and shoot samples in 2 ml Eppendorf tubes were lyophilized for 72 hours. Sterile beads (2 per tube) were added to the dried samples that were ground with a Geno/Grinder 2000 at a speed of 1500 rpm for 3 x 30 seconds. Shoot DNA was extracted using the DNeasy Plant Mini kit (Qiagen, Hilden, Germany) while root sample DNA was extracted following the PowerLyzer™ PowerSoil® DNA Isolation Kit (Mo Bio Laboratories, Carlsbad, CA, USA) using 250 mg of the ground samples. All manufacturer’s instructions for extraction were followed, and DNA concentrations were quantified using a Nanodrop 1000 spectrophotometer (Thermo Scientific, Waltham, MA, USA). DNA samples were stored at -20 ⁰C for downstream processing.

Library preparation and amplicon generation

Bacterial and fungal libraries for Illumina MiSeq sequencing were generated by a two-step dual indexing strategy as previously described [38]. Briefly, PCR amplicons were generated using bacterial specific primers (S-DBact-0341-b-S-17/S-D-Bact-0785-a-A-21) [39] to amplify the V3/V4 region of 16S rRNA while the fITS7 and ITS4 fungal primer pairs was used to amplify the Internal Transcribed Spacer 2 (ITS2) region of the fungal rRNA gene [40]. PCR were performed in a 25 µl reaction mixture consisting of 1 × PCR reaction buffer (Promega Corporation, Madison, USA), 1.5 mM MgCl2, 0.2 mM dNTPs, 1 µM of each of forward and reverse primers, 1 U of GoTaq Flexi polymerase (Promega Corporation) and 1µl of DNA template. The PCR cycle program included 25 cycles (94 ^◦C for 30 sec’’, 55 ^◦C for 30sec’’, 72 ^◦C for 30 sec and a final elongation at 72 ^◦C for 10 min). This was followed by a 15-cycle indexing PCR with a similar PCR reaction mixture and thermal cycling parameters as described above. Fungal PCRs were performed similarly except using an annealing temperature of 57 ^◦C [40]. A dual indexing approach was used for combining samples within each index combination [41, 42]. Primer sequences with internal barcodes and the index combinations are reported in Kudjordjie et al.,[38]. The amplicon size of PCR products was confirmed by visualization in a 1.5% agarose gel using SYBR staining. The barcoded PCR products were equimolarly pooled for equal representation, followed by ethanol precipitation and re-elution in 50 µl TE buffer. The pooled DNA was separated on a 1.5% agarose gel and the DNA band of the expected size (300–450 bp) was extracted using a QIAquick Gel Extraction Kit (Qiagen). The concentration of the extracted product was determined using a Nanodrop 1000 spectrophotometer, and finally shipped to Eurofins MWG (Ebersberg, Germany) for sequencing on a MiSeq platform. Sequence files are deposited in the NCBI sequence read archive under the SRA accession number PRJNA850631.

Pathogen quantification

qPCR was performed to determine fungal biomass accumulated in Col-0 and Ler-0 in the G3 and G5 generations as previously described [43]. Briefly, we used the Fusarium oxysporum specific primers F 5’- CCT GTT CGA GCG TCA TTT CA-3’ and R 3’- GAA TTA ACG CGA GTC CCA ACA C-5’ [38]. The PCR reaction mixture of 12.5 µl consisted of 2.5 µl template, 0.375 µl each of the forward and reverse primers (10 µM stock), 3 µl of water, and 6.25 µl of Bio SyGreen Mix Lo-Rox (PCR Biosystems). 2 µl of purified FOM DNA or sterile water were included as positive and negative controls, respectively. Amplification reactions were performed with 2 replicates per sample in a ViiA ^TM 7 Real-time PCR system (Life Technologies) using 384-well plates. Thermal cycling conditions included an initial denaturation at 95°C for 10 minutes followed by 40 cycles of 95°C for 15s and 58°C at 1 minutes. Standards were included in the run using known dilutions of FOM. A standard curve was obtained by plotting the cycle threshold (C_T) values as a function of log₁₀ of the amount of fungal biomass in a 10-fold serial dilution (1 to 10^− 8) of FOM DNA.

Sequence data analysis

The Illumina MiSeq sequence reads were analyzed in QIIME version 1.9 [44] as previously described [38](Kudjordjie et al., 2019). Briefly, paired-end reads were joined using VSEARCH version 2.6 [45] with an overlapping minimum read length of 30 base pairs. Reads with quality Phred scores less than 30 were discarded. Internal barcodes, forward and reverse primers, and reads of less than 200 base pairs were removed using the command split_libraries.py. The remaining high quality sequences were dereplicated, screened for chimeras and clustered using VSEARCH [45]. For the fungal reads, ITS extraction was done before clustering using ITSx extractor, version 1.0.6 [46]. Taxonomy assignments for the clustered operational taxonomic units (OTUs) were done using the SILVA 132 [47] and UNITE (v7.2) [48] reference databases, respectively, for bacteria and fungi, in QIIME (v1.9) using assign_taxonomy.py [44] (Caporaso et al., 2010). Unassigned OTUs at kingdom level or assigned as chloroplast or mitochondrial sequences were removed.

Statistical analysis

Bacterial and fungal community analysis was performed in R version 3.5.2 [49], primarily, using vegan (v2.5.7) [50] (Oksanen et al., 2021), phyloseq (v1.34.0.) [51], ggplot2 (v3.3.2)[52] packages. Samples with less than 2000 or 500 reads were removed from the bacterial or fungal datasets, respectively, prior to alpha diversity estimation. The OTU tables were rarified 100 times at a depth of 2000 reads for both datasets and the mean of the diversity estimates of 100 trials was used to estimate each alpha diversity metric (observed and Shannon diversity). Fungal alpha diversity analysis was assessed with and without fOTU1 (identified as the inoculated pathogen, FOM). Significant differences between alpha diversities were evaluated using the Kruskal-Wallis rank sum test. OTU tables were transformed to relative abundances before computing beta diversity. Dissimilarity matrices for bacterial and fungal datasets were visualized using unconstrained principal coordinates analysis (PCoA). Variance partitioning and significances of experimental factors was performed by subjecting Bray-Curtis dissimilarity matrices in bacterial and fungal communities to permutational analysis of variance (PERMANOVA) using the ‘adonis’ test from the ‘vegan’ package.

FOM DNA in shoots of Arabidopsis from the different setups was calculated. The normality of qPCR data for each generation was determined using Shapiro-Wilk tests, followed by a Kruskall-Wallis test. A pairwise Wilcoxon test was used to test for differences in biomass for genotypes in different setups in each generation. By using the S1 dataset, we examined whether a linear relationship existed between treatment and succeeding planting cycles.

Biomarker and differential OTUs analysis

Linear discriminant analyses effect size was performed to identify microbial taxa (biomarkers), at taxonomic levels, that characterized the differences between groups of samples, using the ldamarker function in the R package “Microbial” (v.0.022). Prior to the analysis, data was normalized to relative abundance. Significant differentially abundant taxa were detected using the Kruskal-Wallis test at pvalue < 0.05. Results were plotted using the plotLDA function with a LDA score threshold of 4.

Differential analysis was performed using DESeq2 (version1.22.2) [51], to determine bacterial and fungal OTUs with significantly differential abundances within and between generations and setups. We used default parameters, with at least an adjusted p-Value of p < 0.01. Microbial OTUs with significantly different abundances were identified by comparing generations within and between S1, S2 and S3. To visualize differential OTUs, the count data were normalized by the variance-stabilizing transformation (VST) [51, 53].

Cooccurrence networks

Bacterial and fungal networks were constructed using the package “SpiecEasi” in R [54, 55]. Networks were constructed using the most abundant taxa (top bacterial phyla and fungal genera). The “spiec.easi” function was executed with the following specifications to construct all networks: method = “mb”, lambda.min.ratio = 1e-2, nlambda = 20, pulsar.params = list(rep.num = 50). Networks were visualized using ggnet2 function of the Ggally package of R. Network properties, including transitivity (Tr) or clustering coefficient, mean node degree and average path length and the number of positive and negative edges were determined for network robustness comparison.

FOM biomass in shoots of Arabidopsis

qPCR analysis generally showed lower FOM quantities in G5 compared to G3 in Arabidopsis tissues (Fig. 1B). In G3, FOM levels were as expected higher in samples from the FOM inoculated setups S1 and S2 than in the non-inoculated S3. There were higher FOM levels in the un-primed S2 shoots compared to the host-primed S1shoots in G3, although statistically non-significant (Fig. 1B). We further assessed the effects of a continuously FOM-primed microbiome on FOM accumulation in Arabidopsis shoots in the S1 setup and found a significant and decreasing amount of FOM biomass during generations (p < 0.001) and genotypes (p < 0.01) (Fig. 1C).

Figure 1. Schematics of experimental outline and Fusarium oxysporum f.sp. mathioli (FOM) quantification. A) Our experimental system showing the three different setups S1, S2 and S3 and the treatments applied, indicated with a red pipette for FOM application and a blue for BCs. Arrows show the transfer of BCs from one planting generation to the other. B) Boxplots of FOM quantities in shoots of Col-0 and Ler-0 lines harvested from generations 3 (G3) and 5 (G5) in S1, S2 and S3. FOM quantities were determined using qPCR. Significant differences of FOM levels in shoots of G3 and G5 are shown. C) Relationship between FOM quantities and planting generations (G3, G4 and G5) in S1. A significant decreasing effects of both experimental cycle (P < 0.001)) and genotype (P < 0.01) with generation was observed. The trendlines were produced using geom_smooth function in the ggplot2 package.

Sequence data

Sequencing of the rhizosphere samples amplified with bacterial primers yielded 1,784,855 reads in total for all samples (range:1221–76,111; median: 15,595) that clustered into 8,726 OTUs after quality control, chimera filtering, as well as removal of chloroplast and mitochondrial reads. The fungal dataset resulted in 1,524,782 reads (range: 512–106,099; median: 7,146) that clustered into 429 OTUs. For microbial read distribution and rarefaction curves, see Supplementary Figure S1B-E).

Further microbiome analyses were primarily focused on the host-primed (S1) and non-primed (S2) setups in G3 and the FOM-primed and host-primed setups of G5 (Fig. 1A).

Effects of host-primed BCs on root microbiomes (G3)

We compared the S1 and S2 setups in G3 and tested whether host-primed BCs would affect root microbiome structures. Our analysis revealed a higher relative abundance of the bacterial phylum Bacteroidetes in S1, and more specifically the genera Cytophaga and Pedobacter belonging to Bacteroidetes were strongly enriched (Supplementary Figure S2A,B), while a higher abundance of Proteobacteria was observed in S2 (Fig. 2A). After removal of fOTU1 from the analysis (representing the inoculated FOM), it could be observed that Trichoderma, Apiotrichum and Zepfiella were distinctively enriched in S1 (Supplementary Figure S2C).

To further determine microbial OTUs that were significantly sensitive to the host-priming, we performed biomarker and differential abundance analysis. A linear discriminant analysis showed the highest number of biomarker taxa in S1 (Fig. 2B). Alpha-Gamma Proteobacteria including the families Caulobacteraceae and Azospirillaceae (genus Azospirillum) and Gemmatimonadaceae were strongly enriched in S1 while the family Streptomycetaceae was enriched in S2 (Fig. 2B). In addition, the bacterial genera Flavobacterium, Azospirillum, Massilia, and Thermomonas were highly enriched in S1 (Tables S3). The fungal order Helotiales was highly enriched in S1 while the family Bulleribasidiaceae, and Gibberella tricincta (OTU221), strongly characterized S2 (supplementary Figure S2D, Tables S3).

A significantly higher bacterial richness was observed in S1 (Fig. 2E), and the bacterial Shannon diversity was also highest in S1, although not significant. Beta diversity analysis and PCoA visualization revealed a clear clustering of the bacterial communities according to the setups (Fig. 2D), and a PERMANOVA analysis indicated that host-priming effects (Adonis, R² = 0.27, p < 0.001, Table 1) were stronger than genotype effects (Adonis, R² = 0.16, p < 0.001, Table 1). Fungal alpha diversity was not significantly affected by host-priming, whereas community composition, although less distinct when visualized on PCoA plots (Supplementary Figure S2E,F), was significantly affected. Host-priming (Adonis, R² = 0.14, P < 0.001, Table S1) and genotype x host-priming (Adonis, R² = 0.16, P < 0.001, Table S1) contributed most to the variation in the fungal communities compared to the genotype effects (Adonis, R² = 0.12, P < 0.01, Table S1). Expectedly as we transferred BCs, the effects of host-priming were higher on the bacterial communities than on the fungal communities (Table S2).

Table 1

Permutational analysis of variance (PERMANOVA) using the ‘adonis’ test on Bray-Curtis distance matrices for bacterial community dissimilarity assessment using 1000 permutations.
Dataset	Factors	Bacteria (R²)
G3	Genotype	0.16*
	Host-primed BC	0.27 ***
	Genotype* Host-primed BC	---
G5	Genotype	0.10 **
	FOM-primed BC	0.14 ***
	Genotype: FOM-primed BC	0.08 *
Significance of test indicated as * for p < 0.001, p > 0.01, *p < 0.05 and R² for proportion of variation explained. n.s denotes non-significant.

Next, we used co-occurrence network analysis to pinpoint the effects of host-priming on the microbial community structures. Both bacterial and fungal networks associated with the roots that had been host-primed were denser with higher clustering coefficients, mean node degrees and shorter average path lengths (Fig. 2E, F). Transitivity (Tr) or clustering coefficient is a measure of the tendency of the nodes to cluster together, while average path length is the average number of steps along the shortest paths between each node, being a measure of efficiency on a network [56–58]. Similarly, the total number of both positive and negative microbial co-occurrences were higher in host-primed networks (bacteria, 207; fungi, 448) than in non-primed networks (bacteria, 117; fungi, 267) (Table S4).

Figure 2. Host- primed bacterial communities (BC) alters and enhances microbial community networks against Fusarium invasion. A) Bacterial class level relative abundances in the S1 and S2 setups. B) Bacterial biomarkers between S1 and S2 using linear discriminant analysis (LDA). C) Bacterial alpha diversity (observed OTU richness and Shannon diversity) in S1 and S2. D) Bacterial beta diversity visualization using PCoA ordinations of bacterial communities in the different setups of generation 3 (G3), based on Bray-Curtis distance metrics. Compared setups of interest are shown in rectangle. E) Bacterial and F) fungal co-occurrence networks in G3 of S1 and S2. Positive and negative correlations are shown as grey and red edges, respectively. Bacterial and fungal nodes are represented with circle and square symbols in the network, respectively. Network metrics transitivity (Tr) or clustering coefficient, mean node degree (Nm) and average path length (Apl) were computed for each network. Each node represents the OTUs assigned to class (bacteria) or family (fungi), and the size of each node is proportional to the number of connections (degrees). Bacterial and fungal nodes are represented as circle and square symbols in the networks, respectively. Networks were constructed and analyzed using the SpiecEasi and Igraph packages in R.

Effects of FOM-primed BC on root microbiomes (G5)

For an analysis of the effects of FOM-primed BCs on the root microbiome structures, we compared the FOM-primed S2 and host-primed S3 setups in G5. Plants that received FOM-primed BCs had distinct microbial communities compared to those receiving host-primed BCs. Relative abundances of Bacteroidetes were higher in the FOM-primed roots while Actinobacteria abundance was higher in the roots that received host-primed BCs (Fig. 3A). By excluding the inoculated fOTU1 (FOM) from the analysis, it became obvious that Olpidium abundance increased in the FOM-primed roots compared with host-primed roots (Supplementary Figure S3 B, C).

The bacterial family Chthoniobacteraceae was the best marker for the FOM-primed root communities, while Micrococcaceae and Norcardioidaceae were highly present in the host-primed communities (Fig. 3B). The bacterial genus Flavobacterium (OTU112) was highly enriched in the FOM-primed setup (Table S3). The fungal genus Exophiala and family Phaeosphaeriaceae were the most enriched biomarker taxa in the FOM-primed while the families Didymellaceae and Ceratobasidiaceae were strongly enriched in the host-primed setups (Supplementary Figure S3D; Table S3).

Microbial alpha diversity was not significantly different between host- and FOM-primed roots (Fig. 3C). Bacterial beta diversity was significantly distinct between the root microbiomes that had received either host- or FOM-primed BCs as observed in the PCoA plots and PERMANOVA (Fig. 3D, Table 1). A significant but small effect was contributed by genotype (Adonis, R² = 0.07, p < 0.001, Table 1), BC priming (Adonis, R² = 0.06, p < 0.001, Table 1) and genotype x FOM-priming (Adonis, R² = 0.07, p < 0.001, Table 1) on the bacterial communities.

Fungal alpha diversity was not significantly different, and the communities of the different host-primed and FOM-primed setups did not form any notable clusters in PCOA plots (Supplementary Figure S3E,F. Table S1). Priming explained most of the variation observed in the fungal community (Adonis, setup; R² = 0.16 p < 0.001, Table S1).

Network analysis using the G5 dataset revealed distinct co-occurrences with highly destabilized bacterial networks in the FOM-primed roots exhibiting a lower clustering coefficient, a lower mean node degree and a shorter average path length compared to the host-primed networks (Fig. 3E,F). Unlike fungal co-occurrences, positive and negative bacterial co-occurrences were lower in the FOM-primed networks (bacteria, 97; fungi, 245) compared to the host-primed networks (bacteria, 157; fungi, 350) (Supplementary Table S4).

Genotype specific microbiome assembly in host and pathogen -primed BC conditions

Col-0 and Ler-0 exhibit different levels of resistance towards FOM [59, 60] and assemble distinct microbiomes [43, 61]. Thus, we examined whether there were any differences in how the microbiomes of these two genotypes responded to inoculation with host- or FOM-primed BC combined with their response to FOM infection. Data from the individual setups (G3-G5) were pooled for each genotype in this analysis. By examining network metrics, we found that in S1 and S2, bacterial networks were mostly stronger in Col-0 while fungal networks were more robust in Ler-0 (Fig. 4A,B; Supplementary Table S5). The most remarkable network breakdowns in both Col-0 and Ler-0 occurred in the S2 setups that had been challenged with FOM in the first generations (Fig. 4A,B). In contrast, bacterial networks in Ler-0 were stronger than in Col-0 in S3.

We identified distinct microbial biomarker taxa discriminating Col-0 and Ler-0 in both FOM- and host-primed setups (Supplementary Figure S5A,B). In S1, dominant biomarkers included the bacterial families Micrococcaceae, Sphingomonadaceae, Gemmatimonadaceae and the fungal taxon Lasiosphaeriaceae in Col-0, while Chthoniobacteraceae and the fungal taxon Typhulaceae were highly enriched in Ler-0. In S2, the most highly discriminating microbial taxa were Streptomycetaceae and Phaeosphaeriaceae in Col-0, while Pedosphaeraceae and Herpotrichiellaceae were strongly enriched in Ler-0. In S3, the bacterial family Cellvibrionaceae and the fungal class Agaricales were the most highly differentiating taxa in Col-0 while the fungal families Nocardioidaceae and Hypocreaceae dominated in Ler-0 (Supplementary Figure S5A,B).

Figure 4. Microbial co-occurrence networks of A) Col-0 and B) Ler-0 samples in the different setups. Bacterial and fungal nodes represent taxa affiliated at phylum and order levels, respectively. Bacterial and fungal datasets in G3, G4 and G5 were pooled for these analysis. Size of node is proportional to the number of connections (that is, degree). Bacterial and fungal nodes are represented as circle and square symbols in the networks, respectively. Network metrics transitivity (Tr) or clustering coefficient, mean node degree (Nm) and average path length (Apl) were computed for each network. Networks were analyzed using the SpiecEasi and visualized using the ggnet package in R.

Host-associated microbiomes could be modulated by either the host or by pathogens and this could subsequently impact host resilience towards pathogens. Here, we define microbiome priming as the conditioning of the host-associated microbiota by for instance several cycles of growth in association with a specific host or by exposure to a certain pathogen. We established an Arabidopsis-FOM pathosystem to study the effects of host- versus FOM-primed BCs on root microbiota robustness and pathogen infection success. Although disease symptoms were hardly visible in the shoots of the Arabidopsis ecotypes, qPCR analysis revealed significantly higher FOM quantities in shoots that were FOM inoculated (S1 and S2) compared to non-inoculated controls (S3), thus confirming a successful FOM infection of Arabidopsis shoots through the roots. The significantly higher FOM levels in the shoots of the third generation (G3) compared to those of G5 could suggest a suppressive ability of the primed BCs on FOM.

Host- and pathogen-primed BCs distinctively altered root microbiomes and FOM infection

We explored the G3 dataset to examine the effect of host- primed BCs on FOM infection and microbial communities by comparing the S1 and S2 setups. The BC inoculant that we used constitute a subcommunity of bacteria and fungi with a maximum sizes of 5 µm). Generally, using a pore size of 5 um, fungal cells will be excluded whereas the dominant Arabidopsis-associated bacterial taxa [62] will be included from the community [63].

Our result show that bacterial community compositions in G3 were distinct in the roots receiving the host-primed and non-primed BCs, demonstrating an effect of host-primed BCs on the root-associated microbiota. These observed differences in microbial abundance and composition of the individual setups could explain the differences observed in FOM suppression, which were highest in the Arabidopsis shoots from the host-primed BC receiving samples. For instance, Bacteroidetes (genera Cytophaga and Pedobacter) were strongly enriched in the host-primed BC treated roots and could thus potentially contribute to the FOM suppression observed in the plants receiving these BCs. This result supports our hypothesis of enrichment of taxa with potential inhibitory effects on FOM. Bacteroidetes is a core microbial taxon in the plant-associated microbiome and are highly enriched in some plants [64]. Several microbial species belonging to Bacteroidetes, for example Flavobacterium, Chryseobacterium and Pedobacter have been reported to inhibit microbial pathogens [29, 65, 66]. Flavobacterium species that were inoculated to cucumber seedlings suppressed Fusarium [67]. Likewise, a microbial mixture including Pedobacter was found to suppress Fusarium via secreted antibiotics [68].

We found several microbial biomarkers, mostly members of Alpha-Gamma proteobacteria that were differentially abundant between the host- and non-primed BC treated root microbiota. The discriminating bacterial biomarkers in S1 included the families Caulobacteraceae, Gemmatimonadaceae and Azospirillaceae (genus Azospirillum). Several species within these taxa are known for their beneficial functions and are used as biocontrol agents. For instance, Azospirillum was reported to inhibit F. oxysporum [69] and to promote plant growth via nitrogen fixation and synthesis of compounds required for both biotic and abiotic stress tolerance [70]. In another study, an increase in the relative abundance of Caulobacteraceae was observed in soils suppressive towards Rhizoctonia solani [71]. The increased abundance of Streptomycetaceae in the non-primed BC treated roots, especially in the resistant genotype, support previous finding of Streptomyces suppression of Fusarium species [72]. This also shows that not only BC treatment attracted suppressive taxa but also the Col-0 genotype itself.

The significantly higher bacterial alpha diversity observed in the host-primed BC treated roots suggest a strong microbial enrichment, unequivocally, caused by the host- primed BCs in the S1 setup. A higher species richness implies enhanced community robustness and functions [73]. Moreover, the high contribution of host-priming to explain the variance (25%) in the bacterial communities in G3 confirms its potential in structuring bacterial communities in Arabidopsis roots. Together, these findings suggest that host-priming increases the abundance of beneficial microbial taxa able to inhibit FOM.

Disease suppression in soils has been linked to enrichment of specific members of microbial consortia [74, 75]. The FOM-primed BCs (S2, G5) altered microbial community composition with a relatively higher abundance of Bacteroidetes compared with host-primed roots. In contrast, we observed an increase in the abundance of the fungal genus Olpidium that are known to contain pathogenic members, for example, Olpidium brassicae predominantly attacking Brassica plants [76]. We suggest a plausible cooperative interaction, probably via nutritional inter-dependence between FOM and the genus Olpidium, but this requires further studies.

In addition, the most highly enriched bacterial family Chthoniobacteraceae observed in the FOM-primed BC treated root samples was earlier reported to be more abundant in healthy white pine microbiomes, indicative of their protective role in plant health [77]. Similarly, both Micrococcaceae and Norcardioidaceae were identified as some of the highly enriched taxa in the host-primed BC inoculated roots (S3 of G5), showing taxa enrichment also in the healthy host. Altogether, these results support the distinct modulating effects of both host and pathogen primed BCs on the root-associated microbiomes.

Host- primed BC enhances robustness of microbial network structures

Ecological network modeling and visualization enable us to gain insights into microbial interactions [57, 78, 79]. We sought to infer how host- and FOM- primed BC inoculation affected microbial co-occurrence networks. Interestingly, we found that the setups inoculated with the host-primed BC had the strongest microbial networks in comparison with the FOM-primed networks. The observed higher transitivity or clustering coefficient shows that microbial community in the G3S1 network is more densely connected. Also, the smaller average path length depicts a compact microbial community structure [79, 80]. Thus, we infer from our results that treatment with host-primed BCs enhances microbial network robustness and could further explain the observed suppression of FOM. Similarly, Qu et al., [11] found that bio-priming of hypovirulent strain DT-8 of S. sclerotiorum affected individual microbial species and further enhanced fungal and bacterial community network resilience.

Host traits determine microbiome assembly after host- or pathogen- primed BC inoculation.

The differences observed in Col-0 and Ler-0 such as root morphology [32, 33] chemical profiles [34] and disease resistances genes [35] affect microbial community assembly [43]. The observed distinct and highly dissimilar microbial networks in the Col-0 and Ler-0 samples in the different setups suggests an underlying role of host resistance in microbial community structuring [15, 43]. Specifically, in S1 and S2 we found stronger bacterial networks in Col-0 and stronger fungal networks in Ler-0, suggestive of an unequivocal host genotype effect on microbiome assembly. Moreover, the strong network breakdown observed in Col-0 and Ler-0 in the S2 setup is indicative of a strong host-priming effect leading towards resilient microbial communities capable of resisting FOM invasion.

Additional analysis revealed distinct biomarkers of both Col-0 and Ler-0 in the individual setups, confirming that primed BC inoculation influences Arabidopsis microbiomes depending on FOM host resistance [60]. Biomarkers of Col-0 included the bacterial families Micrococcaceae, Sphingomonadaceae and Gemmatimonadaceae in S1 and the family Streptomycetaceae in S2. Micrococcaceae, Gemmatimonadaceae and Streptomycetaceae were reported among the top taxa responsible for Fusarium wilt suppressiveness [81]. Sphingomonadaceae were significantly more abundant in the rhizosphere upon fungal invasion [71] and were also reported to be enriched in leaves of citrus infected by Diaporthe citri, the fungal pathogen causing melanose disease [82]. Distinct fungal biomarkers were also identified in Col-0 (order Sordariales in S1 and Phaeosphaeriaceae in S2) and Ler-0 (order Agaricales in S1 and Herpotrichiellaceae in S2). Studies have previously demonstrated biocontrol activity of Agaricales [83, 84], and Trichoderma [85] which were respectively, enriched in Col-0 and Ler-0 samples in the S3 control setup.

By using a robust experimental setup we have demonstrated the effects of host- and FOM-primed BCs on Fusarium pathogen progression and microbial community structures in Arabidopsis roots. We observed inhibitory effects of the host- primed BC on FOM infection. Both host and FOM-primed BCs distinctively altered microbial communities. Notably, beneficial bacterial families including Flavobacterium, Pedobacter, Azospirillum and families Rhizobiaceae and Caulobacteriaceae were enriched in the host-primed BC setup and could thus possibly be involved in F. oxysporum inhibition. The FOM-primed setup had increased abundance of the microbial classes Bacteroidetes and Exophiala with the families Chthoniobacteriaceae and Phaeosphaeriaceae as major biomarkers. Networks of the host-primed BC primed setup displayed more robustness and resilience compared to the FOM-primed setup, indicative of an ability of host-primed BCs to enhance disease resistance, as also suggested previously [11, 19]. Moreover, the observed distinct microbial networks of Col-0 and Ler-0 in all setups strongly suggests a role of microbial communities in the resistance observed in Col-0. In summary, this study demonstrates both microbiome-modulating and pathogen-suppressive abilities of priming host-associated microbial communities. Likewise, the study suggests a role of root-associated microbiomes in the resistance of Col-0 against FOM. These results could in the future be exploited to engineer sustainable disease suppressive growth systems.

BC Bacterial consortia

Col-0 Columbia - 0

Ler-0 Landsberg erecta

FOM Fusarium oxysporum f.sp. mathioli

ITS2 Internal transcribed spacer – 2

16S rRNA 16S ribosomal RiboNucleic Acid

RFO Resistance to Fusarium oxysporum (RFO)

OTU Operational taxonomic unit

bOTU Bacterial operational taxonomic unit

fOTU Fungal operational taxonomic unit

qPCR quantitative PCR

PERMANOVA Permutational multivariate analysis of variance

NCBI National Center for Biotechnology Information

SRA Sequence Read Archive

QIIME Quantitative Insights Into Microbial Ecology

Ethical approval and consent to participate

Not applicable

Consent for publication

Not applicable

Availability of data and materials

The MiSeq paired end reads for bacterial 16s rRNA gene (V3-V4) and fungal ITS2 regions have been deposited in NCBI SRA database using accession code PRJNA850631. Sequence processing in QIIME and data analysis performed using the pipeline described in Kudjordjie et al., [38].

Competing interests

The authors declare that they have no competing interests.

Funding

This work was financially support by Aarhus University (project number 22550) and the Independent Research Fund Denmark DFF grant number 23844.

Authors' contributions

ENK, RS, MN conceived the study, participated in its design. ENK conducted the experiment and analyzed the data. ENK and MN drafted the manuscript. All authors read, revised and approved the manuscript.

Acknowledgements

We thank H. Corby Kistler for providing the FOM isolate. We thank Simone Ena Rasmussen and Mathilde Schiøtt Dige for their excellent laboratory assistance.

Authors' information

Not applicable

Nihorimbere V, Ongena M, Smargiassi M, Thonart P. Beneficial effect of the rhizosphere microbial community for plant growth and health. Biotechnol Agron Soc Environ. 2011;15:327–37.
Hayat R, Ali S, Amara U. Soil beneficial bacteria and their role in plant growth promotion: a review. Ann Microbiol. 2010;579–98.
Mendes R, Kruijt M, de Bruijn I, Dekkers E, van der Voort M, Schneider JHM, et al. Deciphering the rhizosphere microbiome for disease-suppressive bacteria. Science (80). 2011;332:1097–100.
Turner TR, James EK, Poole PS. The plant microbiome. Genome Biol. 2013;14:209.
Jones DL, Nguyen C, Finlay RD. Carbon flow in the rhizosphere : carbon trading at the soil – root interface. Plant Soil. 2009;5–33.
Venturi V, Keel C. Signaling in the rhizosphere. Trends Plant Sci. 2016;21:187–98.
Agrios GN. Plant pathology. 5th editio. Elsevier Academic Press, Amsterdam; 2005.
Banerjee S, Walder F, Büchi L, Meyer M, Held AY, Gattinger A, et al. Agricultural intensification reduces microbial network complexity and the abundance of keystone taxa in roots. ISME J. 2019;13:1722–36.
Hartman K, Heijden MGA Van Der, Wittwer RA, Banerjee S, Walser J, Schlaeppi K. Cropping practices manipulate abundance patterns of root and soil microbiome members paving the way to smart farming. Microbiome; 2018;1–14.
van der Heijden MGA, Hartmann M. Networking in the plant microbiome. PLoS Biol. 2016;14:1–9.
Qu Z, Zhao H, Zhang H, Wang Q, Yao Y, Cheng J, et al. Bio-priming with a hypovirulent phytopathogenic fungus enhances the connection and strength of microbial interaction network in rapeseed. npj Biofilms Microbiomes. 2020;6:1–13.
Quiza L, St-Arnaud M, Yergeau É. Harnessing phytomicrobiome signaling for rhizosphere microbiome engineering. Front Microbiol. 2015;6:1–11.
Kwak M-J, Kong HG, Choi K, Kwon S-K, Song JY, Lee J, et al. Rhizosphere microbiome structure alters to enable wilt resistance in tomato. Nat Biotechnol (36). 2018; 1100–1109 (2018).
Köberl M, Dita M, Martinuz A, Staver C, Berg G. Members of Gammaproteobacteria as indicator species of healthy banana plants on Fusarium wilt-infested fields in Central America. Sci Rep. 2017;7:1–9.
Yao H, Wu F. Soil microbial community structure in cucumber rhizosphere of different resistance cultivars to Fusarium wilt. FEMS Microbiol Ecol 72. 2010;72:456–63.
Berendsen RL, Vismans G, Yu K, Song Y, Jonge R De, Burgman WP, et al. Disease-induced assemblage of a plant-beneficial bacterial consortium. ISME J. 2018;12:1496–507.
Siegel-hertz K, Edel-hermann V, Chapelle E, Terrat S. Comparative microbiome analysis of a Fusarium wilt suppressive soil and a Fusarium wilt conducive soil from the Châteaurenard Region. Front. Microbiol. 2018;9:1–16.
Weller DM, Landa BB, Mavrodi O V., Schroeder KL, De La Fuente L, Blouin Bankhead S, et al. Role of 2,4-diacetylphloroglucinol-producing fluorescent Pseudomonas spp. in the defense of plant roots. Plant Biol. 2007;9:4–20.
Tao C, Li R, Xiong W, Shen Z, Liu S, Wang B, et al. Bio-organic fertilizers stimulate indigenous soil Pseudomonas populations to enhance plant disease suppression. Microbiome. 2020;8:1–14.
Liu Y, Zhu A, Tan H, Cao L, Zhang R. Engineering banana endosphere microbiome to improve Fusarium wilt resistance in banana. Microbiome; 2019;1–15.
Busby PE, Soman C, Wagner MR, Friesen ML, Kremer J, Bennett A, et al. Research priorities for harnessing plant microbiomes in sustainable agriculture. PLoS Biol. 2017;
Garcia J, Kao-Kniffin J. Microbial group dynamics in plant rhizospheres and their implications on nutrient cycling. Front Microbiol. 2018;9:1–7.
Ryan PR, Dessaux Y, Thomashow LS, Weller DM. Rhizosphere engineering and management for sustainable agriculture. Plant Soil. 2009;321:363–83.
Berendsen RL, Pieterse CMJ, Bakker PAHM. The rhizosphere microbiome and plant health. Trends Plant Sci. 2012;17:478–86.
Mueller UG, Sachs JL. Engineering microbiomes to Improve plant and animal health. Trends Microbiol. 2015;23:606–17.
Panke-Buisse K, Poole AC, Goodrich JK, Ley RE, Kao-Kniffin J. Selection on soil microbiomes reveals reproducible impacts on plant function. ISME J. 2015;9:980–9.
Dudenhöffer JH, Scheu S, Jousset A. Systemic enrichment of antifungal traits in the rhizosphere microbiome after pathogen attack. J Ecol. 2016;104:1566–75.
Berg G, Kusstatscher P, Abdelfattah A, Cernava T, Smalla K. Microbiome modulation—toward a better understanding of plant microbiome response to microbial inoculants. Front Microbiol. 2021;12:1–12.
Jiang G, Zhang Y, Gan G, Li W, Wan W, Jiang Y, et al. Exploring rhizo-microbiome transplants as a tool for protective plant-microbiome manipulation. ISME Commun. 2022;2:1–10.
Jousset A, Lee S. Coming of age for the rhizosphere microbiome transplantation. Soil Ecol. Lett. 2022; https://doi.org/10.1007/s42832-022-0151-5.
Santhanam R, Luu VT, Weinhold A, Goldberg J, Oh Y, Baldwin IT. Native root-associated bacteria rescue a plant from a sudden-wilt disease that emerged during continuous cropping. Proc Natl Acad Sci U S A. 2015;112:E5013–120.
Cajero Sánchez W, García-Ponce B, Sánchez M de la P, Álvarez-Buylla ER, Garay-Arroyo A. Identifying the transition to the maturation zone in three ecotypes of Arabidopsis thaliana roots. Commun Integr Biol. 2017;11:1–7.
Price CG, Knee EM, Miller JA, Shin Di, Mann J, Crist DK, et al. Following phenotypes: an exploration of Mendelian genetics using Arabidopsis plants. Am Biol Teach. 2018;80:291–300.
Mönchgesang S, Strehmel N, Schmidt S, Westphal L, Taruttis F, Müller E, et al. Natural variation of root exudates in Arabidopsis thaliana-linking metabolomic and genomic data. Sci Rep. 2016;6:29033.
Diener AC, Ausubel FM. Resistance to Fusarium oxysporum 1, a dominant Arabidopsis disease-resistance gene, is not race specific. Genetics. 2005;171:305–21.
Kistler HC. Relatedness of strains of Fusarium oxysporum from crucifers measured by examination of mitochondrial and ribosomal DNA. Phytopathology. 1987;77:1289.
Reiss A, Fomsgaard IS, Mathiassen SK, Kudsk P. Silicone tube microextraction for repeated sampling of benzoxazinoids in the root zone and analysis by HPLC /MS-MS. J Allelochem Interact. 2018;4:27–37.
Kudjordjie EN, Sapkota R, Steffensen SK, Fomsgaard IS, Nicolaisen M. Maize synthesized benzoxazinoids affect the host associated microbiome. Microbiome; 2019;1–17.
Klindworth A, Pruesse E, Schweer T, Peplies J, Quast C, Horn M, et al. Evaluation of general 16S ribosomal RNA gene PCR primers for classical and next-generation sequencing-based diversity studies. Nucleic Acids Res. 2013;41:e1.
Ihrmark K, Bödeker ITM, Cruz-Martinez K, Friberg H, Kubartova A, Schenck J, et al. New primers to amplify the fungal ITS2 region–evaluation by 454-sequencing of artificial and natural communities. FEMS Microbiol Ecol. 2012;82:666–77.
Siddique AB, Unterseher M. A cost-effective and efficient strategy for Illumina sequencing of fungal communities: A case study of beech endophytes identified elevation as main explanatory factor for diversity and community composition. Fungal Ecol. 2016;20:175–85.
Wu L, Wen C, Qin Y, Yin H, Tu Q, Van Nostrand JD, et al. Phasing amplicon sequencing on Illumina Miseq for robust environmental microbial community analysis. BMC Microbiol. 2015;15:1–12.
Kudjordjie EN, Hooshmand K, Sapkota R, Darbani B, Fomsgaard IS. Fusarium oxysporum disrupts microbiome-metabolome. Microbiol. Spectr. 2022; 10.1128/spectrum.01226-22.
Caporaso JG, Kuczynski J, Stombaugh J, Bittinger K, Bushman FD, Costello EK, et al. QIIME allows analysis of high-throughput community sequencing data. Nat Methods. 2010;7:335–6.
Rognes T, Flouri T, Nichols B, Quince C, Mahé F. VSEARCH: a versatile open source tool for metagenomics. PeerJ. 2016;4:e2584.
Bengtsson-Palme J, Ryberg M, Hartmann M, Branco S, Wang Z, Godhe A, et al. Improved software detection and extraction of ITS1 and ITS2 from ribosomal ITS sequences of fungi and other eukaryotes for analysis of environmental sequencing data. Methods Ecol Evol. 2013;914–9.
Quast C, Pruesse E, Yilmaz P, Gerken J, Schweer T, Yarza P, et al. The SILVA ribosomal RNA gene database project: Improved data processing and web-based tools. Nucleic Acids Res. 2013;41:590–6.
Nilsson RH, Larsson KH, Taylor AFS, Bengtsson-Palme J, Jeppesen TS, Schigel D, et al. The UNITE database for molecular identification of fungi: Handling dark taxa and parallel taxonomic classifications. Nucleic Acids Res. 2019;47:D259–64.
R Core Team. R : A language and environment for statistical computing. 2017.
Oksanen AJ, Blanchet FG, Friendly M, Kindt R, Legendre P, Mcglinn D, et al. Package ‘vegan’ 2021.
McMurdie PJ, Holmes S, Kindt R, Legendre P, O’Hara R. phyloseq: An R package for reproducible interactive analysis and graphics of microbiome census data. PLoS One. 2013;8:e61217.
Wickham H. ggplot2: Elegant graphics for data analysis. Springer-Verlag New York Media. 2009. Available from: http://link.springer.com/10.1007/978-0-387-98141-3.
Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15:1–21.
Csardi G, Nepusz T. The igraph software package for complex network research. Inter Journal Complex Syst. 2006; http://igraph.sf.net.
Kurtz ZD, Müller CL, Miraldi ER, Littman DR, Blaser MJ, Bonneau RA. Sparse and compositionally robust inference of microbial ecological networks. PLoS Comput Biol. 2015;11:1–25.
Zhou J, Deng Y, Luo F, He Z, Tu Q, Zhi X. Functional molecular ecological networks. MBio. 2010;1:1–10.
Berry D, Widder S. Deciphering microbial interactions and detecting keystone species with co-occurrence networks. Front Microbiol. 2014;5:1–14.
Dunne JA, Williams RJ, Martinez ND. Food-web structure and network theory: The role of connectance and size. Proc Natl Acad Sci U S A. 2002;99:12917–22.
Shen Y, Diener AC. Arabidopsis thaliana resistance to Fusarium oxysporum 2 implicates tyrosine-sulfated peptide signaling in susceptibility and resistance to root infection. PLoS Genet. 2013;9.
Diener AC, Ausubel FM. Resistance to Fusarium oxysporum 1, a dominant Arabidopsis disease-resistance gene, is not race specific. Genetics. 2005;171:305–21.
Kudjordjie EN, Sapkota R, Nicolaisen M. Arabidopsis assemble distinct root-associated microbiomes through the synthesis of an array of defense metabolites. PLoS One. 2021;16:1–19.
Lundberg DS, Lebeis SL, Paredes SH, Yourstone S, Gehring J, Malfatti S, et al. Defining the core Arabidopsis thaliana root microbiome. Nature. 2012;488:86–90.
Luan L, Jiang Y, Cheng M, Dini-Andreote F, Sui Y, Xu Q, et al. Organism body size structures the soil microbial and nematode community assembly at a continental and global scale. Nat Commun. 2020;11:1–11.
Pérez-Jaramillo JE, Carrión VJ, de Hollander M, Raaijmakers JM. The wild side of plant microbiomes. Microbiome. 2018;6:4–9.
Inderbitzin P, Ward J, Barbella A, Solares N, Izyumin D, Burman P, et al. Soil microbiomes associated with Verticillium wilt-suppressive Broccoli and chitin amendments are enriched with potential biocontrol agents. Phytopathology. 2018;108:31–43.
Yin C, Hulbert SH, Schroeder KL, Mavrodi O, Mavrodi D, Dhingra A, et al. Role of bacterial communities in the natural suppression of Rhizoctonia solani bare patch disease of wheat (Triticum aestivum L.). Appl Environ Microbiol. 2013;79:7428–38.
Nishioka T, Marian M, Kobayashi I, Kobayashi Y, Yamamoto K, Hideyuki tamaki3 H suga4 & M shimizu1. Microbial basis of Fusarium wilt suppression by Allium cultivation. Sci Rep. 2019;9:1–9.
De Boer W, Wagenaar AM, Klein Gunnewiek PJA, Van Veen JA. In vitro suppression of fungi caused by combinations of apparently non-antagonistic soil bacteria. FEMS Microbiol Ecol. 2007;59:177–85.
Somers E, Ptacek D, Gysegom P, Srinivasan M, Vanderleyden J. Azospirillum brasilense produces the auxin-like phenylacetic acid by using the key enzyme for indole-3-acetic acid biosynthesis. Appl Environ Microbiol. 2005;71:1803–10.
Hungria M, Campo RJ, Souza EM, Pedrosa FO. Inoculation with selected strains of Azospirillum brasilense and A. lipoferum improves yields of maize and wheat in Brazil. Plant Soil. 2010;331:413–25.
Chapelle E, Mendes R, Bakker PAH, Raaijmakers JM. Fungal invasion of the rhizosphere microbiome. ISME J. 2016;
Gopalakrishnan S, Srinivas V, Vidya MS, Rathore A. Plant growth-promoting activities of Streptomyces spp. in sorghum and rice. Springerplus. 2013;2:1–8.
Wagg C, Schlaeppi K, Banerjee S, Kuramae EE, van der Heijden MGA. Fungal-bacterial diversity and microbiome complexity predict ecosystem functioning. Nat Commun. 2019;10:1–10.
Mendes R, Kruijt M, Bruijn I De, Dekkers E, Der M Van, Schneider JHM, et al. Deciphering the rhizosphere microbiome for disease-suppressive bacteria. Science. 2011;332:1097–100.
Raaijmakers JM, Paulitz TC, Steinberg C, Alabouvette C, Moënne-Loccoz Y. The rhizosphere: a playground and battlefield for soilborne pathogens and beneficial microorganisms. Plant Soil. 2009;321:341–61.
Gkarmiri K, Mahmood S, Ekblad A, Alström S, Högberg N, Finlay R. Identifying the active microbiome associated with roots and rhizosphere soil of oilseed rape. Appl Environ Microbiol. 2017;83.
Lalande BM, Hanna JW, Warwell M, Kim M, Klopfenstein NB, Stewart JE. Assessing soil microbial communities associated with tree health in a western white pine (Pinus monticola). Proceedings of the 66th Western International Forest Disease Work Conference. 2019;27–30.
Banerjee S, Baah-Acheamfour M, Carlyle CN, Bissett A, Richardson AE, Siddique T, et al. Determinants of bacterial communities in Canadian agroforestry systems. Environ Microbiol. 2016;18:1805–16.
Faust K, Raes J. Microbial interactions: From networks to models. Nat Rev Microbiol. 2012;10:538–50.
Kuntal BK, Chandrakar P, Sadhu S, Mande SS. ‘NetShift’: a methodology for understanding ‘driver microbes’ from healthy and disease microbiome datasets. ISME J. 2019;13:442–54.
Cha JY, Han S, Hong HJ, Cho H, Kim D, Kwon Y, et al. Microbial and biochemical basis of a Fusarium wilt-suppressive soil. ISME J. 2015;10:119–29.
Li PD, Zhu ZR, Zhang Y, Xu J, Wang H, Wang Z, et al. The phyllosphere microbiome shifts toward combating melanose pathogen. Microbiome. BioMed Central; 2022;10:56.
Anke T, Besl H, Mocek U, Steglich W. Antibiotics from basidiomycetes. xviii) strobilurin c and oudemansin b, two new antifungal metabolites from xerula species (agaricales). J Antibiot. 1983;36:661–6.
Alves M, Ferreira IFR, Dias J, Teixeira V, Martins A, Pintado M. A review on antimicrobial activity of mushroom (basidiomycetes) extracts and isolated compounds. Planta Med. 2012;78:1707–18.
Ferre FS, Santamarina MP. Efficacy of trichoderma harzianum in suppression of Fusarium culmorum. Ann Microbiol. 2010;60:335–40.

No competing interests reported.

SupplemetaryFile.pdf
Supplementary Figures Figure S1: A) The steps involved in harvesting bacteria microbiomes from the natural soils of growing plants. Roots of i) 6 weeks old Arabidopsis were harvested and ii) pooled into a sterile 50 ml Falcon tubes, iii and iv) centrifuged and v) filtered using 5 um filters. Purified and non-purified bacterial cultures are shown in iv and vi, respectively. The vi) filtrate was used to inoculate vii) 10 days old succeeding Arabidopsis seedlings. Distribution of (B) bacterial and (C) fungal sequence reads and OTUs in samples used in this study. Rarefaction curves showing the coverage of (D) bacterial and (E) fungal OTU richness (species richness in number of OTUs) as a function of sequencing depth (sample size in number of reads). Figure S2. A) Relative abundance of Bacteroidetes top genera between S1 and S2 setups in G3. B) Relative abundances of fungal taxa at genus level in samples from experimental setups in G3. fOTU1 assigned to the inoculated Fusarium, FOM was unsurprisingly highly abundant in S1 and S2. C) Fungal class level relative abundances after removal of fOTU1. D) Fungal biomarkers using linear discriminant analysis (LDA) between host-primed BC (S1) and non-primed (S2) samples in G3 setup. E) Alpha diversity (observed OTU richness and Shannon diversity) and F) beta diversity visualization using PCoA ordinations of bacterial communities in setups of G3, based on Bray-Curtis distance metrics. Compared setups of interest in G3 are shown in rectangle. Figure S3. A) Relative abundance of Bacteroidetes top genera between S2 and S3 setups in G5. B) Relative abundances of fungal taxa at genus level in samples from experimental setups in G5. fOTU1 assigned to the inoculated Fusarium, FOM was unsurprisingly highly abundant in S1 and S2. C) Fungal class level relative abundances after removal of fOTU1. D) Fungal biomarkers using linear discriminant analysis (LDA) between FOM-primed BC (S2) and non-primed (S3) samples in G5setup. E) Alpha diversity (observed OTU richness and Shannon diversity) and F) beta diversity visualization using PCoA ordinations of bacterial communities in setups of G5, based on Bray-Curtis distance metrics. Compared setups of interest in G5 are shown in rectangle. Figure S4: Beta diversity visualization using PCoA ordinations of A) Bacterial and B) fungal community communities in setups of Fusarium inoculated (Foxy) and non-inoculated samples of G3, G4 and G5, based on Bray-Curtis distance metrics. Figure S5. Identification microbial biomarkers in genotypes. Linear discriminant analysis (LDA) to search for A) bacterial and B) fungal biomarkers in Col-0 and Ler-0 root samples in S1, S2 and S3. Supplementary Tables Table S1. permutational analysis of variance (PERMANOVA) using the ‘adonis’ test on Bray-Curtis distance matrices for fungal community dissimilarity assessment using 1000 permutations.Table S2: Pairwise comparisons of microbial community variations among different setups. Table S4: Number of positive and negative co-occurrences in microbial networks of setups.Table S5: Number of positive and negative co-occurrences in microbial networks of setups in genotype co-occurrences networks.
SuppleTableS3DA.xlsx
Table S3: Differential analysis between setups performed using DESeq2 with phyloseq. Differences were considered significant with the p-value (adjusted for false discovery rate using Benjamini–Hochberg correction) cut-off of 0.01, except for fOTUs differential abundance estimation in Col-0 and Ler-0 comparison where a cut-off of 0.05 was used.

Download PDF

Version 1

posted

You are reading this latest preprint version

Host or pathogen primed bacterial consortia alters microbial community structures in Arabidopsis roots and suppresses Fusarium oxysporum

Status:

Version 1

Abstract

Figures

Introduction

Materials And Methods

Plant materials and fungal strains

Experimental design

Purified bacteria consortia preparation and application

FOM culture preparation, quantification, and inoculation

DNA extraction

Library preparation and amplicon generation

Pathogen quantification

Sequence data analysis

Statistical analysis

Biomarker and differential OTUs analysis

Cooccurrence networks

Results

Sequence data

Effects of host-primed BCs on root microbiomes (G3)

Effects of FOM-primed BC on root microbiomes (G5)

Genotype specific microbiome assembly in host and pathogen -primed BC conditions

Discussion

Host- and pathogen-primed BCs distinctively altered root microbiomes and FOM infection

Host- primed BC enhances robustness of microbial network structures

Conclusions

Abbreviations

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1