Addition of plant-growth-promoting Bacillus subtilis PTS-394 on tomato rhizosphere has no durable impact on composition of root microbiome

Representatives of the genus Bacillus are increasingly used in agriculture to promote plant growth and to protect against plant pathogens. Unfortunately, hitherto the impact of Bacillus inoculants on the indigenous plant microbiota has been investigated exclusively for the species Bacillus amyloliquefaciens and was limited to prokaryotes, whilst eukaryotic member of this community, e.g. fungi, were not considered. The root-colonizing Bacillus subtilis PTS-394 supported growth of tomato plants and suppressed soil-borne diseases. Roche 454 pyrosequencing revealed that PTS-394 has only a transient impact on the microbiota community of the tomato rhizosphere. The impact on eukaryota could last up to 14 days, while that on bacterial communities lasted for 3 days only. Ecological adaptation and microbial community-preserving capacity are important criteria when assessing suitability of bio-inoculants for commercial development. As shown here, B. subtilis PTS-394 is acting as an environmentally compatible plant protective agent without permanent effects on rhizosphere microbial community.


Background
Plant growth-promoting rhizobacteria (PGPR) have potential as biocontrol agents that could replace chemical pesticides thereby reducing undesired chemical remnants in agriculture. Plant growth promotion, suppression of plant diseases and rhizosphere competence (colonization and survival on plant roots) by PGPR, especially bacilli, have been considered as critical requisites for the development of commercial products [1,2]. To be effective, PGPRs must establish and maintain a sufficient population in the rhizosphere [3]. It is reported that the robust colonization of plant roots by beneficial microbes directly contributes to the effective biocontrol of soil-borne pathogens [4]. Root colonization and rhizosphere competence is thus a critical prerequisite for the successful use of PGPRs as biocontrol and plant growth-promoting agents [5,6].
The rhizosphere is rich in nutrients due to the accumulation of plant exudates containing amino acids and sugars that provides a rich source of energy and nutrients for colonizing bacteria [7]. This forces the bacterial community to colonize the rhizoplane and rhizosphere [8]. The colonizing capability of PGPR strains has been investigated using GFP labeling and transposon-mediated random mutagenesis approaches, in Pseudomonas and Bacillus strains, in particular [9,10]. Bacterial traits such as chemotaxis, motility, attachment, growth, and stress resistance all appear to contribute to the colonization competence of PGPRs [8].
Rhizosphere competence is essentially a process of niche competition between PGPRs and other microbes present in the vicinity of plant roots, in which resource partitioning, competitive exclusion, and co-existence can all play a part. It is reported that PGPRs colonize more efficient in poorer microbial communities than in richer soils [11]. The indigenous rhizosphere microbial community can be influenced by large-scale application of PGPRs in field trials [12]. Pseudomonas sp. DSMZ 13134 affects the dominant bacterial community of barley roots in the tested soil system [13], for example. By contrast, as revealed by T-RFLP community fingerprinting, application of Bacillus amyloliquefaciens FZB42 did not change the composition of rhizosphere bacterial community in a measurable extent [2]. Similar results were also found for B. amyloliquefaciens BNM122 in soybean [14]. Recently, persistence and the effect of FZB42 on the microbial community of lettuce were more deeply analyzed by 454-amplicon sequencing corroborating that inoculation with B. amyloliquefaciens has no or only transient minor effects on the microbiota in vicinity of plant roots [15]. A slow decrease in the number of inoculated bacteria was also registered. After five weeks only 55% of the initial number of FZB42 DNA was still traceable within the rhizosphere of lettuce in the field [16]. Unfortunately, till now, the effect of Bacillus PGPR on the non-bacterial members of indigenous plant microbiota has not been analyzed. Moreover, no other representatives of the B. subtilis species complex than B. amyloliquefaciens, the main source for commercial biofertilizer and biocontrol agents, have been proven for their impact on plant rhizosphere.
Bacillus subtilis PTS-394, isolated from the rhizosphere of tomato and fully sequenced in 2014 [17], has been shown to suppress tomato soil-borne diseases caused by Fusarium oxysporum and Ralstonia solanacearum which is the main obstacle to continuous cropping of tomato in greenhouse cultivation [18]. In the present study, the colonization behavior and plant growth-promoting capability of PTS-394 were investigated in laboratory and greenhouse experiments. Additionally, the effect of PTS-394 on the rhizosphere microbial community was studied using Roche 454 pyrosequencing of 16S rRNA and ITS partial sequences. The aim was to improve our understanding of the ecological consequences of microbial inoculants and of rhizosphere microbiota ecology.

Strains and growth conditions
Bacillus subtilis PTS-394 and the GFP-tagged strain of B. subtilis PTS-394G containing the plasmid pGFP22 were isolated and constructed by our laboratory [19]. The Bacillus strains were grown in YPG medium containing 0.5% yeast extract, 0.5% peptone and 0.5% glucose. 5 μg/ ml of chloramphenicol was added when necessary for B. subtilis PTS-394G.

The software of Phylogenomic analysis
Phylogenomic analysis of strain PTS-394 was performed taking advantage of availability of the whole genome sequence [17]. JSpeciesWS (http://jspecies.ribohost.com/ jspeciesws/) was used to determine ANIb (average nucleotide identity based on BLAST+) and ANIm (average nucleotide identity based on MUMmer) values by pairwise genome comparisons. Correlation indexes of their Tetra-nucleotide signatures (TETRA) were determined by using the JSpeciesWS software (Richter et al., 2015). Electronic DNA-DNA hybridization (dDDH) is useful to mimic the wet-lab DDH and can be used for genomebased species delineation and genome-based subspecies delineation [20,21] and was applied to finally defined PTS-394 whether belong to B. subtilis subsp. subtilis. Three formulas are applicable for the calculation: Formula: 1 (HSP length / total length), formula: 2 (identities / HSP length) and formula 3 (identities / total length). Formula 2, which is especially appropriate to analyze draft genomes, was used.
The colonization and plant growth promotion of PTS-394 on agar plates B. subtilis PTS-394 was shaken at 180 rpm at 28°C for 36 h, centrifuged at 5000 rpm for 10 min, then the cell pellets were suspended in sterile distilled water and adjusted to an OD 600 of 0.01, 0.1, 1.0, 1.5, and 3.0 (OD 600 = 1.0,~5 × 10 7 CFU/mL). Tomato (tomato cultivar 'Moneymaker') seeds were surface-sterilized using sodium hypochlorite (3%, v/v) for 5 min, washed three times with sterile water and soaked in the above cell suspensions at 25°C for 24 h. Seeds were sown into individual 9 cm diameter MS agar plates [22] at a density of five seeds per plate. Each treatment was performed on 10 plants and was replicated three times. Plates were incubated in a growth chamber with a daytime temperature of 25°C and a night temperature of 18°C, 75% humidity and a 12:12 h light: dark cycle with a light intensity of 8000 lux. At 14 days after sowing, plant fresh weight and the amount of PTS-394 on the root surface were measured as previously described [23].

The colonization and plant growth promotion of PTS-394 in pots
The dynamic colonization of PTS-394 on the tomato rhizoplane and its plant growth-promoting ability were investigated by pot experiments, and the GFP-labeled strain (PTS-394G) was used for colonization analysis. Suspensions of B. subtilis PTS-394 and PTS-394G with an OD 600 = 1.0 were used. In both experiments, tomato seeds were surface-disinfected and sown into nursery soil containing a mixture of vermiculite and organic manure (1:1, w/w) for germination. Tomato seedlings at the 4-leaf stage were transplanted into pots filled with a mixture of vermiculite, rice field soil and organic manure (1:2:1, w/w). 20 mL of bacterial cell suspension was added to each plant and cultivation was continued in a greenhouse under natural conditions with temperatures ranging from 18 to 30°C. Each treatment was performed on 30 plants and replicated three times in a completely randomized block design. In the colonization experiment, tomato root samples were collected from three PTS-394G plants at 0, 1, 2, 3,5,7,9,11,14, and 21 days after transplantation and the amount of PTS-394 on the root surface was determined, as described previously. Meanwhile, tomato roots were examined at 1, 3, 7 and 14 days using fluorescent microscopy in order to detect colonization with PTS-394G. At 30 days after treatment with PTS-394, the fresh weight of the root and the plant height were measured in order to evaluate the growthpromoting effect of PTS-394.

Rhizosphere soil collection and DNA extraction
During the pot experiment described above, rhizosphere soil from untreated and PTS-394-treated pot-grown tomato plants was collected from three plants at 1, 3, 7, 9 and 14 days after treatment. Rhizosphere soil total DNA was extracted using the PowerSoil DNA Isolation Kit (MO BIO Laboratories, Inc., Carlsbad, CA, USA) and the quality was estimated using a NanoDrop spectrophotometer (ND-1000, NanoDrop Technologies, Wilmington, USA).

PCR amplification and Roche 454 pyrosequencing
For analysis of bacteria diversity, partial 16S rRNA genes were amplified using primers 27F (AGAGTTTGATCMT GGCTCAG) and 533R (TTACCGCGGCTGCTGGCAC). Primer pair ITS1 (TCCGTAGGTGAACCTGCGG) and ITS4 (TCCTCCGCTTATTGATATGC) were was used to amplify partial ITS sequences of eukaryotes. Amplifications were performed using the following program: initial denaturation at 95°C for 2 min, 25 (for 16S rRNA) or 33 (for ITS) cycles of denaturation at 94°C for 30 s, annealing at 55°C for 30 s, extension at 72°C for 30 s, and a final extension at 72°C for 5 min. Emulsion PCR was performed with the emPCRAmp-Lib L Kit (Roche) and PCR amplicons were pyrosequenced by Shanghai Majorbio Bio-pharm Technology Co. Ltd. (Shanghai, China) using a

Processing of 454 sequencing data
Sequences were assigned to different samples according to sample-specific barcodes and processed using QIIME (version 1.17). Data were selected based on the following criteria: (i) an almost perfect match with barcode and primers; (ii) a length of at least 200 nucleotides (barcodes and primers excluded); (iii) an average quality score of >25, with no ambiguous bases or homopolymers longer than six bp, and without any primer mismatches. After this procedure, 67,520 and 140,099 high-quality sequences were obtained for ITS and 16S rDNA reads, respectively. The number of clean reads per samples are listed in Additional file 1: Table S1 and Additional file 2: Table S2.
Operational Taxonomic Units (OTUs) were clustered using a 97% similarity cutoff by UPARSE (version 7.1 http://drive5.com/uparse/) and chimeric sequences were identified and removed using UCHIME. The phylogenetic affiliation of each 16S rRNA gene and ITS sequence were analyzed by RDP Classifier (http://rdp.cme.msu.edu/) against the SILVA (SSU111) 16S/18S rRNA database. Taxonomic assignment from phylum level to strain level based on hits, and these were used to plot abundance graphs.

Statistical analysis
Plant fresh weight, height and number of colonizing bacteria were analyzed using Microsoft Excel. Calculation of variance (ANOVA) and mean comparison between treatments was carried out based on the Tukey's test at the 0.05 probability level using SPSS version 19 (IBM Corporation, New York, USA). Principal component analysis (PCA) was performed to analyze the effect of PTS-394 and sampling time on the microbiota community at the OTU level using the R-forge community ecology package (vegan 2.0 was used to generate the PCA Figures).

Phylogenomics of B. subtilis PTS-394
Phylogenomic analysis of strain PTS-394 was performed taking advantage of availability of the whole draft genome sequence. ANIb and ANIm values in comparison to the B. subtilis type strain 168 were determined by using the JSpecies program package. In order to finally decide, whether PTS-394 belong to B. subtilis subsp. subtilis, electronic DNA-DNA hybridization (dDDH) was applied. For calculating dDDH three different formulas can be applied (see Methods), but only results obtained with the recommended formula 2 were used in our analysis. The nearest neighbors of PTS-394 found by using different phylogenomic analyses were representatives of B. subtilis subsp. subtilis including the type strain B. subtilis 168 (Table 1). Tetra results (tetranucleotide signature correlation index) determined for PTS-394 when compared with B. subtilis subsp. subtilis 168(T) were in subspecies range (>0,999). The same comparison revealed ANI values far above the recommended species threshold of 0.96% [24] and dDDH values clearly exceeding subspecies delineation (>79%, [21]

PTS-394 is able to colonize roots and to promote growth of tomato plants
The effect of B. subtilis PTS-394 on tomato growth and root colonization was evaluated using MS agar plates. After 7 and 14 days growth, PTS-394 cells became clearly visible around the tomato primary root (Fig. 1a). Density of PTS-394 on the tomato rhizoplane ranged from 10 6 to 10 7 CFU per gram of fresh root, and was found enhanced with increasing cell density of the soaking suspension (Fig. 1a). These results indicated successful root colonization by PTS-394, which is a key requirement for biocontrol action and plant growth-promoting activity. Seedling fresh weight data showed that PTS-394 exhibited impressive growth-promoting activity at a cell density (OD 600 ) of 0.5 and 1.0 in the soaking suspension, yielding 1.08 × 10 7 and 1.4 × 10 7 CFU/g of fresh root on the tomato rhizoplane, respectively (Fig. 1b). These treatments increased plant growth by 9.46% and 18.36%, respectively, compared to controls. However, tomato growth was slightly inhibited when the PTS-394 cell number exceeded 7 × 10 7 CFU/g of fresh root, suggesting that too high cell numbers at rhizoplane might negatively affect plant growth by that bacterium.

Evaluation of root colonization and tomato growth promotion by PTS-394 in pot experiments
Pot experiments were performed to investigate the colonization capability of PTS-394G at different time intervals. Initially, rate of root colonization decreased sharply before steadily increasing over time (Fig. 2a). Almost 2 × 10 6 CFU per gram of root were detected immediately after inoculation; 48 h after inoculation this number dropped to 3.7 × 10 5 CFU/g root. During next days PTS-394G cell number increased steadily and reached 1.7 × 10 6 CFU/g root nine days after inoculation. Between days 9-21, the PTS-394G cell number ranged between 1 × 10 6 to 2 × 10 6 CFU/g root, and fluorescent microscopy revealed that colonization occurred on the root surface, where PTS-394 cells formed micro-colonies or biofilms 7 days after treatment (Fig. 2b). The average plant height and root weight were 51.25 cm and 2.94 g, respectively, after treatment with PTS-394, which represented an increase of 8.90% and 18.30% compared with untreated control plants (Fig. 2c).

Impact of PTS-394 on bacterial members of the rhizosphere microbiota
Determination of the composition of the bacterial community in vicinity of plant roots was performed by metagenome sequencing (see Methods  Table S1. Global Alignment for Sequence Taxonomy (GAST) was used for taxonomic assignment of 16S rRNA sequences, and 31 phyla were identified by pooling sequences from all samples, with nineteen core phyla present in all ten samples ( Table 2). The three most abundant core phyla were Proteobacteria, Bacteroidetes and Actinobacteria, which were present in all samples, except the one collected on the first day after treatment with PTS-394. Principal component analysis (PCA) was used to analyze the effect of PTS-394 and sampling time on the structure of the bacterial community in the rhizosphere based on OTUs. PC1 and PC2 together accounted for more than 75% of the variation (Fig. 3a). There was a clear shift in bacterial community between 1 and 14 days after inoculation with PTS-394. Treated or untreated samples collected on day 1 were distinct from each other and from the other time points, however control and treated samples could not be distinguished at later stages. These results indicated that (i) changes in the bacterial community profile becomes visible soon after inoculation with B. subtilis, and (ii) application of PTS-394 affected only transiently the root bacteriome. As expected, a sudden rise of Firmicutes was registered due to the large quantity of B. subtilis PTS-394 cells applied. However, in course of the experiment the number of Firmicutes was steadily decreasing and became similar with the control 14 days after inoculation ( Table 2). BLAST analysis revealed that the relative abundance of 40 different groups (based on genus level under inclusion of some uncertain groups) was temporarily affected after inoculation (Fig.3b). A total of 19 groups were altered by PTS-394 treatment, occurrence of 10 groups was enhanced, including Thermomonas, Pseudomonas, Lactococcus and Rhodanobacter, while nine groups were suppressed, including Flexibacter, Planctomyces and Sneathiella. However, as in Firmicutes (see above), these groups were only temporarily affected and 14 days after inoculation no significant differences to the control were registered. The relative abundance and variation in the composition of these 19 groups is presented in Additional file 3: Figure S1.

Impact of PTS-394 on eukaryotic members of the rhizosphere microbiota
Presence of eukaryotic groups in vicinity of plant roots was determined by using 18S rRNA sequences in a similar way as described for bacteria. In total, 84,579 raw ITS sequences were obtained and the number of clean sequences per sample varied ranged from3, 897 to 8402, with an average of 6752. A total of 1491 Operational Taxonomic Units (OTUs) were obtained at a distance of 0.03 using the TSC algorithm. The Shannon index of diversity was determined for 10 samples, and ranged from 0.28 to 4.15. The Good's coverage per sample ranged from 0.96 to 0.99. Although there was some fluctuation of reads over time, analysis of these data corroborated their credibility. The statistical indexes and richness estimates of per sample are summarized in Additional file 2: Table S2. The eukaryotic OTUs could be arranged into three kingdoms (Fungi, Metazoa and Viridiplantae) along with the no rank Eukaryota group, a group of unclassified Eukaryota and a group of unclassified organisms. Altogether, thirteen phyla were identified. Seven core phyla were detected in every sample. The most abundant phyla were Streptophyta, Chlorophyta Ascomycota and Basidiomycota (Table 3). After the application of PTS-394, the composition of Fungi, Viridiplantae, the no rank Eukaryota group and the unclassified group were altered, compared to the untreated control. The relative abundance of Fungi, of the no rank Eukaryota group, and of the group of unclassified organisms was lower than that of the control at the same sample time. However, the relative abundance of Viridiplantae (Solanoideae) was found higher than in the control.
All OTUs were subjected to PCA analysis and PC1 and PC2 together accounted for more than 90% of the variation (Fig. 4a). A significant temporal change in the community profile of Eukaryota is visible in the biplot. Profiles of control and PTS-394-treated samples collected at identical time points revealed slight differences from day 1 until day 9 after treatment with PTS-394, but could not be distinguished 14 days after inoculation. These results are in accordance to those found for the rhizosphere bacteriome, and indicate that (i) PTS-394 had a transient influence on composition of the eukaryotic members of the rhizosphere microbiome, but (ii) the community recovered to its original state 14 days after treatment. BLAST analysis revealed that the relative abundance of 29 groups (based on the eukaryota genus level and including some uncertain groups) was transiently affected after inoculation with PTS-394 (Fig. 4b). A total of 19 groups were altered by PTS-394 treatment; five groups were enhanced, including Solanoideae, Chlamydomonas and Chloromonas, and 14 groups were decreased. The variation in these 19 groups is presented in Additional file 4: Figure S2.
PCA on OTUs classified as fungi revealed a temporal variation in the fungal community in dependence of PTS-394 inoculation (Fig. 5). The Control and the PTS-394-treated samples were found different during the first four time points, suggesting that PTS-394 inoculation affects the fungal community transiently, similar to its effect on the bacterial community. Sequence homology and analysis of the relative abundance of OTUs showed that abundance of fungi was generally suppressed during  (Table 3). As expected, the relative abundance of Fusarium oxysporum was inhibited after treatment with PTS-394 indicating antagonistic activity of PTS-394 against F. oxysporum (Additional file 5: Figure S3).

Discussion
Plant growth promotion and biocontrol activities are important features of commercial agents used in sustainable agriculture. To date members of the genus Bacillus are preferred for preparing bioformulations with beneficial impact on plant growth and health [1]. Especially, representatives of the B. subtilis species complex are known for their beneficial action on plants, especially B. amyloliquefaciens and B. subtilis [25]. These Bacilli stimulate plant growth: (1) directly, by increasing nutrients through the production of phytohormones, siderophores, organic acids involved in P-solubilisation and/or fixation of nitrogen [26]; and (2) indirectly, by producing antagonistic substances or by inducing the plant resistance against pathogens [7]. Here, we found that plant root-colonizing B. subtilis subsp. subtilis PTS-394G is able to persist on roots and to promote growth of tomato plants. Moreover, the plant pathogen Fusarium This suggests that sterile environment favors root colonization of PTS-394, which is consistent with previous reports [11,27]. It is interesting to note that the plant growth promoting effect exerted by PTS-394 was dependent on the amount of B. subtilis cells used for inoculating tomato plants. Whilst cell numbers of 1.08 × 10 7 and 1.4 × 10 7 CFU/g fresh root supported plant growth, higher cell numbers such as 7 × 10 7 CFU/g fresh root, led to a slight growth inhibition. Without genetic experiments it remains questionable whether this different pattern of growth stimulation could be explained by a concentration dependent effect of the phytohormone indole-3-acetic acid (IAA). The PGPR B. amyloliquefaciens FZB42 is able to produce IAA and it is possible that higher concentrations of IAA exerts an inhibitory effect on plant growth [1]. In fact, ysnE, a gene encoding a putative IAA acetyltransferase, which is involved in IAA synthesis pathway, was detected in the genome of PTS-394. Our results demonstrate that, besides B. amyloliquefaciens and B. pumilus, plant-associated representatives of the B. subtilis subsp. subtilis taxon are promising candidates for developing of efficient biocontrol and biofertilizer agents. In this context it is important to know, whether addition of PTS-394 has an effect on composition of rhizosphere microbial community.
Here, the effects of PTS-394 on the entire rhizosphere microbiota community was investigated by taxonomic profiling of metagenome sequences. In accordance with the results obtained for B. amyloliquefaciens FZB42 [2,15,16], only a transient effect on composition of the root microbiome was detected after adding the PGPR bacterium. Similarly, independent of its mode of application, applying B. amyloliquefaciens BNM122 to soy bean plants did not shift the composition of rhizosphere bacterial community in a measurable extent [14], suggesting that inoculation of crops with different representatives of the genus Bacillus has no durable impact on the bacteria living in vicinity of plant roots. By contrast, external addition of Gram-negative PGPR, such as Pseudomonas spp. [15,28], Enterobacter cowanii [29], and Sinorhizobium meliloti [30] alter composition of root bacterial community in a similar extent as reported for fungal plant pathogens, e.g. Ralstonia solani [2].
Whilst the impact of PGPRs on the rhizosphere bacterial community is fairly investigated, relatively little is known about the effect of PGPRs on the overall eukaryotic community [8]. Analysis of 18S ribosomal ITS sequences showed that PTS-394 influences Eukaryota in the rhizosphere microbial community. However, similar to the effects exerted by rhizosphere bacteria, the impact of PTS-394 on Eukaryota was only transient, although longer-lasting than its effect on bacterial community.
The rhizosphere contains amino acids and sugars nutrients due to the accumulation of root exudates that provides a rich source of energy and nutrients for rhizosphere microbiome. This determines composition of the microbiome community [7,8]. After application of PTS-394, relative abundance of Fungi, Viridiplantae, the no rank Eukaryota group and the unclassified group was similar as in the control, but changes in their relative abundance compared to control were registered. Here, a significant increase of Viridiplantae relative abundance was observed which might be due to enhanced root development caused by PTS-394 as seen by enhanced number of streptophyta counts largely corresponding to tomato cells. Chlorophyte, another group of the Viridiplantae, the relative abundance was decreased significantly in the PTS-394 treatment, which correspond to the variation trend of main genus Chlamydomonas. In addition, PGPR have the ability to secrete several antagonistic compounds, such as antifungal acting lipopeptide antibiotics. Nihorimbere et al. reported that surfactin, iturin and fengycin were detected when Bacillus amyloliquefaciens S499 colonized tomato rhizosphere [31]. We found that suspensions of B. subtilis PTS-394 contained lipopeptides and polyketides. Whilst PTS-394 colonizes plant root, lipopeptides and polyketides might be secreted by the strain into the environment. These secondary compounds will either directly inhibit members of the microbiome community or stimulate the ISR response in plants. The suppressing effect of PTS-394 on plant pathogen Fusarium oxysporum in the rhizosphere is a typical example (Additional file 5: Figure S3). We assume that the transient effect of Bacillus subtilis PTS-394 on composition of the rhizosphere microbiome is caused by a sudden rise of Bacillus cells after inoculation, which is soon compensated by the indigenous microbial community. In case of eukaryota this restoration process is somewhat delayed due to their longer generation time compared to bacteria.