Bacillus amyloliquefaciens FH-1 significantly affects cucumber seedlings and the rhizosphere bacterial community but not soil

Plant growth-promoting bacteria (PGPB) inoculants have been applied worldwide. However, the ecological roles of PGPB under different soil conditions are still not well understood. The present study aimed to explore the ecological roles of Bacillus amyloliquefaciens FH-1 (FH) on cucumber seedlings, rhizosphere soil properties, and the bacterial community in pot experiments. The results showed that FH had significant effects on cucumber seedlings and the rhizosphere bacterial community but not on soil properties. The FH promoted cucumber seedlings growth, reduced the rhizosphere bacterial diversity, increased Proteobacteria, and decreased Acidobacteria. Linear discriminant analysis (LDA) effect size (LEfSe) revealed that FH enriched two taxa (GKS2_174 and Nannocystaceae) and inhibited 18 taxa (mainly Acidobacteria, Actinobacteria, BRC1, Chloroflexi, Plantctomycetes, and Verrucomicrobia). Co-occurrence network analysis demonstrated that FH increased bacteria-bacteria interactions and that Bacillus (genus of FH) had few interactions with the enriched and inhibited taxa. This might indicate that FH does not directly affect the enriched and inhibited taxa. Correlation analysis results displayed that cucumber seedlings’ weight and height/length (except root length) were significantly correlated with the 18 inhibited taxa and the enriched taxa Nannocystaceae. It was speculated that FH might promote cucumber seedling growth by indirectly enriching Nannocystaceae and inhibiting some taxa from Acidobacteria, Actinobacteria, BRC1, Chloroflexi, Plantctomycetes, and Verrucomicrobia.

www.nature.com/scientificreports/ In recent years, with the recognition of the importance of the rhizosphere microbiome, research on the effect of B. amyloliquefaciens on the rhizosphere microbiome has increased. Rhizosphere microbiomes play key roles in the disease, health, growth, and development of their host [21][22][23][24] . Many reports have indicated that the application of microbial inoculants could influence resident microbial communities 8,[25][26][27] . The effects of B. amyloliquefaciens on the rhizosphere microbial communities of tomato, rice, lettuce, banana, tobacco, and cucumber were investigated (Table S3). However, most studies focus on community composition and diversity, while only a few focus on co-occurrence network analysis. Co-occurrence network analysis of taxon co-occurrence patterns might help identify potential biotic interactions between inoculants and soil indigenous microorganisms and increase the understanding of how inoculants affect microbial communities 13,28,29 . In addition, many studies on cucumbers are based on peat and vermiculite 13,19 . This may be different from the results based on soil. Moreover, the ecological roles of B. amyloliquefaciens under soil conditions are not well understood. The comprehensive effects of B. amyloliquefaciens on crops, soil, and microorganisms still lack systematic and in-depth study.
To better understand the ecological roles of B. amyloliquefaciens under soil conditions, we investigated the effects of B. amyloliquefaciens FH-1 (FH), which could significantly promote rice growth in field experiments 18 , on cucumber seedlings, rhizosphere soil properties, and the bacterial community in soil by using high-throughput sequencing technology, network analysis, and multivariate statistical methods. This will provide theoretical guidance for the development and application of PGPB inoculants in future.

Results
FH had significant effects on cucumber seedlings. The cucumber seedlings' weight and height were significantly affected by FH (Table 1). FH significantly increased the fresh weight of plants, shoots, and roots and increased the plant dry weight and shoot height of cucumber seedlings compared to those drenched with sterile deionized water (CK).
FH had no significant effect on rhizosphere soil properties. FH had no significant effect on soil pH, total organic carbon, total nitrogen, total phosphorus, nitrate nitrogen, or available phosphorus (Table 2). However, the soil total nitrogen, total phosphorus, nitrate nitrogen, and available phosphorus in FH were generally higher than that in CK.  (Fig. 1). These dominant phyla accounted for more than 94% of the bacterial sequences from each soil sample. Deltaproteobacteria (P = 0.01) was significantly increased, while Acidobacteria (P = 0.00) was significantly decreased by FH (Table S1). LEfSe analysis showed that a total of 20 bacterial groups were distinct between FH and CK treatments using the logarithmic (LDA) value of 2 (Fig. 2). The bacterial taxa enriched in FH were GKS2-174 and Nannocystaceae. Acidobacteria-6 (the class and its order CCU21 and iii1-15, the order and its family mb2424), MB-A2-108 (the class and its order 0319-7L14), Rubrobacteria (the class and its order Rubrobacterales, the order and its family Rubrobacteraceae), PRR-11, C0119, Gitt-GS-136, Gemmataceae (the family and its genus Gemmata), Pseudonocardiaceae, Leucobacter, and Prosthecobacter were enriched in CK, which also could be regarded as inhibited taxa in FH.  www.nature.com/scientificreports/ The relative abundances of both B. amyloliquefaciens and Bacillus spp. were slightly higher in FH than in CK (Fig. S1). This suggested that B. amyloliquefaciens FH-1 might slightly colonize cucumber rhizosphere soil.
Principal coordinate analysis (PCoA) revealed that the rhizosphere bacterial communities of FH were distinct from those of CK (Fig. 3). ANOSIM analysis (global R = 0.488, P = 0.008) and PERMANOVA analysis (R 2 = 0.326, P = 0.009) demonstrated that the structure of bacterial communities was significantly changed by FH.
FH modified rhizosphere bacterial networks. Whether FH affected the interaction of bacterial communities and whether FH interacted with enriched or inhibited taxa at the genus level were determined using co-occurrence network analysis based on a strong (Spearman's r > 0.6) and significant (P < 0.05) correlation. The calculated modularity index was larger than 0.4, and the random modularity index (Table 4) indicated a typical module structure 30 . Overall, the FH showed a remarkable influence on the co-occurrence networks in bacterial communities (Fig. 4). The number of positive correlations was higher than that of the negative correlations in both networks. FH had higher edges, negative correlations, and an average degree and modularity but lower positive correlations than CK (Table 4). There were more species interacting with Bacillus in FH than that in CK. There were 19 genera that interacted with Bacillus, and seven of them had positive interactions in FH. In CK, only seven genera interacted with Bacillus, and six of them had positive interactions ( Fig. 4 and Table S2). Bacillus only had positive interactions with the inhibited taxa Leucobacter in CK and the inhibited taxa MB-A2-108 in FH.
Cucumber seedling characteristics were significantly correlated with the bacteria inhibited and enriched by FH. Correlation analysis showed that cucumber seedlings' weight and height/length (except root length) had a significant correlation with the bacteria taxa inhibited and enriched by FH (Fig. 5). All 18 inhibited taxa (mainly Acidobacteria, Actinobacteria, BRC1, Chloroflexi, Plantctomycetes, and Verrucomicrobia) were significantly and negatively correlated with some cucumber seedlings' characteristics. These inhibited taxa had a closer relationship with cucumber shoots than roots. Enriched taxa Nannocystaceae had a significant positive correlation with cucumber shoot height.

Discussion
In this study, the ecological roles of inoculant B. amyloliquefaciens FH-1 on cucumber seedlings, rhizosphere soil, and the bacterial community were investigated. The results illustrated that FH had a significant effect on cucumber seedlings and the rhizosphere bacterial community but not on soil.
Rhizosphere bacterial communities play a key role in the disease, health, growth, and development of plants [31][32][33] . The effect of PGPB on the bacterial community is still unclear. As a well-known PGPB, the effect of B. amyloliquefaciens on the bacterial community has been widely studied (Table S3). Some studies have shown that B. amyloliquefaciens has no influence on rhizosphere bacteria, while some have a significant influence. Some increased diversity, while some decreased diversity, and some improved Proteobacteria, while some improved Firmicutes [13][14][15][16][17][18][19][34][35][36][37][38] . In this study, we found that FH inoculation significantly reduced bacterial diversity, increased Proteobacteria that may belong to r-strategies, and decreased Acidobacteria that may belong to k-strategies 14 . The influence of B. amyloliquefaciens on the bacterial community may be attributed to different strains, plant species, soil types, and environmental factors. Therefore, it is necessary to investigate the influence of B. amyloliquefaciens  LEfSe was used to identify taxa that were inhibited and enriched by FH. The inhibited taxa were mainly Acidobacteria, Actinobacteria, Chloroflexi, BRC1, Planctomycetes, and Verrucomicrobia. The inhibited Acidobac-teria_6 was universal in soil, yet our knowledge of the role of these diverse organisms remained rudimentary 39,40 . Some Actinobacteria are pathogenic to plants. The vast majority of Actinobacteria are important saprophytes capable of decomposing plant and animal debris 41 . Species of Pseudonocardiaceae are recognized as emerging opportunistic pathogens of plants and animals 42 . Rubrobacter spp. are extremophiles with radioresistant characteristics 43,44 . Leucobacter spp. are generally identified as chromium reducers 45,46 . Certain subspecies of Leucobacter have the potential for pathogenic interactions with nematodes 47 . Chloroflexi usually exist in some extreme environments 48,49 . Planctomycetes, Verrucomicrobia, and BRC1 belong to the PVC superphylum. This group appeared to be ubiquitous and contained pathogenic species 50 . Prosthecobacter has been found in freshwater, activated sludges, and Panax notoginseng 51,52 . Gemmataceae and its genus Gemmata have been found in wastewater plants, acid bogs, swamps, and the soil 53,54 . The inhibited taxa were not common disease bacteria. They may be potentially harmful bacteria or useless bacteria on cucumbers. The enriched taxa contained GN02_GKS2_174 and Nannocystaceae. At present, little is known about the ecological function of GN02. Nannocystaceae was reported to promote Brassica napus growth 55 .
In addition, network analysis was used to understand how B. amyloliquefaciens FH-1 affected the interactions of bacterial communities and to explore whether B. amyloliquefaciens FH-1 interacted with the enriched and inhibited taxa. The relative abundance of B. amyloliquefaciens was too low; so, a co-occurrence network at the genus level was constructed. The results showed that FH increased the complexity of the whole network, especially the negative interactions. The higher complexity of networks was more resilient to environmental stressors, as different species can complement each other 56 . A negative correlation may mean competition 57 . It is possible that FH increased the competition among bacterial communities and led to the decrease of some taxa and the α-diversity. FH also increased the interactions of Bacillus with other genera, especially negative interactions. Except for MB-A2-108 and Leucobacter in Actinobacteria, Bacillus had no interaction with other inhibited or enriched taxa. This might indicate that B. amyloliquefaciens FH-1 does not directly affect the enriched and inhibited taxa. In CK and FH, Bacillus and the taxa it interacted with belonged to a module (Fig. 4). Modules reflect the heterogeneity of habitats, aggregation of closely related species in phylogeny, niche overlap, and coevolution of species 58 . The genera interacting with Bacillus in FH (nine phyla) were completely different from those in CK (four phyla) (Table S2). It was speculated that the function of the module with Bacillus as the core changed after being inoculated with B. amyloliquefaciens FH-1.
Correlation analysis showed that except for GN02_GKS2_174, all inhibited and enriched taxa had significant correlations with cucumber seedlings' weight and height/length, except root length. This indicated that Figure 5. Heatmap of Spearman's correlation coefficients between cucumber seedlings and bacteria inhibited and enriched by FH. The colors represent the correlation, with red being more positive and blue being more negative. Significance is given as *(P < 0.05) and **(P < 0.01). www.nature.com/scientificreports/ B. amyloliquefaciens FH-1 might promote cucumber seedling growth by regulating the bacterial community and indirectly enriching Nannocystaceae and inhibiting some taxa from Acidobacteria, Actinobacteria, BRC1, Chloroflexi, Plantctomycetes, and Verrucomicrobia. This result was roughly supported by many previously published studies 18,59 . Regulating the rhizosphere microbiome is an important mechanism for PGPB to promote plant growth. Whole genome data showed that B. amyloliquefaciens FH-1 had no complete pathway for nitrogen fixation or secretion of IAA, gibberellin (GA), abscisic acid (ABA), or ethylene but had a complete pathway to secrete organic acids (malic acid, acetic acid, succinic acid, and gluconic acid), phytase, zeatin, and siderophore (data not shown). This study showed that FH had no significant effect on soil properties, suggesting that the ability of B. amyloliquefaciens FH-1 to dissolve phosphorus and potassium did not play a role in soil characteristics. In our next work, we will verify whether FH promotes cucumber seedling growth by secreting zeatin and siderophore.

Materials and methods
Bacterial inoculum preparation. B. amyloliquefaciens FH-1 was grown at 30 °C for 48 h in Luria-Bertani (LB) broth on a rotary shaker (180 rpm). The cells were harvested by centrifugation (5000 × g for 10 min), and the bacterial pellet was washed three times with 0.9% NaCl and finally resuspended in sterile deionized water at 1 × 10 8 CFU/ml.
Pot experiment for the cultivation of cucumber seedlings. For future applications in coastal salinealkali land, soil (pH 8.14, 4.1 g/kg organic matter, 655 mg/kg total N, 18 mg/kg available N, 250 mg/kg total P, 155 mg/kg available P, 4893 mg/kg total K, and 124 mg/kg available K) was collected from the upper 30 cm of a weed field in an airport economic area in Tianjin, China. The sampled soil was air dried and mixed thoroughly, followed by a sieving step (0.5-cm mesh) to remove plant debris. Cucumber seeds (Jin you NO.1, Tianjin Kerun Agricultural Science Technologies Inc., Tianjin, China) were procured from the local market (Fig. S2). Two cucumber seeds were sown in each plastic pot (diameter 8 cm; height 10 cm) containing 300 g of soil. Pot soils were drenched with 300 ml of the prepared inoculums or equivalent sterile deionized water. In total, there were two treatments: (1) soil drenched with B. amyloliquefaciens FH-1 (FH), and (2) soil drenched with sterile deionized water (CK). Five replications of each treatment were set up during the entire experimental period. Pots were placed randomly in a growth chamber at 28 °C day/17 °C night, 75% relative humidity, and 9 h light, and watered weekly. All methods were carried out in accordance with relevant guidelines and regulations.
Plant characteristics and soil chemical properties. At 35 days after sowing, plants of each pot were harvested and carefully separated into roots and shoots to determine the growth parameters, including length, fresh weight, and dry weight, using rulers and balances. Meanwhile, rhizosphere soil was shaken off of the roots of each treatment to be collected and stored at 4 °C and − 80 °C, respectively. The rhizosphere soil pH, total organic carbon, total nitrogen, total phosphorus, nitrate nitrogen, and available phosphorus were determined using commercial chemical assay kits (Suzhou Comin Biotechnology Co. Ltd., Suzhou, China) following the manufacturer's instructions.
DNA extraction, PCR amplification, and Hiseq sequencing. Soil metagenomic DNA was isolated from 10 soil samples by the PowerSoil DNA isolation kit (MO BIO Laboratories Inc., Carlsbad, CA, USA) according to the manufacturer's instructions. DNA purity and concentration were monitored by 1% agarose gels and NanoDrop ND-2000 spectrophotometry (NanoDrop Technologies, Wilmington, DE, USA), respectively. The bacterial hypervariable regions (V4) of the 16S rRNA genes were amplified using primer 515F-806R with a barcode 60 . PCR products were purified and sequenced using the Miseq platform at Novogene Co. Ltd (Tianjin, China). The raw sequence data were deposited in the NCBI Sequence Read Archive as accession PRJNA544608 for bacteria. Raw data were processed and analyzed as previously described using the QIIME 60 . The relative abundance of B. amyloliquefaciens was determined by local BLAST.
Data analyses. All statistical analyses were performed using R (version 3.1.1) 61 . The cucumber seedlings' characteristics, soil properties, bacterial α-diversity indices, and relative abundance of taxa in different treatments were compared using Independent Sample t tests. Principal coordinate analysis (PCoA), analysis of similarity (ANOSIM), and permutational multivariate analysis of variance (PERMANOVA) with the ADONIS function based on weighted UniFrac distance were performed to evaluate the overall differences in the bacterial community 62 . Linear discriminant analysis (LDA) effect size (LEfSe) analysis was used to identify taxa that were enriched and inhibited by FH 63 . Network analysis was used to explore whether FH affected bacteria-bacteria interactions and whether FH directly interacted with the enriched or inhibited taxa at the genus level. The cooccurrence network was inferred based on the Spearman correlation matrix constructed with the 'Hmisc' and 'igraph' package in R. We generated network images and calculated network properties with Gephi 64,65 . Spearman's rank correlation coefficient was used to evaluate the relationships between cucumber seedlings and the taxa enriched and inhibited by FH. Heatmaps that illustrate correlation data were generated using the 'pheatmap' package in R. www.nature.com/scientificreports/ Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.