Biosynthetic Gene Clusters in Sequenced Genomes of Four Contrasting Rhizobacteria in Phytopathogen Inhibition and Interaction with Capsicum annuum Roots

Abstract

Through screening of rhizobacteria, species that effectively suppress phytopathogens and/or promote plant growth are found. Genome sequencing is a crucial step in obtaining a complete characterization of microorganisms for biotechnological applications. This study aimed to sequence the genomes of four rhizobacteria that differ in their inhibition of four root pathogens and in their interaction with chili pepper roots to identify the species and analyze differences in the biosynthetic gene clusters (BGCs) for antibiotic metabolites and to determine possible phenotype-genotype correlations. Results from sequencing and genome alignment identified two bacteria as Paenibacillus polymyxa, one as Kocuria polaris, and one that was previously sequenced as Bacillus velezensis. Analysis with antiSMASH and PRISM tools showed that B. velezensis 2A-2B, the strain with the best performance of referred characteristics, had 13 BGCs, including those related to surfactin, fengycin, and macrolactin, not shared with the other bacteria, whereas P. polymyxa 2A-2A and 3A-25AI, with up to 31 BGCs, showed lower pathogen inhibition and plant hostility; K. polaris showed the least antifungal capacity. P. polymyxa and B. velezensis had the highest number of BGCs for nonribosomal peptides and polyketides. In conclusion, the 13 BGCs in the genome of B. velezensis 2A-2B that were not present in the other bacteria could explain its effective antifungal capacity and could also contribute to its friendly interaction with chili pepper roots. The high number of other BGCs for nonribosomal peptides and polyketide shared by the four bacteria contributed much less to phenotypic differences. IMPORTANCE To advance the characterization of a microorganism as a biocontrol agent against phytopathogens, it is highly recommended to analyze the potential of the profile of secondary metabolites as antibiotics that it produces to counteract pathogens. Some specific metabolites have positive impacts in plants. By analyzing sequenced genomes with bioinformatic tools, such as antiSMASH and PRISM, outstanding bacterial strains with high potential to inhibit phytopathogens and/or promote plant growth can be quickly selected to confirm and expand our knowledge of BGCs of great value in phytopathology.

Keywords: Bacillus velezensis; antibiotics; biocontrol of phytopathogens; biosynthetic gene clusters; secondary metabolites.

FIG 1

Bacteria and phytopathogens dual confrontations in PDA medium after 6 to 12 days of bacterial growth and 7 to 14 days of pathogen growth. In the top row, the pathogen Rhizoctonia solani without contact (1st column) or in contact with, from left to right (2nd to 5th column), 3A-25AI, 2A-2A, 2A-2B, and MS50-16 bacterial strains. In the second row, the pathogen Fusarium solani in contact with the same bacteria and the same order described above. In the third row, the pathogen Fusarium oxysporum in contact with the same bacteria and the same order. In the bottom row, the pathogen Phytophthora capsici in contact with the same bacteria.

FIG 2

Growth inhibition of four root pathogens in pepper plant by four bacterial strains from the rhizosphere or soil. Pathogens include Phytophthora capsici, Rhizoctonia solani, Fusarium solani, and Fusarium oxysporum. ANOVA statistical analysis and Tukey test were performed.

FIG 3

Behavior of bacterial strains when inoculated in roots of 15-day-old chili pepper seedlings, under in vitro conditions, and statistical results. Results 6 days postinoculation. (A) Plant-bacteria interactions, control treatment in image a. Friendly plant-bacteria interactions in images b and c; in b, strain 2A-2B; in c, strain MS50-16. Hostile plant-bacteria interactions in images d and e; in d, strain 2A-2A; in e, strain 3A-25AI. (B) ANOVA statistical analysis and Tukey test were performed.

FIG 4

Protection in plant infected with P. capsici and R. solani via root preinoculation with bacterial strain 2A-2B. The experiment was conducted in 18-day-old plants and went on for 6 days. (A) The top row shows plants inoculated with pathogens P. capsici and R. solani; a, time zero; b, c, and d, 48, 72, and 96 h postinoculation, respectively. (A) The bottom row shows plants preinoculated with bacterial strain 2A-2B and immediately inoculated with strains of the pathogens P. capsici and R. solani. The observation times for the images e, f, g, and h were as mentioned above. (B) ANOVA statistical analysis and Tukey test were performed.

FIG 5

Phylogenetic tree based on the genome of bacterial strain MS50-16 and 10 other accessions with major identity in the NCBI data bank. Phylogenetic tree constructed using Neighbor-joining method.

FIG 6

Representation of 28 or more local colinear blocks (LCBs) between chromosomal sequences of Bacillus velezensis 2A-2B and the reference genome B. velezensis JS25R, in addition the GFP and AD8 reference genomes from B. velezensis. (A) Alignment of bacterial strain 2A-2B and B. velezensis JS25R as reference genome. (B) Multiple alignment of genomes between 2A-2B strain with respect to the other three genomes of B. velezensis strains with the highest identity in the gene bank. The alignment was performed in Mauve tool, version snapshot 2015 (47).

FIG 7

Conserved biosynthetic gene clusters (BGCs) for biosynthesis of secondary metabolites in the genomes of the four bacterial strains studied. In each case, to the right side, in the green background, BGCs with similarity between 80 and 100% and, in red background, BGCs with similarity less than 80%. These data correspond to results of the antiSMASH algorithm analysis.

FIG 8

Comparisons of structural arrangement and number of genes in the BGC for fengycin biosynthesis, found in the genomes of bacterial strain 2A-2B and Bacillus velezensis FZB42 as reference BGC, obtained from the repository of BGCs in the MIBiG online server. (A) Well-described 87.4-MB reference BGC for fengycin biosynthesis in B. velezensis FZB24; in the red background are five core genes, fenA to fenE; 10 modules and other regulatory genes are in the gray background. (B, C, and D) BGCs for fengycin biosynthesis in the genome of 2A-2B strain. (B) A 12.8-kb BGC for fengycin biosynthesis with 26% similarity. (C) A 87.4-kb BGC for fengycin with 80% similarity. (D) A 9.4-kb BGC with 20% similarity. This information was obtained using the online antiSMASH tool.

FIG 9

Comparisons of structural arrangement and number of genes in the BGC for surfactin biosynthesis found in genomes of bacterial strain 2A-2B and Bacillus velezensis FZB42 as reference BGC, obtained from the repository of BGCs in the MIBiG online server. (A) Well-described reference 41.8-kb BGC for surfactin in B. velezensis FZB24; in red background are three central core genes, srfAA to srfAC and 19 core genes, and about 30 modules and regulatory genes are in the gray background. (B, C, and D) BGCs for surfactin biosynthesis in the genome of the 2A-2B strain. (B) A BGC for surfactin biosynthesis with 52% similarity. (C) A 2nd BGC for fengycin with 80% similarity. (D) A 3rd BGC with 20% similarity. This information was obtained using the online antiSMASH tool.

FIG 10

Circular map of genomes of the four bacteria with indications of genome size, GC content, contigs, coding sequences, and open reading frames (ORFs) in the DNA. Inner ring, thousands of bp of the genome; 2nd ring, GC content; 3rd ring, location and size of contigs; 4th and 5th ring, open reading frames (OPF) in sense and antisense DNA strands; 6th and 7th rings, coding sequences on sense and antisense DNA strand. (A) Paenibacillus polymyxa 2A-2A; (B) Bacillus velezensis 2A-2B; (C) P. polymyxa 3A-25AI; (D) strain MS-5016. The maps were constructed using the CGView tool (45).

FIG 11

Antimicrobial secondary metabolite biosynthesis gene clusters (BGCs) in each of the four bacterial genomes studied, 2A-2B, 2A-2A, 3A-25AI, and MS50-16. With data obtained using the antiSMASH tool, antiSMASH 7 beta version.

FIG 12

Antimicrobial secondary metabolite biosynthesis gene clusters (BGCs) in each of the four bacterial genomes studied, 2A-2B, 2A-2A, 3A-25AI, and MS50-16. With data obtained using the PRISM tool, version 4.