Analysis of the complete genome sequence of Paenibacillus sp. lzh-N1 reveals its antagonistic ability

Background Plant diseases caused by pathogenic fungi are devastating. However, commonly used fungicides are harmful to the environment, and some are becoming ineffective due to fungal resistance. Therefore, eco-friendly biological methods to control pathogenic fungi are urgently needed. Results In this study, a strain, Paenibacillus sp. lzh-N1, that could inhibit the growth of the pathogenic fungus Mycosphaerella sentina (Fr) Schrorter was isolated from the rhizosphere soil of pear trees, and the complete genome sequence of the strain was obtained, annotated, and analyzed to reveal the genetic foundation of its antagonistic ability. The entire genome of this strain contained a circular chromosome of 5,641,488 bp with a GC content of 45.50%. The results of species identification show that the strain belongs to the same species as P. polymyxa Sb3-1 and P. polymyxa CJX518. Sixteen secondary metabolic biosynthetic gene clusters were predicted by antiSMASH, including those of the antifungal peptides fusaricidin B and paenilarvins. In addition, biofilm formation-related genes containing two potential gene clusters for cyclic lactone autoinducer, a gene encoding S-ribosylhomocysteine lyase (LuxS), and three genes encoding exopolysaccharide biosynthesis protein were identified. Conclusions Antifungal peptides and glucanase biosynthesized by Paenibacillus sp. lzh-N1 may be responsible for its antagonistic effect. Moreover, quorum sensing systems may influence the biocontrol activity of this strain directly or indirectly.


Background
Plant diseases caused by pathogenic fungi are devastating, which may lead to severe crop losses.A variety of diseases caused by fungi have been controlled by chemical fungicides.However, most of these fungicides are harmful to the environment, and some are not effective in eradicating pathogens as many pathogens have become resistant to antifungal drugs [1].Therefore, biological methods to protect crops from plant diseases have received considerable attention.Plant-growth-promoting rhizobacteria (PGPR), a group of rhizosphere bacteria, can improve plant growth through various mechanisms, such as biological nitrogen fixation, siderophore production, antibiotic secretion, and systemic resistant induction [2], making them the most promising biocontrol agents to control diseases affecting crop production.
Among PGPR, bacteria in the genus Paenibacillus (e.g., P. polymyxa) have attracted global interest as they have native beneficial properties to apply in the modern agriculture.Paenibacillus sp. can promote plant growth by improving nitrogen fixation, iron acquisition, and phosphorus solubilization [3].More importantly, Paenibacillus sp. can secrete diverse beneficial bioactive metabolites including antimicrobial polypeptides, volatile organic compounds, and hydrolytic enzymes, which can reduce the reliance on chemical antimicrobial agents.Furthermore, Paenibacillus sp. has shown extensive environmental adaptability [4][5][6].These characters enable the bacteria to be widely used in the biofertilizer for improving plant growth and for maintaining a sustainable agroecosystem simultaneously.
Paenibacillus sp.lzh-N1 was isolated from the rhizosphere of pear in Taian, Shandong Province of China.It has excellent inhibitory effects on the pathogen Mycosphaerella sentina (Fr) Schrorter.In the present study, the complete genome sequence of Paenibacillus sp.lzh-N1 was obtained, and the corresponding genes were annotated.In addition, the secondary metabolite biosynthetic gene clusters (BGCs), genes encoding hydrolytic enzymes, and potential quorum sensing (QS) systems were analyzed to elucidate the genetic background and molecular mechanism underlying the antagonistic ability of this strain.

The biocontrol efficacy of Paenibacillus sp. lzh-N1
The antagonistic activity of Paenibacillus sp.lzh-N1 against M. sentina (Fr) Schrorter, a destructive pathogenic fungus for pear in China, was evaluated.Paenibacillus sp.lzh-N1 showed excellent antagonistic activity against the pathogen (Fig. 1), indicating that this strain is promising to improve pear growth as a biocontrol agent.Paenibacillus sp.lzh-N1 has been employed as a microbial fertilizer to increase pear production.

General genome features
The assembled genome of Paenibacillus sp.lzh-N1 was found to be comprised of a single circular chromosome of 5,641,488 bp in length with a GC content of 45.50%, including 5,018 genes, 4,860 protein-coding genes, 109 tRNA, 45 rRNA, and 4 ncRNA (Table 1; Fig. 2).The genome data have been deposited in GenBank under the accession number of CP025696.1.
The protein-coding genes of Paenibacillus sp.lzh-N1 were classified into 20 different functional categories of Cluster of Orthologous Groups of proteins (COG, Fig. 3).The top three categories enriched by the genes included transcription (11.9%), carbohydrate transport and metabolism (11.1%), and amino acid transport and metabolism (8.8%).However, about 13.4% of the protein-coding genes were only function predicted, and about 7.6% were still poorly characterized.

Phylogenetic analyses
Full-length 16 S ribosomal RNA gene (16 S rDNA) sequences (1425 nt) of Paenibacillus sp.lzh-N1 were obtained and deposited in GenBank under the accession number of KX865137.1.The gene sequence comparison with sequences in the GenBank database indicated that the strain lzh-N1 belonged to the genus Paenibacillus.The phylogenetic analysis of Paenibacillus sp.lzh-N1 with related Paenibacillus species based on 16 S rDNA sequences is shown in Fig. 4. Paenibacillus sp.lzh-N1 was clustered to P. polymyxa, and the closest relatives of the strain were P. polymyxa Sb3-1 and P. polymyxa CJX518.The phylogenetic analysis based on whole genome sequences is displayed in Fig. 5. Consistent with the results of phylogenetic analysis based on 16 S rDNA sequences, Paenibacillus sp.lzh-N1 was closest to P. polymyxa Sb3-1 and P. polymyxa CJX518.

Average nucleotide identity (ANI) and genome-to-genome distance calculation (GGDC) analyses
Two independent "digital DNA-DNA hybridization (dDDH)" methods, ANI and GGDC, were used to estimate the overall similarity between the genomes of two strains.The heatmap analysis of ANI values is shown in Fig. 6A.Genome comparison between Paenibacillus sp.lzh-N1 and P. polymyxa Sb3-1 and between Paenibacillus sp.lzh-N1 and P. polymyxa CJX518 generated an ANI value of 98.75% and 98.76%, respectively.It is higher than the 95% threshold for same species designation.The ANI values between Paenibacillus sp.lzh-N1 and other strains were less than 91%.The heatmap analysis based on the predicted DDH values is presented in Fig. 6B.For Paenibacillus sp.lzh-N1 and P. polymyxa Sb3-1 and for Paenibacillus sp.lzh-N1 and P. polymyxa CJX518, the predicted DDH values were 88.9% and 89.0%, respectively, much higher than the recommended 70% species-delimiting DDH value.The DDH values between Paenibacillus sp.lzh-N1 and other strains were less than 42%.The values of ANI and DDH are consistent, and it can be stated that Paenibacillus sp.lzh-N1 belongs to the same species as P. polymyxa Sb3-1 and P. polymyxa CJX518.

Genetic foundation for the biosynthesis of antimicrobial metabolites
Paenibacillus sp.lzh-N1 exhibited significant antagonistic activity against M. sentina (Fr) Schrorter, indicating   2).Among these secondary metabolites, paenicidin [7], polymyxin B [8], aurantinin B [9], brevicidine [10], paenilipoheptin, and tridecaptin [11] have been proven to be active against bacteria, while fusaricidin B [8] and paenilarvins [12] have excellent antifungal activities.However, the functions of another four potential secondary metabolic BGCs are still unknown.These RiPP and betalactone metabolites may be related to the antimicrobial activity of the strain.Apart from these antimicrobial peptides, hydrolytic enzyme glucanase which can degrade the cell walls of various phytopathogens has antifungal effect [4].Six genes encoding glucanase were found in the genome, and their locus_tags were C0638_03115, C0638_11415, C0638_14795, C0638_14995, C0638_20700, and C0638_20790, respectively.These findings provide a

Regulation of biofilm formation
It has been reported that biofilm formation can enhance the antagonistic efficacy of many biocontrol agents, including P. polymyxa [13] and Bacillus subtilis [14].Numerous studies have demonstrated that biofilm formation is regulated by QS systems in various bacteria [15].Additionally, QS can regulate the production of some secondary metabolites in a cell density-dependent manner [16].Therefore, QS may be important for the biocontrol activity of Paenibacillus sp.lzh-N1.There were two potential gene clusters producing cyclic lactone autoinducer (Table 2).Genes encoding the accessory regulator AgrB were found in two clusters with the locus_tags of C0638_23545 and C0638_25055, respectively.Furthermore, gene encoding S-ribosylhomocysteine lyase (LuxS, AUO05948.1)was also found in the genome.These results suggest that there are one or two QS systems in the strain.However, further studies are needed to clarify the roles of QS systems in regulating the biofilm formation and antagonistic activity of Paenibacillus sp.lzh-N1.
Exopolysaccharide (EPS) is an important component of the biofilms, which can influence the attachment and colonization of microbes to roots and root appendages [4].Three genes were found to encode the EPS biosynthesis protein in the genome.The locus_tags of these genes were C0638_00325, C0638_11875, and C0638_20035, respectively.However, their functions in biofilm formation still await further investigations.

Discussion
The pear leaf spot caused by M. sentina (Fr) Schrorter has received worldwide attention.If the attack of the fungus is intense, it will lead to premature defoliation and consequently yield losses.In this study, a strain, Paenibacillus sp.lzh-N1, was isolated from the rhizosphere soil of pear trees.The results of antagonistic experiments showed that Paenibacillus sp.lzh-N1 could significantly inhibit the growth of the pathogenic fungus.The results of species identification show that the strain belongs to the same species as P. polymyxa Sb3-1 and P. polymyxa CJX518.The biocontrol properties of Paenibacillus sp.lzh-N1 revealed its importance as a PGPR, which could provide great potential for agricultural applications.
The complete genome sequence of the strain was obtained and annotated to reveal the genetic foundation of its antagonistic ability.Fourteen antimicrobial peptides related to BGCs were successfully identified.The peptides included RiPPs, NRPs, PKs, and the NRPs-PKs hybrid.Among these peptides, fusaricidins and paenilarvins have outstanding antifungal activities [8,12], which might be responsible for the antagonistic effects of Paenibacillus sp.lzh-N1 against the pathogenic fungus.In addition, several other peptides such as paenicidin, polymyxin, aurantinin, brevicidine, paenilipoheptin, and tridecaptin exhibit great antibacterial activities [7][8][9][10][11].Therefore, it is speculated that Paenibacillus sp.lzh-N1 may have inhibitory activities against a wide range of microorganisms, making this strain a promising biocontrol agent.Besides, six genes encoding glucanase were found in the genome of Paenibacillus sp.lzh-N1.Hydrolytic enzymes (e.g., glucanase and chitinase) can degrade the cell walls of many phytopathogens [4].Overexpression of glucanase from P. polymyxa A21 in Streptomyces lydicus has been reported to improve its biocontrol effect against the pathogenic fungus Botrytis cinerea [17].Glucanase may also participate in the antifungal effects of Paenibacillus sp.lzh-N1.
Biofilm formation plays an important role in the attachment and colonization of the bacteria to the roots and can promote the antagonistic efficacy of biocontrol agents [4].Generally, biofilm formation is strictly regulated by bacteria and affected by many environmental factors.QS is one of the crucial factors that regulate the formation of biofilm in various bacteria.Bacteria can monitor population density and regulate gene expression through QS [15].There are two typical QS systems in Gram-positive bacteria.One QS system uses the autoinducing peptides (AIP) as signal molecules, and cyclic lactone-peptides are one group of AIP [18].Two potential BGCs for the cyclic lactone autoinducer were found in the genome, suggesting that the strain contains this type of QS system.The other QS system uses autoinducer-2 (AI-2) as a signal molecule.The biosynthesis of AI-2 depends mainly on the expression of LuxS, which is widespread among both Gram-positive and Gramnegative bacteria [19].Therefore, LuxS is the core of the AI-2-mediated QS system.In this study, a gene encoding LuxS was found in the genome.This indicates that the AI-2 mediated QS system exists in Paenibacillus sp.lzh-N1.The exact roles of QS systems in biofilm formation and whether QS can regulate the biosynthesis of antimicrobial peptides will be further elucidated.

Conclusions
In the present study, a strain, Paenibacillus sp.lzh-N1, that could prevent pear leaf spots was isolated from the rhizosphere soil of pear trees.Analysis of its genome revealed the genetic foundation of its biocontrol properties, that is, antifungal peptides and glucanase might be responsible for the antagonistic effects of this strain.Moreover, potential QS systems were found in the bacterium, which might influence the biocontrol activity of Paenibacillus sp.lzh-N1 directly or indirectly.The potential antifungal molecules identified in this study lay a foundation for further studies on the genetic and biochemical pathways of Paenibacillus sp.lzh-N1 to improve its antagonistic efficiency and commercial application.

Strain isolation
Paenibacillus sp.lzh-N1 was isolated from the rhizosphere soil samples of pear trees with brown spots in Taian, Shandong Province of China.This strain was chosen due to its antagonistic effects against the pathogen.The rhizosphere soil samples were collected and delivered to the laboratory at low temperature.The samples were diluted with sterile H 2 O and spread on the plates of LB and Gorodkowa′s medium.Colonies were selected after 24 h of incubation, and then inoculated on the plate of M. sentina (Fr) Schrorter.The colonies that could inhibit the growth of the pathogen were screened.

Evaluation of antagonistic ability
The antagonistic ability of Paenibacillus sp.lzh-N1 was tested against M. sentina (Fr) Schrorter as follows.M. sentina (Fr) Schrorter was incubated on a plate of PDA medium at 30℃ for 48 h.Fresh colonies of Paenibacillus sp.lzh-N1 were inoculated on the sides of M. sentina (Fr) Schrorter and incubated at 30℃ for another 48-72 h.

Genome sequencing, assembly and annotation
The genomic DNA of Paenibacillus sp.lzh-N1 was extracted and sequenced using the PacBio Sequel system.Raw data were filtered and assembled using SPAdes software (version 3.9.0(10)) to generate 1,307 Mb of total clean data, with a genome coverage of 232.0 ×.Gene annotation was performed using the NCBI Prokaryotic Genomes Automatic Annotation Pipeline.The function of each gene was further analyzed using four databases (Kyoto Encyclopedia of Genes and Genomes, KEGG; Gene Ontology, GO; COG; and Swiss-Prot).tRNA, rRNA, and sRNA were analyzed using the tRNAscan-SE 1.3.1,rRNAmmer 1.2, and Rfam, respectively.The circular genome map was constructed by TBtools [20], including general genome features and gene function annotation analyzed by KEGG [21][22][23].The COG distribution was conducted by the Bacterial Pan Genome Analysis tool (BPGA) [24].The potential BGCs were predicted with antiSMASH version 6.1.1 (http://antismash.secondarymetabolites.org/) [25].

Phylogenetic analyses
Total genomic DNA for the analysis of 16 S rDNA sequence was isolated using the Bacterial Genomic DNA Isolation Kit (Tiangen, China) according to the manufactuter's instructions.The 16 S rDNA sequence was amplified using the universal primers 27 F and 1492R as described previously [26].The purified amplicons were sequenced by Sangon Biotech (Shanghai, China).The obtained 16 S rDNA sequence was compared with those available from the GenBank database (www.ncbi.nlm.nih.gov/genbank/).The phylogenetic trees based on 16 S rDNA and entire genome sequences were constructed with some species of the genus Paenibacillus by the approximately-maximum-likelihood method using FastTree software [27].Whole genome sequences of the strains were downloaded from the GenBank database (https://www.ncbi.nlm.nih.gov/genome), and the GenBank accession numbers for these 16 S rDNA and genome sequences are listed in Table 3.

Fig. 4
Fig. 4 Phylogenetic tree of Paenibacillus sp.lzh-N1 based on 16 S rDNA sequences.The numbers in each branch points denote the percentages supported by bootstrap; Bar = 0.2% nucleotide divergence

Fig. 6
Fig. 6 Heatmap analyses of ANI (A) and DDH (B) values.The color represents the ANI or DDH values.The green indicates the highest ANI or DDH value, and the red indicates the least ANI or DDH value

Table 1
The general genome feature of Paenibacillus sp.lzh-N1

Table 2
The potential gene clusters encoding secondary metabolites in Paenibacillus sp.lzh-N1

Table 3
The GenBank accession numbers of 16 S rDNA and genome sequences for Paenibacillus spp