Introduction

The genus Pseudomonas in the family Pseudomonadaceae is one of the most ubiquitous environmental bacterial genera due to its metabolic versatility. Some strains can produce water-soluble fluorescent pigments. In addition, the genus Pseudomonas has received much attention as biological control agents (Raaijmakers and Mazzola 2012). Several studies have suggested that antimicrobial compounds produced by Pseudomonas spp. are involved in the biological control of plant diseases. P. corrugate GFBP 5454 and P. mediterranea GFBP 5447 inhibited the growth of plant pathogenic fungi and bacteria due to the production of diffusible compounds and cyclic lipopeptides (Strano et al. 2017). P. fluorescens 2P24 produced a variety of antimicrobial compounds, and 2,4-diacetylphloroglucinol (2,4-DAPG) is the key determinant of its protection against various soilborne diseases (Wei et al. 2004). Some biocontrol agents may produce a wide range of antifungal compounds. For example, P. chlororaphis PA23 produces multiple secondary metabolites, including phenazine-1-carboxylic acid (PCA), PRN, HCN, proteases, lipases and siderophores, some of which have been shown to contribute to antagonism (Poritsanos et al. 2006). The antibiotics 2,4-DAPG and PLT are essential for the biocontrol of soil borne diseases of cotton caused by Rhizoctonia solani and Pythium ultimum by P. fluorescens Pf-5 (Nowak-Thompson et al. 1994; Howell and Stipanovic 1980). Hu et al. (2005) reported that PLT and PCA are responsible for the ability of Pseudomonas sp. M18 to suppress many soilborne phytopathogenic fungi. PCA derived from strain M18 has been registered as a new biogenic pesticide to prevent sheath blight of rice in 2011, illustrating the potential prospects of biocontrol pseudomonads in agricultural production.

The antibiotic 2,4-DAPG, a phloroglucinol derivative, can suppress the activities of many bacteria, fungi, oomycetes, and nematodes (Haas and Defago 2005). The biosynthesis of 2,4-DAPG begins with the synthesis of phloroglucinol (PG) by the type III polyketide synthase PhlD (Achkar et al., 2005). PG is subsequently acetylated through the action of PhlABC to form DAPG. These four structural genes constitute a single operon phlACBD (Bangera and Thomashow, 1999). Pyoluteorin is a broad spectrum polyketide that is composed of a resorcinol and a pyridine containing two chlorine atoms. The structural genes pltLABCDEFG for biosynthesis constitute a single operon (Nowak-Thompson et al. 1999). The Pro-S-PltL intermediate is oxidized by PltE and chlorinated by the halogenase enzymes PltA to synthesize 4,5-dichloropyrrolyl-S-PltL (Dorrestein et al. 2005). PltA and PltD are FADH2-dependent halogenases that chlorinate secondary metabolites. PltD lacks a putative conserved FAD-binding site and is likely not functional. Tn5 insertion in pltD disrupts pyolutrorin production, suggesting a role of pltD in synthesis of the antibiotic (Nowak-Thompson et al. 1999). The PltD is a rate-limiting enzyme in PLT biosynthesis. The pyoluteorin production of pltD overexpression mutant was increased by 77.5% (Li et al. 2012). The physiological role of pltD in pyoluteorin assembly is not clear. Mutation in genes for antibiotics biosynthesis has provided strong evidence of the role of these antibiotics in pathogen suppression.

Pseudomonas fluorescens FD6 has shown antagonistic activity against various plant pathogenic fungi. The mechanisms of action for the biocontrol activity of strain FD6, which include 2,4-DAPG, PLT, PRN, extracellular proteases, siderophores and HCN, have been identified (Chang et al. 2011). Certain isolates of the Pseudomonas genus produce 2,4-DAPG but not PLT (Wei et al. 2004; Weller 2007). Ramette et al. (2011) proposed the new species name P. protegens for these fluorescent Pseudomonas strains, which could produce both 2,4-DAPG and PLT. At present, only certain strains of P. protegens are known to produce these two antibiotics (Loper et al. 2012). Therefore, P. fluorescens FD6 was redesignated as P. protegens FD6 based on the types of antibiotics it produces. Only indirect evidence was analysed to determine the role of antibiotic production by FD6 in its biocontrol efficacy (Chang et al. 2014). To shed some light on the biocontrol traits of P. protegens FD6, its genome was sequenced and genomic analysis of secondary metabolites was performed. The aim of this study was to investigate the role of the pltD and phlC genes in biocontrol activity of P. protegens FD6 on the basis of the genomic analysis.

Materials and methods

Strains, plasmids, and growth conditions

The strains, plasmids, and primers used in this study are listed in Table 1. Escherichia coli strains were grown in Luria-Bertani (LB) media with shaking overnight at 37 °C. P. protegens strains were cultured in LB liquid media at 28 °C. Antibiotics were used as necessary at the following concentrations: kanamycin at 50 μg/mL and ampicillin at 50 μg/mL.

Table 1 Strains, plasmids and primers used in this study

Genomic sequencing, gene annotation and bioinformatic analysis

The complete genome sequence of P. protegens FD6 was performed with Pacific Biosciences RS II (Pacific Biosciences, Menlo Park, CA, USA) sequencing platform, providing a coverage depth of approximately 175. De novo assembly of PacBio Reads was carried out with the RS_HGAP_Assembly v3.0 protocol included in the SMRT analysis pipeline v2.3, and additional assembly was executed by Minimus2. Open reading frames were identified and annotated automatically with GeneMarks and proceeded with manual curation. RNAs were predicted using ARAGORN, tRNAscan-SE, RNAmmer and Barrnap (Laslett and Canback 2004; Lowe and Eddy 1997; Lagesen et al. 2007). Other non-coding RNAs were predicted and classified by Infernal and Rfam (Nawrocki and Eddy 2013; Nawrocki et al. 2015). The genome sequence of strain FD6 was retrieved from the GenBank database and submitted to RAST and AntiSMASH 4.0 (Weber et al. 2015) for genomic analysis of secondary metabolites. The orthologous coding sequences (CDSs) of each group were identified using Inparanoid software.

DNA manipulations and sequence analysis

Plasmid DNA isolation and other molecular assays were performed using standard protocols (Russell and Sambrook 2001). Plasmids were introduced into Pseudomonas strains via electroporation (Wei and Zhang 2006). DNA sequencing was performed by Nanjing Qing Ke Biological Technology Co., Ltd. (Nanjing, China). Nucleotide sequences were analysed with BLASTn databases.

Construction of P. protegens mutations in strain FD6

To delete pltD from the chromosome of strain FD6, two DNA fragments flanking the pltD gene were PCR amplified using primers pltF1–240/pltR1–1291 and pltF2–2164/pltR2–2748 (Table 1). These two fragments were fused together by PCR and digested using BamHI/HindIII to generate a 1635 bp DNA fragment containing pltD with an 873 bp internal deletion. This DNA fragment was ligated to pEX18Km to create the construct pEX18-ΔpltD. This plasmid was integrated into the chromosome of strain FD6 by electroporation. A second cross-over mutant was isolated by loss of kanamycin resistance. An analogous gene replacement strategy was performed to obtain the 541 bp ΔphlC deletion mutant. To generate the phlC mutant ΔphlC, two DNA fragments flanking the phlC gene were PCR amplified using primers phlA138-F1/phlC1222-R1 and phlC1763-F2/phlB2743-R2 (Table 1). These two fragments were fused together by PCR and digested using BamHI/HindIII to generate a 2064 bp DNA fragment containing phlC with a 1084 bp internal deletion. The transformation and selection of this deletion construct, pEX18-ΔphlC, were performed as described above. The deletion of phlC and pltD was confirmed by PCR using primers pltF1–240/pltR2–2748 and phlA138-F1/phlB2743-R2, respectively. The resulting PCR products were also sequenced to confirm that the products were incorporated into the pltD and phlC genes.

Antifungal assays

Radial diffusion assays were performed to assess fungal inhibition in vitro. A 5 mm fungal plug of B. cinerea or M. fructicola was placed on the centre of a plate, and aliquots (3 μL) of overnight bacterial cultures were spotted on potato sucrose agar plates, 2.5 cm away from the fungal plug. The plates were incubated at 25 °C, and the antifungal activity was assessed after 5–7 d by measuring the distance between the edges of the colony and the fungal mycelium. Three replicates were performed for each strain, and the antifungal assays were repeated twice.

Extraction and detection of PLT, 2,4-DAPG and its phloroglucinol

Extraction and HPLC (high-performance liquid chromatography) detection of 2,4-DAPG and PLT from P. protegens FD6 were performed as described previously (Liu et al. 2010; Costa et al. 2009; Zhang et al. 2015). Antibiotics extracted from the culture suspension were analysed by C18 reverse-phase high-performance liquid chromatography (HITACHI L-2000). Aliquots of 10 μL were injected directly into the HPLC system for the determination of PLT and 2,4-DAPG. Three replicates were used for each treatment, and the experiment was performed twice. In the HPLC analysis, standard 2,4-DAPG and PLT were used as controls. PG, a 2,4-DAPG intermediate, quantification was performed as previously described (Kidarsa et al. 2011). Briefly, 150 μL bacterial supernatants was collected, and then 50 μL concentrated HCl and 200 μL 0.2% 4-hydroxy-3-methoxy-cinnamaldehyde were added. The mixture was allowed to react for 5 min, then recording the absorbance at 600 nm.

Detection of other extracellular metabolites

Extracellular protease was assessed using skim milk agar plates (Reimmann et al. 1997). Assays for HCN production on KB agar medium were performed qualitatively using HCN-indicator paper (Castric and Castric 1983). Qualitative siderophore assays were performed on a siderophore detection plate as previously described (Schwyn and Neilands 1987). These tests were repeated twice.

Efficacy of P. protegens FD6 and its derivatives in controlling tomato grey mould disease

Tomato cultivars “Sufen” were harvested from Shatou Greenhouse in Yangzhou and quickly brought to the laboratory. To evaluate the biocontrol effect of P. protegens FD6 and its derivatives on controlling postharvest decay, we determined the development of tomato grey mould disease after treatment with different bacterial cultures as previously described by Zhang et al. (2018). There were three replicate trials of two fruits. The test was repeated twice.

GenBank accession

The complete annotated genomic sequence of P. protegens FD6 is available under GenBank accession number CP031396. Accession numbers for the complete genome sequences of Pf-5 and 2P24 are CP000076 and NZ_CP025542, respectively.

Statistical analysis

Data were analysed and compared by performing Fisher’s least significant difference (LSD) test (p < 0.05 was considered significant) using SPSS software.

Results

Genome features and comparative genomics

General characteristics of the genomes of P. protegens FD6 and Pf-5, and Pseudomonas fluorescens 2P24

The P. protegens FD6 genome is composed of one circular chromosome of 6.7 Mb. A total of 6264 CDSs were identified within the P. protegens FD6 genome. The genome of P. protegens FD6 is larger than that of the biocontrol strain P. fluorescens 2P24 isolated from wheat take-all decline soil (Zhang et al. 2014). However, the genome of P. protegens FD6 is smaller than that of P. fluorescens Pf-5 isolated from soil in College Station, Texas, USA (Kraus and Loper 1995). The complete genome sequence included one phage region, and no insertion element was identified. No genomic islands or CRISPR units are present in the genome. A total of seven 5S rRNAs are present on the FD6 chromosome, whereas the genomes of both Pf-5 and 2P24 contain six 5S rRNAs (Table 2). The complete genomes of Pf-5 and 2P24 were performed by Sanger sequencing and early Pacbio sequencing. These two sequencing systems may misassemble genomes due to read length limitations and poor assembly accuracy for series repeats. Three-generation PacBio sequencing was applied in the study of the complete genome sequence of FD6. All protein sequences from the three species were clustered into 3431 orthologous groups through multiparanoid. Based on additional analyses, 511 genes unique to P. protegens FD6 were identified that participate in many biological processes, such as signal transduction and amino acid and carbohydrate metabolism (Fig. 1, Fig. S1).

Table 2 General characteristics of the genomes of Pseudomonas fluorescens 2P24 and P. protegens Pf-5 and FD6
Fig. 1
figure 1

Venn diagram showing the number of predicted proteins in Pseudonmonas protegens FD6 with significant homology to the predicted proteins from the sequenced genomes of P. protegens Pf-5 and P. fluorescens 2P24. The orthologous CDSs for each group were identified using Inparanoid software. This diagram was created using Venn Diagram

Secondary metabolite analysis

The results from antiSMASH 4.0 showed twelve secondary metabolite gene clusters in the genome of P. protegens FD6 (Fig. 2). These include gene clusters for PLT, 2,4-DAPG, PRN and HCN. The gene cluster responsible for the biosynthesis of the antibiotic 2,4-DAPG, phlHGFACBDE, was identified in the FD6 genome. The structural genes pltLABCDEFG are also present in the PLT biosynthetic gene cluster. Two gene clusters for the siderophores pyoverdine and pyochelin are present, and these genes are conserved in some Pseudomonas spp. isolates (Gross and Loper 2009). Two previously unknown gene clusters encoding secondary metabolites were identified. Orfamide synthases, which contain three structural genes named ofaA, ofaB and ofaC, are present in the whole genome of P. protegens FD6. Orfamide lipopeptides produced by P. protegens show different biological activities against phytopathogenic fungi (Ma et al. 2016). Another 5964 bp gene cluster that contains six ORFs, MbnTPHABC, was identified as the methanobactin gene cluster. The operon includes the genes mbnA, mbnB and mbnC, which are involved in the biosynthesis of methanobactin, mbnT, which is involved in export of methanobactin, and mbnP and mbnH of unknown function (Kenney et al. 2018). Dispirito et al. (2007) reported that methanobactin isolated from methanotrophic bacteria had antagonistic activity against gram-positive bacteria (Dispirito et al. 2007). However, orfamide and methanobactin have not been extracted from P. protegens FD6 culture filtrate or cells. P. protegens FD6 and Pf-5 had seven common compound biosynthesis clusters, and the gene clusters for methanobactin and rhizoxins are unique to FD6 and Pf-5, respectively. Rhizoxins from cultures of P. fluorescens Pf-5 have demonstrated antifungal, phytotoxic, and cytotoxic properties (Loper et al. 2008). The gene clusters for syringomycin, fragin and mupirocin are unique to P. fluorescens 2P24 (Table S1).

Fig. 2
figure 2

Circular representation of the genome of Pseudomonas protegens FD6. The outer scale designates coordinates in base pairs. The two circles show predicted genes on the plus strand and minus strand. The third circle shows twelve secondary metabolite biosynthesis gene clusters, with the structures and names of the corresponding compound indicated with arrows. Three gene clusters whose metabolic products are unknown are indicated with a question mark. The genomic analysis of secondary metabolites was performed using AntiSMASH bacterial version (https://antismash.secondarymetabolites.org/#!/start), and the artwork was created by Circos software

Genes involved in plant growth promotion

Pseudomonas sp. UW4 is a plant growth-promoting bacterium that can grow under different environmental stresses. A number of genes involved in plant growth promotion were identified in Pseudomonas sp. UW4, such as indole-3-acetic acid (IAA) synthesis and acetoin biosynthesis (Duan et al. 2013). A search for FD6-, 2P24- or Pf-5-like IAA pathway-related genes revealed the presence of four orthologous genes in these three strains, suggesting similar pathways in comparison to UW4. Certain Pseudomonas isolates promote plant growth through the production of a volatile compound, acetoin (Ryu et al. 2003). The associated genes, including acetolactate synthase and zinc-containing alcohol dehydrogenase, were identified in the genomes of FD6, Pf-5, and 2P24. Pyrroloquinoline quinine (PQQ) is a plant-growth promotion factor, and biosynthetic genes for PQQ are clustered in an operon pqqABCDEF (Choi et al. 2008). This complete operon was present in the genomes of Pf-5, FD6, and 2P24. The deaminase 1-aminocyclopropane-1-carboxylate (ACC) produced by Pseudomonas strains promotes root elongation and controls plant diseases (Wang et al. 2000). The genome of 2P24 contained the putative ACC deaminase structural gene acdS and its regulatory gene acdR, whereas these two genes are absent from both the Pf-5 and FD6 genomes (Table 3).

Table 3 Putative genes involved in plant growth promotion in Pseudomonas spp. genomes

Rhizosphere colonization

Biocontrol agents usually show certain competitive colonization traits, including motility and the ability to attach to the root surface. Many genes related to chemotaxis and motility were found in the genomes of the three Pseudomonas isolates. FD6 contained eight genes associated with chemotaxis traits. In this study, we also found eight genes responsible for flagella biosynthesis, such as the fli and flh operons in the genomes of Pf-5, FD6, and 2P24. Attachment to the root surface is another competitive colonization trait (Shen et al. 2013; Periasamy et al. 2015). A large number of genes involved in attachment were predicted in the pseudomonad genomes, such as genes associated with type IV pili, adhesion, and agglutination (Table S2).

Construction and isolation of P. protegens FD6 derivatives

The details of the construction of the pltD and phlC gene mutations are shown in Fig. 3. The 1051 bp upstream and 584 bp downstream regions of pltD were amplified by PCR from chromosomal DNA of P. protegens FD6 and fused together by overlap PCR. The resulting fragment was used to construct the pEX18-ΔpltD recombinant plasmid (Fig. 3a). More than 4000 transconjugants were screened for in-frame deletions in pltD on LB agar medium and LB agar medium supplemented with kanamycin. From this library, one mutant, ΔpltD, was obtained and identified by PCR. Using the primers pltF1–240 and pltR2–2748, a 1635 bp fragment with an 873 bp in-frame deletion of pltD was obtained as expected.

Fig. 3
figure 3

Construction scheme of recombinant plasmids pEX18-pltD and pEX18-phlC. The bars designate the fragments cloned into the vector pEX18 to give pEX18-ΔpltD and pEX18-ΔphlC. *, restriction enzyme; Δ, gene deletion

An analogous gene replacement strategy was performed to screen the phlC deletion mutant (Fig. 3b). A ΔphlC deletion mutant was obtained from 12,000 transconjugants. Using the primers phlA138-F1 and phlB2743-R2, a 2064 bp fragment with a 541 bp in-frame deletion of phlC was obtained as expected.

Role of PLT and 2,4-DAPG in growth inhibition of phytopathogenic fungi

To further clarify the relative importance of 2,4-DAPG and PLT production by the FD6 strain in the inhibition of fungal growth, the ΔpltD and ΔphlC mutants were assayed for growth inhibition of the pathogens B. cinerea and M. fructicola. The results showed that the ΔphlC mutant showed reduced inhibition of M. fructicola. However, the ΔpltD mutant had reduced antagonistic activity compared to the wild-type strain, and no fungal inhibition was observed suggesting that pyoluteorin is required for inhibition (Fig. 4).

Fig. 4
figure 4

Inhibitory activity of Pseudomonas protegens FD6 and its derivatives against Botrytis cinerea and Monilinia fructicola. Error bars represent standard deviation, and different letters indicate statistically significant differences (p < 0.05). FD6 is the wild strain of P. protegens FD6, ΔpltD is the mutant defective in production of pyoluteorin, and ΔphlC is the mutant defective in production of 2,4-diacetylphloroglucinol

Detection of antifungal compound production in tested Pseudomonas strains

As shown in Fig. 5, we clearly observed that ΔphlC did not biosynthesize 2,4-DAPG in the PSA medium, whereas the PLT yield markedly increased by more than sixtyfold compared with wild-type FD6. The PLT-deficient strain ΔpltD could not produce PLT, and 2,4-DAPG production dropped by approximately sixteenfold compared to strain FD6 (Fig. 5a, b). Also, the PG production of mutant ΔpltD and ΔphlC was significantly decreased compared with wild-type FD6 (Fig. 5c).

Fig. 5
figure 5

Detection of pyoluteorin (a), 2,4-DAPG (b) and phloroglucinol (c) production in Pseudomonas strains. P. protegens FD6 and variants were assayed for 2,4-DAPG and pyoluteorin concentration by HPLC (Zhang et al. 2015). Phloroglucinol was quantified by colorimetric method (Kidarsa et al. 2011). The experiment was repeated twice. Error bars represent standard deviation, and different letters indicate statistically significant differences (p < 0.05). FD6 is the wild strain of P. protegens FD6, ΔpltD is the mutant defective in production of pyoluteorin, and ΔphlC is the mutant defective in production of 2,4-diacetylphloroglucinol

In addition to secondary metabolite biosynthesis, other biocontrol factors were detected, which are presented in Table 4. The results showed that there was no difference in the production of HCN, proteases or siderophores between ΔpltD and ΔphlC compared with strain FD6.

Table 4 Phenotypic characterization of Pseudomonas protegens FD6 and its derivatives

Efficacy of strain FD6 and its derivatives in controlling grey mould of tomato

Treatment of bacterial cultures of strain FD6 showed an obvious effect for controlling grey mould in tomato fruit. The largest lesion diameters were observed on tomato treated with LB broth. There were no visible disease lesions at 5 days post inoculation after treatment with strain FD6. The lesion diameters of grey mould were inhibited by 65% and 30% after treatment with ΔphlC and ΔpltD bacterial cultures, respectively (Fig. 6).

Fig. 6
figure 6

Biocontrol effect of Pseudomonas protegens FD6 and its derivatives on Botrytis cinerea. Each treatment was composed by three replicates with two fruits per replicate. Error bars represent standard deviation, and different letters indicate statistically significant differences (p < 0.05). FD6 is the wild strain of P. protegens FD6, ΔpltD is the mutant defective in production of pyoluteorin, and ΔphlC is the mutant defective in production of 2,4-diacetylphloroglucinol

Discussion

A wide range of biocontrol agents, such as P. fluorescens A506 and 1629RS, has been in commercial production to control fungal and bacterial diseases in crops over the past decade (O’Brien 2017). At least seven types of compounds, including pyoluteorin, phloroglucinols, phenazines, pyrrolnitrin, lipopeptides, siderophores, and hydrogen cyanide, which were obtained from some isolates of Pseudomonas spp., have received attention because they naturally suppress multiple fungal pathogens (Liu et al. 2015; Yan et al. 2017). Both 2,4-DAPG and PRN have been found to exist alone in one antibiotic-producing bacterium so far as we know. However, PLT is produced with other antibiotics in several isolates, and no isolate has been obtained that only produces PLT (Liu et al. 2006).

Our previous study showed that ΔprnA mutant was as effective as the wild-type strain FD6 at inhibiting hyphal growth, suggesting that inhibition is not due to production of pyrrolnitrin (Zhang et al. 2016). In addition to pyrrolnitrin, 2,4-DAPG is another antibiotic produced by P. protegens FD6. To determine the role of PLT and 2,4-DAPG in strain FD6, we generated ΔpltD and ΔphlC mutants with an in-frame deletion of pltD and phlC, respectively. The HPLC data indicated that the PLT yield in ΔphlC was markedly increased compared with strain FD6. The pltD gene mutation led to a decrease in 2,4-DAPG production compared with strain FD6. In agreement with the above research on the quantity of antibiotics, ΔphlC mutant still displayed weak inhibition of M. fructicola and B. cinerea due to overproduction of pyoluteorin. It has been demonstrated that PG is converted into PG-Cl and PG-Cl2, which serves as signals to activate the expression of pyoluteorin biosynthetic genes (Yan et al. 2017).

The results of the HPLC analysis show that both ΔpltD and ΔphlC mutants produce less phloroglucinols. The low level of phloroglucinols biosynthesis was also observed in the ΔphlA mutant, a derivative of Pf-5. The decrease of PG production in the ΔphlC mutant might be attributed to its inability to acetylate PG to MAPG and DAPG, along with the overproduction of pyoluteorin, leading to loss of auto-induction of the phlACBD operon by DAPG (Kidarsa et al. 2011). These results suggested that the absence of PLT influenced the production of another antibiotic: 2,4-DAPG in strain FD6. The pltD mutant which produces no pyoluteorin or 2,4-DAPG, still exhibits certain biological control activity against tomato grey mould disease. We supposed P. protegens FD6 promotes the biocontrol activity through other mechanisms, such as induction of systemic resistance and competition.

In some biocontrol agents that produce a number of antibiotics, mutants lacking one antibiotic might not lose their inhibitory ability due to homeostatic regulatory systems or metabolic co-regulation. Metabolic co-regulation is usually observed in microorganisms and plays a great role in microbial interactions. Moreover, there is a metabolic co-regulation between the biosynthetic pathways for 2,4-DAPG and PLT in P. protegens Pf-5. Phloroglucinol, as an intermediate in the 2,4-DAPG biosynthetic pathway, is converted into mono- and di-chlorinated phloroglucinols by a halogenase that is encoded in one of the pyoluteorin gene clusters. The chlorinated phloroglucinols serve as signals that induce the expression of PLT biosynthetic genes (Yan et al. 2017). In short, 2,4-DAPG repressed the biosynthesis of PLT in P. protegens FD6. However, it is not clear why the pltD mutation led to the reduction in 2,4-DAPG production. The current study indicated that 2,4-DAPG and PLT play a crucial role in P. protegens FD6 biocontrol traits.

Pseudomonads promote their biocontrol activities through the production of multiple secondary metabolites with distinct roles in competitive and cooperative microbial interactions (O’Brien 2017). Orfamide-type cyclic lipopeptides consist of 10 amino acids and three fatty acid residues, and show different biological activities against phytopathogenic fungi (Gross and Loper 2007; Ma et al. 2016). BLAST searches revealed that P. protegens FD6 may be a potential producer of orfamide and methanobactin, but this prediction has yet to be experimentally verified. In addition, P. protegens FD6 displayed stronger antagonistic activity against B. cinerea in potato agar than in potato dextrose agar or potato sucrose agar. The MALDI-TOF analysis showed that P. protegens FD6 may produce another antifungal lipopeptide (data not shown). More research on this compound is underway.