Characteristics of biological control and mechanisms of Pseudomonas chlororaphis zm-1 against peanut stem rot

Peanut stem rot is a serious plant disease that causes great economic losses. At present, there are no effective measures to prevent or control the occurrence of this plant disease. Biological control is one of the most promising plant disease control measures. In this study, Pseudomonas chlororaphis subsp. aurantiaca strain zm-1, a bacterial strain with potential biocontrol properties isolated by our team from the rhizosphere soil of Anemarrhena asphodeloides, was studied to control this plant disease. We prepared extracts of Pseudomonas chloroaphis zm-1 extracellular antibacterial compounds (PECEs), determined their antifungal activities by confrontation assay, and identified their components by UPLC-MS/MS. The gene knockout strains were constructed by homologous recombination, and the biocontrol efficacy of P. chlororaphis zm-1 and its mutant strains were evaluated by pot experiments under greenhouse conditions and plot experiments, respectively. P. chlororaphis zm-1 could produce extracellular antifungal substances and inhibit the growth of Sclerotium rolfsii, the main pathogenic fungus causing peanut stem rot. The components of PECEs identified by UPLC-MS/MS showed that three kinds of phenazine compounds, i.e., 1-hydroxyphenazine, phenazine-1-carboxylic acid (PCA), and the core phenazine, were the principal components. In particular, 1-hydroxyphenazine produced by P. chlororaphis zm-1 showed antifungal activities against S. rolfsii, but 2-hydroxyphenazine did not. This is quite different with the previously reported. The extracellular compounds of two mutant strains, ΔphzH and ΔphzE, was analysed and showed that ΔphzE did not produce any phenazine compounds, and ΔphzH no longer produced 1-hydroxyphenazine but could still produce PCA and phenazine. Furthermore, the antagonistic ability of ΔphzH declined, and that of ΔphzE was almost completely abolished. According to the results of pot experiments under greenhouse conditions, the biocontrol efficacy of ΔphzH dramatically declined to 47.21% compared with that of wild-type P. chlororaphis zm-1 (75.63%). Moreover, ΔphzE almost completely lost its ability to inhibit S. rolfsii (its biocontrol efficacy was reduced to 6.19%). The results of the larger plot experiments were also consistent with these results. P. chlororaphis zm-1 has the potential to prevent and control peanut stem rot disease. Phenazines produced and secreted by P. chlororaphis zm-1 play a key role in the control of peanut stem rot caused by S. rolfsii. These findings provide a new idea for the effective prevention and treatment of peanut stem rot.


Introduction
Peanut is one of the world's most important oil crop species, as its cultivation area is second only to that of rapeseed, and plays an important role in oil production worldwide. Peanut stem rot is an important soilborne fungal disease occurring in plants that leads to yield losses of many important crop species, including cultivated peanut [2,15]. This fungal disease has been reported in many peanut-producing regions of the world, such as India, the United States, Argentina, Indonesia, the Philippines, Thailand, Vietnam and South Africa [18]. In China, peanut stem rot disease occurred widely in peanut-producing areas of Anhui Province from 1957 to 1959, and the incidence area surpassed 70% in three years [30]. In contrast, in recent decades, the occurrence of peanut stem rot disease has become exacerbated; the distribution area of the disease has been expanding annually; and the disease has been widespread in some producing areas of Shandong, Liaoning, Henan, Guangdong, Jiangxi and other provinces in China [1].
Sclerotium rolfsii, a pathogenic fungus, is the main pathogen causing peanut stem rot disease [13]. Unfortunately, S. rolfsii is very difficult to control, as it produces sclerotia that overwinter in the soil and cause disease in the following season [13,19]. On the other hand, S. rolfsii can cause stem rot in more than 600 plants, especially cash crops such as peanut, onion, cotton, potato, soybean, tomato, wheat, and cucurbits [4,11]; Therefore, it is a pathogenic fungus with diverse hosts [29]. Once peanut plants are infected with S. rolfsii, branch wilting and even whole-plant wilting can occur. Peanut stem rot caused by S. rolfsii is also known as southern stem rot, southern blight, white mould, and Sclerotium rot [29], which causes severe economic losses to current agricultural production and development.
At present, chemical fungicides were often used to control peanut stem rot, and agronomic measures, such as rotation with nonhost crops or coverage of infected crop debris via deep ploughing were also adopt [7]. Although these strategies can be employed to control this disease, it was still easy to get rid of control, and then spread. This is because chemical fungicides are not effective in the late diagnosis of the disease. Horticultural measures can only focus on prevention and cannot deal with the situation of diseased plants, too. For another, S. rolfsii has wide range of hosts, profuse mycelia, abundant persistent sclerotia, and genetic variability, once an infection occurs, it is difficult to eliminate the damage of this disease [25]. Furthermore, applying chemical fungicides causes many serious problems, such as fungicide residue, pathogen resistance and environmental pollution [13]. Hence, we urgently need an effective and eco-friendly control method to prevent the spread of S. rolfsii. The search for biological control of plant pests and diseases holds great promise as a safer and more environmentally friendly alternative to the use of chemical pesticides and others.
Pseudomonas has been reported to have good inhibitory activity against plant pathogens and has proved to be an important candidate of biocontrol strains [28]. Most plant-beneficial Pseudomonad strains were selected for their ability to suppress plant diseases. It was found that these Pseudomonad strains can produce a variety of secondary metabolites with antifungal activity [10]. The antifungal compounds produced by these biocontrol agents included the secretion of phenazine-1-carboxylic acid (PCA), 2,4-diacetylphloroglucinol, pyrrolnitrin, hydrogen cyanide, siderophores, and hydrolytic enzymes such as chitinases, proteases, cellulase, and β-glucanases [3]. According to previous reports, phenazine and its derivatives, a nitrogen-containing heterocyclic redox agents with broad-spectrum activity against gram-positive and gram-negative bacteria and fungi, are produced mainly by Pseudomonads and Streptomyces species [3,12]. However, there are still few reports on biocontrol strains with excellent performance that can be used for the control of peanut stem rot, and the screening of efficient biocontrol strains is still the focus of current research. In this study, to find an effective way to control peanut stem rot plant disease, we aimed to screen biocontrol strains with activities against S. rolfsii. P. chlororaphis has also been reported to be a biocontrol strain of bacteria due to its production of various antibacterial substances, but its biocontrol characteristics and mechanisms are still unclear. In this paper, we further discussed the basic biocontrol of P. chlororaphis zm-1 and its ability to promote the control of peanut stem rot.

Materials and methods
Materials, test strains, and culture media P. chlororaphis zm-1 was isolated from the rhizosphere soil of healthy plants and identified via determining its 16S rRNA sequence and sequence alignment analysis. The derived strains were constructed by homologous recombination. P. chlororaphis zm-1 and its derived strains were cultured in Luria-Bertani (LB) media, which included 10 g of peptone, 5 g of yeast extract, and 10 g of NaCl per litre. S. rolfsii was cultured on potato Keywords: Pseudomonas chlororaphis, Peanut stem rot, Sclerotium rolfsii, Phenazine, Biocontrol efficacy dextrose agar (PDA). The peanut Yuhua 12 used in this experiment is a cultivated variety, which was identified to belong to the Angiospermae, Leguminosae, Arachis, and its scientific name is Arachis hypogaea Linn. Its seeds are from the Plant Germplasm Resources and Genetic Engineering Center of Henan University, and it has been cultivated for 10 years. The peanut plants are grown in a greenhouse using pot culture, and plotbased experiments were carried out by simulating field planting.

Confrontation assay under laboratory conditions
The antifungal activity of P. chlororaphis zm-1 against S. rolfsii (24 h old) was determined by a dual confrontation assay as described by Chen [2]. For details, S. rolfsii was inoculated on PDA plates and cultured at 25°C until the mycelia expanded and covered the whole plate. Agar blocks with a diameter of 6 mm were taken from the activated S. rolfsii culture with a sterilized hole punch and then inoculated onto the centre of a new PDA plate. Logarithmic-stage P. chlororaphis zm-1 (2 μL) was plated at a distance of 2 cm from the centre of the S. rolfsii disc. The suspension of P. chlororaphis zm-1 was inoculated at a distance of 2 cm from the centre of the plate at a volume of 10 μL, and a plate with the same amount of methanol was used as the control. The plates were incubated at 25 ± 2°C for 3 days and monitored regularly after 24 h of incubation for the zone of inhibition. The fungal inhibition rate is equal to the mycelial extension length after inhibition compared to that without inhibition. The differences in fungal inhibition rate were analysed by oneway analysis of variance (ANOVA) followed by Tukey's pairwise post-hoc comparisons.

Preparation of P. chlororaphis zm-1 extracellular compounds extract (PECE)
The extracellular antibacterial compounds of P. chlororaphis zm-1 were extracted from the culture media of the tested strains by the method of Peng et al. [22]. Briefly, cultures were grown for two days in 100 ml of potato dextrose broth (PDB) at 25°C. After cultivation for 3 days, the fermented broth was extracted with an equal volume of ethyl acetate three times. Finally, 10 g of crude extract was harvested after the organic phase was evaporated under vacuum pressure at 37°C. The concentrated crude metabolites were dissolved in methanol, the major compounds were purified by preparative scale chromatography equipped (Waters, USA) with a Reversed-phase C 18 column (5.0 × 250 mm, with 2.5 μm coating layer), and their antibacterial activity was evaluated by confrontation assays.

Components analysed and prepared by HPLC
The components separated and prepared by HPLC (Shimadzu, Japan) refer to the method of previous report [33]. 1mL of sample was loaded, and the temperature of the column box was set to 30 °C. The mobile phase was A: HPLC grade H 2 O (0.1% formic acid); B: HPLC grade acetonitrile. The chromatographic separation program was: 10% acetonitrile elution for 2 min, then was increased to 80% within 10 min, and finally increased to 100% within 3 min. The flow rate was 0.3 mL/min.

Components identified by UPLC-MS/MS
According to our previous report [34], composition analyses were carried out by UHPLC-MS/MS using a Q-Exactive Plus mass spectrometer (Thermo Fisher, Waltham, USA) coupled to a Vanquish Flex system equipped with a Hypersil GOLD column (2.1 × 100 mm, with a 1.9 μm coating layer). 1mL of sample was loaded, and the temperature of the column box was set to 25°C. The A and B mobile phases consisted of H 2 O (including 0.1% formic acid) and acetonitrile, respectively. The chromatographic separation programme was as follows: 10% acetonitrile elution for 2 min, an increase to 80% within 10 min, and, finally, an increase to 100% within 3 min. The flow rate was 0.3 mL/min.
The mass spectrometric conditions were described as previously reports [34]. For detail, The mass spectrometer was operated in full scan mode and in both positive and negative modes in a range of m/z 50-1050. The resolution of the MS data was set to 70,000; the spray voltage was set to 3.2 kV for negative mode; the capillary temperature was set to 320°C; the auxiliary gas heater temperature was 350°C; the sheath gas flow rate was 35 L/h; the auxiliary gas flow was set to 15 L/h; the S-lens RF level was 50; the maximum injection time used was 100 ms; and the resolution of MS/MS acquisition was set to 17,500. The top eight ions in each full scan were isolated within a 1.0 Da window and then fragmented with stepwise collision energies of 20, 40 and 60 units and a maximum injection time of 50 ms, with an automatic generation control target of 10 6 . The acquired raw data were analysed using Compound Discover 3.2 (Thermo Fisher, Waltham, USA) in conjunction with information from several metabolite databases (mzCloud, mzVault, MassList, and ChemSpider).

Construction of the gene deletion strains ΔphzH and ΔphzE
Whole-genome sequencing results (GenBank ID: CP048051) and subsequent gene functional annotations showed that the P. chlororaphis zm-1 genome encodes two proteins, PhzH (Protein ID: WP_038575848.1) and PhzE (Protein ID: WP_081359794.1), which are annotated as a phenazine biosynthesis-related protein and a phenazine-specific anthranilate synthase, respectively. We constructed gene knockout mutants of ∆phzH and ΔphzE by allelic exchange according to a previous report, with some modifications [32]. In detail, two fragments suitable for allelic exchange were created by cloning two EcoRI-HindIII DNA fragments containing locus-specific flanking regions into the EcoRI site of pK18 plasmid. These fragments were created via PCR using the primers listed in Table S1. Ligation of the two fragments led to the precise deletion of the respective open reading frame from the start to the stop codons and to generation of a HindIII site at the locus. Then, the pK18 plasmid was cloned into the Escherichia coli 116 competent cells. The plasmids were digested by EcoRI-HindIII enzymes and then amplified by PCR to verify the correctness of the cloned DNA fragment. The plasmid that contained the correct DNA fragment was demethylated and cloned into E. coli S17-1 by electroporation. The correct DNA fragment was purified again and then transformed into competent P. chlororaphis zm-1 cells. After recovery, the electroporation-competent cells were spread onto LB agar plates with 20 μg/mL kanamycin. After incubation at 30°C overnight, the correct clones were obtained and used for allelic exchange. Then, the correct mutant strains were screened on plates for two types of resistance.

Biological control assays under greenhouse conditions
To evaluate the efficiency of P. chlororaphis zm-1 in controlling peanut stem rot, pot experiments were carried out under greenhouse conditions according to the previous reports by Chen et al. [2] and Dai et al. [5] with some modifications. Specifically, wheat grain sand medium with S. rolfsii were prepared as follows: Wheat grains were put into a triangular flask and soaked in distilled water for 3 h. The water was removed, and a 1/6 volume of sand was added. The mixture was autoclaved and then inoculated with S. rolfsii. Afterwards, the mixture was cultured at 30°C for one week until mycelia grew on all the wheat grains and then stored at 4°C. The peanut seeds were transferred to petri dishes for germination for 12 h and subsequently sown in pots containing 150 g of sterilized sandy soil. After growing for 14 days at 28±1°C, peanut seedlings with consistent growth were chosen for pot experiments. Then, the sand with the wheat grains and with S. rolfsii was ground with an aseptic mortar and sprinkled evenly onto the surface of the culture pots of 14 d old peanut seedlings, at a rate 10 g per pot. A 2 mL aliquot of the P. chlororaphis zm-1 suspension, suspended in sterile water and diluted to 1 ×10 8 CFU/mL, was prepared and evenly sprayed onto the peanut shoots. Sterile water was used as a blank control, and carbendazim was used as a fungicide control. Each treatment involved 10 pots and was performed in triplicate. All the pots were kept at 28°C with the day 14 h and night 10 h in a sunlit greenhouse and watered regularly. After 28 days, all the plants were removed, washed in running tap water to remove soil particles and evaluated for disease incidence.

Statistics and calculations of plant disease
The plant disease incidence under each treatment was calculated as the percentage of diseased plants. Disease severity was scored on a 0-4 scale (0 = peanut growth was normal, and the base of the stem had apparent disease spots; 1 = peanut growth was normal, and there were apparent disease spots at the base of the stem; 2 = less than 25% of the whole peanut plant had wilted or died; 3 = 26~50% of the whole peanut plant had wilted or died; 4 = greater than 50% of the whole peanut plant had wilted or died). The disease index and the relative biocontrol efficacy of the antagonists were calculated as follows [5,14]: The differences in data statistics and calculation were analysed by one-way analysis of variance (ANOVA) followed by Tukey's pairwise post-hoc comparisons.

Plot experiments
Plot experiments were employed to evaluate the biocontrol performance of P. chlororaphis zm-1 and its mutants against peanut stem rot. We carried out experiments according to a previously reported method [27] with some modifications. Specifically, the test plot was divided into 6 sections, each with a size of 4.0 × 3.2 m and surrounded by cement walls, after which the sifted soil was sterilized and added to the cement-divided plots. The soil consisted of Tifton loamy sand (fine-loamy, siliceous and thermic Plinthic Kandiudult; pH of 6.1) artificially infested with S. rolfsii. Peanut seeds were sown in each divided plot at a spacing of 40 × 20 cm. The main plots (12.8 × 7.6 m) consisted of four divided treatment plots and two untreated plots. Data were collected from the treated plots. The relative biocontrol efficacy was ultimately calculated as shown above.

P. chlororaphis zm-1 inhibits the growth of S. rolfsii through extracellular compounds
As shown in Fig. 1, P. chlororaphis zm-1 strongly inhibited S. rolfsii under our laboratory conditions (Fig. 1B). Due to the growth and antagonism of P. chlororaphis zm-1, the mycelium of S. rolfsii extended only 1.38 cm in the transverse direction. At the same time, they extended 3.78 cm along the vertical axis. In the control group without any interference from antagonistic strains, the mycelium of S. rolfsii extended 4.06 cm when they were cultured for the same time (Fig. 1A). PECE was prepared to investigate whether the extracellular products produced by P. chlororaphis zm-1 were the active substances inhibiting the growth of S. rolfsii.
The inhibitory effect of PECE on S. rolfsii is shown in Fig. 1C. The aerial hyphae of S. rolfsii extended only 1.97 cm in the direction where a dish loaded with 30 μL of PECE was placed, but in the undisturbed direction, the hyphae stretched 3.24 cm. As shown in Fig. 1C, PECE significantly (P < 0.05) inhibited the mycelial expansion of S. rolfsii, while methanol, a negative control (NC), had no effect on the expansion of mycelia. These results indicated that P. chlororaphis zm-1 has great potential for use as a biocontrol agent to control peanut stem rot caused by S. rolfsii.

Analysis of the antibacterial components of PECE via HPLC
The components of PECE as analysed by HPLC are shown in Fig. 2. There were 5 principal peaks in the liquid chromatogram. To confirm which component plays a key role in inhibiting the growth of S. rolfsii, we isolated the five components via HPLC and collected them separately (referred to as 1, 2, 3, 4, and 5) for identification of their antifungal activities. The results of our identification experiments were shown in Fig. 3. Components 2, 4 and 5 all inhibited the growth of S. rolfsii. We purchased three standard phenazines, 1-hydroxyphenazine, phenazine-1-carboxylic acid, and the core phenazine compound, each of their retention times was determined under the same chromatographic conditions. On the basis of the determination of retention time compared with the liquid chromatograms of the three standard reagents in Fig. S1, the components 2, 4 and 5 were very likely to be 1-hydroxyphenazine, phenazine-1-carboxylic acid, and the core phenazine compound, respectively.

Identification of phenazines in PECE via UPLC-MS/MS
To define component of PECE, UPLC-MS/MS was used to display the total ion liquid chromatogram. In Fig. 4, more than fifteen compounds were presented in the PECE. There are more than fifteen compounds present in the PECE. According to the above HPLC results, we focused on the determination of phenazine compounds in the PECE and identified three principal components, 1-hydroxyphenazine, phenazine-1-carboxylic acid, and the core phenazine, via alignment with database information. The details of the identification information were shown in Table 1, and the chromatograms obtained by mass spectrometry are shown in Supplementary Document 2. According to the UPLC-MS/MS results, peak 1 could be attributed to 1-hydroxyphenazine or 2-hydroxyphenazine. We further measured the antibacterial activity of standards 1-hydroxyphenazine and 2-hydroxyphenazine and found that 1-hydroxyphenazine could inhibit the growth of S. rolfsii, but 2-hydroxyphenazine could not (Fig. 5). Thus, combined with the experimental results in section 3.2, the peak 1 corresponded to 1-hydroxyphenazine. Therefore, P. chlororaphis zm-1 could produce three phenazines, 1-hydroxyphenazine, phenazine-1-carboxylic acid, and the core phenazine. We proposed that P. chlororaphis zm-1 antagonizes S. rolfsii by producing and secreting phenazine compounds.

Deletion of the phzE or phzH gene blocked phenazine compound production by P. chlororaphis zm-1
According to the whole genome of P. chlororaphis zm-1 (GenBank ID: CP048051) and subsequent gene functional annotations, the P. chlororaphis zm-1 genome contain two gene products of phenazine biosynthesis protein (Protein ID: WP_038575848.1) and phenazine-specific anthranilate synthase (Protein ID: WP_081359794.1), respectively. Through blast on NCBI, we found the phenazine biosynthesis protein is highly homologous (100%) to an anthranilate synthase family protein from P. chlororaphis, named PhzH, which reported to be related to the phenazine-1-carboxylic acid-producing Pseudomonas spp. strains. And the phenazine-specific anthranilate synthase is highly homologous (100%) to phenazine biosynthesis protein from PhzE (100%), which were reported Fig. 4 The total ion liquid chromatogram of the PECE. "1" responds 1-hydroxyphenanzine, "2" responds phenazine-1-carboxylic acid, "3" responds phenazine, all of them were identified by standard control to be a highly conserved phenazine biosynthetic operon both in P. chlororaphis and P. aeruginosa. Therefore, we constructed the two gene mutants of phzH and phzE, and examined the biocontrol effects of P. chlororaphis zm-1 on abnormal synthesis of extracellular phenazine compound. Thus, we analysed the components of their extracellular compounds by UPLC-MS/MS. The total ion liquid chromatograms of the PECE from P. chlororaphis zm-1 and its mutants are shown in Fig. 6. As a result, phzH gene deletion inhibited the production of 1-hydroxyphenazine, but the synthesis of phenazine-1-carboxylic acid and phenazine was unaffected. More interestingly, ΔphzE could not produce any of the three kinds of phenazine compounds. This result indicated that both PhzE and PhzE -especially PhzE -were closely related to the synthesis of phenazine compounds in P. chlororaphis zm-1.

Deletion of the phzE gene prevents P. chlororaphis zm-1 from antagonizing S. rolfsii
To illustrate the mechanism through which P. chlororaphis zm-1 antagonizes S. rolfsii, we further determined the antifungal activities of ΔphzH and ΔphzE, and the results are shown in Fig. 7. Contrast to wild-type P. chlororaphis zm-1 (42.86%), the antifungal activities of the ΔphzH mutant decreased to 19.23%, and ΔphzE completely lost its antifungal activity. These results were in accordance with the results of the composition analysis above. Since ΔphzE could not produce any kind of phenazine compound (Fig. 6C), it lost antifungal activity against S. rolfsii. While ΔphzE lost the abilities to produce 2-OH-phenazine, but it still had PCA production, thus, its antifungal activity was just decreased. Therefore, we conclude that P. chlororaphis zm-1 antagonizes S. rolfsii by producing and secreting phenazine compounds under our experimental conditions.

Biocontrol efficacy of P. chlororaphis zm-1 and its mutants in the greenhouse
The biocontrol efficacy of ΔphzE and ΔphzH in the greenhouse was determined, and the results were shown in Table 2. Compared to the sterile water control and fungicide control treatments, the P. chlororaphis zm-1 and ΔphzH treatments significantly reduced both the disease incidence and severity of peanut stem rot (P < 0.05). The biocontrol efficacy of P. chlororaphis zm-1 was significantly (P < 0.05) higher than that of carbendazim (70.60%), reaching 75.63%. However, the biocontrol efficacy of ΔphzH dramatically declined to 47.21%. The biocontrol efficacy of ΔphzE was reduced to 6.19%, and ΔphzE almost lost its ability to inhibit the growth of S. rolfsii. These results were in strongly in accordance with  Table 2 The biocontrol efficacy of P. chlororaphis zm-1 and its mutants in greenhouse Values are the mean ± SD. Different lowercase letters in the same column indicated a significant difference between the treatments (P < 0.05)

Plots of the biocontrol efficacy of P. chlororaphis zm-1 and its mutants
Because the greenhouse assay results showed differences between P. chlororaphis zm-1, ΔphzH and ΔphzE, we further determined the biocontrol efficacy of ΔphzH and ΔphzE in actual plots, The results in Table 3 were in accordance with the trend of the greenhouse assay. The P. chlororaphis zm-1 and ΔphzH treatments significantly reduced both the disease incidence and the severity of peanut stem rot (P < 0.05), although the biocontrol efficacy of ΔphzH decreased, contrast to wild-type P. chlororaphis zm-1. The biocontrol efficacy of P. chlororaphis zm-1 was significantly (P < 0.05) higher than that of carbendazim (49.87%), reaching 57.09% in the plot experiments. Similarly, the biocontrol efficacy of ΔphzE was reduced to 1.88%; in fact it almost lost its biological control activity.

Discussion
Peanut stem rot is a serious plant disease that causes great economic loss and has become increasingly severe in recent years. Biological control is one of the most promising plant disease control measures. In this study, we screened biocontrol P. chlororaphis strains from the rhizosphere soil of A. asphodeloides and identified P. chlororaphis zm-1 isolate against peanut stem rot. We sequenced its whole-genome sequence which was uploaded to the NCBI database (GenBank ID: CP048051.1). Under laboratory conditions, we confirmed that P. chlororaphis zm-1 had a strong antagonistic effect on S. rolfsii (Fig. 1A). The biological control of plant pathogens by Pseudomonas works through three main types of modes of action: direct interactions with the pathogen via antibiosis; interactions with the host plant via induction of resistance; and competition with the pathogen for nutrients and niches [21]. To investigate whether P. chlororaphis zm-1 antagonizes S. rolfsii by producing extracellular active substances, we extracted the supernatant of P. chlororaphis zm-1 fermentation broth with ethyl acetate and methanol, and identified that PECE significantly inhibited the growth of S. rolfsii, but methanol, a negative control (NC), had no effect on the expansion of mycelia. These results indicated that P. chlororaphis zm-1 has great potential for use as a biocontrol agent to control peanut stem rot caused by S. rolfsii.
Since the extracellular metabolites of P. chlororaphis zm-1 contain anti-micronuclear chemicals, we analysed the components of PECE by HPLC. To confirm which component plays a key role in inhibiting the growth of S. rolfsii, we isolated the five identified components via HPLC and collected them separately (referred to as 1, 2, 3, 4, and 5) for identification of their antifungal activities. Components 2, 4 and 5 all inhibited the growth of S. rolfsii. According to previous reports, phenazines, which compose a class of diffusible, heterocyclic compounds, are among the well-studied secondary metabolites shown to be important for biological control [16,24,31]. With standard phenazines, the three components, 1-hydroxyphenazine, phenazine-1-carboxylic acid (PCA), and the core phenazine compound were identified by UPLC-MS/ MS. According to previous reports, natural phenazine compounds, which compose a class of diffusible, heterocyclic compounds, comprise a large group of pigmented, heterocyclic, and nitrogen-containing aromatic secondary metabolites produced and secreted almost exclusively by eubacteria [8,16]. In another related development, more than 180 different naturally derived phenazines have been identified, and most of them are among the well-studied secondary metabolites shown to be important for biological control [9,16,20,31]. Pseudomonas and Streptomyces have been reported to produce phenazine compounds [26,31], but the role of components phenazine components 2-hydroxyphenazine, PCA, and 1-OH-PCA was unclear. Thus, we speculate that the mechanism through which P. chlororaphis zm-1 antagonizes S. rolfsii may be related to its production of extracellular phenazine compounds. In a previous report by Le et al. [13], phenazine-producing P. chlororaphis strain Phz24 significantly inhibited the hyphal growth of S. rolfsii and suppressed stem rot disease in peanut, which was Table 3 The biocontrol efficacy of P. chlororaphis zm-1 and its mutants in plot Values are the mean ± SD. Different lowercase letters in the same column indicated a significant difference between the treatments (P < 0.05) Our results further confirmed the antifungal activities of 1-hydroxyphenazine, PCA, and phenazine against S. rolfsii. In particular, 1-OH-PHZ produced by P. chlororaphis zm-1 also showed antifungal activities against S. rolfsii, but the previously reported 2-OH-PHZ did not (Fig. 5).
The metabolic regulatory pathways of phenazines have been extensively studied during the past several years. Through comparisons of the phz operons in different species, five conserved sets of enzymes, PhzE, PhzD, PhzF, PhzB, and PhzG, have been confirmed to be present in all phenazine-producing bacteria [8,20,23]. By reports, in most phenazine produced pseudomonas, the phenazine biosynthetic operon is flanked by one or more coding genes whose coding products are responsible for converting PCA to different phenazine derivatives by methylation, transamination, hydroxylation, or decarboxylation [6,16]. According to the report of Müller and Behrendt [21], PhzE plays a key role in the synthesis of phenazine compounds from Pseudomonas spp. In another two reports of Mavrodi et al. [16,17], the genome of the opportunistic human pathogen P. aeruginosa PAO1 contained several phenazine-modifying genes whose products could convert PCA into three additional phenazine derivatives, i.e., PCN, 5-methyl-phenazine-1-carboxylic acid, and 1-hydroxy-phenazine, via the activity of phzH, phzM, and phzS, respectively. PhzH is likely to participate in converting PCA into additional phenazine derivatives [16,17].
We sequenced the whole genome of P. chlororaphis zm-1, the sequencing results (GenBank ID: CP048051) and subsequent gene functional annotations showed that the P. chlororaphis zm-1 genome contain two gene products which are annotated as a phenazine biosynthesis protein (Protein ID: WP_038575848.1) and phenazine-specific anthranilate synthase (Protein ID: WP_081359794.1), respectively. We downloaded their amino acid sequence and blast on NCBI, then, we found that the phenazine biosynthesis protein is highly homologous (100%) to an anthranilate synthase family protein from P. chlororaphis, named PhzH, which could extend the range of biocontrol ability of phenazine-1-carboxylic acid-producing Pseudomonas spp. strains by the previous report of Chin-A-Woeng et al. [3]. And the phenazine-specific anthranilate synthase is highly homologous (100%) to phenazine biosynthesis protein, PhzE, which were reported to be a highly conserved phenazine biosynthetic operon both in P. chlororaphis and P. aeruginosa [31]. Therefore, we constructed the two gene mutants of ΔphzH and ΔphzE, and examined the biocontrol effects of P. chlororaphis zm-1 on abnormal synthesis of extracellular phenazine compounds. First, we constructed two gene mutants, ΔphzH and ΔphzE, and determined their antifungal activities (Fig. 7). In contrast to those of wild-type P. chlororaphis zm-1, the antifungal activities of the ΔphzH mutant decreased, and ΔphzE lost its antifungal activity. We further analysed the composition of their extracellular compounds by UPLC-MS/MS and found that ΔphzE could not produce all three kinds of phenazine compounds and that deletion of the phzH gene inhibited the production of 1-hydroxyphenazine; however, the synthesis of phenazine-1-carboxylic acid and phenazine was unaffected. A previous report indicated that phzH, a phenazine biosynthesis-related gene, was identified in P. chlororaphis PCL1391 and shown to be required for the presence of the 1-carboxamide group of PCN because a phzH mutant of strain PCL1391 accumulated PCA [3]. However, our results showed that the ΔphzH mutant of P. chlororaphis zm-1 also accumulated PCA, but PhzH was required for the production of 1-hydroxyphenazine, which is not quite consistent with the findings concerning P. chlororaphis PCL1391. This is very likely due to the differences in extracellular metabolites of different strains under different culture conditions. Therefore, we conclude that P. chlororaphis zm-1 antagonizes S. rolfsii by producing and secreting phenazine compounds under our experimental conditions.
To confirm the mechanism through which P. chlororaphis zm-1 antagonizes S. rolfsii, we further determined the biocontrol efficacy of ΔphzH and ΔphzE via greenhouse and plot experiments. As a result, P. chlororaphis zm-1 significantly reduced both the disease incidence and the severity of peanut stem rot (P < 0.05) in both greenhouse assay and plot experiments. The biocontrol efficacy of P. chlororaphis zm-1 was significantly (P < 0.05) higher than that of carbendazim. Moreover, ΔphzH treatments also significantly reduced both the disease incidence and the severity of peanut stem rot (P < 0.05), but the biocontrol efficacy of ΔphzH dramatically declined, contrast to the wild-type strain. And ΔphzE mutant almost completely lost the abilities to inhibit S. rolfsii. These results were strongly in accordance with those of the above experiments, which confirmed that P. chlororaphis zm-1 antagonizes S. rolfsii by producing and secreting phenazine compounds.
The biocontrol mechanism of P. chlororaphis zm-1 is an important theoretical basis for its application in the control of peanut stem rot disease. Biocontrol mechanisms are one of the main ways for biocontrol strains to exert their control. In addition, it has been reported that P. chlororaphis can induce hosts to form a self-defence mechanism against pathogens. Whether the biocontrol effect of P. chlororaphis zm-1 on peanut stem rot is also related to this mechanism is still unclear. This will be our main topic in investigate in terms of the degree of the biocontrol of this strain in future experiments.

Conclusion
In this study, we found P. chlororaphis zm-1 can antagonize S. rolfsii, which is the main causal agent of peanut stem rot through producing extracellular antifungal substances, Thus, P. chlororaphis zm-1 could prevent and control peanut stem rot disease. Furthermore, we identified the extracellular antifungal substances were 1-hydroxyphenazine, phenazine-1-carboxylic acid, and the core phenazine compound. Moreover, 1-hydroxyphenazine inhibited the growth of S. rolfsii, but 2-hydroxyphenazine did not. ΔphzE did not produce any phenazine compounds, and its antagonistic ability was almost completely abolished. Relatedly, ΔphzH no longer produced 1-hydroxyphenazine but could still produce PCA and phenazine, and its antagonistic ability declined. Therefore, our results showed phzE in P. chlororaphis zm-1 was essential for production of the active phenazine compounds, while phzH was mainly associated with PCA and phenazine productions. With potting experiments under greenhouse conditions and plot experiments, P. chlororaphis zm-1 all showed the biocontrol ability in controlling peanut stem rot antagonize S. rolfsii. Our findings provide a new idea for the effective prevention and treatment of peanut stem rot.