Introduction

Pseudomonas corrugata was first described by Scarlett et al. [29] as the causal agent of tomato pith necrosis in Britain. The bacteria has subsequently been reported in many other countries, including the United States, Canada, Europe, South Africa, New Zealand, Israel [18], and China [8]. In addition to water [29] and soil [31], P. corrugata was also isolated from many parts of the world, such as pepper [17], tomato [18], chrysanthemum [10], geranium [21] (associated with induced pith necrosis), healthy alfalfa roots [19], and grapevine stem [2], which suggested that the bacterium could be considered ubiquitous.

P. corrugata has shown effective ability in suppressing damping-off in tomato and maize [12, 24], but there have been no reports on controlling tomato grey mildew caused by Botrytis cinerea. In this article, we report the isolation and characterization of a nonpathogenic P. corrugata strain P94 that showed inhibition activity to grey mildew of tomato; we discuss its potential antagonistic activity against phytopathogenic micro-organisms; and we describe the spectrum, the biochemical characteristics, and some phenotypes related to biologic control.

Materials and Methods

Bacterial and Fungal Strains

The antagonistic bacterial strain P94 was isolated from Beijing agricultural soil and identified in this study as P. corrugata. P. corrugata strain ICMP 5819 was kindly provided by Zhao Tingchang (Plant Protection Institute, The Chinese Academy of Agricultural Sciences, Beijing, China). B. cinerea, Ceratocystis fimbriata, Magnaporthe grisea, Rhizoctonia cerealis, Phytophthora capsici, and Alternaria solani were kindly provided by Liu Xili (Department of Plant Pathology, China Agricultural University, Beijing, China). Monilinia laxa, Pythium aphanidermatum, Fusarium oxysporum f. sp. cucumerinum, F. oxysporum f. sp. vasinfectum, and F. oxysporum f. sp. lilii were kindly provided by Qiu Jiyan (Plant and Environment Protection Institute, The Beijing Academy of Agricultural and Forestal Sciences, Beijing, China). P. syringae and Ralstonia solanacearum were kindly provided by Feng Jie (Plant Protection Institute, The Chinese Academy of Agricultural Sciences, Beijing, China). Acidovorax avenae was kindly provided by Dr. Zhang Liqun (Department of Plant Pathology, China Agricultural University, Beijing, China). Agrobacterium rhizogenes K27 and A. vitis K308 were kindly provided by Kerr (Department of Plant Pathology, Waite Agricultural Research Institute, University of Adelaide, South Australia). Xanthomonas oryzae, X. campestris, and A. tumefaciens Cy4 were collected by our laboratory.

Bacterial Isolation and Screening

Soil samples were collected from the rhizospheres of different tomatoes infected with B. cinerea. After removing approximately 5 cm of soil from the surface layer, the samples were placed in sterile plastic bags. The bacteria were isolated from soil according to Mew et al. [22]. Each isolated colony was transferred to a potato dextrose agar (PDA) plate on which was placed a 5-mm mycelial plug of B. cinerea. Bacteria were situated approximately 2.5 cm from the plug. After incubation at 25°C for 4 to 7 days, the most effective strain that inhibited the fungal growth was selected.

Assessment of Antiphytopathogenic Activity In Vitro

The antagonistic ability of P. corrugata P94 strain was checked against 11 phytopathogenic fungi and 9 phytopathogenic bacteria. The inhibition activity of strain P94 against fungi was tested using the modified method of Daayf et al. [9]. Two replicates of 5-µl aliquots of suspension of P94 were spotted 180° apart and 2.5 cm from the center of the PDA plate. A 5-mm mycelial plug of each pathogenic fungus was transferred to the center of the plate. A single plug of fungus placed on a PDA plate without P94 inoculation served as control. The plates were incubated at 25°C, and duplicate diameters of fungal colonies were measured in a beeline of the bacteria spotted. The inhibitory effect of strain P94 against fungus was estimated based on the percent relative growth:

$$ {\rm Relative\,growth}\,(\% ) = (T/C) \times 100, $$

where T is the growth of the fungi when challenged with the strain P94, and C is the growth of the control fungi. The activity against phytopathogenic bacteria was examined using the double-layer method of Stonier [32].

Test for Pathogenicity

Tomato and tobacco seedlings were grown in a greenhouse. P. corrugata strain P94 was inoculated using 2-month-old tomato and tobacco seedlings, 10 each of which were used for bacterial inoculation; P. corrugata strain ICMP 5819 and sterile distilled water served as positive and a negative controls, respectively. The plants were inoculated according the method of Sutra et al. [33]. External and internal symptoms were recorded 35 days after inoculation; each plant was longitudinally cut, and pith necrosis was observed.

Assays for Biologic Control Grey Mildew of Tomato

The inhibition activity of tomato grey mildew by P. corrugata strain P94 was examined using the method of Weeds et al. [36] with some modifications: Tomato leaves of the same age were picked. The leaves were washed thrice with sterile distilled water and placed in a large Petri dish (270 mm) with five layers of humid filter article on the bottom. P94 was cultured for 48 hours at 28°C in liquid potato dextrose (PD) medium, and the bacteria were diluted to optical density (OD)600 of 0.6 (approximately 1.0 × 108 colony-forming units/ml) in sterile buffered saline. Then the bacteria suspension was sprayed onto the surface of tomato leaves. Twenty minutes later, 5-mm mycelial plugs of B. cinerea were inverted onto the upper leaf surface when there were no beads on the leaves. Leaves onto which sterile distilled water was sprayed served as negative control. The plates were closed to maintain high humidity and were incubated for 3 days at 20° to 25°C under a diurnal cycle of daylight and white fluorescent light. The area of disease blot was measured by staffs with small square grids (1 mm2/grid), and grey mildew inhibition activity was calculated as follows:

$$ {{\rm{Grey\,mildew\,inhibition\,(\% ) = [(C }} - {\rm{ T)/C] \times 100,}}} $$

where C is the average area of the disease blot of the negative control group, and T is the average area of the disease blot of tomato leaves treated with P94.

Identification of Strain P94: Biochemical and Physiologic Characterization

The ability to use different carbon sources was tested using the Biolog GN Microplate method (Biolog, Hayward, CA).The morphologic, cultural, biochemical, and physiologic characteristics of P94 were examined according to the methods previously reported by Lopez et al. [18].

16S rDNA Gene Sequencing

Genomic DNA of P94 was extracted and purified according to the method described by Sambrook et al. [28]. 16S rDNA was amplified from P94 genomic DNA with the primers 63F 5′-CAGGCCTAACACATGCAAGTC-3′ and 1494R 5′-GGYTACCTTGTTACGACTT-3′. PCR amplification was performed as follows: one cycle of 5 minutes at 94°C, followed by 30 cycles of 30 seconds each at 94°C, 30 seconds at 56°C, and 1 minute at 72°C, followed by one cycle of 5 minutes at 72°C. PCR products were purified using a gel extraction kit (Omega). The clean 16S rDNA fragments were cloned into pMD18-T (TaKaRa) and sequenced with an ABI 3730 DNA sequencer at the Invitrogen Corporation in Beijing. DNA sequence homology searches were performed using the online BLAST search engine in GenBank (available at: http://www.ncbi.nlm.nih.gov).

Amplification of Specific PCR Products

P. corrugata was identified and distinguished into two genomic groups (I or II) by using two pairs of specific primers (PC 1/1 and PC 1/2 and PC 5/1 and PC 5/2) within a single PCR reaction for amplification of two specific PCR products (either 1100 or 600 bp) [6]. PCR was performed as previously described by Catara et al. [6]. The strain was distinguished by the presence of either of these two possible bands: 1100 (group I) or 600 bp (group II).

Transmission Electron Microscopy

For observation of stain P94 flagella by transmission electron microscopy (TEM), cells were grown on a King’s B (KB) plate and suspended in sterile buffered saline. A small drop of the suspension was placed on a carbon-coated copper grid, and the cells were negatively stained with 0.5% phosphotungstic acid (pH 4.0) for observation under the electron microscope (JEM-100CX; Jeol).

Detection of HCN and Siderophore Production

Production of HCN was observed according to the method of Castric and Castric [5] with the following modification: A single colony of strain P94 and non-HCN producing P. fluorescens strain P12 (as a control) were inoculated into an Eppendorf tube containing 0.5 ml of solid Luria-Bertani medium, then a small piece of detective paper was hung over the medium, and the tube was sealed with Parafilm (Pechiney Plastic Packaging). The tube was incubated at 28°C for 48 hours, and the result was investigated. The change in color to blue of the HCN-sensitive paper indicated the production of HCN. The amount of HCN released by strain P94 was quantified according to the method of Lambert et al. [16]. Siderophore production was determined by chrome azurol S (CAS) assay [30]. A yellow halo around a colony after 4 to 5 days of incubation on blue agar is indicative of siderophore excretion.

Determination of Hydrolytic Activity and IAA Production

Protease production was determined using skim-milk agar plates [15]. Chitinase activity was evaluated on chitin agar (CA) plates [7]. Cellulase production was detected as described by Teather and Wood [34]. Phosphatase activity was detected using Pikovskaya’s agar plates [25], and quantitative analysis of soluble phosphate was assessed according to the method of King [13]. The production of IAA was determined using the method of Bric et al. [3] in nutrient broth with and without tryptophan (0.5 g/l).

Nucleotide Sequence Accession Number

The 16s rDNA sequence of 1462 bp has been deposited in GenBank under accession number EF153018.

Results and Discussion

Six hundred twenty-eight bacterial strains were isolated from agricultural soil samples collected from a greenhouse in Beijing, China. Of all the strains tested, 11 bacterial isolates (5 Pseudomonas sp., 3 Bacillus sp., 2 Rahnella sp., and 1 Achromobacter sp.) inhibiting B. cinerea growth were selected, and the most potent one was strain P94 (data not show). Strain P94 exhibited the most obvious antimicrobial activity in vitro and showed significant growth-inhibitory activity against a range of phytopathogenic fungi and phytopathogenic bacteria (Table 1), and the antifungal activity was efficient compared with that of P. corrugata (inhibition of Pythium ultimum) [11] and P. aeruginosa (inhibition of B. cinerea and M. grisea) [15] previously reported, and the antibacterial activity was as much as that of other antagonistic bacteria [35, 37]. Other isolates in this work differed from P94 regarding their spectrum of antagonism to phytopathogenic fungi and were defective in antibacterial activity (data not shown). However, strain P94 and the P. corrugata strains described previously also showed differences in spectrum-of-inhibition activity against phytopathogenic fungi and bacteria. P. corrugata was isolated from rhizosphere of maize (strains P. corrugata 1 and 7) and tomato P. corrugata strains 1 and 3 as a potential biologic control agent [12, 23]. Strains P. corrugata 1 and 7 inhibited the growth of Alternaria alternata, Fusarium moniliforme, and Rhizoctonia [23], but P94 did not exhibit inhibitory activity against the phytopathogenic fungi mentioned above. The antibacterial activity of strains P. corrugata 1 and 7, as well as tomato strains 1 and 3, was not mentioned in the previous study [12, 23]. Strain P94 may produce bacteriocin because it was only active against closely related classical Pseudomona and was inactive against Xanthomonas sp. and Agrobacterium sp. A common trait of bacteriocins is their narrow spectrum of activity, and bacteriocins are only toxic to bacteria closely related to the producing strain [14]. Previously, many Pseudomonas sp. strains had been reported to show growth inhibition in vitro of pathogenic bacteria by bacteriocin [35].

Table 1 Antagonistic activity of P. corrugata P94 to the indicator phytopathogenic fungi and phytopathogenic bacteria
Full size table

In the pathogenicity test, plants inoculated with P. corrugata strain ICMP 5819 produced typical symptoms of necrosis and pith hollowing, but they showed no difference when inoculated with strain P94 (Fig. 1). Strain P94 showed the ability to protect tomato against B. cinerea in tomato leaf testing in vitro (Fig. 2), and the biologic control effect of P94 was 55.3% by testing 90 tomato leaves (30 leaves treated with P94, 30 leaves as negative control, and 30 leaves treated with water only as blank control). This result was in accordance with recent studies showing that P. fluorescens (strains PTA-268 and PTA-CT2) suppressed grapevine grey mildew caused by B. cinerea (45% to 70%) [20]. We had not previously found the reports about P. corrugata biocontrol of B. cinerea. P. corrugata was isolated from rhizosphere of maize, which had shown effective ability in suppressing damping-off [11, 12, 24], and further work should be done to determine if strain P94 could also suppress damping-off.

Fig. 1
figure 1

Tomato stem split longitudinally to show symptoms of pith necrosis. Lane 1 = tomato plants inoculated with P. corrugata ICMP 5819. Lane 2 = tomato plants inoculated with P. corrugata P94. Lane 3 = tomato plants inoculated with water as control.

Full size image
Fig. 2
figure 2

Inhibition activity to tomato grey mildew. Lane 1 = tomato plants inoculated with B. cinerea treated with distilled water. Lane 2 = tomato plants inoculated with B. cinerea treated with P94. Lane 3 = no B. cinerea as control.

Full size image

Strain P94 was identified as P. corrugata using the Biolog identification system, and that was confirmed by 16S rDNA sequence analysis and biochemical and physiologic characteristics. This procedure in identification was more accurate than the traditional method. The 16S rRNA gene sequence of this strain exhibited 99.7% similarity to that of P. corrucata. The results of biochemical and physiologic testing is listed in Table 2. Strain P94 cells were Gram negative, rod shaped (0.7 × 1.8 µm), and motile by ≥2 polar flagellum. The bacteria was aerobic and produced a diffusible nonfluorescent, yellowish-green pigment. This strain showed positive reactions for oxidase, gelatin hydrolysis, poly-beta-hydroxybutyrate (PHB) accumulation, and nitrate reduction, but it was negative for starch hydrolysis and levan production, which corresponds with P. corrugata reports in the literature [18, 33]. The biochemical and physiologic characteristics of strain P94 produced similar results to those of P. corrucata strain NCPPB 2445, with the exception of arginine dihydrolase and lipase. Arginine dihydrolase activity was positive for strain P94 and negative for strain NCPPB 2445. Lipase activity was negative for strain P94 and positive for NCPPB 2445. However, Sutra et al. [33] found that not all P. corrugata strains were negative for arginine dihydrolase. The size of specific PCR product with the mixed primers was 600 bp, indicating that strain P94 belonged to the P. corrugata genomic group II (Fig. 3), and further supporting evidence for this result came from biochemical tests. Genomic groups I and II corresponded, respectively, to the subphenas 1a and 1b of Sutra et al. [33], and the genetic groups were not correlated with differences in pathogenicity or virulence of the strains [6]. Use of meso-tartrate, isobyturate, and histamine was the distinguishable characteristic of P. corrugata subphenas 1a and 1b [33]. Strain P94 fell into P. corrugata subphena 1b because it could assimilate meso-tartrate and histamine (Table 2).

Table 2 Biochemical and physiologic characteristics of P. corrugat a strain 94 and strain NCPPB 2445 reported by Lopez et al. (1994)
Full size table
Fig. 3
figure 3

PCR amplification with a mixture of primers (PC1/1, PC1/2, PC5/1, and PC5/2). Lanes 1 = marker. Lane 2 = P. fluorescens P12 as negative control. Lane 3 = P. corrugata P94 belonged to P. corrugata genomic group II. Lane 4 = P. corrugata ICMP 5819 belonged to P. corrugata genomic group I

Full size image

It has been documented (1) that HCN and cell wall–degrading enzymes–such as cellulases, proteases, and chitinases–could be involved in antagonism toward phytopathogens and contribute to biocontrol of plant diseases [1, 4] and (2) that production of IAA and phosphatase by plant-beneficial bacteria enhance the development of the host plant root system and thus help in the growth of plants [1, 27]. Strain P94 showed a positive reaction for production of HCN, protease, phosphatase, and IAA and a negative reaction for production siderophore, chitinase, and cellulase. These characteristics of P. corrugata have not been described previously. Quantitative dating confirmed the production of 21.8 and 7.1 µg/ml HCN in the presence and absence of glycine, respectively. However, the amount of HCN released by P94 was greater than that achieved with P. aeruginose PAO1 or Pseudomonas sp. NJ-101, whatever the glycine condition [1, 26]. Strain P94 produced 15.97 µg/ml and 6.97 µg/ml IAA in nutrient broth with and without tryptophan, respectively. Quantitative analysis revealed the release of 82.4 µg/ml soluble phosphate from tricalcium phosphate in the medium. The phosphate-solubilizing activity of P. corrugata P94 was stronger than that of Pseudomonas sp. NJ-101 [1]. However, IAA production was low compared with that reported in the earlier study [1, 15].

In this study, for the first time, we reported P. corrugata P94 as a new nonpathogenic strain of tomato rhizosphere soil origin with the production of HCN, IAA, phosphatase, and protease and without the production of siderophore, chitinase, and cellulase. P. corrugata P94 showed the ability to control tomato grey mildew caused by B. cinerea, inhibit phytopathogenic fungi and bacteria, and produce plant growth–promoting substances, such as IAA and phosphatase, that would enhance the potential use of P94 as an effective biocontrol agent.