Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Journal of Applied Microbiology ISSN 1364-5072 ORIGINAL ARTICLE Isolation and characterization of antagonistic Bacillus subtilis strains from the avocado rhizoplane displaying biocontrol activity F.M. Cazorla1,2, D. Romero2, A. Pérez-Garcı́a2, B.J.J. Lugtenberg1, A. de Vicente2 and G. Bloemberg1 1 Leiden University, Institute of Biology Leiden, Clusius Laboratory, Wassenaarseweg, AL Leiden, The Netherlands 2 Departamento de Microbiologı́a, Facultad de Ciencias, Universidad de Málaga, Campus Universitario de Teatinos, s ⁄ n, Malaga, Spain Keywords antifungal metabolite, Dematophora necatrix, Dematophora root rot, Persea americana, phytopathogenic fungi. Correspondence F.M. Cazorla, Departamento de Microbiologı́a, Faculdad de Ciencias, Universidad de Málaga, Campus Universitario de Teatinos, s ⁄ n, 29071 Málaga, Spain. E-mail: [email protected] 2007 ⁄ 0076: received 17 January 2007, revised 6 March 2007 and accepted 4 April 2007 doi:10.1111/j.1365-2672.2007.03433.x Abstract Aim: This study was undertaken to isolate Bacillus subtilis strains with biological activity against soil-borne phytopathogenic fungi from the avocado rhizoplane. Methods and Results: A collection of 905 bacterial isolates obtained from the rhizoplane of healthy avocado trees, contains 277 gram-positive isolates. From these gram-positive isolates, four strains, PCL1605, PCL1608, PCL1610 and PCL1612, identified as B. subtilis, were selected on the basis of their antifungal activity against diverse soil-borne phytopathogenic fungi. Analysis of the antifungal compounds involved in their antagonistic activity showed that these strains produced hydrolytic enzymes such as glucanases or proteases and the antibiotic lipopeptides surfactin, fengycin, and ⁄ or iturin A. In biocontrol trials using the pathosystems tomato ⁄ Fusarium oxysporum f.sp. radicis-lycopersici and avocado ⁄ Rosellinia necatrix, two B. subtilis strains, PCL1608 and PCL1612, both producing iturin A, exhibited the highest biocontrol and colonization capabilities. Conclusions: Diverse antagonistic B. subtilis strains isolated from healthy avocado rhizoplanes have shown promising biocontrol abilities, which are closely linked with the production of antifungal lipopeptides and good colonization aptitudes. Significance and Impact of the Study: This is one of the few reports dealing with isolation and characterization of B. subtilis strains with biocontrol activity against the common soil-borne phytopathogenic fungi F. oxysporum f.sp. radicis-lycopersici and R. necatrix. Introduction Recent, intensive agricultural production encourages greater attention to crop protection from pathogens that lessen yields, as well as to the microbial quality of these crops as raw materials. From the initial implementation of sustainable agriculture, the availability of alternative protective strategies has been reassessed and consequently, the development of environment-friendly and food-hygienically-safe plant-protecting methods based on biological agents has been greatly emphasized (Warrior 2000). 1950 Avocado (Persea americana Mill.) is a crop that was introduced into Europe fairly recently, although it has mainly been implemented in Spain and Portugal because of optimal weather conditions. However, the rising threat of root fungal diseases that seriously affect health and crop yields has led to focusing more attention on devising reliable disease control programmes (Pérez-Jiménez 2006). Management of avocado Dematophora root rot is difficult because any preplanting treatment must have a long-term effect and any postplanting treatment must not adversely affect the crop. Diverse approaches to control ª 2007 The Authors Journal compilation ª 2007 The Society for Applied Microbiology, Journal of Applied Microbiology 103 (2007) 1950–1959 F.M. Cazorla et al. Rosellinia necatrix before planting have been tested over the past two decades; these techniques include soil fumigation (Sztejnberg et al. 1987), or specially soil solarization (López-Herrera et al. 1999) which may induce disease suppressive activity in soils by increasing microbial activities (Greenberger et al. 1987) and biological control using the antagonistic fungus Trichoderma harzianum (Sztejnberg et al. 1987). In the context of biocontrol, a successful exploratory attempt to repress avocado root rot caused by R. necatrix by using antagonistic pseudomonads has been recently reported (Cazorla et al. 2006); however, the use of bacterial strains as biocontrol agents against avocado soil-borne phytopathogens remains an issue to be further explored. Bacteria belonging to the genera Bacillus are considered to be safe micro-organisms and hold the remarkable abilities of synthesizing a vast array of beneficial substances for agronomical and industrial purposes (Stein 2005) and producing endospores, which warrant the prevalence of Bacillus under different environmental cues, its long-term storage and easy development of reliable formulations (Collins and Jacobsen 2003). Many of these bacilli are soil-inhabiting bacteria and exist as epiphytes or endophytes (Sneath 1986; McSpadden Gardener 2004) in environments as diverse as spermosphere (Walker et al. 1998), Brassica leaves (Leifert et al. 1992) or compost (Phae et al. 1990), and may provide plants with protection against pathogen attacks by a blend of diverse modes of action (Rampach and Kloepper 1998; Shoda 2000; Romero et al. 2004). These features have led to the increased devising and implementation of antimicrobial active biological products based on Bacillus species or their metabolites as alternative or supplementary methods to chemicals for plant disease control (Fravel 1988; Potera 1994; Leifert et al. 1995; Raaijmakers et al. 2002; Schisler et al. 2004; Ongena et al. 2005). The success of biocontrol strategies will depend to a large extent on the seeking and selection process of potential biological agents, which consider the pathogen to be the target and the cropping system. Therefore, the main aims of this study were to isolate gram-positive bacilli with antifungal activity from the rhizoplane of healthy avocado trees in order to evaluate their biocontrol potential against diverse soil-borne phytopathogenic fungi, and to obtain insights into the putative mode of action involved in their protective activity. Materials and methods Micro-organisms and culture conditions The micro-organisms used in this study are listed in Table 1. The bacterial strains were kept for long-term Characterization of antagonistic Bacillus storage at )80C in Luria-Bertani broth (LB) with 15% glycerol (v ⁄ v). Routinely, fresh bacterial cultures were obtained from frozen stocks before each experiment and were grown at 24C in nutrient agar medium (NA; Difco Laboratories, Detroit, MI, USA) or in optimal medium for lipopeptide production (MOLP) at 37C (Ahimou et al. 2000). The fungal strains were stored at 4C immersed in water. They were routinely grown on potato dextrose agar (PDA; Difco Laboratories) or King-B (KB) agar at 24C and checked for viability. Isolation and characterization of antagonistic rhizobacterial strains A collection of rhizobacterial isolates was obtained from 28 healthy avocado roots of 20-year-old plants collected from 17 avocado orchards affected by white root rot and located in Algarrobo, Fuengirola and Vélez-Málaga (Málaga, Spain) (Cazorla et al. 2006). Briefly, roots were sampled at a distance of 1 m from the trunk and 10 cm from the soil surface. The root samples were gently shaken to remove loosely adhering soil and aseptically transferred to storage bags, then maintained on ice for transportation and kept at –80C until processing. Twenty-four root samples were washed twice in tap water, weighed and homogenized in a lab blender for 3 min with 10 ml of sterile phosphate-buffered saline (PBS; pH 7Æ2) per gram of fresh root material. Bacterial counts were determined by serial dilutions and plated onto tryptic soy agar (TSA; Oxoid, Perth, UK) or NA amended with cycloheximide (100 lg ml)1) to prevent fungal growth, after 48 or 72 h of incubation at 24C. Bacillus-like colonies were roughly identified on the basis of their morphology and gram reaction (Powers 1995). Bacterial isolates showing a broad spectrum of antifungal activity were subjected to further identification according to phenotypic and physiological tests using API20NE (BioMerieux, Mercy L’etoyle, France) and BIOLOGTM (BIOLOG Inc., California, USA) systems, and analysis of the 16S rDNA sequence. To obtain 16S rDNA sequences, colony polymerase chain reaction (PCR) was carried out using the primers 41 F (5¢-GCTCAGATTGAACGCTGGCG-3¢) and 1486r-P (5¢-GCTACCTTGTTACGACTTCACCCC-3¢) and the PCR amplification protocol previously described (Cazorla et al. 2006). The resulting PCR fragments were purified (QIAquick PCR Purification Kit 50, Westburg, Leusden, The Netherlands) and sequenced (Baseclear, Leiden, The Netherlands). The sequences were analysed using DNAman software (Lynnon Biosoft, Quebec, Canada) and compared with 16S rRNA gene sequences in the public database using the NCBI Genebank Blast software (NCBI, USA). ª 2007 The Authors Journal compilation ª 2007 The Society for Applied Microbiology, Journal of Applied Microbiology 103 (2007) 1950–1959 1951 Characterization of antagonistic Bacillus F.M. Cazorla et al. Table 1 Micro-organisms used in this study Strain designation Fungal strains Fusarium oxysporum f.sp. radicis-lycopersici ZUM2407 Phytophthora cinnamomi 273 301 344 Pythium ultimum LBOP17 Rhizoctonia solani 3R4FNA Rosellinia necatrix 358 397 400 CECT2817 CECT2818 Sclerotium rolfsii 151Æ31 Bacterial strains Bacillus subtilis PCL1605 PCL1608 PCL1610 PCL1612 UMAF6614 UMAF6639 Relevant characteristics Reference or source Causes crown and foot rot of tomato IPO-DLO* Isolated from Prr, high virulence Isolated from Prr, medium virulence Isolated from Prr, low virulence Causes avocado root rot Causes Rhizoctonia seed and root rot Pérez-Jiménez (1997) Pérez-Jiménez (1997) Pérez-Jiménez (1997) IPO-DLO IPO-DLO Isolated from Wrr, low virulence Isolated from Wrr, medium virulence Isolated from Wrr, high virulence Isolated from avocado pear rot Isolated from avocado pear rot Causes avocado seedling blight Pérez-Jiménez (1997) Pérez-Jiménez (1997) Pérez-Jiménez (1997) CECT§ CECT CBS** Wild-type, isolated from avocado rhizoplane Wild-type, isolated from avocado rhizoplane Wild-type, isolated from avocado rhizoplane Wild-type, isolated from avocado rhizoplane Producer of bacillomycin, surfactin and fengycin Producer of iturin A, surfactin and fengycin This study This study This study This study Romero et al. (2007) Romero et al. (2007) *Institute for Plant Protection-Agriculture Research Department, Wageningen, The Netherlands. Avocado trees infected by Phytophthora root rots. Avocado trees showing white root rot symptom. §Spanish Type Culture Collection. **Fungal Biodiversity Center, Utrecht, The Netherlands. Antagonistic activity assays in vitro The rhizobacterial isolates were tested for their ability to inhibit the growth of diverse soil-borne fungal pathogens using the in vitro dual-culture analysis (Romero et al. 2004). A plug of 0Æ6-cm diameter containing mycelium taken from 5-day-old target fungi was placed at the centre of PDA or KB dual plates, and single bacterial colonies were patched at a distance of about 3 cm from the fungus. Plates were incubated for 5 days at 25C and inhibition of fungal growth was monitored by recording the diameter of the inhibition zone (mm). The antifungal activity of cell-free supernatants of the different B. subtilis antagonistic strains was evaluated against the target fungi Fusarium oxysporum and R. necatrix using the in vitro test described elsewhere (Romero et al. 2007). Bacteria were grown on MOLP at 37C, and after 5 days of incubation, cells were removed by centrifugation at 2500 g for 15 min. Thereafter, the supernatants were extracted with n-butanol and once the organic phase had evaporated, the remaining residue was dissolved in sterile distilled water. The antifungal activity was finally 1952 evaluated using the dual-culture analysis described before, but the bacterial colonies were replaced with the antagonistic solutions. Production of hydrolytic enzymes and lipopeptides The ability of B. subtilis strains to produce hydrolytic enzymes such as proteases, lipases, b-glucanases and cellulases was analysed following general procedures previously described (Gerhardt et al. 1994). The presence of antifungal lipopeptides in bacterial culture supernatants was analysed by thin-layer chromatography (TLC) and reverse phase high-performance liquid chromatography (RPHPLC) as described previously (Romero et al. 2007). Bacterial cultures were grown in MOLP for 5 days at 37C. Cell-free supernatants were obtained by centrifugation at 2500 g for 15 min, and then extracted with n-butanol. Once n-butanol layers were evaporated to dryness under a vacuum, the residues were dissolved in methanol and fractionated by TLC followed by RP-HPLC analysis using an analytical reverse phase C18 column ultrasphere, 4Æ6-mm diameter · 250mm long (Beckman Instrument ª 2007 The Authors Journal compilation ª 2007 The Society for Applied Microbiology, Journal of Applied Microbiology 103 (2007) 1950–1959 F.M. Cazorla et al. Inc., Fullerton, CA, USA) and solutions of 0Æ05% trifluoroacetic acid in acetonitrile and milliQ water, with a flow rate of 1 ml min)1. For identification of these antifungal compounds, methanolic extracts of the B. subtilis strains UMAF6614 and UMAF6639 were obtained following the same procedure described before, and used as controls together with purified standards of iturin A, fengycin and surfactin lipopeptides (Romero et al. 2007). Biocontrol assays against tomato foot and root rot Biocontrol trials in the pathosystem tomato ⁄ F. oxysporum f.sp. radicis-lycopersici were set up as previously described (Chin-A-Woeng et al. 1998). A third part of a 10-day-old PDA plate culture of F. oxysporum f.sp. radicis-lycopersici was homogenized and inoculated in 1-l Erlenmeyer flasks containing 200 ml of Czapek-Dox medium. After growth for 3 days at 28C under aeration (110 rev min)1), the fungal material was placed on top of sterile glass wool and the filtrate was adjusted to 5 · 105 spores ml)1. For soil inoculation, spore suspensions were mixed thoroughly with potting soil to a final concentration of 3Æ2 · 106 spores kg)1. Tomato (Solanum lycopersicum L.) seeds (cv. Carmello) were coated with bacteria by dipping the seeds in a mixture of 1% (wt ⁄ v) methylcellulose (Sigma, St. Louis, MO, USA) and 109 CFU ml)1 bacteria in PBS buffer. Coated seeds were dried in a sterile stream of air. One seed was sown in each pot of approximately 1Æ5-cm depth and containing 25 g of soil. Ten sets of 10 plants each were included in each treatment. Seedlings were grown in a greenhouse at 24C with 70% relative humidity, 16-h daylight and were watered from the bottom. The number of diseased plants was determined when a considerable fraction of untreated plants (above 60%) used as control showed symptoms, usually 18 days after sowing. Plants were removed from the soil, washed and the plant roots were examined for tomato crown and root rot symptoms, indicated by root browning and lesions. Roots without any disease symptoms were designated as healthy. Biocontrol experiments of Bacillus subtilis strains against avocado white root rot Biocontrol assays against avocado white root rot were carried out in the avocado ⁄ R. necatrix system previously described (Cazorla et al. 2006). Avocado plants were obtained by growing seedlings from avocado embryos (Pliego-Alfaro et al. 1987). After 4 weeks of incubation, the seedling shoots were removed from the tubes, washed with tap water, transferred into pots containing 30 g of wet perlite (Agra-perlite, Maasmond Westland B.A., Rijnsburg, The Netherlands) and kept in a growth cham- Characterization of antagonistic Bacillus ber for 20–30 days to harden the roots before biocontrol experiments. Plants with hardened roots were removed from the perlite and the roots were washed with tap water in order to remove residual perlite. The roots of the seedlings were surface disinfected by immersion in 0Æ1% NaOCl for 20 min, and were washed and bacterized following the method previously described (Lugtenberg et al. 1994) with slight modifications. The roots of avocado seedlings were immersed into a suspension of the bacterial isolate (109 CFU ml)1) or into sterile tap water for 20 min. Any excess of the bacterial suspension was allowed to drip off and the seedlings were placed into pots containing 30 g of wet potting soil (Jongkind Grond B.V., Aalsmeer, The Netherlands) and infected with R. necatrix using wheat grains (four infected grains per pot) as described by Freeman et al. (1986). Five sets of 10 avocado seedlings each were tested per treatment. The seedlings were grown in a growth chamber at 24C and 70% relative humidity with 16-h daylight and were watered twice per week from above. The number of diseased seedlings was determined 21 days after bacterization. As monitoring of symptoms on hardened avocado roots was difficult because of overgrowth of R. necratix, aerial symptoms were recorded on a 0–3 scale of values: 0 (healthy plant), 1 (yellowing and wilting of the leaves), 2 (overall drying of the plant) and 3 (dead plant). Disease index (DI) was then calculated using the previously described formula (Cazorla et al. 2006): DI ¼ 100 ðax0Þ þ ðbx1Þ þ ðcx2Þ þ ðdx3Þ 3xn where a, b, c and d correspond to the number of plants showing disease values of 0, 1, 2 and 3, respectively, and n is the total number of plants tested. Survival of Bacillus subtilis strains on avocado roots The persistence of B. subtilis bacteria on avocado roots was studied using the plants and methods described before for biocontrol experiments. The roots of individual plants were disinfected and bacterized with vegetative cells of B. subtilis strains (109 CFU ml)1). A total of 20 plants were included for each treatment and maintained in a growth chamber at 24C with 70–80% relative humidity, 16-h daylight and were watered from the bottom. After 1 and 30 days of incubation, 10 plants per treatment were sampled, gently shaken to remove loosely adhering soil to root systems and processed as follows. Root samples were washed twice in water, weighed and homogenized in a lab blender for 3 min with 10 ml of sterile PBS (pH 7Æ2) per gram of fresh root material. In order to select the sporulating bacteria in the resulting suspensions, they were ª 2007 The Authors Journal compilation ª 2007 The Society for Applied Microbiology, Journal of Applied Microbiology 103 (2007) 1950–1959 1953 Characterization of antagonistic Bacillus F.M. Cazorla et al. heated at 80C for 10 min (Marten et al. 2000), and serially diluted and plated on NA. A bacterial count was carried out after 4–5 days of incubation at 25C. Statistical methods Data were statistically analysed by analysis of variance followed by Fisher’s least significant difference test (P = 0Æ05) using SPSS software (SPSS Inc., Chicago, IL, USA). All experiments were performed at least twice. Identification of antifungal compounds produced by Bacillus subtilis antagonistic strains Results Selection and identification of rhizobacterial strains with antifungal activity The total cultured bacterial counts obtained on NA from 28 different avocado root samples ranged from 2 · 105 to 3Æ2 · 106 CFU g)1 of fresh weight of avocado roots. The presence of Bacillus-like bacteria was indicated by growth of white-creamy and dry colonies and ranged from 6Æ3 · 102 to 3Æ2 · 104 CFU g)1 fresh weight of roots. Among the collection of 905 bacterial isolates, up to 30% (n = 277) were characterized as gram-positive Bacilluslike bacteria and four of them (PCL1605, PCL1608, PCL1610 and PCL1612), belonging to different root samples located in different geographical areas were selected on the basis of their antifungal activity against the highly virulent R. necatrix strain Rn400 and F. oxysporum f.sp. racidis-lycopersici ZUM2407. The antifungal activity of the four selected bacteria against several soil-borne phytopathogenic fungi in the dual-plate assays is summarized in Table 2. It was remarkable that two Bacillus strains, PCL1608 and PCL1612, inhibited the growth of all target fungi, especially F. oxysporum f.sp. radicis-lycopersici. Table 2 Screening of antagonistic Bacillus subtilis rhizobacteria isolated from healthy avocado rhizoplanes for ability to inhibit the growth of diverse soil-borne fungal phytopathogens in dual plate assays B. subtilis strains Target fungi PCL1605 PCL1608 PCL1610 PCL1612 Fusarium oxysporum Phytophthora cinnamomi (n = 3) Pythium ultimum Rosellinia necatrix (n = 5) Rhizoctonia solani Sclerotium rolfsii +* ) ++ + + – ++ + – ++ + – + + + + – + + (+) + + + + *Diameter of inhibition zones. ++, inhibition >20 mm; +, zone of inhibition 8–20 mm; (+), <8 mm; –, no antifungal activity. Number of different target fungal isolates. 1954 Identification of these four rhizobacterial isolates using the API20NE and BIOLOGTM tests, along with homology analyses of nucleotide sequences derived from PCRamplified 16S rDNA fragments, revealed that these strains could be designated as B. subtilis. The 16S rDNA sequences of B. subtilis PCL1605, PCL1608, PCL1610 and PCL1612 were submitted to the Genbank database under accession numbers DQ779882, DQ779883, DQ779884 and DQ779885, respectively. The four selected B. subtilis strains were analysed for their ability to produce hydrolytic enzymes (Table 3). All of them showed b-glucanase activity, and B. subtilis PCL1605 and PCL1608 additionally exhibited protease activity; however, no lipase or cellulase activities were detected for these B. subtilis strains. In order to evaluate the involvement of other putative antimicrobial compounds in their suppressive effect, the antifungal activity of cell-free supernatants from the four B. subtilis strains was analysed. To do this, we looked for inhibition of the growth of F. oxysporum f.sp. radicis-lycopersici and R. necatrix in the plate assay. Antagonistic suspensions obtained from butanolic extracts of the four B. subtilis cell-free filtrates were able to inhibit the growth of both target fungi. In an attempt to identify the putative antifungal compounds occurring in B. subtilis supernatants, methanolic fractions derived from the butanolic extracts of cell-free culture filtrates of the four B. subtilis strains were initially separated in silica TLC plates, using purified iturin A, Table 3 Production of antifungal metabolites by antagonistic Bacillus subtilis strains B. subtilis strains Secreted products Hydrolytic enzymes b-glucanase Cellulase Lipase Protease Lipopeptides Fengycin Iturin A Surfactin PCL 1605 PCL 1608 PCL 1610 PCL 1612 + – – + + – – + + – – – + – – – + – + + + + + – + + + + Hydrolytic activities were determined by plate assays, and the lipopeptides produced were analysed from supernatants extracted from bacterial cultures. Lipopeptides purified from B. subtilis strains UMAF6614 and UMAF6630 were used as controls for identification. +, presence; –, absence. ª 2007 The Authors Journal compilation ª 2007 The Society for Applied Microbiology, Journal of Applied Microbiology 103 (2007) 1950–1959 F.M. Cazorla et al. fengycin and surfactin as standards. The four B. subtilis strains produced spots with Rf values similar to fengycin (Rf = 0Æ09) and surfactin (Rf = 0Æ7), whereas B. subtilis strains PCL1608 and PCL1612 also produced spots with Rf values similar to iturin A (Rf = 0Æ3). These findings were further confirmed by comparison of RP-HPLC profiles derived from the methanolic extracts with the profiles for standard lipopeptides purified from the wellcharacterized lipopeptide-producing B. subtilis strains Characterization of antagonistic Bacillus UMAF6614 and UMAF6639 (Romero et al. 2007). Three main groups of peaks were monitored at elution times comparable with those observed for standard lipopeptides, confirming that the four B. subtilis strains simultaneously produced surfactin and fengycin in broth cultures. The B. subtilis strains PCL1608 and PCL1612 also produced iturin A (Fig. 1). Biocontrol activity against soil-borne fungal pathogens and root colonization abilities of Bacillus subtilis strains The four selected B. subtilis strains were tested for their biocontrol abilities in the pathosystem tomato ⁄ F. oxysporum f.sp. radicis-lycopersici. As shown in Fig. 1, up to 75% of untreated plants were diseased after 16 days of growth in soil infested with F. oxysporum, whereas treatment of tomato seeds with cells of B. subtilis strains induced a significant reduction in the percentage of diseased plants in comparison with the nonbacterized plants. It was notable that the highest protection against tomato crown and root rot was provided by B. subtilis strains PCL1608 and PCL1612, which reduced the disease incidence by around 40–53% (Fig. 2). Similarly, in biocontrol experiments with the avocado ⁄ R. necatrix Figure 1 Reverse phase high-performance liquid chromatographic analysis of lipopeptides occurring in butanolic extracts from liquid cultures of: (a) the iturin-producing antagonistic Bacillus subtilis strains PCL1608 and PCL1612; B. subtilis strain UMAF6639 was used as reference, and (b) the nonproducing iturin B. subtilis strains PCL1605 and PCL1610. Peaks corresponding to iturin (It), fengycin (Fen) and surfactin (Sf) are indicated. Figure 2 Biocontrol of tomato foot and root rot caused by Fusarium oxysporum f.sp. radicis-lycopersici. Four Bacillus subtilis strains isolated from the avocado rhizosphere were tested in a F. oxysporum ⁄ tomato test system using potting soil. Tomato seedlings were scored as healthy or diseased 21 days after bacterization. Data were analysed for significance after arcsine square root transformation with analysis of variance, followed by Fisher’s least significant difference test (P = 0Æ05). Values of bars with different letter indications represent a statistically significant difference. ª 2007 The Authors Journal compilation ª 2007 The Society for Applied Microbiology, Journal of Applied Microbiology 103 (2007) 1950–1959 1955 Characterization of antagonistic Bacillus F.M. Cazorla et al. A remarkable persistence of B. subtilis strains in the avocado rhizosphere was observed after 1 month. Therefore, samples taken from bacterized avocado roots showed bacterial levels above 104 CFU g)1 of root for the strains PCL1605 and PCL1608 (Table 4). The roots of nonbacterized control plants showed high levels of microbial populations before heating treatment (Table 4), although colonies mostly showed gram-negative-like morphology. Despite the high microbial population levels detected prior to heating treatment, no growth of sporulating microorganisms was detected in nonbacterized plants (Table 4). Discussion Figure 3 Biocontrol of white root rot of avocado plants caused by Rosellinia necatrix. Four Bacillus subtilis strains isolated from the avocado rhizosphere were tested in a R. necatrix ⁄ avocado test system in potting soil. Avocado seedlings were scored as diseased or healthy after 18–21 days of growth after bacterization. Data were analysed for significance after arcsine square root transformation with analysis of variance, followed by Fisher’s least significant difference test (P = 0Æ05). Values of bars with different letter indications denote a statistically significant difference. pathosystem (Fig. 3), the B. subtilis strains PCL1608 and PCL1612 gave significant reductions in disease incidence (33% to 17%, respectively) compared with untreated controls, whereas B. subtilis strain PCL1605 failed to control the disease (Fig. 3). Bacillus subtilis PCL1605 and PCL1608 were selected for persistence studies on avocado roots based on their different patterns of antifungal production and their differential performance in biocontrol experiments. The rhizosphere of healthy avocado trees present in areas affected by soil-borne phytopathogenic fungi could represent a feasible and useful source for isolation of microorganisms with promising antagonistic ability (Schroth and Hancock 1982; Cazorla et al. 2006). The percentage of gram-positive bacterial isolates (30%) and the presence of antagonistic B. subtilis strains among them suggest that B. subtilis could be one of the habitual culturable bacteria associated with avocado roots, which is consistent with the well-established ability of this group of bacteria to secrete antibiotics or hydrolytic enzymes capable of modifying the environment for self-benefit and persisting as highly resistant endospores (Földes et al. 2000; Shoda 2000). The in vitro prescreening test of dual culture allowed us to select four B. subtilis strains, PLC1605, PCL1608, PCL1610 and PCL1612, with noticeable antifungal activity against R. necatrix and other soil-borne phytopathogenic fungi (Table 2). The selected antagonistic B. subtilis strains showed a moderate range of antagonistic activity against many target fungi (Table 2), with variable responses against Phytophthora cinnamomi, an oomycete which causes the most serious fungal disease in avocado plants worldwide (Erwin and Ribeiro 1996), and Log CFU g)1 root* Log CFU cm)1 root* Root sample (treatment) 1 day 30 days 1 day 30 days Nonbacterized (prior heating) Nonbacterized (a.h.) B. subtilis PCL1605 (a.h.) B. subtilis PCL1608 (a.h.) 6Æ75 ± 0Æ45a <2 6Æ33 ± 0Æ01a 6Æ04 ± 0Æ13a 7Æ42 ± 0Æ40a <2 4Æ14 ± 0Æ30c 5Æ70 ± 0Æ29b 5Æ03 ± 0Æ21a <2 3Æ67 ± 0Æ31b 3Æ93 ± 0Æ58b 5Æ50 ± 0Æ87a <2 2Æ42 ± 0Æ21c 4Æ16 ± 0Æ19b Table 4 Bacterial counts (log CFU g)1 root) on avocado roots bacterized with Bacillus subtilis strains PCL1605 and PCL1608 Samples were analysed for total heterotrophic bacteria present in the root (prior heating), and total sporulated bacteria, estimated after a heat treatment (80ºC, 10 min) at 1 and 30 days after bacterial inoculation. Nonbacterized avocado seedlings were used as controls. *Each value represents mean ± standard error from three independent samples of 10 avocado root tips (one root tip from 10 independent plants). Data were analysed for significance with analysis of variance, followed by Fisher’s least significant difference test (P = 0Æ05). Values with different letter indications represent a statistically significant difference. After heating treatment of root sample (80C, 10 min). 1956 ª 2007 The Authors Journal compilation ª 2007 The Society for Applied Microbiology, Journal of Applied Microbiology 103 (2007) 1950–1959 F.M. Cazorla et al. R. necatrix, which causes most of the crop losses in the Mediterranean area (Pérez-Jiménez 2006). These preliminary results initially suggested the differential production of antifungals in vitro by these B. subtilis strains. It is well known that bacteria commonly produce cell wall-degrading enzymes and secondary metabolites to hinder the growth of other micro-organisms (Shoda 2000); further analysis proved that the four selected B. subtilis strains produced different cell wall hydrolytic enzymes as well as antibiotic lipopeptides, mainly surfactin or fengycin, and to a lesser extent iturin (Table 3). Lipopeptides represent a group of compounds with relatively low molecular weight that are synthesized in a nonribosomal manner and exhibit amphiphilic features that have engaged attention because of their surfactant and antimicrobial activities. Genes for lipopeptides are common to multiple biocontrol strains that have been commercialized and that genomes with such genes have an enhanced capacity to produce antibiotics that inhibit the growth of fungal root pathogens (Joshi and McSpadden Gardener 2006). The involvement of these lipopeptides in the protective action provided to plants by Bacillus has been reported under pre- and postharvest conditions, either hampering the pathogenic fungi directly or inducing systemic resistance of host plants (Asaka and Shoda 1996; Koumoutsi et al. 2004; Ongena et al. 2005; Romero et al. 2007). The antifungal profiles obtained for these B. subtilis strains differed in such a way that only two strains, PCL1608 and PCL1612 produced high levels of lipopetidic antibiotics, especially iturin A (Fig. 1) and concomitantly gave the best biocontrol results against F. oxysporum (Fig. 2) and R. necatrix (Fig. 3) in biocontrol trials, in contrast to the iturin A nonproducing strains, also displaying a low production of lipopeptidic antibiotics, which failed to successfully control both diseases. These findings, along with previously reported role of iturin A in the biocontrol of several filamentous fungi (Asaka and Shoda 1996; Munimbazi and Bullerman 1998; Chitarra et al. 2003) supported a major role of iturin A in the antifungal activity of these B. subtilis against the target fungi. However, the involvement of surfactin or fengycin in their biocontrol acitivity should not be ruled out, considering the previously reported synergistic effect between iturin A and surfactin that results in a considerable increase in surfactant and antimicrobial activities (Magnet-Dana et al. 1992). Additionally, the presence of surfactin has been reported to be crucial in biofilm formation, consequently playing an essential role in the general process of colonization (Bais et al. 2004). Despite the antifungal activity showed by the rest of the B. subtilis strains on agar plate tests (Table 2) and TLC analysis, they failed to prevent disease progression in the rhizosphere environment, which could reflect the scarcity or lack of delivery Characterization of antagonistic Bacillus of antifungal metabolites into the surroundings and thereby reduced colonization aptitudes of root systems (Chin-A-Woeng et al. 1998; Lugtenberg et al. 2001). These findings are in close agreement with previous biocontrol reports (Shoda 2000; Raaijmakers et al. 2002), which emphasized the fact that promising biocontrol agents also need to retain the ability to efficiently colonize the habitat in which they are expected to perform in order to successfully control diseases (Raaijmakers et al. 2002; Collins and Jacobsen 2003; Bais et al. 2004). The slight decrease in population experienced by both antagonistic B. subtilis strains PCL1605 and PCL1608 during the 30 days of biocontrol experiments suggests that they may be considered good colonizers of avocado roots. Persistence and colonization are also supported by their glucanase or cellulase activities (Table 3), which should allow them to efficiently feed on compounds naturally occurring in this environment (Sessitsch et al. 2004). This feature indeed represents another efficient way to exclude micro-organisms from the occupied habitat and therefore is considered of great relevance for biocontrol purposes against necrotrophic fungi because spare root susceptibility to pathogen invasion is reduced, achieving reasonable disease reduction (Lugtenberg et al. 1994; Shoda 2000; Collins and Jacobsen 2003). All of these features should come together to arrest pathogen development, avoiding plant invasion and warranting biocontrol agent survival, which should provide more efficient disease control. This is one of the few reports of B. subtilis strains showing biocontrol traits against the soil-borne phytopathogenic fungi R. necatrix and F. oxysporum f.sp. radicis-lycopersici. In conclusion, two B. subtilis strains, PCL1608 and PCL1612, showed important requirements for valuable biological agents, such as production of fungal cell wall hydrolytic enzymes and antibiotic lipopeptides or efficient colonization and persistence in treated soil and plant roots, which turn them into promising candidates to be included in biocontrol programmes against avocado root rot, tomato foot and root rot and other diseases caused by soil-borne phytopathogenic fungi. Although antibiosis seems to be the strategy involved in their biocontrol activity, further analysis of their modes of action are currently being undertaken in order to improve their use and implementation under real culture conditions. Acknowledgements This research was supported by a Marie Curie Fellowship of the European Community Programme Improving Human Potential under contract number HPMF-CT1999-00377 and a grant from Plan Nacional de Recursos y Tecnologı́as Agroalimentarias from the Ministerio de ª 2007 The Authors Journal compilation ª 2007 The Society for Applied Microbiology, Journal of Applied Microbiology 103 (2007) 1950–1959 1957 Characterization of antagonistic Bacillus Educación y Ciencia of Spain (AGL 2005-06347-CO3-01). The authors thank C. López-Herrera and R. Pérez-Jiménez for supply of some R. necatrix and P. cinnamomi strains, and A. Barceló for technical assistance in growing the avocado plants in vitro. References Ahimou, F., Jacques, P. and Deleu, M. (2000) Surfactin and iturin A effects on Bacillus subtilis surface hydrophobicity. Enzy Microb Technol 27, 749–754. Asaka, O. and Shoda, M. (1996) Biocontrol of Rhizoctonia solani damping-off of tomato with Bacillus subtilis RB14. Appl Environ Microbiol 62, 4081–4085. Bais, H.P., Fall, R. and Vivanco, J.M. (2004) Biocontrol of Bacillus subtilis against infection of Arabidopsis roots by Pseudomonas syringae is facilitated by biofilm formation and surfactin production. Plant Physiol 134, 307–319. Cazorla, F.M., Duckett, S.B., Bergström, E.T., Noreen, S., Odijk, R., Lugtenberg, B.J.J., Thomas-Oates, J. and Bloemberg, G.V. (2006) Biocontrol of avocado dematophora root rot by antagonistic Pseudomonas fluorescens PCL1606 correlates with the production of 2-hexyl 5-propyl resorcinol. Mol Plant–Microbe Interact 19, 418–428. Chin-A-Woeng, T.F.C., Bloemberg, G.V., van der Bij, A.J., van der Drift, K.M.G.F., Schripsema, J., Kroon, B., Scheffer, R.J., Keel, C., et al. (1998) Biocontrol by phenazine-1-carboxamide-producing Pseudomonas chlororaphis PCL1391 of tomato root rot caused by Fusarium oxysporum f.sp. radicis-lycopersici. Mol Plant–Microbe Interact 11, 1069–1077. Chitarra, G.S., Breeuwer, P., Nout, M.J.R., van Aelst, A.C., Rombouts, F.M. and Abee, T. (2003) An antifungal compound produced by Bacillus subtilis YM10-20 inhibits germination of Penicillium roquefortii conidiospores. J Appl Microl 94, 159–166. Collins, D.P. and Jacobsen, B.J. (2003) Optimizing a Bacillus subtilis isolate for biological control of sugar beet cercospora leaf spot. Biol Control 26, 153–161. Erwin, D.C. and Ribeiro, O.K. (1996) Phytophthora Diseases Worldwide. St. Paul, Minnesota, USA: APS Press. Földes, T., Bánhegyi, I., Herpai, Z., Varga, L. and Szigeti, J. (2000) Isolation of Bacillus strains from the rhizosphere of cereals and in vitro screening for antagonism against phytopathogenic, food-borne pathogenic and spoilage microorganisms. J Appl Microl 89, 840–846. Fravel, D.R. (1988) Role of antibiosis in the biocontrol of plant diseases. Annu Rev Phytopathol 26, 75–91. Freeman, S., Sztejnberg, A. and Chet, I. (1986) Evaluation of Trichoderma as a biocontrol agent for Rosellinia-necatrix. Plant Soil 94, 163–170. Gerhardt, P., Murray, R.B.E., Costilow, R.N., Nester, E.W., Wood, W.A, Krieg, N.R. and Briggs Phillips, G. (1994) Methods for General and Molecular Bacteriology. Washington: ASM. 1958 F.M. Cazorla et al. Greenberger, A., Katan, J. and Yogen, A. (1987) Induced supressivenes in solarized soils. Phytopathology 77, 1663–1667. Joshi, R. and McSpadden Gardener, B. (2006) Identification of genes associated with pathogen inhibition in different strains B. subtilis. Phytopathology 96, 145–154. Koumoutsi, A., Chen, X.-H., Henne, A., Liesegang, H., Hitzeroth, G., Franke, P., Vater, J. and Borriss, R. (2004) Structural and functional characterization of gene clusters directing nonribosomal synthesis of bioactive cyclic lipopeptides in Bacillus amyloliquefaciens strain FZB42. J Bacteriol 186, 1084–1096. Leifert, C., Sigee, D.C., Epton, H.A.S., Stanley, R. and Knight, C. (1992) Isolation of bacteria antagonistic to postharvest fungal diseases of cold stored Brassica spp. Phytoparasitica 20, 143–149. Leifert, C., Li, H., Chidburee, S., Hampson, S., Workman, S., Sigee, D., Epton, H.A. and Harbour, A. (1995) Antibiotic production and biocontrol activity by Bacillus subtilis CL27 and Bacillus pumilus CL45. J Appl Bacteriol 78, 97–108. López-Herrera, C.J., Pérez-Jiménez, R.M., Basallote-Ureba, M.J., Zea-Bonilla, T. and Melero-Vara, J.M. (1999) Loss of viability of Dematophora necatrix in solarized soils. Eur J Plant Pathol 105, 571–576. Lugtenberg, B.J.J., de Weger, L.A. and Schippers, B. (1994) Bacterization to protect seeds and rhizosphere against disease. BCPC Monogr 57, 293–302. Lugtenberg, B.J.J., Dekkers, L.C. and Bloemberg, G.V. (2001) Molecular determinants of rhizosphere colonization by Pseudomonas. Annu Rev Phytopathol 39, 461–490. Magnet-Dana, R., Thimon, L., Peypoux, F. and Ptak, M. (1992) Surfactin ⁄ iturin A interactions may explain the synergistic effect of surfactin on the biological properties of iturin A. Biochemie 74, 1047–1051. Marten, P., Smalla, K. and Berg, G. (2000) Genotypic and phenotypic differentiation of an antifungal biocontrol strain belonging to Bacillus subtilis. J Appl Microbiol 89, 463–471. McSpadden Gardener, B.B. (2004) Ecology of Bacillus and Paenibacillus spp. in agricultural systems. Phytopathology 94, 1252–1258. Munimbazi, C. and Bullerman, L.B. (1998) Isolation and partial characterization of antifungal metabolites of Bacillus pumilus. J Appl Microl 84, 959–968. Ongena, M., Jacques, P., Touré, Y., Destain, J., Jabrane, A. and Thonart, P. (2005) Involvement of fengycin-type lipopeptides in the multifaceted biocontrol potential of Bacillus subtilis. Appl Microbiol Biotechnol 69, 29–38. Pérez-Jiménez, R. (1997) Podredumbres radiculares del aguacate (Persea americana Mill.) en el Sur de Andalucı́a. PhD Thesis, University of Málaga, Málaga, Spain. Pérez-Jiménez, R.M. (2006) A review of the biology and pathogenicity of Rosellinia necatrix – the cause of white ª 2007 The Authors Journal compilation ª 2007 The Society for Applied Microbiology, Journal of Applied Microbiology 103 (2007) 1950–1959 F.M. Cazorla et al. root rot disease of fruit trees and other plants. J Phytopathol 154, 257–266. Phae, C.G., Saski, M., Shoda, M. and Kubota, H. (1990) Characteristics of Bacillus subtilis isolated from composts suppressing phytopathogenic microorganisms. Soil Sci Plant Nutr 36, 575–586. Pliego-Alfaro, F., López-Encina, C. and Barceló-Muñoz, A. (1987) Propagation of avocado rootstocks by tissue culture. S Afr Avocado Growers’ Assoc Yrbk 10, 36–39. Potera, C. (1994) From bacteria: a new weapon against fungal infection. Science 265, 605. Powers, E.M. (1995) Efficacy of the Ryu nonstaining KOH technique for rapidly determining gram reactions of foodborne and water-borne bacteria and yeast. Appl Environ Microbiol 61, 3765–3758. Raaijmakers, J.M., Vlami, M. and de Souza, J.T. (2002) Antibiotic production by bacterial biocontrol agents. Antonie van Leeuwenhoek 81, 537–547. Rampach, G.S. and Kloepper, J.W. (1998) Mixtures of plant growth-promoting rhizobacteria enhance biological control of multiple cucumber pathogen. Phytopathology 88, 1158–1164. Romero, D., Pérez-Garcı́a, A., Rivera, M.E., Cazorla, F.M. and de Vicente, A. (2004) Isolation and evaluation of antagonistic bacteria towards the curcurbit powdery mildew fungus Podosphaera fusca. Appl Microbiol Biotechnol 64, 263–269. Romero, D., de Vicente, A., Rakotoaly, R.V., Dufour, S.E., Veening, J.W., Arrebola, E., Cazorla, F.M., Kuipers, O.P., et al. (2007) The iturin and fengycin families of lipopep- Characterization of antagonistic Bacillus tides are key factors in antagonism of Bacillus subtilis towards Podosphaera fusca. Mol Plant–Microbe Interact 20, 430–440. Schisler, D.A., Slininger, P.J., Behle, R.W. and Jackson, M.A. (2004) Formulation of Bacillus spp. for biological control of plant diseases. Phytopathology 94, 1267–1271. Schroth, M.N. and Hancock, J.G. (1982) Disease-suppressive soil and root-colonizing bacteria. Science 216, 1376–1381. Sessitsch, A., Reiter, B. and Berg, G. (2004) Endophytic bacterial communities of field-grown potato plants and their plant growth-promoting and antagonistic abilities. Can J Microbiol 50, 239–249. Shoda, M. (2000) Bacterial control of plant diseases. J Biosci Bioeng 89, 515–521. Sneath, P.H.A. (1986) Endospore-forming gram-positive rods and cocci. In Bergeys’s Manual of Systematic Bacteriology ed. Sneath, P.H.A., Mair, N.S., Sharpe, M.E. and Holt, J.G. pp. 1104–1207. Baltimore, MD: Williams & Wilkins. Stein, T. (2005) Bacillus subtilis antibiotics: structures, syntheses and specific functions. Mol Microbiol 56, 845–857. Sztejnberg, A., Freeman, S., Chet, I. and Katan, J. (1987) Control of Rosellinia necatrix in soil and apple orchard by solarization and Thrichoderma harzianum. Plant Dis 71, 365–369. Walker, R., Powell, A.A. and Seddom, B. (1998) Bacillus isolates from the spermosphere of peas and dwarf French beans with antifungal activity against Botrytis cinerea and Pythium species. J Appl Microbiol 84, 791–801. Warrior, P. (2000) Living systems as natural crop-protection agents. Pest Manag Sci 56, 681–687. ª 2007 The Authors Journal compilation ª 2007 The Society for Applied Microbiology, Journal of Applied Microbiology 103 (2007) 1950–1959 1959