Journal of Applied Microbiology 2004, 97, 12471256

doi:10.1111/j.1365-2672.2004.02408.x

Genetic and functional characterization of a Bacillus sp. strain excreting surfactin and antifungal metabolites partially identied as iturin-like compounds
G.I. Souto, O.S. Correa, M.S. Montecchia, N.L. Kerber, N.L. Pucheu, M. Bachur and A.F. Garca
tedra de Microbiologa, Facultad de Agronoma, UBA and Instituto de Investigaciones Bioqumicas y Fisiolo gicas (IBYF-CONICET), Ca noma de Buenos Aires, Argentina Ciudad Auto
2004/0328: received 23 March 2004, revised 28 June 2004 and accepted 29 June 2004

ABSTRACT
G.I. SOUTO, O.S. CORREA, M.S. MONTECCHIA, N.L. KERBER, N.L. PUCHEU, M. BACHUR AND A . F . G A R C A . 2004. I

Aims: A bacterial strain producing antifungal compounds active against the plant pathogenic fungi Fusarium, Rhizoctonia and Sclerotinia has been characterized and shown to control Rhizoctonia root rot of soya bean. Methods and Results: The metabolites excreted by Bacillus BNM 122 remained active after autoclaving, were resistant over a wide pH range and to hydrolytic enzymes. By 1H-NMR and thin-layer chromatography analyses surfactin and iturin-like compounds were partially identied. Moreover, soya bean seeds bacterization with BNM 122 in a compost-based formulation was as effective controlling Rhizoctonia solani as pentachloronitrobenzene. According to its 16S rDNA sequence BNM 122 was closely related to Bacillus amyloliquefaciens and Bacillus subtilis. PCR analysis of the 16S-23S rRNA intergenic spacer region and repetitive sequence-based PCR (rep-PCR) genomic ngerprinting revealed a close genetic relationship to B. amyloliquefaciens. However, by physiological characterization using API tests, this strain resembled more B. subtilis. Conclusions: This is the rst report describing the co-production of surfactin and iturin-like compounds by a putative strain of B. amyloliquefaciens. The synergistic effect of both lipopetides is a remarkable trait for a candidate biocontrol agent. Signicance and Impact of the Study: This kind of research has relevance in order to minimize the use of synthetic fungicides and surfactants, contributing to the preservation of the environment. Keywords: Bacillus amyloliquefaciens, biological control, compost-based formulation, fungal phytopathogens, surfactin and iturin co-production.

INTRODUCTION Plant fungal diseases reduce yield and productivity of several economical crops all over the world. Resistant plant cultivars, cultural practices and chemical applications are
Correspondence to: Olga S. Correa, Catedra de Microbiologa, Facultad de Agronoma, UBA and Instituto de Investigaciones Bioqumicas y Fisiologicas (IBYF-CONICET), Av. San Martn 4453, C1417DSE, Ciudad Autonoma de Buenos Aires, Argentina (e-mail: correa@agro.uba.ar).

routinely used to provide disease control. However, resistant cultivars for every disease are not available and cultural practices are not always economically or technologically feasible. Moreover, available chemical fungicides are often expensive and also have adverse effects on human beings. Therefore, biological control appears to constitute an alternative strategy for controlling diseases, perhaps as part of an integrated control system, thus reducing the use of chemical products and contributing to the preservation of the environment.

2004 The Society for Applied Microbiology

1248 G . I . S O U T O ET AL.

The use of bacteria as biocontrol agents has been extensively studied (Expert and Digat 1995; Asaka and Shoda 1996; Podile and Prakash 1996; Kim et al. 1997; Mao et al. 1997; Singh et al. 1998; de Vrije et al. 2001). The heatand desiccation-resistant structures of spore-producing Gram-positive bacteria can be readily formulated into stable products (Handelsman and Stabb 1996). In particular, different Bacillus species excrete peptides and lipopeptides to the culture medium, such as fungicine, iturin, bacillomicine and others, having antifungal properties (Katz and Demain 1977; Jacques et al. 1993; Zuber et al. 1993; Lebbadi et al. 1994; Eshita et al. 1995; Kajimura et al. 1995; Yakimov et al. 1995; Yu et al. 2002; Chitarra et al. 2003; Cho et al. 2003). These antifungal peptides inhibit the growth of a large number of fungi, including Aspergillus, Penicillium and Fusarium species (Munimbazi and Bullerman 1998), as well as some yeasts (Thimon et al. 1995). In this paper, we show that a Bacillus sp. strain, designated as BNM 122, was effective in vitro against several plant fungal pathogens. We also investigated the mechanisms of biological activity and we tested a compost-based formulation of BNM 122 against soya bean damping-off caused by Rhizoctonia solani. Furthermore, BNM 122 has been characterized by means of its phenotypic characteristics as well as by sequencing its 16S rDNA, PCR analysis of the 16S-23S rRNA intergenic spacer region (IGS-PCR) and repetitive sequence-based PCR (rep-PCR) genomic ngerprinting.

Fungus isolation and culture conditions The fungal strains were obtained from strain collection of the Banco Nacional de Microorganismos, Catedra de Microbiologa Agrcola, Facultad de Agronoma, Universidad de Buenos Aires, Argentina. These fungi were routinely grown on PDA at 28C and stored on the same medium at 4C. Antifungal activity of the whole culture of Bacillus sp. BNM 122 A bacterial culture grown in NB to a concentration of ca 108 CFU ml)1 was streaked in a straight line on one side of a Petri dish (3 cm from the centre) containing NA, PDA or a mix (1 : 1, v/v) of NA : PDA. Simultaneously a 9 mm diameter agar plug containing fungal mycelium of Fusarium oxysporum f. sp. lycopersici, F. solani, R. solani, or S. sclerotiorum, grown in PDA for 48 h, was placed in the centre of the plate. After 7 days at 28C the inhibitory effect on fungal growth was evaluated. All in vitro antagonism assays were made in triplicate. Activity of cell-free supernatants on S. sclerotiorum ascospore germination Cell-free BNM 122 cultures grown for 72 h were concentrated by freeze-drying to recover the excreted antifungal metabolites. The lyophilized metabolites were dissolved in distilled water to obtain a vefold concentrated preparation (5) and were sterilized by ltration (02 lm pore-size membrane). The evaluation of their activity on the germination of S. sclerotiorum ascospores was performed by mixing on microslides 25 ll of the concentrate preparation with 25 ll of ascospores suspension (2 108 ascospores ml)1 in 13% sucrose in sterile distilled water). Controls consisted of 25 ll of NB lter-sterilized, plus 25 ll of the same ascospores suspension. After incubation for 16 h at 28C in a humidity chamber, germination of the ascospores was microscopically evaluated. All assays were performed in triplicate. Stability of cell-free supernatants The concentrated preparation of antifungal metabolites was tested for resistance to temperature, pH and hydrolytic enzymes. All assays were performed as described by Lebbadi et al. (1994). To evaluate the residual antifungal activity after each treatment, 50 or 100 ll of the treated antifungal solutions were placed into wells made in opposite sides of PDA plates while, in the centre of each plate, a plug of actively growing fungal mycelium was simultaneously inoculated. Inhibition strength was

M A T E R I A LS A N D M E T H O D S Bacterial strains and culture conditions The Bacillus strain used in this paper was isolated from a sclerotium of Sclerotinia sclerotiorum withdrawn from a sunower (Heliantus annus L.) capitulum. The sclerotium was supercially disinfected with 2% (v/v) NaClO for 5 min and exhaustively washed with sterile distilled water. It was placed on potato glucose agar (PDA) and incubated at 30C for 5 days. There was an abundant growth of mycelia from the sclerotium except in a region of the plate where bacterial growth was evident. A Gram-positive, aerobic, endospore-forming, rod-shaped bacterium was isolated from that plate and it was designated as BNM 122. The reference strains used in the present study were: Bacillus amyloliquefaciens DSM 7T, B. amyloliquefaciens DSM 1060, B. subtilis subsp. subtilis DSM 10, B. subtilis DSM 1088, B. licheniformis DSM 1913 and Bacillus sp. DSM 1325, from the Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH (DSMZ). All the strains were routinely grown aerobically at 28C in nutrient broth (NB) or nutrient agar (NA). All the strains were stored at )50C in NB with 30% glycerol.

2004 The Society for Applied Microbiology, Journal of Applied Microbiology, 97, 12471256, doi:10.1111/j.1365-2672.2004.02408.x

GENETIC AND FUNCTIONAL CHARACTERIZATION OF A BACILLUS SP.

1249

assumed to be proportional to the clear, mycelia-free zones, appearing around the wells after 5 days at 28C. All assays were performed in triplicate against all fungal pathogens. Preparation of antifungal metabolites from bacterial cultures BNM 122 was grown in NB for 72 h. Cell-free supernatant (15 000 g for 15 min at 4C) was precipitated at 80% ammonium sulphate saturation and the precipitate was dissolved in distilled water. The remaining ammonium sulphate was removed by extensive dialysis against distilled water (MWCO 12 KDa, cellulose dialysis tubing from Sigma Chemical Co.). The solution was freeze-dried and stored at )20C.
1

Control seeds were treated in the same way that those bacterized but without LP. Seeds were pregerminated in moist chambers at 25C for 48 h. Plant growth chamber assays The soil used was a commercial mix (7% organic matter, 139 C/N, pH 50). Soil was articially infested by R. solani to evaluate the potential biocontrol of BNM 122. Fungal inoculum was prepared from R. solani grown for 2 weeks at 28C on autoclaved wheat seeds. Sterilized soil by tyndallization was infested with 05% fungal inoculum. Ten pregerminated seeds were planted in 1 l plastic pots lled with the soil mix with or without fungal inoculum. Plants were conducted in plant growth chamber at 25C and 12 h light (15 000 lux) and maintained at eld capacity with tap water. Treatments were: (i) nonbacterized seeds in soil without R. solani; (ii) bacterized seeds in soil with 05% fungal inoculum; (iii) nonbacterized seeds in soil with 05% fungal inoculum. Ten replicates of each treatment were performed in a completely randomized block design. Growth chamber assays were repeated three times. Greenhouse experiment Greenhouse experiments were conducted under natural temperature and light conditions. Seeds were surface disinfected in the same way as above and were sown immediately after treatments (not pregerminated). Commercial compost was sterilized by autoclaving in polypropylene bags. An LP suspension in sterile distilled water was added to 200 g of compost and kept 1 week at 28C. This formulated product with a BNM 122 concentration of 1 109 CFU g)1 was mixed with soya bean seeds plus a water suspension of 05% CMC as adhesive. Before sowing the mean number of BNM 122 adhering to the seeds (5 107 CFU per seed) was determined on NA. Treatments were: (i) seeds coated with sterile compost and sowed in soil without R. solani; (ii) seeds coated with the BNM 122 compost-based formulation and sowed in soil without fungal inoculum; (iii) seeds coated with the BNM 122 compost-based formulation in soil with 05% fungal inoculum; (iv) seeds coated with sterile compost sowed in soil with 05% R. solani containing 500 ll of pentachloronitrobenzene (PCNB) water suspension (05%, w/v) applied to soil where seed was located; (v) seeds coated with sterile compost in soil with 05% R. solani seeds coated with sterile compost in soil with 05% R. solani. Ten seeds were sowed in 1 l plastic pots lled with the soil mix. Pots were watered with tap water and maintained at eld capacity. Ten replicates of each treatment were performed in a completely randomized block design.

H-NMR analysis

A freeze-dried sample of antifungal metabolites was dissolved in acetone-d6 and 1H-NMR, correlated spectroscopy (COSY90), total correlation spectroscopy (TOCSY) and heteronuclear correlations (HMQC) were recorded on a Bruker Avance DPX 400 spectrometer operating at 400 MHz (Bruker BioSpin, Ontario, Canada). The spectra were compared with those produced by commercially available surfactin from B. subtilis (Sigma Chemical Co.). Thin-layer chromatography analysis (TLC) Cell-free supernatants from 72 h grown bacterial cultures were precipitated with 70 mM MnCl2 as described by Feignier et al. (1995). The pellet was dialysed against distilled water and freeze-dried. Antifungal metabolites were extracted from the lyophilized material with chloroform : methanol 2 : 1 (v/v) and the extract was applied to silica gel 60 TLC plates (Merck, Darmstadt, Germany) and run in chloroform : methanol : water 65 : 25 : 4 (v/v/v) as described by Sandrin et al. (1990). The Rf of the compounds were compared with those of pure lipopeptides surfactin and iturin A from B. subtilis (Sigma). Soya bean seed bacterization BNM 122 cultures grown in NB for 72 h were harvested by centrifugation and the pellet was lyophilized (LP) and kept at room temperature pending its use. Soya bean seeds were surface disinfected 2 min with 2% NaClO and exhaustively washed with sterile tap water. The bacterization was made by mixing seeds with a suspension of LP in sterile distilled water plus 05% carboxy-methyl-cellulose (CMC). The mean number of BNM 122 adhering to the seeds (4 107 CFU per seed) was determined by dilution plating.

2004 The Society for Applied Microbiology, Journal of Applied Microbiology, 97, 12471256, doi:10.1111/j.1365-2672.2004.02408.x

1250 G . I . S O U T O ET AL.

Statistical analysis Analysis of variance was performed using the general linear models procedure of SAS and mean values were compared using Fishers protected least signicant difference (SAS Institute, Cary, NC, USA). Bacterial identication Identication of BNM 122 strain was carried out using the API 50CHB and API 20E tests (bioMerieux, Marcy lEtoile, France) as recommended by the manufacturer and the sequence of its 16S rDNA was determined. These studies were complemented by 16S-23S rRNA IGS-PCR analysis and rep-PCR genomic ngerprinting (Jensen et al. 1993; Versalovic et al. 1994). The determination of the 16S rRNA gene sequence of strain BNM 122 was performed by MIDI Labs (Newark, DE, USA). For PCR ngerprinting, total genomic DNA of the bacteria was prepared from NB cultures using the Wizard genomic DNA purication kit (Promega Inc., Madison, WI, USA) and adjusted to a concentration of 50 ng ll)1. All DNA preparations were stored at 4C. To amplify the 16S-23S rDNA intergenic spacer region, primers FGPS1490 and FGPS132 were used (Laguerre et al. 1996). The reactions were carried out in a total volume of 50 ll containing 50 ng of DNA, 5% dimethyl sulphoxide, 15 mM MgCl2, 02 mM of each dNTP, 03 lM of each primer, 125 U of Taq polymerase (Promega) and the buffer provided with the enzyme. Amplications were carried out in an MJ Research PTC100 thermocycler with the following temperature programme: initial denaturation for 5 min at 95C followed by 30 cycles each consisting of denaturation (94C, 1 min), annealing (55C, 40 s) and extension (72C, 2 min) with a nal extension step at 72C for 8 min. Five ll of the PCR products were loaded onto 10 cm-long 2% Metaphor agarose gels (FMC Bioproducts, Rockland, ME, USA) and run at room temperature in TBE buffer (89 mM Tris, 89 mM Boric acid, 2 mM EDTA, pH 80) at 5 V cm)1 for 25 h. As size control, a 100 bp DNA ladder (Promega) was included. Rep-PCR genomic ngerprinting was performed with BOXA1R, REP (REP1R-I and REP2-I) and ERIC (ERIC1R and ERIC2) primers, as previously described by Versalovic et al. (1994). Eight ll of the PCR products were run in 15% agarose gels in TBE buffer at 5 V cm)1 for 2 h. As reference, 1 kb DNA ladder (Promega) was used. Gels were stained with ethidium bromide (06 lg ml)1), and photographed with a Polaroid type 667 lm. Nucleotide sequence accession number The 16S rDNA sequence determined for strain BNM 122 was submitted to the GeneBank database under accession number AF411118.

RESULTS Antifungal activity of Bacillus sp. BNM 122 Mycelia growth of F. oxysporum f. sp. lycopersici, F. solani, R. solani and S. sclerotiorum was inhibited using the dual culture technique in all media tested. Figure 1 shows the myceliar growth and sclerotia production of S. sclerotiorum

Fig. 1 Effect of Bacillus sp. BNM 122 on S. sclerotiorum growth and sclerotia production in dual culture. Plate A shows a pure culture of S. sclerotiorum and its abundant sclerotia formation. Plate B shows the fungal growth inhibition by BNM 122 (PDA : NA medium) displaying a scarce sclerotia production. An arrow points to a clear zone showing growth inhibition of mycelia

2004 The Society for Applied Microbiology, Journal of Applied Microbiology, 97, 12471256, doi:10.1111/j.1365-2672.2004.02408.x

GENETIC AND FUNCTIONAL CHARACTERIZATION OF A BACILLUS SP.

1251

in PDA : NA medium (Fig. 1A). When S. sclerotiorum was challenged with the bacterial culture (Fig. 1B) growth inhibition of mycelia occurred accompanied by a decreased sclerotia production. Effect of cell-free supernatants on S. sclerotiorum ascospores germination The germination of ascospores and the inhibitory effect of supernatant are shown in Fig. 2. There was total inhibition on S. sclerotiorum ascospores germination using the bacterial culture supernatant of 72 h (Fig. 2). We observed inhibition of ascospores germination also with 12-, 24- and 48-h supernatant (data not shown).

Stability of excreted antifungal metabolites The effects of autoclaving, pH and hydrolytic enzymes upon the antifungal activity of cell-free supernatants are shown in Table 1. No signicant difference in the size of the inhibited zone was found comparing the control and the treatment at 121C for 20 min. Treatments at pH from 40 to 100 did not affect the activity but it was completely lost at pH 20. Proteinase K, trypsin and lipase A had no effect on the antifungal activity. The same results were observed with the precipitated metabolites (data not shown). The antifungal activity of freeze-dried extracts was also stable when dissolved in chloroform : methanol 2 : 1 (v/v) or 90% acetone.
1 1

H-NMR analysis

H-NMR spectra of B. subtilis surfactin and the metabolites excreted to the culture medium by Bacillus sp. BNM 122 were identical, and in addition an unidentied minor compound was also detected in our preparation (Fig. 3). By HMQC, peak position of the seven amino acids of commercial surfactin was compared with the main metabolite of BNM 122, showing complete identity between both samples (data not shown). Analysis of antifungal components by TLC The TLC proles of the lyophilized organic extract revealed two main fractions, one showing an identical chromato-

Table 1 Stability of the antifungal metabolites excreted by Bacillus sp. BNM 122 to heat, hydrolytic enzymes and pH, tested against Sclerotinia sclerotiorum Treatments Temperature Control 121C for 20 min. Enzymes Control Proteinase K Tripsin Lipase A pH Control pH 20 pH 40 pH 80 pH 100 Inhibition zone diameter* (mm)

300 02 295 04 270 265 262 272 275 NI 273 273 270 03 05 06 03

03 02 04 03

Fig. 2 S. sclerotiorum ascospores germination and their inhibition by antifungal metabolites excreted by Bacillus sp. BNM 122. Photomicrographs (40) of ascospores germination in the presence of NB (control) and total inhibition by BNM 122 excreted metabolites (72 h supernatant). Bars 11 lm

*Diameter of mycelia-free zone (mm) around the wells inoculated with excreted antifungal metabolites. NI, no inhibition. Values are mean S.D. (n 3).

2004 The Society for Applied Microbiology, Journal of Applied Microbiology, 97, 12471256, doi:10.1111/j.1365-2672.2004.02408.x

1252 G . I . S O U T O ET AL.

2226 2075

(a) 2791

3764

0986

the results obtained in a representative growth chamber experiment. Soya bean seed treatments with strain BNM 122 (biocontrol treatment) resulted in a signicant (P 005) increase in mean stand per pot compared with the pathogen check. However, the mean stand per pot from biocontrol treatment was signicantly lower (P 005) than that of the healthy check. A reinforcement of bacterial treatment through irrigation applied to the shoot base, immediately after emergence, did not increase signicantly the number of plants per pot (data not shown). Greenhouse assay. In greenhouse, soya bean seed treatment with the compost-based formulation resulted in a mean stand per pot that was without signicant difference (P 001) with the healthy check and it was as effective as the PCNB treatment (Table 2). Soya bean seed treatment with BNM 122 together with the pathogen resulted in mean plant weight and mean plant height that were signicantly greater (P 005) than the pathogenic check but lesser in weight than the healthy and bacterial check (Table 2). On the other hand, there was no evidence of phytotoxicity to soya bean due to seed treatment with BNM 122 (bacterial check). The smaller stand of plants per pot obtained in greenhouse assay compared with the stand in growth chamber (Table 2) was because of the use of non-pregerminated seeds. Identication of strain BNM 122 Molecular and biochemical assays were used to identify the strain BNM 122. The biochemical proling obtained by using the API systems did not produce conclusive results. By using API 50CHB only 47% identity (Id) with B. subtilis and 36% Id with B. licheniformis were obtained. By combining the results of that kit with those produced using API 20E, our isolate showed 767% Id with B. subtilis, 181% Id with B. licheniformis and 37% Id with B. amyloliquefaciens. The almost complete 16S rRNA gene sequence determined for strain BNM 122 consisted of 1545 nucleotides. On the basis of 16S rDNA sequence analysis, this strain appeared to belong to the genus Bacillus, B. subtilis group, being closely related to B. subtilis and B. amyloliquefaciens. These species show a very high similarity level of their 16S rDNA sequences and are characterized by a strict phylogenetic relationship (Ash et al. 1991). The assessment of a more variable region of the rRNA operon, enabling the differentiation of these closely related Bacillus species, revealed identical IGS-PCR patterns for strain BNM 122 and also for two reference strains of B. amyloliquefaciens, and differing, from the patterns displayed by B. subtilis, B. licheniformis and Bacillus sp. (Fig. 4). Therefore, BNM 122 and the B. amyloliquefaciens strains

1911

3 p.p.m.

1308

0874

2830

2075

(b)

1301 2221 2 1911 0986

4015

0868

3 p.p.m.

Fig. 3 1H-NMR spectra of B. subtilis surfactin (a) and the antifungal metabolites excreted to the culture medium by the strain BNM 122 (b). The bar points to the main difference between both spectra

graphic mobility to B. subtilis surfactin (Rf 068) and the other one to iturin A (Rf 051). Furthermore, two additional compounds (Rf 041 and Rf 038) were detected, whereas an intensely pigmented fraction remained at the origin (data not shown). Biological control on soya bean seeds Assays in plant growth chamber. The number of plants per pot was recorded 15 days after sowing. Table 2 shows

2004 The Society for Applied Microbiology, Journal of Applied Microbiology, 97, 12471256, doi:10.1111/j.1365-2672.2004.02408.x

GENETIC AND FUNCTIONAL CHARACTERIZATION OF A BACILLUS SP.

1253

Table 2 Biocontrol of damping-off of soya bean caused by Rhizoctonia solani with Bacillus sp. BNM 122 Treatments

Growth chamber Stand* % 100

Greenhouse Stand* % Weight* (mg) % Height* (cm) % 100 97 107 98 79

Healthy check 98a Bacterial check n.d. BNM 122 + R. solani 82b PCNB + R. solani n.d. Pathogenic check 64c

81a 82a 83 83a 82a 65 55b

100 157a 101 154a 102 134b 101 127b 68 65c

100 95a 98 92a 85 102a 80 93a 41 75b

*Mean values (n 100) followed by the same letter in each column are not signicantly different (P 005). Stand: mean plant per pot, weight: mean plant dry weight per pot, height: mean plant height per pot. n.d., not determined.

Fig. 4 IGS-PCR ngerprinting of Bacillus strains. Lane: M, 100 bp DNA ladder; lane 1, B. amyloliquefaciens DSM 1060; lane 2, B. amyloliquefaciens DSM 7T; lane 3, B. licheniformis DSM 1930; lane 4, Bacillus sp. BNM 122; lane 5, B subtilis subsp. subtilis DSM 10; lane 6, B. subtilis DSM 1088; lane 7, Bacillus sp. DSM 1025

Fig. 5 rep-PCR-generated genomic ngerprints of Bacillus sp. BNM 122 and B. amyloliquefaciens strains. Lane: M, 1 kb DNA ladder; lanes 13, BOX-PCR; lanes 46, ERIC-PCR; lanes 79, REP-PCR of B. amyloliquefaciens DSM 1060, B. amyloliquefaciens DSM 7T and Bacillus sp. BNM 122 respectively

were analysed by rep-PCR in order to further determine the identity of the biocontrol strain and to distinguish those strains from each other. Figure 5 shows the ngerprints obtained with BOX, REP and ERIC primers. The strain DSM 1060 had unique genomic ngerprints with each primer set, but strains BNM 122 and DSM 7T showed identical BOX and REP-PCR ngerprints, indicating their closely genetic relationship. However, strain BNM 122 could be distinguished from B. amyloliquefaciens DSM 7T by using ERIC primers (Fig. 5).

DISCUSSION Bacillus strain BNM 122 isolated in our laboratory excreted metabolites with antifungal activity against mycelia growth of F. oxysporum f. sp. lycopercisi, F. solani, R. solani and S. sclerotiorum. Those compounds efciently inhibited in vitro ascospore germination and sclerotia production of S. sclerotiorum. The antifungal activity was resistant to high temperature, a wide range of pH and the action of many hydrolytic

2004 The Society for Applied Microbiology, Journal of Applied Microbiology, 97, 12471256, doi:10.1111/j.1365-2672.2004.02408.x

1254 G . I . S O U T O ET AL.

enzymes. These characteristics indicate that the antifungal compounds may belong to the iturin group of antibiotics (Chitarra et al. 2003). Bacteria of the genus Bacillus are known as producers of a number of peptides with antibiotic properties effective against bacteria, fungi and yeasts (Katz and Demain 1977) and also with a high stability attributable to their structure. They are small or cyclic lipopeptides having uncommon amino acids as constituents, such as ornithine or D-amino acids (Lebbadi et al. 1994; Munimbazi and Bullerman 1998). B. subtilis is considered the major producer of those antibiotic peptides and a B. subtilis strain producing iturin A and surfactin was shown to be effective for the control of damping-off caused by R. solani in tomato plants (Asaka and Shoda 1996). These previous results prompted us to search for small lipopeptides as responsible for the high antifungal activity of cell-free extracts obtained from Bacillus sp. BNM 122. 1 H-NMR analysis pointed to the presence of surfactin as a main product excreted to the medium by that strain, while the presence of some minor compounds was also evident. When we analysed these excreted metabolites using TLC, the presence of one compound with the same Rf of iturin A was revealed. In addition, two additional compounds were detected showing Rf values identical with those of iturin B and C from B. subtilis as determined by Peypoux et al. (1973). We assayed the pure lipopeptides, iturin A and surfactin, an inhibitory effect against fungal mycelium growth was produced only by iturin A (data not shown). The co-production of surfactin, which has surfactant properties, and iturin, with antifungal activity, by the same bacterial strain could be advantageous as a synergistic effect of surfactin on the activity of iturin A was earlier demonstrated (Thimon et al. 1992). This study suggests that antibiotic production was involved in the disease-suppression by BNM 122. Soya bean seed-coating with BNM 122 induced signicant protection against R. solani, under growth chamber and greenhouse conditions. Moreover, a compost-based formulation delivered to soya bean seeds was as effective as soil application of the fungicide PCNB in controlling Rhizoctonia damping-off. Bacillus sp. BNM 122 was characterized by various phenotypic and genotypic methods. With the API identication systems, strain BNM 122 showed the highest percentage of identication (76% Id) with B. subtilis and only 37% Id with B. amyloliquefaciens. However, comparison of 16S-23S rDNA IGS patterns generated by PCR suggests that strain BNM 122 could be assigned to B. amyloliquefaciens species, as the IGS-PCR ngerprints are unique for each species among members of the B. subtilis group (Wunschel et al. 1994; Daffonchio et al. 1998a,b). Moreover, rep-PCR ngerprinting was a different source of molecular evidence of the close genetic relatedness existing

between strains BNM 122 and B. amyloliquefaciens DSM 7T, supporting the assignation of strain BNM 122 to B. amyloliquefaciens. However, DNA-DNA hybridization data are required to denitely assign strain BNM 122 as belonging to that species. The antifungal activity of B. amyloliquefaciens DSM 7T was conrmed in in vitro assays and the same excreted compounds were revealed by TLC when DSM 7T and BNM 122 extracts were run on the same plate (de Estrada 2003). The commercial use of micro-organisms as biocontrol agents requires physiological and molecular ngerprints for characterization, registration, patenting and identication of introduced biocontrol strains from native microbial populations. Rep-PCR has been successfully used to identify and differentiate among different strains of the genus Bacillus (Herman et al. 1998; da Silva et al. 1999; Herman and Heyndrickx 2000; Marten et al. 2000), thus ERIC-PCR ngerprinting could be condently used for the genotyping of the strain BNM 122. The characterization of strain BNM 122 showed that it lied genetically closer to B. amyloliquefaciens than to B. subtilis although it physiologically resembled more the latter species. Our results support recent evidence that B. amyloliquefaciens strains produce iturins (Yoshida et al. 2001; Yu et al. 2002). We also determined the synthesis of a surfactant compound that was identied as surfactin. Although the iturins production by B. amyloliquefaciens has been reported, the co-production of surfactin and iturins has been reported only in B. subtilis strains (Sandrin et al. 1990; Thimon et al. 1992; Asaka and Shoda 1996; Ahimou et al. 2000). This two species are closely related and it may be that some B. amyloliquefaciens strains were declassied as B. subtilis. The results obtained with the type strain of B. amyloliquefaciens species permit us to speculate that the co-production of surfactin and iturins-like compounds could not be an uncommon trait among B. amyloliquefaciens strains. The co-production is an interesting characteristic with potential practical applications. ACKNOWLEDGEMENTS AFG wishes to acknowledge support from the Program of SETCIP: Cooperation Argentina-Germany. The authors wish to acknowledge nancial support from the Centro Argentino-Brasileno de Biotecnologa (CABBIO), Grant 13AR-07BR. REFERENCES
Ahimou, F., Jacques, P. and Deleu, M. (2000) Surfactin and iturin A effects on Bacillus subtilis surface hydrophobicity. Enzyme Microbiology Technology 27, 749754.

2004 The Society for Applied Microbiology, Journal of Applied Microbiology, 97, 12471256, doi:10.1111/j.1365-2672.2004.02408.x

GENETIC AND FUNCTIONAL CHARACTERIZATION OF A BACILLUS SP.

1255

Asaka, O. and Shoda, M. (1996) Biocontrol of Rhizoctonia solani damping-off of tomato with Bacillus subtilis RB14. Applied Environmental and Microbiology 62, 40814085. Ash, C., Farrow, J.A.E., Wallbanks, S. and Collins, M.D. (1991) Phylogenetic heterogeneity of the genus Bacillus revealed by comparative analysis of small-subunit-ribosomal RNA sequences. Letters of Applied Microbiology 13, 202206. 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 YM 1020 inhibits germination of Penicillium roqueforti conidiospores. Journal of Applied Microbiology 94, 159166. Cho, S-J., Lee, S.K., Cha, B.J., Kim, Y.H. and Shin, K-S. (2003) Detection and characterization of the Gloeosporium gloeosporioides growth inhibitory compound iturin A from Bacillus subtilis strain KS03. FEMS Microbiology Letter 223, 4751. Daffonchio, D., Borin, S., Consolandi, A., Mora, D., Manachini, P. L. and Sorlini, C. (1998a) 16S-23S rRNA internal transcribed spacers as molecular markers for the species of the 16S rRNA group I of the genus Bacillus. FEMS Microbiology Letters 163, 229236. Daffonchio, D., Borin, S., Frova, G., Manachini, P.L. and Sorlini, C. (1998b) PCR ngerprinting of whole genomes: the spacer between the 16S and 23S rRNA genes and of intergenic tRNA gene regions reveal a different intraspecic genomic variability of Bacillus cereus and Bacillus licheniformis. International Journal of Systematic Bacteriology 48, 107116. Eshita, S.M., Roberto, N.H., Beale, J.M., Mamiya, B.M. and Workman, R.F. (1995) Bacillomycin Lc, a new antibiotic of iturin group: isolation, structures and antifungal activities of the congeners. Journal of Antibiotics 48, 12401247. de Estrada, M. (2003) Mecanismos involucrados en la actividad antifungica de Bacillus amyloliquefaciens DSM7T. Thesis, Catedra de Microbiologa Agrcola, Facultad de Agronoma, UBA, Buenos Aires. Expert, J.M. and Digat, B. (1995) Biocontrol of Sclerotinia wilt of sunower by Pseudomonas uorescens and Pseudomonas putida strains. Canadian Journal of Microbiology 41, 685691. Feignier, C., Besson, F. and Michel, G. (1995) Studies on lipopeptide biosynthesis by Bacillus subtilis: isolation and characterization of iturin), surfactin+ mutants. FEMS Microbiology Letters 127, 1115. Handelsman, J. and Stabb, E.V. (1996) Biocontrol of soilborne plant pathogens. Plant Cell 8, 18551869. Herman, L. and Heyndrickx, M. (2000) The presence of intragenically located rep-like elements in Bacillus sporothermodurans is sufcient for rep-PCR typing. Research in Microbiology 151, 255261. Herman, L., Heyndrickx, M. and Waes, G. (1998) Typing of Bacillus sporothermodurans and other Bacillus species isolated from milk by repetitive element sequence based PCR. Letters in Applied Microbiology 26, 183188. Jacques, Ph., Hbid, C., Vanhentenryck, F., Destain, J., Bare, G., Razandralambo, H., Paquot, M. and Thonart, Ph. (1993) Quantitative and qualitative study of the production of broad-spectrum antifungal lipopeptide from Bacillus subtilis S499. In Proceedings of the 6th European Congress on Biotechnology, Florence, Italy, Vol. 9 ed. Albergina, L., Frontali, L. and Sensi, P. pp. 10671070. Amsterdam, the Netherlands: Elsevier Science. Jensen, M.A., Webster, J.A. and Straus, N. (1993) Rapid identication of bacteria on the basis of polymerase chain reaction-amplied

ribosomal DNA spacer polymorphisms. Applied and Environmental Microbiology 59, 945952. Kajimura, Y., Sugiyama, M. and Kaneda, M. (1995) Bacillopeptins, new cyclic lipopeptide antibiotics from Bacillus subtilis FR-2. Journal of Antibiotics 48, 10951103. Katz, E. and Demain, A.L. (1977) The peptide antibiotics of Bacillus: chemistry, biogenesis, and possible functions. Bacteriological Review 41, 449474. Kim, D., Cook, R.J. and Weller, D. (1997) Bacillus sp. L324-92 for biological control of three root diseases of wheat grown with reduced tillage. Phytopathology 87, 551558. Laguerre, G., Mavingui, P., Allard, M.R., Charnay, M.P., Louvrier, P., Mazurier, S.I., Rigottier-Gois, L. and Amarger, N. (1996) Typing of rhizobia by PCR DNA ngerprinting and PCR-restriction fragment length polymorphism analysis of chromosomal and symbiotic gene regions: application to Rhizobium leguminosarum and its different biovars. Applied and Environmental Microbiology 62, 20292036. Lebbadi, M., Galvez, A., Maqueda, M., Martnez-Bueno, M. and Valdivia, E. (1994) Fungicin M4: a narrow spectrum peptide antibiotic from Bacillus licheniformis M-4. Journal of Applied Bacteriology 77, 4953. Mao, W., Lewis, J.A., Hebbar, P.K. and Lumsden, R.D. (1997) Seed treatment with fungal or a bacterial antagonist for reducing corn damping-off caused by species of Pythium and Fusarium. Plant Disease 81, 450454. Marten, P., Smalla, K. and Berg, G. (2000) Genotypic and phenotypic differentiation of an antifungal biocontrol strain belonging to Bacillus subtilis. Journal of Applied Microbiolology 89, 463471. Munimbazi, C. and Bullerman, L.B. (1998) Isolation and partial characterization of antifungal metabolites of Bacillus pumilus. Journal of Applied Microbiology 84, 959968. Peypoux, F., Guinand, M., et Michel, G. (1973) Isolement de lacide 3-amin 12-methytetradecanoique a partir de liturine, antibiotique de Bacillus subtilis. Tetrahedron 29, 34553459. Podile, A.R. and Prakash, A.P. (1996) Lysis and biological control of Aspergillus niger by Bacillus subtilis AF1. Canadian Journal of Microbiology 42, 533538. Sandrin, C., Peypoux, F. and Michel, G. (1990) Co-production of surfactin and iturin A, lipopeptides with surfactant and antifungal properties, by Bacillus subtilis. Biotechnology and Applied Biochemistry 12, 370375. da Silva, K.R., Rabinovitch, L. and Seldin, L. (1999) Phenotypic and genetic diversity among Bacillus sphaericus strains isolated in Brazil, potentially useful as biological control agents against mosquito larvae. Research Microbiology 150, 153160. Singh, P.P., Shin, Y.C, Park, C.S. and Ching, Y.R. (1998) Biological control of Fusarium wilt of cucumber by chitinolytic bacteria. Phytophatology 89, 9299. Thimon, L., Peypoux, F., Maget-Dana, R., Roux, B. and Michel, G. (1992) Interactions of bioactive lipopeptides, iturin A and surfactin from Bacillus subtilis. Biotechnology and Applied Biochemistry 16, 144 151. Thimon, L., Peypoux, F., Wallach, J. and Michel, G. (1995) Effect of lipopeptide antibiotic, iturin A, on morphology and membrane ultrastructure of yeast cells. FEMS Microbiology Letters 128, 101 106.

2004 The Society for Applied Microbiology, Journal of Applied Microbiology, 97, 12471256, doi:10.1111/j.1365-2672.2004.02408.x

1256 G . I . S O U T O ET AL.

Versalovic, J., Schneider, M., de Bruijn, F.J. and Lupski, J.R. (1994) Genomic ngerprinting of bacteria using repetitive sequence-based polymerase chain reaction. Methods of Molecular and Cell Biology 5, 2540. de Vrije, T., Antoine, N., Buitelaar, R.M., Bruckner, S., Dissevelet, M., Durand, A., Gerlagh, M., Jones, E.E. et al. (2001) The fungal biocontrol agent Coniothyrium minitans: production by solid-state fermentation, application and marketing. Applied Microbiology and Biotechnology 56, 5868. Wunschel, D., Fox, K.F., Black, G.E. and Fox, A. (1994) Discrimination among the B. cereus group, in comparison to B. subtilis, by structural carbohydrate proles and ribosomal RNA spacer region PCR. Systematic of Applied Microbiology 17, 625635. Yakimov, M.M., Timmis, K.N., Wray, V. and Fredrickson, H.L. (1995) Characterization of a new lipopeptide surfactant produced by

thermotolerant and halotolerant subsurface Bacillus licheniformis BAS50. Applied and Environmental Microbiology 61, 17061713. Yoshida, S., Hiradate, S., Tsukamoto, T., Hatakeda, K. and Shirata, A. (2001) Antimicrobial activity of culture ltrate of Bacillus amyloliquefaciens RC-2 isolated from mulberry leaves. Phytopathology 91, 181187. Yu, G.Y., Sinclair, J.B., Hartman, G.L. and Bertagnolli, B.L. (2002) Production of iturin A by Bacillus amyloliquefaciens suppressing Rhizoctonia solani. Soil Biology and Biochemistry 34, 955963. Zuber, P., Nakano, M. and Marahiel, M.A. (1993) Peptide antibiotics: Bacillus subtilis and other Gram-positive bacteria. In Biochemistry, Physiology, and Molecular Genetics ed. Sonenshein, A.L., Hoch, J.A. and Losick, R. pp. 897916. Washington, DC: American Society of Microbiology.

2004 The Society for Applied Microbiology, Journal of Applied Microbiology, 97, 12471256, doi:10.1111/j.1365-2672.2004.02408.x