The plant-associated Bacillus amyloliquefaciens strains MEP218 and ARP23 capable of producing the cyclic lipopeptides iturin or surfactin and fengycin are effective in biocontrol of sclerotinia stem rot disease

Authors

  • F. Alvarez,

    1.  Departamento de Ciencias Naturales, Facultad de Ciencias Exactas, Físico-Químicas y Naturales, Universidad Nacional de Río Cuarto, Córdoba, Argentina
    Search for more papers by this author
  • M. Castro,

    1.  Departamento de Ciencias Naturales, Facultad de Ciencias Exactas, Físico-Químicas y Naturales, Universidad Nacional de Río Cuarto, Córdoba, Argentina
    Search for more papers by this author
  • A. Príncipe,

    1.  Departamento de Ciencias Naturales, Facultad de Ciencias Exactas, Físico-Químicas y Naturales, Universidad Nacional de Río Cuarto, Córdoba, Argentina
    Search for more papers by this author
  • G. Borioli,

    1.  Centro de Investigaciones en Química Biológica de Córdoba, (CIQUIBIC, UNC-CONICET), Departamento de Química Biológica, Facultad de Ciencias Químicas, Universidad Nacional de Córdoba, Córdoba, Argentina
    Search for more papers by this author
  • S. Fischer,

    1.  Departamento de Ciencias Naturales, Facultad de Ciencias Exactas, Físico-Químicas y Naturales, Universidad Nacional de Río Cuarto, Córdoba, Argentina
    Search for more papers by this author
  • G. Mori,

    1.  Departamento de Ciencias Naturales, Facultad de Ciencias Exactas, Físico-Químicas y Naturales, Universidad Nacional de Río Cuarto, Córdoba, Argentina
    Search for more papers by this author
  • E. Jofré

    1.  Departamento de Ciencias Naturales, Facultad de Ciencias Exactas, Físico-Químicas y Naturales, Universidad Nacional de Río Cuarto, Córdoba, Argentina
    Search for more papers by this author

Edgardo Jofré, Departamento de Ciencias Naturales, Facultad de Ciencias Exactas, Físico-Químicas y Naturales, Universidad Nacional de Río Cuarto, Ruta 36 Km 601 (X5804BYA) Río Cuarto, Córdoba, Argentina. E-mail: ejofre@exa.unrc.edu.ar

Abstract

Aims:  This work was conducted to identify the antifungal compounds produced by two previously isolated Bacillus sp. strains: ARP23 and MEP218. Both strains were subjected to further analysis to determine their taxonomic position and to identify the compounds responsible for their antifungal activity as well as to evaluate the efficiency of these strains to control sclerotinia stem rot in soybean.

Methods and Results:  The antifungal compounds were isolated by acid precipitation of cell-free supernatants, purified by RP-HPLC and then tested for antagonistic activity against Sclerotinia sclerotiorum. Mass spectra from RP-HPLC eluted fractions showed the presence of surfactin C15, fengycins A (C16–C17) and B (C16) isoforms in supernatants from strain ARP23 cultures, whereas the major lipopeptide produced by strain MEP218 was iturin A C15. Alterations in mycelial morphology and sclerotial germination were observed in the presence of lipopeptides-containing supernatants from Bacillus strains cultures. Foliar application of Bacillus amyloliquefaciens strains on soybean plants prior to S. sclerotiorum infection resulted in significant protection against sclerotinia stem rot compared with noninoculated plants or plants inoculated with a nonlipopeptide-producing B. subtilis strain.

Conclusions:  Both strains, renamed as B. amyloliquefaciens ARP23 and MEP218, were able to produce antifungal compounds belonging to the cyclic lipopeptide family. Our data suggest that the foliar application of lipopeptide-producing B. amyloliquefaciens strains could be a promising strategy for the management of sclerotinia stem rot in soybean.

Significance and Impact of the Study:  Sclerotinia stem rot was ranked as one of the most severe soybean disease in Argentina and worldwide. The results of this study showed the potential of B. amyloliquefaciens strains ARP23 and MEP218 to control plant diseases caused by S. sclerotiorum.

Introduction

Soybean (Glycine max (L.) Merril) production in Argentina has increased in the last years, reaching 27 million hectares in 2006/2007, with a record production of almost 85 million tons of grain, making Argentina the world’s third largest producer (SAGyP 2007). The expansion of single-crop agriculture has favoured the occurrence of diseases caused by fungal phytopathogens. Sclerotinia stem rot of soybean (also known as white mould) caused by Sclerotinia sclerotiorum was ranked as one of the most severe soybean diseases in Argentina and the second most important disease in the United States during the 1990s (Wrather et al. 2001). Yield loss was estimated at 83 to 335 kg ha−1 for each 10% increment in diseased plants (Danielson et al. 2004).

The most obvious symptoms of plants infected by S. sclerotiorum are necrotic tissues covered with patches of fluffy white mycelia, which produce sclerotia when nutrition is not sufficient or other conditions are favourable for sclerotial development (Christias and Lockwood 1973). The difficulty of controlling the disease resides in the ability of the fungus to overwinter in the field as sclerotia (resting bodies) that can germinate carpogenically or myceliogenically, depending on environmental conditions (Coley-Smith and Cooke 1971; Grau 1998; Bolton et al. 2006). Current cropping practices related to germplasm uniformity, lack of crop rotation and tillage allow sclerotia to survive and to reproduce. It has been demonstrated that oxalic acid is a key factor in the pathogenicity of S. sclerotiorum, and enzymes capable of degrading this compound have been utilized to produce transgenic resistant plants (Donaldson et al. 2001; Cunha et al. 2010). However, no truly resistant germplasms are available yet. Moreover, the effectiveness of fungicides in achieving disease control is quite limited because sclerotia often survive for long periods in the soil and are relatively inaccessible.

The application of chemical pesticides remains the main method for disease control (Mueller et al. 2002). Considering the problems that these compounds cause to the environment and human health, alternative and more sustainable strategies are required. Several micro-organisms have been reported as effective biocontrol agents for diseases caused by S. sclerotiorum, particularly mycoparasitic species such as Coniothyrium minitans (Whipps et al. 2008), Trichoderma harzianum (Menendez and Godeas 1998) and Sporidesmium sclerotivorum (del Rio et al. 2002), among others. In addition to these mycoparasitic species, strains of the Bacillus subtilis group have been reported as effective for the biocontrol of multiple plant diseases caused by soilborne (Asaka and Shoda 1996; Kita et al. 2005; Cazorla et al. 2007) or postharvest pathogens (Touréet al. 2004; Arrebola et al. 2009). The major fraction of the Bacillus sp. antibiotics suppressing plant pathogens are nonribosomally synthesized peptide derivatives, mainly cyclic lipopeptides. These amphiphilic compounds share a common cyclic structure consisting of a β-amino or β-hydroxy fatty acid integrated into a peptide moiety (Stein 2005). Large multienzymes, named nonribosomal peptide synthetases (NRPSs), catalyse all the necessary steps in peptide biosynthesis, including the selection and ordered condensation of amino acid residues (Sieber and Marahiel 2005). Based on their structure, cyclic lipopeptides can be generally classified into three families or groups: surfactin, iturin and fengycin (Stein 2005; Ongena and Jacques 2007).

Surfactin is a heptapeptide linked to a β-hydroxy fatty acid consisting of 13–15 carbon atoms to form a cyclic lactone structure. The surfactin operon contains four open reading frames (ORFs) encoding three multifunctional proteins SrfA-C and an external thioesterase/acyltransferase enzyme SrfD (Chen et al. 2009). Surfactin behaves as a very powerful biosurfactant and possesses several other interesting biological activities, mainly by altering membrane integrity as a consequence of the establishment of strong interactions with the constituents of the phospholipid membrane (Ahimou et al. 2000; Carrillo et al. 2003). The presence of cholesterol in the phospholipid bilayer attenuates the destabilising effect of surfactin, suggesting that the susceptibility of biological membranes to surfactin might vary in a specific manner depending on the sterol content of the target organism (Carrillo et al. 2003). This could explain why surfactin displays haemolytic, antiviral, antimycoplasma and antibacterial activities but no marked fungitoxicity (Ongena and Jacques 2007).

The iturin family comprises iturin A, C, D, E, bacillomycin D, F, L, bacillopeptin and mycosubtilin (Moyne et al. 2004). They are heptapeptides linked to a β-amino fatty acid chain with a length of 14–17 carbons. The operon consists of four ORFs called ituDABC for iturin A in B. subtillis strain RB14 (Tsuge et al. 2001). The iturin A gene cluster is inserted at exactly the same position as the bacillomycin (bmy) operon in Bacillus amyloliquefaciens strain FZB42 (Chen et al. 2009). Iturins display a strong in vitro antifungal action but limited antibacterial and no viral activities. This fungitoxicity is based on osmotic perturbation owing to the formation of ion-conducting pores (Maget-Dana and Peypoux 1994; Aranda et al. 2005) and not membrane disruption or solubilization as caused by surfactin (Heerklotz and Seelig 2001; Francius et al. 2008).

Fengycin A and B, and the closely related plipastatin, are cyclic lipodecapeptides containing a β-hydroxy fatty acid with 16–19 carbon atoms. They are synthesized by NRPSs encoded by an operon with five ORFs, fenA-E (or ppsA-E) (Steller et al. 1999). Fengycins are specifically active against filamentous fungi (Koumoutsi et al. 2004; Hu et al. 2007).

In a previous study, we showed that Bacillus sp. ARP23 and MEP218, two plant growth-promoting rhizobacteria (PGPR) isolated from soils of the Córdoba province of Argentina, were able to suppress the growth of Fusarium spp. and Sclerotinia spp. in vitro (Principe et al. 2007). In the present work, the antifungal compounds produced by Bacillus sp. ARP23 and MEP218 were characterized, and these strains were evaluated as biocontrol agents against sclerotinia stem rot on soybean.

Materials and methods

Micro-organisms and growth conditions

Bacillus strains ARP23 and MEP218 were selected according to PGPR and biocontrol characteristics from a bacterial collection obtained from a saline soil of the south of Córdoba province, Argentina (Principe et al. 2007).

The Bacillus subtilis laboratory strain JH642 (ΔtrpC2, pheA1), a derivative from B. subtilis strain 168, was kindly provided by Mansilla and De Mendoza (1997). This strain, a nonlipopeptide producer (genotypically sfp0), was used as a control for the in vitro and in vivo assays. Genomic DNA of the Bacillus strain QST713 isolated from the commercial product Serenade WPO (AgraQuest Inc., Davis, CA), kindly provided by Dr B. McSpadden Gardener, was employed as the template DNA for PCR assays (Joshi and McSpadden Gardener 2006).

All Bacillus strains were grown in Luria broth (LB) or in medium optimal for lipopeptide production (MOLP) (Gu et al. 2005) at 30°C and 150 rev min−1. Strains were routinely grown at 30°C on nutrient agar (NA) plates and stored at −80°C in LB plus 20% glycerol. Production of lipase was detected on egg-yolk medium (Willis and Growland 1960). Fermentation of lactose was evaluated according to Gordon et al. (1973).

Sclerotia of S. sclerotiorum, Sclerotinia minor and Sclerotium rolfsii were kindly supplied by Dr A. Marinelli from the laboratory of phytopathology of the Universidad Nacional de Río Cuarto, Argentina. Sclerotia, stored at 4°C, were germinated at 22°C on potato dextrose agar (PDA) (Whipps 1987) supplemented with 200 μg ml−1 of streptomycin. The fungal strains were routinely grown on carnation leaf agar plus streptomycin (Nelson et al. 1983).

Isolation of lipopeptides

Lipopeptides were isolated by the acid precipitation method (Vater et al. 2002; Kim et al. 2004). Bacillus strains were grown in LB broth until the stationary-phase growth. Bacterial cells were removed by centrifugation at 12 000 g for 15 min. Lipopeptides were precipitated from the remaining supernatants by adding 3 N HCl to a final pH 2.0 and stored for 30 min at 4°C. The precipitates were collected by centrifugation at 10 000 g for 20 min and stored at 4°C.

RP-HPLC analysis

Precipitated lipopeptides, obtained as described earlier, were suspended in 100% methanol and injected into a high-performance liquid chromatography (HPLC) system equipped with a C18 column (Varian Microsorb-100 C18, 250 × 4·6 mm, particle size 5 μm, CP30714) and a UV detector (Hewlett-Packard). The compounds were eluted at a flow rate of 1 ml min−1 with an increasing gradient of methanol from 5 to 100%. Elution was performed as follows: 5–10% methanol for 0–5 min, 10–60% for 5–20 min, 60–80% for 20–40 min and 80–100% for 40–45 min. The elution was detected at 215 nm. Eluted fractions were collected, concentrated in a centrifugal vacuum concentrator (SC110 Savant), suspended in 20 μl of 100% methanol and challenged against S. sclerotiorum. Controls consisted of LB medium and the B. subtilis strain JH642. Commercial Iturin A (Sigma-Aldrich, St Louis, MO) and surfactin (Sigma-Aldrich) were used as standards at concentrations of 0·5 and 2·5 mg ml−1, respectively.

Mass spectrometry analysis

Eluted fractions collected by RP-HPLC and showing antifungal activity were analysed in positive ion mode using a MALDI-TOF mass spectrometric instrument (Micromass; Waters, Milford, MA) equipped with a 337-nm nitrogen laser for desorption and ionization of analytes. Samples (1 μl) were mixed with an equal volume of matrix solution (5 mg ml−1α-cyano-4-hydroxycinnamic acid matrix in 50% aqueous acetonitrile containing 0·1% (v/v) trifluoroacetic acid). Mass spectra were accumulated over 20 individual laser shots with energy increasing from 5 to 80%, and peptide masses in the range 800–4000 Da were measured. Adrenocorticotropic hormone was used as the lock mass (2465·2 Da).

PCR assay

Genomic DNA preparations and agarose gel electrophoresis were performed according to standard protocols (Sambrook et al. 1989). Fragments corresponding to the ituC (465 bp) and srfA-A (1300 bp) genes involved in iturin A and surfactin biosynthesis, respectively, were amplified by PCR using the following primers: ituC-Fw (5′-AAAGGATCCAAGCGTGCCTTTTACGGGAAA-3′) and ituC-Rv (5′-AAAAAGCTT AATGACGCCAGCTTTCTCTT-3′), srfAA-Fw (5′-AAAGGATCCAGCCGAAGGGTG TCATGGT-3′) and srfAA-Rv (AAAAAGCTTGTTTTTCTCAAAGAACCAGCG-3′). Genes related to the synthesis of bacillomycin D (1200 bp, bmyA) and fengycin (293 bp, fenD) were detected by PCR employing the primers reported by Koumoutsi et al. (2004) and Joshi and McSpadden Gardener (2006), respectively.

PCR amplifications were carried out in a 10-μl reaction mixture containing 2 μl of genomic DNA, 1X PCR buffer, 50 μmol l−1 of each deoxynucleoside triphosphate (Promega, Madison, WI), 50 pmol of each primer (forward and reverse), 1·5 mmol l−1 of MgCl2 and 1.0 U of Taq DNA polymerase (Invitrogen, Carlsbad, CA). The PCR programme consisted of an initial denaturation at 95°C for 3 min, followed by 30 cycles at 94°C for 1 min, 56°C for 1 min and 70°C for 1 min. A final extension step at 70°C for 5 min was followed by a 4°C soak (Joshi and McSpadden Gardener 2006). In all cases, PCR products were sequenced (Macrogen Inc., Seoul, Korea), and the resultant sequences were compared with those present in the GenBank Database using the algorithm blastn (Altschul et al. 1997).

The recA gene was amplified by PCR using the primers and conditions described by Arguelles-Arias et al. (2009).

Antifungal compound production during growth of Bacillus strains

To determine the relationship between growth and production of antifungal compounds, Bacillus strains were grown in MOLP. Overnight bacterial cultures (500 μl) were inoculated on 250-ml flasks containing 50 ml of MOLP. The flasks were incubated on a rotator shaker at 30°C and 150 rpm. Culture samples of 1 ml were taken at different times and were used to determinate the cell density (OD600nm) as well as the antifungal activity of precipitated lipopeptides present in the cell-free supernatants. To test the antifungal activity, lipopeptides were suspended in 80% (v/v) methanol and spotted onto a paper disc on PDA plates containing a fungal plug grown 2 days before. Plates were incubated for 7 days, and the diameter of the inhibition halos was measured. One millilitre of MOLP was subjected to acid precipitation and the resulting precipitate was suspended in 80% (v/v) methanol and used as a control.

Biosurfactant activity of lipopeptides produced by Bacillus strains

For surface tension measurements, precipitated lipopeptides from the culture supernatants of the Bacillus strains grown in MOLP for 24, 48 and 72 h were suspended in 600 μl Milli-Q® water, and after 20 min, the surface tension was measured by the Wilhelmy method (MicroTroughX; Kibron Inc., Espoo, Finland) (Vater et al. 2002). The surface tension of water (72 mN m−1) was used as a reference.

In vitro antagonism of Bacillus strains cell-free supernatants on sclerotia germination

Cell-free supernatants from Bacillus strains cultures, grown at 30°C and 150 rpm for 48 h in MOLP, were filtered through a 0·45-μm pore-size filter and concentrated by freeze-drying. The lyophilized supernatants were dissolved in distilled water to obtain a fivefold (5×) concentrated preparation and further sterilized by filtration (0·2-μm pore-size filter). Sclerotia were surface sterilized by immersion in 70% (v/v) ethanol for 1 min, 10% (v/v) sodium hypochlorite for 5 min and washed seven times with sterile distilled water. Once they dried on filter paper, ten sclerotia were placed in a Petri dish containing 5-ml PDA plus 200 or 500 μl (final concentration in the medium 0·2×, or 0·5×, respectively) of concentrated cell-free supernatants obtained as described earlier. Controls consisted of PDA plates containing either concentrated MOLP or cell-free supernatants from B. subtilis JH642 cultures. Plates were incubated for 6–10 days at 28°C, and the percentage of sclerotia germination was determined.

Effect of antifungal compounds on mycelia morphology

Sclerotinia sclerotiorum was grown on PDA medium containing paper discs embedded with 100 μl of fivefold (5×) cell-free supernatants from the Bacillus strains grown in MOLP medium or 100 μl of 5× concentrated MOLP (control). After 5 days of incubation at 22°C, mycelia were taken from the colony edge and evaluated under light microscopy (Zeiss Axiophot). Images were recorded with an AxioCam HRc camera (Zeiss) and axio vision software (Zeiss).

Fungal inoculum preparation and biocontrol assays

Mycelial plugs obtained by cutting 5-mm diameter discs with a sterile cork borer from actively growing margins of S. sclerotiorum were transferred to Erlenmeyer flasks containing 300 ml of potato dextrose broth with streptomycin. The inoculated flasks were cultured at 22°C and 100 rev min−1 for 5 days. Before plant inoculation, mycelia were homogenized in the culture media by blending for 10 s, and the optical density at 600 nm (OD600) of the suspension was adjusted to 1·5 to obtain a uniform mycelia concentration (Chen and Wang 2005).

Soybean seeds (Don Mario 4600) were surface sterilized by treatment with 95% (v/v) ethanol for 5 min, 10% (v/v) sodium hypochlorite for 15 min and then rinsed six times with sterilized distilled water. Surface-sterilized seeds were germinated aseptically on water-soaked filter paper in Petri dishes. Seedlings were transferred to polyethylene bags with sterile soil and volcanic sand (1 : 1) and further cultivated in a plant growth chamber (24°C, light 16 h per day).

For inoculation of soybean plants, Bacillus strains ARP23, MEP218 and B. subtilis JH642 grown in MOLP at 30°C and 150 rev min−1 for 48 h (approximately 108 CFU ml−1) were diluted 20-fold and amended with 0·004% humectant adherent (NITRAP; Bioverde, Buenos Aires, Argentina). The preparations were applied to soybean plants at the V3 stage (Fehr et al. 1971) by spraying, until run-off, with a hand-pump sprayer at a rate of 5 ml per plant (2·5 × 107 CFU per plant). After air-drying for 1 h, each plant was treated with 5 ml of a S. sclerotiorum suspension prepared as described earlier. Treatments consisted of plants inoculated with the strains ARP23, MEP218 and JH642 or water before spraying of S. sclerotiorum. Polyethylene bags containing plants of each treatment were arranged in glass boxes (45 × 60 × 39 cm) with a relative humidity of 100%. Disease severity was determined at 7 and 21 days after inoculation rated on a scale of 0–5, where 0 =  no disease, 1 = lesion <5 mm on leaf, 2 = lesion >5 mm on leaf, 3 = lesion >5 mm on leaf and start of stem colonization, 4 = severe stem colonization and 5 = plant death (Mueller et al. 2002). The experiment was performed three times with eighteen plants per treatment.

Statistical analysis

The program statgraphics Plus version 4.1 was used for the statistical analysis of the biocontrol data. The Kruskal–Wallis one-way analysis of variance by ranks, which is a nonparametric method, was used to test the equality of population medians among treatments (P < 0·05).

Results

Taxonomic position of the Bacillus strains APR23 and MEP218

In a previous work, we described the isolation and characterization of two PGPR strains capable of inhibiting the growth of the phytopathogenic fungi Sclerotinia spp. and Fusarium spp. in vitro (Principe et al. 2007). According to the analysis of the 16S rDNA nucleotide sequences, these strains, designed MEP218 (formerly Bacillus sp. LZ) and ARP23 (GenBank accession numbers: DQ343613 and DQ343615, respectively), were included into the B. subtilis group, which contains the closely related B. subtilis and B. amyloliquefaciens species, among others (Wattiau et al. 2001; Fritze 2004). To accurately characterize these strains taxonomically, we performed a more extensive analysis including recA sequencing (Zeigler 2003; Arguelles-Arias et al. 2009) and biochemical tests based on the observation that B. amyloliquefaciens strains are distinguished from B. subtilis by their ability to produce lipase as well as acid from media containing lactose as the only carbon source (Priest et al. 1987; Idriss et al. 2002). The obtained phylogenetic tree constructed on the recA gene (GenBank accession numbers: JN024700 and JN024701 for strains MEP218 and ARP23, respectively), which encodes for a DNA repair and recombination protein, demonstrated that these strains were more related to B. amyloliquefaciens than to B. subtilis (Fig. 1). Moreover, both strains were capable of producing lipase as well as acid from lactose, unlike the reference strain B. subtilis JH642 (data not shown). According to these results, strains MEP218 and ARP23 were reassigned as B. amyloliquefaciens MEP218 and B. amyloliquefaciens ARP23.

Figure 1.

 Comparative sequences analysis of 16S rRNA (a) and recA (b) genes among Bacillus amyloliquefaciens strains ARP23 and MEP218 and other representatives of the Bacillus genus. The tree was constructed through the use of the Jukes-Cantor distance and neighbour-joining methods. The significance of each branch is indicated by a bootstrap for 1000 subsets. Strain ARP23: ARP3; Strain MEP218: MEP18.

Identification of antifungal compounds produced by B. amyloliquefaciens strains ARP23 and MEP218

The antifungal activity shown by B. amyloliquefaciens strains ARP23 and MEP218 was previously detected in cell-free culture supernatants (Principe et al. 2007). To identify the compounds responsible for in vitro antifungal activity, reverse-phase high-performance liquid chromatography (RP-HPLC) and matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF-MS) analyses were carried out. Antifungal compounds were isolated by acid precipitation from cell-free supernatants of the Bacillus strains cultures, suspended in 100% methanol and subjected to RP-HPLC on a C18 column with a linear gradient from 5% to 100% methanol. Fractions were collected and assayed for in vitro antagonism against S. sclerotiorum. At least three and two active fractions, which showed strong antifungal activity, were found in the supernatants from Bacillus amyloliquefaciens ARP23 and MEP218, respectively. Additionally, a major fraction from strain ARP23 without antifungal activity was also characterized. These fractions eluted at 60–75% of methanol with retention times from 17 to 23 min, comparable with the retention times observed for the main fractions of iturin A and surfactin lipopeptides used as controls (data not shown). As expected, HPLC analysis of culture media (LB) or cell-free supernatants from B. subtilis JH642 did not show any active fraction (data not shown).

MALDI-TOF mass spectrometry analysis of the purified active fractions led to the identification of cyclic lipopeptides belonging to the three known groups: iturin, surfactin and fengycin. By comparing the mass data obtained from individual fractions with the mass numbers reported for the cyclic lipopeptides from other Bacillus subtilis group strains, the major lipopeptidic products of strain ARP23 were identified as surfactin (C14–C15), fengycin A (C16–C17) and fengycin B (C16), whereas iturin A (C15) was the only lipopeptide detected in culture supernatants from strain MEP218 (Fig. 2a–f). The mass spectra of these lipopeptides showed peaks consistent with the protonated forms, as well as with the sodium and potassium adducts. The mass spectra of the eluted fractions with antifungal activity from strain ARP23 showed peaks at m/z 1058·5, 1080·5 and 1096·5, which were attributed to [M + Na]+, [M + 2Na]+ and [M + Na + K]+, respectively, of the C15 isoform Leu/Ile-7 of surfactin (Fig. 2). In the same strain, the HPLC peak eluted at 19·9 min, which did not show antifungal activity corresponded to the C14 isoform of surfactin.

Figure 2.

 MALDI-TOF mass spectrometric analysis of cyclic lipopeptides present in fractions eluted from RP-HPLC (detection at 220 nm) of Bacillus amyloliquefaciens strains. Mass spectra corresponded to surfactin isoform Leu/Ile-7 C15 in fractions eluted at 17·7 (a) and 18·9 (b) min, surfactin isoform Leu/Ile-7 C14 eluted at 19·9 (c) min, fengycin isoforms Ala-6 C16-17 and Val-6 C16 eluted at 22·4 min (d) from strain ARP23, and iturin A isoform Asn-1 C15 in fractions eluted at 21·1 (e) and 21·7 (f) min from strain MEP218.

PCR detection of nonribosomal lipopeptide synthetases

PCR assays using specific primers were conducted to detect genes related to the biosynthesis of cyclic lipopeptides in the genomes of the B. amyloliquefaciens strains. The fenD, bmyA and srfA-A genes involved in fengycin, bacillomycin D and surfactin synthesis, respectively, were detected in strain ARP23. The same amplicons were also detected in strain MEP218, although the ituC gene, involved in iturin A production, was detected instead of bmyA. In strain QST 713, the active constituent of the biocontrol agent ‘Serenade®’, used as a positive control, showed all genetic markers, ituC, fenD, bmyA and srfA-A, whereas strain JH642 only showed the presence of the surfactin biosynthetic gene (srfA-A) (data not shown). DNA sequences resulting from the sequencing of the PCR-amplified products showed a high identity to the peptide synthetase genes of Bacillus subtilis and Bacillus amyloliquefaciens strains deposited in the GenBank database. Thus, sequence of the ituC gene from B. amyloliquefaciens MEP218 showed 100% homology with the ituC gene from B. subtilis RB14 (Tsuge et al. 2001), while the srfA-A, bmyA and fenD genes from B. amyloliquefaciens ARP23 showed 98, 100 and 99% homology with the srfA-A and bmyA genes from B. amyloliquefaciens FZB42 (Koumoutsi et al. 2004) and the ppsD gene from B.  subtilis 168 (Kunst et al. 1997), respectively. These results confirm the presence of operons involved in lipopeptide production in the strains ARP23 and MEP218.

Relationship between cell growth, antifungal activity and biosurfactant activity

Once it was determined that the antifungal compounds produced by Bacillus strains were lipopeptides, the dynamics of lipopeptide production were monitored along with the growth of the Bacillus strains in MOLP. At different time intervals, samples were taken, and lipopeptides were precipitated and assayed against S. sclerotiorum by measuring the diameter of the inhibition halo. As shown in Fig. 3, the antifungal activity started to become evident during the exponential-phase growth, reaching the maximum at the stationary-phase growth. For both strains, the strongest fungal inhibition was observed at 48 h of incubation and then the activity declined over the next 24 h (Fig. 3a,b).

Figure 3.

 Relationship between the cell density of Bacillus amyloliquefaciens strains ARP23 (a) and MEP218 (b) and the production of antifungal compounds over 72 h in a medium optimal for lipopeptide production. Means of the growth curve measured by optical density at 600 nm (•). Metabolites were purified from different growth phases by acid precipitation and suspended in 80% methanol to evaluate the diameters of the inhibition halos against Sclerotinia sclerotiorum (bsl00001). Data represent the average of three repetitions ± SE.

Taking into account these results, the antifungal activity of lipopeptides produced at 24, 48 and 72 h of culture by strains ARP23 and MEP218 was evaluated against S. minor and S. rolfsii (Table 1). Additionally, the biosurfactant activity of these lipopeptides was also determined.

Table 1.   Growth inhibition of Sclerotinia spp. and Sclerotium rolfsii by lipopeptides produced by Bacillus strains at different time points of culture
FungiB. amyloliquefaciens ARP23B. amyloliquefaciens MEP218B. subtilis JH642
24 h48 h72 h24 h48 h72 h24 h48 h72 h
  1. +/−, inhibition halo ≤0·7 cm; −, inhibition halo = 0 cm.

  2. References: Sclerotinia sclerotiorum (Ss), Sclerotinia minor (Sm) and Sclerotium rolfsii (Sr).

  3. Data are mean ± SE of the diameter of the inhibition halos in cm from three independent experiments (n = 3).

Ss3·23 ± 0·035·23 ± 0·184·76 ± 0·272·03 ± 0·092·60 ± 0·212·23 ± 0·15
Sm2·67 ± 0·093·40 ± 0·382·75 ± 0·272·8 ± 0·103·83 ± 0·443·60 ± 0·50
Sr+/−+/−+/−+/−1·17 ± 0·091·25 ± 0·05

Sclerotinia sclerotiorum was more inhibited by lipopeptides produced by strain ARP23, whereas the iturin-producer strain MEP218 showed the strongest activity against S. minor and S. rolfsii (Table 1). In all cases, the highest antifungal activity (as determined by the diameter of the inhibition halos) was achieved at 48 h of culture.

It is well known that cyclic lipopeptides decrease water surface tension. To evaluate the biosurfactant activity of lipopeptides present in cell-free supernatants of the Bacillus strains, different mixtures of lipopeptides and water were made, and the surface tension of the resulting mixtures was measured. According to the values shown in Table 2, the presence of lipopeptides produced by strains ARP23 and MEP218 after 24 h of incubation decreased the surface tension of water from 72 mN m−1 to approximately 32 and 28 mN m−1, respectively, whereas lipopeptides produced at 48 h or 72 h of incubation decreased the surface tension of water to approximately 40 mN m−1. Surprisingly, the time of culture in which the highest antifungal activity was observed did not correlate with that in which the highest biosurfactant activity was determined. This result suggests that other surfactant compounds with no antifungal activity are being released to the medium before 48 h of culture or that the change in the composition of lipopeptides during growth may induce micellization with a consequent lowering of biosurfactant activity, but not of antifungal activity (Thimon et al. 1992). As expected, the JH642 strain showed neither antifungal nor biosurfactant activity.

Table 2.   Biosurfactant activity of lipopeptides produced at different times of Bacillus strains cultures
Strain*Surface tension (mN m−1)
24 h48 h72 h
  1. Data are mean ± SE from three determinations.

  2. *Bacillus strains were grown in medium optimum for lipopeptide production at 30°C. The surface tension of this medium was 72 mN m−1.

B. subtilis JH64270·05 ± 1·0269·43 ± 1·7271·56 ± 0·52
B. amyloliquefaciens ARP2332·20 ± 1·140·33 ± 2·0644·0 ± 0·4
B. amyloliquefaciens MEP21828·36 ± 0·7741·4 ± 1·043·5 ± 0·5

Effect of cell-free supernatants on sclerotia germination and fungal mycelium morphology

To evaluate whether the antifungal metabolites present in culture supernatants from Bacillus strains were capable of inhibiting the germination of sclerotia produced by S. sclerotiorum, S. minor and S. rolfsii, cell-free supernatants from 48-h cultures of strains ARP23 and MEP218 grown in MOLP were included in fungal growth media (PDA medium) and the percentage of sclerotia germination was assessed. Sclerotia of S. sclerotiorum showed a 100% inhibition when PDA medium was supplemented with 200 μl (data not shown) and 500 μl of fivefold concentrated supernatants from strains ARP23 and MEP218, respectively (Fig. 4A3 and A4). In contrast, sclerotial germination of S. minor and S. rolfsii was inhibited by 50% and 25%, respectively, when the medium was supplemented with 500 μl of concentrated supernatants from both strains (Fig. 4B3, B4, C3 and C4). In all fungal species, the germination of sclerotia was not affected on PDA medium (Fig. 4A1, B1 and C1) or PDA containing MOLP 5X (Fig. 4A2, B1 and C1).

Figure 4.

 Inhibition of sclerotia germination of Sclerotinia sclerotiorum (A1–A4), Sclerotinia minor (B1–B4) and Sclerotinia rolfsii (C1–C4) by concentrated cell-free supernatants from Bacillus amyloliquefaciens ARP23 (A3, B3 and C3) and B. amyloliquefaciens MEP218 (A4, B4 and C4) on potato dextrose agar (PDA) after 10 days of incubation. PDA (A1, B1 and C1) or PDA-containing concentrated medium optimum for lipopeptide production (A2, B2 and C2) was used as controls. Final concentration of cell-free supernatants in the medium was 0·5×.

On the other hand, S. sclerotiorum mycelia grown in the presence of antifungal compounds produced by strains ARP23 and MEP218 displayed a deformed morphology characterized by swollen hyphal cells and extensive vacuolization, corresponding to disorganization of the cytoplasm that became more granulated (Fig. 5c,d). Moreover, some parts of the swollen hyphae exhibited extreme morphological alterations, resulting in the formation of swollen bodies of spherical or irregular shape, especially in mycelia treated with the iturin-producing strain MEP218 (Fig. 5d). As expected, neither MOLP nor B. subtilis JH642 supernatants induced morphological alterations of fungal hyphae (Fig. 5a,b).

Figure 5.

 Effect of antifungal compounds present in concentrated cell-free supernatants from Bacillus amyloliquefaciens strains on Sclerotinia sclerotiorum mycelia. Untreated (a) and treated mycelia with 100 μl of fivefold concentrated medium optimum for lipopeptide production medium (b) or 100 μl of fivefold concentrated cell-free supernatants from B. amyloliquefaciens strains ARP23 (c) and MEP218 (d) were evaluated by light microscopy after 5 days of incubation. Bars: 10 μm.

Biocontrol of sclerotinia stem rot by Bacillus amyloliquefaciens strains ARP23 and MEP218

The effectiveness of the lipopeptide-producing B. amyloliquefaciens strains ARP23 and MEP218 to control sclerotinia stem rot of soybean was evaluated under growth chamber conditions. 48-h-old stationary cultures of strains ARP23 and MEP218 were inoculated by spraying on V3-stage soybean plants that were further infected with S. sclerotiorum. At 7 and 21 days post-treatment, there were no significant differences (P < 0·05) among the inoculated plants (treatments ARP23 + Ss and MEP218 + Ss) with respect to untreated plants (healthy control). In contrast, those plants that were not inoculated and infected with the fungal pathogen (treatment Ss) showed the greatest disease severity (Fig. 6a). The symptoms were characterized by expanded foliar lesions and fungus-colonized stems after the first week (average disease severity of 3·7) and dead plants at the end of the experiments. To verify that the biocontrol effect was caused mainly by lipopeptides produced by strains ARP23 and MEP218, soybean plants were inoculated with the B. subtilis strain JH642 (sfp0). At 7 days after application, plants inoculated with strain JH642 and infected with S. sclerotiorum showed significant disease severity compared with plants inoculated with strains ARP23 or MEP218 and infected with S. sclerotiorum. At 21 days postinoculation, most of the plants inoculated with strain JH642 showed severe symptoms of disease (mean disease severity of 4·8) similar to those plants that were not inoculated and infected with S. sclerotiorum (Fig. 6b). These results showed that foliar application of B. amyloliquefaciens strains ARP23 or MEP218 before pathogen challenge resulted in a protective effect of soybean plants against sclerotinia stem rot disease.

Figure 6.

 Severity of Sclerotinia stem rot symptoms after treatment with Bacillus amyloliquefaciens strains ARP23 and MEP218. (a) Soybean plants at the V3 stage were evaluated after seven (light grey bars) and 21 days (dark grey bars) postinoculation for disease symptoms. The severity index was rated on a scale of 0–5 (see Materials and methods). Bars represented the average ± SE per treatment (n = 18) of three independent experiments. Except for the uninfected control plants, leaves were first sprayed with diluted (1/20) 48-h-old cultures of Bacillus strains (2·5 × 107 CFU per plant) and one hour later with Sclerotinia sclerotiorum mycelia adjusted to an optical density of about 1·5 at 600 nm. References: S. sclerotiorum (Ss), B. amyloliquefaciens ARP23 (ARP3) B. amyloliquefaciens MEP218 (MEP18) and B. subtilis JH642 (JH642). Treatments sharing a common letter were not significantly different (P = 0·05) using the Kruskal–Wallis test. (b) Representative soybean plants of each treatment 21 days postinoculation.

Discussion

Members of the Bacillus genus are among the beneficial bacteria mostly exploited as biopesticides (Fravel 2005). Their protective effect may rely on different mechanisms to directly antagonize pathogen growth. In this context, Bacillus subtilis group produces a variety of bioactive metabolites that could be involved in antibiosis (Stein 2005; Chen et al. 2009).

Antibiosis is one mechanism of biological control that is well characterized genetically as well as biochemically in Bacillus strains (Stein 2005). In recent years, cyclic lipopeptide antibiotics have been reported to inhibit a wide range of phytopathogenic fungi (Ongena and Jacques 2007; Nagórska et al. 2007).

In this study, antifungal compounds produced by novel Bacillus amyloliquefaciens strains ARP23 and MEP218 were identified, and the ability of these strains to suppress sclerotinia stem rot of soybean caused by Sclerotinia sclerotiorum was evaluated under growth chamber conditions. In a previous work, we demonstrated the ability of these strains to inhibit the growth of sclerotium-forming phytopathogenic fungi by an antibiosis mechanism through the release of protease-resistant and thermo-stable compounds into the culture medium (Principe et al. 2007). Here, analysis by HPLC and mass spectrometry of cell-free supernatants from B. amyloliquefaciens strains proved the occurrence of three families of known cyclic lipopeptides: surfactins, fengycins and iturins. According to the mass spectra, isoforms of surfactins (C14 and C15) as well as fengycins A (C16 and C17) and B (C16) were coproduced by strain ARP23, while iturin A C15 was detected in strain MEP218. Strong antifungal activity was observed in the strains producing fengycins and iturin A, which is coincident with previous reports (Asaka and Shoda 1996; Cho et al. 2003; Koumoutsi et al. 2004; Touréet al. 2004; Hu et al. 2007; Arrebola et al. 2009), but only surfactin isoform C15 showed bioactivity against S. sclerotiorum. It is well documented that surfactin is a powerful biosurfactant (Maget-Dana et al. 1992; Heerklotz and Seelig 2001), antimicrobial agent (Bais et al. 2004) and it has a role in biofilm formation (Hofemeister et al. 2004), but has weak antifungal activity by itself (Ongena and Jacques 2007). The antifungal effect of surfactin has been more connected with some synergistic effect with iturin A (Thimon et al. 1992; Souto el al. 2004; Joshi et al. 2007) or fengycin (Koumoutsi et al. 2004; Romero et al. 2007). However, relatively few Bacillus strains exhibit antifungal activity through the production of surfactin C15 (1035 Da) (Tendulkar et al. 2007; Snook et al. 2009). Interestingly, in this study, we identified by MALDI-TOF mass spectrometry one peak eluted from HPLC, showing antifungal activity as surfactin C15 in B. amyloliquefaciens strain ARP23.

In addition to the identification of cyclic lipopeptides, the presence of nonribosomal peptide synthetase genes involved in the synthesis of cyclic lipopeptides was detected in B. amyloliquefaciens strains ARP23 and MEP218 by PCR-based assays. The use of genetic markers such as lipopeptide synthetases genes, associated with biological control activities, has been proposed as a tool for the identification and selection of novel biocontrol agents from environmental samples (Joshi and McSpadden Gardener 2006; Athukorala et al. 2009; Tapi et al. 2010). In our study, in all cases, it was possible to associate the presence of lipopeptides in the culture supernatants of B. amyloliquefaciens strains with the corresponding biosynthetic genes. However, the opposite situation did not happen. Thus, the peptide synthetase genes involved in the production of bacillomycin, in strain ARP23, or fengycin and surfactin, in strain MEP218, were detected by PCR, but these lipopeptides were not found in the culture supernatants from these strains.

It has been reported that the presence of a particular gene is not necessarily a good predictor of the biocontrol capacity in a Bacillus strain (Joshi and McSpadden Gardener 2006). For example, Bacillus subtilis JH642 contains an intact surfactin operon, but it is unable to produce the antibiotic because of a frameshift mutation on the sfp gene encoding for 4′-phosphopantetheinyl transferase, which is required to convert peptide synthetases into their active holoforms (Mootz et al. 2001).

The maximum production of antifungal compounds by Bacillus strains was observed during the stationary growth phase in an optimal medium for lipopeptide production. However, this maximum production of antifungal compounds did not correlate with the highest biosurfactant activity, which was registered during the exponential-phase growth, suggesting that in B. amyloliquefaciens strains ARP23 and MEP218, other surface-active metabolites produced during the exponential-phase growth have a more predominant role as biosurfactants than as antifungals.

It has been reported that the production of lipopeptides with strong antifungal activity, such as iturin or fengycin, is articulately delayed (late stationary-phase growth) as compared with surfactin (Stein 2005). Surfactin molecules exhibit higher cell-surface hydrophobicity and reduction in surface tension of the medium compared with iturin molecules (Maget-Dana et al. 1992). Concordantly, Murkherjee and Das (2005) found a surfactin-producer strain that exhibited greater surface tension reduction in the growth medium but was shown to be poor in inhibiting micro-organisms. However, at a later stage, a possible increase in the more biocidal active but less surface-active iturin may account for our observations.

Massive reproductive potential along with the capability for long-term survival makes sclerotia central components in the epidemiology of S. sclerotiorum diseases. Sclerotia can remain viable for up to 8 years in soil, germinate myceliogenically or carpogenically, and start a new cycle of infection on soybean or other host plants. The lipopeptides produced by B. amyloliquefaciens strains showed a significant inhibition on the germination of Sclerotinia spp. and Sclerotium rolfsii sclerotia. This ability to inhibit the germination of sclerotia is particularly important because soybean sclerotinia stem rot disease is caused by infection from ascospores produced from germinating sclerotia. Therefore, decreased sclerotial production or delayed germination of sclerotia resulting from the field application of B. amyloliquefaciens strains could be considered as a strategy to reduce primary infection.

In addition to the inhibition of sclerotia germination, lipopeptides produced by B. amyloliquefaciens strains notably affected S. sclerotiorum mycelial morphology, revealing a granulated and vesicular cytoplasm compared with the hyaline and healthy cytoplasm of control untreated hyphae. Similar effects have been reported by other authors on phytopathogenic fungi, suggesting that biosurfactant lipopeptides can traverse the fungal cell wall and induce serious alterations in plasma membrane permeability (Romero et al. 2007; Tendulkar et al. 2007). Differences observed in mycelial morphology in the presence of lipopeptides produced by strain ARP23 or strain MEP218 could be associated with their different mode of action; surfactin and fengycin, produced by strain ARP23, act through their detergent-like properties (Francius et al. 2008, Deleu et al. 2008), whereas iturin, produced by strain MEP218, causes an osmotic perturbation owing to the formation of ion-conducting pores (Maget-Dana and Peypoux 1994; Aranda et al. 2005).

Bacillus amyloliquefaciens strains ARP23 and MEP218 were sprayed on the foliar area of soybean plants before pathogen challenge. This application method resulted in effective suppression of disease severity under greenhouse conditions. As expected, unlike strains ARP23 and MEP218, the nonlipopeptide producer B. subtilis strain JH642 did not show a protective effect against sclerotinia stem rot. These results are consistent with those observed in vitro, demonstrating a clear contribution of iturin A, surfactin and fengycin in the antagonism of B. amyloliquefaciens strains towards S. sclerotiorum. The application of broth cultures provides advantages over the inoculation of cell or lipopeptide suspensions. In agreement with this observation, Zhang and Xue (2010) observed that application of broth cultures from B. subtilis strain SB24 provided slightly better efficacy in terms of control of sclerotinia stem rot than cell suspensions followed by cell-free filtrates. Moreover, lipopeptide production has been demonstrated for Bacillus populations growing on plant tissues such as roots, leaves and fruits (Touréet al. 2004; Romero et al. 2007). Romero et al. (2007) detected in situ the production of surfactin, fengycin and iturin A in Bacillus-treated melon leaves, which contributed to a reduction in disease symptoms related to powdery mildew. Recently, it was demonstrated that surfactin is the only cyclic lipopeptide synthesized during colonization of different plant roots (Fan et al. 2011).

Increased concern over the impact of chemical pesticides on the environment has resulted in increased interest in biocontrol strategies for the management of S. sclerotiorum. The most predominantly studied biocontrol agents for the management of S. sclerotiorum are mycoparasites, such as Coniothyrium minitans and Sporidesmium sclerotivorum (Bolton et al. 2006). However, few attempts have been made to explore the possibility of bacterial biocontrol agents for the management of sclerotinia diseases (Fernando et al. 2007; Abdullah et al. 2008; Zhang and Xue 2010). To our knowledge, this is the first report of antibiosis through iturin- or surfactin- and fengycin-producing B. amyloliquefaciens strains in phyllosphere biocontrol of sclerotinia stem rot disease in soybean. Foliar application of strains ARP23 and MEP218 provided a significant suppression of sclerotinia stem rot under controlled conditions, indicating that they could be potential strains for the control of disease under field conditions.

Further research is needed to verify the production of lipopeptides by B. amyloliquefaciens over time. Field evaluations as well as proper strain formulations need to be conducted to ascertain the potential of B. amyloliquefaciens strains ARP23 and MEP218 for large-scale use to control diseases caused by S. sclerotiorum.

Acknowledgements

This investigation was supported by the Secretaría de Ciencia y Técnica de la Universidad Nacional de Río Cuarto, the Agencia Nacional de Promoción Científica y Tecnológica and the Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET), Argentina. We thank Dr Hugo García Ovando for providing the HPLC equipment. We are also grateful to María Isabel Mora and Fernando Corrales from the Proteomics and Bioinformatics Unit of the Applied Medical Research Centre (CIMA) for their intensive cooperation in the MALDI-TOF-MS analysis.

Ancillary