OUP user menu

Characterization of a free-living maize-rhizosphere population of Burkholderia cepacia: effect of seed treatment on disease suppression and growth promotion of maize

Annamaria Bevivino , Sabrina Sarrocco , Claudia Dalmastri , Silvia Tabacchioni , Cristina Cantale , Luigi Chiarini
DOI: http://dx.doi.org/10.1111/j.1574-6941.1998.tb00539.x 225-237 First published online: 1 November 1998

Abstract

A Burkholderia cepacia population naturally occurring in the rhizosphere of Zea mays was evaluated by metabolic and molecular profiling and for some traits associated with biocontrol and plant growth promoting (PGP) activity. The purpose was to investigate the potential of this bacterial species closely associated with maize to act as a PGP inoculant. The bacterial strains, isolated on semiselective PCAT medium, were assigned to the species B. cepacia by an analysis of the restriction patterns produced by amplified DNA encoding 16S rRNA (16S rDNA) (ARDRA) with the enzyme AluI. Biodiversity among the 14 B. cepacia isolates was analyzed by the Biolog GN system and by the random amplified polymorphic DNA (RAPD) technique with two 10-mer primers. The analysis of the Biolog data revealed that all rhizosphere B. cepacia strains formed a tight phenetic cluster which includes B. cepacia LMG 11351, a reference strain isolated from the maize rhizosphere, and allowed us to distinguish single isolates from one another. The analysis of RAPD patterns allowed us to identify two principal groups in this bacterial population. Other tests included in vitro inhibition of the maize pathogens Fusarium spp., analysis of siderophore production, bioassay using maize seeds coated with B. cepacia in soil artificially infested with the maize pathogens Fusarium spp., and greenhouse-based plant growth promotion experiments with maize. The data obtained demonstrated that all B. cepacia strains displayed a wide antibiosis against the phytopathogenic fungi studied and produced, under low iron conditions, hydroxamate-like and thiazo-like siderophores. Moreover, the bioassay allowed us to select six and eight B. cepacia strains with a potential for the biological control of F. proliferatum ITEM-381 and F. moniliforme ITEM-504, respectively. Growth promotion experiments showed that the effect of seed bacterization with B. cepacia isolates on maize growth depended on the potting medium used. When a sand-peat/manure mixture was used, almost all B. cepacia isolates promoted maize growth; whereas, when the soil collected from the field of bacterial isolations was used, only four strains exerted a positive effect on maize growth.

Keywords
  • Zea mays
  • Burkholderia cepacia
  • Biolog
  • Random amplified polymorphic DNA
  • Plant growth promoting rhizobacterium
  • Fusarium

1 Introduction

The use of plant growth promoting rhizobacteria (PGPR) has great promise in agricultural crop production systems [1]. Growth promotion caused by these organisms may result either from indirect action, such as biocontrol of soilborne diseases through competition for nutrients, siderophore-mediated competition for iron, antibiosis, and induction of systemic resistance (ISR) activity in the plant [2], or from direct action, providing the host plant with fixed nitrogen, phosphorus and iron solubilized from the soil, and phytohormones [3]. In order to be effective, bacterial inoculants have to colonize plant roots and survive in the rhizosphere at least for the time they exert their positive effects on plants. This is possible when the inoculated bacteria are able to compete with the well-established and highly competitive indigenous microflora. This property, known as rhizosphere competence [4], appears to be a prerequisite for successful biological control of root diseases by means of microorganisms applied to seeds.

With the growing interest in the introduction of specific PGP organisms into the rhizosphere of maize [57], it becomes of particular importance to characterize the indigenous bacterial communities naturally associated with maize root systems, and to investigate the effects of individual rhizosphere isolates on plants. Such information is needed for the development of effective strategies for delivery and maintenance of bacterial inoculants in association with maize root systems, and for evaluating the impact produced by bacteria released into the environment. Several works on the composition of bacterial populations in the rhizosphere of maize pointed out the large numbers of diverse types of bacteria associated with maize root surfaces and the significant differences in size and composition of bacterial populations found among different soil types and during plant development [811]. These works highlighted the complexity of relationships between roots and their associated microbes within an individual grass root system and are indicative of the influence of the plant on its rhizosphere.

Among the microorganisms occurring in the rhizosphere of maize, Burkholderia cepacia represents probably one of the predominant bacterial species [12]. Previous studies revealed that B. cepacia is present in large numbers associated with the roots and the rhizosphere of maize, both in maize monoculture soils [6] and in a field with no previous cropping history of maize [13]. Furthermore, B. cepacia has been reported to compete, survive, and colonize roots of various maize cultivars [14,15], to enhance the yield of several crop plants [6,1619], and to antagonize and repress all the major soilborne fungal pathogens of maize, such as those belonging to the genus Fusarium[20,21].

However, a more detailed characterization of maize associated B. cepacia populations is needed in order to select and successfully inoculate PGPR strains of this species in the maize rhizosphere. In a previous work [13] we investigated the genetic diversity of a B. cepacia population naturally associated with maize roots during plant development in one growing season. We observed that the biodiversity of this rhizosphere bacterial population decreased over time, being higher in the first stages of plant growth, i.e. at the end of germination, than during the last stages of plant growth, when maturation is complete. This suggested that it may be more advantageous to select B. cepacia strains to be used as maize inoculants from the late plant growth stages, as bacterial genotypes better adapted to the host plant tend to establish during plant growth.

In the present work, a B. cepacia population naturally occurring in the rhizosphere of maize and isolated at the elongation stage of plant growth was investigated in order to: (i) analyze the metabolic and genetic profiles of bacterial isolates using Biolog automated analysis [22] and the random amplified polymorphic DNA (RAPD) method [23]; (ii) investigate traits involved in biocontrol activity, such as the ability to inhibit the in vitro growth of several Fusarium spp. phytopathogens of maize, and to synthesize Fe(III) chelator(s); (iii) analyze the increased growth response of maize plants due to seed bacterization with B. cepacia strains, in soil artificially infested with F. proliferatum ITEM-381 and F. moniliforme ITEM-504; (iv) evaluate the ability of isolated strains to promote maize growth.

2 Materials and methods

2.1 Bacterial and fungal strains

B. cepacia LMG 11351 (also named PHP7), isolated in France from the rhizosphere of maize, was kindly provided by T. Heulin, Centre Pédologie Biologique, CNRS, Nancy, France, and cryopreserved at −75°C in 30% (v/v) glycerol.

F. moniliforme ISPVF44, F. graminearum G1 ISPV218, F. graminearum G2 ISPV523, F. graminearum G2 ISPV784, F. graminearum G2 ISPV511, F. proliferatum ISPV447, F. subglutinans ISPV817, and F. solani ISPV291 were kindly provided by L. Corazza, Phytopathology Institute, MAF, Rome, Italy. F. solani EF5 was obtained from the Plant Pathology Institute, University of Naples, Italy. F. moniliforme (Sheldon) DSM 62264 was provided by the Collection of the Deutsche Sammlung für Mikroorganismen, Braunschweig, Germany. F. proliferatum (Matsushima) Nirenberg ITEM-381 and F. moniliforme (Sheldon) ITEM-504, isolated from preharvest maize ear rot and corn stalks respectively, were obtained from the Collection of the Istituto Tossine e Micotossine da Parassiti Vegetali, CNR, Bari, Italy. Mycelia and conidia from strains grown on carnation leaf agar [24] were freeze-stored in sterile 18% (w/v) glycerol (−75°C).

2.2 Isolation of B. cepacia strains

B. cepacia strains were isolated in May 1995 from the rhizosphere of Zea mays L. (cv. Pactol) grown in an experimental field with no previous cropping history of maize (free from the incidence of any fungal infection for the past 5 years) at Santa Maria di Galeria, Rome, Italy. The soil had a loamy sand texture and the following composition, as determined by standard soil analysis methods [25]: sand 47%, clay 39%, silt 14%, organic C 2.38% (w/w), organic matter 4.1%, total N 0.21%, available K 82.0 mg kg−1, available P 10.2 ppm, available Mg 37.5 mg kg−1; the pH was 6.04, and the moisture holding capacity was 0.28 ml g−1. After 40 days of plant growth, i.e. at the elongation growth stage, two plants were randomly harvested, roots were excised, and loosely adhering soil was removed. Afterwards, each root was weighed, blended, and then resuspended in phosphate buffered saline (PBS, Flow Laboratories). Serial dilutions of these suspensions were plated onto semiselective PCAT medium [26] to isolate microorganisms belonging to B. cepacia species. B. cepacia colonies on PCAT medium were white or pale yellow, smooth, with flat edges and an elevated center. After 72 h of bacterial growth at 27°C, 14 B. cepacia-like colonies were selected from the same dilution of root samples, i.e. 100-fold dilution containing 50–100 colonies. The isolates from PCAT plates were subjected to single-colony isolation on the same medium and cryopreserved at −75°C in 30% (v/v) glycerol.

2.3 ARDRA

An analysis of the restriction patterns of DNA encoding 16S rRNA (16S rDNA) amplified by means of PCR (ARDRA) was performed directly using a single whole colony for each of the 14 isolates, according to the procedure described by Di Cello et al. [13].

2.4 Metabolic fingerprinting

Substrate utilization analysis was performed with Biolog GN plates (Biolog, Hayward, CA, USA) for each of the 14 MCI isolates and the reference strain B. cepacia LMG 11351, isolated from the maize rhizosphere. Strains were grown overnight on Biolog Universal Growth Medium (BUGM) at 27°C, and characterized by the Biolog GN system. For each strain, cells were removed from the agar plate with a sterile swab, and released into a 20 mm diameter tube containing 18 ml of sterile saline solution. Using the Biolog turbidimeter, the cell suspension was adjusted to a transmittance level of about 56%, corresponding to a cell density of about 3×108 colony forming units (cfu) ml−1, as estimated by colony counts on nutrient agar (NA, Difco). Then, 150 μl of this suspension was added to each of the 96 wells in Biolog GN microtiter plates. Plates were inoculated in triplicate with each of the bacterial isolates and incubated at 27°C. The percent change in optical density versus control wells was recorded at 590 nm after 24 and 48 h of incubation using a Biolog microplate reader (Molecular Devices), with the Biolog MicroLogTM 3, Release 3.50 computer software.

The reaction data (positive, negative, and borderline responses) from the 24-h readings of the Biolog microplates were recorded for each strain and the metabolic profiles were compared. Results obtained were also compared with the substrate utilization profiles of the following reference strains belonging to the β-subclass of the proteobacteria, included in the Microlog GN data base files: B. cepacia (BUR.CEP) clinical category, ATCC 25416; B. gladioli (BUR.GLA) clinical category; B. pseudomallei (BUR.PMA) clinical category; B. caryophylli (BUR.CARY) environmental category; Ralstonia pickettii (BUR.PIK) clinical category; R. solanacearum A (BUR.SOL A) environmental category; R. solanacearum B (BUR.SOL B) environmental category.

2.5 RAPD fingerprinting

Amplification reactions were performed as previously described by Di Cello et al. [13]. Two 10-mer primers, AP12 (5′-CGGCCCCTGC-3′) and AP5 (5′-TCCCGCTGCG-3′), with GC contents of 90 and 80%, respectively, were used.

2.6 Screening for in vitro antibiosis

An in vitro assay was performed to test the ability of B. cepacia strains to suppress fungal phytopathogens on both potato dextrose agar (PDA, Difco), a C-, N- and Fe(III)-rich medium, and King's B (KB), an iron limiting medium [27]. Bacterial strains were streaked near the edge of a Petri dish (9 cm diameter) at fixed positions and incubated in the dark at 28°C for 48 h. An agar plug with mycelium (7-mm block removed with a No. 2 cork borer from the actively growing margins on PDA cultures) was then placed in the center of these preinoculated plates, which were reincubated in the dark at 28°C. The growth of fungal phytopathogens was measured at 48-h intervals and compared with that of fungal colonies emerging from plugs not challenged with bacterial isolates. The percent inhibition of the fungal growth due to the presence of the bacterium was calculated with the following formula: [(R1R2)/R1]×100, where R1 is the radial distance grown by fungus in plates without bacteria (control value), and R2 is the distance grown on a line between the inoculation positions of the fungal phytopathogen and the antagonist strain of B. cepacia (inhibition value) [28]. One bacterial strain was inoculated per agar plate, and three replicates of each bacterial isolate and pathogen combination were evaluated in each experiment. Each experiment was done twice, with similar results that were combined for statistical analysis.

2.7 Siderophore synthesis

Siderophore production was preliminarily screened by a plate assay using chrome azurol-S (CAS) agar plates [29]. Siderophore production was also tested in culture supernatants by means of specific colorimetric tests. Cultures of B. cepacia were grown with or without 100 μM FeCl3 for about 48 h at 37°C and 150 rpm (Clim-O-Shake, Adolf Kühner, Switzerland) either in Fe(III)-deprived casamino acids (DCAA) or in sodium succinate M9 (SM9) medium, as described by Visca et al. [30,31]. Hydroxamate-like compounds were detected according to Csàky [32], using hydroxylamine hydrochloride as the standard. Catechol-type compounds were measured using the test developed by Arnow [33], with 2,3-dihydroxybenzoic acid as the standard. Siderophores (pyochelin, cepabactin, salicylic acid or related compounds) were extracted with ethyl acetate or chloroform, following the procedure reported by Bevivino et al. [34]. The dry residues obtained were dissolved in a small volume of methanol and applied to a silica gel G thin layer chromatography plate (TLC), using chloroform/acetic acid/ethanol (90:5:2.5, v/v) as the development solvent. Siderophores were characterized by chromatographic mobility (RF), fluorescence emission under UV light, and chemical reactivity when sprayed with 0.1 M FeCl3 in 0.1 M HCl or with the ammoniacal silver nitrate reagent for thiazolidine rings [30,31].

2.8 Fungal inoculum preparation

Twenty colonized agar plugs (7 mm diameter) of F. proliferatum ITEM-381 and F. moniliforme ITEM-504 were removed with a cork borer from the actively growing margins of PDA cultures and transferred individually to autoclaved cornmeal-sand mixtures (270 g of riverbed sand, 30 g of cornmeal, 60 ml of distilled water) in 1-l Erlenmeyer flasks. After 2 weeks of incubation at 25°C, the contents of each flask was hand-mixed with 5 kg of non-autoclaved sand-soil mixture (1:4) and placed in 30×24×10 cm plastic boxes. Sand contained only a trace of N and no organic C; the pH was 6.0. The soil had been collected from the upper 30 cm of the same field from which B. cepacia strains were isolated (see above) and mixed thoroughly with sand. After introduction of fungal inoculum, the boxes were watered and transferred in a greenhouse at 29°C during the day and 18°C at night, with a 16-h light period (30 000 lux) followed by an 8-h dark period. One week later, 1-g samples of each infested sand-soil mixture were evaluated for propagule numbers of either F. proliferatum or F. moniliforme by conducting a dilution series; 1-ml aliquots were pipetted into plastic Petri dishes, and 20 ml of molten modified Nash and Snyder's medium (PCNB) [35] was added to each dish. Each Petri dish was swirled to distribute the soil suspension evenly in the medium. The inoculum concentrations of fungal pathogens before planting were approximately 3×104–4×104 and 1×104–2×104 cfu g−1 of soil for F. proliferatum and F. moniliforme, respectively. An autoclaved cornmeal-sand mixture without fungus was added to the sand-soil mixture for the control treatment (non-infested soil).

2.9 Seed bacterization

Cultures of B. cepacia were grown in 100 ml of nutrient broth (NB, Difco) for about 24 h at 28°C and 150 rpm, washed in sterile saline solution, and resuspended in 50 ml of 1% methyl cellulose (MC) (∼1010 cfu ml−1). Fifty Zea mays L. seeds (cv. Pactol) were submerged in 50 ml of bacterial inoculum in 200-ml Erlenmeyer flasks. Control seeds were submerged in 50 ml of 1% MC. Flasks were incubated at 25°C on a rotary shaker at 70 rpm for 2 h, to allow bacterial cells to adhere to seeds. After incubation, excess inoculum was removed and seeds were immediately planted in the potting medium. Bacteria were applied to the seed at approximately 108 cfu seed−1, as determined by plate counts on PCAT.

2.10 Greenhouse screening of rhizosphere B. cepacia isolates for biocontrol activity

The effect of F. proliferatum ITEM-381 and F. moniliforme ITEM-504 on maize growth was preliminarily investigated under greenhouse conditions, in five experimental trials. Treatments were as follows: (i) control seeds planted in non-infested soil; (ii) seeds planted in soil artificially infested with F. proliferatum ITEM-381; (iii) seeds planted in soil artificially infested with F. moniliforme ITEM-504.

B. cepacia strains were then tested in a greenhouse screening. Treatments were as follows: (i) seeds planted in soil artificially infested with F. proliferatum ITEM-381; (ii) seeds treated separately with each of the 14 B. cepacia isolates, and planted in soil artificially infested with F. proliferatum ITEM-381; (iii) seeds planted in soil artificially infested with F. moniliforme ITEM-504; (iv) seeds treated separately with each of the 14 B. cepacia isolates, and planted in soil artificially infested with F. moniliforme ITEM-504.

The above experiments were carried out in a greenhouse at 29°C during the day and 18°C at night, under the light/dark period previously described. Maize seeds were planted in 30×24×10 cm plastic boxes at a depth of approximately 2.5 cm and water was added to each box on alternate days. After 21 days of plant growth, 24 plants for each treatment were collected from two replicates boxes, and fresh weight of root and shoot was calculated.

2.11 Greenhouse screening of isolates for maize growth promoting activity

In the first screening, a non-sterile sand-peat/manure (Letamplus KB, Rhône Poulenc) mixture (4:1) was used. The sand characteristics have been reported above. Letamplus KB composition was as follows: organic matter 25%, total N 5%, and organic C 16%. Treatments were as follows: (i) control seeds; (ii) seeds treated separately with each of the 14 B. cepacia isolates. The plants were grown in 30-l pots with a diameter of 27 cm (1 seed per pot), and watered three times a week with a mineral solution containing: 1 g KNO3, 0.49 g MgSO4·7H2O, 3 g K2HPO4, 2 g KH2PO4, 0.007 g FeCl3-EDTA, 0.15 g CaCl2·2H2O, 0.01 g MnSO4·H2O, 1 g NaCl, 0.002 g ZnSO4, 0.003 g H3BO3, 0.025 mg CuSO4·5H2O, 0.25 mg Na2MoO4·2H2O in 1 l of distilled water. Each treatment was replicated 24 times. Plants were collected 55 days after sowing, and root and shoot fresh weight were measured.

Bacterial strains were then tested in the second greenhouse experiment at 21 days after sowing, using as potting medium a non-sterile sand-soil mixture (1:4) (see above). Greenhouse treatments were as follows: (i) control seeds; (ii) seeds treated separately with each of the 14 B. cepacia isolates. Maize seeds were planted in 30×24×10 cm plastic boxes (two replicate boxes for each treatment) at a depth of approximately 2.5 cm. Tap water was added to each box three times a week. After 21 days of inoculation, 24 maize plants for each treatment were collected, and root and shoot fresh weight were measured.

In the third trial, the plants were grown in 30-l pots with a diameter of 27 cm (1 seed per pot), containing the same sand-soil mixture used in the second greenhouse experiment. Greenhouse treatments were as follows: (i) control seeds; (ii) seeds treated separately with each of the 14 B. cepacia isolates. Each treatment was replicated 24 times. Pots were watered three times a week with tap water. Maize plants were collected 55 days after sowing, and root and shoot fresh weight were measured.

Both pots and boxes were arranged in a completely randomized design in a greenhouse at 29°C during the day and 18°C at night, under the light/dark period previously described.

2.12 Statistical analysis

The metabolic profiles of the 14 MCI isolates and the reference strains belonging to the β-subclass of the proteobacteria were compared. The Euclidean distance was calculated between all possible combinations of samples taken in pairs. These distances were hierarchically clustered to produce a tree by using the programs FITCH (for applying the Fitch-Margoliash method), NEIGHBOR (for applying the neighbor-joining method), and CONSENSE (for computing consensus trees) of the PHYLIP 3.57 c software package [36]. A total of 1003 bootstrap replications were conducted to determine the statistical significance of the obtained branches. The TREEVIEW program for displaying and printing the obtained trees was used [37].

Data from the in vitro antagonism were Logit transformed prior to statistical analysis. The results obtained were analyzed using one-way ANOVA (StatView 512+, BrainPower, CA, USA). Greenhouse experiments were repeated at least twice, unless otherwise specified. Variances between experimental trials were homogeneous, and, thus, data from repeated experimental trials were pooled and analyzed using one-way ANOVA.

3 Results

3.1 ARDRA

Fourteen B. cepacia-like colonies were isolated from the maize rhizosphere on PCAT medium, which provided a degree of selectivity for B. cepacia of more than 70%[13]. The isolates were designated MCI followed by progressive numbers of isolation, and they were all assigned to the B. cepacia species by restriction analysis of amplified 16S rDNA with the enzyme AluI (data not shown). ARDRA patterns were the same as that obtained with the 16S DNA from reference strain B. cepacia LMG 11351, isolated from the maize rhizosphere.

3.2 Biolog GN profiles

In order to characterize the physiological profiles of the 14 bacterial isolates, their ability to assimilate the 95 different carbon sources of the Biolog GN system (including carbohydrates, organic acids, amino acids, polymers, esters, alcohols, amides, amines, aromatics, and phosphorylated and brominated compounds) was investigated. All bacterial isolates were identified as B. cepacia with a similarity index above 75%. The Biolog analysis showed that all rhizosphere B. cepacia isolates, including the reference strain B. cepacia LMG 11351, were able to utilize 79% of the carbon sources present in the GN microplates, confirming the ability of B. cepacia to multiply using a wide range of organic compounds; furthermore, the profiles of all rhizosphere B. cepacia isolates differed for the ability to utilize several substrates, such as cellobiose, gentiobiose, α-d-lactose, lactulose, β-methyl d-glucoside, turanose, acetic acid, α-ketoglutaric acid, and propionic acid. To investigate the relationships between the metabolic profiles of these 14 strains and the reference strains belonging to the β-subclass of the proteobacteria, we analyzed the Biolog data by the Fitch-Margoliash method. The unrooted tree obtained (Fig. 1) revealed that all rhizosphere B. cepacia strains were grouped together in a principal cluster, showing a high degree of relatedness to the reference strain B. cepacia LMG 11351. In addition, these rhizosphere strains were not closely related to B. cepacia ATCC 25416 of clinical origin and other Burkholderia spp. and Ralstonia spp., which were assigned originally to the pseudomonad rRNA similarity group II and are now affiliated with the β-subclass of the proteobacteria [3840]. The strict consensus tree obtained by using the neighbor-joining method (not shown) indicated that the clusters of strains were basically the same when compared with those obtained by using the Fitch-Margoliash method.

1

Unrooted tree based on the examination of Biolog GN fingerprints showing relationships between B. cepacia MCI isolates and representatives of the β-subclass of the proteobacteria. Clustering was performed using the Fitch-Margoliash method and displayed with the TREEVIEW program.

3.3 RAPD analysis

To evaluate the genetic profiles of bacterial isolates, the DNAs of lysed cell suspensions of the 14 B. cepacia strains were amplified by the RAPD technique with two 10-mer primers, AP5 and AP12. The reproducibility of the results was verified by independent experiments (data not shown). Analysis of the amplification patterns allowed us to recognize two main groups, regardless of the primers used: the first group including the B. cepacia isolates MCI 2, MCI 11, MCI 14, MCI 15, and MCI 23; the second group including the B. cepacia isolates MCI 4, MCI 12, MCI 13, MCI 16, MCI 18, MCI 19, MCI 22, MCI 26, and MCI 27. According to Di Cello et al. [13], the amplification patterns obtained with primer AP5 exhibited a higher level of polymorphism than did those obtained with primer AP12. Amplification of genomic DNAs of B. cepacia strains with primer AP5 gave rise to nine bands, whose dimensions ranged from 280 to 1300 bp. Amplification of genomic DNAs of B. cepacia strains with primer AP12 gave rise to five different bands, whose dimensions ranged from 350 to 2100 bp. As an example, Fig. 2 shows the amplification patterns obtained with primer AP5.

2

Electrophoretic patterns obtained by RAPD analysis of B. cepacia strains with primer AP5. Lanes 1–8, B. cepacia MCI 2, MCI 4, MCI 27, MCI 11, MCI 12, MCI 13, MCI 14, and MCI 15, respectively. Lane m, 123-bp molecular size marker ladder. Lanes 9–14, B. cepacia MCI 16, MCI 18, MCI 19, MCI 22, MCI 23, and MCI 26, respectively.

3.4 In vitro antagonism

The antagonistic activity of B. cepacia isolates against maize pathogenic strains of Fusarium was tested in a plate assay using KB and PDA media. It is generally assumed that the inhibition of fungal pathogens on PDA, a C-, N- and Fe(III)-rich medium, may be due to the production of antibiotic-like substances, and on KB, an iron limiting medium, the inhibition may occur by production of siderophores in addition to antibiotics. The results obtained indicated that: (a) all B. cepacia strains were able to restrict the growth of the Fusarium isolates on both KB and PDA, although to differing degrees; (b) there were no significant differences in the level of antagonism between bacterial strains (P>0.05); (c) the inhibition was more evident on KB with a percentage of inhibition of 64% whereas on PDA the effect was significantly reduced (49%) (P<0.001); (d) the fungi tested showed an increasing sensitivity to B. cepacia strains as follows: on KB (P<0.001), F. solani spp. (43% inhibition), F. moniliforme spp. (58%), F. proliferatum spp. (66%), F. subglutinans ISPV 517 (74%) and F. graminearum spp. (74%); whereas on PDA (P<0.05) the order was: F. subglutinans ISPV 517 (43%), F. graminearum spp. (48%), F. moniliforme spp. (49%), F. proliferatum spp. (51%) and F. solani spp. (51%).

3.5 Synthesis of siderophores

Since B. cepacia strains displayed generally higher inhibition of fungal growth in Fe(III)-poor medium, we reasoned that siderophores could be involved in the antagonistic response. Therefore, we investigated the existence of Fe(III) transport systems in our B. cepacia isolates. We found that all MCI strains appeared to synthesize Fe(III) chelator(s) as shown by the formation of a large yellow-gold halo around colonies growing on CAS agar plates [29]. When grown in DCAA or SM9 media, all B. cepacia strains produced a Fe(III)-repressible, hydroxamate-like molecule, the level of which ranged from 437 to 535 μM, depending on the isolate and the growth medium. As expected for a siderophore-like compound, the addition of 100 μM FeCl3 to the Fe(III)-poor media strongly repressed hydroxamate synthesis. TLC analysis did not allow identification of this hydroxamate-positive compound as cepabactin [41]. Under iron-deficient conditions, none of the 14 MCI strains produced catechol-type siderophores.

TLC analysis also revealed that under Fe(III)-depleted conditions all B. cepacia strains produced a compound which reacted positively with the ammoniacal silver nitrate reagent for thiazolidine rings, which are known to form black silver mercaptide salts during alkaline hydrolysis in the presence of Ag(I). The compound did not exhibit gray-green fluorescence emission under UV light and did not turn brown-red when sprayed with acidic FeCl3. Because of its peculiar chromatographic characteristics (RF=0.32 in chloroform/acetic acid/ethanol) and chemical reactivity we cannot relate this compound to any known B. cepacia siderophore (i.e. pyochelin [42] or salicylic acid [43]).

3.6 Biological control of Fusarium spp. by B. cepacia isolates

Preliminary experiments were conducted under greenhouse conditions to test the effect of two Fusarium strains, F. proliferatum ITEM-381 and F. moniliforme ITEM-504, on root and shoot fresh weight of maize, after 21 days of plant growth. The introduction of F. proliferatum ITEM-381 into the potting medium caused a significant reduction (P<0.001) in both root (0.97±0.71 g) and shoot (1.49±0.98 g) fresh weight values as compared to root (1.31±0.60 g) and shoot (1.96±1.01 g) fresh weight values of control plants growing in non-infested soil. The inoculation with F. moniliforme ITEM-504 caused a significant reduction (P<0.05) in root (1.00±0.67 g) fresh weight value, as compared to root (1.17±0.62 g) fresh weight value of control plants, whereas it did not cause any significant effect on shoot fresh weight (P>0.05).

Greenhouse experiments were then performed to test the efficacy of the bacterial isolates as biocontrol agents of both F. proliferatum ITEM-381 and F. moniliforme ITEM-504. In soil infested with F. proliferatum ITEM-381, treatment of maize seeds with B. cepacia MCI 12, MCI 15, and MCI 27 resulted in a significant increase in both root and shoot fresh weight, seed bacterization with MCI 2 and MCI 16 significantly increased the root fresh weight only, and seed treatment with MCI 26 resulted in significantly higher values of shoot fresh weight only (Table 1a). In soil infested with F. moniliforme ITEM-504, B. cepacia isolates MCI 23, MCI 26, and MCI 27 exerted a positive effect on both root and shoot fresh weight, B. cepacia isolates MCI 14, MCI 18, and MCI 19 on root fresh weight only, whereas B. cepacia isolates MCI 13, and MCI 16 exerted a positive effect on shoot fresh weight only (Table 1b). Seed bacterization with B. cepacia MCI 12 resulted in significantly lower fresh weight value of plant roots, as compared to the untreated control. The significance level of treatments ranged from 5% to 0.1%, as indicated in Table 1.

View this table:
1

B. cepacia strains showing a positive effect on maize plants grown in soil infested with F. proliferatum ITEM-381 (a) or F. moniliforme ITEM-504 (b)

3.7 Plant growth promotion of B. cepacia isolates

In these experiments we tested the plant growth promoting activity of all 14 B. cepacia isolates, separately, in different potting media. In a non-sterile sand-peat/manure mixture, bacterization of maize seeds with each B. cepacia isolate generally resulted in significant increases of root and shoot fresh weight in comparison with the uninoculated control, with a significance level ranging from 5% to 0.1% (Table 2a). Only one isolate, B. cepacia MCI 16, did not have a significant influence on maize plant growth. When a non-sterile sand-soil mixture was used as potting medium, only four strains brought about a significant increase in root and/or shoot fresh weight at 21 days of plant growth (Table 2b). The isolates MCI 19 and MCI 22 came out as the strains with the greatest effect on both root (P<0.001) and shoot (P<0.01) fresh weight, B. cepacia isolate MCI 2 exerted a positive effect on root fresh weight (P<0.05) only, whereas B. cepacia isolate MCI 13 exerted a positive effect on shoot fresh weight (P<0.05) only. After 55 days of plant growth, seed bacterization with B. cepacia isolates did not have a significant influence on maize plants grown in the same non-sterile sand-soil mixture previously used (P>0.05) (Table 2b).

View this table:
2

Effect of B. cepacia strains on maize plants grown in a sand-peat/manure (a) and in sand-soil (b) mixtures

4 Discussion

A study of a population of B. cepacia naturally associated with maize roots was carried out, which included metabolic and genetic profiling, and investigation of some of the traits associated with the biocontrol and plant growth promoting activity. The present study indicates that maize-rhizosphere isolates of B. cepacia can act as PGP microorganisms. Of particular interest is the finding that 43% and 57% of B. cepacia isolates tested have a potential as maize inoculants for the biological control of pathogenic strains of F. proliferatum ITEM-381 and F. moniliforme ITEM-504, respectively.

In order to characterize the metabolic and genetic profiles of the B. cepacia isolates and to obtain information on the relatedness between the rhizosphere isolates, Biolog and RAPD analyses were carried out. The Biolog analysis, a commercial system for identifying bacterial isolates by means of sole carbon source utilization [22,44], confirmed the ability of B. cepacia to multiply using a wide range of organic compounds and indicated that rhizosphere isolates differed in some biochemical traits. The unrooted tree representing the phenotypic relatedness between the rhizosphere B. cepacia isolates, the type strain B. cepacia ATCC 25416 of clinical origin and B. cepacia LMG 11351 isolated from the maize rhizosphere, and the reference strains of closely related species belonging to the β-subclass of the proteobacteria revealed that all rhizosphere isolates were grouped together in a principal cluster, which included the reference strain B. cepacia LMG 11351, and were clearly distinct from the other reference strains examined (i.e. B. cepacia of clinical origin and other Burkholderia spp. and Ralstonia spp.). The RAPD analysis, a methodology which has been successfully used to evaluate the genetic polymorphism of strains [13,23], indicated that the rhizosphere B. cepacia strains displayed two principal RAPD patterns, regardless of the primer used for amplification, revealing a low level of polymorphism within these strains. Moreover, B. cepacia strains which were grouped very closely in the unrooted tree obtained from the Biolog analysis showed the same RAPD pattern.

A major objective of this study was to investigate some physiological traits related to biocontrol and PGP activity in order to examine the potential use of the B. cepacia isolates as field inoculants. We preliminarily investigated the ability of the B. cepacia isolates to inhibit the in vitro growth of 12 fungal isolates belonging to the Fusarium section Liseola, which are the most common fungal associates of maize plants causing diseases of seedlings, roots, stalks, and kernels. The results obtained showed that all 14 B. cepacia isolates were able to restrict the in vitro growth of the Fusarium strains, exerting higher inhibition of fungal growth on KB, an iron limiting medium, than on PDA, a C-, N- and Fe(III)-rich medium. It may therefore be hypothesized that Fe(III) deficiency might enhance the synthesis of anti-fungal compounds or, alternatively, that siderophores might contribute to the in vitro inhibition of fungal pathogens. Siderophores, low-molecular-mass chelators, are produced by many microorganisms as a means of sequestering limiting iron. We found that all B. cepacia isolates displayed a strong chelating reactivity in the Schwyn and Neilands [29] assay and released, under low iron conditions, hydroxamate-like and thiazo-like compounds. TLC analysis did not allow identification of these compounds as cepabactin [41], pyochelin [42], or salicylic acid [43], which have previously been described for B. cepacia.

It is generally recognized that expression of antagonism by a microorganism towards a pathogen in culture media cannot be regarded an evidence that the microorganism will have a functional role in controlling the pathogen in the field [45]. This is not unexpected as the nutritional environment and many other factors that affect growth and survival of biocontrol agents in nature are considerably different from those in nutrient-rich culture media. In view of their potential application as biocontrol agents, seed inoculation studies with the maize-rhizosphere B. cepacia isolates were carried out to assess in a rhizosphere environment their antagonism against two pathogenic fungi of maize, F. proliferatum and F. moniliforme. Seed treatment is an attractive method for introducing biological control agents into the soil-plant environment. Application of bacteria to seeds has been utilized for the biological control of soilborne plant pathogens affecting many hosts plants [4]. In particular, B. cepacia has been used for the biological control of foliar diseases [46], soilborne diseases [21,47], and storage fruit rots [48]. The results obtained in the present study indicated a very poor correlation between antagonism recorded in vitro and the effectiveness of B. cepacia isolates to reduce the infection due to Fusarium spp. in the greenhouse, indicating that eight and six MCI strains that displayed in vitro antibiosis against F. proliferatum ITEM-381 and F. moniliforme ITEM-504, respectively, were not effective in inhibiting them in greenhouse trials. These findings suggest that multiple mechanisms are likely to account for the biological control.

Finally, the ability of our bacterial isolates to promote the growth of maize plants was investigated in different potting media. Direct growth promotion may be considered an indirect mechanism of biological control [49]. Indeed, promotion of early emergence and increased early root development may serve to allow the plant to escape disease. Kloepper et al. [1] reported that the bacterial effects on plant growth may be influenced by multiple environmental variables such as soil type, nutrition, moisture and temperature. Our results performed under non-sterile greenhouse conditions indicated that: (i) bacterization with B. cepacia isolates never resulted in negative effects on plant development; (ii) the substrate composition influenced the PGP activity of the B. cepacia isolates; in fact, in a sand-peat/manure mixture almost all B. cepacia strains exerted a positive effect on maize growth, whereas in a sand-soil mixture the positive effect of seed bacterization with B. cepacia was more evident on shoots and roots of young plants than of mature plants. A possible explanation for this latter finding is that in a sand-soil mixture bacteria colonize the rhizosphere of maize rapidly, exerting their positive effect on maize, but as the plant grows the inoculant bacteria lose their competitiveness with the endogenous rhizosphere microorganisms.

The exact mechanisms by which our B. cepacia isolates promote maize growth remain unclear. Microbial siderophores may be partly responsible for biocontrol and enhanced plant growth by competitively inhibiting the growth of plant pathogens or other harmful rhizosphere microorganisms with less efficient Fe uptake systems or by increasing the availability of Fe in the soil surrounding the root [50]. Additional research is needed to investigate the factors influencing the rhizosphere competitiveness of these strains and to verify the mechanisms of plant growth stimulation.

In conclusion, several maize-rhizosphere isolates of B. cepacia have a potential as maize inoculants for the biological control of Fusarium spp. Currently, research is continuing to develop these bacteria as biocontrol agents, with further investigations of their interactions with Fusarium spp. The acquired knowledge can be used to improve their consistency and level of effectiveness. Finally, a more complete comprehension of the dynamics of bacterial rhizosphere populations is an important factor for a more effective management of biocontrol microbial agents.

Acknowledgements

We thank Dr. M. Broglia for the collection of maize plants, Dr. A. Logrieco for providing fungal strains, and Dr. Secco of Novartis Seeds for providing maize seeds (cv. Pactol). We are grateful to C. Nacamulli, G. Seri, and A. Angeli for valuable assistance in carrying out the greenhouse experiments. The contribution of Dr. P. Visca and L. Ferrandi was also greatly appreciated. The authors also wish to thank Dr. Fiona Leckie for the revision of the manuscript. Part of this work was presented at the Fifth IOBC/EFPP Workshop on Molecular Approaches in Biological Control, held in Delemont, Switzerland, 15–18 September 1997. This work was partially supported by a contribution of MIRAAF – Piano Nazionale Biotecnologie Vegetali, D.M. 125/7240/96.

References

  1. [1]
  2. [2]
  3. [3]
  4. [4]
  5. [5]
  6. [6]
  7. [7]
  8. [8]
  9. [9]
  10. [10]
  11. [11]
  12. [12]
  13. [13]
  14. [14]
  15. [15]
  16. [16]
  17. [17]
  18. [18]
  19. [19]
  20. [20]
  21. [21]
  22. [22]
  23. [23]
  24. [24]
  25. [25]
  26. [26]
  27. [27]
  28. [28]
  29. [29]
  30. [30]
  31. [31]
  32. [32]
  33. [33]
  34. [34]
  35. [35]
  36. [36]
  37. [37]
  38. [38]
  39. [39]
  40. [40]
  41. [41]
  42. [42]
  43. [43]
  44. [44]
  45. [45]
  46. [46]
  47. [47]
  48. [48]
  49. [49]
  50. [50]
View Abstract