Paola Lavermicocca, Istituto Tossine e Micotossine da Parassiti Vegetali, C.N.R., V. le L. Einaudi 51, I-70125, Bari, Italy (e-mail: email@example.com).
Pseudomonas syringae pv. ciccaronei strain NCPPB2355 was found to produce a bacteriocin inhibitory against strains of Ps. syringae subsp. savastanoi, the causal agent of olive knot disease. Treatments with mitomycin C did not substantially increase the bacteriocin titre in culture. The purification of the bacteriocin obtained by ammonium sulphate precipitation of culture supernatant fluid, membrane ultrafiltration, gel filtration and preparative PAGE, led to the isolation of a high molecular weight proteinaceous substance. The bacteriocin analysed by SDS-PAGE revealed three protein bands with molecular weights of 76, 63 and 45 kDa, respectively. The bacteriocin was sensitive to heat and proteolytic enzymes, was resistant to non-polar organic solvents and was active between pH 5·0–7·0. Plasmid-DNA analysis of Ps. syringae ciccaronei revealed the presence of 18 plasmids; bacteriocin-negative variants could not be obtained by cure experiments.
Strains of many bacterial species can produce proteinaceous antimicrobial substances called bacteriocins. Bactericidal specificity, restricted to species closely related to their producer, and chemical composition distinguish bacteriocins from other classic antibiotics ( Tagg et al. 1976 ; Vidaver 1983; Gross & Vidaver 1990).
In previous investigations ( Lavermicocca et al. 1996 ), it was reported that Ps. syringae pv. ciccaronei, isolated from a leaf spot of the carob tree ( Ercolani & Caldarola 1972), was able to inhibit specifically the growth of several strains of Ps. syringae subsp. savastanoi, the causal agent of olive knot disease. This pathogen induces overgrowths and/or cankers on several hosts (olive, oleander, privet, etc.) and causes severe damage in most olive-growing regions. Lavermicocca et al. (1997) observed that treatment of the inoculation sites with a cell-free culture of Ps. syringae ciccaronei reduced symptom expression in olive plants inoculated with Ps. syringae subsp. savastanoi. Inhibition of the multiplication of the pathogen was correlated with the presence of a bacteriocin in the cell-free culture of Ps. syringae ciccaronei.
In this paper, purification, physical and chemical properties and spectrum of activity of the antimicrobial substance produced by Ps. syringae ciccaronei strain NCPPB2355 are described.
Materials and methods
Bacterial strains and culture conditions
Pseudomonas syringae pv. ciccaronei NCPPB2355 (National Collection of Plant Pathogenic Bacteria, Harpenden, UK) was maintained on nutrient agar containing 2% v/v glycerine (NGA) at 4 °C and was grown on King’s B medium (KB) ( King et al. 1954 ). The bacterial strains used in this study are listed in Table 1.
Table 1. Antagonistic activity of Pseudomonas syringae pv. ciccaronei NCPPB2355 and antimicrobial activity of the culture filtrate of the bacterium
Strains of Ps. syringae subsp. savastanoi belong to the Collection of Istituto Tossine e Micotossine da Parassiti Vegetali (ITM), Bari, Italy or to the Collection of Dipartimento di Patologia Vegetale (PVBa), University of Bari, Italy; NCPPB: National Collection of Plant Pathogenic Bacteria, Harpenden, UK; ICMP: International Collection of Microorganism from Plants, Auckland, New Zealand; strains of Ps. syringae syringae are: NCPPB 1053, 1242, 2106, 2326, 2786, 2813; strains B301, B359, B427, B452, were kindly supplied by J.E. DeVay, University of California, Davis, CA, USA.
Values represent the mean difference between the diameter (mm) of the inhibition zone and the diameter of the colony. When more than one strain is reported, values represent the smallest and the largest diameter registered.
‡Arbitrary units in culture filtrate needed to inhibit the growth of the test strains. When more strains are reported, values represent the minimum and maximum units ml −1 recorded.
Other bacteria used were: Ps. syringae ciccaronei NCPPB2355, Ps. fluorescens NCPPB1795, Escherichia coli K12 ITM103, Bacillus megaterium ITM100, B. subtilis ITM101, Rhodococcus fascians NCPPB2551, Streptococcus lactis NCPPB2939 and Listeria innocua ITM99.
NI, growth of the test strain was inhibited neither by Ps. syringae ciccaronei colony nor by 3200 AU ml−1. The values are the mean of three replicates.
The following media were compared for bacteriocin production: NBY ( Vidaver 1967); KB; Luria-Bertani (LB) medium ( Carlton & Brown 1981); Woolley (W) ( Woolley et al. 1955 ); Woolley-Peptone (WP), i.e. Woolley’s medium supplemented with Bacto-Peptone (Difco), 15 g l−1; potato-dextrose broth–casamino acids (PDB–CA), i.e. PDB (Difco) supplemented with Bacto Casamino-acids (Difco) 4 g l−1; and IMM ( Surico et al. 1988 ). Pseudomonas syringae ciccaronei bacterial suspension (200 μl; A600 nm, 0·3) was added to 250 ml Erlenmeyer flasks containing 100 ml media. The cultures were shaken (100 rev min−1) at 26 °C. Every 24 h, for a period of 7 d, aliquots (1 ml) were taken from the cultures, centrifuged (9000 g, 10 min, 4 °C) and filter-sterilized (0·45 μm cellulose acetate, Millipore Corp., Bedford, MA, USA).
In order to quantify bacteriocin production in the culture filtrates, an agar-spot assay was used. The bacteriocin titres were determined by a quantitative serial dilution test. Twofold serial dilutions of Ps. syringae ciccaronei culture filtrate (10 μl) were spotted onto the surface of KB agar plates seeded with the indicator strain Ps. syringae subsp. savastanoi PVBa204 (Collection of Dipartimento di Patologia Vegetale, University of Bari, Italy). After 24 h incubation at 26 °C, inhibition of growth of the indicator strain was expressed in arbitrary units (AU) of activity, i.e. the amount of bacteriocin in an end-point dilution which completely inhibited the growth of the test strain in the area of application of a 10 μl droplet.
Mitomycin C induction.
Pseudomonas syringae ciccaronei was grown in WP and NBY. Cultures were induced for bacteriocin production with mitomycin C (final concentration 0·1 μg ml−1) when the cell concentration reached about 3 × 108 cells ml−1. After induction, cultures were shaken at 26 °C for 4 h and then stored at 4 °C overnight. Cell debris was removed by centrifugation (9000 g, 10 min, 4 °C) and the supernatant fluids filter-sterilized and assayed for antimicrobial activity.
Spectrum of antimicrobial activity
To detect the antagonistic activity of Ps. syringae ciccaronei, an agar-spot deferred method assay was used ( Vidaver et al. 1972 ). A bacterial suspension (5 μl; A600 nm, 0·3) of the strain was spot-seeded on the surface of KB agar and plates incubated at 26 °C for 3 d. The colonies were killed with chloroform vapours and the plates overlaid with 3 ml soft agar (agar 0·7% w/v) seeded with 100 μl of a bacterial suspension (A600 nm, 0·15) of the strains listed in Table 1. After 24 h incubation at 26 °C, inhibition was scored positive if the width of the clear zone around Ps. syringae ciccaronei colonies was 1·0 mm or larger. Assays were performed twice in triplicate.
To exclude bacteriophage activity, agar plugs were picked out from the area of inhibition and tested for the presence of bacteriophages by standard procedures ( Pugsley & Oudega 1987).
To evaluate the antimicrobial activity of Ps. syringae ciccaronei culture filtrate, the bacterium was grown in WP medium for 4 d. Aliquots (10 μl) of twofold serial dilutions of the culture filtrate were spotted onto KB agar plates and tested in the agar-spot assay against the strains listed in Table 1.
Production and purification of bacteriocin
Pseudomonas syringae ciccaronei strain NCPPB2355 was grown in l litre Erlenmeyer flasks containing 400 ml WP. Bacterial suspension (A600 nm, 0·3; 800 μl per flask) was used as inoculum. After 4 d incubation in shaken culture (100 rev min−1, 26 °C), the cells were removed by centrifugation (9000 g, 10 min, 4 °C) and the supernatant fluid filter-sterilized. The antimicrobial activity of the preparation was determined against the indicator strain Ps. syringae subsp. savastanoi PVBa204, using the agar-spot assay. The culture filtrate (1800 ml) was precipitated with ammonium sulphate (66% w/v saturation) and stored overnight at 4 °C. The precipitate and the surface pellicles were removed by centrifugation (12 000 g, 20 min, 4 °C), resuspended in Tris-phosphate buffer (0·01 mol l−1, pH 6·5; 1/10 of the initial volume) and dialysed against the same buffer for 48 h at 4 °C with several changes (dialysis tube, porosity 24, cut-off 17 kDa; Union Carbide Corporation, Danbury, CT, USA). The dialysed precipitate, designated as crude bacteriocin, was sequentially ultra-filtered through a series of Spectra/Por membranes (Spectrum Medical Industries, Laguna Hills, CA, USA) of decreasing pore sizes (300, 100, 50 and 20 kDa). The volumes and the antimicrobial activities of the retentates and filtrates were recorded. The retentate recovered from the 100 kDa ultrafiltration was lyophilized, and the residue (1·20 g) was dissolved in 6 ml ultrapure Milli-Q water and applied to a Sephadex G-150 fine column (Pharmacia, Uppsala, Sweden; 4·5 × 40 cm; flow rate 2·5 ml min−1). The column fractions (7 ml each) were collected in homogeneous groups according to the u.v. diagram obtained by monitoring at 267 nm, the maximum for absorption of ultrafiltered crude bacteriocin. Fractions were lyophilized and tested for protein content and antimicrobial activity against the indicator strain.
Analytical gel electrophoresis.
Non-denaturing discontinuous PAGE of the most active fraction recovered from gel filtration (fraction IV) was performed according to the method of Gross & DeVay (1977) with minor modifications, in a 7·5% (w/v) discontinuous gel with a Mini-Protean II electrophoresis system (BioRad Laboratories, Hercules, CA, USA). Bovine serum albumin (BSA; Sigma) was used as a protein marker. Samples were run at a constant voltage of 200 V 1 mm−1 gel for 1 h. After electrophoresis, the gel was cut in half, leaving identical marker and sample lanes on each half. One half was stained with Coomassie brilliant blue R-250 (Sigma) and the other was rinsed in deionized water for 20 min with frequent changes, laid on KB agar plates and overlaid with soft KB-agar (0·7% w/v) seeded with the indicator strain ( Bhunia et al. 1987 ). The plate was incubated overnight at 26 °C. Following growth of the indicator strain, the location of inhibiton halo(s) on the unstained gel was correlated with the protein band(s) on the stained gel.
Gel preparation and electrophoresis conditions were the same as those described for analytical PAGE. A 5 mg sample of fraction IV was resuspended in 200 μl distilled water and applied to the gel. After electrophoresis, the stained band with the Rf corresponding to the location of the inhibition halo in the unstained lane was cut out of the gel using a razor blade. The protein band was recovered with Micropure 0·22 μm separator located in a Microcon-30 centrifugal microconcentrator (Amicon, Beverly, MA, USA) according to the manufacturer’s instructions.
The protein recovered from the preparative gel was analysed by SDS-PAGE performed as described by Laemmli (1970). Protein standards and their molecular weights included the following: thyroglobulin 330 kDa; ferritin (half unit) 220 kDa; albumin 67 kDa; catalase 60 kDa; lactate dehydrogenase 36 kDa; ferritin 18·5 kDa (Pharmacia).
Protein estimation and antimicrobial activity
The concentration of protein in the samples was determined either by the method of Bradford (1976), or using a BioRad (Bradford) Low Protein Assay Kit. Bovine serum albumin fraction V (Merck) was used as protein standard. The antimicrobial activity of bacteriocin preparations was assayed in the agar-spot assay against Ps. syringae subsp. savastanoi strain PVBa204.
Chemical and physical stability of bacteriocin
Aliquots (1 ml) of crude bacteriocin (3200 AU ml−1) were treated with the following enzymes (Sigma): proteinase K (EC 220.127.116.11), protease (EC 18.104.22.168), trypsin (EC 22.214.171.124), α–chymotrypsin (EC 126.96.36.199), ficin (EC 188.8.131.52), lipase (EC 184.108.40.206) and α–amylase (EC 220.127.116.11). The assays were performed at a final concentration of 1 mg ml−1 at pH 6·5. Samples with and without enzymes were held at the appropriate temperature (depending on the enzymes) for 1 h.
To test stability at various temperatures and pH values, crude bacteriocin aliquots (1 ml, 3200 AU ml−1, pH 6·5) were adjusted to pH values ranging from 2·0 to 10·0 with 1 mol l−1 HCl or NaOH; they were kept for 1 h at 25 and 50 °C, and for 15 min at 60 and 100 °C, then rapidly cooled and brought back to the initial pH value.
To test the effect of organic solvents on the stability, freeze-dried crude bacteriocin aliquots (1 ml; 3200 AU ml−1) were treated with various organic solvents ( Table 3); they were kept at room temperature for 1 h and the solvents were then evaporated under vacuum. The dried samples were reconstituted with sterile distilled water.
Table 3. Effect of enzymes and organic solvents on the antimicrobial activity of Pseudomonas syringae pv. ciccaronei NCPPB2355 crude bacteriocin
Arbitrary units of activity recovered after treatments. Pseudomonas syringae subsp. savastanoi PVBa204 was used as an indicator in the agar-spot assay.
Formaldehyde (10% v/v)
Chloroform (10% v/v)
Acetone (10% v/v)
Isopropanol (10% v/v)
Ethyl alcohol (25% v/v)
Acetonitrile (70% v/v)
After each treatment, samples were assayed for antimicrobial activity against the indicator strain Ps. syringae subsp. savastanoi strain PVBa204.
Mode of action
To study the mode of action of the bacteriocin, the viability and lysis of sensitive cells of Ps. syringae subsp. savastanoi strain PVBa229 were examined. An overnight culture of the strain was diluted 1/10 into fresh Woolley’s medium and aliquots of Ps. syringae ciccaronei culture filtrate, crude bacteriocin or fraction IV from Sephadex G150 column (1200 AU ml−1, final concentration) were added to the culture which was incubated at 26 °C. The optical density (A600) of the cultures was measured at intervals (Beckman DU-65 spectrophotometer) and the number of viable bacterial cells was determined on KB agar plates.
Plasmid–DNA preparation and curing
Pseudomonas syringae ciccaronei NCPPB2355 was grown in shaking culture in King’s B medium at 26 °C. After 24 h incubation, 0·5 ml aliquots were centrifuged (9000 g, 10 min, 4 °C), the pellets were washed twice with 0·85% w/v NaCl, suspended in a minimal volume of the above solution and stored overnight at −20 °C. Plasmid–DNA preparations were obtained following the procedure of Hansen & Olsen (1978) with some modifications. The plasmid–DNA preparations were filtered through Ultrafree MC 0·45 μm (Millipore) and analysed by pulsed field gel electrophoresis (PFGE, Mapper BioRad) in 1% w/v agarose (Pulsed Field Certified Agarose, Sigma) in Tris borate buffer (45 mmol l−1 Tris; 45 mmol l−1 borate; 1 mmol l−1 EDTA; pH 8·3). Switch time was ramped from 0·1 to 2·0 s over the 16 h run time with a 180 V forward voltage and a 120 V reverse voltage.
Plasmid curing was performed by growing Ps. syringae ciccaronei cells at a high temperature (33 °C) in Nutrient Broth (Difco) (8 g l−1) as reported by Carlton & Brown (1981) with minor modifications. After two consecutive transfers, one every 48 h, the number of viable cells was determined by plating on KB agar plates. As soon as individual colonies became visible (72–96 h at 26 °C), bacteriocin production was determined in the agar-spot deferred method. Colonies were collected and some characterized for plasmid–DNA profile.
Bacteriocin production in liquid media
Amongst the liquid media tested for the production of bacteriocin, the highest titre (3200 AU ml−1 referring to the most sensitive test strain Ps. syringae subsp. savastanoi PVBa204) was obtained in the medium WP after 4 d growth at 26 °C.
The addition of mitomycin C to exponentially-growing cultures did not increase bacteriocin titre more than twofold.
Spectrum of antimicrobial activity
Fifty-six strains were used in both the agar-deferred method and the agar-spot assay to evaluate the antagonistic activity of Ps. syringae ciccaronei and the spectrum of antimicrobial activity of the bacteriocin produced in culture. Most of the strains were plant pathogenic bacteria; the others were used to obtain an overall view of the antimicrobial activity. Growth of 28 strains was inhibited by a Ps. syringae ciccaronei colony and its culture filtrate ( Table 1). Twenty-six of the inhibited strains were Ps. syringae subsp. savastanoi isolated from different host plants. In addition, two pathovars of Ps. syringae, namely persicae and apii, were found to be inhibited by Ps. syringae ciccaronei. None of the other bacteria, including several strains of nine Ps. syringae pathovars, were inhibited either by the bacterium or by the highest bacteriocin concentration tested (3200 AU ml−1). The bacteriocin-producing strain did not show susceptibility to its own active principle. No plaques were formed when agar plugs picked out of the inhibition area around a Ps. syringae ciccaronei colony were tested for the presence of bacteriophage.
Purification of bacteriocin
The purification steps are summarized in Table 2. The elution pattern at 267 nm of ultrafiltered crude bacteriocin on a Sephadex G150 column, and the specific activities of seven fraction groups, namely I (1–30), II (31–53), III (54–95), IV (96–124), V (125–150), VI (151–196) and VII (197–216), are reported in Fig. 1. The highest specific activity was associated with fraction group IV. Specific activity increased 852-fold in the final purification. Fraction IV, analysed by PAGE, revealed eight protein bands when gel was stained with Coomassie Blue. The location of a single protein band (Rf 0·41) corresponds to an inhibition halo on the unstained gel overlaid with KB agar containing the indicator strain ( Fig. 2). Bacteriocin was further purified by preparative PAGE and, when the active band recovered from the gel was subjected to SDS-PAGE, three protein bands were visible ( Fig. 2). The molecular weights of these three bands, calculated from a standard plot of log molecular weight vs relative electrophoretic mobility, were 76, 63 and 45 kDa, respectively.
Table 2. Purification of bacteriocin from Pseudomonas syringae pv. ciccaronei NCPPB2355
Arbitrary units of activity tested in the agar-spot assay against the indicator strain Ps. syringae subsp. savastanoi PVBa204.
†Arbitrary units μg −1 of lyophile.
Ammonium sulphate precipitate
Ultrafiltrate (retentate 100 kDa)
Sephadex G150 (fraction IV)
Chemical and physical stability of bacteriocin
Table 3 shows the amount of antimicrobial activity recovered after treating crude bacteriocin with several enzymes and organic solvents. The activity was completely lost after treatment with proteolytic enzymes proteinase K, protease, α–chymotrypsin and ficin, and was slightly reduced by treatment with trypsin; lipolytic or glycolytic enzymes had no effect on antimicrobial activity. Among the organic solvents tested, activity was reduced only by treatment with ethyl alcohol and acetonitrile. Table 4 reports the effects of pH and temperature on bacteriocin activity. The incubation of crude bacteriocin for 15 min at a temperature of 60 °C and a pH ranging from 2·0 to 10·0, resulted in a reduction of antimicrobial activity; treatment at 100 °C for 15 min led to complete loss of activity. Activity was stable between pH 5·0 and 7·0.
Table 4. Effect of pH and temperature on the antimicrobial activity of Pseudomonas syringae pv. ciccaronei NCPPB2355 crude bacteriocin
*Arbitrary units of activity recovered after treatments of crude bacteriocin solution (3200 AU ml −1). Pseudomonas syringae subsp. savastanoi PVBa204 was used as an indicator in the agar-spot assay.
Mode of action
The addition of aliquots of Ps. syringae ciccaronei culture filtrate, crude bacteriocin or fraction IV to growing cells of Ps. syringae subsp. savastanoi strain PVBa229 had dramatically reduced the number of viable bacteria in the culture after only 4 h of incubation ( Fig. 3); a further viable count decline was then observed during the following hours. No appreciable change in the number of viable cells occurred between 8 and 24 h growth. The optical density of the culture to which bacteriocin preparations had been added remained constant throughout the experiment ( Fig. 3). A consistent increase in turbidity of control cells was observed after 8 h incubation.
Figure 4 shows the electrophoretic pattern of Ps. syringae ciccaronei NCPPB2355 plasmid–DNA. Eighteen plasmid species, ranging in size from 4·9 to 98 kb, were resolved. Bacteriocin-negative variants were not obtained by plasmid curing at high temperature and subsequent screening of colonies on KB agar plates, and 3% of the presumptive cured colonies randomly collected and analysed for plasmid composition showed an unmodified plasmid–DNA profile.
The bacteriocin produced by Ps. syringae ciccaronei was observed to be a heat-sensitive, high molecular weight, proteinaceous compound with a narrow spectrum of activity. Its activity was partially inactivated by trypsin and completely by protease and other proteolytic enzymes. Bacteriocin was resistant to lipase, α-amylase and non-polar organic solvents, and was stable within a pH range of 5·0–7·0. These properties are more similar to those reported for bacteriocins produced by Ps. syringae subsp. savastanoi ( Iacobellis et al. 1995 ) than to those reported for syringacin W-1 ( Smidt & Vidaver 1986) and syringacin 4-A ( Haag & Vidaver 1974) produced by Ps. syringae syringae. The bacteriocins of Ps. syringae ciccaronei and Ps. syringae subsp. savastanoi actually belong to the group of thermolabile, proteolytic, enzyme-sensitive bacteriocins ( Vidaver 1976); on the other hand, although syringacins W-1 and 4-A were heat-sensitive, they were unaffected by proteolytic enzymes ( Haag & Vidaver 1974; Smidt & Vidaver 1986). Moreover, treatment with mitomycin C did not substantially increase the antimicrobial activity of the Ps. syringae ciccaronei culture; similarly, cultures of several strains of Ps. syringae subsp. savastanoi were not mitomycin C-inducible ( Iacobellis et al. 1995 ). In the case of syringacins, mitomycin C induction of Ps. syringae syringae cultures resulted in a 100-fold increase of bacteriocin concentration ( Haag & Vidaver 1974; Smidt & Vidaver 1986).
Pseudomonas syringae ciccaronei bacteriocin was purified by ammonium sulphate precipitation, membrane ultrafiltration, gel filtration and gel electrophoresis. All the purification steps gave a satisfactory purification yield of at least 62%. The estimated molecular weight for bacteriocin based on ultrafiltration was approximately 100 kDa; this was confirmed by the data from gel filtration through Sephadex G150. However, a sample of bacteriocin purified using preparative gel electrophoresis revealed three bands on SDS-PAGE with molecular weights of 76, 63 and 45 kDa, respectively. Further investigations are required to clarify whether bacteriocin activity is associated with one or more bands. The different molecular weights of the bands could suggest that they are unrelated proteins and that Ps. syringae ciccaronei could produce three distinct bacteriocins in culture.
The evidence that Ps. syringae ciccaronei harbours plasmids prompted us to investigate the location of genetic determinants for this bacteriocin. Cure experiments in which the bacteriocin-producing strain was grown at high temperatures did not generate bacteriocin-negative variants or colonies with a modified plasmid–DNA profile. Although attempts to cure Ps. syringae ciccaronei were unsuccessful, this does not rule out the fact that determinants for bacteriocin production may be located on the plasmid. Bacteriocin-coding genes are plasmid located in many bacterial species ( Vidaver 1983; Jack et al. 1995 ) but, to our knowledge, genetic determinants of bacteriocin production by pathovars or subspecies of Ps. syringae are still unknown.
Pseudomonas syringae ciccaronei and its bacteriocin strongly inhibited the Ps. syringae subsp. savastanoi strains isolated from several host plants, and affected the growth of pathovars apii and persicae of Ps. syringae. However, they were unable to inhibit the other pathovars of Ps. syringae and the other unrelated bacterial strains ( Table 1). The Ps. syringae ciccaronei culture supernatant fluid and the partially purified bacteriocin were effective in inhibiting the multiplication of Ps. syringae subsp. savastanoi in culture. Cell death did not appear to be associated with lysis as no changes occurred in the O.D. of the culture. Moreover, inhibition halos of the indicator strain tested with bacteriocin preparations in an agar-spot assay remained clear for several weeks at 26 °C. These results suggest that this bacteriocin may act bactericidally. In fact, when Ps. syringae subsp. savastanoi strains PVBa229 or PVBa304 were inoculated into the wounds made in the bark of olive plants, and Ps. syringae ciccaronei culture supernatant fluid was applied to the wounds, overgrowth formation was inihibited and the inoculated wounds healed as well as the uninoculated control ( Lavermicocca et al. 1997 ).
The findings reported here on the characterization of Ps. syringae ciccaronei bacteriocin may lay the foundations for its use in the control of olive knot disease. As Ps. syringae subsp. savastanoi has an epiphytic resident population on olive twigs, leaves and drupes ( Lavermicocca & Surico 1987), the treatment of olive plants with bacteriocin preparations can reduce the epiphytic survival of the pathogen. Furthermore, because of the high specificity of its action against Ps. syingae. subsp. savastanoi, the use of this bacteriocin as a biological control agent will result in a reduced alteration in the microbial ecology of the micro-organisms present on the olive phylloplane.
The authors thank Mr Francesco De Marzo for photographic assistance. This investigation was supported by grants from the Italian Ministry of University and Scientific and Technological Research.