FEMS Microbiology Letters 252 (2005) 19–23 www.fems-microbiology.org
Isolation and characterization of a Staphylococcus warneri strain producing an anti-Legionella peptide Yann He´chard a
, Se´bastien Ferraz a, Emilie Bruneteau a, Michael Steinert b, Jean-Marc Berjeaud a
Equipe de Microbiologie, Laboratoire de Chimie de lÕEau et de lÕEnvironnement, UMR CNRS 6008, Universite´ de Poitiers, 40 Avenue du Recteur Pineau, 86022 Poitiers, France b Institut fu¨r Molekulare Infektionbiologie, Universita¨t Wu¨rzburg, 97070 Wu¨rzburg, Germany Received 26 January 2005; received in revised form 24 March 2005; accepted 28 March 2005 First published online 27 April 2005 Edited by M. Schembri
Abstract Legionella pneumophila is a pathogenic bacterium found in freshwater environments that is responsible for pneumonia. People become infected through inhalation of contaminated droplets from water devices, such as cooling towers and showers. It is important to ﬁnd new treatments that decrease the development of Legionella. We found a Staphylococcus warneri strain that inhibits Legionella growth. This activity is due to a molecule secreted by S. warneri. This molecule displayed a high heat-stability and its activity was lost after protease treatments, suggesting that it might be a bacteriocin. Its puriﬁcation led us to conclude that this anti-Legionella molecule is an highly hydrophobic peptide. It has an original and very speciﬁc spectrum of activity, directed only toward the Legionella genus. This is the ﬁrst description of an antibacterial peptide active against Legionella. 2005 Federation of European Microbiological Societies. Published by Elsevier B.V. All rights reserved. Keywords: Bacteriocin; Legionella; Staphylococcus; Antibacterial
1. Introduction Legionella pneumophila is found in freshwater environments associated with bioﬁlms and free-living amoebae [1,2]. This Gram-negative bacterium is responsible for a severe bacterial pneumonia called LegionnaireÕs disease . People become infected with L. pneumophila after inhaling aerosols of contaminated water droplets . The aerosols may be generated by various contaminated devices (e.g., cooling towers, showers). In cases of severe contamination, treatments might be necessary *
Corresponding author. Tel.: +33 5 49 45 40 07; fax: +33 5 49 45 35
03. E-mail address: [email protected]
such as thermal disinfection, UV irradiation, use of chlorine and derivatives or metal ionization . However, these treatments are not fully eﬃcient and after a while L. pneumophila might re-establish in the treated systems. This ability to resist a variety of treatments has been mainly associated with L. pneumophila development within amoebae . The latter may protect the bacteria from treatments and amoeba-grown bacteria have been shown to be more resistant to chemical disinfectants and biocides . Therefore, it is important to ﬁnd new molecules active against Legionella that might be useful for disinfection. Bacteriocins are antibacterial proteins produced by bacteria . Many bacteriocins are produced by lactic acid bacteria and are active against foodborne pathogens such as Listeria monocytogenes or
0378-1097/$22.00 2005 Federation of European Microbiological Societies. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.femsle.2005.03.046
Y. He´chard et al. / FEMS Microbiology Letters 252 (2005) 19–23
Clostridium botulinum. They are used mainly in food protection . To our knowledge, no bacteriocin against Legionella has been described. Staphylococcus warneri has been reported to produce only one bacteriocin, nukacin ISK-1, which is active against several Gram-positive bacteria including Pediococcus acidilactici . Our study describes the partial characterization of a new bacteriocin from S. warneri RK that is active against L. pneumophila.
2. Materials and methods 2.1. Bacterial strains The bacterial strains used in this study are listed in Table 1. The Legionella strains were grown at 37 C, 5% CO2 on BCYE agar plates. All other strains were grown at 37 C on BHI agar plates. S. warneri RK, which produced the anti-Legionella molecule, was grown at 37 C in BHI liquid medium with agitation (230 rpm). 2.2. Bacterial identiﬁcation The anti-Legionella strain was identiﬁed by the ID 32 Staph and the Vitek galleries according to the manufacturerÕs instructions (Biome´rieux). 2.3. Antibacterial assays The spot on lawn assay was performed as follows. The target strain was spread onto a nutrient agar plate. Samples to be tested (10 ll) were spotted onto the surTable 1 Antibacterial spectrum of S. warneri raw extracts, assessed by well diﬀusion assays Bacterial strain
Bacillus megaterium F04 Enterobacter cloacae D03a Enterococcus faecalis V583 Escherichia coli E01a Hafnia alveia Legionella bozemanii ATCC 33217 Legionella dumoﬁi ATCC33279 Legionella longbeachae ATCC 33484 Legionella micdadei ATCC 33218 Legionella pneumophila Corby Legionella pneumophila Lens Legionella pneumophila Paris Leuconostoc mesenteroides Y105a Listeria monocytogenes EGDe Pediococcus acidilactici G06a Pseudomonas aeruginosa B06a Serratia liquefaciensa Staphylococcus aureus A15a Staphylococcus epidermidis A26a a
+ + + + + + +
face of the agar plate. The plate was incubated until the target strain lawn developed. Antibacterial activity was revealed by a zone of inhibition around the test samples. The well diﬀusion assay was performed as follows. A bacterial suspension (100 ll at OD600 0.1) was spread onto a BCYE agar plate. Wells were punched into the agar. Samples to be tested (100 ll) were placed in the wells. Plates were incubated until the target strain grew. Antibacterial activity was revealed by a zone of inhibition around the well. A twofold serial dilution of samples was performed to measure antibacterial activity with the well diﬀusion assay. The activity, in arbitrary units (AU), of a sample was deﬁned as the highest dilution that gave a clear zone of inhibition. 2.4. Puriﬁcation In order to characterize the anti-Legionella peptide, a three-step puriﬁcation was conducted as described below. S. warneri RK was grown for 24 h at 37 C in BHI medium. The culture was centrifuged (6000g, 15 min) to remove the bacteria, then the supernatant was heated at 70 C for 15 min. This sample was termed the raw extract (RE). Firstly, the raw extract (200 ml) was diluted (50:50) with sodium acetate buﬀer (20 mM, pH 5) and applied to a weak cation exchange chromatography column (HiprepTM 16/10 Carboxy-Methyl FF, Amersham Biosciences). The column was washed successively with 100 ml of sodium acetate buﬀer (20 mM, pH 5) containing 0, 0.1 and 1 M NaCl. The anti-Legionella peptide was eluted with 1 M NaCl. Secondly, the latter fraction was applied onto a solid phase extraction C18 cartridge (Sep-pak plus, Waters), washed successively with 5 ml of 0%, 10%, 20%, 30%, 40% and 80% acetonitrile containing 0.1% triﬂuoroacetic acid. Each fraction was concentrated under vacuum, lyophilized and resuspended in 1 ml of water. The anti-Legionella peptide was eluted with 80% acetronitrile. Thirdly, the active fraction was lyophilized, solubilized with 1 ml of 40% acetonitrile and injected on a Kromasil C8 reverse-phase ˚ , 4.6 · 250 mm, HPLC analytical column (5 lm, 100 A A.I.T.). Reverse phase HPLC was conducted on a Perkin–Elmer series 200 LC pump, ﬁtted with a Perkin–Elmer 785A detector. Elution was monitored for absorbance at 220 nm. Separation was carried out using a water/acetonitrile/triﬂuoroacetic acid 0.1% (v/v) solvent system. After an initial 5 min wash with 20% acetonitrile, elution was achieved in 40 min at a ﬂow rate of 0.8 ml min 1 with a 30 min linear gradient from 20% to 100% acetonitrile, followed by a 10 min wash with 100% acetonitrile. After evaporation under vacuum and lyophilization, each fraction was resuspended in 1% ml water and tested for anti-bacterial activity. The active fractions were lyophilized, resuspended in 50% acetonitrile-0.1% formic acid, and analyzed by
Y. He´chard et al. / FEMS Microbiology Letters 252 (2005) 19–23
mass spectrometry, on a Perkin–Elmer Sciex API 165 mass spectrometer equipped with an ion spray source. The sample analysis was carried out by infusion injection at a ﬂow rate of 5 l min 1. The instrument scale for the mass-to-charge (m/z) ratio was calibrated with the ions of the ammonium adduct of polypropylene glycol. Scan data were obtained with LC2-Tune and mass calculation was done with Biomultiview 1.2 (Software package Sciex).
Table 2 Stability of the anti-Legionella activity. The raw extract was submitted to various treatments before measurement of its activity by the well diﬀusion assay Treatment
None Trypsin 20 mg ml 1 Proteinase K 20 mg ml 70 C, 15 min 90 C, 15 min 121 C, 15 min
16 0 0 16 16 8
3. Results and discussion 3.1. Characterization of the producing strain A bacterial strain, isolated from the environment, displayed an anti-Legionella activity by the spot on lawn test (Fig. 1). Two diﬀerent galleries were used to identify this strain; they both resulted in reliable identiﬁcation of the S. warneri species. The strain was named S. warneri RK. 3.2. Biochemical characterization In order to characterize the anti-Legionella molecule, the raw extract (see above) was subjected to various treatments. This extract was treated with proteases: proteinase K 1 mg ml 1 or trypsin 1 mg ml 1 for 1 h at 37 C; or by heating 15 min at 70 C, 15 min at 90 C or 15 min at 121 C (autoclaving). The activity of each sample was compared to the non-treated extract by the well diﬀusion assay. The results are presented in
Table 2. The protease treatments led to a total loss of inhibitory activity, showing that the anti-Legionella molecule is proteinaceous. The heat treatments, 70 and 90 C, did not aﬀect the inhibitory activity whereas the treatment at 121 C led to a 50% decrease. According to the results from protease activity, the molecule might be a bacteriocin, deﬁned as an antibacterial protein produced by bacteria. The results from the heat treatments suggest that the molecule might be a peptide rather than a protein since peptides are less sensitive to high temperature treatments, mainly because they are able to spontaneously refold. 3.3. Anti-bacterial spectrum Various bacterial strains, Gram positive and Gram negative, were used as target organisms to test their sensitivity to S. warneri RK. The spot on lawn test was used to assess the activity of both raw extracts and S. warneri RK cultures. The results are described in Table 1. All the Legionella strains tested, from diﬀerent species, were sensitive. None of the other strains, except P. acidilactici, were sensitive. P. acidilactici has been reported to be a target of nukacin ISK-1, the only bacteriocin known to be produced by S. warneri . Therefore, a nukacin-like bacteriocin might be produced by S. warneri RK that explains its anti-Pediococcus activity. This led us to question whether only one bacteriocin, active against Legionella and P. acidilactici, or two diﬀerent bacteriocins are produced by S. warneri RK.
Table 3 Puriﬁcation of the anti-Legionella peptide produced by S. warneri RK
Fig. 1. Spot on lawn assay against L. pneumophila Lens. The two spots correspond to S. warneri RK colonies, around which a zone of inhibition of Legionella growth can be seen.
Total activity (AU)a
Raw extract Carboxymethyl eluate Sep-Pack C18 eluate RP-HPLC eluate
200 100 1 1
8 16 512 256
1600 1600 512 256
100 100 32 16
Y. He´chard et al. / FEMS Microbiology Letters 252 (2005) 19–23
Fig. 2. Reverse phase HPLC proﬁle. Fraction 1 was active against P. acidilactici and fraction 2 was active against L. pneumophila.
3.4. Puriﬁcation The anti-Legionella activity was monitored during a three-step puriﬁcation procedure. Results are given in Table 3. Based on activity measurement, 100% of the anti-Legionella activity present in the raw extract was eluted from the carboxy-methyl column. This result underlines that the peptide was positively charged at pH 5. In the ﬁnal step, 16% of the anti-Legionella activity was recovered from the reverse-phase HPLC column. The chromatogram is presented in Fig. 2. Fraction 1, corresponding to the peak at 21 min, was active toward P. acidilactici but not toward L. pneumophila. On the contrary, fraction 2, corresponding to the peak at 38 min, was active toward L. pneumophila but not toward P. acidilactici. These results clearly demonstrate that two diﬀerent bacteriocins are produced by S. warneri RK. The anti-Legionella activity was eluted at 100% acetonitrile, indicating that the corresponding peptide is highly hydrophobic. The mass spectrometry analysis of the anti-Pediococcus fraction led to a molecular mass of 5466.4 Da. The nukacin ISK-1 molecular mass is 2960.1 Da . Accordingly, the anti-Pediococcus bacteriocin produced by S. warneri RK is clearly diﬀerent from nukacin ISK1. The anti-Legionella active fraction was subjected to mass spectrometry analysis giving two molecular masses of 2613.8 and 2449.4 Da. This result shows that two different peptides were present in the fraction, one of which corresponds to the anti-Legionella bacteriocin. Finally, database screening revealed that the molecular mass of none of the previously described bacteriocins corresponds to the molecular masses of these three peptides. The anti-Legionella and the anti-Pediococcus bacteriocins are therefore novel molecules. To conclude, we have described two new bacteriocins secreted by S. warneri RK. No anti-Legionella peptide has been described so far in the literature. This emphasizes the originality of the anti-Legionella bacteriocin,
the activity of which is directed towards, and seems restricted to the Legionella genus. This spectrum of activity is intriguing since bacteriocins are usually directed toward species related to the producer. Not only was it not active against the closest strains we tested, but it was active against Legionella, which is a Gram negative genus. This bacteriocin needs to be fully characterized and sequenced to initiate genetic studies. The mode of action is also worth exploring since this bacteriocin might have a novel mode of action and could be used as an alternative treatment against Legionella.
Acknowledgments We acknowledge Claire Souchaud and Silke Hammer for technical help. The work was supported by the Deutsche Forschungsgemeinschaft DFG (SFB 630).
References  Molofsky, A.B. and Swanson, M.S. (2004) Diﬀerentiate to thrive: lessons from the Legionella pneumophila life cycle. Mol. Microbiol. 53, 29–40.  Steinert, M., Hentschel, U. and Hacker, J. (2002) Legionella pneumophila: an aquatic microbe goes astray. FEMS Microbiol. Rev. 26, 149–162.  Fields, B.S., Benson, R.F. and Besser, R.E. (2002) Legionella and LegionnairesÕ disease: 25years of investigation. Clin. Microbiol. Rev. 15, 506–526.  Muder, R.R., Yu, V.L. and Woo, A.H. (1986) Mode of transmission of Legionella pneumophila. A critical review. Arch. Intern. Med. 146, 1607–1612.  Kim, B.R., Anderson, J.E., Mueller, S.A., Gaines, W.A. and Kendall, A.M. (2002) Literature review – eﬃcacy of various disinfectants against Legionella in water systems. Water Res. 36, 4433–4444.  Abu Kwaik, Y., Gao, L.Y., Stone, B.J., Venkataraman, C. and Harb, O.S. (1998) Invasion of protozoa by Legionella pneumophila and its role in bacterial ecology and pathogenesis. Appl. Environ. Microbiol. 64, 3127–3133.
Y. He´chard et al. / FEMS Microbiology Letters 252 (2005) 19–23  Barker, J., Brown, M.R., Collier, P.J., Farrell, I. and Gilbert, P. (1992) Relationship between Legionella pneumophila and Acanthamoeba polyphaga: physiological status and susceptibility to chemical inactivation. Appl. Environ. Microbiol. 58, 2420–2425.  Riley, M.A. and Wertz, J.E. (2002) Bacteriocins: evolution, ecology, and application. Annu. Rev. Microbiol. 56, 117–137.  Cleveland, J., Montville, T.J., Nes, I.F. and Chikindas, M.L. (2001) Bacteriocins: safe, natural antimicrobials for food preservation. Int. J. Food Microbiol. 71, 1–20.
 Sashihara, T., Kimura, H., Higuchi, T., Adachi, A., Matsusaki, H., Sonomoto, K. and Ishizaki, A. (2000) A novel lantibiotic, nukacin ISK-1, of Staphylococcus warneri ISK-1: cloning of the structural gene and identiﬁcation of the structure. Biosci. Biotechnol. Biochem. 64, 2420–2428.  Kimura, H., Sashihara, T., Matsusaki, H., Sonomoto, K. and Ishizaki, A. (1998) Novel bacteriocin of Pediococcus sp. ISK-1 isolated from well-aged bed of fermented rice bran. Ann. N. Y. Acad. Sci. 864, 345–348.