Diversity of bacteriocinogenic lactic acid bacteria ...

Diversity of bacteriocinogenic lactic acid bacteria isolated from Mediterranean fish viscera Sarra Migaw, Taoufik Ghrairi, Yanath Belguesmia, Yvan Choiset, Jean-Marc Berjeaud, Jean-Marc Chobert, Khaled Hani & Thomas Haertlé World Journal of Microbiology and Biotechnology ISSN 0959-3993 World J Microbiol Biotechnol DOI 10.1007/s11274-013-1535-6

1 23

Your article is protected by copyright and all rights are held exclusively by Springer Science +Business Media Dordrecht. This e-offprint is for personal use only and shall not be selfarchived in electronic repositories. If you wish to self-archive your article, please use the accepted manuscript version for posting on your own website. You may further deposit the accepted manuscript version in any repository, provided it is only made publicly available 12 months after official publication or later and provided acknowledgement is given to the original source of publication and a link is inserted to the published article on Springer's website. The link must be accompanied by the following text: "The final publication is available at link.springer.com”.

1 23

Author's personal copy World J Microbiol Biotechnol DOI 10.1007/s11274-013-1535-6

ORIGINAL PAPER

Diversity of bacteriocinogenic lactic acid bacteria isolated from Mediterranean fish viscera Sarra Migaw • Taoufik Ghrairi • Yanath Belguesmia • Yvan Choiset • Jean-Marc Berjeaud • Jean-Marc Chobert Khaled Hani • Thomas Haertle´

•

Received: 17 July 2013 / Accepted: 21 October 2013 Ó Springer Science+Business Media Dordrecht 2013

Abstract Nine lactic acid bacteria strains showing bacteriocin-like activity were isolated from various fresh fish viscera. The following species were identified based on 16S rDNA sequences: Enterococcus durans (7 isolates), Lactococcus lactis (1) and Enterococcus faecium (1). These strains were active against Listeria innocua and other LAB. Random amplified polymorphic DNA analyses showed four major patterns for the E. durans species. PCR analyses revealed a nisin gene in the genome of the Lc. lactis strain. Genes coding enterocins A, B and P were found in the genome of the E. faecium isolate. Enterocins A and B genes were also present in the genome of E. durans GM19. Hence, this is the first report describing E. durans strains producing enterocins A and B. Electrospray ionization mass spectrometry revealed that the purified bacteriocin produced by the E. durans GMT18 strain had an exact molecular mass of 6,316.89 Da. This bacteriocin was designated as durancin GMT18. Edman sequencing failed to proceed; suggesting that durancin GTM18 may contain terminal lanthionine residues. Overall, the results obtained revealed the presence of a variety of enterococci in

S. Migaw T. Ghrairi (&) K. Hani De´partement de Biochimie, Faculte´ de Me´decine Ibn El Jazzar, Unite´ de Recherche U12-ES03, 4002 Sousse, Tunisia e-mail: [email protected] Y. Belguesmia Y. Choiset J.-M. Chobert T. Haertle´ UR 1268, Biopolyme`res Interactions Assemblages, Equipe FIP, INRA, B.P. 71627, 44316 Nantes Cedex 03, France J.-M. Berjeaud Ecologie et Biologie des Interactions, Equipe Microbiologie de l’Eau, UMR CNRS 7267, 40 Avenue du Recteur Pineau, 86022 Poitiers, France

Mediterranean fish viscera, as evidenced by their genetic profiles and abilities to produce different bacteriocins. These strains could be useful for food biopreservation or as probiotics. Keywords Enterococcus durans Enterococcus faecium Bacteriocins Listeria Virulence factors Antibiotic resistance

Introduction The intestinal microflora of fish is often reported to be highly variable (Asfie et al. 2003) and to depend on several factors, including environmental and temperature conditions. These factors have, for instance, influential effects on the selection pressure exerted on the microflora by harsh conditions in the gut-intestinal tract. This raises several questions, including whether there are specific microfloras associated with the digestive tract of specific fishes or whether these bacteria are only a reflection of the populations existing in the environment. In fact, several species of lactic acid bacteria (LAB) have been reported to constitute inherent components of the natural intestinal microflora of healthy fishes (Leroi 2009). LABs have been used for the preservation of several food products for many years, including fermented foods. They are able to limit contamination from opportunistic microflora and limit the growth of undesired bacteria under various conditions in fermented and in non-fermented food products such seafood. Some LAB strains have also been successfully used as probiotics (Agerholm et al. 2000). The major groups of LAB used as starters in fermented foods belong to the Lactobacillus, Lactococcus, Pediococcus and Streptococcus species.

123

Author's personal copy World J Microbiol Biotechnol

LABs emerge as part of the normal intestinal microbiota of fish within the first days of birth (Ringø 2008). Lactobacilli (Lb.), notably Lb. plantarum, have been detected in Atlantic salmon, pollock, Arctic char and cod. Carnobacterium (C.), including C. maltaromaticum, C. divergens and C. gallinarum, have also been isolated from these species as well as from rainbow trout (Huber et al. 2004). LAB strains with antimicrobial activities against fish pathogens have been extensively investigated to explore their potential applications as probiotics in aquaculture (Munõz-Atienza et al. 2011). Selected antagonistic LAB, or their bacteriocins, was used for the control of pathogens in certain foods (El-Ghaish et al. 2010). Bacteriocins are defined as proteinaceous substances displaying bactericidal activities against several closely related species. Based on their structural and biochemical properties, bacteriocins produced by LAB are subdivided into three major classes (Klaenhammer 1993). Class I bacteriocins consists of lantibiotics, which are small posttranslationally modified peptides that include unusual amino acids, such as lanthionine, methyllanthionine, dehydrobutyrine and dehydroalanine. Nisin produced by Lc. lactis ssp. lactis (Cotter et al. 2005) is the most prominent bacteriocin, being the only lantibiotic applied in the food industry. Class II bacteriocins contain unmodified bacteriocins, which are further subdivided into three subclasses: class IIa (anti-Listeria peptides), class IIb (two-peptide bacteriocins) and class IIc (one-peptide bacteriocins or non-pediocin-like bacteriocins). Class III bacteriocins is a group of large, heat-labile proteins with antimicrobial activity that include helveticin J and lactococcin MMT05. Enterocins, bacteriocins that are primarily produced by enterococci, have been isolated from various fermented food products, including enterocins A and B (Casaus et al. 1997), enterocin P (Cintas et al. 1997), enterocin L50A and L50B (Cintas et al. 1998) and enterocin I (Floriano et al. 1998). They are generally known to be active against various Gram-positive bacteria, such as spoilage and foodborne pathogens. E. faecalis and E. faecium are the two most prominent bacteriocin-producing species in the enterococci genus. Recent studies have also discovered more unusual bacteriocinogenic species, including E. hirae (Sańchez et al. 2007), E. mundtii (Campos et al. 2006), E. gallinarum (Jennes et al. 2000), E. casseliflavus (Sabia et al. 2002) and E. durans (Hu et al. 2008). Despite this large volume of data, little is known on the LAB community within the viscera of Mediterranean fish. This study reports on the isolation and characterization of new bacteriocinic strains of E. durans, E. faecium and Lc. lactis from a number of fish species captured along the Tunisian coastline.

123

Materials and methods Culture media and reagents The culture media (M17, MRS, APT) used in this work were purchased from Difco Laboratories (Detroit, USA); all chemicals were obtained from Sigma-Aldrich (Sigma Aldrich Handels GmbH, Vienna, Austria). Sampling and screening for antimicrobial activities Freshly captured fish samples were bought to the laboratory from the local markets of the east Mediterranean coast of Tunisia (Sousse, Mahdia and Monastir). Seven species were collected: Spicara smaris (Picarel), Mullus surmutelus (Red Mullet), Trachurus trachurus (Horse mackerel), Boops boops (Bogue), Mugil cephalus (Stripped Mullet), Pagellus erythrinus (Pandora) and Scomber scombrus (Atlantic Mackerel). Five grams of viscera were removed, fragmented in sterile conditions and homogenized in 40 ml sterile physiological water (0.9 % NaCl). The obtained suspension was serially diluted. Aliquots (200 ll) of the 10-4, 10-5, 10-6 and 10-7 dilutions were spread in different media (M17, MRS, APT) and incubated at 30 °C for 24–48 h. Ten representative colonies with different morphology from the plates 10-6 and 10-7 were selected for characterization as previously described elsewhere (Ghrairi et al. 2004). Catalase-negative strains were screened for bacteriocin production by agar-well diffusion assay (Tagg et al. 1976). The presence of distinct inhibition zones around the well was considered a positive antagonistic effect. For this step, Listeria (L.) ivanovii BUG496 and Lc. lactis subsp. cremoris ATCC 111603 were used as indicator strains. Listeria ivanovii and Listeria innocua were tested in this study since these Listeria strains have the same biochemical characteristics of L. monocytogenes, except for pathogenicity. Characterization of the antimicrobial activities To evaluate the sensitivity of the antimicrobial metabolites to lytic enzymes, aliquots of the cell-free supernatant were treated with proteinase K, trypsin, a-chymotrypsin, lipase and a-amylase at a final concentration of 0.1 mg ml-1 for 2 h at 37 °C. The enzymes were dissolved in buffers as recommended by the manufacturer (Sigma-Aldrich). They were finally inactivated by a heating at 100 °C for 5 min. Residual bacteriocin activities were determined using the agar well diffusion assay against the indicator strain Lb. bulgaricus 340 as described above. The stabilities of the active substances were studied at different pH values by adjusting the pH of the supernatant


to values ranging from 2 to 10 using sterile 5 M NaOH or 5 M HCl. After 2 h of incubation at 30 °C, the pH was adjusted to 6.5 and substances were tested for antimicrobial activity using the agar well diffusion method. The thermostability of the inhibitory substances produced by the isolates under investigation was checked at different temperatures: 60 °C/30 min, 100 °C/30 min and 121 °C/15 min at pH 6.0. Residual activities were assayed by the agar well diffusion method and compared to those of the corresponding controls. Kinetics of bacteriocin production In order to determine the kinetics of bacteriocin production, 10 ml of MRS broth was inoculated with 1 % (v/v) of an overnight culture of the producer strain. The culture was incubated at 30 °C and samples were taken every hour to determine bacteriocin production (AU/ml) and biomass (OD at 600 nm). One arbitrary unit (AU) of bacteriocin was defined as the reciprocal of the highest dilution yielding a zone of growth inhibition on the indicator strain. Identification of bacterial isolates Bacteriocin-producing strains were identified according to their physiological and biochemical properties, as described by Schillinger and Lu¨cke (1987). Further identification was made by PCR and 16S rRNA sequence analyses as described by Weisburg et al. (1991). The genomic DNA of the bacteria was extracted using the DNeasy Blood Tissue Kit (Qiagen GmbH, Hilden, Germany) and served as a template for 16S rRNA gene amplification. Amplicons for sequencing were purified with a QIAprep Spin Miniprep Kit (Qiagen S.A., Courtabœuf, France). All sequence analyses were performed on an ABI 370 automated sequencer (Perkin–Elmer, Boston, MA, USA). The resulting sequences were compared to those available at the Genebank using the BLASTN alignment software (http:// www.ncbi.nlm.nih.gov/blast). The Mega 5 computer software program (2011) was used for sequence alignment and phylogenetic tree analysis. Random amplified polymorphic DNA (RAPD) analysis RAPD analysis was carried out as previously described elsewhere (Weisburg et al. 1991). Three random microsatellites primers OPL, M13 and E1 (Invitrogen, France) were used for RAPD amplification. PCR amplification was carried out using a final volume of 25 ll containing 30–40 ng of DNA, 3 mM MgCl2 and 0.15 units (U) of Taq DNA polymerase (Promega). Amplifications were performed in an Eppendorf Thermal Cycler (Barloworld scientific, Cambridge, UK) with initial denaturation at 94 °C

for 2 min, 25 cycles of 1 min denaturation at 94 °C, 1 min annealing at 46 °C and 2 min extension at 72 °C, followed by final extension for 7 min at 72 °C. The amplification products were separated by agarose gel electrophoresis and visualized after ethidium bromide staining. Each primer was amplified at least twice and only reproducible products were taken. Fragment lengths were estimated by comparison with standard size markers. The RAPD bands were analysed according to the presence or the absence of bands. Bacteriocin purification and molecular mass determination The purification of bacteriocins was carried out as previously described (Ghrairi et al. 2004). It involved ammonium sulfate precipitation, followed by solid phase chromatography on Sep-pack C18 cartridges (Waters, Millipore) and reversed phase-high performance liquid chromatography (RP-HPLC). The fractions with highest bacteriocin activities were concentrated by lyophilization, resuspended in deionized water and stored at -20 °C. The molecular weights of bacteriocins were checked by tricineSDS-PAGE. Gel solutions were prepared as described by Scha¨gger and von Jagow (1987). The molecular masses of the peptides were also determined by electrospray ionization mass spectrometry (ESI– MS) using a Xevo Q-TOF (Waters) mass spectrometer. Each sample was suspended in 50 % acetonitrile/0.2 % formic acid (v/v) and analysed in a positive mode. The spray voltage was set at 3.0 kV and the source and desolvation temperatures were 120 °C and 250 °C, respectively. The data were processed using the BioLynx tool of the MassLynx software (Waters). Antibiotics sensitivity The isolates were assayed for their susceptibility to various commercial antibiotics used at various concentrations, namely gentamicin (MIC B32 lg ml-1), ampicillin (MIC B4 lg ml-1), tetracycline (MIC B2 lg ml-1), kanamycin (MIC B1,024 lg ml-1), vancomycin (MIC B4 lg ml-1) and penicillin (MIC B 4 lg ml-1), using the disk diffusion antibiotic sensitivity method. The sensitivity thresholds were defined according to the method published by the National Committee for Clinical Laboratory Standards (NCCLS 2003). The presence of a clear zone of growth inhibition of indicator strain around the disk was considered as a positive reaction. Identification of the inhibitory substances by PCR The DNA of the isolates was further screened by PCR for the presence of known genes coding enterocins using

123


specific primers as described by Cintas et al. (2000) and de Vuyst et al. (2003). PCR was carried out using primers for 7 enterocin-encoding genes: bacteriocins class IIa (entA, entP, ent31), class IIb (entB, entQ), class I (Cyl) and class II (entAS48). Amplification reactions were carried out using the DNA solutions described above as templates in a DNA thermal cycler (Mastercycler-personal, Eppendorf, Hamburg, Germany), followed by an increase to 72 °C for 60 s. The extension of the amplified product was at 72 °C for 7 min. The amplified products were separated by electrophoresis in 1 % (w/v) agarose gels in TBE buffer. A 100-bp DNA ladder (Invitrogen) was used as a molecular weight marker. PCR detection of virulence factors The genomic DNA of selected enterococci isolates was used as a template for the identification of genes coding the following virulence factors: asal (aggregation substance), cylA and cylB (cytolysin), ace (collagen binding protein), efaAfs (enterococcal endocarditis antigen) and espfm (enterococcal surface protein). Their presence was verified by gene specific amplification primers using classic PCR. E. faecalis MM4594 was used as a positive control (El-Ghaish et al. 2010).

Results Detection of LAB strains with antimicrobial activities Thirty-nine LAB strains were isolated from the viscera of the seven fish samples and screened for their ability to produce antibacterial substances by the well diffusion method. Nine isolates (GM11 to 15 and 17–20) exhibited marked inhibition activities and were, therefore, selected for further analyses. These strains were Gram-positive and catalase-negative. All strains have been isolated on MRS medium except GM11 strain which was isolated on M17 medium. They were then tested for their antimicrobial activities against Gram-positive and Gram-negative indicator strains and the results are summarized in Table 1. Most of the LAB isolates produced well defined inhibition zones against at least one of the indicator strains tested. In fact, all of the nine bacteriocin genic isolates produced large inhibition zones against Lc. lactis ssp. cremoris ATCC 11603 and E. faecalis JH.2.2. Two Lactobacillus species, namely Lb. brevis F145 DSM2011 and Lb. bulgaricus 340, were inhibited. The Lb. delbrueckii DSM20081 strain was, however, insensitive to the cell-free supernatants. Among the pathogenic and food-spoilage bacteria tested, L. ivanovii, L. innocua, B. cereus and

Table 1 Antimicrobial activity of strains cell-free supernatant Fish species¥ Indicator strain

A GM11

B GM12

C GM13

D GM14

E GM15

D GM17

F GM18

G GM19

C GM20

MMT21*

Lc. lactis ATCC 11603

111

111

111

111

111

111

111

111

111

111

Lc. lactis MMT24

2

2

2

2

2

2

2

2

2

2

Lc. garvieae ATCC 43921

2

2

2

2

2

2

11

11

11

2

Lc. cremoris

2

2

2

2

2

2

2

2

2

2

Lb. brevis F145

111

11

11

11

1

111

1

111

11

2

Lb. delbrueckii DSM20081

2

2

2

2

2

2

2

2

2

2

Lb. bulgaricus 340

111

11

11

11

1

111

111

111

11

11

E. faecalis JH.2.2

111

111

111

111

111

111

111

111

111

111

Mc. luteus ATCC 10240

2

2

2

2

2

2

2

2

2

2

L. ivanovii BUG496

111

11

1

1

1

1

11

11

11

111

L. innocua CIP

111

11

1

2

1

1

11

11

11

11

B. megaterium ATCC 10778

2

2

2

2

2

2

2

2

2

2

B. cereus E. coli XL1

1 2

1 2

1 2

1 2

1 2

1 2

1 2

1 2

1 2

1 2

S. enterica CIP 8297

2

2

2

2

2

2

2

2

2

2

Saccharomyces cerevisiae

111

1

11

1

1

11

1

1

1

2

Candida pseudotropicalis

111

11

11

1

11

11

11

11

11

nd**

- No activity, ? radius of inhibition zone (\6 mm), ?? radius of inhibition zone (6–12 mm), ??? radius of inhibition zone ([12 mm) * **

Producer of enterocins A and B (Ghrairi et al. 2004) nd not determined

¥

A Spicara smaris (Picarel); B Mullus surmutelus (Red Mullet); C Trachurus trachurus (Horse mackerel); D Boops boops (Bogue); E Mugil cephalus (Stripped Mullet); F Pagellus erythrinus (Pandora); G Scomber scombrus (Atlantic Mackerel)

123


Candida pseudotropicalis were inhibited, but no inhibition was observed against B. megaterium. Moreover, none of the cell-free supernatants were able to inhibit Gram-negative bacteria such as E. coli XL1 and S. enterica CIP 8297. The GM18, -19 and -20 strains displayed good antibacterial activity against Lc. garviae. Only strains GM15, -18 and 20 were inhibited by the cell-free supernatant of E. faecium MMT21, the enterocin A and B producer strain (data not shown). Taken together, the findings indicated that, among the nine isolates, isolate GM11 was the most active against the indicator strains used. Biochemical characterization of the antibacterial substances The effects of enzymes, pH and heat on the activity of the bacteriocins produced by the isolates under investigation are shown in Table 2. The antimicrobial substances produced by the isolates were inactivated only by the proteolytic enzymes proteinase K, trypsin and a-chymotrypsin, which is a characteristic of bacteriocin-like-compounds. Other non-proteolytic enzymes had no effect on bacteriocin activities. The bacteriocins were active over a wide range of pH from 2 to 10 (Table 2). They were also stable after heating for 30 min at 100 °C and even after autoclaving for

15 min at 121 °C. This level of thermostability is, in fact, highly valued in the food industry, particularly in applications involving cooked/fried food preservatives. Growth kinetics and bacteriocin biosynthesis Figure 1 shows the growth curve and profile of bacteriocin production by E. durans GM18 at 30 °C. Bacteriocin production was first recorded during the mid-exponential phase of growth (i.e. after 15 h) in MRS broth. This was then followed by a gradual increase, with maximum levels of antimicrobial activity (5,600 AU ml-1) being detected at the beginning of the stationary phase of growth. Bacteriocin production was noted to remain stable until the end of growth. Genotyping of the isolates The isolates were identified to species level by 16S rDNA gene sequencing. The determined sequences were directly compared to those available at the Genebank database. The isolates (GM13, -14, -15, -17, -18, -19 and -20) were identified as belonging to the E. durans species. Isolate GM12 was identified as E. faecium, while GM11 was phylogenetically placed within the Lc. lactis family. Thus,

Table 2 Effect of enzymes, pH and heat treatment on antibacterial activity of selected LAB strains Treatments

GM11

GM12

GM13

GM14

GM15

GM17

GM18

G19

GM20

Proteinase K

2

2

2

2

2

2

2

2

2

Lipase

1

1

1

1

1

1

1

1

1

Trypsin

2

2

2

2

2

2

2

2

2

Amylase

1

1

1

1

1

1

1

1

1

a-chymotrypsin

2

2

2

2

2

2

2

2

2

2–6

111

111

111

111

111

111

111

111

111

6–8

111

111

111

11

111

11

111

111

111

11

11

11

11

11

11

11

11

11

100 °C/30 min

1

1

1

1

1

1

1

1

1

121 °C/20 min

1

1

1

1

1

1

1

1

1

Enzymes

pH

8–10 Heat

Antibiotics Gentamicin

R

S

S

S

S

S

S

S

S

Ampicillin

R

R

S

S

R

S

S

R

S

Tetracycline

S

R

R

S

S

R

R

S

R

Kanamycin

S

R

R

R

S

R

S

S

R

Vancomycin

S

R

S

S

S

S

S

S

S

Penicillin

R

S

S

S

R

S

S

S

S

- No activity, ? radius of inhibition zone (\6 mm), ?? radius of inhibition zone (6–12 mm), ??? radius of inhibition zone ([12 mm) R resistant, S sensitive

123


differentiation between strains with relatively low significance (data not shown).

PCR analysis of genes of bacteriocins

Fig. 1 Growth and bacteriocin activity of E. durans GM18 at uncontrolled pH at 30 °C

among the coccoid LAB, the most frequently isolated species is E. durans. The phylogenetic trees illustrating the relationships between the nine strains with their corresponding similar sequences from other closely related bacteria are shown in Fig. 2. Three distinct branches were revealed. Branch 1 comprised the 7 E. durans species. Branches 2 and 3 included E. faecium and Lc. lactis, respectively. RAPD analysis RAPD was used for the genotyping of E. durans strains. Figure 3 shows examples of RAPD profiles obtained with oligonucleotide primers OPL, M13, seven strains of E. durans (GM13, -14, -15, -17, -18, -19 and -20) and one strain of E. faecium (GM12). A clear discrimination between the two species investigated was obtained. At least four different profiles were observed among the seven profiles from the E. durans isolates. While E. durans GM14, -15, -17 and -20 displayed similar profiles, E. durans GM13, -18 and -19 showed different profiles. Two other E. durans, namely primers E1 and E2, showed Fig. 2 Phylogenetic tree of active lactic acid bacteria (LAB) strains isolated from digestive tractus of Tunisian fishes. Nucleotide sequence accession numbers are in brackets. Phylogenetic trees were generated using UPGMA, neighbor-joining and maximum likelihood analysis. Bootstrap probabilities are given in parentheses at branch nodes (UPGMA). Coxellia sp. was used as an outgroup taxon

123

PCR amplification was performed to detect whether the strains under investigation carried the structural genes of known bacteriocins. PCR fragment generated by strain Lc. lactis GM11 corresponded to the gene encoding nisin (510 bp), namely the nisin Z-producing strain Lc. lactis MMT01 (Fig. 4). The E. faecium GM12 strain yielded into three PCR fragments corresponding to the expected sizes of the gene fragments of enterocin A (130 bp), enterocin B (180 bp) and enterocin P (132 bp), respectively (Fig. 4). The E. durans GM19 strain produced amplification fragments corresponding to the enterocin A and enterocin B genes (Fig. 4). No products were obtained when the same primers were used with the other E. durans isolates. Bacteriocin purification and molecular mass determination The bacteriocin produced by the E. durans GM18 strain was purified using a three-step protocol, including ammonium sulfate precipitation, Sep-pack C18 chromatography and RP-HPLC. After the last step, the bacteriocin was eluted as a single peak of activity corresponding to a chromatographic retention time of 31.5 min (data not shown). The collected active fraction was checked on Tricine-SDS-PAGE (Fig. 5). A single protein band was also observed when stained with Coomassie blue, which corresponded well with the prominent inhibition zone observed on gel overlaid with soft agar seeded with the indicator strain. This result indicated that the molecular weight of the bacteriocin of E. durans GM18 was below 3,300 Da. However, mass spectrometry analysis revealed that the molecular mass of the purified peptide was 6,316.89 ± 0.64 Da (Fig. 6).

Author's personal copy World J Microbiol Biotechnol 1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

200 bp 100 bp

Fig. 3 RAPD-PCR profiles of strains of E. durans and E. faecium obtained with OPL primer (lanes 1–8) and M13 primer (lanes 10–17). Lane 1 E. durans GM20; lane 2 E. durans GM19; lane 3 E. durans GM18; lane 4 E. durans GM17; lane 5 E. durans GM15; lane 6 E. durans GM14; lane 7 E. durans GM13; lane 8 E. faecium GM12; lane

Fig. 4 Agarose gel electrophoresis of PCR analysis of known bacteriocins. A nisin gene PCR analysis of isolate GM11: lane 1 DNA GM11 isolate; lane 3 nisin positive control (Lc. lactis MMT01); B enterocins genes screening for GM12 isolate: lane 1 ent A; lane 2 ent B; lane 3 ent P; C enterocins genes screening for GM19 isolate: lane 1 ent A; lane 2 ent B. Lanes A2 and B4, 100 bp DNA Ladder (invitrogen)

A

Screening for virulence factors and susceptibility to antibiotics The isolated strains were investigated for the presence of six virulence factors (asal, cyl A and B, ace, efaAfs and espfm). The results from PCR analyses revealed that only the E. faecium GM12 strain harbored a cylA gene, with no other virulence genes being recorded (data not shown). This suggests that the occurrence of virulence genes in enterococci from the fish viscera was very limited. The antibiotic resistance of the enterococcal strains showing significant antibacterial activities was determined by the disk diffusion method as described above. The

9 100 bp DNA Ladder (invitrogen). Lane 10 E. durans GM20; lane 11 E. durans GM19; lane 12 E. durans GM18; lane 13 E. durans GM17; lane 14 E. durans GM15; lane 15 E. durans GM14; lane 16 E. durans GM13; lane 17 E. faecium GM12

B

C

lowest concentration without visible growth was determined and the antibiotic activity distributions are presented in Table 2. All the isolates showed resistance to one or more of the antibiotics tested. High levels of tetracycline and kanamycin resistance were observed among the enterococci isolates. The screening test revealed that E. durans GM15 and Lc. lactis GM11 were resistant to both ampicillin and penicillin (b-lactam). Moreover, Lc. lactis GM11 and E. faecium GM12 showed multiple resistance profiles to three and four antibiotics, respectively. However and except for Lc. lactis GM11, all of the isolates were susceptible to gentamicin. Interestingly, only E. faecium GM12 showed resistance to vancomycin, an amino-

123


1

2

3

MM (Kda) 97.0 45.0 30.0

20.1

14.4

3.3

A

B

Fig. 5 Tricine-sodium dodecyl sulfate–polyacrylamide gel electrophoresis (Tricine-SDS-PAGE) of purified bacteriocin from GM18 strain and direct detection of its antimicrobial activity on the gel. A Coomassie blue-stained gel. B Bioassay of inhibitory activity on gel against Lb. bulgaricus 340. Lane 1 Low range molecular weight marker; lanes 2 and 3 purified bacteriocin after HPLC separation

glycoside antibiotic. This feature is, in fact, of special interest particularly for medical applications involving the use of this and similar agents.

Discussion The present study aimed the exploration of the intestinal microflora of wild Tunisian fishes and the investigation of their antimicrobial activities. It reports the isolation and subsequent characterization of new bacteriocinic strains from seven fish species captured along the Tunisian coastline using the well diffusion method. The isolates selected for the antagonism assays presented typical LAB characteristics (Axelsson 2004), with a predominance of Gram-positive cocci. Among the thirty-nine LAB isolates

123

that displayed antagonist properties, nine showed relatively high activity (GM11 to 15 and 17–20) and were, therefore, selected for further studies. The antimicrobial compounds from the supernatants of the nine isolates were sensitive to proteolytic enzymes, thus indicating their proteinaceous nature. Bacteriocin activities were stable at pH values ranging from 2 to 9 and temperature values of up to 100 °C for 15 min. Similar results were previously reported for enterocin EJ97 produced by E. faecalis (Galvez et al. 1998) and plantaricin ST31 produced by Lb. plantarum (Todorov et al. 1999). Bacteriocin production started at the midexponential phase and increased gradually with bacterial growth, reaching maximal yields after 28 h of incubation. It is worth noting in this context that bacteriocin production has often been reported to depend on biomass concentration. The isolates under investigation exhibited a broad spectrum of inhibitory activity that included several Listeria strains. The GM11 isolate showed a broader inhibitory spectrum than the other isolates. It also displayed important inhibitory activity against Saccharomyces cerevisiae and C. pseudotropicalis. Based on 16S rRNA sequencing, the bacteriocinogenic cocci strains were identified as Lc. lactis (strain GM11), E. faecium (strain GM12) and E. durans (strains GM13 to 20). It was observed that E. durans represented the most common species occurring in the gastro-intestinal tract (GIT) of the fish species investigated. To our knowledge, this is the first report of a bacteriocinogenic E. durans strain isolated from such a biotope. In fact, E. durans were previously mostly associated with poultry infections and described to constitute one of the several enterococci species sporadically reported in human clinical studies (Knijff et al. 2001). However, Batdorj et al. (2006) reported the isolation and characterization of E. durans strain used in the fabrication of fermented Mongolian mare’s milk. Considering that the E. durans isolates were obtained in our experiments from the 10-7 dilution of fish viscera, they appear to represent a dominant species with beneficial roles in the GIT of fish. In fact, this species has frequently been isolated from human feces but was less prevalent in livestock (Franz et al. 1999). Few studies have detected the prevalence of E. durans in aqueous environments (di Cesare et al. 2011). A previous study by de Oliveira et al. (2008) reported that only 3.8 % of E. durans were identified in seawater, while other strains were more represented, including E. faecalis (34.6 %) and E. faecium (26.9 %). Little is known about the digestive or protective roles of this species in fish GIT. Bacteriocin production may be advantageous to the producing strain, particularly for its establishment and maintenance in the GIT. Moreover, several bacterial diseases that build up in the GIT may disturb the normal flora, which when

Author's personal copy World J Microbiol Biotechnol Fig. 6 Electrospray-ionization mass spectrometry analysis of purified bacteriocins from E. durans GM18 with multiple charged molecular ions [M ? nH?]n. The molecular mass of Durancin GTM18 is indicated

outbalanced by pathogens may suffer from septicemia. Bacteriocinic LAB strains could hinder this disruption and colonization by undesirable bacteria. Accordingly, further research is needed to clarify the complexity of this ecosystem. The studied isolates were submitted to RAPD typing to investigate the diversity of the enterococci species isolated from fish viscera. All of the isolates were successfully genotyped, with four distinct profiles being recognized among the seven E. durans isolated strains, which indicated their high genetic variability. Strains with identical RAPD profiles were found in the same sample and in fish from different species. This could presumably be attributed to the predominance of a particular strain among the enterococci population in seawater. Hence, no fish-specific LAB species were identified. The DNA products generated from RAPD depend on the primer used, with different primers producing different banding patterns. OPL and M13 primers showed significant differences between strains. The Lc. lactis GM11 strain produced nisin, a highly valued bacteriocin for the long preservation of food products. Other lactococci strains have previously been reported to be significant bacteriocin producers (Ayad et al. 2002). Moreover and using the primers specific for several well-known enterocins, E. faecium GM12 strain was shown to yield PCR products corresponding to the genes of enterocins A, B and P, suggesting its production of bacteriocins identical to enterocins A, B and P. The co-production of enterocins A, B and P has already been reported in the literature (Sańchez et al. 2007; Strompfova´ et al. 2008). Furthermore, the structural genes of enterocins A and B were detected in E. durans GM19. In this respect, this study is the first to report on E. durans strains producing enterocins A and B. Previous studies

have reported on the detection of enterocin A and B genes in E. faecium and E. faecalis strains from different sources (Strompfova´ et al. 2008). The findings of the present study also confirmed that none of the six other E. durans strains (GM13, -14, -15, -17, -18 and -20) carried genes coding enterocins A, B or P. It was also observed that E. durans GM15, -18 and -20 were inhibited by enterocins A and B from the supernatant of the E. faecium MMT21 culture, indicating that those strains did not produce enterocins A and B. Accordingly, further studies are required to characterize these strains. The present work also aimed to establish the virulence traits and antibiotic resistance profile of the eight Enterococcus strains and Lc. lactis strain isolated from Tunisian fish viscera. The findings revealed that tetracycline-resistance was most prevalent among the strains. This is in accordance with the results previously reported on enterococci from aquaculture sites (Tamminem et al. 2011). Resistance to multiple antimicrobial agents was observed for the Lc. lactis GM11 and E. faecium GM12 strains. The latter was resistant to multiple antimicrobial agents, including beta-lactams, amino-glycosides and vancomycin. Vancomycin-resistant enterococci posed a particular challenge to clinicians since this antibiotic has traditionally been the ‘‘drug of last resort’’. On the other hand, E. durans strains were resistant to at least two antibiotics, with the exception of two isolates, namely E. durans GM14 and -19, which were resistant only to kanamycin and ampicillin, respectively. E. durans has not been considered as particularly pathogenic to humans. However, isolates from fish GIT exhibited more multi-drug resistance than isolates from food origin (Belicova´ et al. 2007). Considering that the ability to transfer resistance gene to other species was not

123


investigated in the isolates, further investigations are needed before their application as food biopreservatives. Furthermore, it was shown that only E. faecium GM2 harbored potential virulence factors. Clinical enterococci strains generally have more virulence determinants than food strains, which, in turn, may have more than starter strains (Eaton and Gasson 2001). The E. durans strains described in this study do not show any amplification by PCR of virulence factor genes. These strains could, therefore, be useful for future application as biopreservatives for the biocontrol of food, particularly seafood products. Nevertheless, further studies are required to confirm their candidacy for application in this domain. Enterococcus durans GM18 showed a highly interesting range of antibacterial and anti-fungal activities. This strain did not exhibit genes coding the most common enterocins. This was also supported by the fact that it was not immunized toward the cell-free supernatant of strain E. faecalis MMT21 producing enterocins A and B. The antimicrobial activity of the purified bacteriocin produced by strain GM18 was confined to a single protein band and its molecular weight was estimated below 3.3 kDa, which was quite low compared to the molecular weights of other bacteriocins such as enterocin A (4,828 Da) and enterocin B (5,463 Da). The exact molecular mass of the purified bacteriocin produced by E. durans GM18 was determined by ESI–MS as 6,316.89 ± 0.64 Da. Upon database screening (Bactibase or DAMPD), none of the currently available enterococci-produced bacteriocins was noted to display this molecular mass. Taken together, the results presented above provided strong support to assume that the GM18 strain produced a novel anti-Listeria bacteriocin designated as durancin GTM18. The durancin GTM18 was submitted to automate Edman degradation to determine its N-terminal amino acid sequence. The first residue was, however, blocked (data not shown), suggesting the presence of an unusual residue, a characteristic feature of lantibiotics. This hypothesis was further supported by the high migration of durancin GTM18 on tricine-SDS-PAGE. Hu et al. (2008) have recently reported on a new bacteriocin produced by E. durans, termed durancin TW-49 M, with high homology with enterocin B. Further studies to fully determine the primary structure of durancin GMT18 are currently underway in our laboratories. In conclusion, bacteriocinogenic strains of E. durans, E. faecium and Lc. lactis are common members of natural population of several fish GIT. These strains and their antimicrobial compounds have a number of attractive properties and attributes that might open new promising opportunities for the development of novel preservative agents that inhibit the growth of pathogens and spoilage microorganisms for application in seafood processing or aquaculture operations.

123

References Agerholm LL, Bell ML, Grunwald GK, Astrup A (2000) The effect of a probiotic milk product on plasma cholesterol: a meta-analysis of short-term intervention studies. Eur J Clin Nutr 54:856–860 Asfie M, Yoshijima T, Sugita H (2003) Characterization of the goldfish fecal microflora by the fluorescent in situ hybridization method. Fish Sci 69:21–26 Axelsson L (2004) Lactic acid bacteria: classification and physiology. In: Salminen S, Von Wrigtht A, Ouwehand A (eds) Lactic acid bacteria-microbiological and functional aspects. Marcel Dekker Inc, New York, pp 1–66 Ayad EHE, Verheul A, Wouters JTM, Smit G (2002) Antimicrobialproducing wild lactococci isolated from artisanal and non-dairy origins. Int Dairy J 12:145–150 Batdorj B, Dalgalarrondo M, Choiset Y, Pedroche J, Me´tro F, Pre´vost H, Chobert JM, Haertle´ T (2006) Purification and characterization of two bacteriocins produced by lactic acid bacteria isolated from Mongolian airag. J Appl Microbiol 101:837–848 Belicova´ A, Krzkova´ L, Krajcovic J, Jurkovic D, Sojka M, Ebringer L, Duinsky´ R (2007) Antimicrobial susceptibility of Enterococcus species isolated from Slovak Bryndza cheese. Folia Microbiol 52:115–119 Campos CA, Rodrıguez O, Calo-Mata P, Prado M, Barros-Velazquez J (2006) Preliminary characterization of bacteriocins from Lactococcus lactis, Enterococcus faecium and Enterococcus mundtii strains isolated from turbot (Psetta maxima). Int Food Res 39:356–364 Casaus P, Nilsen T, Cintas LM, Nes IF, Hernandez PE, Holo H (1997) Enterocin B, a new bacteriocin from Enterococcus faecium T136 which can act synergistically with enterocin A. Microbiology 143:2287–2294 Cintas LM, Casaus P, Havarstein LS, Hernandez PE, Nes IF (1997) Biochemical and genetic characterization of enterocin P, a novel secdependent bacteriocin from Enterococcus faecium P13 with a broad antimicrobial spectrum. Appl Environ Microbiol 63:4321–4330 Cintas LM, Casaus P, Holo H, Hernandez PE, Nes IF, Havarstein LS (1998) Enterocins L50A and L50B, two novel bacteriocins from Enterococcus faecium L50, are related to staphylococcal hemolysins. J Bacteriol 180:1988–1994 Cintas LM, Casaus P, Herranz C, Havarstein LS, Holo H, Hernandez PE (2000) Biochemical and genetic evidence that Enterococcus faecium L50 produces enterocins L50A and L50B, the secdependent enterocin P and a novel bacteriocin secreted without an N-terminal extension termed enterocin Q. J Bacteriol 182:6806–6814 Cotter PD, Hill C, Ross RP (2005) Bacterial antibiotics: strategies to improve therapeutic potential. Curr Protein Pept Sci 6:61–75 De Oliveira FC, Watanabe AJ, Pinhata JM (2008) Antimicrobial resistance and species composition of Enterococcus spp. isolated from waters and sands of marine recreational beaches in Southeastern Brazil. Water Res 42:2242–2250 De Vuyst L, Foulquie´-Moreno MR, Revets H (2003) Screening for enterocins and detection of hemolysin and vancomycin resistance in enterococci of different origins. Int J Food Microbiol 84:299–318 Di Cesare A, Vignaroli C, Luna GM, Pasquaroli S, Biavasco F (2011) Antibiotic-resistant enterococci in seawater and sediments from a coastal fish farm. Microb Drug Resist 18:502–509 Eaton TJ, Gasson MJ (2001) Molecular screening of Enterococcus virulence determinants and potential for genetic exchange between food and medical isolates. Appl Environ Microbiol 67:1628–1635 El-Ghaish S, Dalgalarrondo M, Choiset Y, Sitohy M, Ivanova I, Haertle´ T (2010) Screening of strains of lactococci isolated from

Author's personal copy World J Microbiol Biotechnol Egyptian dairy products for their proteolytic activity. Food Chem 120:758–764 Floriano B, Ruiz-Barba JL, Jimenez-Diaz R (1998) Purification and genetic characterization of enterocin I from Enterococcus faecium 6Tia, a novel antilisterial plasmid-encoded bacteriocin with does not belong to the pediocin family of bacteriocins. Appl Environ Microbiol 64:4883–4890 Franz CMAP, Holzapfel WH, Stiles ME (1999) Enterococci at the crossroads of food safety. Int J Food Microbiol 47:1–24 Galvez A, Valdivia E, Abriouel H (1998) Isolation and characterization of enterocins EJ97, a bacteriocin produced by Enterococcus faecalis EJ97. Arch Microbiol 171:59–65 Ghrairi T, Manai M, Berjeaud JM, Frere J (2004) Antilisterial activity of lactic acid bacteria isolated from rigout, a traditional Tunisian cheese. Appl J Microbiol 97:621–628 Hu CB, Zendo T, Nakayama J, Sonomoto K (2008) Description of durancin TW-49 M, a novel enterocin B-homologous bacteriocin in carrot-isolated Enterococcus durans QU 49. J Appl Microbiol 105:681–690 Huber I, Spanggaard B, Appel KF, Rossen L, Nielsen T, Gram L (2004) Phylogenetic analysis and in situ identification of the intestinal microbial community of rainbow trout (Oncorhynchus mykiss, Walbaum). J Appl Microbiol 96:117–132 Jennes W, Dicks LMT, Verwoerd DJ (2000) Enterocin 012, a bacteriocin produced by Enterococcus gallinarum isolated from the intestinal tract of ostrich. J Appl Microbiol 88:349–357 Klaenhammer TR (1993) Genetics of bacteriocins produced by lactic acid bacteria. FEMS Microbiol Rev 12:39–86 Knijff E, Dellaglio F, Lombardi A, Biesterveld S, Torriani S (2001) Development of the specific and random amplification (SARA)PCR for both species identification of enterococci and detection of the vanA gene. J Microbiol Meth 43:233–239 Leroi F (2009) Bacte´ries lactiques et applications alimentaires. Partie 2 : Les produits de la mer. In « Physiologie, Me´tabolisme, Geńomique et Applications Industrielles des Bacte´ries Lactiques » Djamel Drider et Herve´ Pre´vost coordonnateurs. E´ditions Economica. ISBN. 978-2-7178-5676-7 Makanera A, Arlet G, Gautier V, Manai M (2003) Molecular epidemiology and characterization of plasmid-encoded b-lactamases produced by Tunisian clinical isolates of Salmonella enterica Serotype Mbandaka resisitant to broad-spectrum cephalosporines. J Clinical Microbiol 41:2940–2945

Munõz-Atienza E, Landeta G, De las Rivas B, Go´mez-Sala B, Munõz R, Hernańdez E, Pablo M, Cintas L, Herranz C (2011) Phenotypic and genetic evaluations of biogenic amine production by lactic acid bacteria isolated from fish and fish products. Int J Food Microbiol 146:212–216 NCCLS (2003) Methods for antimicrobial susceptibility testing of anaerobic bacteria. Approved standard, 6th ed. Document M11A6. NCCLS, Wayne, PA Ringø E (2008) The ability of carnobacteria isolated from fish intestine to inhibit growth of fish pathogenic bacteria: a screening study. Aquac Res 39:171–180 Sabia C, Manicardi G, Messi P, de Niederhausern S, Bondi M (2002) Enterocin 416K1, an antilisterial bacteriocin produced by Enterococcus casseliflavus IM 416K1 isolated from Italian sausages. Int J Food Microbiol 75:163–170 Sańchez J, Basanta A, Go´mez-Sala B, Herranz C, Cintas LM, Hernańdez PE (2007) Antimicrobial and safety aspects and biotechnological potential of bacteriocinogenic Enterococci isolated from mallard ducks (Anas platyrhynchos). Int J Food Microbiol 117:295–305 Scha¨gger H, von Jagow G (1987) Tricine-sodium dodecyl sulfatepolyacrylamide gel electrophoresis for the separation of proteins in the range from 1 to 100 kDa. Ann Biochem 166:368–379 Schillinger U, Lu¨cke K (1987) Identification of lactobacilli from meat to meat products. Food Microbiol 4:199–208 Strompfova´ V, Laukova´ A, Simonova´ M, Marcina´kova´ M (2008) Occurrence of the structural enterocin A, P, B, L50B genes in enterococci of different origin. Vet Microbiol 132:293–301 Tagg JR, Dajani AS, Wannamaker LW (1976) Bacteriocins of grampositive bacteria. Bacteriol Rev 40:722–756 Tamminem M, Karkman A, Lohmus A, Muziasari WI, Takasu H, Wada S, Suzuki S, Virta M (2011) Tetracycline resistance genes persist at aquaculture farms in the absence of selection pressure. Environ Sci Technol 45:386–401 Todorov S, Onno B, Sorokin O, Chobert JM, Ivanova I, Dousset X (1999) Detection and characterization of a novel antibacterial substance produced by Lactobacillus plantarum ST31 isolated from sourdough. Int J Food Microbiol 48:167–177 Weisburg WG, Barns SM, Pelletier DA, Lane DJ (1991) 16S ribosomal DNA amplification for phylogenetic analysis. J Bacteriol 173:697–703

123