Detection of bacteriocin production and virulence traits in vancomycin-resistant enterococci of different sources

Authors


Moreno Bondi, Dipartimento di Scienze Biomediche, Università di Modena e Reggio Emilia, via Campi, 287, 41100, Modena, Italy.
E-mail: bondi.moreno@unimore.it

Abstract

Aim:  Three hundred and two enterococci were isolated from food, animal and clinical samples in order to evaluate the incidence of vancomycin-resistant enterococci (VRE) and bacteriocin, cytolysin, haemolysin, gelatinase production.

Methods and Results:  Among the isolates, 27 (8·9%) were VRE, and 17 (63%) of these showed, by the deferred antagonism method, bacteriocin production against Gram-positive and some Gram-negative indicators. Eight bacteriocin producers displayed by polymerase chain reaction an enterocin structural gene: six Enterococcus faecium the Enterocin A, two Enterococcus faecalis the Enterocin P genes. The enterocins AS-48, 31, L50 and 1071A/B genes were not found. Regarding the virulence factors, two VRE produced gelatinase and seven were haemolytic. Gelatinase gelE gene was found in 19 strains and cytolysin cylLL gene in eight. Among the strains showing the cylLL gene, only two E. faecalis expressed a β-haemolysis.

Conclusions:  Our results showed the persistence of VRE in food, animal and clinical samples. Many of these strains displayed antibacterial activity and sometimes different components of virulence, which could emphasize their pathogenicity.

Significance and Impact of the Study:  This work indicates the need of a constant monitoring of enterococci in order to assess their possible pathogenic properties. The strains of interest in the food industry or used as probiotics should be tested for antibiotic resistance and virulence traits.

Introduction

Enterococcus is an important genus belonging to the group of lactic acid bacteria (LAB), which play a significant role in the ripening of traditional Mediterranean country products (sausages, cheeses etc.) (de Vuyst et al. 2003; Foulquié Moreno et al. 2006). Enterococci are generally recognized as safe (GRAS) in food production (Devriese and Pot 1995; Schillinger et al. 1996), and they show favourable metabolic activities (lipolytic and esterolytic activity, citrate utilization, etc.) contributing to the typical taste and flavour (Centeno et al. 1996; Giraffa and Carminati 1997;Manolopoulou et al. 2003). Many enterococci have also the capability to produce bacteriocins (enterocins) responsible for the inhibition of a narrow range of strains closely related to the producer (Tagg et al. 1976), and in some instances, endowed with a broad-spectrum of antibacterial activity (Klaenhammer 1988; Ennahar and Deschamps 2000). As enterococci are commonly present in many food systems and their technological and probiotic benefits are widely recognized (Giraffa and Carminati 1997; Franz et al. 1999), they could be good candidates for potential application of bacteriocin-mediated antagonism against spoilage bacteria and food-borne pathogens, such as Listeria monocytogenes and Clostridium spp. (Giraffa 1995; Muriana 1996; Sabia et al. 2003; García et al. 2004; Ananou et al. 2005; Muñoz et al. 2007).

On the other hand, the control of enterococci in food processing has assumed a different importance in the recent years, as the conviction of the harmlessness of these bacteria has been partly changed by the increase of their incidence in nosocomial infections. Actually, they have been implicated as an important cause of endocarditis, bacteraemia and urinary tract, central nervous system, intra-abdominal and pelvic infections (Foulquié Moreno et al. 2006). Enterococci do not possess the common virulence factors found in pathogenic bacteria, but they display other important biological characteristics. The presence of vancomycin-resistant enterococci (VRE), partly owing to the use of avoparcin as feed additive in animal husbandry, represents a significant threat to public health, as glycopeptides are considered to be the last-resort drugs in the treatment of enterococcal infections. Moreover, VRE are frequently resistant to many standard antibiotics commonly used to treat these illnesses (Landman and Quale 1997).

In addition to antibiotic resistance, bacteriocin/cytolysin and gelatinase may constitute components of virulence (Jett et al. 1994; Franz et al. 1999; Eaton and Gasson 2001). Gelatinase, e.g., an extracellular metallo-endopeptidase involved in the hydrolysis of gelatin, collagen, haemoglobin and other bioactive peptides, seems to contribute to enterococci pathogenesis in some animal models (Qin et al. 2000). At the same time, the antibacterial activity is assumed to be important in the regulation of population dynamics in bacterial ecosystems, favouring the bacteriocin-producing micro-organisms in the competition for the colonization of a particular ecological niche (Kalmokoff et al. 1999). Finally, the antibiotic resistance, the bacteriocin/cytolysin and other biological characteristics contributing to virulence in enterococci are frequently associated with pheromone-responsive conjugative plasmids (Martínez-Bueno et al. 1994; Booth et al. 1996; Tomita et al. 1996). Therefore, these traits have the potential of being transferred by conjugation, in particular, where the density of bacteria and the variety of species are high (Sengelov and Sorensen 1998; Lilley and Bailey 2002). For example, the sources of resistance genes for Listeria spp. actually appear to be enterococci, and mating experiments indicated that transposons and plasmids are responsible for the emergence of drug resistance in these micro-organisms (Leclercq et al. 1989; White et al. 2002; de Niederhäusern et al. 2004).

Considering the controversial aspect displayed by enterococci in these last years, the aim of the present work was to study the bacteriocin production and the presence of some virulence components, such as cytolysin, haemolysin and gelatinase in VRE isolated from food, animal and clinical samples. The expression of these biological traits could justify the higher virulence that some enterococci can show in nosocomial infections.

Materials and methods

Bacterial strains and sample collection

Enterococci were collected from clinical, food and animal samples (Table 1) in the period between March and November 2005. Clinical enterococci were kindly supplied by the Departments of Microbiology of Bolzano and Modena ‘S. Agostino’ hospitals (Italy), and were isolated from urine, wounds and stools.

Table 1.   Enterococci collection and vancomycin-resistant enterococci (VRE) isolates
IsolatesSourceVRE incidenceSpecies
  1. E., Enterococcus.

154 clinical isolates52 urine13 (8·4%)E. faecium (3)
E. faecalis (4)
E. casseliflavus (1)
11 wounds E. faecium (3)
91 stools E. faecium (1)
E. faecalis (1)
56 food isolates56 minced meat5 (8·9%)E. faecium (1)
E. faecalis (3)
E. casseliflavus (1)
92 animal isolates79 swine9 (9·8%)E. faecium (4)
E. casseliflavus (4)
13 equine E. faecium (1)

Food enterococci were collected from samples of raw minced meat of beef and pork. The samples were bought from different supermarkets of our region (Emilia Romagna, Italy), stored at +4°C and transferred to the laboratory. A 25-g portion of each sample was placed in sterile plastic bags with 225-ml buffered peptone water added (Oxoid, Milan, Italy) and then homogenized for 1 min in Stomacher (Lab Blender, Seward, London, UK).

Animal samples were collected from equine and swine rectal swabs. In particular, the equine samples originate from Italian, Polish and Hungarian horse breeding, and all the swine samples were from pig breeding of our region. Rectal swabs were stirred on a vortex mixer (Vortex Genie II Model G-560; Scientific Industries, Bohemia, NY, USA) for 1 min to release the cells into 10 ml of Ringer solution, which was then serially diluted and plated on selective agar.

Enterococci were determined by streaking with a 10-μl loop serially diluted samples on Kennel Fecal (KF)-Streptococcus agar (Difco Laboratories, Detroit, MI, USA). The plates were incubated for 24 h at 37°C aerobically. For each sample, a red colony with the typical enterococcal morphology was isolated and presumptively identified to genus level by Gram’s stain, catalase test and bile-esculin reaction.

Food, animal and clinical isolates were successively subcultured on KF-streptococcus agar added with vancomycin (4 μg ml−1), and the grown enterococci were identified to species level on the basis of their biochemical properties (API 20 Strep; bioMérieux, Marcy l’Etoile, France). Species identification was confirmed by polymerase chain reaction (PCR) as described by Dutka-Malen et al. (1995).

Minimum inhibitory concentration (MIC)

The MIC of vancomycin and teicoplanin (1–128 μg ml−1) was determined for all VRE isolated according to standard procedures recommended by the National Committee for Clinical Laboratory Standard (NCCLS 2000). Mueller-Hinton agar (Oxoid), supplemented with 5% of sheep blood, was used as test medium.

Plasmid DNA analysis

Small-scale preparations of plasmid DNA were performed by the Rapid Mini-Prep procedure of O’ Sullivan and Klaenhammer (1993) on the strains that showed glycopeptide resistance. DNA plasmid was analysed in 0·7% agarose gel electrophoresis at 3·5 V cm−1 for 8 h in a Tris-borate-EDTA buffer. Purified plasmids of Escherichia coli V517 (Macrina et al. 1978) were used as a size reference for molecular weight determinations.

vanA and vanB gene amplification

Purified DNA plasmid from VanA and total DNA from VanB enterococci were subjected to amplification assays employing the specific oligonucleotide primers A1 (5′-GGGAAAACGACAATTGC-3′), A2 (5′-GTACAATGCGGCCGTTA-3′) and B1 (5′-ATGGGAAGCCGATAGTC-3′), B2 (5′- GATTTCGTTCCTCGACC-3′) and testing conditions as described by Dutka-Malen et al. (1995). The PCR products were resolved by electrophoresis on a 1% agarose-Tris-borate-EDTA gel containing 5 μg ml−1 of ethidium bromide.

Antibacterial activity evaluation

The strains identified as VRE were cultured in tryptic soy broth (TSB; Oxoid) and incubated at 37°C for 24 h. Bacteriocin production was screened in tryptic soy agar (TSA; Oxoid) by the deferred antagonism method (Kekessy and Piguet 1970), using as indicators the same strains and other Gram-positive and Gram-negative bacteria: L. monocytogenes, Listeria ivanovii, Aeromonas hydrophila, Pseudomonas putida, Bulkholderia cepacia, Enterobacteriaceae (E. coli, Serratia marcescens, Proteus mirabilis, Klebsiella pneumoniae and Enterobacter agglomerans). To eliminate inhibition owing to hydrogen peroxide production, a first incubation was performed anaerobically.

Haemolysis and gelatinase activity assays

Vancomycin-resistant enterococci were cultured in TSB and incubated at 37°C for 24 h. Cultures were spotted onto blood agar base plates containing 5% of defibrinated horse blood (Oxoid) and incubated overnight at 37°C. The haemolytic activity was determined by observing a clear zone of haemolysis (β-haemolysis), a partial and greening haemolysis zone (α-haemolysis) or no activity (γ-haemolysis) around the spots.

Gelatinase production was evaluated using Todd–Hewitt agar plates (Oxoid) containing 3% gelatin. After overnight incubation at 37°C, colonies that had opaque zones around them were considered positive.

PCR detection of cytolysin, gelatinase and enterocin structural genes

The total DNA of VRE, isolated by the rapid alkaline method, as described by Dutka-Malen et al. (1990), was subjected to amplification assay for the cylLL and gelE genes. The DNA of bacteriocin producers was subjected to amplification assay for the structural genes of Enterocin 1071 A/B, Enterocin 31, Enterocin AS-48, Enterocin L50, Enterocin A and Enterocin P employing the specific oligonucleotide primers listed in Table 2.

Table 2.   Polymerase chain reaction primers and their products
PrimersSequenceProduct size (bp)
1071 A/B for5′-ATATTTAGGGGGACCGATAA-3′405
1071 A/B rev5′-TATACATTCTTCCACTTATTTTT-3′
31 for5′-GAAAAAGAAATTAGTTATTTGTG-3′202
31 rev5′-CTATCTAGGAGCCCAAGG-3′
AS-48 for5′-GAGGAGGAGTATCATGGTTAAAG-3′340
AS-48 rev5′-CATATTGTTAAATTACCAAGCAATAA-3′
L50 for5′-GGGAGCAATCGCAAAATTAG-3′96
L50 rev5′-ATTGCCCATCCTTCTCCAAT-3′
A for5′-ATGAAACATTTAAAAATTTTGTCTA-3′198
A rev5′-TTAGCACTTCCCTGGAATTG-3′
P for5′-GTTAGTTGTGACAAATTTTGG-3′382
P rev5′-GGTGTGGAAAAGCCGTTTC-3′
cylLL for5′-GATGGAGGGTAAGAATTATGG-3′252
cylLL rev5′-GCTTCACCTCACTAAGTTTTATAG-3′
gelE for5′-ACCCCGTATCATTGGTTT-3′419
gelE rev5′-ACGCATTGCTTTTCCATC-3′

Polymerase chain reaction was performed on a DNA Thermal Cycler (Hy-Baid CELBIO, Milan, Italy) in a final volume of 70 μl containing the following: 250 ng of DNA plasmid as template; 50 pmol of each oligodeoxynucleotide primer; 500 μmol l−1 (each) dATP, dCTP, dGTP and dTTP; Tris 10 m mol l−1 pH 8·3, KCl 50 mmol l−1, MgCl2 1·5 m mol l−1 and 2 U of Taq DNA polymerase. The cycles used were 94°C for 2 min for the first cycle; denaturation at 94°C for 1 min, annealing at appropriate temperature for 1 min, and extension at 72°C for 1 min, for the next 25 cycles; and 72°C for 10 min for the last cycle. PCR products were resolved by electrophoresis on a 1% agarose-Tris-borate-EDTA gel containing 5 μg ml−1 of ethidium bromide.

Results

VRE isolation

A total of 302 enterococci were isolated from clinical (154), food (56) and animal (92) samples, and 13 (8·4%), five (8·9%) and nine (9·8%), respectively, showed different levels of glycopeptide resistance and were considered VRE.

Among these, the most frequently isolated species was Enterococcus faecium (13), followed by Enterococcus faecalis (eight) and Enterococcus casseliflavus (six) (Table 1). The preliminary identification of VRE isolates to species level by their biochemical properties fitted well with the PCR results. Indeed, only one strain showed a different identification.

MIC

On the basis of the different levels of glycopeptide resistance (Table 3), the VRE were presumptively ascribed to three different phenotypes: 21 were VanA, characterized by high-level resistance to vancomycin (MIC >128 μg ml−1) and teicoplanin (MIC ≥32 μg ml−1), four were VanB with lower level of acquired inducible resistance to vancomycin (MIC ≥64 μg ml−1), but not to teicoplanin (MIC ≤8 μg ml−1) and two were VanC with low-level resistance only to vancomycin (MIC =16 μg ml−1). In particular, all animal isolates belonged to the VanA phenotype, while among the clinical isolates, we ascribed eight strains to VanA, four to VanB and one to VanC phenotype. Finally, the food isolates were four VanA and one VanC.

Table 3.   Minimum inhibitory concentration (MIC) against vancomycin and teicoplanin of clinical, food and animal vancomycin-resistant enterococci isolates
 Vancomycin Break point ≥32 μg ml−1Teicoplanin Break point ≥32 μg ml−1Phenotype
  1. E., Enterococcus.

Clinical isolates
E. faecium B03>128>128VanA
E. faecalis B04>128>128VanA
E. faecium B06>128>128VanA
E. faecium B07>128>128VanA
E. faecium B08>128>128VanA
E. faecalis M09>128>128VanA
E. faecalis B10>12864VanA
E. casselifavus M11162VanC
E. faecium B12>128>128VanA
E. faecium B13>1288VanB
E. faecalis B14641VanB
E. faecium B15>1284VanB
E. faecalis B16642VanB
Food isolates
E. faecalis C1>128>128VanA
E. faecalis C2>128>128VanA
E. faecium C3>128>128VanA
E. faecalis C4>12864VanA
E. casseliflavus C516<1VanC
Animal isolates
E. faecium A2>128>128VanA
E. casselifavus A3>128>128VanA
E. faecium A4>128>128VanA
E. casselifavus A6>12832VanA
E. faecium A8>128>128VanA
E. casseliflavus A15>128>128VanA
E. faecium A16>128>128VanA
E. casselifavus A18>128>128VanA
E. faecium A102>128>128VanA

vanA and vanB gene amplification

Figures 1 and 2 show an example of 732- and 635-bp fragment amplification in DNA of VanA and VanB isolates, respectively. For all strains, the phenotype emerged by the MIC evaluation was confirmed as vanA or vanB genotype by PCR amplification.

Figure 1.

 Amplification of the 732-bp fragment in plasmid DNA of some VanA isolates. Lane 1: 100-bp DNA ladder; lane 2: Enterococcus faecium B03 clinical isolate; lane 3: Enterococcus faecalis C1 food isolate; lanes 4 and 5: E. faecium A2 and Enterococcus casseliflavus A3 animal isolates; lane 6: negative control.

Figure 2.

 Amplification of the 635-bp fragment in DNA of VanB isolates. Lane 1: 100-bp DNA ladder; lanes 2–5: Enterococcus faecium B13, Enterococcus faecalis B14, E. faecium B15, E. faecalis B16 clinical isolates; lane 6: negative control.

Considering that PCR in VanA strains was performed using purified plasmid DNA as template, the results obtained demonstrated the extrachromosomal location of the VanA glycopeptide-resistance determinants.

Plasmid DNA analysis

All VRE, independently of their origin, had more than one plasmid with different molecular weights. No correlation was observed between plasmids and sample matrixes. Figure 3 shows an example of the typical plasmid profiles of clinical, animal and meat VRE isolates.

Figure 3.

 Example of the typical plasmid profiles of clinical, animal and meat vancomycin-resistant enterococci isolates. Lane 1: molecular size markers; lanes 2 and 3: clinical isolates; lane 4: food isolate; lane 5: animal isolate.

Among the observed plasmids, those ranging from 19 and 36 MDa were the most interesting, as the transposon Tn1546, responsible for the transferability of vanA genes, is generally harboured in a plasmid with a similar molecular weight (Leclercq et al. 1988; Chang et al. 2003).

Antibacterial activity evaluation

Seventeen (63%) of the 27 VRE, screened by the deferred antagonism method for bacteriocin production, showed antibacterial activity against one or more Gram-positive indicators, including L. monocytogenes. The activity was also observed against Gram-negative bacteria, such as P. putida, A. hydrophila and some Enterobacteriaceae (Table 4).

Table 4.   Antibacterial activity, haemolytic and gelatinase expression, bacteriocin and gelatinase structural genes, in vancomycin-resistant enterococci
 E. faecium (14)E. faecalis (8)E. cassel* (5)L. mon† (4)L. ivan‡ (2)A. hydr§ (3)P. put¶ (2)B. cep** (2) Ent†† (48) Hly‡‡Bacteriocin structural genesGelatinase expression/structural genes
  1. E., Enterococcus.

  2. *Enterococcus casseliflavus.

  3. Listeria monocytogenes.

  4. Listeria ivanovii.

  5. §Aeromonas hydrophila.

  6. Pseudomonas putida.

  7. **Bulkholderia cepacia.

  8. ††Enterobacteriaceae.

  9. ‡‡Haemolysis.

Clinical isolates
E. faecium B03222  11  αAgelE
E. faecalis B0496442 2  γ gelE
E. faecium B0612       αAgelE
E. faecium B07         γcylLLgelE
E. faecium B08421      αA 
E. faecalis M09 3 2     γcylLLGel/gelE
E. faecalis B10108342    βcylLLgelE
E. casseliflavus M11         α  
E. faecium B128724222  γcylLL 
E. faecium B13         γcylLLgelE
E. faecalis B1413554232  γ gelE
E. faecium B15         γcylLLgelE
E. faecalis B1614754232 6γ gelE
Food isolates
E. faecalis C126142 2  βcylLLgelE
E. faecalis C2 6 22 2  γPgelE
E. faecium C31134  1  α  
E. faecalis C41114  1  γPgelE
E. casseliflavus C511234222  γ gelE
Animal isolates
E. faecium A2   42 1  γA, cylLLgelE
E. casseliflavus A3         γ  
E. faecium A4         γ gelE
E. casseliflavus A6         γ  
E. faecium A81145423  11γAgelE
E. casseliflavus A15         γ gelE
E. faecium A16         γ Gel/gelE
E. casseliflavus A18         γ  
E. faecium A102117342 1  γA 

Bacteriocin production seemed to be correlated to the species: all E. faecalis were producers unlike E. faecium that showed 61·5% of producers and E. casseliflavus that showed only one producer (16·6%). No correlation was observed between bacteriocin production and resistance phenotype.

Haemolysis and gelatinase activity assays

We observed two different haemolytic reactions (Table 4). A β-haemolysis under aerobic condition, which increased in anaerobiosis, was observed for E. faecalis B10 and C1. An α-haemolysis under aerobic condition that became a slight, but distinct β-haemolysis under anaerobiosis, indicating an oxygen regulatory effect, was observed for E. faecium B03, B06, B08, C3 and E. casseliflavus M11. All the remaining strains were nonhaemolytic. Regarding the gelatinase production, only two strains, E. faecalis M09 and E. faecium A16, expressed this biological character.

PCR detection of cytolysin, gelatinase and enterocin structural genes

The PCR results are showed in Table 4. Cytolysin cylLL gene was present in eight (29%) VRE, six of clinical source, one of animal and one of food origin, but only E. faecalis B10 and C1, clinical and food isolates, respectively, expressed the cytolytic activity. Gelatinase gelE gene was found in 19 (70%) strains, 10 of clinical, four of food and five of animal origin.

Among the bacteriocin producers, eight (47%) showed an enterocin structural gene different from cylLL gene using the primers listed in Table 2. Enterococcus faecium B03, B06 and B08 clinical isolates, and E. faecium A2, A8 and A102 of animal origin presented the structural gene for Enterocin A. Only two food isolates, E. faecalis C2 and C4, contained the Enterocin P gene. The genes for the Enterocin AS-48, Enterocin 31, Enterocin L50 and Enterocin 1071A/B were not found in any bacteriocin producer strain.

Discussion

Enterococcus spp. is the most controversial genus of LAB group. For their metabolic and bacteriocinogenic capability, enterococci are often associated with fermented foods, but the selection of strains to be involved in food processes presents remarkable problems owing to their potential risk in human health (Klein et al. 1998; Vancanneyt et al. 2002; Peters et al. 2003; de Vuyst et al. 2003). The control of enterococci in foods is important for the whole community, but especially for those consumers who are at highest risk as the elderly, pregnant, hospitalized and immunocompromised persons. This situation is made worse by the increasing degree of antibiotic resistances shown by enterococci, especially to vancomycin, which leaves few options for medical treatment. In European countries, the community reservoir of VRE has been related to the use of avoparcin as growth promoter in animal husbandry, a veterinary practice banned since April 1997.

Bacteriocin production also can present a double face, because strains that produce bacteriocins seem to have an ecological advantage when compared with other nonproducing bacteria which inhabit the same ecosystem (Poeta et al. 2006) or which concur for colonization and for invasion in a particular ecological niche. Besides, enterococci are good acidifiers in the presence of available sugars, and the inhibitory effect owing to acid production could be considered an advantage in addition to bacteriocin production, and is equally responsible for survival and colonization.

In this work, we isolated 302 enterococci, 27 (8·9%) were found resistant to glycopeptides and among these, the VanA phenotype (78%) was clearly the most represented. The incidence of VRE did not show evident correlations with the different origin of the strains, and the rates of food (8·9%), animal (9·8%) and clinical (8·4%) VRE isolates were similar to those reported by Pantosti et al. (1999), Aarestrup et al. (2000) and Bouchillon et al. (2004), respectively.

In our VanA strains, as observed by PCR, the vanA cluster was localized in the extrachromosomal DNA. This result is important because, according to many authors, the spread of VanA resistance, which reaches the human enterococcal flora, is carried on the Tn1546 transposon, frequently integrated into conjugative plasmids across a variety of different strains (Witte 2001). As the vanB genes also can be linked to a mobile genetic element which transpose into conjugative plasmids, both phenotypes could facilitate the spread of glycopeptide resistance to more pathogenic micro-organisms (Woodford et al. 1995) even in the absence of selective pressure.

Our data showed a high rate (63%) of VRE endowed with antibacterial activity not uniformly distributed among the different species; in particular, the bacteriocin producer strains were mainly found among E. faecalis. These results, in agreement with those already reported by del Campo et al. (2001) and de Vuyst et al. (2003), could justify the high incidence of E. faecalis in nosocomial infections (Kayser 2003; de Vuyst et al. 2003; Cookson et al. 2006).

In addition to bacteriocin production, our VRE displayed other biological traits, such as haemolysis and gelatinase expression, which are considered virulence factors in animal models (Ike et al. 1984; Jett et al. 1992; Chow et al. 1993; Singh et al. 1998; Qin et al. 2000; Foulquié Moreno et al. 2006), even if their role in pathogenicity is not always demonstrated (Archimbaud et al. 2002). The molecular analysis performed for cytolysin and gelatinase evaluation showed, in many cases, the lack of the phenotypic expression also in the presence of the cylLL and gelE determinants, and these silent genes were equally distributed in strains of different sources. Therefore, as already reported (Eaton and Gasson 2001; Semedo et al. 2003; Lopes et al. 2006), the presence of cytolysin and gelatinase genes does not imply the capability to express the correlate phenotype. Eaton and Gasson (2001) reported that the lack of phenotypic activity may be explained by low levels or down-regulation of gene expression or by an inactive gene product. As environmental factors could strongly influence gene expression (Finlay and Falkow 1997), it is necessary to take into account that the ‘in vitro’ conditions used to test the phenotypic characters are different from those found in the human host. For these reasons, enterococci should be evaluated not only for the ‘in vitro’ expression of virulence traits, but also using molecular assays which can put in evidence silent genes which could be activated changing these bacteria into potential pathogens or emphasizing their pathogenicity.

In conclusion, our results, limited to the number of samples analysed, are in agreement with other authors that report similar rates of VRE isolation (Klein et al. 1998; Pantosti et al. 1999; Aarestrup et al. 2000; Bouchillon et al. 2004), and seem to confirm the persistence of a low incidence of VRE in food, animal and clinical samples, after 8 years from the avoparcin ban.

Many of our VRE showed a broad spectrum of antibacterial activity, sometimes associated with further components of virulence which can enhance their survival in humans and make them effective opportunists in nosocomial infections (Giraffa 2002; Kayser 2003; Inoue et al. 2006).

Therefore, this work indicates the need of a constant monitoring of enterococci in animals and in meat products, considering their importance as vectors for the transfer of the micro-organisms via food chain to humans. Moreover, as the demands for traditional fermented foods are in continuous increase, strains proposed for food biotechnological applications should be carefully tested for the presence of antibiotic resistance and virulence traits.

Ancillary