Enterococci (mostly Enterococcus faecalis and Enterococcus faecium) are natural inhabitants of the gastrointestinal tract of humans and animals,1 but are also found in other anatomical sites, including the vagina and oral cavity,2 and in the soil, water, plants, and food.3–5 Several reports have documented that the two most important species are among the leading causes of opportunistic human infections,6 including urinary tract infections,7 infections of the surgical site and burn wound infections,8,9 bacteremia and sepsis,10 endocarditis,11 cholecystitis,12 peritonitis,13 neonatal meningitis,14 and others.

The severe infections caused by Enterococcus spp. are difficult to treat due to the organism's capacity to survive in the hospital environment, their intrinsic resistance to many antimicrobials, their remarkable ability to acquire further resistance mechanisms toward strategic antibiotics during the treatment period and a variety of virulence factors.15

Virulence factors contribute to the pathogenesis of enterococcal infections through mediation of adhesion, colonization and invasion into the host tissues, modulation of the host immunity, and extracellular production of enzymes and toxins, which enhance the severity of the infection.16 Many virulence factors, such as the enterococcal surface protein (Esp), aggregation substance, capsule formation and gelatinase, are involved in bacterial adherence to host cells and/or in biofilm formation on abiotic surfaces in hospital environment.17–21 Biofilm production has an important role in the pathogenesis of enterococcal infections and also favors disease sustenance because of restricted penetration of antimicrobials.19 Invasion is usually facilitated by damage to host tissues and presence of virulence factors, including hyaluronidase and gelatinase, which assist in the advancement and further survival in newly infected places.21

Currently, our knowledge about the possible relationship between the presence of virulence factors and their implication in human enterococcal infections is still limited.

The aim of the present study was to examine the prevalence of genes encoding cell-associated and extracellular virulence factors in clinically relevant isolates of Enterococcus spp. and to assess the association of those genes with species, infection site, and patient population (inpatients or outpatients).

In this study, we found a significantly higher prevalence (p<0.001) of genes encoding some virulence factors involved in biofilm formation (the enterococcal surface protein (esp), aggregation substance (agg) and gelatinase (gelE)) in E. faecalis, compared to E. faecium. A previous study indicated higher prevalence of biofilm formation among E. faecalis strains compared to other enterococcal species.30E. faecalis has been recognized as the enterococcal species most commonly infecting indwelling medical devices.17,30 In a study conducted by Dupré et al. with clinically relevant enterococcal isolates from a hospital in Sardinia, Italy, in the period of 1998–2001 there was a complete absence of genes for gelatinase and aggregation substance in E. faecium, but contrary to our results, higher frequency of esp in E. faecium vs. E. faecalis was reported (71.9%/60%).23

The prevalence of efaA, eep, gelE, and cylA in E. faecalis isolates in our study was significantly higher than previously published data on clinical strains of E. faecalis from Brazil in 2004 – as it follows: 84.9%/58.9%; 95.7%/58.9%; 82.2%/45.3%; 64.9%/16.8% (p<0.001), while the gene frequency for agg and esp had similar rates in both studies (44.3%/36.8%, and 59.5%/57.9%).29 Detection of genes for adhesins (esp, agg, ace and efaA) in newer clinical E. faecalis isolates from Southern Brazil was more frequent than the rates found in our study (respectively, 68.4% in Brazil vs. 59.5% in our study, 57.9%/44.3%, 73.7%/61.1% and 100%/84.9%). In contrast, genes for gelE and cylA were more frequent in our isolates than among the Brazilian isolates (77.2%/82.2% and 54.4%/64.9%, respectively).31 A recent study of clinical isolates of E. faecalis in Italy showed a significantly lower frequency of cylA compared to our isolates (26.4%/64.9%, p<0.001), and close to our distribution of esp, agg, ace and efaA (52.7%/59.5%; 42.8%/44.3%; 67%/61.1%; 96.7%/84.9%).32

A multicenter study on the distribution of virulence determinants in fecal E. faecium isolates from patients in 13 hospitals from nine European countries showed total absence of asa1 and gelE, lower frequency of hyl in comparison to E. faecium isolates (17%/24.2%) of our study and significantly higher prevalence of esp (65% vs. 4.3%, p<0.001).26

The significantly higher prevalence of clinical E. faecalis isolates with multiple virulence determinants (over 4) compared to E. faecium, and the significant predominance of E. faecium isolates with one gene compared to E. faecalis, demonstrated in this study, was also recently found by Padmasini et al.33 According to another study of enteroccoci from clinical isolates and starter cultures, all strains of E. faecalis were characterized by a high number of virulence genes (between 6 and 11), while the E. faecium strains had no such gene determinants.25

In Italy, Creti et al. reported wider distribution of esp in non-invasive E. faecalis isolates (56.2%) compared to invasive isolates (38.4%).28 In our study esp was found in 33.3% of infections caused by E. faecalis, with bacteremia in the clinical course, while the frequency of esp in non-invasive infections varied from 54.3% (in GTIs) to 64.8% (in UTIs), which confirms the important role of Esp as a colonizing factor in UTIs.34 The frequency of esp among our UTI E. faecalis isolates (64.8%) was very similar to the rate reported by Creti et al. (66.7%). The findings for cylA in both studies were similar: 44.4% in invasive isolates from bacteremia in our data vs. 50% in the Italian study, as well as the results for cylA in non-invasive isolates – varying from 62.5% to 77.1% in our study and 62.5% in the Italian study.28

Recently Padilla et al. published their data on incidence of virulence genes amongst strains of E. faecalis isolated from UTIs and bacteremia in Chile, for the period of 2008–2009.35 The rates for the following genes from urine samples in our study were significantly higher than those reported in Chile: efaA (94.3%/62%, p<0.001), eep (98.9%/72%, p<0.001), gelE (83%/28%, p<0.001), cylA (62.5%/40%, p<0.01); while the presence of agg was significantly lower (43.2%/78%, p<0.001). Our isolates from bacteremia showed a significantly higher frequency of efaA compared to the Chilean isolates (66.7%/28%, p<0.001).

E. faecalis isolates from patients with UTIs and wound infections in the present study showed higher prevalence of numerous virulence genes, compared to similar isolates from Brazil, studied by Bittencourt de Marques and Suzart – esp (64.9% and 60.4% vs. 62% and 46.1%), agg (43.2% and 43.8%/38% and 38.5%), efaA (94.3% (p<0.001) and 72.9%/54% and 69.2%), eep (98.9% (p<0.001) and 93.8% (p<0.001)/58% and 53.8%), gelE (83% (p<0.001) and 81.3% (p<0.001)/50% and 34.6%) and cylA (62.5% (p<0.001) and 64.6% (p<0.001)/22% and 3.8%).29

An interesting finding from the present study was the higher dissemination rates of gene determinants for some virulence factors among Enterococcus spp. causing infections in outpatients, compared to those infecting inpatients. The review of published medical literature could not find a similar comparative study. Hospitalized patients may have a greater incidence of enterococcal infections not only because of virulence, but because the hospital itself is a risky hub. On the other hand, the intestinal tract is an important reservoir for opportunistic pathogens such as enterococci and allows them to access infected sites especially in outpatients through different virulence factors.21 A weakness of this comparative analysis results from the difficulties to obtain epidemiological data; often UTIs, occurring in inpatients, begin before admission to the hospital.

In the present study clinical E. faecalis isolates were characterized by a significantly higher frequency of genes encoding some adhesins and gelatinase in comparison to E. faecium isolates. The higher virulence potential of E. faecalis vs. E. faecium isolates was proven (on the basis of a significantly higher prevalence of E. faecalis isolates with multiple virulence factor genes vs. E. faecium and significant predominance of E. faecium isolates without or with only one gene compared to the E. faecalis).

The genes for some adhesins (esp and agg/asa1) were more prevalent in non-invasive Enterococcus spp. isolates than in invasive isolates from bacteremia, while ace/acm showed a significantly higher frequency among invasive isolates.

For the first time, we proved that Enterococcus spp. isolates from outpatients with symptomatic infections had significantly higher frequency of the following virulence genes: agg/asa1, eep, gelE and cylA, in comparison to isolates causing infections in inpatients.

To the best of our knowledge, this is the first study in Bulgaria and in the Balkans region of the genes associated cellular and extracellular virulence factors among clinically relevant Enterococcus spp. isolates, in which a comparative analysis with respect to different features (species of microorganism, infection site, and origin) was made. Our results may serve as a basis for additional research of the pathogenesis of infections caused by Enterococcus spp.

The authors declare no conflicts of interest.