HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via CrossRef Citing Articles via Google Scholar Google Scholar Articles by Creti, R. Articles by Baldassarri, L. Search for Related Content PubMed PubMed Citation Articles by Creti, R. Articles by Baldassarri, L. Agricola Articles by Creti, R. Articles by Baldassarri, L.

Survey for virulence determinants among Enterococcus faecalis isolated from different sources

^1,2Laboratorio di Batteriologia¹ and Laboratorio di Ultrastrutture², Istituto Superiore di Sanità, Viale Regina Elena, 299-00161 Rome, Italy ³Dipartimento di Medicina Clinica, Università ‘La Sapienza', P. le Aldo Moro, 5-00185 Rome, Italy

Correspondence Roberta Creti roberta.creti{at}iss.it

Received June 19, 2003
Accepted October 20, 2003

A collection of Enterococcus faecalis strains from clinical isolates, healthy individuals and the environment was screened for the presence of virulence factor genes, such as those for collagen-binding protein (ace), endocarditis antigen (efaA), haemolysin activator (cylA), gelatinase (gelE), aggregation substances (asa1 and asa373), a surface protein (esp) and two novel putative surface antigens (EF0591 and EF3314). Apart from some genes that were present in all strains (ace, efaA and EF3314), the gelE gene was the most common factor, although its presence did not correlate with its expression. The genes that encode Esp and CylA were never detected in endocarditis isolates, whereas an association was noted between the esp gene and isolates from urinary tract infection (UTI) and bacteraemia. An aggregation substance gene was always present in commensal strains. As for gelatinase, the presence of the cylA and asa genes did not correlate completely with their phenotypic expression. Generally, isolates from endocarditis, biliary stents and the environment were equipped with fewer virulence factors than isolates from other sources. UTI strains possessed the highest number of factors.

	INTRODUCTION

TOP INTRODUCTION METHODS RESULTS AND DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
Enterococci are commensals of the intestinal tract of humans and animals and have emerged in recent decades as a major cause of nosocomial infections (Murray & Weinstock, 1999). Enterococcal infections represent the second most common cause of bacteraemia and endocarditis in US hospitals, whereas in Europe, epidemiological evaluations have been less comprehensive (Bonten et al., 2001). In Italian hospitals where active surveillance is operative, enterococci are the third most common cause of infections and cause mainly urinary tract infections (UTIs) and blood infections (Moro et al., 2001).

Mechanisms of acquisition of antibiotic resistance and spread have been well studied (Gilmore, 2002; Shepard & Gilmore, 2002), but enterococcal virulence and pathogenic mechanisms are still largely unknown. Previous studies have analysed the distribution of factors such as haemolysin, gelatinase, aggregation factor or Esp surface protein, although none has been found to be associated significantly with isolates from infection rather than with faecal strains, or with mortality in patients with bacteraemia (Coque et al., 1995; Huycke & Gilmore, 1995; Elsner et al., 2000; Eaton & Gasson, 2001; Archimbaud et al., 2002; Vergis et al., 2002; Waar et al., 2002). Little information is available on the distribution of virulence factor patterns.

The purpose of the present report was to perform a molecular epidemiological survey by investigating the presence of known and novel potential virulence factors in Enterococcus faecalis isolated from different sources, as well as to study possible correlations between potential virulence factors possessed by strains and their source of isolation.

	METHODS

TOP INTRODUCTION METHODS RESULTS AND DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
Bacterial strains and growth conditions.

Seventy-four E. faecalis strains, isolated from invasive (26 strains) and non-invasive (32 strains) hospital infections, as well as from the environment (six strains) and the faeces or throat of healthy individuals (10 strains), were analysed. Invasive infections included endocarditis (nine strains), sepsis (11 strains), endovascular infections (four strains), ascites (one strain) and meningitis (one strain). Isolates from UTIs (n = 12), blocked biliary stents (n = 12), surgical wounds (n = 2), vaginal infections (n = 2), eye infections (n = 1), pneumonia (n = 1) and orthopaedic prostheses (n = 2) were considered to be from non-invasive infections. Strains considered in this study had been characterized previously for biofilm formation, antibiotic susceptibility and PFGE profile (Baldassarri et al., 2001b; Dicuonzo et al., 2001).

Bacteria were grown in trypticase soy broth or agar (TSA) or in Todd–Hewitt broth (THB) at 37 °C in 5 % CO₂ without agitation.

PCR.

Enterococcal DNA was prepared by suspending a loop of overnight colonies in a tube that contained 500 µl sterile distilled water, boiling for 10 min and then centrifuging at 14 000 g for 5 min. An aliquot of the supernatant (5 µl) was used as the template in a final volume of 25 µl PCR mixture, which contained: 1x PCR buffer, 2 mM MgCl₂, 200 µM each dNTP, 400 nM each primer and 0.25 U Taq DNA polymerase (Life Technologies).

Samples were amplified on a DNA thermal cycler (MJ Research) by heating for 5 min at 95 °C, followed by 30 cycles of 95 °C for 60 s, 58 °C for 60 s (52 °C for gelE and 63 °C for esp, as indicated by Shankar et al., 1999) and 72 °C for 60 s, and a final step of 72 °C for 10 min.

PCR products were analysed by gel electrophoresis in 0.8 % (w/v) agarose gel (Life Technologies).

Primers utilized.

Oligonucleotides were synthesized by a custom primer service (Life Technologies) and are described in Table 1.

Phenotypic assays.

Production of gelatinase was determined by using TSA supplemented with 1.5 % skimmed milk; a clear halo around colonies after 18 h at 37 °C was considered to be a positive result (Coque et al., 1995).

Haemolysin production was evaluated on Columbia agar base supplemented with 5 % (v/v) fresh human blood. Zones of clearing around colonies after 24 h at 37 °C indicated production of ß-haemolysin (Franz et al., 2001).

Expression of aggregation substance was determined for PCR-positive strains by the clumping assay (Dunny et al., 1979) in the presence of sex pheromone, obtained by growing the pheromone-producing E. faecalis strain JH2-2 in THB at 37 °C for 18 h. Strains to be tested for aggregation substance expression were grown in THB for 18 h at 37 °C; 20 µl of each culture was used to inoculate 96-well microtitre plates that contained 200 µl pheromone preparation per well, which was diluted serially in THB. Cell clumping was checked at 2, 4, 6 and 18 h. E. faecalis OG1XpAD1 and JH2-2 were used as positive and negative controls, respectively.

	RESULTS AND DISCUSSION

TOP INTRODUCTION METHODS RESULTS AND DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
The presence of genes that encode aggregation substance (asa1 and asa373), cytolysin activator (cylA), surface protein (esp), gelatinase (gelE), collagen-binding protein (ace) and endocarditis antigen (efaA) was investigated by PCR. The frequency of two novel putative surface antigen genes (EF0591 and EF3314), detected by screening of the E. faecalis strain V583 genome (Paulsen et al., 2003), available at the website of The Institute for Genomic Research (http://www.tigr.org/), were also investigated.

In addition, E. faecalis-specific 16S rRNA gene primers were included as a control when the biochemical identification for E. faecalis was doubtful (Baldassarri et al., 2001b).

Gelatinase, haemolysin and aggregation substance production was tested by phenotypic assays. Individual virulence determinant content of each strain, incidence of virulence factors and their distribution among different classes are reported in Table 2, Table 3 and Table 4, respectively.

Genes that encode Ace and EfaA were found in all strains. For efaA, this is in accordance with a previous study in which the gene was always present in medical E. faecalis isolates, whereas the majority (89 %) of E. faecalis strains from food possessed the efaA determinant (Eaton & Gasson, 2001). Both the efaA and ace genes display allelic sequence variation (with the latter being more variable), suggesting that these variations can influence the ability of surface characteristics among strains of diverse sources. Due to their sequence variability, these genes have recently been used as possible markers for multilocus sequence typing of E. faecalis (Nallapareddy et al., 2002).

The esp gene was present in 44.6 % of isolates. It was associated more frequently with non-invasive infections (56.2 %, vs 38.4 % of invasive infection isolates) and particularly with UTIs, in accordance with other studies that indicated a possible role of Esp as a colonization factor in UTI (Shankar et al., 2001). Among invasive infections, esp was present in most isolates from bacteraemia (72.7 %), but it was never detected in isolates from endocarditis in this study. This is in contrast with the findings of Archimbaud et al. (2002), who reported the presence of esp in all endocarditis and bacteraemia isolates tested (ten strains in each group).

Even if esp is considered to be an infection-associated virulence factor (Shankar et al., 1999), we detected the gene in 20 % of isolates from healthy individuals and in 50 % of environmental strains.

During this study, we also identified esp-positive strains among Enterococcus faecium isolates (Baldassarri et al., 2001a) that had been identified erroneously as E. faecalis. This gene is an esp variant (esp_fm), the sequence of which is in accordance with the E. faecium esp gene sequence determined by others (Willems et al., 2001; Woodford et al., 2001).

The cylA gene was present in 23 % of all isolates and was distributed equally among healthy individuals and those with invasive and non-invasive infections. It was never detected in isolates from endocarditis, biliary stents or the environment. Its absence or low prevalence in endocarditis strains has already been reported by other authors (Huycke & Gilmore, 1995; Archimbaud et al., 2002). Haemolytic activity was detected in 64.7 % of cylA-positive strains, with a tendency to be present more often among non-invasive (62.5 %) and commensal (100 %) strains than in invasive strains (50 %). The lack of cytolysin phenotypic/genotypic congruence may suggest the occurrence of missing genes in the cyl operon among cylA-positive/haemolysin-negative strains.

The aggregation substance gene asa1 was present in 63.5 % of all isolates. It was associated more frequently with non-invasive infections (68.7 %, vs 53.8 % of invasive strains) and present in almost all strains that were derived from healthy individuals. Also, one-third of isolates from the environment were positive for the presence of the asa1 gene. Previous studies of the incidence of asa1 in enterococcal isolates are contradictory: some studies indicated a high prevalence of asa1 in clinical isolates compared to strains from healthy individuals (Coque et al., 1995; Waar et al., 2002), whereas in others, asa1 did not occur more frequently in invasive than in commensal strains (Huycke & Gilmore, 1995; Archimbaud et al., 2002).

To our knowledge, this is the second study that investigates the occurrence of the aggregation substance gene asa373 in clinical and commensal isolates. In the previous study (Waar et al., 2002), incidence of this gene was lower (5 and 6 % of isolates from blood cultures and faeces of healthy individuals, respectively) than ours (18.2 and 10 %, respectively). Waar et al. (2002) speculated about possible linkage between the asa1, asa373 and esp genes, as they were co-present in all isolates tested. Also, in another study of the incidence of the asa373 gene among enterococci isolated mostly from cheeses, Franz et al. (2001) found that the gene always occurred when the asa1 gene was present. We found no such correlation: nine strains (12.2 %) were positive for asa373 and the gene was associated with asa1 and esp in six and five strains, respectively. The three genes were co-present in three strains.

Interestingly, we determined the sequence of our asa373 amplicons and found that they displayed 72 % deduced amino acid similarity with the Asa373 protein (Muscholl-Silberhorn, 1999; De Boever et al., 2000). Further studies on the nature of the plasmid that bears this asa373 gene variant are ongoing.

By phenotypic assay (Table 3), 66 % of the pAD1 and/or pAM373 PCR-positive strains gave a positive clumping reaction in the presence of sex pheromone produced by E. faecalis JH2-2. No correlation could be found between the strength of the clumping reaction and the simultaneous or individual presence of the asa1 and/or asa373 genes.

The gelE gene was detected in 74.3 % of all isolates and was thus the most common of the factors that we tested. GelE-positive isolates were significantly more frequent among clinical isolates (75–86 %) when compared to commensal strains (40 %), as already reported by Waar et al. (2002) and Archimbaud et al. (2002). The presence of the gelE gene was, however, not strictly correlated with its expression, as gelatinase was produced by only 36.4 % of gelE-positive strains.

Generally, a higher proportion of invasive strains (45.5 %) compared to non-invasive strains (33.3 %) produced gelatinase; half of commensal strains expressed it and, more strikingly, even though the majority of environmental isolates possessed the gene (83.3 %), it was never expressed under our conditions. All endocarditis isolates possessed the gelE gene, but only 66.7 % of them expressed it.

Silent gelE genes have been observed both in E. faecalis isolates from food and in clinical strains, with a higher incidence of silent genes in the latter group (Eaton & Gasson, 2001). Most studies (Huycke & Gilmore, 1995; Elsner et al., 2000; Franz et al., 2001; Archimbaud et al., 2002; Vergis et al., 2002; Waar et al., 2002) investigated only gelatinase production, perhaps thereby underestimating the real incidence of the gene that encodes GelE among strains tested.

In our hands, results obtained by phenotypic tests always revealed a lower percentage of strains that produced haemolysin, gelatinase or aggregation substance, compared to genotypic characterization. This may be due to the presence of silent genes that are expressed only under in vivo conditions, to the presence of undetected gene mutations or to the fact that detection by PCR of a single gene inside an operon, as is the case of cylA for haemolysin production, may overlook the absence of other genes that are necessary for phenotypic expression. Techniques such as RT-PCR may provide useful information on the level of expression of the target DNA.

We also focused our attention on two novel putative E. faecalis cell-associated antigens by screening the V583 genome. One gene (EF0591) has also recently been found in an E. faecalis clinical isolate inside a pathogenicity island associated with the esp gene (Shankar et al., 2002). The incidence of EF0591 in our collection was rather low (16.2 %), but was associated exclusively with clinical isolates. The EF0591 gene was always associated with the asa1 gene, but was only associated with the esp gene in four cases, indicating that in these strains, a spontaneous deletion of the portion that contains the esp gene, already observed by Shankar et al. (2002), may have occurred.

EF3314 was chosen because of its significant similarity (31.9 %) with biofilm-associated proteins (Cucarella et al., 2001). The PCR survey indicated that this gene is always present and specific for E. faecalis; its role as a possible novel E. faecalis-restricted antigen is currently under investigation.

On the whole, omitting from the analysis the ever-present genes ace, efaA and those for the novel putative antigens EF0591 and EF3314, a characteristic distribution in type and number of virulence factors among strains could be noted.

E. faecalis isolates from endocarditis and the environment possessed only one or two factors. Strains that belonged to other categories had between one and all four factors but the total number of factors differed, depending on the source of isolation (Fig. 1). Most strains isolated from biliary stents possessed one or two factors, while commensal isolates generally had only one factor. Strains from bacteraemia did not show any particular propensity for having a particular number of factors, whereas isolates from UTIs usually possessed two to four factors.

View larger version (19K): [in this window] [in a new window]   Fig. 1. Incidence of single or multiple virulence determinants possessed by E. faecalis isolates. Numbering on the x axis represents number of factors; the percentage of strains that possess a certain number of factors is reported on the y axis.

Commensal isolates always had a gene for aggregation substance. Of significance, cylA was always associated with aggregation substance genes asa1 or asa373, whereas the reverse was not observed.

The presence of virulence factors or their association with a strain from a particular isolation source did not seem to depend on clonal spread of a few enterococcal genotypes. In a recent paper (Dicuonzo et al., 2001), we analysed the PFGE patterns of our E. faecalis collection and observed extreme genetic heterogeneity in our isolates. Only in one case could a small cluster that was derived from the same hospital be grouped clonally, whereas the majority of strains, even when isolated from the same hospital unit, clustered in different subtypes and, notably, some environmental strains had the same genotype as clinical isolates.

The ability to form biofilm on inert surfaces was common in the majority of our E. faecalis isolates; this characteristic was affected strongly by growth conditions (Baldassarri et al., 2001b). No association between biofilm production and presence of other virulence factors was noted in any of several growth media.

A recent study (Duprè et al., 2003) examined the diffusion of some putative virulence factors in a small collection of enterococcal clinical isolates (15 E. faecalis and 32 E. faecium) from an Italian region. They found slightly different levels of incidence, particularly of ace and efaA, which were detected in only 60 and 86.6 % of isolates, respectively.

In conclusion, our data indicate that E. faecalis strains isolated from different sources possess distinctive patterns of potential virulence factors, with a larger number of genes that encode potential virulence factors among isolates from UTIs. Further investigations are needed to evaluate the expression of such factors, which may not be revealed by in vitro phenotypic tests during the course of infection.

	ACKNOWLEDGEMENTS

TOP INTRODUCTION METHODS RESULTS AND DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
This work was presented in part at the First International ASM Conference on Enterococci held in Banff, Alberta, Canada, on 27 February–2 March, 2000. We want to thank P. Boccardi for editorial assistance and Federico Giannoni for helpful discussions. This work was supported in part by a grant from CNR (no. CNRC003CE9_004) to G. O. and a grant from the Italian Ministry of Health, Project 1 % no. OAD/F to L. B.

	REFERENCES

TOP INTRODUCTION METHODS RESULTS AND DISCUSSION ACKNOWLEDGEMENTS REFERENCES