Comparison of Antibiotic Resistance and Virulence between Biofilm-producing and Non-producing Clinical Isolates of Enterococcus Faecium*

An increase in the antibiotic resistance among Entero-coccus faecium strains has been observed worldwide. Moreover, this bacteria has the ability to produce several virulence factors and to form biofilm that plays an important role in human infections. This study was designed to compare the antibiotic resistance and the prevalence of genes encoding surface protein (esp), ag-gregation substance (as), surface adhesin (efaA), colla-gen adhesin (ace), gelatinase (gelE), and hialuronidase (hyl) between biofilm-producing and non-producing E. faecium strains. Therefore, ninety E. faecium clinical isolates were tested for biofilm-forming ability, and then were assigned to two groups: biofilm-positive (BIO + , n =70) and biofilm-negative (BIO – , n = 20). Comparison of these groups showed that BIO + isolates were resistant to β-lactams, whereas 10% of BIO – strains were susceptible to ampicillin (statistically significant difference, p = 0.007) and 5% to imipenem. Linezolid and tigecycline were the only antibiotics active against all tested isolates. Analysis of the virulence factors revealed that ace, efaA, and gelE genes occurred more frequently in BIO – strains (ace in 50% BIO + vs. 75% BIO – ; efaA 44.3% vs. 85%; gelE 2.9% vs. 15%, respectively), while hyl gene appeared more frequently in BIO + isolates (87.1% BIO + vs. 65% BIO –). These differences were significant (p < 0.05). We concluded that BIO + strains were more resistant to antibiotics than BIO – strains, but interestingly, BIO – isolates were characterized by possession of higher virulence capabilities.


INTRODUCTION
Enterococcus, gut commensals in a wide variety of hosts, are today among the leading causative agents of nosocomial infections due to their multiresistance to many antimicrobials.These bacteria are responsible for bacteremia, endocarditis, urinary tract and wound infections (Amyes, 2007;Bronk & Samet, 2008).For a long time, these infections were mostly caused by E. faecalis.In the last two decades, E. faecium has become one of the most prevalent nosocomial pathogens, increasing the total number of enterococcal infections and partially replacing E. faecalis as a cause of hospital-associated infections (Top et al., 2007;Diani et al., 2014).This change is related to the fact that E. faecium has a number of mechanisms of intrinsic resistance to cephalosporins, lincosamides, low levels of aminoglycosides, and many β-lactams (Sieńko et al., 2014).Moreover, this species is also able to acquire resistance by means of mutations or as a result of the transfer of genes located on plasmids, transposons, or due to the incorporation of integrons (Chen et al., 2009;Sieńko et al., 2014).In the standard treatment of enterococcal infections, the use of a cell wall active agent (β-lactam, glycopeptide) with an aminoglycoside results in synergistic bactericidal activity (Rodriguez-Bano et al., 2005).E. faecium has high-level resistance to many β-lactams as a consequence of overproduction and modification of penicillin-binding proteins (PBP), particularly PBP5 (Rodriguez-Bano et al., 2005).Although rare, this resistance is mediated by the production of a β-lactamase enzyme (Rodriguez-Bano et al., 2005;Comerlato et al., 2013).Bacteria resistant to glycopeptides produce cell wall precursors with decreased affinity for the drug administered to treat an infection, which prevents the antibiotic from blocking cell wall synthesis (Sacha et al., 2008;Cheng et al., 2014).High-level aminoglycoside resistance (HLAR), caused by production of aminoglycoside-modifying enzymes (AMEs), makes standard therapy with aminoglycosides and ß-lactams or vancomycin ineffective (Vakulenko et al., 2003).
Additionally, E. faecium strains have the ability to produce several virulence factors and have the ability to form biofilm that plays an important role in human infections (Di Rosa et al., 2006;Sava et al., 2010).The most prominent of the virulence determinants are aggregation substance (as), collagen adhesin (ace), surface adhesins (efaA), hialuronidase (hyl), enterococcal surface protein (esp), and gelatinase (gelE) (Fisher & Philips, 2009;Özden Tuncer et al., 2013).As, encoded by a plasmid as gene, causes binding to the host epithelium and mediates bacterial aggregation during conjugation (Fisher & Philips, 2009).Ace (ace gene), which binds to collagen types I and IV, and efaA (efaA) have been identified as the principal virulence factors associated with infective endocarditis (Fisher & Philips, 2009).Hyl (hyl) degrades hyaluronic acid and is associated with tissue damage (Wu et al., 2007).Esp (esp), which mediates the colonization, and GelE (gelE), a zinc metalloprotease, have been suggested to be involved in the process of biofilm formation (Di Rosa et al., 2006;Heikens et al., 2007;Fisher & Philips, 2009;Diani et al., 2014).
Biofilm is an assemblage of bacterial cells attached to a biotic or abiotic surface and enclosed in a self-produced polysaccharide matrix (Mohamed & Huang, 2007).Its structure provides an optimal microenvironment for growth and facilitates transmission of mobile genetic elements between bacteria.Data suggest that microorganisms in biofilms are more highly resistant to antibiotics than others, extremely difficult to eradicate, and are a source of many chronic infections (Heikens et al., 2007, Paganelli et al., 2013).Among Enterococcus, a multistep process of biofilm formation has been reported to occur less frequently among E. faecium strains compared to E. faecalis species (Baldassari et al., 2001;Almohamad et al., 2014).It has been suggested that several virulence determinants are implicated in biofilm formation among Enterococcus (Di Rosa et al., 2006;Paganelli et al., 2013).Nevertheless, there are many conflicting literature reports about their contribution in biofilm production (Di Rosa et al., 2006;Mohamed & Huang, 2007;Heikens et al., 2007;Fisher & Philips, 2009).Moreover, in the case of E. faecium, data about biofilmforming ability are still very limited (Almohamad et al., 2014).This prompted us to determine the prevalence of biofilm-forming ability among E. faecium clinical isolates.We focused on the search for differences in virulence between biofilm-producing (BIO + ) and non-producing (BIO-) E. faecium strains.Moreover, due to the alarming increase in resistance among Enterococcus in Poland (Dworniczek et al., 2014; Antimicrobial resistance surveillance in Europe 2013.Stockholm: ECDC; 2014), the next goals of our study were to compare the susceptibility of tested strains to antibiotics between BIO + and BIO -isolates, to determine their resistance profiles, and to indicate the antibiotic with the highest activity.

MATERIALS AND METHODS
Tests were performed on ninety E. faecium strains, randomly selected from the collection of the Department of Microbiological Diagnostics and Infectious Immunology (Medical University of Białystok, Poland).Strains were isolated from clinical specimens of patients hospitalized at the University Hospital in Białystok (Poland) from December 2013 to January 2015.Isolates were recovered from various clinical materials, mostly rectal swabs, faeces, blood, urine, and pus.Most of collected isolates were gathered from the intensive care unit and a hematology clinic.
Identification and susceptibility testing.The identification and susceptibility testing of study isolates were conducted on the automated VITEK 2 system (bioMérieux, France) according to the manufacturer's instructions using VITEK 2 GP and AST-P516 cards, respectively.Susceptibility to ampicillin, imipenem, gentamicin, streptomycin, vancomycin, teicoplanin, linezolid, and tigecycline was interpreted according to the European Committee on Antimicrobial Susceptibility Testing (EU-CAST) recommendations (breakpoint tables for interpretation of minimum inhibitory concentrations (MIC) and zone diameters; version 5.0, 2015; http://www.eucast.org.).E. faecalis ATCC 29212 was used as a reference strain.Later, identification to the species level was confirmed by polymerase chain reaction (PCR) with primers targeted to specific sequences in the ddl (d-Ala-d-Ala ligase) chromosomal genes (Table 1).
Biofilm production.The tube method (Christensen et al., 1982;Oliveira & Cunha, 2010) and Congo red agar (CRA) method (Freeman et al., 1989;Cabrera-Contreras et al., 2013) were used to assess biofilm-forming ability.Each experiment was repeated three times for each strain.In the first test, trypticase soy broth (Emapol, Poland) with sucrose (WARCHEM, Poland) was inoculated with a loopful of the bacteria from an overnight culture and incubated for 24 h at 37ºC.The tubes were then decanted and washed with phosphate-buffered saline (PBS) (pH 7.2) (BTL, Poland) to remove non-adherent cells.The tubes were then dried and stained with 0.1% crystal violet (Graso, Poland).After 30 minutes, the excess stain was washed off with distilled water and the tubes were left to dry in an inverted position and observed for biofilm formation.A visible film lining the bottom and the sides of the tube was considered to be indicative of biofilm production.In the second method, for preparing CRA medium, 50 g/l sucrose and 0.8 g/l Congo red Isolates that demonstrated the ability to produce biofilm by both methods were identified as BIO + strains, others as BIO -isolates.β-lactamase and hemolysin production.Strains were tested for β-lactamase production by a chromogenic cephalosporinase method (Pitkälä et al., 2007) using nitrocefin discs (OXOID, United Kingdom) as per manufacturer's instruction.Staphylococcus aureus ATCC 29213 was used as a positive control.Hemolysin production was evaluated on Columbia blood agar supplemented with 5% sheep blood (OXOID, United Kingdom), as previously described (Vergis et al., 2002).
DNA extraction.Genomic DNA was extracted from overnight E. faecium cultures using a Genomic Mini Kit (A&A Biotechnology, Poland) according to the manufacturer's protocol.
PCR detection of virulence genes.PCR assays were performed to detect the following virulence genes: gelE, ace, hyl, esp, as, and efaA.The primers used in this assay were selected from the literature and their sequences are listed in Table 1.PCR amplification was performed in 25 µl mixtures using 2 µl of DNA solution, 1 µl of each primer, 8.5 µl of nuclease-free water, and 12.5 µl of PCR master mix (DNA Gdańsk, Poland).Samples were subjected to an initial cycle of denaturation at 94ºC for 5 min, followed by thirty cycles of denaturation at 94ºC for 1 min, annealing at an appropriate temperature for 1 min, and elongation at 72ºC for 1 min using a programmable DNA thermocycler (SensoQuest GmbH, Germany).
PCR products were separated electrophoretically on Sub-Cell GT apparatus (Bio-Rad, USA) at 5V/cm for 100 min on a 1.5% agarose gel (Sigma-Aldrich, USA) containing 0.5% ethidium bromide (MP Biomedicals, USA) in Tris-borate-EDTA (ethylenediaminetetraacetic acid) buffer.Then, amplicons were visualized and pho-tographed using the ChemiDoc XRS imaging system and Quantity One 1-D analysis software (Bio-Rad).The positions of obtained products were estimated with the molecular weight marker, Perfect TM 100-1000 bp DNA ladder (EURx, Poland).To confirm the presence of the above-mentioned virulence genes, DNA sequencing was carried out on selected PCR products by GENOMED S.A. company in Poland.The sequences were aligned and compared with reference sequences achieved using GenBank with the Basic Local Alignment Search Tool (BLAST) algorithm.
Statistical analysis.STATA 13.1 (StataCorp LP, USA) was used for statistical analysis.Differences in the prevalence of antibiotic resistance and virulence determinants between BIO + and BIO -strains were assessed by the Chi-square test and Fisher's exact test.Results with p < 0.05 were considered significant.

RESULTS
We observed an ability to produce biofilm in seventy of ninety E. faecium strains (77.8%).As mentioned earlier, these isolates were classified as BIO + , and the remaining twenty strains (22.2%) as BIO -.An exact comparison of antibiotic resistance between BIO + and BIO -E.faecium isolates is presented in Fig. 1.All of the BIO + isolates showed phenotypic resistance to tested β-lactams, whereas 10% of BIO -strains were susceptible to ampicillin (statistically significant difference, p = 0.007), and 5% were susceptible to imipenem.None of the investigated isolates had the ability to produce the β-lactamase enzyme.Both groups showed high-level resistance to aminoglycosides: resistance to gentamicin was detected in 75.7% of BIO + and 60% of BIO -strains, to streptomycin in 91.4% of BIO + and 85% of BIO -strains, respectively (insignificant differences, p > 0.05).Interestingly, more than half of the tested strains in both groups was resistant to vancomycin and teicoplanin.Linezolid and tigecycline had the highest activity against all studied isolates (100% susceptibility).Comparative analysis (Fig. 2) of the prevalence of virulence genes among BIO + and BIO + strains revealed that ace, efaA, and gelE genes occurred more frequently in BIO -strains (ace in 50% BIO + vs. 75% BIO-, efaA in 44.3% vs. 85%, gelE in 2.9% vs. 15%, respectively), while hyl gene appeared more frequently in BIO + isolates (87.1% BIO + vs. 65% BIO -).These differences were statistically significant (p < 0.05).No statistically significant differences were found in the case of esp and as genes (p > 0.05).
Resistance and virulence patterns among all BIO + and BIO -E.faecium strains are shown in Table 2.The most frequent antibiotic resistance profile among BIO + strains was AMP R IMP R CN R S R (resistance to ampicillin, imipenem, gentamicin, streptomycin, respectively), which was detected in ten strains.These strains had hyl and esp virulence genes and the ability to hemolyze.The most frequent resistance and virulence profile of BIO -isolates, which occurred in three strains, was AMP R IMP R CN R S R TEC R VA R (resistance to ampicillin, imipenem, gentamicin, streptomycin, teicoplanin, and vancomycin, respectively) with ace, efaA, and hyl genes.All (100%) BIO + strains carried resistance to three or more antibiotics and had the ability to α-hemolyze, whilst three (15%) BIO - strains showed resistance to only two antibiotics, and a smaller number of these strains (n = 12; 60%) exhibited α-hemolysis.BIO + isolates showed a greater variety of resistance and virulence determinants than BIO -strains.

DISCUSSION
The present study focused on determining the prevalence of biofilm-forming ability among E. faecium clinical isolates and on comparison of the antibiotic resistance and the prevalence of genes encoding selected virulence factors between biofilm-producing (BIO + ) and non-producing (BIO -) strains.In this study we observed very high incidence of the ability to form biofilm among randomly selected E. faecium clinical strains (77.8%).A similar proportion was observed in Spain (75%) (Latasa et al., 2006).However, studies by other authors showed different results; in India, Italy, and Turkey, E. faecium isolates were able to produce biofilm less frequently or even could not form this structure (0%, 28.8%, and 48%, respectively) (Prakash et al., 2005;Di Rosa et al., 2006;Diani et al., 2014).These results indicate that the level of the ability to form biofilm among E. faecium, as well as the factors conducive to its formation, vary with geographic location.
The notion that bacteria in biofilms are more resistant to antibiotics than planktonically grown microorganisms (Lewis, 2001;Heikens et al., 2007, Paganelli et al., 2013) was not fully confirmed in our survey.BIO + strains were, admittedly, slightly more resistant than BIO -isolates, but the statistically significant difference between these groups was found only in the case of ampicillin.We were also very astounded due to the fact that in the literature there are only a few reports about the differences in resistance between biofilm-producing and nonproducing isolates belonging to E. faecium species (Raad et al., 2005;Akhter et al., 2014), compared to the amount of data about E. faecalis strains (Chai et al., 2007;Mohamed et al., 2007;Yayanthi et al., 2008;Lins et al., 2013;Frank et al., 2015).
In our study, the majority of E. faecium isolates (> 90%) exhibited resistance to β-lactams.None of the tested strains had β-lactamase activity; therefore, we can assume that this resistance is associated with changes in PBPs.These results are in agreement with other studies (Simonsen et al., 2003;Iris et al., 2014).However, researchers from Spain (Rodriguez-Bano et al., 2005) obtained only 28.6% ampicillin-resistant E. faecium strains, whereas in the Netherlands (Schouten et al., 1999) 24% of E. faecium strains were susceptible to imipenem.It should be noted that in this study most of the tested strains that were resistant to β-lactams were also resistant to gentamicin or streptomycin.Similar results were observed by Simonsen et al. (2003).The high rate of co-resistance between ampicillin and aminoglycosides among E. faecium, especially in vancomycin-resistant enterococci (VRE) isolates, is worrisome since it eliminates the synergistic effect between β-lactams and aminoglyco- sides in the treatment of patients (Simonsen et al., 2003;Sieńko et al., 2014).Unfortunately, E. faecium isolates resistant to β-lactams, aminoglycosides, and glycopeptides, considered as multidrug resistant (MDR), are now widespread across Europe (Hryniewicz et al., 2009).According to a recent multicenter report (Antimicrobial resistance surveillance in Europe 2013.Stockholm: ECDC; 2014), resistance to glycopeptides has significantly increased over the last four years, including Poland.The high prevalence of E. faecium strains with AMP R IMP R CN R S R TEC R VA R resistance patterns, obtained in this study, confirms that the scale of the problem with MDR Enterococcus strains is large, and that changing in epidemiology of these strains remains a major infection control challenge throughout Europe.This study showed that linezolid and tigecycline were the most active antibiotics toward all tested strains.Many authors confirm that they are a valuable therapeutic option in infections caused by E. faecium, including VRE (Franiczek et al., 2008;Freitas et al., 2011;Praharaj et al., 2013;Sieńko et al., 2014).However, cases of resistance to these antibiotics have been recently reported (Werner et al., 2008;Baldir et al., 2013).This may indicate that resistance to newer antimicrobials is also developing; therefore, new strategies, including combination therapies, are urgently needed.
Despite the fact that the biofilm-formation process has an essential impact on the course of enterococcal infections, our knowledge of the mechanisms and factors involved in this process is still insufficient (Almohamad et al., 2014).Therefore, many studies have sought to find the relation between biofilm formation and virulence genes, especially esp and gelE, among Enterococcus (Dupre et al., 2003;Dworniczek et al., 2005;Raad et al., 2005;Di Rosa et al., 2006;Heikens et al., 2007;Fisher & Philips, 2009;Diani et al., 2014).We reported that the prevalence of esp gene was higher in BIO + isolates (71.4%) than in BIO -strains (55%), but that this difference was statistically insignificant, and that many esp-positive isolates did not form biofilm.These findings suggest that this gene has no connection with biofilm-forming ability.Similar proportions and lack of significant differences were seen by other researchers (Dupre et al., 2003;Dworniczek et al., 2005;Raad et al., 2005;Di Rosa et al., 2006;Almohamad et al., 2014).Nevertheless, many authors found that there is a strong relationship between the esp gene and biofilm formation (Heikens et al., 2007;Fisher & Philips, 2009;Top et al., 2013;Diani et al., 2014).Undoubtedly, these varied and conflicting results indicate that Esp may require interaction with other virulence factors to result in biofilm enhancement.Interestingly, researchers from Sweden (Bilström et al., 2008) concluded that E. faecium strains that carry the esp gene demonstrate higher resistance to β-lactams.Likewise, in our study, we observed the coexistence of resistance to β-lactams and the esp gene in the majority of tested strains, and two strains from the BIO -group that were susceptible to ampicillin did not have this gene.However, more research is definitely needed in this area, particularly studies concerning the expression of virulence genes.
The presence of gelE and as genes among E. faecium strains is very rare, whereas they are widely present in E. faecalis isolates (Vankerckhoven et al., 2004;Comerlato et al., 2013).In our study, we observed small percentages of strains with gelE and as genes (< 15%).Similar results have also been reported by other researchers (Kowalska-Krochmal et al., 2011;Hasani et al., 2012;Comerlato et al., 2013), but two studies (Vankerckhoven et al., 2004, Diani et al., 2014) did not find any of them with PCR in large groups of E. faecium isolates.We found that gelE and as genes occurred more frequently in BIO -groups than in BIO + groups, which suggests that these genes have a greater role in the pathogenicity of E. faecium strains unable to form biofilm.This result is not in concordance with those from other studies (Hancock & Perego, 2004;Mohamed et al., 2007;Kafil et al., 2015); however, one study (Di Rosa et al., 2006) confirmed that there is no association between gelatinase and biofilm in E. faecium strains.
In the case of other virulence factors, our findings that BIO -strains had significantly more ace and efaA genes than BIO -strains indicate that isolates carrying these genes prefer a planktonic rather than a biofilm lifestyle.We also found that the hyl gene occurred more frequently in BIO + strains.Different results were observed in one recently published study (Kafil et al., 2015); the authors showed that isolates with the efaA gene produced more biofilms than negative ones, while strains with the hyl gene had a significantly lower biofilm-forming ability.
In conclusion, our data demonstrated that BIO + E. faecium strains were slightly more resistant to antibiotics than BIO-strains, but, interestingly, BIO -isolates were characterized by a higher virulence potency.Nevertheless, these observations are not in agreement with many previously published reports.Our attempts to understand these large numbers of contradictory results have allowed us to conclude that the ability to form biofilm cannot be unambiguously linked to increased virulence and resistance in E. faecium strains.This stresses the need to perform more research on regulation and expression of virulence and resistance genes, how to prevent the spread of MDR enterococcal nosocomial infections, and on treatment alternatives.Novel approaches, including the use of metabolomics, proteomics, and genomics, may improve our knowledge of E. faecium biofilm, in the light of changing epidemiology and increasing resistance to antibiotics.Novel drugs targeted at specific virulence factors may play a preventative or even therapeutic role in the elimination of MDR E. faecium strains.

Table 1 . PCR primers, annealing temperatures, and product sizes for detection of ddl gene and virulence genes.
efaA AAC AGA TCC GCA TGA ATA CAT TTC ATC ATC TGA TAG TA 735 50 Özden Tuncer et al., 2013 ddl GGC AGA GCA TGA AGT GTC CA CTT CTG GGT TTT CTG CTT TTG TA 557 56 Dutka-Mahlen et al., 1995 Comparison of biofilm-producing and non-producing E. faecium strains stain (WARCHEM) were added to brain heart infusion agar (Sigma-Aldrich, USA) and autoclaved at 121ºC for 15 m.Strains were inoculated onto CRA and incubated at 37ºC for 24 h.Reading was done after 24 h and 48 h.A positive result was indicated by black colonies with black crystalline morphology; non-biofilm producers produced pink-colored colonies.
α 1 Comparison of biofilm-producing and non-producing E. faecium strains