Abstract
Resistance to disinfectants is well investigated in staphylococci and pseudomonads but nearly unexplored in bacteria of the genus Enterococcus, despite their rising significance as nosocomial pathogens. In this study, Enterococcus faecalis (n=585) from blood (n=42) and stool (n=109) of hospitalized humans, from faeces of farm animals (n=226), and from food (milk and dairy products, n=96; meat and meat products, n=112) were screened for the presence of qac-genes (qacA, qacB, qacC, smr [qacC+qacD], qacEΔ1, qacG, qacH, qacJ) via PCR. The isolates' susceptibility to a quaternary ammonium compound (didecyldimethylammoniumchloride, DDAC) and antibiotics was assessed by microdilution. Four E. faecalis strains were positive for qac-genes: qacA/B was found in one isolate from cattle and one isolate from human blood; smr (qacC+qacD) was detected in one isolate from human stool and in one isolate from cheese (“Camembert”). The sequences of the qacA/B-amplicons differed in two basepairs. DDAC had an elevated minimum inhibitory concentration (MIC) of 2.45–3.5 mg/L in one qacA/B-positive strain from human blood, whereas the other qac-gene carriers had wild-type MIC-values for DDAC (1.05 mg/L). This is the first detection of qacA/B in the genus Enterococcus.
Introduction
Materials and Methods
Isolation and identification of Enterococcus faecalis
All E. faecalis strains were collected in the frame of the second Bavarian antimicrobial resistance monitoring project (B4VMAPII). E. faecalis from human blood were isolated during clinical routine diagnostic testing in Bavarian hospitals between 2008 and 2009 and were sent into the research laboratory on Amies-Medium (BD, Heidelberg, Germany). E. faecalis from human stool were isolated as follows within the same period: Diagnostic stool samples were cultured on citrate azide tween carbonate (CATC)-agar (VWR, Darmstadt, Germany) at the hospital or the diagnostic laboratory and sent to the research laboratory. Diagnostic stool samples of animals were sent by veterinarians and were cultured on CATC-agar in the research laboratory. Food samples were taken within routine food control and streaked on CATC-agar after a 1:10 preenrichment in 1% peptone water. All isolates were subcultured on Standard I Nutrient agar (Merck, Darmstadt, Germany) containing 7% defibrinated sheep blood (Fiebig, Idstein-Niederauroff, Germany). For species-identification, results of the biochemical test procedure (API Rapid ID 24 Strep; biomeriéux, Nürtingen, Gerrmany) were compared with the resistance profiles (see below) of the isolates; in case of discrepancies (e.g., synercid-susceptibility in presumptive E. faecalis), the species was confirmed by PCR/agarose-gel-electrophoresis. For this purpose, DNA was extracted from Enterococcus isolates using cetyltrimethylammoniumbromide (CTAB). The CTAB-extraction was mainly carried out as described elsewhere, 17 but directly began with the inoculation of bacterial colonies into 500 μl Tris-EDTA. The PCR-protocol was modified from Jackson et al. 14 for E. faecalis and from Cheng et al. 4 for E. faecium, using a primer concentration of 0.5 μM for E. faecalis and 0.1 μM for E. faecium; as a further modification, an annealing step was introduced in the two-temperature-PCR of Cheng et al. 4 Primers and annealing temperatures are given in Table 1.
Antimicrobial resistance
Antimicrobial resistance was assessed in a standardized microdilution procedure according to DIN 58940-18 8 as previously described.30,13 See Hölzel et al. 13 for concentration ranges and breakpoints (mainly DIN 58940-4 7 ) of ampicillin, amoxicillin+clavulanic acid, imipenem, enrofloxacin, ciprofloxacin, moxifloxacin, doxycycline, gentamicin high level, erythromycin, vancomycin, and strepromycin high level. For the following substances the concentration ranges (mg/L) were modified, compared to Hölzel et al. 13 : mezlocillin (0.5–64), meropenem (0.25–32), chloramphenicol (1–32), teicoplanin (0.125–16), linezolid (0.25–32), florfenicol (0.5–64), and tylosin (0.25–8). Levofloxacin (0.0625-8), telithromycin (0.0625-8), and tigecycline (0.03125-4) were additionally introduced into the testing procedure.
Susceptibility to didecyldimethylammonium-chloride
The susceptibility to didecyldimethylammoniumchloride (DDAC) (Tradename Sokrena, Bode Chemie, Hamburg, Germany) of type strains and qac-positive reference strains (Table 1) was assessed in a microdilution procedure. The microdilution procedure was adapted from the macrodilution test described in the guidelines of the German Veterinary Medical Society (DVG 10 ). CS-Broth (30 g cold filterable Tryptone Soya Broth per 1 L distilled water) containing 1×108 to 1×109 cfu of the test strain per ml was diluted 1:10 with water of standardized hardness (WSH, consisting of 0.89 g waterfree NaCl2 and 0.5 g MgCl2×7 H2O per 3.3 L of distilled water), resulting in bacterial concentrations of 1×107 to 1×108 cfu per ml. Then, 100 μl of a double concentrated DDAC–WSH solution was manually placed into each well of a 96 well microtitre plate. A total of 11 log4-scaled concentrations of DDAC and a growth control were eight times placed on each microtitre plate. Then, 13 ml of CS-broth was inoculated with 226 μl bacterial suspension, and 100 μl of the resulting suspension was pipetted into each (DDAC-WSH filled) well of the microtitre plate by a semi-automatic dispenser (Micronaut Sprint, Genzyme-Virotech). Microtitre plates were covered with transparent plastic films and incubated at 37°C for 72 hr, as specified in the DVG-instructions 10 ; the microbial growth was investigated after 24 hr and 72 hr. The investigated DDAC concentrations ranged from 0% (growth control) to 0.5% (corresponding to 0 to 350 mg DDAC/L). These concentrations included the minimum inhibitory concentration (MIC) value of DSM type strains and the concentration recommended for use by the manufacturer (0.5%).
Resulting from the data of DSM-type strains and 585 field strains, the epidemiological cutoff value (ECOFF) was set at 1.4 mg DDAC/L. The ECOFF is defined as the value which separates microorganisms without (wild type) and with acquired resistance mechanisms (non-wild type) to the agent in question (see the European Committee on Antimicrobial Susceptibility Testing, EUCAST, http://217.70.33.99/Eucast2/SearchController/index.jsp?action=initAdvanced). Values below or equal to the ECOFF are referred to as wild-type MIC-values. Subsequently, strains with MIC-values >1.4 mg/L and DSM type strains (Table 1) were additionally tested in steps of 0.35 mg/L for concentrations of 0.35 to 2.8 and in steps of 0.7 mg/L for concentrations from 2.8 to 5.6 mg/L. MIC-results of all strains with MIC-values >1.4 mg/L were confirmed twice with freshly defrosted subcultures.
Detection of qac-genes
For the detection of qac-genes, endpoint-PCR (Biometra, Göttingen, Germany) was performed as described elsewhere.2,24 For primer sequences and annealing temperatures see Table 1. PCR-products were visualized in a 1% agarose gel (45 min, 200 V) containing 0.285 μg/ml ethidium bromide (Sigma-Aldrich, Hamburg, Germany). Gels were loaded with the target PCR-amplicons, PCR-amplicons of the positive controls (Table 1), a negative control (consisting of all PCR-ingredients apart from DNA extract), and a size marker. Detected fragments of the right size were extracted from the gel using the QIAquick Gel-Extraction Kit (Quiagen, Hilden, Germany); the fragments were subsequently purified with the QIAquick PCR Purification Kit (Quiagen) and externally sequenced (Sequiserve, Vaterstetten, Germany). Sequences were compared with GenBank reference sequences using the BLAST-program (Basic Local Alignment Search Tool; www.ncbi.nlm.nih.gov/). All positive results were repeated with freshly extracted cultures.
Results
DDAC-MIC-values of type strains and of qac-positive reference strains
All Enterococcus type strains were inhibited at 1.05 mg DDAC/L. S. aureus DSM 2569 was inhibited at 0.7 to 1.05 mg/L, Escherichia coli DSM 1103 at 4.2 to 5.6 mg/L, and P. aeruginosa DSM 1117 at 21.0 mg/L. Carriers of qac-genes were also tested for their susceptibility to DDAC; all qac-positive S. aureus/MRSA strains had DDAC-MIC-values between 1.75 to 2.45 mg/L, except S. aureus-qacB, which had a MIC of 1.05 mg DDAC/L, and S. aureus-qacA, which had MIC of 1.05 to 1.75 mg/L. The MIC of the qacEΔ1–transformed E. coli did not differ from the E. coli-DSM type strain (4.2 to 5.6 mg/L).
Detection of qac-genes
Biocide resistance genes of the qac-type were detected in 4 of 585 E. faecalis strains (0.7%). Two E. faecalis strains were positive for qacA/B: one strain from cattle (Efa-AB-1) and one strain from human blood (Efa-AB-2). Two E. faecalis strains were positive for the smr-element containing qacC+qacD: Efa-smr-1 from the feces of a stationary patient in a supra-regional hospital and Efa-smr-2 from a dairy product (“Camembert”). qacEΔ1, qacG, qacH, and qacJ were not detected. Both smr-positive strains and Efa-AB1 from cattle had a wild-type MIC for DDAC (1.05 mg/L). By contrast, Efa-AB-2 had an elevated MIC-value of 2.45–3.5 mg DDAC/L.
Efa-AB-1 and Efa-AB-2 were resistant to Doxycycline and Erythromycin. Wild-type MICs were seen for all other substances, except for Streptomycin High Level (1024 mg/L). Efa-smr-1 and Efa-smr-2 were pan-susceptible.
Sequences of the smr- and qacC-amplicons were completely identical to smr (qacC) reference sequences on Genbank (smr: GI:27461330; position 1963–2157; qacC: GI:2598266, position 143-299). As shown in Table 2, the amplicon of Efa-AB-2 (MIC 2.45 to 3.5 mg/L) differed in only one base pair from qacA (GI:300492196, position 300–670), but in three base pairs from qacB (GI:3327946, position 300–670), although it was completely identical to another qacA-subject in the Genbank database (GI:298103888). By contrast, the amplicon of Efa-AB-1 (wild-type MIC) differed in three base pairs from qacA (GI:300492196), and in three base pairs from qacB (GI3327946). On the protein level, these differences resulted in complete identity between the translated Efa-AB-2-amplicon and both QacA-Proteins encoded by the Genbank reference sequences mentioned above. By contrast, the translated Efa-AB-1-amplicon had one substitution of a conserved amino acid (148V→L) compared with the products of qacA (GI:300492196) and qacB (GI3327946). Compared with QacB, a second substitution of a conserved amino acid (167I→L) was detected. Compared to QacA, the second substituted amino acid was a non-conserved one (152A→V).
Three points (…) symbolize all parts with 100% identity between all compared sequences.
Numbers in parentheses: position within the amplicon.
Discussion
With four positive findings, qac-resistance genes were only rarely detected (0.7%). There are only two single reports on qac-genes in enterococci up to now, concerning the integron-associated qacEΔ116 and the smr (=ebr)-element which contains qacC+D. 29 Kazama et al. 16 found the qacE Δ1 gene in clinical Enterococcus-isolates with a comparatively high prevalence (18.8%), whereas we did not find qacEΔ1 in any of the 298 clinical E. faecalis isolates (neither in E. faecalis of other sources). Different investigation pools might explain the difference, as Kazama et al. 16 investigated Japanese strains. This hypothesis of regional differences is supported by the fact that the authors report also a much higher prevalence of qacEΔ1 in Pseudomonas spp., 15 compared with a study carried out in Germany. 18 Moreover, the situation might have changed in the course of time, since the isolates of Kazama et al. 16 dated from 1995 to 1996.
Instead of qacEΔ1, we detected qacA/B and smr (=ebr) in two E. faecalis isolates, each. To our knowledge, this is the first report of qacA/B in enterococci.
Enterococcus-isolates with smr had wild-type MIC-values of 1.05 mg DDAC/L. The same was reported by Sasatsu et al. 29 for ebr (=smr)-positive Enterococcus strains; the authors described that the number of ebr-copies per cell needs to increase for the expression of QAC-resistance. One isolate which harboured qacA/B had a wild-type MIC-value for DDAC (1.05 mg/L). A second qacA/B positive isolate from human blood had an elevated MIC value of 2.45 to 3.5 mg DDAC/L. Benzalkonium chloride MIC-values for qacA/B-positive Staphylococcus spp. ranged from 2.0 to 3.5 mg/L in a previous study. 2 In the present study the MIC-value of DDAC for the control strains with qacA and qacB was 1.05 (qacB) and 1.05 to 1.75 (qacA), respectively. The genes qacA and qacB are closely related and differ in seven basepairs only21,27; qacA is thought to cause more pronounced tolerance to QAC than qacB. 27 Thus, qacA might have developed from qacB by spontaneous mutation and might have subsequently been selected by biocide use. 27 The higher MIC-value of Efa-AB-2, compared with Efa-AB-1, might therefore result from the presence of different genes (qacA and qacB). In fact, the sequences of the amplicons differed in two base pairs. Although the sequence of the translated Efa-AB-2 amplicon was identical to the protein products of qacA-reference sequences as deposited in Genbank, the sequence of the Efa-AB-1-amplicon differed from both qacA and qacB in three base pairs, resulting in one or two exchanges of conserved amino acids, compared with QacA or QacB, respectively.
In S. aureus, qac-genes have been described to be linked with antimicrobial resistance genes. 31 In the present study, no phenotypic antimicrobial resistance was associated with the presence of smr in E. faecalis, since both positive strains were pan-susceptible. For qacA/B, an association with antimicrobial resistance cannot be excluded, but is not necessarily present, since both phenotypic antimicrobial resistances (against doxycycline and erythromycin) of the qacA/B-positive strains were frequently seen in E. faecalis from cattle and humans (data not shown).
According to the literature, qac-genes are regularly found on plasmids. 20 GenBank reference sequences of smr belong to chromosomal DNA and plasmids from S. aureus, Staphylococcus epidermidis, Staphylococcus pasteuri, and Staphylococcus warneri. GenBank reference sequences of qacA/B belong to 27 different plasmids of S. aureus, 4 plasmids of S. epidermidis, and also to genomic DNA of Staphylococcus haemolyticus and S. aureus. We did not investigate the location or transferability of qac-genes in the present study. However, it is interesting that the only qac-gene which we found in enterococci from food–qacC (smr)–is the most prevalent qac-gene of staphylococci from food (54.2%), 12 whereas qacA/B is the most frequent gene in staphylococci of clinical origin (63.0% in MRSA). 23 The emergence of qacC in the genus Enterococcus in food and of qacA/B in enterococci from the clinical environment might reflect the high prevalence of these genes in the respective surrounding gene pools.
Conclusions
The gene qacA/B was detected in the genus Enterococcus for the first time. This finding should not be ignored, though the prevalence is low up to now: in methicillin-resistant S. aureus, the prevalence of qacA/B was initially low, too, 26 but increased to 42% in the same area within the following seven years. 25 The presence of smr seems not yet to be selectable in enterococci by biocide use, since its presence was not related to phenotypic resistance (neither in the present study nor in a previous study). 29 However, phenotypic resistance might be caused by an increased copy-number of smr per cell. The increase of copies from a genetic element which is already present in the genome might be more easily achieved than the acquisition of high copy-numbers of a completely new element. The detection of qacA/B in one clinical strain of this study was already accompanied by an elevated MIC of a quaternary ammonium compound. Thus, qac-positive enterococci might be further selected in the future by biocide use, therefore deserving close attention.
Footnotes
Acknowledgments
This study was funded by the Bavarian State Ministry of the Environment and Public Health. Thanks are due to Renate Heckel for expert technical assistance and to J. Bjorland, N. Noguchi, and K. Smith for the kind provision of reference strains.
Disclosure Statement
No competing financial interests exist.
