OUP user menu

Conjugal transfer of aminoglycoside and macrolide resistance between Enterococcus faecium isolates in the intestine of streptomycin-treated mice

Camilla H Lester, Niels Frimodt-Moller, Anette M Hammerum
DOI: http://dx.doi.org/10.1111/j.1574-6968.2004.tb09614.x 385-391 First published online: 1 June 2004

Abstract

The purpose was to study conjugal transfer of resistance genes between a multi-resistant Enterococcus faecium isolate and a sensitive E. faecium isolate. Co-transfer of erm(B)-Tn5405-like element and aac(6′)-Ie-aph(2′)-Ia was obtained in both in vivo and in vitro. Plasmid profiles and Southern blots showed that both the erm(B)-Tn5405-like element and aac(6′)-Ie-aph(2′)-Ia were placed on the same large plasmid (>147 kb). These data show to our knowledge the first co-transfer of the erm(B)-Tn5405-like element and aac(6′)-Ie-aph(2′)-Ia. The in vivo study also indicates that transfer of resistance genes between enterococci might occur under natural conditions in the gut of animals.

Keywords
  • Conjugation
  • erm(B)
  • aac(6′)-Ie-aph(2′)-Ia
  • Intestine

1 Introduction

Enterococcus spp. is a common bacterial genus encountered in clinical specimens; this includes being the second most common among urine isolates and third most common among blood isolates in the USA [1]. Enterococcal infections are often treated with a combination of an aminoglycoside and a cell-wall-active agent such as a penicillin or a glycopeptide. In both the USA and the United Kingdom, vancomycin- and ampicillin-resistant Enterococcus faecium has become a major clinical problem [1]. In Denmark, vancomycin-resistant enterococci still have a very low prevalence, whereas high-level resistance to gentamicin is common [2]. Resistance to high-level gentamicin (MIC > 500 mg l−1) can be a problem because combination treatment with aminoglycoside and vancomycin/penicillin will fail. High-level gentamicin resistance in enterococci is predominantly mediated by aac(6′)-Ie-aph(2′)-Ia, which encodes the bifunctional aminoglycoside-modifying enzyme AAC(6′)-APH(2′) [3].

Resistance to streptomycin, kanamycin and erythromycin is often detected in Danish high-level gentamicin-resistant E. faecium isolates [4]. The genes aadE, aphA-3 and erm(B), encoding resistance to streptomycin, kanamycin and erythromycin, respectively, are found together on the erm(B)-Tn5405-like element in both E. faecium and Staphylococcus intermedius[5,6].

Resistant enterococci are known to exist outside humans especially in food animals and the same resistance genes are found in humans and food animals [7]. Enterococci can harbor different transferable genetic elements such as conjugative transposons and different types of plasmids with resistance genes. Conjugal transfer from enterococci to other bacterial species has also been demonstrated [8,9]. An ideal place for gene transfer could be the gut of different animal species including humans. Transient colonization of enterococci in the human gut may easily take place after ingestion of food containing the bacteria [10,11]. Different models have been used to study gene transfer between Gram-positive bacteria in the gut of different animal species [1216].

The main goal of the present study was to investigate the transfer of resistance genes from a multi-resistant human E. faecium isolate to a sensitive E. faecium isolate. The transfer was studied both in vitro and in vivo in the intestine of streptomycin-treated mice.

2 Materials and methods

2.1 Bacterial strains

A Danish human ascites E. faecium isolate 160/00 resistant to erythromycin, streptomycin, kanamycin, penicillin, tetracycline and gentamicin was used as donor. The strain used as recipient was E. faecium 64/3-RFS, which was a spontaneously streptomycin-resistant mutant of the rifampin- and fusidic acid-resistant E. faecium 64/3. The strain was plasmid-free [17].

2.2 Antimicrobial susceptibility testing

The donor and the transconjugants were tested for susceptibility to the antimicrobial agents erythromycin, gentamicin, kanamycin, penicillin, streptomycin, tetracycline, vancomycin, linezolid, penicillin, quinupristin/dalfopristin and chloramphenicol using a commercially prepared, dehydrated panel (Sensititre®, TREK, United Kingdom) as previously described [18].

2.3 PCR

PCR-detection was done according to Hammerum et al. [19]. All primers are listed in Table 1. S. intermedius MLS-2 was used as a positive control for the erm(B)-Tn5405-like element [6] and E. faecalis SF350 for aac(6′)-Ie-aph(2′)-Ia[20].

View this table:
1

Primerlist

Primer namePrimer sequencePrimer positionGenBank Accession No.Reference
ermB-15′-CAT TTA ACG ACG AAA CTG GC-3′386–405AF299292[35]
ermB-25′-GGA ACA TCT GTG GTA TGG CG-3′810–791AF299292[35]
Macjonc-A5′-AGC AAT GAA ACA CGC CAA AG-3′901–920AF299292[6]
Macjonc-B5′-CTG CCA CTG ATT AAG CCA CT-3′1990–1971AF299292[6]
Kanint-F5′-CGC ACA AGC TTT CAT TCT TA-3′3095–3076AF299292[6]
Kanint-N5′-ACA GAG CAC GGA ATT GTT AC-3′2972–2991AF299292[6]
MLS-4B5′-AGG AAT CCA TCC GGT AGG T-3′4958–4940AF299292This study
MLS-3B5′-GAT ATC ATG GAA GGT CGG CA-3′4831–4850AF299292This study
Strkan-B5′-GTA ATC CAA TTC GGC TAA GCG-3′6233–6213AF299292[6]
Strkan-L5′-GAT CTG GCC GAT GTG GAT TG-3′6246–6265AF299292[6]
Kanint-C5′-TCT GAA TTG CTG CAA TAA CG-3′7304–7285AF299292[6]
aph2-Ia-15′-GAG CAA TAA GGG CAT ACC AAA AAT C-3′731–755M13771[20]
aph2-Ia-25′-CCG TGC ATT TGT CTT AAA AAA CTG G-3′1235–1211M13771[20]

2.4 In vitro matings

Filter matings with E. faecium recipients were performed according to Clewell et al. [21] with the following modifications: 100 μl of overnight donor culture was mixed with 100 μl of overnight recipient culture. One hundred microliters of the mixed culture was dispensed onto a 47-mm sterile filter (0.2 μm, MFS, California, USA) that was placed on a 5% bovine blood agar plate. After allowing the mixed culture to dry into the filter, the plates were incubated at 37 °C. Transferability was determined after 24 h of incubation by placing the filter in 5 ml of 0.9% saline, suspending the culture by vortex mixing and plating on selective media. Plates contained 10 mg l−1 erythromycin, 25 mg l−1 fusidic acid and 25 mg l−1 rifampin for detecting transconjugants; 10 mg l−1 erythromycin for detecting donors; and transconjugants and plates contained 25 mg l−1 fusidic acid and 25 mg l−1 rifampin for detecting recipients and transconjugants. Two overnight cultures of approximately 108 CFU ml−1 with, respectively, donors and recipients were plated on selective plates containing 10 mg l−1 erythromycin, 25 mg l−1 fusidic acid and 25 mg l−1 rifampin. No growth were observed on these plates. Fusidic acid was provided by Leo Pharma (Denmark). All other antimicrobial agents were purchased from Sigma Chemical Company (St. Louis, USA).

2.5 In vivo matings

Six female NMRI mice (20 ± 2 g) (Taconic M&B, Ry, Denmark), caged two by two, with an intact normal flora were treated with 5000 mg l−1 streptomycin sulphate in the drinking water according to Hentges and co-workers [2224]. After 24 h, fecal samples were taken, homogenized in 0.9% (w/v) NaCl and plated on Bile Aesculin Azide agar plates (Difco, Detroit, USA) with either 10 mg l−1 erythromycin, 25 mg l−1 fusidic acid, 25 mg l−1 rifampin or 10 mg l−1 erythromycin or 25 mg l−1 fusidic acid and 25 mg l−1 rifampin. These samples were taken as negative controls to make sure that none of the mice contained enterococci resistant to any of the antibiotics used in the study. Bacteria were grown overnight in Todd–Hewitt broth (Difco, Detroit, USA) prior to inoculation. Mice were fed orally via tubes with a suspension of the recipient strain, E. faecium 64/3-RFS, (2 × 109 CFU in 200 μl 0.9% NaCl suspension). The recipient was allowed to establish in the intestine for 5 days before the donor strain E. faecium 160/00 (6 × 108 CFU in 200 μl 0.9% NaCl suspension) was given to the mice. The experiment proceeded for 11 days after inoculation of the donor strain. The cages were changed daily, and the mice continuously received drinking water containing streptomycin (5000 mg l−1). Fecal samples were collected on days 1, 3, 4, 7, 8, 9, 10, 11, 14, 16 and 18. The samples were taken from each cage, so one sample represented feces from both mice in one particular cage. Feces was diluted by 0.9% NaCl and spread on selective agar plates as specified above.

2.6 Plasmid profiles

Small-scale preparations of enterococcal plasmid DNA were purified by a modified alkaline lysis procedure [25]. Escherichia coli 39R861 was used as size marker. It has four plasmids (147, 63, 36 and 7 kb) [26]. The E. coli plasmids were purified by a method described by Kado and Liu [27].

2.7 Pulsed-field gel electrophoresis

Whole cell DNA was prepared according to Jensen [28]. A small slice of the agarose plug was digested with 20 U of the restriction enzyme SmaI (New England BioLabs Inc., Medinova Scientific A/S, Denmark) for a minimum of 4 h. Digested DNA was electrophoresed in a 1% agarose gel in 0.5× TBE (45 mM Tris, 45 mM boric acid and 1 mM EDTA) by use of a CHEF DRII apparatus (Bio-Rad, Munich, Germany). DNA fragments were separated with pulse times of 5–35 s for 24 h (temperature, 12 °C; voltage, 6 V cm1; angle, 120°). Lambda Ladder PFG Marker NO350S (New England BioLabs Inc., Medinova Scientific A/S, Denmark) was used as a molecular size-marker.

2.8 Southern hybridizations

Plasmid DNA was transferred to a nylon membrane (Hybond-N, Amersham Pharmacia Biotech UK Limited) by vacuum blotting. Digoxigenin-labeled DNA probes for erm(B) and aac(6′)-Ie-aph(2′)-Ia were prepared by PCR amplification, using primers ermB-1, ermB-2, aph2-Ia-1, and aph2-Ia-2 (Table 1) and the PCR Dig Probe Synthesis Kit (Roche Diagnostics GmbH, Mannheim, Germany). Hybridization was carried out at 60 °C and detection performed using the DIG Nucleic Acid Detection Kit (Roche).

3 Results

3.1 Characterizaton of the donor

Susceptibility testing of the donor showed the following MIC values: erythromycin >32 mg l−1, gentamicin >2048 mg l−1, kanamycin >2048 mg l−1, penicillin >128 mg l−1, streptomycin >2048 mg l−1 and tetracycline >32 mg l−1. The donor strain was PCR-positive for tet(M), aac(6′)-Ie-aph(2′)-Ia, aadE, aphA-3 and erm(B). The genes aadE, aphA-3 and erm(B) were found together on the erm(B)-Tn5405-like element.

Both donor and recipient were sensitive to vancomycin, linezolid, penicillin, quinupristin/dalfopristin and chloramphenicol.

3.2 In vitro mating

In the in vitro mating experiments, a mean transfer frequency of 2 × 10−5 transconjugants/donor was obtained. The in vitro transfer frequency was calculated as the number of CFU on the erythromycin/rifampin/fusidin acid plate divided by the number of CFU on the erythromycin plate minus numbers of CFU on the erythromycin/rifampin/fusidin acid plate. The mating experiments were repeated 3 times.

3.3 In vivo conjugation

On day 2, the mice were inoculated with the recipient strain 64/3-RFS. After 5 days when the strain had established in the mouse intestine at a level about 107 CFU g−1 feces (Fig. 1), the donor strain 160/00 was inoculated to the mice. In two of the three cages transconjugants were observed 24 h after inoculation of the donor, the amount was 1000 and 1500 CFU g−1 feces, respectively. In the third cage transconjugants were not observed until 48 h after inoculation of the donor. This explains the large SEM at the first point of the graph for the transconjugants (Fig. 1). The transconjugants established in the intestine and the number of transconjugants was from about day 10 and throughout the experiment almost stable at about 106 CFU g−1 feces. Control experiments showed that no resistant enterococci were present in the intestine of the mice before and 24 h after the streptomycin treatment. No transconjugants were found in the period where the mice had only received the recipient strain.

1

Concentrations of recipients, donors and transconjugants in fecal samples. At day 2, the recipient was introduced (marked with the black arrow). From day 2 to day 7 the dots only represent recipients. After day 7, the dots represents both recipients and transconjugants. At day 7, the donor was introduced (marked with the white arrow). The figure shows the average values from three cages. Error bars represent SDs of the mean values.

3.4 Antimicrobial susceptibility testing

In total, 30 in vitro transconjugants (10 isolates were randomly selected from each of the three in vitro matings) and 40 in vivo transconjugants (two per cage per day) were tested for their antimicrobial susceptibility. All transconjugants were resistant to erythromycin (MIC > 32 mg l−1) and kanamycin (MIC > 2048 mg l−1) and sensitive to penicillin (MIC=4 mg l−1) and tetracycline (MIC=1 mg l−1). 26 out of 30 in vitro transconjugants and all in vivo transconjugants were also resistant to high levels of gentamicin (MIC > 500 mg l−1).

3.5 PCR verification of transconjugants

All the transconjugants were examined for the presence of erm(B) and aac(6′)-Ie-aph(2′)-Ia. The results were in accordance with the susceptibility testing. In all in vivo transconjugants and in 26 out of 30 in vitro transconjugants both erm(B) and aac(6′)-Ie-aph(2′)-Ia were present. In the last four in vitro transconjugants only erm(B) was present. Five in vivo and five in vitro transconjugants resistant to erythromycin and kanamycin were tested for the entire erm(B)-Tn5405-like element with the primers listed in Table 1. They were all PCR-positive for the entire erm(B)-Tn5405-like element.

3.6 Pulsed-field gel electrophoresis

PFGE was performed on the donor, the recipient and the aforementioned 30 in vitro and 40 in vivo transconjugants. All the transconjugants had a genetic profile similar to the recipient. All in vivo transconjugants and half of the in vitro transconjugants had an additional band at approximately 190 kb. One transconjugant had an additional band at approximately 210 kb. The remaining transconjugants had the same genetic profile as the recipient (Fig. 2).

2

PFGE profiles of donor, recipient and selected transconjugants: 1, molecular size marker; 2, donor; 3, recipient and 4–9, transconjugants. All transconjugants had a profile similar to the recipient. The transconjugant in lane 4 had a profile with an additional band at approximately 210 kb marked with a black arrow. PFGE profiles of the transconjugants in lanes 5, 6, 8 and 9 had an additional band at approximately 190 kb marked with a white arrow.

3.7 Plasmid profiles and Southern hybridizations

Plasmid profiles followed by Southern hybridization were performed on the donor, the recipient and two selected transconjugants. The two transconjugants were chosen based on their different antibiotic resistance profiles. One was both erythromycin- and gentamicin-resistant, while the other was erythromycin-resistant only. Both the donor and the two transconjugants contained several plasmids (Fig. 3). This indicates that more than one plasmid was transferred during the conjugation experiments. Hybridization showed that the donor contained two very large plasmids (>147 kb), designated pCAL1 and pCAL2, which hybridized with the erm(B) probe. The largest of the two plasmids, pCAL1, also showed hybridization to the aac(6′)-Ie-aph(2′)-Ia probe. The exact size of these two very large plasmids is unknown as the largest band in the molecular size marker (E. coli 39R861) was 147 kb and the plasmids were larger than this. The transconjugant that was both erythromycin- and gentamicin-resistant showed hybridization to both the erm(B) probe and the aac(6′)-Ie-aph(2′)-Ia probe to a plasmid with the same size as pCAL1. The transconjugant that only was erythromycin-resistant showed hybridization to the erm(B) probe on a plasmid with the same size as pCAL2 (Fig. 3).

3

(a) Plasmid profiles of the donor and two transconjugants. (b) Southern hybridization of the gel in (a) with erm(B) probe: 1, molecular size marker plasmids from E. coli 39R861; 2, donor; 3, erythromycin-resistant transconjugant; and 4, erythromycin and gentamicin-resistant transconjugant. (c) Plamid profiles of two transconjugants and the donor. (d) Southern hybridization of the gel in panel (c) with aac(6′)-Ie-aph(2′)-Ia probe. 5 Molecular size marker plasmids from E. coli 39R861; 6, erythromycin-resistant transconjugant; 7, erythromycin and gentamicin-resistant transconjugant and 8, donor.

4 Discussion

High-level gentamicin-resistant enterococci have been isolated from retail meat and the same PFGE-type has been isolated from a grocery store chicken and a human indicating the spread of high-level gentamicin-resistant enterococci from animals to humans through the food supply [29]. Conjugative transfer of resistance genes might also take place in the gut of animals and humans since enterococci are known to harbor transferable genetic elements.

In vivo transfer experiments were carried out in the intestine of streptomycin-treated mice as previously described by Hentges and co-workers [2224]. The recipient was allowed to establish in the intestine for 5 days before the donor strain was inoculated. Twenty four hour after inoculation of the donor, transconjugants could be recovered from feces in two out of three cages. A possible explanation for the delay in the last cage could be that the mice have a different transit time of the bacteria through the intestine. After 48 h, the number of transconjugants established at a level about 106 CFU g−1 feces in all three cages. This level of transconjugants remained almost stable throughout the experiment that lasted for 18 days. Conjugal transfer of resistance genes between enterococci in gnotobiotic animals has previously been described [14,16]. In both of these studies the level of recipients, donors and transconjugants was higher than in this study. A possible explanation for the higher number of transconjugants could be the use of gnotobiotic animals. Gnotobiotic animals have no colonization resistance, whereas streptomycin-treated mice still have some colonization resistance even though it is decreased [22].

The donor strain established at a higher level in the intestine of the mice than the recipient strain (Fig. 1). A possible explanation for this could be that the two strains did not display the same growth rate in vitro. The donor strain had a shorter generation time than the recipient strain (data not shown).

The high ability to colonize and the high level of transconjugants observed in the intestine of the streptomycin-treated mice indicates that transfer of resistance genes might also take place in the intestine of antibiotic-treated humans and animals. Whether this transfer also can occur in humans or animals with a undisturbed normal flora is, however, not known. Transient colonization of a vancomycin- and streptogramin-resistant E. faecium strain from retail meat has been shown in the intestine of humans, but this study was not designed to measure the transfer of resistance genes [11]. Other studies have also shown transient colonization of enterococci after ingestion of food containing these bacteria [10,30].

Co-transfer of the erm(B)-Tn5405-like element and the aac(6′)-Ie-aph(2′)-Ia gene was observed in all transconjugants from this in vivo conjugation experiment. To our knowledge this is the first co-transfer of the erm(B)-Tn5405-like element and aac(6′)-Ie-aph(2′)-Ia. Transfer of several resistance genes at the same time by in vitro and in vivo conjugation in enterococci has previously been shown. Co-transfer of erm(B) and vanA, vat(D), and vat(E), respectively, has been described [16,31,32]. Co-transfer of vancomycin and gentamicin resistance by in vitro conjugation has been described too [33].

PFGE showed that the transconjugants had a similar genetic profile as the recipient, which indicates that conjugation took place. An additional band at approximately 190 kb was observed in many of the transconjugants. It was not possible to see any connection between the transconjugants different plasmid profiles and different resistance- or PFGE-profiles. Very large plasmids containing several resistance genes and in vitro co-transfer of these resistance genes has previously been described in E. faecium isolates by Werner et al. [34].

In our study erm(B)-Tn5405-like element and aac(6′)-Ie-aph(2′)-Ia were found on plasmids and co-transfer of erm(B)-Tn5405-like element and aac(6′)-Ie-aph(2′)-Ia was shown both in vitro and in vivo in the intestine of streptomycin-treated mice. Several resistance genes were transferred at the same time. This indicates that the passing of enterococci carrying conjugative plasmids represents a potential source of spread of resistance genes to the indigenous intestinal flora in animals as well as humans. This could cause treatment problems if resistant enterococci from retail meat could transfer their resistance genes to hospital-adapted and may be more virulent enterococci in the intestine of humans.

Acknowledgements

Britta Pedersen, Michael Schou Agnild, Stine Frese-Madsen, Mette Skafte Thomsen, Frank Hansen, Tina Sommer Bisgaard and Dorte Truelsen are thanked for their technical assistance. This work is part of The Danish Integrated Antimicrobial Resistance and Research program (DANMAP).

References

  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
  20. [20].
  21. [21].
  22. [22].
  23. [23].
  24. [24].
  25. [25].
  26. [26].
  27. [27].
  28. [28].
  29. [29].
  30. [30].
  31. [31].
  32. [32].
  33. [33].
  34. [34].
  35. [35].
View Abstract