Faecal carriage of extended-spectrum β-lactamase-producing and AmpC β-lactamase-producing bacteria among Danish army recruits


Corresponding author: A. M. Hammerum, Department of Microbiological Surveillance and Research, Statens Serum Institut, Artillerivej 5, DK-2300 Copenhagen S, Denmark
E-mail: ama@ssi.dk


Clin Microbiol Infect 2011; 17: 566–568


During May and June 2008, 84 Danish army recruits were tested for faecal carriage of extended-spectrum β-lactamase (ESBL)-producing and AmpC β-lactamase-producing bacteria. Three ESBL-producing (CTX-M-14a) Escherichia coli isolates, two AmpC-producing (CMY-2) E. coli isolates and one AmpC-producing (CMY-34) Citrobacter freundii isolate were detected. Two of the CTX-M-14a E. coli isolates had similar pulsed-field gel electrophoresis and multilocus sequence typing profiles, indicating the same origin or transmission between the two army recruits. The blaCTX-M-14a genes were transferable to an E. coli recipient. These commensal bacteria therefore constitute a reservoir of resistance genes that can be transferred to other pathogenic bacteria in the intestine.

Extended-spectrum β-lactamases (ESBLs) are increasingly being reported in Europe. In human patients, ESBLs are mostly found in Escherichia coli isolates causing community-acquired infections; in particular, CTX-M enzymes are increasingly being detected [1]. As ESBL-producing bacteria have also been detected in food and animals, there might be a possibility of transfer of the ESBL-producing bacteria, or their resistance genes, from animals to humans, either by direct contact or via the food chain [2–8]. It is therefore of interest to look for faecal carriage of ESBL-producing or AmpC-producing bacteria in healthy humans in the community. However, only a few studies have investigated this [9–12].

The aim of the present study was to screen for ESBL-producing or AmpC-producing bacteria among Danish healthy humans not recruited through the healthcare system, and to characterize the resistant isolates in relation to virulence genes and transferability of the resistance genes.

Between 20 May and 17 June 2008, 84 of 120 army recruits (70.0%) from one military barracks in Denmark agreed to participate in the study. Each person mailed a faecal sample to the Statens Serum Institut (SSI) and filled in a questionnaire with information on gender, age, medications, gastrointestinal symptoms and diseases, travel activities, food consumption and animal contact. The Scientific Ethics Committee for the Copenhagen and Frederiksberg municipalities approved the protocol prior to the investigation (KF 01-006/02). Only eight of the 84 recruits were women, and the median age was 21 years (range: 18–31 years). The recruits had been resident in the barracks for 6 months for training, but went home each weekend.

Half a gram of faeces was suspended in 5 mL of saline (0.9%). One hundred microlitres of the diluted faecal sample was spread on two SSI Enteric agar plates (SSI Diagnostika, Hillerød, Denmark), one supplemented with 2 mg/L cefotaxime and the other with 2 mg/L ceftazidime. Both plates were incubated in ambient air for 18 h at 35°C. Presumptive third-generation cephalosporin-resistant Enterobacteriaceae were subcultured on 5% blood agar plates (SSI Diagnostika). Only one colony per sample was investigated further. The isolates were identified with API 20E (BioMérieux, Marcy-l’Etoile, France) and retested for susceptibility to cefotaxime and ceftazidime by Etest (AB Biodisk, Solna, Sweden).

ESBL and/or AmpC phenotypes were tested with the NeoSensitabs double-disk method according to the manufacturer’s guidelines (Rosco Diagnostica A/S, Taastrup, Denmark). On the basis of the obtained phenotype, the presence of TEM, SHV, CTX-M, OXA, MOX, CIT, DHA, ACC, EBC and FOX β-lactamase genogroups was investigated by PCR [13] (F. Hansen, unpublished data). Positive amplicons were sequenced. E. coli isolates were typed by pulsed-field gel electrophoresis and multilocus sequence typing [14,15]. MICs were determined for chloramphenicol, ciprofloxacin, gentamicin, nalidixic acid, streptomycin, sulphamethoxazole, tetracycline and trimethoprim (Trek Diagnostic, East Grinstead, UK). Data were interpreted with the EUCAST breakpoints.

The phylogenetic background (A, B1, B2, D) of the E. coli isolates was determined by triplex PCR with three DNA markers [16]. The results obtained allowed classification of the isolates into the four major E. coli phylogenetic lineages or as non-typeable isolates according to Gordon et al. [17].

The E. coli isolates were investigated for the extraintestinal pathogenic E. coli (ExPEC)-related virulence genes kpsMII, papA, papC, iutA, sfaS, focG, afa and hlyD (unpublished data).

In vitro matings were obtained according to Trobos et al. [15]. The recipient strain was a rifampin-resistant mutant of E. coli K-12 (MG1655). The selective plates used were Luria–Bertani agar (Sigma-Aldrich, Brøndby, Denmark) supplemented with: (i) 2 mg/L cefotaxime and 50 mg/L rifampin; (ii) 2 mg/L cefotaxime; or (iii) 50 mg/L rifampin.

From faecal samples of six male army recruits, five E. coli isolates and one Citrobacter freundii isolate were obtained (one isolate per person) (Table 1). Three of the five E. coli isolates had an ESBL phenotype and were positive for CTX-M-14a. The two other E. coli isolates had an AmpC phenotype and were positive for CMY-2. The C. freundii isolate had an AmpC phenotype and was positive for CMY-34 (Table 1). Two of the ESBL-producing E. coli isolates (3807X08 and 3827X08) had identical pulsed-field gel electrophoresis, multilocus sequence typing and resistance profiles. Both isolates were obtained from faecal samples sent within 1 week; it is therefore likely that these isolates had the same origin or that there had been transmission between these two army recruits. Only one of the recruits had been in contact with animals. CTX-M-14 has also been detected in faecal samples from healthy humans in Spain, indicating a community reservoir of this enzyme in other countries [10–12].

Table 1.   Description of six army recruits with extended-spectrum β-lactamase (ESBL)-producing or AmpC-producing bacteria
IDBackground dataSpeciesPhenotypeGenes encoding the β-lactamasesMLST typePFGE typeResistence to non-β-lactamsPhylotypeExPEC-related genesTransfer rated
GenderAge (years)Intake of antimicrobial agentsaMeat eatenbContact with animalsc
  1. ExPEC, extraintestinal pathogenic E. coli; MLST, multilocus sequence typing; NT, non-typeable; PFGE, pulsed-field gel electrophoresis.

  2. aWithin 6 months before the faecal sample was taken.

  3. bPork, poultry meat and beef on a reguar basis.

  4. CWithin 1 week before the faecal sample was taken.

  5. dTransconjugants/donor.

3786Z08Male22PenicillinYesDogEscherichia coliAmpCbla CMY-2ST18221Chloramphenicol, spectinomycin, streptomycin, sulphamethoxazoleDiutA<1 × 10−8
3823Z08Male21DicloxacillinYesDogE. coliAmpCbla CMY-2ST18002Ciprofloxacin, nalidixic acid, spectinomycin, streptomycin, sulphamethoxazole, trimethoprimB1iutA<1 × 10−8
3826Z08Male20NoYesDog, cat, wild birds, hunting animalsCitrobacter freundiiAmpCbla CMY-34 None <1 × 10−8
3792X08Male21NoYesDogE. coliESBLbla CTX-M-14aST6413NoneB17.6 × 10−4
3807X08Male19NoYesNoneE. coliESBLbla CTX-M-14aST16314NoneNT4.7 × 10−4
3827X08Male19NoYesDog, cat, reptiles, parrotE. coliESBLbla CTX-M-14aST16314NoneNT4.9 × 10−4

Two isolates (3786Z08 and 3823Z08) contained the iutA gene, whereas the other E. coli isolates were negative for all virulence genes tested. None of the five E. coli isolates was an ExPEC isolate according to the molecular definition of ExPEC isolates [18]. Urinary tract infections are most often associated with phylogroup B2 and, to some extent, with phylogroup D. One of the tested E. coli isolates belonged to phylogroup D, whereas none of them belonged to phylogroup B2 (Table 1). Rates of transfer of phenotypic resistance to the E. coli and C. freundii donors are presented in Table 1. The blaCTX-M-14a genes had a high transfer rate.

The six army recruits had not been hospitalized within 1 month before the faecal samples were taken, and nor had they travelled abroad within 3 months prior to sampling. All had been eating meat (pork, poultry meat and beef) on a regular basis (Table 1). Two of the six army recruits had been treated with a penicillin compound (penicillin and dicloxacillin, respectively) within 6 months before sampling. Such treatment might have selected for the resistant E. coli isolates. Five recruits had been in contact with companion animals or hunting animals, but not with food animals, within the last week before sampling.

It was not possible to screen the meat that the recruits had been eating or the animals that they had been in contact with for ESBL/AmpC-producing bacteria, but a common source seems likely.

The investigated E. coli isolates were not virulent, and the ESBL-producing isolates were susceptible to all of the non-β-lactam antibiotics tested, but, because of their excellent ability to transfer resistance genes, they may act as donors of ESBL genes for other more virulent and resistant E. coli isolates.

In summary, even though the ESBL prevalence in hospitals is low in Denmark as compared with southern European countries, six of the 84 studied Danish army recruits (7%) harboured ESBL-producing or AmpC-producing bacteria in their gut, which indicates a human faecal reservoir in the community.


Part of this study was presented in the DANMAP report 2008. K. S. Pedersen, F. Hansen and K. Racz are thanked for excellent technical assistance at SSI. We thank H. O. Jørgensen (Danish Armed Forces Health Services), I. M. Giversen (Danish Armed Forces Health Services) and E. Døssing (Infirmeriet Holstebro, Jyske Dragonregiment) for enabling contact with the recruits. We are grateful to the army recruits for their participation.

Transparency Declaration

This work was supported by the Danish Ministry of Science, Technology and Innovation and the Danish Ministry of Health and Prevention as part of the Danish Integrated Antimicrobial Resistance Monitoring and Research Programme (DANMAP).