OUP user menu

Low level of cross-resistance between triclosan and antibiotics in Escherichia coli K-12 and E. coli O55 compared to E. coli O157

Maria Braoudaki, Anthony Craig Hilton
DOI: http://dx.doi.org/10.1111/j.1574-6968.2004.tb09603.x 305-309 First published online: 1 June 2004


Misuse of biocides has encouraged the emergence of resistance and cross-resistance in certain strains. This study investigated resistance of triclosan-adapted Escherichia coli K-12 and E. coli O55 to antimicrobial agents and compared these to E. coli O157:H7. Cross-resistance in E. coli K-12 and E. coli O55 was observed however to a lesser extent than in E. coli O157:H7. Triclosan-adapted E. coli K-12 demonstrated cross-resistance to chloramphenicol, whereas triclosan-adapted E. coli O55 exhibited resistance to trimethoprim. In comparison, E. coli O157:H7 was resistant to chloramphenicol, tetracycline, amoxicillin, amoxicillin/clavulanic acid, trimethoprim, benzalkonium chloride and chlorohexidine suggesting strain specific rather than general resistance mechanisms.

  • Triclosan
  • Resistance
  • Escherichia coli
  • Antibiotic
  • Biocide

1 Introduction

Antibiotics and biocides used to be extremely effective in combating bacterial pathogens, however, their current effectiveness is compromised due to sustained misuse. It is estimated that between 1992 and 1999 over 700 consumer products with antibacterial properties, the vast majority of them containing triclosan entered the consumer market [1]. Triclosan is a broad spectrum antimicrobial agent used in such products as hand soaps, lotions, fabrics, plastics and toothpastes [2,3] and over the last 30 years has become the most widely used bisphenol [4]. In the case of triclosan especially, current reports strongly suggest that inappropriate administration could select for a more generalised resistance. This has been demonstrated in a variety of different strains including Pseudomonas aeruginosa[5], Escherichia coli[6,7] and Salmonella enterica[8] among others. In 1998, McMurry et al. [2] suggested that triclosan acted on a specific bacterial target rather than as a non-specific biocide, which could facilitate the acquisition of bacterial resistance. As a consequence, bacterial pathogens do not only survive the threat of antibiotics and biocides they may also thrive [9].

The basic mechanisms of action of antibiotics are generally well documented compared with those of biocides [10]. However, with respect to Gram-negative bacteria it has been proposed that antibiotics and biocides share common mechanisms of resistance [5,9,11,12]. Research in this area suggests that the up-regulation of multidrug efflux systems are recognised as resistance determinants [13,14], capable of accommodating both antibiotics and biocides [12] and can provide cross-resistance to other drugs [8,14].

Previous work on E. coli O157 [8] suggested that strains acquired increased levels of resistance to triclosan following only two sub-lethal exposures. It was also found that triclosan-resistant strains repeatedly demonstrated decreased susceptibility to a wide panel of antimicrobial agents including chloramphenicol, tetracycline, amoxicillin, amoxicillin/clavulanic acid, and trimethoprim as well as to biocides benzalkonium chloride and chlorhexidine. Given the additional genetic material known to be harboured by E. coli O157 compared to ancestral strains of E. coli[15] it was interesting to investigate if the rapid development of triclosan-resistance was E. coli O157 specific or a property expressed by other E. coli strains. In this study data is presented on cross-resistance of triclosan-adapted E. coli K-12, E. coli O157, E. coli O111 and E. coli O55 to a range of antimicrobial agents.

2 Materials and methods

2.1 Bacterial strains

E. coli O157:H7 were VT-negative strains obtained from the National Collection of Type Cultures; NCTC 12900 and NCTC 43888. E. coli O55:H7, O55:H29 and O111:H24 were clinical strains from Birmingham Heartlands Hospital, Birmingham, UK, isolated from stool specimens. E. coli K-12 (W3110) was obtained from the American Type Culture Collection (ATCC 27325) and E. coli K-12 (MRE600) was obtained from National Collection of Industrial and Marine Bacteria (NCIMB 10115). All strains were stored on Microbank beads (Pro-lab Diagnostics, Neston, UK) at −70 °C and cultured at 37 °C on nutrient agar (Oxoid, Basingstoke, UK) and in nutrient broth (Lab M, Lancashire, UK) where appropriate.

2.2 Antimicrobial agents and biocides

Antimicrobial agent tablets were supplied by Adatab, Merseyside, UK, and disks supplied by Oxoid, Basingstoke, UK. These included amoxicillin (AMX – 25 μg/ml), amoxicillin/clavulanic acid (AMC – 30 μg/ml), chloramphenicol (CHL – 30/ml), ciprofloxacin (CIP – 1 μg/ml), clindamycin (CLI – 2 μg/ml), colistin sulfate (CS – 25 μg/ml), fusidic acid (FD – 10 μg/ml), gentamycin (GEN – 10 μg/ml), rifampicin (RIF – 5 μg/ml), tetracycline (TET – 10 μg/ml), trimethoprim (TMP – 1.25 μg/ml), vancomycin (VAN – 5 μg/ml). Biocides benzalkonium chloride (BKC) (Sigma–Aldrich, London, UK) and chlorhexidine (CHX) (Sigma–Aldrich, Poole, UK) were supplied as laboratory standard powders of known potency and triclosan (TLN) (Aquasept, Oldham, UK) was purchased as a laboratory standard solution. All solutions were filter sterilised using a 0.2 μm cellulose syringe filter (Nalgene, Leicester, UK).

2.3 Minimum inhibitory concentration

The minimum inhibitory concentration (MIC) was determined using a standard broth microdilution method using a two-fold dilution of each antibacterial agent [8,16] and was established as the lowest concentration of the antibiotic/biocide inhibiting growth. Pre- and post-adapted strains were characterised by Random Amplification of Polymorphic DNA [17] assay to confirm strain continuity throughout strain passage.

2.4 Cross-resistance to antimicrobial agents and biocides

Cross-resistance towards a panel of antibiotics and biocides was determined by the Stokes' method [18]. Suspensions of the parent E. coli O55 and E. coli K-12 strains were inoculated over the central portion of the surface of separate Mueller–Hinton plates using a rotary plater (Denley Instruments Ltd., Sussex, UK) leaving an outer 1 cm ring. Adapted strains were inoculated onto the remaining perimeter of the plate and antibiotic discs placed at the interface. Plates were incubated overnight at 37 °C and examined for cross-resistance and antibiotic susceptibility by comparing zones of clearance around the disks. A difference of >2 mm was taken to indicate cross-resistance, which was confirmed using a standard broth microdilution method.

A range of antimicrobial agents was used, including broad spectrum antibiotics and those clinically against E. coli (CIP and TMP). In addition, VAN and FD were used as control agents as these are not active against E. coli. Cross-resistance of TLN-adapted strains to BKC and CHX was also determined.

3 Results

3.1 Bacterial adaptation

The progress of adaptation to TLN for all isolates investigated is shown in Fig. 1. E. coli O157:H7 (12900) data are reproduced from Braoudaki and Hilton [8]. All adapted strains showed an elevated level of resistance to TLN. The MIC of TLN increased from 0.25 to 2048 mg/l in E. coli K-12 W3110 and from 0.5 to 2048 mg/l in E. coli K-12 MRE 600. In E. coli O55:H7 the MIC increased from 1 to 2048 mg/l, whereas in E. coli O55:H29 the MIC increased from 0.25 to 2048 mg/l. The MIC of E. coli O111:H24 increased from 0.25 to 2048 mg/l. Both E. coli O157: H7 strains shared similar adaptation profiles and demonstrated increased resistance after one sub-lethal exposure to TLN, whereas all other strains tested required supplementary exposures. Molecular fingerprinting of all isolates by RAPD confirmed strain continuity throughout passage (data not shown).

Figure 1

Adaptation profiles of E. coli O157:H7 (12900), E. coli O157:H7 (43888), E. coli O55:H7, E. coli O55:H24, E. coli K-12 W3110 and E. coli K-12 MRE 600 to TLN. Both strains of E. coli O157 rapidly acquire increased levels of resistance following only one sub-lethal exposure to TLN, whereas other E. coli strains require additional exposure.

3.2 Cross-resistance

Resistance or sensitivity to an antibiotic/biocide for a representative isolate of each strain was determined by the zone of inhibition (measured in millimetres) around the impregnated disc which was placed at the interface between the pre- and post-adapted strains. Cross-resistance was recorded in incidences where a greater than 2 mm difference in the zone of clearing between parent and adapted strains was observed. In addition, the MIC of the antimicrobial agent towards the parent and adapted strain was determined by a standard broth dilution method. All cross-resistance data are summarised in Table 1. Cross-resistance to antibiotics and biocides was demonstrated only in a minority of cases; triclosan-resistant E. coli K-12 demonstrated decreased susceptibility only to CHL from a panel of different antimicrobial agents and triclosan-resistant E. coli O55 strains exhibited a decreased sensitivity to just TMP. E. coli O157 data is reproduced from a study employing identical experimental procedures to allow direct comparison [8]. Triclosan-resistant E. coli O157:H7 showed a decreased susceptibility to seven of the 15 antimicrobials tested. No activity, or resistance to VAN and FD was observed as expected.

View this table:
Table 1

MICs (mg/l) and zones of inhibition (mm) of parent and adapted strains of E. coli O157 adapted to TLN

E. coli K-12(mm)6/57/73/39/67/712/100/08/60/010/120/07/614/010/80/0
E. coli O55:H7(mm)12/1213/130/010/104/417/170/010/100/015/153/316/1416/014/100/0
E. coli O157:H7(mm)11/013/06/013/56/014/140/09/100/012/125/517/1411/413/00/0
  • Boldfaced data indicate cross-resistance.

  • a E. coli K-12 (W3110) and E. coli O157:H7 (12900) were tested.

4 Discussion

In this study resistance in E. coli O55 and E. coli K-12 was readily achieved by repeated passage in sub-lethal concentrations of TLN. Exposure to relatively low concentrations of TLN led to a high-level of resistance within four passages for both strains tested. In general, the adaptation profiles followed by E. coli K-12 and E. coli O55 share similarities to TLN-adapted E. coli O157 [8] as regards the fact that all strains were initially extremely sensitive to low concentrations of TLN. Subsequently, all acquired high level of resistance following only four sub-lethal exposures. The difference between the adaptation profiles of the non O157:H7 E. coli isolates compared to that of E. coli O157 is the speed with which E. coli O157 acquired resistance. It was repeatedly observed that following the first passage E. coli O157 became resistant to extremely high concentrations (2048 mg/l) of triclosan; however, this was not observed in E. coli K-12, E. coli O55 or E. coli O111.

Cross-resistance of triclosan-resistant E. coli to a panel of antibiotics and biocides was investigated in both E. coli K-12 W3110 and E. coli O55:H7. The data generated suggested that there was a degree of cross-resistance in E. coli K-12 W3110 and E. coli O55:H7, however to a lesser extent than that observed in E. coli O157:H7 [8]. More specifically, E. coli K-12 W3110 exhibited reduced susceptibility to CHL, whereas E. coli O55:H7 demonstrated cross-resistance to TMP, which is a clinically important used drug against E. coli. In comparison, TLN-resistant E. coli O157:H7 (12900) strains repeatedly showed decreased susceptibility to a range of antimicrobial agents, including CIP, TET and TMP, as well as to the biocides BKC and CHX. The lack of cross-resistance of the TLN-adapted E. coli K-12 W3110 and E. coli O55:H7 also suggested that the mode of action of TLN is not shared with the other biocides tested.

Differences in cross-resistance profiles between E. coli O157, E. coli K-12 and E. coli O55 suggest that strain specific rather than general mechanisms are underlying the resistance observed, some of which may be facilitated by the additional genes E. coli O157 is known to possess over E. coli K-12, accounting for an additional 1387 genes [15,19]. It is possible that some product of this additional coding capacity is contributing to the increased resistance observed.

To our knowledge the genome sequence of E. coli O55 has not yet been determined, however, it is noteworthy that evolutionary studies have shown that E. coli O157 strains are closely related and share a common ancestry with pathogenic E. coli O55 strains [15,20]. It is interesting, therefore, that E. coli O55 showed an adaptation and cross-resistance profile more similar to that of E. coli K-12 than E. coli O157. This suggests that if the enhanced resistance demonstrated by E. coli O157 was obtained by horizontal acquisition of DNA, it is something that has occurred relatively recently.

Another possible explanation for the rapid development of resistance is based on the mutator hypothesis, which may account for a speeded up evolutionary process in E. coli O157. LeClerc et al. [21] demonstrated that more than 1% of E. coli O157 strains had spontaneous rates of mutation 1000-times higher than typical E. coli strains [21].

Taken in consideration with work of McMurry et al. [2] who showed that mutations in the fabI gene leading to amino acid substitutions conferred resistance to triclosan it is possible these two mechanisms may contribute to the increased adaptation rate observed in E. coli O157. Contrary to this idea, Whittam et al. [22] found no evidence of a genome-wide increase in the mutation rate in pathogenic E. coli compared with E. coli K-12, however, they could not rule our the effectiveness of the mutator phenotype during short-term evolutionary periods.

The cross-resistance observed could be a result of the presence of active efflux pumps as has been proposed previously. According to Schweizer [1], TLN and antibiotics not only share multidrug efflux systems as common mechanism of resistance also cause expression of these efflux systems by selecting similar mutations in the respective regulatory loci. This is supported by other studies in Pseudomonas aeruginosa[5] in addition to E. coli[14], in which strains repeatedly express elevated levels of resistance to a wide range of structurally unrelated antibiotics which has been shown to result from increased levels of active efflux.

In conclusion, these results add to the growing body of evidence linking resistance to antibiotics and biocides, especially TLN, likely as a result of their continuous mis- and over-use. In addition, our data suggests that E. coli O157 not only possesses an enhanced virulence compared to closely related E. coli strains, but also an increased capacity to become resistant to the activity of TLN and other antimicrobial agents.


  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].
View Abstract