Introduction

Antibiotics, including bacitracin and tetracycline, are used as growth promoters in livestock because they improve feed conversion, increase animal growth, and reduce morbidity and mortality.

Favourable antimicrobial properties and the absence of major adverse side effects led to tetracycline’s extensive use in human and animal medicine. However, the use of this and other antibiotics as animal growth promoters is becoming increasingly controversial due to concerns that such use may lead to the emergence of antibiotic resistance in human pathogens.

Numerous tetracycline resistance (Tet^r) genes have been identified; however, those associated with efflux of tetracycline have been the most studied (McMurry et al. 1980). Efflux-associated resistance gene products are cytoplasmic membrane proteins that are energydependent tetracycline transporters, but their mechanism of cell protection is unknown (Speer et al. 1992).

In 1980, Walton and Laerdal reported that the percentage of Escherichia coli resistant to tetracycline that were isolated from piglets fed 100 ppm zinc bacitracin steadily dropped over an 11-week period from 50 per cent to 18 per cent. No such change occurred among pigs fed a diet devoid of antibiotics.

In another, study reported the same year, Bochner et al. (1980) described a method for the positive selection of loss of Tet^r. Certain lipophilic chelating agents, including picolinic, quinaldic and fusaric acids, when employed along with a Tet^r-inducing agent (heat-inactivated chlortetracycline) preferentially selected for tetracycline sensitive (Tet^s) E. coli. Bochner et al. suggested that efflux proteins acting as Tet^r factors lowered the effective concentration of one or more species of metal ions in the bacterial cytoplasmic membrane. By lowering the effective concentration of tetracycline-binding metal ions, the efflux protein would greatly slow tetracycline permeation of the bacterial cell. However, these membrane-bound metal ions likely serve an essential physiological function. Thus, induced Tet^r bacterial cells, having a lower effective metal ion concentration, will be more vulnerable than their Tet^s counterparts to chelating agents.

The purpose of the current study was to confirm Walton’s original observations under laboratory conditions, and to begin to identify the mechanism by which bacitracin selects for an increase in microbial tetracycline sensitivity. Because bacitracin, which is a mixture of various microbial compounds, has chelating ability, the author sought to determine whether the antibiotic had biological activities similar to the lipophilic chelators employed by Bochner et al. in their studies in 1980.

Methods

To determine whether culture in the presence of bacitracin provided a selective growth advantage to the Tet^s strains in a mixed population of Tet^s and Tet^r organisms, Tet^s and Tet^r E. coli or Salmonella strains were cultured overnight in Luria-Bertani (LB) or LB/T (15 μg/mL of tetracycline) broth respectively at 37°C, mixed 1:1, serially diluted, then cultured on LB agar, and LB agar containing 195-780 μg/mL bacitracin [¼ to ½ the minimal inhibitory concentration (MIC) of each strain].

Since tetracycline resistance is an inducible trait, in some experiments autoclaved (thus inactivated) chlortetracycline was included as an inducing agent in the selection medium which also contained bacitracin to generate the membrane modifications associated with tetracycline resistance as described by Bochner et al. (1980).

Approximately 120 colonies were randomly selected from each medium and tested for sensitivity to tetracycline (15 μg/mL). Strains used in this study included commercially available E. coli strain JM 109 and porcine non-pathogenic isolate G58-1 (both Tet^s) and their pBR322 plasmid-transformed constructs (both Tet^r).

The author also utilized a number of field isolates of Tet^s and Tet^r E. coli and Salmonella. For some experiments, Tet^s clones of field strains were derived from Tet^r isolates by the methods of Bochner et al. (1980) to create isogenic strain pairs.

To assess cumulative temporal effects of exposure to bacitracin, cultures in one experiment were serially passaged on antibiotic-containing and antibiotic-free medium for 14 days, then tested for tetracycline sensitivity.

In a second study, 11 pigs (mean starting weight 1.86 kg) were weaned, and challenged with enterotoxigenic E. coli (ETEC) strain 3030-2 (O157:K88/LT/STb; Tet^r), then fed a standard starter diet containing 0.25 kg/ton bacitracin (5 pigs), or no antibiotic (6 pigs). Pigs were monitored by changes in weight, signs of clinical disease and the shedding of ETEC for 16 days post-weaning.

Results

Mixtures of Tet^s and Tet^r isogenic E. coli strains cultured in the presence of bacitracin in selective medium (containing inactivated chlortetracycline as an inducer) resulted in a significant reduction in the Tet^r isotype.

Similar results were observed when wild-type field isolates were either randomly paired or pooled (Figure 1). Similar results were not obtained for identically cultured Salmonella strains (data not shown). However, when the same mixtures of Salmonella were cultured in LB broth containing bacitracin at ¼ or ½ its MIC, selection for tetracycline sensitivity did occur (data not shown). When the JM109 E. coli and its pBR322 (Tet^r) isogenic partner were serially passaged in medium containing bacitracin, JM109 gradually came to dominate the mixture. Relative concentrations of the isogens remained the same when the bacteria strains were cultured in antibiotic-free medium (Figure 2).

Figure 1. Relative concentration (ratio) of tetracycline sensitive colonies to total E. coli colonies grown out from 1:1 mixtures of Tet^s and Tet^r strains when cultured in the presence or absence of selection medium containing 780 ìg/mL bacitracin. Bars represent portion of total bacteria represented by Tets strains: 1: G58-1 and G58-1/pBR322; 2: JM109 and JM109/ pBR322; 3-4: Tet^r strain and isogenic Tet^s clones; 5: culture mixture of 9 (Tet^s) and 9 (Tet^r) strains. In each case, relative concentrations of Tet^s cultures were higher in selective medium containing bacitracin; T-test (P < 0.05); 3 replications per experiment. Legend: control = Luria-Bertani broth without antibiotics; selection medium with bacitracin and inactivated chlortetracycline.

Figure 2. The ratio of tetracycline-sensitive colonies and resistant colonies for JM109 and JM109/pBR322 grown in LB broth with or without ½ of MIC bacitracin over a 14 day experimental period. Note: straight lines represent regression lines while curved lines represent actual data.

No clinical disease was observed in piglets on starter feed either containing bacitracin or no antibiotic. There was no statistically significant difference in weight between pigs in the bacitracin and non-bacitracin groups during the 16-day observation period, but differences approached significance (P<0.10) by day 16. The concentration of the ETEC strain was about the same in animals of the two treatment groups for the first four days post-challenge. Thereafter, ETEC in the unmedicated pigs was significantly higher (P<0.05; Figure 3). ETEC became undetectable in faeces from pigs of either group after day 11.

Figure 3. ETEC Strain 3030-2 shedding curve in feeder pigs fed or not fed bacitracin. No difference days 1-4 post-challenge; significant difference days 5-11 post-challenge (t-test, P<0.05).

Conclusions

In this study, the author found that cultivation of a wide variety of E. coli and Salmonella strains in the presence of sub-inhibitory concentrations of bacitracin favoured selection of Tet^s over Tet^r organisms.

These findings support observations of Walton made nearly 30 years ago in the animal studies that the feeding of bacitracin may have the beneficial effect of decreasing the percentages of populations of enteric organisms possessing tetracycline resistance genes.

Further, this animal study may suggest that bacitracin in piglet feed at growth promoting concentrations may have an inhibitory effect on Tet^r E. coli. However, further studies are required to establish the relationship between Tet^r and microbial inhibition.

References