Genetic Characterisation of Resistance Markers in Sentinel Escherica Coli and Enterococcus in Farm Animals

Progress: Extensive use of antibiotics in the latter half of the 20th Century in both human medicine and animal husbandry has left a legacy of resistant bacteria that seriously threaten the continued efficacy of antibacterial chemotherapy. Bacteria can become antibiotic resistant in two principal ways; either through mutation of existing genes located on bacterial chromosomes or by acquiring additional DNA. Such additional DNA is normally acquired from other bacteria in the form of plasmids or transposons, which may encode one or more antibiotic-resistance genes. Since the main driver of the evolution of resistant-bacteria is the use of antibiotics, any action taken to reverse the seemingly inexorable increase in the numbers of resistant bacteria must include more sparing use of antibiotics. However, there is insufficient information to allow a confident prediction as to the consequences on resistant populations of bacteria of limiting the use of antibiotics. <P>
The project was designed to reduce this paucity of information; with specific reference to determining what types of antibiotic-resistance are the most stable in the absence of antibiotic use, by using both laboratory and animal models, as well as investigating naturally-occurring bacteria isolated from farm animals. Most of the work undertaken studied Escherichia coli, a common, mostly harmless bacterium that inhabits human and animal guts. However, certain pathogenic variants of this species can cause infections in humans such as diarrhoeal disease, urinary tract infections and invasive disease. Some work was also carried out on Enterococcus spp., bacteria that are usually harmless inhabitants of animal and human guts, but which can also cause serious infections in hospitalised patients and on Campylobacter spp., the most common cause of gastroenteritis in the UK. <P>
One of the central concepts surrounding the persistence of antibiotic resistance in the absence of antibiotic selective pressure, is the notion of biological fitness, a measure of how well a bacterium can grow and survive in its environment and compete with other micro-organisms. Conventional scientific thinking assumes that in the absence of antibiotic use, antibiotic-resistant bacteria will have a fitness disadvantage because expression of antibiotic resistance requires the use of extra resources and often alters the normal metabolism of the bacterial cell. Hence, it has been hypothesised that in antibiotic-free environments resistant bacteria will be out-competed and displaced by their sensitive counterparts, resulting in a reduction in their prevalence. In this project, we developed a series of experimental models to study this hypothesis. At first, we introduced nine well-characterised antibiotic-resistance coding genetic elements into an E. coli strain isolated from a pig and developed specifically to fit the requirements of our models. We measured the effects on the overall fitness of the strain resulting from the acquisition of these nine elements, by investigating the ability of the resistant variants to compete-against their sensitive counterpart in laboratory culture and by measuring the impact of antibiotic-resistance on the ability of the strain to colonise and persist in the pig gut. We found that 8/9 different antibiotic resistance elements studied had little or no negative effects on the ability of the bacterium to compete in the laboratory or persist in the pig gut, and in the case of a transposon conferring ampicillin resistance, acquisition of the element actually improved the bacterium’s fitness. In our second pig gut model, we investigated the ability of naturally occurring antibiotic-resistant E. coli, Enterococcus spp. and Campylobacter spp. to persist in the guts of pigs, after removal of antibiotic selection, and found that in general resistant-organisms representing all three bacterial genera persisted over the eight-week study period. Some decline in numbers of resistant bacteria was observed, but curiously, these were bacteria resistant to antibiotics other than those the pigs had been fed prior to the start of the experiment. These decreases appeared to be due to the displacement of certain strains of gut bacteria with others. <P>
We also used our pig-gut models to investigate the molecular reasons behind any loss of antibiotic resistance. In particular, we determined rates of retention of the antibiotic-resistance encoded by the genetic-elements we had introduced into our E. coli strain in the pig gut; and in the case of loss, the molecular factors responsible. On the whole, the introduced antibiotic resistance elements were extremely stable, and almost no outright loss of plasmids or transposons was detected. However, we found evidence that in some cases bacteria were able to switch off plasmid-encoded antibiotic resistance whilst retaining the plasmid and the intact genes required for resistance. This is a previously unreported phenomenon that we have termed gene-silencing, and appears to represent a previously uncharacterised mechanism of genetic control in bacteria. In our experiments, gene-silencing was triggered in the pig gut and was a property of the host bacterium rather than the acquired plasmid itself. It is relatively worrying as it suggests a reservoir of antibiotic-resistance may exist which cannot be detected by phenotypic screening, the most common method used for identifying antibiotic resistance. We therefore screened a number of E. coli and E. faecium isolated from commercial farm animals for the presence of expressed as well as silent antibiotic resistance genes. <P>We found some evidence that unexpressed antibiotic resistance genes are present in naturally occurring bacteria, although at a relatively low frequency of approximately 2%. Some of these genes contained mutations that explained why they were not expressed, while others appeared to be silent without any obvious genetic reason, as observed in our animal model. <P>
The results of this work suggest that mobile genetic elements carrying antibiotic-resistance genes impose little or no disadvantage on the bacteria carrying them. Hence, once established in the population, antibiotic-resistant bacteria are likely to persist in the absence of antibiotic selection. In addition to direct compensation of the costs associated with carriage of antibiotic-resistance, bacteria are also able to switch off resistance genes when not required, although we do not yet know whether such gene silencing is an evolutionary response designed to combat the fitness costs of resistance or whether it is a coincidental phenomenon. However, once characterised, it may be possible to chemically trigger the silencing mechanism as to switch off resistance genes at will, a strategy that would be of enormous and obvious veterinary and clinical benefit. The most imperative future aspects of this study, i.e. investigation of why silencing occurred and what the underlying mechanisms is, has already been funded by DEFRA (project code OD2022) and is currently under investigation. However, our studies on the persistence of antibiotic resistance also warrant further study. Although the models described here have provided valuable data concerning the factors that govern the persistence of antibiotic-resistant bacteria they are still somewhat preliminary and further detailed studies need to be undertaken. The obvious first steps would be to extend studies on the biological fitness of resistant bacteria, using other strains and species of bacteria and contemporary mobile genetic elements encoding antibiotic resistance, which would first need to be characterised in detail. <P>Once a deeper understanding of the biological factors governing the persistence of antibiotic resistance has been has been gained, focus can then shift to the molecular mechanisms underlying it, as well as the factors governing the transmission of antibiotic resistance between animals and between animals and humans.

<P>