Escherichia coli is a species of bacteria that forms an essential part of the gut microbiome of many warm-blooded animals, including humans. Most strains are completely harmless to us, but some cause diseases including food poisoning and urinary tract infections. A group of antibiotics known as cephalosporins are often used to treat harmful E. coli infections, but some strains of the bacterium produce substances called extended-spectrum beta-lactamases (ESBL) that confer immunity to these antibiotics.
ESBL-producing E. coli used to be most commonly found in hospital environments, but they are now increasingly causing infections outside of hospitals, as well. The presence of such bacteria in the meat we consume is thought to contribute to the rising prevalence of community-acquired infections. After all, antibiotics are used heavily in the farming industry, so the development of resistant bacterial strains is to be expected.
Previous studies concluded that the ESBL-producing E. coli strains found in farm animals, retail meat and humans were extremely similar, suggesting that consumption of contaminated meat had allowed the bacteria to spread to humans. However, as Mark de Been, a bioinformatician at the Utrecht Medical Centre, explains, this research was based on studies of only a small part of the bacterial genomes. de Been is the lead author of a new study recently published in the journal PLOS Genetics, which shows that the bacteria found in animals and humans are actually significantly different.
More detail, better insights
Working with a team from the Netherlands, Spain, China, Norway and Denmark, Dr de Been sequenced the whole genomes of ESBL-producing E. coli from live pigs and poultry, supermarket chicken, and humans. This was feasible as genetic sequencing is increasingly cheap and fast to do – but sequencing methods still do not always generate complete sets of DNA. Instead, the output of sequencing consists of large numbers of short DNA sequences called ‘short reads’, leaving the researchers to puzzle out how these individual pieces fit together.
Software facilitates the assembly of short reads into a cohesive genome, but generally not all molecules fit together perfectly. This is because DNA contains many repeated elements that can either be part of the bacterial genome or belong to much smaller DNA molecules known as plasmids. Plasmids are small units of DNA that can replicate independently inside a suitable host cell; they can contain a wide range of genes, including ones that confer resistance to one or more antibiotics. Since plasmids can move between bacteria relatively easily, they can cause the transmission of antibiotic resistance to new bacterial populations.
Same message, different messenger
Once they had assembled the E. coli DNA from the various samples, Dr de Been and his colleagues found that the strains were actually quite different: it is highly unlikely that bacteria were transmitted between farm animals and humans, or between supermarket meat and humans. This is contrary to the findings of older studies, which had investigated just a few DNA segments that happened to be similar between ESBL-producing E. coli strains from animals and humans.
But if the bacterial strains are different, where did the genes for ESBL production in human E. coli come from?
Dr de Been and his team found that while the DNA of E. coli populations in farm animals, supermarket meat and humans was significantly different, the plasmids they found in each of the samples were remarkably similar. It appears that genes encoding ESBL production could, in fact, have come from animals – but it was plasmids, not bacteria themselves, that were transmitted. The plasmids could enter E. coli in the human gut and encode the ability to produce ESBL in their DNA.
ESBL production could come from animals. But does it?
The new study has showed that plasmids found in E. coli from human, farm animal and supermarket meat samples are very similar, suggesting that this is the route of transmission for the genes encoding ESBL production.
However, Dr de Been admits that we do not know if animals and meat are the only place that these plasmids could have come from. He notes that it is possible that they may be widely present elsewhere in the environment as a result of antibiotic overuse. That is, the similarity of the plasmids present in farm animals, meat and humans indicates correlation in the development of ESBL production – but further research is needed to ascertain whether there is clear causation, too.
We cannot turn back time on the decades of heavy antibiotic use in farming and medicine alike. Different plasmids could possibly encode many different forms of antibiotic resistance, not just the production of ESBL. If antibiotics are used less, then plasmids conferring antibiotic resistance will be less likely to be essential to an organism’s survival, so they may become less common. The Netherlands have recently cut back on their antibiotic usage significantly, and Dr de Been looks forward to measuring the impacts of this change by conducting a study similar to the current one in the future.
de Been, M., Lanza, V., de Toro, M., Scharringa, J., Dohmen, W., Du, Y., Hu, J., Lei, Y., Li, N., Tooming-Klunderud, A., Heederik, D., Fluit, A., Bonten, M., Willems, R., de la Cruz, F., & van Schaik, W. (2014). Dissemination of Cephalosporin Resistance Genes between Escherichia coli Strains from Farm Animals and Humans by Specific Plasmid Lineages PLoS Genetics, 10 (12) DOI: 10.1371/journal.pgen.1004776