Placing drug-resistant strains of E. coli into a broader context

DATE: 18/07/17
By: Teemu Kallonen, Julian Parkhill and Sharon Peacock

Escherichia coli is commonly carried in the human gut, and is the leading cause of bloodstream infection in England, elsewhere in Europe and the United States.

Several observations have made the study of E. coli an urgent priority. There has been a marked increase in the rate of E. coli bloodstream infection in recent years. For example, annual rates increased in England by 80% between 2003 and 2011 (from 16,542 to 29,777 cases), after which mandatory surveillance documented a further 10% increase between 2012/13 and 2014/15 (from 32,309 to 35,676 cases).

There has also been an emergence of E. coli strains that are resistant to multiple antibiotics. As the problem of antibiotic resistance and how to tackle this has become a global priority, this has gained increasing research attention. One of the highest profile drug-resistant E. coli strains to be studied so far is the bacterial type classified as sequence type (ST) 131.

A drawback of focusing on drug-resistant E. coli is that this only represents part of the story, since drug-resistant strains co-exist alongside their susceptible counterparts. In our study, we used whole genome sequencing to characterize an unbiased E. coli collection of over 1500 isolates, most of which had been collected by the British Society of Antimicrobial Chemotherapy over 11 years from patients with bloodstream infection in 10 hospitals across England.

Our study captured the year (2002) in which ST131 emerged in England (see figure below). We noted that within a short space of time, the number of ST131 isolates reached an equilibrium with other types. Around the same time, another type (ST69, not a multidrug resistant strain) also emerged, and again quickly reached an equilibrium within the overall population. These findings draw a sharp contrast with other drug-resistant pathogens such as methicillin-resistant Staphylococcus aureus (MRSA) or vancomycin-resistant E. faecium, where one or a limited number of types dominate the population that cause human infection.


Proportions of STs by year of isolation for our collection of E. coli from patients with bloodstream infection. Percentage is plotted by year ordered by frequency at the start of the study (most common at the bottom). The emergence of ST131 and ST69 can be observed in 2003 and 2002, respectively.

These findings suggest that the emergence of new types of E. coli may occur quite frequently, but does not necessarily indicate that these will out-compete other types to become a dominant cause of infection in the human population. This includes drug-resistant strains, a finding which has important implications for the control of drug-resistant infections. In particular, the development of control strategies should not restrict their focus to just drug-resistant strains.

The reason for this equilibrium may relate to the fact that all bacteria are constantly competing with others to survive. For example, E. coli has to compete with other bacterial strains of E. coli and other bacterial species in the human gut. It is likely that some bacteria carry a genetic repertoire (beyond antibiotic genes) that provide a fitness advantage. The pattern of equilibrium suggests that this advantage is stronger when the bacterium is rare, but is reduced as they become more common, a process called negative frequency-dependant selection.

E. coli type ST73 was the most common in our collection (ST131 was second commonest). ST73 was largely antibiotic-susceptible. We looked at whether drug resistance changed in these two types over time, and found this to be largely stable. This suggests that not every type is likely to develop resistance; furthermore, our findings indicated that most types were stably susceptible over the 11 years of study.

The fact that ST131 and ST73 were the most common types in the collection allowed us to compare whether there were specific genes that could explain their biological success. We found genes that were specific to ST131 and highly related types, and different genes that were specific to ST73. Their link to success can only be speculative based on sequence data, and these findings warrant further experimental studies to test whether they go some way to explaining their ability to be maintained in the human gut and elsewhere.

We did not look at E. coli types being carried by humans but not causing disease, which limits our ability to relate these findings to healthy people. But our findings provide some reassurance, at least in our setting, that ST131 associated with the severe end of the infectious disease spectrum (bloodstream infection) is in equilibrium with the overall population of E. coli strains, rather than increasing over time.

About the authors:

Dr Teemu Kallonen previously worked as a senior bioinformatician/postdoctoral researcher at the Wellcome Trust Sanger Institute and is currently working as a postdoctoral fellow in the Department of Biostatistics at the University of Oslo, Norway. He works with whole genome sequenced Enterobacteriaceae to investigate their evolution, virulence and resistance to antimicrobials.

Professor Julian Parkhill is Head of Infection Genomics and Senior Group Leader at the Wellcome Trust Sanger Institute, and Honorary Professor of Microbial Genomics at the University of Cambridge. His group is using high-throughput genomic approaches to understand the evolution of bacterial pathogens on short and long timescales; how they transmit between hosts on a local and global scale, how they adapt to different hosts and how they respond to natural and human-induced selective pressures.

Professor Sharon Peacock is an Honorary Faculty member at the Wellcome Trust Sanger Institute, an honorary Senior Research Fellow at the University of Cambridge and Professor of Microbiology at the London School of Hygiene & Tropical Medicine and University College Hospital. She has a long-term interest in bacterial infection, including antimicrobial resistance. Her group are currently using whole genome sequencing to study the reservoirs from which humans may acquire antibiotic-susceptible and drug-resistant bacteria.

Related publication:

Teemu Kallonen et al. (2017) Systematic longitudinal survey of invasive Escherichia coli in England demonstrates a stable population structure only transiently disturbed by the emergence of ST131. Genome Research. DOI 10.1101/gr.216606.116

Further links: