22 April 2016
The news late last year that resistance to an antibiotic of last resort, colistin, had been found in China in two species of disease-causing bacteria collected from humans, animals and samples of raw meat naturally caused great alarm.
These bacteria, Escherichia coli and Klebsiella pneumoniae, belong to the Enterobacteriaceae, a large group of Gram-negative bacteria that include many species capable of causing disease in humans; Salmonella also belong to this group.
Even more concerning was the implication that the resistance gene, called mcr-1, had spread between these bacteria via plasmid transfer. Other antibiotic resistance genes are also found on plasmids, circles of DNA present in bacteria that can be transferred between members of the same species, or between different species. The concern is that Enterobacteriaceae that already have plasmid resistance genes to other last-resort antibiotics, particularly the carbapenems, might also gain mcr-1 carrying plasmids, leaving the medical system to deal with the nightmare scenario of untreatable bacterial infections.
Understanding the scale of the resistance problem requires a range of approaches including surveillance and access to a large selection of samples taken from patients, animals and food products. Fortunately, organisations within the UK are able to study these issues due to the fact that Public Health England (PHE) has been whole genome sequencing (WGS) Salmonella spp. isolates since 2012 and a range of other bacteria including E. coli and Klebsiella spp. since 2014. A WGS study published this week in the Journal of Antimicrobial Chemotherapy is the first to assess the presence of plasmids containing mcr-1 in Enterobacteriaceae taken from patients and food samples in the UK.
The PHE team studied the genome sequences of over 24,000 isolates of Salmonella, E. coli, Klebsiella spp. and a small number of other gram-negative bacteria collected since 2012. Of these, 15 isolates contained the mcr-1 gene, 13 isolates from 12 humans and two from poultry meat. Three of the human isolates were E. coli while all of the other isolates were species of Salmonella. Eight of the human patients had recently travelled – two to Egypt, the rest to south-east Asia. The genetic backgrounds of the bacterial isolates and the mcr-1 plasmids were varied, indicating that these rare cases were mostly imported 'one-offs' rather than part of a wider outbreak. None of the carbapenem resistant bacteria analysed carried mcr-1, suggesting this nightmare scenario would be very rare if or when it occurs.
This study has highlighted many of the advantages of using WGS to understand antibiotic resistance and using this information to carry out surveillance of emerging threats. By storing the genome and associated patient data since the sequencing service began in 2012, researchers now have a valuable resource that they can use to explore new questions as they emerge, including whether new resistance mutations were present all along, now that we know what to look for. The fact that colistin resistance has been present at extremely low levels in the UK since at lea st 2012 highlights the need for vigilance: by continuing to sequence pathogen genomes to monitor the situation, the health system will be in a position to put extra infection control measures in place immediately should this resistance mutation become more common.
This data set will only improve as new samples are added, and enable us to quickly respond to new threats by allowing us to answer questions such as:
In order to fully realise the potential of this information, we must continue to support efforts to collect and sequence disease causing organisms, including strengthening the associated data and computational infrastructure that allow us to analyse these data and share it across national and international borders. T his work shows that pathogen whole genome sequencing is here and is important for allowing the health system to carry out vital surveillance that will allow it to rapidly respond to the emergence of unpredictable threats.