A new weapon for clinical microbiology
In 2003 the first whole human genome was sequenced. It contained over three billion base pairs. This achievement rested on Sanger's pioneering DNA sequencing technique and marked the culmination of many years of work by more than 200 scientists around the world. The sequencing had been carried out as part of the Human Genome Project, officially launched in 1990 with US$3 billion from the US National Institutes of Health. Many of those who had campaigned for the Project believed it would pave the way to understanding the genetic code of life and open up a new chapter in medicine. It could take many more years, however, before researchers fully understand how instructions encoded in DNA shape health and disease. Such work will necessitate the investigation of genome samples taken from thousands, probably millions, of people. Nor will it be straightforward. Not only does the genetic profiling of individuals raise major ethical and legal concerns; many question the importance of genes on health compared to other factors such as environment and lifestyle. Often the genomic data does not reveal a neat relationship between genes and health. coct would perention.
One area, however, where DNA sequencing is already having an impact is in the field of clinical microbiology, a branch of medicine concerned with the prevention, diagnosis and treatment of infectious diseases. Clinical microbiology depends on the rapid characterisation of pathogen samples so that infected patients can be treated individually and the outbreak and spread of infectious diseases can be monitored. Until very recently, DNA sequencing was highly expensive, limiting its use on a routine basis in clinical microbiological laboratories. This situation, however, is beginning to change as a result of the fall in the cost of whole-genome sequencing (WGS).
A tool to track antimicrobial resistance
Where the technology is set to have the most dramatic impact is in the fight to control drug-resistant bacterial strains, which are increasingly a threat to public health globally. Traditionally most bacterial pathogens are identified by growing a patient's specimen in a culture, testing its susceptibility to antimicrobial drugs and comparing it with other bacterial strains. All these steps rely on the knowledge of the clinical microbiologist and specialised, species-specific methodologies. The whole process is labour intensive and time-consuming. Identification and susceptibility testing can take days, in the case of rapidly growing bacteria such as Escherchia-coli, and months in the case of slower growing bacteria such as Mycobacterium tuberculosis.
This image shows a culture of bacteria in a petri dish with four white paper discs impregnated with antibiotics. The clear concentric circles around the discs show where the antibiotics has leached out into the agar covering on the dish and inhibits bacterial growth. This is the conventional technique used to determine the susceptibility of bacteria isolated from infected patients to antibiotics. Credit: Wellcome Library.
A number of advances were made in clinical microbiology in the first decade of the 21st century to help in the identification of bacteria and in defining their susceptibility pattern, but the technology still has certain shortcomings. Typing pathogens, for example, remains difficult. This is important for working out whether a new outbreak of an infectious disease has occurred. Monitoring the mutation of bacterial strains is also difficult as is tracking their routes of transmission (Peacock, 2015).
Some headway, however, has been made as a result of the increasing availability of rapid sequencers and the development of large reference databases. Today the average cost of bench-top sequencers is US$125,000. Such machines can sequence several bacterial genomes in a day at a cost of about US$150 per sample. The new technology is providing the means to sequence whole microbial genomes from clinical samples and to pick up antibiotic resistance quickly across many different samples. Indeed, the sequencers and reference databases are increasingly becoming significant tools in the surveillance of drug-resistant pathogens and efforts to contain them (Didelot, Bowden, Wilson, et al 2012; Peacock, 2014).
WGS has proved useful in the management of the human immunodeficiency virus (HIV). Diagnostic laboratories are already sequencing the virus in samples taken from patients over their lifetimes to detect the emergence of drug resistant strains. This tool will be invaluable in future for determining and monitoring the outbreaks of other infectious diseases. One of the major difficulties when assessing an outbreak of disease is knowing whether the cases presented are caused by a single strain or signify an increased incidence of infection involving multiple strains. Distinguishing between these different types dictates what action is necessary. Hospital staff, for example, who face a disease outbreak in a ward caused by cross infection from a single strain need to undertake a deep cleaning of wards and the stringent isolation of patients. On the other hand, where the infection is not associated with an outbreak, other control measures are required (Robinson, Walker, Pallen, 2013; Peacock, 2014).
This shows the Wellcome Trust Sanger Institute which officially opened its doors in 1993. The Institute was funded by the Wellcome Trust and the Medical Research Council and was named after Fred Sanger. Since its launch the Institute has played a pivotal role in mapping, sequencing and decoding both the human genome and the genomes of other organisms. Credit: Wellcome Trust Sanger Institute.
One of the earliest applications of WGS was in tracking the spread of tuberculosis, a chronic disease that primarily affects the lungs. It is caused by a microorganism known as Mycobacterium tuberculosis. By the mid-twentieth century it was widely predicted that tuberculosis would become a disease of the past due the widespread success of antibiotics and vaccination. The disease, however, has made a dramatic resurgence since the 1990s, caused in part by the emergence of HIV and a rise in multidrug-resistant Mycobacterium tuberculosis strains. It had become so rampant by 1993 that the World Health Organisation (WHO) declared a global emergency in an effort to raise public and political awareness of the problem. TB remains a major global health problem to this day and is the second leading cause of death from an infectious disease worldwide. What is of major concern is the continuing rise of drug-resistant TB strains. A report put out by WHO in 2013 indicated that at least 3.6% of new cases and 20.2% of previously treated cases were caused by drug-resistant Mycobacterium tuberculosis Mycobacterium tuberculosis (WHO, 2013).
Since the mid-1990s automated DNA sequencing has been an important tool for investigations into the genetic basis of drug resistance in Mycobacterium tuberculosis. Research in this field was given a major boost in 1998 by the publication of the complete genome sequence for the Mycobacterium tuberculosis strain known as H37Rv. The sequence of the genome, which is 4,411,529 nucleotides, was determined by an international team headed by researchers from the Wellcome Trust Sanger Institute in Cambridge. This sequence has become an important reference point for researchers who continue to sequence other strains to trace the rise and spread of multidrug-resistant tuberculosis. Such sequencing is providing important insights into the biology, metabolism and evolution of the Mycobacterium tuberculosis pathogen (Kapur, Li, Iordanescu, 1994; Cole, Brosch, Parkhill, 1998; Ioerger, Feng, Ganesula, Chen, 2010; Wang, Jiao, Xu et al, 2013).
Drug-resistant tuberculosis is just one of the many infectious diseases where DNA sequencing is being used to track the rise of antimicrobial resistance. Another such case is that of Staphylococcus aureus, a bacterial species that causes many different health problems in humans. The species is associated with carbuncles and food-poisoning, as well as more serious complications arising from medical devices and wounds. It is also linked to life-threatening conditions such as blood poisoning (bacteremia), rare and potentially fatal heart infections (endocarditis), and a form of pneumonia which results in the death of lung tissue (necrotising pneumonia). Unfortunately it has become resistant to many widely used antibiotics, including penicillin, methicillin and more recently vancomycin. Known as MRSA, these bacteria are a common cause of many intractable infections in community and hospital settings.
In 2004 researchers from the Wellcome Trust Sanger Institute, Cambridge, and fellow researchers completed the first sequences of two disease-causing Staphylococcus aureusis strains. The first was a hospital-acquired MRSA strain, isolated from a 64- year old woman who had died after an operation, and the second was a community-acquired strain isolated from a nine-year old boy who had survived. Soon after this another three strains were sequenced. The sequencing was part of a long-term project to follow the evolution and virulence of the bacterium and identify the genetic mechanism behind its resistance to antibiotics (Holden, Feil, Lindsay, Peacock, 2004).
Electron micrograph of clusters of MRSA bacteria. Credit: Wellcome Library.
Knowledge from such sequencing is already being used in the clinical context. Its power is illustrated by the handling of a MRSA outbreak in a Special Care Baby Unit (SCBU) at the Rosie Hospital in Cambridge in 2011. All babies in the Unit are tested for MRSA on admission, and then once a week during their stay. In 2011 the hospital's infection control team spotted three babies in the unit who all tested positive for MRSA at the same time, and this prompted them to investigate its source. They combed through the hospital's testing records from the previous six months, and found that a number of babies had tested positive sporadically in this time, with some months passing between cases. From this information, it was difficult to determine whether what they were seeing was an outbreak of MRSA or merely a series of unrelated MRSA infections (Peacock, 2015).
Unable to answer the question with conventional epidemiological data and laboratory tests, the hospital's infection control team called on WGS support from a group of researchers at the Sanger Institute. Headed by Sharon Peacock, the team sequenced all the MRSA specimens isolated from the babies, and quickly established that the samples were all related at the genome level which suggested they were part of an outbreak. To understand what had happened, Peacock's team trawled through the hospital's microbiology laboratory database to see if they could find any other bacteria reported to have the same antibiotic susceptibility profile as the outbreak strain. Their aim was to find specimens that could be sequenced among those stored since the start of the putative outbreak. The database held records from about 800,000 clinical specimens processed each year, 30% of them originating from primary care. The team discovered that nearly 3,000 of these samples had been shown to be MRSA positive in 2011. More than a third came from 1,500 patients seen in outpatient and emergency departments and primary care surgeries. On sequencing these bacteria, they found that there had been an outbreak of a new strain, that it was twice as big as originally believed, and that it was also prevalent in the wider community. Without the sequencing the team would not have been able to link the cases in the hospital with those picked up in the community. The key issue now was to work out how the transmission had occurred (Peacock, 2012; Peacock, 2015).
Sharon Peacock, who headed up the investigation into the outbreak of MRSA at the Rosie Hospital. Credit: Peacock. Prior to her work in Cambridge, Peacock spent seven years directing a wide-ranging programme researching the prevalence of bacterial disease in South-East Asia at the Mahidol-Oxford Tropical Medicine Research Unit. Since her work in Cambridge she has been appointed the first director of the Bloomsbury Research Institute which is leading a programme to find new treatments, vaccines and diagnostics for the prevention of infectious diseases.
To pinpoint what was going on the team put the SCBU under close surveillance to see if they could identify any new carriers or people infected with MRSA. Soon after this, another baby tested positive for MRSA. The case was particularly striking because it occurred immediately after the SCBU had had a thorough clean and two months after the last case had tested positive. To find out whether one of the Unit's 154 staff members could be carrying some strain of the bacterium, the researchers secured their agreement to be swabbed. On sequencing the bacteria from these swabs, the team discovered a particular individual was carrying the MRSA that directly matched the outbreak strain. This explained the persistent outbreak over several months and the gaps between the positive testing of cases, as well as why another case had appeared following the deep clean of the SCBU. Once the staff member had been treated, the outbreak came to an end(Harris, Cartwright, Torok et al, 2013).
The work on the MRSA cases at the Rosie Hospital was one of the first occasions that WGS had been used to track down and halt the outbreak of an antimicrobial infection in a hospital. One of the advantages of the method was that it made it possible to reconstruct the history of the outbreak, and helped to identify the transmission route. The sequencing not only showed that the outbreak resulted from a new strain of MRSA, but also managed to link it to earlier infections on the ward and to trace its complex transmission pathways from babies to their mothers, from these mothers to mothers in the postnatal ward, and to partners of affected mothers. What the sequencing research also highlighted was the pitfall of concentrating merely on hospital-based infection control and pointed to the need to develop strategies that could take on board the important transmission dynamics between the community and healthcare institutions (Harris, Cartwright, Torok et al, 2013).
The MRSA case at the Rosie Hospital not only showed how effective WGS could be for controlling the spread of an antimicrobial infection, but also how cost-effective it could be. Overall the healthcare costs attributable to the outbreak in the Rosie Hospital were estimated to exceed £10,000. Yet, the cost of doing WGS on each MRSA isolated was £95 (Harris, Cartwright, Torok et al, 2013).
Since sequencing the MRSA outbreak at the Rosie Hospital, Peacock's team, together with researchers from Thailand and Australia have demonstrated the utility of WGS to track and identify MRSA on two intensive care units in an under-resourced hospital in north-east Thailand. One of the team’s striking findings from this project was how many different variants it detected circulating through the Thailand hospital at the same time, something conventional typing had failed to pick up. It also helped identify clinically important genes such as those coding for antiseptic resistance and antibiotic resistance. (University of Cambridge, 2014).
The work undertaken in Thailand is particularly important because of the especially heavy burden of MRSA and other similar antimicrobial resistant bacteria in middle and low income countries. The huge burden of resistant bacteria reflects the low barriers to transmission in such countries, brought about in part by the widespread use of antibiotics, lack of formal screening procedures for picking up MRSA and minimal measures in place to control it, such as effective hand-washing (Tong, Holden, Nickerson, Cooper, et al, 2014).
WGS has not only been demonstrably helpful in ending an outbreak of MRSA; this has also been shown in the case of Acinetobacter baumannii which plagued the Queen Elizabeth Hospital in Birmingham, England between July 2011 and February 2013. Commonly known as 'Iraquibacter' due to what seemed its sudden appearance in military treatment centres during the Iraq War that started in 2003, Acinetobacter baumannii is an opportunistic bacteria which causes ventilator-associated pneumonia and bloodstream infection in critically ill patients with compromised immune systems. A number of its strains are now drug-resistant.
The Queen Elizabeth hospital is a public hospital with 1,200 beds and often receives repatriated British military casualties. As in the MRSA case at the Rosie Hospital, WGS proved highly successful in tracking down the source of the Acinetobacter baumannii in the Queen Elizabeth Hospital. The first case detected at the hospital was that of a repatriated military patient who had suffered a blast injury in Afghanistan. Laboratory tests showed the strain to be resistant to multiple classes of antimicrobial drugs. Over the next 80 weeks a further 51 patients, civilians and military casualties, tested positive for the same bacterial strain. Sequencing work quickly established that the strain was distinct from other well-characterised strains. This indicated the infection was caused by a new strain (Halachev, M R, Chan, J Z M, Constantinidou, et al, 2014).
Initial epidemiological and genomic analyses suggested the infection was transmitted between patients on the same ward. Further sequencing, however, identified the same strain in samples taken from patients treated in other wards. This forced the hospital staff to look for alternative routes of transmission. They soon noticed that most of the affected patients had made numerous visits to an operating theatre specialising in burns treatment. Based on this, they closed the operating theatre down for deep cleaning with a chlorine-based treatment to eradicate any sources of the outbreak strain. No new theatre-acquired cases were noted in the subsequent six weeks and for a time it appeared the outbreak had been stopped. Soon, however, another case was picked up in a patient although initial epidemiological investigations did not indicate that this had been caused by direct transmission from any particular ward or operating theatre. However, further investigation of the epidemiological evidence along with the sequencing data revealed that the patient had picked up the infection after occupying a specialised burns-care bed previously occupied by an infected patient. This prompted the hospital to develop a decontamination protocol for this specialised type of bed. Over the next nine weeks the outbreak spread to a further dozen patients. Again WGS, used in conjunction with conventional laboratory and epidemiological investigations, tracked the route of transmission to the burns theatre which was again put through deep cleaning. Following this no new cases were found (Halachev, M R, Chan, J Z M, Constantinidou, et al, 2014).
As had been the case with MRSA at the Rosie Hospital, WGS, alongside epidemiological and laboratory analyses, proved indispensable in tracking the transmission route of the Acinetobacter baumannii in the Birmingham Hospital, and in halting its spread. Importantly it had the ability to follow the rapid evolution of the genome of the Acinetobacter baumannii in the hospital setting during the course of the outbreak (Halachev, M R, Chan, J Z M, Constantinidou, et al, 2014).
WGS as diagnostic tool?
WGS is clearly proving invaluable as a research tool for identifying and monitoring drug-resistant bacteria, and many researchers are now beginning to show how it might be used to improve diagnostics within the clinical setting. For the past hundred years the diagnosis of infectious diseases has remained largely the same, with samples of the micro-organism taken from patients and examined under a microscope. While this has proved effective at detecting a particular bacterial infection and can provide some guidance in terms of their infectivity and responsiveness to treatment, it is technically demanding and requires biocontainment facilities (Doughty, Sergeant, Adetifa, et al 2014).
Some progress has been made through the recent introduction of molecular-based tests which are helping in the identification of different pathogen species and strains. Such tests are nonetheless far from perfect and still rely on the isolation of the pathogen in pure culture which takes time. One way forward is thought to lie with the application of WGS directly to clinical samples. It is still early days, however, and much more work is needed to determine whether this approach can offer a way forward. What is limiting the wide-spread adoption of WGS is the difficulty of automating the interpretation of the sequence data. Once a pathogen has been sequenced, the data still need to be translated into a form that can be understood by non-specialists (Peacock, 2014).
Cole, S T, Brosch, R, Garnier, T, et al (1998) 'Deciphering the biology of Mycobacterium tuberculosis from the complete genome sequence', Nature, 393: 537-44. Back
Didelot, X, Bowden, R, Wilson D J, Peto, T E, Cook, D W (2012) 'Transforming clinical microbiology with bacterial genome sequencing', Nature Review, 13: 601-11. Back
Doughty, E L, Sergeant, M J, Adetifa, I, Antonio, M, Pallen, M J (2014) 'Culture-independent detection and characterisation of Mycobacterium tuberculosis and M. africanum in sputum samples using shotgun metagenomics on a benchtop sequencer', PeerJ, 23/2:e585, doi: 10.7717. Back
Halachev, M R, Chan, J Z M, Constantinidou, C I, et al (2014) 'Genomic epidemiology of a protracted hospital outbreak caused by multidrug-resistant Acinetobacter baumanni in Birmingham, England', Genome Medicine, 6: 70, doi:10.1186/s13073-014-0070-x. Back
Harris, S R, Cartwright, E J P, Torok, M E, et al (2013), 'Whole-genome sequencing for analysis of an outbreak of meticillin-resistant Staphylococcus aureus: a descriptive study', Lancet Infectious Diseases, 13/2:130-36. Back
Holden, M T G, Feil, E J, Lindsay, J A, Peacock, S J (2004) 'Complete genomes of two clinical Staphylococcus aureus strains: Evidence for the rapid evolution of virulence and drug resistance', PNAS, 101/26: 9786-91. Back
Ioerger, T R, Feng, Y, Ganesula, Chen, X, et al (2010) 'Variation among Genome Sequences of H37Rv Strains of Mycobacterium tuberculosis from Multiple Laboratories', Journal of Bacteriology, 192/14, 3845-53. Back
Kapur, V, Li, L L, Iordanescu, S, et al (1994) 'Characterization by automated DNA sequencing of mutations in the gene (rpoB) encoding the RNA polymerase beta subunit in rifampin-resistant Mycobacterium tuberculosis strains from New York City and Texas', Journal Clinical Microbiology, 32/4: 1095-8. Back
Peacock, S (2012), 'DNA sequencing stops MRSA spread', interview, The Naked Scientist. Back
Peacock, S (2014) 'Bring microbial sequencing to hospitals', Nature, 509: 557-9. Back
Peacock, S, Interview with Lara Marks, 3 May 2015, notes. Back
Robinson, E R, Walker, T M, Pallen, M J (2013) 'Genomics and outbreak investigation: From sequence to consequence', Genome Medicine, 5: 36. Back
Tong, Y C, Holden M T G, Nickerson, E K, Cooper, B S, et al (2014) 'Genome sequencing defines phylogeny and spread of methicillin-resistant Staphylococcus aureus in a high transmission setting', Genome Research, http://dx.doi.org/10.1101/gr.174730.114 Back
University of Cambridge (2014) 'Using genome sequencing to track MRSA in under-resourced hospitals'. Back
Wang, X, Jiao, J, Xu, W, et al (2013) 'A simple, rapid and economic method for detecting multidrug-resistant tuberculosis', The Brazilian Journal of Infectious Diseases, 17/6: 667-71. Back
WHO (2013), Global Tuberculosis Report 2013. Back