Skip to main content
Premium Trial:

Request an Annual Quote

Houston Hospital Tests Adding RNA-seq to WGS Protocol to Improve Outbreak Monitoring


NEW YORK (GenomeWeb) – A clinical research team at the Houston Methodist Research Institute that has been using clinical whole-genome sequencing for infectious disease outbreak monitoring recently tested adding RNA-seq to its protocol.

In a paper published in September in the Journal of Clinical Microbiology, the team reported that RNA-seq and animal virulence studies added important data about regulatory genes to a whole-genome sequencing analysis of a mock Streptococcus outbreak and surmised that such a combined analysis could be performed rapidly to contribute to patient care and public health decisions in the event of an actual outbreak.

The group first began using whole-genome sequencing for outbreak monitoring at Houston Methodist Hospital in 2011 when a patient died from an unknown Bacillus species.

Since then, the group has published a number of studies using whole-genome sequencing to investigate outbreaks and has implemented a protocol within its hospital to track potential outbreaks, James Musser, director of the Center for Molecular and Translational Human Infectious Diseases Research at HMRI, told Clinical Sequencing News.

The hospital runs whole-genome sequencing of microbes about once a week as a routine clinical test, the authors wrote in their paper. In addition, Musser said that it is used when needed, such as if "there is concern that there is a potential nosocomial transmission, or unusual disease severity, unusual antimicrobial resistance, or, for instance, a cluster of Listeria that could be food-borne."

Recently, the group wanted to see what other types of data it could generate rapidly that would bear on the issues that arise during the course of a potential outbreak, Musser said. The group also wanted to see how quickly the clinical laboratory could respond to a potential outbreak scenario while still maintaining normal patient care activities.

In its JCM study, the team described a protocol that includes whole-genome sequencing, transcriptome sequencing, and animal virulence studies to understand potential outbreaks. They completed the investigation in nine days.
The team constructed a mock outbreak scenario using serotype emm59 group A Streptococcus as a model pathogen because the lab has an extensive genome sequence database on that pathogen. While the strain is typically uncommon, a hypervirulent clone recently arose in Canada, causing hundreds of severe infections, the researchers wrote. In addition, infections have also been increasing in some areas of the US, so the researchers wanted to test whether the organisms in Canada and US are related and if so, whether they could identify potential ways in which the infection was spread.

The group sequenced the whole genomes of 84 strains that were recovered by the US Centers for Disease Control and Prevention's Active Bacterial Core surveillance program from patients infected in the US between 2010 and 2013.

The team first sequenced the genomes of the 67 strains recovered between 2010 and 2012 using a 75-bp paired-end protocol on the Illumina MiSeq. For an additional 17 strains collected in 2013, the group used a 250-bp single-end sequencing protocol. For the RNA-seq experiments, they used a 50-bp single-end sequencing strategy.

For the first 67 strains, the group generated a phylogenetic tree within two days showing that the infections in the US were related to the Canadian epidemic clone. The 67 strains differed by an average of 17 polymorphisms. Five sets of genomically indistinguishable strains were identified, indicating that those likely arose from direct human-to-human transmission. In addition, the team identified geographical relationships between the strains; those recovered from the same states were more closely related, and even those from the same county tended to cluster more closely together.

Further analysis revealed that the strains in Minnesota likely had two distinct origins in British Columbia and Manitoba. Meanwhile strains from New Mexico were related to strains originating in either British Columbia or Alberta, suggesting that the Canadian clone came to the US through multiple events, the researchers wrote.

Next, they chose three different strains collected in Minnesota to evaluate via RNA-seq. The strains were chosen to mimic a regional outbreak scenario, so were only collected from one state. In addition, the group collected strains from different branches of the phylogenetic tree and only those with wild-type alleles for the major transcription regulatory genes. The transcriptomes showed a high degree of similarity, which was to be expected. In addition, they identified differential expression in five genes that encode virulence factors or transcription regulators.

In one strain, hyaluronic acid synthase, a key virulence factor, was significantly upregulated. Upon re-examining the genomic data, the team identified a 38-bp deletion in the promoter region that had previously not been recognized as important. Another unexpected finding from the transcriptome data was altered expression of the YesM/N genes, which are known regulatory genes in a different serotype but had not been studied in the serotype responsible for the outbreak. Reanalysis of the genome data with a new bioinformatics pipeline found a mutation in the promoter region of the YesM/N genes.

"Taken together, the RNA sequence data add to the list of regulatory genes identified by whole-genome sequencing that are undergoing diversifying selection in epidemic emm59 GAS," the authors wrote.

Finally, the researchers performed virulence studies in mouse models, infecting mice after sequencing the pathogens and then observing them throughout the course of the study, confirming that each strain was highly virulent.

Going forward, Musser said that the laboratory would likely include RNA-seq on an as-needed, case-by-case basis. For instance, he said, if 10 patients have an unusually severe form of some infection like Escherichia coli or Staphylococcus aureus, "the question could become: Are these patients having an unusually severe form of the infection because their strain is a super bug? And, if so, what's responsible at the molecular level?" Another scenario in which he envisioned RNA-seq being helpful is if a patient or group of patients displayed a clinical spectrum that was unusual for a particular pathogen.

Throughout the study, the team also made process improvements to its standard whole-genome sequencing analysis, including implementing a new bioinformatics pipeline that includes read correction, read mappers, and variant callers that can handle mixed read populations. In addition, the team is creating tools for analyzing important pathogens such as S. aureus and Klebsiella pneumoniae so if there is a future outbreak, the time it takes to analyze those strains' genomes will be greatly reduced.

Houston Methodist is not the only hospital that has begun using whole-genome sequencing to investigate potential outbreaks in the wake of the 2011 E. coli outbreak in Europe in which several teams sequenced and analyzed the strains responsible. The National Institutes of Health's Clinical Center has been testing the technology for outbreak monitoring after it used it to track an outbreak of drug-resistant K. pneumoniae, and a consortium in the UK is looking to make sequencing routine in pubic health microbiology for surveillance and outbreak monitoring.

Musser said that while he did not think it would become standard in most clinical labs, at least in the immediate future, for a large 2,000-bed hospital like Houston Methodist, "it makes sense."