Skip to main content
Premium Trial:

Request an Annual Quote

Sequencing of Serial Clinical Pathogen Isolates IDs SNV Responsible for Drug Resistance


NEW YORK (GenomeWeb) – A research team at Mt. Sinai School of Medicine has shown that sequencing clinical isolates from a patient suffering from a bacterial infection can pinpoint the single variant responsible for the development of drug resistance.

In addition, by sequencing other patients with the same bacterial infection, the group was able to figure out that the cases had not been caused by a single transmission event, but rather represented a random uptick in Stenotrophomonas maltophilia infection.

The team published their results recently in the journal Antimicrobial Agents and Chemotherapy.

Whole-genome sequencing was "pretty crucial to make the findings that we made," Theodore Pak, lead author of the study and an MD/PhD student in Andrew Kasarkis' lab at Mt. Sinai, told GenomeWeb. He added that the case was one of the team's first experiences in using sequencing to understand infection, the cause of drug resistance, and whether an outbreak was occurring, so the goal was not to use the sequencing and analysis to inform patient treatment. In addition, the patient had already been successfully treated with a different antibiotic. But, he said that the team is now working to automate the bioinformatics portion in order to reduce the turnaround time to less than one week.

The researchers used Pacific Biosciences' RS II for the sequencing because of its long reads, enabling the researchers to "assemble the entire genome from scratch," Pak said, "which is important when there are not a lot of reference genomes," like in the case of S. maltophilia.

Pak said a main goal for the study was to determine whether a recent uptick in S. maltophilia infection represented an outbreak, caused by one clonal strain, or separate events. In addition, one patient had initially been responding to the drug levofloxacin but developed resistance, and the team wanted to see if they could figure out why.

The researchers sequenced an isolate from the patient before and soon after he developed resistance to levofloxacin, a broad-spectrum fluoroquinolone antibiotic. Little is known about the genetic and molecular underpinnings of resistance to quinolones. Often the quinolone-resistance-determining region of topoisomerase genes in the bacteria is unaltered, the authors wrote.

Aside from the patient who had become resistant to quinolones, the team sequenced isolates from five other patients that had become infected with S. matophilia during the same two-month period in 2013.

The PacBio sequence reads were de novo assembled using the firm's HGAP software. Comparing the genomes from the same patient before and after the development of quinolone resistance, the team identified one SNV and five single-base indels that differed between the two. The five single-base indels turned out to be homopolymer assembly errors, but Sanger sequencing confirmed the SNV. The variant was located inside the gene smeT, a repressor located upstream of an operon that encodes a multi-drug efflux pump. Previous in vitro studies had found that strains with the same mutation were also resistant to quinolone.

None of the remaining clinical isolates had that same mutation, but one isolate that was also resistant to levofloxacin had a different mutation in the same gene that was predicted to be damaging.

The gene smeT appears to play a main role in quinolone resistance, the authors noted. "Since any mutation that inactivates this protein would be able to repress smeDEF and confer resistance, smeT is under intense selective pressure in the presence of these drugs," the authors wrote.

Based on the study, Pak said, the gene appears to "easily evolve to produce the resistant phenotype." The patient was ultimately successfully treated with a different antibiotic, trimethoprim/sulfamethoxazole (SXT). Pak said that it is possible that the mechanism for developing resistance to SXT may not occur as easily as the mutation to the smeT gene, so treating patients with SXT could be a better option. However, he said, there is so far only anecdotal evidence in support of that theory, so a clinical trial would need to be done comparing the two antibiotics.

Another aspect of the study that was important, Pak noted, was that the resistant isolate was analyzed very early on in the development of drug resistance and had not yet acquired a significant number of mutations. As a result, the researchers were able to identify a single mutation that occurred early on to confer resistance.

Now, said Pak, they would like to study additional cases of "spontaneous drug resistance in greater detail." For instance, he said, one option would be to search through medical records to identify other cases where a patient initially responded and then suddenly became resistant. Often, such samples are banked, and the researchers could perform the same sequencing and assembly technique to try to determine the early mutations that lead to resistance.

"In hospitals, bacteria are steadily evolving and becoming more resistant," Pak said. "It's very concerning." His team would like to be able to catch that initial mutation and hopefully prevent drug resistance.

Pak said that the Mt. Sinai team recently began a study in which they plan to sequence clinical cases of Clostridium difficile infection on the PacBio system. The goal is to reduce the turnaround time to just a number of days by automating the bioinformatics, Pak said.

"Since patients are often in the hospital for 10 to 20 days, the sequence data could actually be used for clinical care," he said. The project just began and will run for at least four years, Pak said. In a pilot, the team sequenced around 40 isolates.

The goal is to develop protocols that other hospitals and laboratories will be able to adopt and use, Pak said. In the future, he anticipated that whole-genome sequencing of bacterial genomes isolated from patient samples would become common in hospitals in order to monitor for potential outbreaks and to impact patient care, if the patient develops resistance. The main hurdle is the bioinformatics analyses, which need to be automated, reproducible, and faster.

Already, Pak said, it is possible to sequence a bacterial genome for around $100, which is "well within the budget of hospitals."