The doctor enters the patient’s room for the fourth visit in three days. Her immunosuppressed patient was transferred from a small local clinic to this large research facility to provide more intensive treatment. Despite this, the patient is deteriorating, and suffering from multiple infections. A culture has just come back from the microbiology lab, which shows the Klebsiella pneumoniae isolated from the patient is resistant to ertapenem, imipenem, and meropenem – the isolate is a carbapenem-resistant Enterobacteriaceae (CRE). The drug susceptibility testing directs therapeutic decisions, but an important question remains: did the patient pick up this resistant strain in this hospital or was it brought from the small local clinic?
This is the type of situation that Drs. Nicole Pecora and Lynn Bry hope to address with their most recent research. These clinical scientists, along with their collaborators, have used whole-genome sequencing to address epidemiological questions regarding the sources of antibiotic resistant bacteria. Their research focused on two types of CRE found in the gastrointestinal tract: Klebsiella pneumoniae, a common secondary infection in healthcare settings, and Enterbacter cloacae, which can be pathogenic and has growing resistance to carbapenems. The results of their three-year study are now published in mBio.
To examine where antibiotic resistance comes from, the doctors “took a broad range of multidrug resistant organisms, not just specific outbreaks, to investigate what genomically was going on,” explains Bry. The isolates were first characterized by drug-resistant phenotypes on media containing one of three carbapenem drugs (ertapenem, imipenem, and meropenem). Those able to grow in the presence of all three drugs were classified as CRE and were subsequently submitted for whole genome sequencing.
Many genes have been associated with carbapenem resistance. The (Klebsiella pneumoniae carbapenemase) KPC genes (KPC-2, -3, and -4), as well as OXA-48, NDM, and VIM are carbapenemases that break down carbapenem drugs before the drug has a chance to inhibit bacterial growth. Other genes, including extended-spectrum beta lactamases and mutated (dysfunctional) porin genes, can confer resistant phenotypes as well. Sequencing the entire genome allowed Pecora and Bry to connect the resistant clinical isolates with the genes that conferred resistance.
Using this genomic information helps the hospital determine if they are dealing with an internal outbreak or an external source. Spread from internal contamination will often have a similar genomic pattern, while external sources may use entirely different genes for the same growth pattern. Differentiating between the two initiates different sets of operational procedures.
Several important results arose from this research. Bry and Pecora recognized that closely linked isolates were collected from patients treated at the same time and place, suggesting spread between patients. But they also found sporadic introduction of resistant strains with new genomic signatures, suggesting contaminant entry as part of a patient’s microbiota or from a different clinic. Because sequencing entire bacterial sequences takes about a week, this isn’t a viable alternative for diagnostic tests (yet). However, this turnaround time gives infection control departments enough lead time to put in proper preventative measures.
The results showed that all of the KPC genes found were associated with transposons. “The KPC genes are well characterized to travel on transposons, so we expected to see (transposon) Tn4401 isoforms on different plasmid backbones,” Pecora says. Transposons are short pieces of DNA that can ‘jump’ in and out of DNA sequences, often traveling between different plasmids. What Pecora hadn’t expected was to find chromosomal integrations of the transposon, where the KPC gene had jumped from a plasmid into the bacterial chromosome. These chromosomally-located resistance transposons may potentially jump into a second plasmid, conferring resistance to any bacteria to which the plasmid is passed (see left). Different plasmids exist in different bacterial populations, so this may lead to the further spread of resistance into currently susceptible bacterial species.
The authors’ research also adds to the complicated nature of defining resistant strains. Under current CDC guidelines, carbapenem resistance is defined by the three phenotypic resistances, meaning all of the strains sequenced in this study fall under the resistant category. However, phenotypic resistance can involve multiple gene involvement, as was seen in several of the E. cloacae isolates. These strains didn't contain KPC genes, but instead had both beta-lactamases and porin mutations, which allowed them to grow in the presence of the three carbapenems tested. Asks Bry, “What is the predictive capacity of saying you have a true KPC resistant strain? Should we revisit or adjust the guidelines on the best way to call them KPC resistant strains?”
Both authors hope that this research will set the stage for a larger database. “This was done in a single institution, but with an eye to a regional, national, and even international setting,” says Bry, “to leverage what we know and stem the spread of multidrug resistant organisms.”
-- Julie Wolf