The emergence of multidrug resistant (MDR) bacteria has greatly challenged the effective management for both community and hospital acquired infections. Many MDR bacteria are bred inside hospitals, where they further disseminate by contaminated devices and via the hands of health care workers. Therefore, surveillance on outbreaks of infections caused by MDR bacteria is crucial for the understanding of their emergence and the design of effective management strategies. For several years, our group has been working closely with hospital surveillance systems to unravel how MDR pathogens evolved.
In 2016, we got a dozen of MDR Pseudomonas aeruginosa strains isolated from patients hospitalized in National University Hospital of Singapore. These strains can produce the metallo-beta-lactamase NDM-1, which causes their extremely high resistance to carbapenems, the first-line agents to treat P. aeruginosa infections. They initially drew our attention because although we’ve been working on P. aeruginosa clinical isolates for several years, the NDM-1-producing P. aeruginosa have not been identified in Singapore. We were interested in how these “local” strains acquired the blaNDM-1 gene and if they also possess other antimicrobial and virulence determinants that can explain their successful spreading. To our surprise, a novel integrative and conjugative element (ICE), for which we named ICETn43716385, was found to integrate into the genomes of these strains and is responsible for transmitting blaNDM-1 into P. aeruginosa.
It is uncommon that ICEs carry acquired antibiotic resistance genes in P. aeruginosa, while it turned out that other than blaNDM-1, ICETn43716385 carries two more: msr(E) and floR. Msr(E) is responsible for resistance to macrolides and is a member of the ABC-F family proteins. The identification of msr(E) in P. aeruginosa genomes raised us two more questions. First, P. aeruginosa is intrinsically resistant to macrolide-mediated growth inhibition, does acquisition of msr(E) confers other selective advantages on P. aeruginosa? Second, although several studies suggested that ABC-F proteins mediate macrolide resistance by ribosomal protection, there was still no study showing the structure of Msr(E) bound ribosomes. It remained unclear how Msr(E) binds the ribosome to mediate macrolide resistance.
To address the first question, we exogenously expressed Msr(E) in laboratory strains and tested if the Msr(E) expressing and non-expressing strains respond differently to sub-inhibitory levels of a macrolide antibiotic azithromycin, which was reported to inhibit the quorum sensing system of P. aeruginosa and can be used as an anti-virulence drug for P. aeruginosa infections. Our results showed that Msr(E) abolished macrolide-mediated quorum sensing inhibition in vitro and anti-Pseudomonas effect in vivo. Meanwhile we were investigating the functional roles of Msr(E) in P. aeruginosa, our collaborators uncovered how Msr(E) interacts with the ribosome. Their Cryo-EM structure clearly showed that Msr(E) binds to the ribosomal exit site and extents its “needle” to the peptidyl-transferase centre to displace macrolides from the ribosome, thereby conferring macrolide resistance on the host bacteria.
While we finally got our answers, even more questions arise: how ribosomal protection can link to the relief of macrolide-mediated quorum sensing inhibition? Do other ribosomal protection proteins have the same effect as Msr(E)? ICETn43716385 “chooses” to carry blaNDM-1 and msr(E) simultaneously, is it by chance or by selection, and if so, what is the selective force? Finding the answers to these questions will deepen our understanding on the ultimate question: how to deal with MDR pathogens promptly and effectively?
Links to our papers: