The paper in Nature is here: https://go.nature.com/2J29mXV
Discovery of new antibiotics and elucidation of their mechanism often require interdisciplinary collaboration. We would like to share our story about a successful collaboration with a group of clinicians, geneticists, physicists, engineers, and chemists from Brown, Harvard, and Emory University.
As we are part of an infectious diseases division, we are completed by the frequent relapses and the prolonged treatment associated with Staphylococcus aureus infections. This motivated us to study not only the ability of S. aureus to develop antibiotic-resistance, but also its ability to shift to non-growing dormant subpopulation of “persisters”. Indeed, S. aureus persisters formed in biofilms are known as a major reason for high relapse rates of endocarditis and osteomyelitis. Therefore, new classes of antibiotics are urgently required to treat these complex S. aureus infections.
To identify the new antibiotics effective against MRSA, we conducted a high-throughput screening project in collaboration with Frederick M. Ausubel. The major advantage of the C. elegans-based screening method is to identify anti-infectives while simultaneously excluding compounds that are toxic to the worm or have no in vivo efficacy. Using this screening methodology, we screened ~82,000 small molecules and identified 185 compounds that rescue C. elegans from MRSA infections.
Among hit compounds, we found that two synthetic retinoids (vitamin A analogues), CD437 and CD1530 that are formerly known bioactive compounds and share structural similarity to each other. Interestingly, their analogue, adapalene (a FDA-approved synthetic retinoid acne drug) was not identified as a hit. We found that CD437 and CD1530 kill both growing and persistent MRSA by inducing rapid membrane permeabilization without detectable resistance development. By comparing CD437 and CD1530 with adapalene, two polar branch groups, we easily noted that carboxylic and a hydroxyl moieties play key roles in antimicrobial activity and membrane-permeability.
To understand the mode of action of these retinoids, we collaborated with geneticists, biophysicists and engineering scientists. First, Daria Van Tyne and Michael M. Gilmore analyzed the whole genome sequences of S. aureus mutants exhibiting modest retinoid-resistance and identified the mutated genes resulting in modest retinoid resistance. Interestingly, the identified mutated genes were related to bacterial membrane physiology. Based on this results, we hypothesized that our synthetic retinoids target S. aureus membranes.
In 2013, our lab moved from Harvard Medical School to Warrant Alpert Medical School of Brown University. We were able to continue to meet wonderful collaborators in this new place. To confirm our hypothesis, we collaborated with Brown membrane biophysicists, Petia Vlahovska (Currently Northwestern University) and her postdoc Nico Fricke who are experts on artificial lipid bilayers such as giant unilamellar vesicles to mimic bacterial lipid bilayers. Using the giant unilamellar vesicles they made, we were able to confirm that CD437 and CD1530, but not adapalene, interact with and disrupt lipid bilayers. We then questioned how our reitniods interact with lipid bilayers at molecular levels. To address this question, we reached out to Brown engineering scientists, Huajian Gao and his postdoc Wenpeng Zhu, experts on computational simulations of cell-nanomaterials interactions. By using molecular dynamic simulations, we found that our retinoids persistently bind to hydrophilic heads via two polar branch groups and then penetrate the bilayers using interactions between retinoids’ hydrophobic backbone and lipid tails. As a result, the retinoids embedded into outer membrane leaflet, causing substantial disruption of the membrane lipid bilayers.
To further confirm the functionality of each chemical groups of retinoids and assess the possibility of further optimization the expertise of medicinal chemists was needed. Coincidentally, Bill Wuest from Temple University (currently Emory University) was invited to Brown and presented his work. We discussed our results with Bill, and he was willing to collaborate with us. Bill and his graduate students, Andrew M. Steele and Colleen E. Keohane synthesized 16 analogues. We were able not only to confirm our simulation results and but also to synthesize ‘analogue 2’ that exhibits significantly reduced cytotoxicity while retaining antimicrobial activity (another lesson here is to never miss a seminar).
In addition to the scientific experiences through this process, we truly enjoyed working with scientists from 7 different countries (South Korea, China, United States, Germany, India, Bulgaria, and Greece). In addition to new compounds, scientific discovery can provide a social paradigm how different cultural and disciplinary backgrounds can synergize effectively.
Written with Eleftherios Mylonakis