Since the 1970s, the discovery of current antibiotics has been in a period of stagnation. Now the World Health Organization decided antimicrobial resistance crisis as one of the 10 greatest threats to global public health.
When an infection is treated repeatedly, doctors run the risk of bacteria becoming resistant to antibiotics. But why would an infection return after adequate antibiotic treatment? One well-documented possibility is that the bacteria become metabolically inert, evading detection by established antibiotics that respond only to metabolic activity. Once the danger has passed, the bacteria come back to life, and the infection reappears.
“Drug resistance builds over time, and recurrent infections are a result of that dormancy,” says Jackie Valeri, a former MIT-Takeda Scholarship (based at the MIT Abdul Latif Jameel Clinic for Machine Learning in Health) who recently earned a PhD in biological engineering from the Collins Lab. Valeri is the first author new paper published in the current print issue of the journal, which shows how machine learning can support study compounds that kill dormant bacteria.
Tales of dormant-like bacterial immunity are nothing novel to the scientific community – old strains of bacteria dating back 100 million years have been discovered in recent years living in an energy-efficient state on the bottom of the Pacific Ocean.
James J. Collins, chair of the MIT Jameel Clinical Department of Life Sciences, the Termeer Professor of Medical Engineering and Science in the Institute of Medical Engineering and Science and the MIT Department of Biological Engineering, recently made headlines for using AI to discover a novel class of antibiotics. This is part of a broader mission for the group to utilize AI to dramatically expand the range of antibiotics available today.
According to a 2019 paper published by , 1.27 million deaths could be prevented if infections were drug-sensitive, and one of the many challenges scientists face is finding antibiotics that can target bacteria in a metabolically dormant state.
In this case, Collins Lab scientists used AI to speed up the process of finding antibiotic properties in known drug compounds. With millions of molecules, this process can take years, but the researchers were able to identify the compound, semapimod, in a weekend, thanks to the AI’s ability to perform high-throughput screening.
Researchers found that semapimod, an anti-inflammatory drug commonly used to treat Crohn’s disease, is also effective in treating stationary-phase and .
Another discovery was the ability of semapimod to disrupt the cell membranes of so-called Gram-negative bacteria, which are known for their high intrinsic resistance to antibiotics due to their thicker, less easily penetrated outer membrane.
Examples of Gram-negative bacteria are , , and , for which finding novel antibiotics is complex.
“One of the ways we discovered the mechanism of sema [sic] was that its structure was really big and reminded us of other things that target the outer membrane,” Valeri explains. “When you start working with lots of small molecules… for our eyes, it’s a pretty unique structure.”
By destroying a component of the outer membrane, semapimod sensitizes Gram-negative bacteria to drugs that usually only act on Gram-positive bacteria.
Waleri cites a quote from a 2013 article published in the journal: “For Gram-positive infections, we need better drugs, but for Gram-negative infections, we need any drugs.”