Connect with us

Fitness

The bacteria-fighting skills of non-antibiotic drugs

Published

on

The bacteria-fighting skills of non-antibiotic drugs

Researchers at the University of Massachusetts led by cancer biologist Amir Mitchell, wondered if non-antibiotic drugs might also have antibacterial potential (1). By screening anticancer drugs for antibacterial effects, the researchers identified new bacterial targets for antibiotics, opening the potential for new treatments in the fight against infections. 

“We started off doing some work on trying to understand the toxicity of different chemotherapies on bacteria,” Mitchell said. “This led us to a bigger question of how many drugs that were never designed to be antibiotics were impacting bacteria.” 

Deciphering how non-antibiotic drugs harm bacteria could be a medical game-changer. If they target bacteria differently than standard antibiotics, they offer exciting possibilities for new antibiotic development. However, if they work similarly to existing drugs, their long-term use in chronic illnesses might unintentionally contribute to antibiotic resistance. 

Mitchell’s team used pooled genetic screening of an Escherichia coli strain collection and high-throughput barcode gene sequencing to analyze almost two million instances of toxicity between 200 drugs and thousands of bacteria with single gene deletions. They used a machine learning algorithm to group the drugs based on how they killed the bacteria. The analysis showed that known antibiotics grouped together by their known classes of killing mechanisms, and non-antibiotic drugs formed separate hubs from antibiotics, indicating that non-antibiotic and antibiotic drugs have different ways of killing bacterial cells.

Traditional antibiotics often work by interfering with cell wall synthesis, protein production, or DNA replication (2). In contrast, the non-antibiotic drugs identified in the study disrupted various other bacterial functions such as translation initiation. Because non-antibiotics affect different bacterial pathways compared to conventional antibiotics, bacteria are less likely to develop resistance to non-antibiotic drugs. 

The team then investigated whether they could identify new drug targets in bacteria using non-antibiotic drugs. They grew hundreds of generations of bacteria exposed to different non-antibiotic drugs. The group then sequenced the genomes of the bacteria that adapted to the drugs and applied their machine learning algorithm. 

This approach pinpointed the specific bacterial proteins targeted by different non-antibiotic drugs. For example, they identified a bacterial protein targeted by triclabendazole, a drug used to treat parasite infections. Currently available antibiotics don’t target this protein, highlighting the potential of machine learning approaches to identify drugs with unknown bacteria-killing mechanisms.

“You really need an efficient method to quickly tell you the mechanisms of action,” said Mitchell.  “AI gives you a way out of this to develop new applications,” he added. “We are at the watershed moment in history. Everything is going to change.”

Joao Xavier, a bioinformatician from the Memorial Sloan Kettering Cancer Center who was not part of this study, said, “This is a type of work with very modern principles of analyzing drug effects on bacteria at a very broad scale while looking at the impact of every single gene and their interactions with a vast library of molecules. It’s really the type of work that we need to advance our understanding of novel antibiotics.”

Mitchell and his group now plan to test if the non-antibiotic drugs they identified can serve as useful antibiotics in animal models and human samples. “For us, the most exciting thing is really pushing the envelope and increasing the throughput of the research we can do,” Mitchell said.

Continue Reading