Artificial Intelligence (AI) model shows how disease-causing bacteria, like E. coli for example, cause infection by injecting a cocktail of many different proteins knows as effectors into the host cell. As the worst part, bacteria are quite resilient and some of these proteins can actually be removed, but the infection still proceeds, according to a study published in Science (1). Researchers believe this work could help predict infection outcomes and develop new treatments in the future.
Pathogens like E. coli and Salmonella may be small, but they have incredible ways to make sure that, when they infect a host, the infection is successful (from the pathogen’s point of view, obviously).
One of these mechanisms uses tiny molecular ‘syringes’ to inject a special type of proteins known as effectors directly into host cells. These effectors aim to ultimately cause havoc with the cell’s metabolism, wiping out any possible immune response. Curiously, they’re not intended to kill the host cell. Their activities are often very subtle and more “tuned” to modulate cellular functions rather than irreversibly disrupt cellular mechanisms.
In the past, researchers have only focused on one effector at a time, but this approach has offered limited benefits. Now, a team from Imperial College and The Institute of Cancer Research, in London, decided to look at various effectors in different combinations. In particular, the team looked at how mice reacted to infection with their version of E. coli – known as Citrobacter rodentium – which can produce 31 different effectors.
The team gathered data from live animals for more than 100 different combinations of the 31 different effectors to conduct this analysis, which was then used to build an artificial intelligence (AI) algorithm. Researchers know that it’s impossible to test in the lab all the possible combinations effectors can form, so an AI model is the next best thing to predict the outcomes of infection. Using an AI model is the only viable approach to understand biological systems this complex.
It turned out effectors don’t need to work individually but instead get together to launch a plan of attack as a network. What’s worse, even if one or more components are eliminated, the infection can still proceed. Astonishingly, this network could be reduced up to 60% and still produce an infection.
“Our study shows that we can predict how a cell will respond when attacked by different combinations of bacterial effector proteins. The research will help us to understand better how cells, the immune system and bacteria interact, and we can apply this knowledge to diseases like cancer and inflammatory bowel disease where bacteria in the gut play an important role”, said Professor Jyoti Choudhary, from the Functional Proteomics Lab at The Institute of Cancer Research, London.
“We hope, through further study, to build on this knowledge and work out exactly how these effector proteins function and how they work together to disrupt host cells. In future, this enhanced understanding could lead to the development of new treatments.”
“The AI allows us to focus on creating the most relevant combinations of effectors and learn from them how bacteria are counteracted by our immune system. These combinations would not be obvious from our experimental results alone, opening up the possibility of using AI to predict infection outcomes.”, continued Dr David Ruano-Gallego from the Department of Life Sciences at Imperial.
Looking to the future, AI and synthetic biology are in an ideal position to understand which cell functions are affected during infection. It’s not unreasonable to suggest a scenario where doctors could treat patients not by killing the pathogen with antibiotics but by modulating and improving our own natural defence mechanisms against effectors and infection.
(1) Ruano-Gallego D, Sanchez-Garrido J, Kozik Z team al. (2021) Type III secretion system effectors form robust and flexible intracellular virulence networks. Science, 371: eabc9531