Using AI, they get the most complete map of interactions for bacterial survival.

Bacteria perform many functions that are key to their survival, such as producing the energy they need, replicating DNA and dividing cells for reproduction, or synthesizing the cell membrane for protection and interaction with the environment, among others. .

Complexes interfere with all these processes, requiring the coordinated action of a set of proteins that are necessary: ​​without them, the processes do not develop and the bacterium dies. Therefore, detailed knowledge of how these basic processes are regulated, which proteins are involved and how they interact is necessary to understand the mechanisms of bacterial growth, reproduction and survival.

The experimental methods used so far have identified millions of interactions between proteins and thousands of their structures, but they represent raw data that produces a very large number of false positives, interactions that are really of no value.

With the help of recently developed artificial intelligence (AI) models such as AlphaFold2, it is possible to obtain protein structures with accuracy similar to experimental methods, distinguishing between genuine interactions between proteins and spurious interactions.

Thus, researchers at the Department of Biochemistry and Molecular Biology at the Autonomous University of Barcelona (UAB) used AlphaFold2 to predict a set of interactions between proteins that are essential for the survival of bacteria, a total of 1,402 possible interactions that make up the most complete map of what is called the essential interactome of bacteria.

All of these interactions expand our knowledge of the mechanisms of action that bacteria need to survive and allow us to identify which protein interactions may be targets for the development of new antibiotics.

“We have a map that captures all the fundamental interactions that allow bacteria to live and reproduce, and we have characterized them structurally using new artificial intelligence tools, particularly AlphaFold,” explains Mark Torrent, director of the newly published study in eLife. magazine.

“We believe that these structures are a starting point for the development of new antibiotics, since molecules that can inhibit these interactions will behave like antibiotics with unusual mechanisms of action,” he adds.

Mark Torrent Burgas and Jordi Gomez Borrego (fourth and fifth from left respectively) with the rest of the research team: Carmen Mesas Vaz, Enea Sancho Vaello, Roberto Bello Madruga and Alba Gembe Mühlberger. / COMPANY

Great predictive power

Between 4,000 and 5,000 proteins are involved in bacterial activity. This set is called the bacterial proteome, resulting in an interactome that may have 20 million possible interactions.

However, it is estimated that the number of interactions that occur in a single species, such as Escherichia coli, is limited to about 12,000. And not all of these interactions are necessary for the bacteria to survive.

To isolate significant interactions, the researchers only looked at those in which the two proteins interacting to form a complex were present in at least two different bacterial species. Using these filters and an artificial intelligence model, the researchers obtained a set of 1,402 significant interactions between proteins.

To test the reliability of AlphaFold2, the team compared its predictions to 140 protein interactions that had previously been obtained experimentally.

The result was a prediction ability that the authors describe as excellent, as 113 of these experimental interactions (81%) were predicted by the AI ​​with great accuracy. The researchers believe that many of the protein interaction complexes that can be found in experimental databases may be false positives.

Method for obtaining new antibiotics

Experts highlight the discovery with this method of a set of interactions between proteins that were unknown until now and that are involved in 9 different important processes: biosynthesis of fatty acids in the cell membrane, synthesis of lipopolysaccharides in the outer membrane, lipid transport, transport of proteins and lipoproteins of the outer membrane, cell division, maintenance of elongated bacilli, DNA replication for bacterial propagation, and ubiquinone synthesis.

A detailed understanding of the structure of newly discovered protein complexes provides new knowledge about the molecular mechanisms occurring in these vital processes for bacteria and opens the way to the production of new antibiotics.

Link:

Gomez Borrego and Torrent Burgas. “Structural assembly of the bacterial essential interactome”. electronic life 2024