AI tools help uncover enzymatic mechanisms of lasso peptides

Lasso peptides are natural products made by bacteria. Their unusual lasso shape gives them remarkable stability, protecting them from extreme conditions. In a new study, published in Nature Chemistry BiologyThe researchers built and tested models explaining how these peptides are made and demonstrated how this information could be used to advance Lasso peptide-based drugs toward the clinic.

“Lasso peptides are interesting because they are linear molecules that have been tied together in a noose,” said Susanna Barrett, a graduate student in the Mitchell lab (MMG). “Because of their incredible stability and ingenuity, they have great therapeutic potential. They have also been shown to have antibacterial, antiviral and anticancer properties.”

Lasso peptides are ribosomally synthesized and post-translationally modified molecules. Peptide chains are formed by joining amino acids together in a chain, which is accomplished by the ribosome. Two enzymes, a peptidase and a cyclase, then work together to convert a linear precursor peptide into the characteristic knotted lasso structure. Since their discovery more than three decades ago, scientists have been trying to understand how the cyclase folds the lasso peptide.

“One of the main challenges in solving this problem is that enzymes are difficult to use. They are usually insoluble or inactive when you try to purify them,” Barrett said.

A rare counterexample is fusilassine cyclase, or FusC, which the Mitchell lab characterized in 2019. Former members of the group were able to purify the enzyme, and it has since served as a model for understanding the lasso-tying process. Yet the structure of FusC remained unknown, making it impossible to understand how the cyclase interacts with the peptide to fold the knot.

In the current study, the group used the artificial intelligence program AlphaFold to predict the structure of the FusC protein. They used the structure and other artificial intelligence-based tools, such as RODEO, to identify residues in the cyclase active site that were important for interacting with the lasso peptide substrate.

“FusC is made up of about 600 amino acids and the active site contains 120. These programs were essential to our project because they allowed us to do ‘structural studies’ and determine which amino acids are important in the active site of the enzyme,” Barrett said.

They also used molecular dynamics simulations to computationally understand how the lasso is folded by the cyclase. “With the computing power of Folding@home, we were able to collect a lot of simulation data to visualize the interactions at the atomic level,” said Song Yin, a graduate student in the Shukla lab. “Prior to this study, there were no MD simulations of the interactions between lasso peptides and cyclases, and we believe this approach will be applicable to many other peptide engineering studies.”

Using their computational efforts, the researchers discovered that among the different cyclases, the back wall region of the active site seemed particularly important for folding. In FusC, this corresponded to the helix 11 region. The researchers then performed cell-free biosynthesis where they added all the cellular components needed to synthesize lasso peptides to a test tube with enzyme variants that had different amino acids in the helix 11 region. Eventually, they identified a version of FusC with a mutation on helix 11 that could fold lasso peptides that could not be made by the original cyclase. These data confirm the model of lasso peptide folding that the researchers developed with their computational approaches.

“How enzymes form a lasso knot is a fascinating question. This study provides the first insight into the biophysical interactions responsible for producing this unique structure,” said Diwakar Shukla, associate professor of chemical and biomolecular engineering.

“We also showed that these molecular contacts are the same in several different cyclases across different phyla. Although we haven’t tested all the systems, we think this is a generalizable model,” Barrett said.

In collaboration with San Diego-based Lassogen, the researchers showed that the new knowledge can guide cyclase engineering to generate lasso peptides that cannot be made otherwise. To prove the concept, they engineered another cyclase, called McjC, to efficiently produce a potent inhibitor of a cancer-promoting integrin.

“The ability to generate a diversity of Lasso peptides is important for optimizing drug discovery,” said Mark Burk, CEO of Lassogen. “Nature’s enzymes don’t always allow us to produce the Lasso peptides we’re interested in, and the ability to engineer Lasso cyclases greatly expands the therapeutic utility of these amazing molecules.”

“Our work would not have been possible without access to powerful computational tools and recent advances in artificial intelligence and cell-free biosynthesis methods,” said Douglas Mitchell, the John and Margaret Witt Professor of Chemistry. “This work is an extraordinary example of how interdisciplinary collaborations are catalyzed at the Carl R. Woese Institute for Genomic Biology. I am grateful to the IGB MMG theme and our external colleagues at Lassogen for their participation in solving this complex problem.”