A pair of newly discovered enzymes could change drug manufacturing, reducing both costs and environmental harm.
Researchers at Duke University School of Medicine showed that one of the enzymes is sufficient to make azetidine, an organic compound and a building block of a wide range of therapeutics, including antibiotics, antiviral drugs, and cancer treatments.
Rather than relying on harsh solvents and toxic chemicals, the team discovered enzymes, PolF and PolE, that produce azetidine from an inexpensive precursor compound in water.
“The two enzymes that we discovered are functionally and structurally novel,” said Kenichi Yokoyama, PhD, associate professor of biochemistry at Duke and senior author of the study. “We're interested in applying these enzymes for biocatalytic reactions to produce structurally diverse azetidine amino acids, and other cyclic compounds.”
Their work is published October 21, 2025 in Nature Chemistry.
A previously reported enzymatic process for producing azetidine required a precursor compound that’s difficult to produce and thus expensive to buy; one supplier sells a one-gram vial (about half the weight of a U.S. penny) for $1,390. That’s about 1000-fold more than it costs to buy the precursor used in the Duke study, Yokoyama said.
Azetidine is a small, “strained” ring of four atoms: one nitrogen atom and three carbons. It’s a challenge to make because of its small size, Yokoyama explained. “As you make the ring smaller, you have to bend the bonds, and that requires a lot of energy. Somehow, PolF can do that job.”
The second enzyme they discovered, PolE, helps by boosting the supply of a necessary intermediate.
PolF’s capabilities don’t stop there. The researchers found that it can also produce an even smaller compound, aziridine, which is a nitrogen-containing, three-membered ring.
“One enzyme that can make both a four-membered and a three-membered ring — that's unique and unprecedented,” Yokoyama said.
The researchers discovered the newfound enzymes by studying how azetidine is made naturally in an antifungal compound called polyoxin.
Collaborators at Pennsylvania State University, including Amie Boal, PhD, Karsten Krebs, PhD, and J. Martin Bollinger, PhD, provided key expertise in structural biology, bioinorganic chemistry, and spectroscopy, Yokoyama said.
Former postdoctoral associate Ya-Nan Du, PhD, conducted much of the hands-on experimental work, both in Yokoyama’s lab at Duke and at Pennsylvania State, where she learned more about how the PolF enzyme works by using stopped-flow and Mössbauer spectroscopy, which allow chemists to monitor the fast, enzyme-catalyzed reactions.
These detailed experiments revealed the iron-containing reactive species at the center of the PolF enzyme, responsible for breaking the strong carbon-hydrogen bonds and forming the strained four- and three-membered rings, Yokoyama said.
Additional authors: Anyarat Thanapipatsiri, Jesús José Blancas Cortez, Xavier Enrique Salas Solá, and Chi-Yun Lin.
Funding: National Institutes of Health