TL;DR: University of Illinois Urbana-Champaign researchers engineered a flavin-dependent enzyme to form carbon-phosphorus bonds with oxygen, and used it to make L-glufosinate with very high precision.

Engineered enzyme makes herbicide ingredient in one step

Researchers at the University of Illinois Urbana-Champaign created a laboratory enzyme that builds bonds between carbon and phosphorus, a linkage that enzymes in nature almost never assemble. The team demonstrated the redesigned catalyst by producing L-glufosinate, the active portion of a widely used broad-spectrum herbicide, according to the Center for Advanced Bioenergy and Bioproducts Innovation, or CABBI.

The outcome was reported in the journal Science on Oct. 8, 2026, in a paper titled Asymmetric enzymatic hydrophosphorylation through O2 activation. The authors listed on the study are Yi Zhou, Yifei Ge, Wesley Harrison and Huimin Zhao, all of the University of Illinois Urbana-Champaign, according to CABBI.

Zhao served as corresponding author and leads the Conversion Theme for CABBI, the U.S. Department of Energy bioenergy center based at Illinois. A photograph released with the announcement shows, from left, researcher Yi Zhou, Professor Huimin Zhao and researcher Yifei Ge in Urbana, Ill.

The engineered system centers on variants of OYE1, a flavin-dependent enzyme. Through laboratory evolution and condition tuning, the variants gained the ability to carry out asymmetric hydrophosphorylation through oxygen activation under both light and dark conditions, according to Science.

Mechanistic work described in Science indicates that reactive oxygen species formed from molecular oxygen are redirected into productive phosphorus-centered radicals. Those radicals add to an olefin, after which enzymatic hydrogen-atom transfer sets the chiral product, yielding high enantioselectivity.

The Illinois group framed the advance as an expansion of biocatalytic manufacturing. Enzymes are valued as manufacturing tools because they operate with high precision under relatively mild conditions, CABBI reported. The new work suggests that such mild, selective catalysis can now be extended to a class of phosphorus-containing molecules important for agriculture and medicine.

Why carbon-phosphorus bonds are difficult for biology

Carbon-phosphorus, or C-P, compounds appear in medicines, agricultural chemicals and advanced materials. Enzymatic formation of C-P linkages in nature is extremely rare, which has limited biocatalytic access to phosphorus-containing products used in pharmaceuticals and agrochemicals, according to Science.

CABBI explained the same constraint in practical terms. Bonds between carbon and phosphorus occur in many useful products, from medicines to agricultural chemicals, yet enzymes capable of creating them are extremely rare in nature.

Chemists have alternative routes to C-P bonds through phosphorus-centered radicals, but those chemical approaches often demand harsh oxidants or specially activated starting materials and provide limited selectivity, according to Science. CABBI similarly reported that nonenzymatic approaches can require harsh chemicals and multiple carefully controlled steps, making production complex and resource-intensive.

The scarcity of suitable natural catalysts has therefore restricted the use of enzymes in industrial processes for phosphonate-containing small molecules, even though such molecules are useful in medicine, agriculture and other fields. An editor's summary in Science by Michael A. Funk noted that the few natural enzymes that make molecules with carbon-phosphorus bonds are not amenable for use in bioindustrial processes.

The Illinois team aimed to overcome that barrier with a biocatalytic method that directly incorporates common phosphorus compounds into alkenes. Such a route would provide a mild and selective path to value-added, structurally diverse products, according to Science.

In the reported system, a flavin enzyme activates molecular oxygen to generate superoxide or hydroxyl radical intermediates that oxidize a phosphite precursor. After addition to an olefin, the enzyme directs hydrogen-atom transfer to generate a chiral product, according to the summary in Science.

How directed evolution and photocatalysis created new chemistry

Zhao is a professor of chemical and biomolecular engineering and a faculty member at the Carl R. Woese Institute for Genomic Biology at Illinois. His group integrated enzyme catalysis with photocatalysis and applied directed evolution to confer a new function on an existing enzyme, CABBI reported.

Directed evolution, as described by CABBI, means making variations of an enzyme and selecting those that perform better. The Illinois researchers tested more than 100 variations and combined the most promising changes.

The team then used machine learning to identify additional modifications that could improve performance. CABBI reported that this combination of experimental variation, combination of beneficial changes, and computational guidance eventually produced an enzyme able to use oxygen to help join carbon and phosphorus.

Science described the development as semirational and machine-learning-guided directed evolution of OYE1 variants. Experimental and computational studies were used to clarify the role of reactive oxygen species generated in place, according to the journal.

Molecular oxygen is an attractive and abundant oxidant, but its partial reduction normally generates reactive oxygen species that damage enzymes and disrupt radical reactions. Flavin-dependent enzymes naturally manage oxygen activation within confined active sites, suggesting that those reactive species might be redirected toward new reactivity, according to Science.

The researchers reasoned that an engineered flavin enzyme could harness reactive oxygen species to generate phosphorus-centered radicals from phosphorus-hydrogen donors and control product enantioselectivity through enzymatic hydrogen-atom transfer. By combining condition optimization, directed evolution and mechanistic studies, they sought a mild biocatalytic route to diverse chiral organophosphorus compounds, according to Science.

Computational results further revealed how enzymatic hydrogen-atom transfer controls enantioselectivity and how the mutations affect the process, Science reported. Three pathways were proposed to supply phosphorus-centered radicals from phosphorus-hydrogen donors in the active site or in bulk solution.

L-glufosinate test shows very high precision

The team put the new capability to the test by using it to make L-glufosinate, the active ingredient of a widely used broad-spectrum herbicide. CABBI reported that producing this form efficiently has long presented a chemistry challenge.

The engineered enzyme successfully produced L-glufosinate with very high precision, demonstrating that the new approach could be used to make an agriculturally important chemical, according to CABBI.

Science reported a parallel result for the related product form. Value-added organophosphorus compounds, such as the broad-spectrum herbicide glufosinate-ammonium, could be synthesized through this one-step biocatalytic transformation.

The paper describes the transformation as a one-step biocatalytic conversion of readily available alkenes and phosphorus-hydrogen compounds under mild conditions. The enzyme accommodates diverse phosphorus-hydrogen donors that pose challenges to chemical catalysis, enabling the biosynthesis of valuable phosphorus-containing scaffolds, according to Science.

CABBI quoted Zhao on the broader intent:

“Nature has evolved enzymes to carry out an enormous number of chemical reactions, but it doesn’t necessarily give us the enzymes we need to make the products we want,” said Huimin Zhao, the study’s corresponding author and CABBI Conversion Theme Lead. “Our work shows that we can engineer an existing enzyme to perform new chemistry and use that chemistry to make a commercially important product.”

That statement frames the L-glufosinate result as a demonstration case rather than an isolated synthesis. The findings, published in Science, could expand the range of useful chemicals that scientists can make using enzymes, including compounds used in agriculture and medicine, CABBI reported.

Beyond one herbicide: broad range of phosphorus products

The enzyme's usefulness was not limited to glufosinate. In tests with a variety of materials, the researchers successfully produced a range of phosphorus-containing compounds, many with similarly high precision, CABBI reported.

Science provided matching detail on substrate scope. The evolved top three enzymes showed excellent activity and stereoselectivity across a broad range of alkenes and phosphorus-hydrogen substrates, including phosphites, phosphorous acid and H-phosphinates.

The enzyme accommodates diverse phosphorus-hydrogen donors that pose challenges to chemical catalysis, according to Science. That tolerance is significant because chemical catalysis has struggled with several of those donor classes.

The results show that the approach could extend beyond a single product, giving researchers a new way to make phosphorus-containing compounds used in agriculture, medicine and other applications, CABBI reported.

The paper's rationale section likewise states that a biocatalytic method that directly incorporates common phosphorus compounds into alkenes would provide a mild and selective route to value-added, structurally diverse products. The conclusion describes the resulting system as providing access to structurally diverse chiral organophosphorus compounds with high enantioselectivity.

CABBI quoted Zhao on future use:

“Our repurposed enzymes represent a new-to-nature class of enzymes and potentially have many applications,” said Zhao. “We will further optimize and scale up the synthesis process of L-glufosinate and extend the same enzyme engineering strategy to other types of chemical reactions of both scientific significance and practical importance.”

The planned next steps described in that statement are further optimization and scale-up of the L-glufosinate synthesis process and extension of the same enzyme-engineering strategy to other reaction types.

What oxygen activation changes for biocatalysis

A central concept in the Science paper is the productive use of oxygen reactivity. Rather than treating reactive oxygen species only as damaging byproducts, the engineered system channels them into synthesis.

Mechanistic studies revealed that reactive oxygen species are converted into productive phosphorus-centered radicals, followed by radical addition and enzymatic hydrogen-atom transfer, achieving high enantioselectivity, according to the abstract in Science.

The work expands the scope of biocatalysis to programmable carbon-phosphorus bond formation and establishes a paradigm for channeling oxygen reactivity in enzymes, the abstract states.

The conclusion in Science elaborates that point. By engineering a flavin-dependent enzyme, the team redirected reactive oxygen species from damaging side reactions toward productive phosphorus-centered radical formation and used enzymatic hydrogen-atom transfer to control stereochemistry.

More broadly, the paper states that the work shows that reactive oxygen species can be used, rather than merely suppressed, to initiate enzymatic radical reactions and expand the synthetic capabilities of biocatalysis.

CABBI placed the same idea in manufacturing context. Enzymes are attractive tools for manufacturing because they can perform chemical reactions with high precision under relatively mild conditions. The new oxygen-using enzyme retains that advantage while adding a reaction that almost never happens in nature.

Support, computing and authorship

The work was supported by the U.S. Department of Energy Center for Advanced Bioenergy and Bioproducts Innovation, or CABBI, based at Illinois, CABBI reported.

The research was funded by CABBI through the U.S. Department of Energy's Office of Science, Biological and Environmental Research program under award No. DE-SC0018420, according to CABBI.

The work also used the Delta supercomputer at the National Center for Supercomputing Applications at Illinois through the National Science Foundation-supported Advanced Cyberinfrastructure Coordination Ecosystem: Services and Support program, CABBI reported.

The study's authors are Yi Zhou, Yifei Ge, Wesley Harrison and Zhao, all of the University of Illinois Urbana-Champaign, according to CABBI. The Science author listing likewise identifies Yi Zhou, Yifei Ge, Wesley Harrison and Huimin Zhao.

Contact information published with the CABBI announcement lists 1206 W. Gregory Drive (IGB), Urbana, IL 61801, telephone (217) 265-4228, and email [email protected].

Sources referenced