Enzyme engineers tackle protein damage from aging

As humans get older, the proteins in our bodies accrue chemical damage. Scientists at the biotechnology firm Revel Pharmaceuticals have now engineered an enzyme to reverse one type of chemical aging—glycation (Nat. Commun. 2026, DOI: 10.1038/s41467-026-75141-2). The researchers hope to develop the enzyme into a therapy to treat currently irreversible damage to long-lived proteins that accumulate in diabetes and aging.

Glycation is the chemistry behind the Maillard reaction, which contributes new flavors during cooking. But amino acids can react with the carbonyl groups of lipids or sugars in living tissues as well as in the pan. “Instead of cooking at 400 degrees [Farenheit] for an hour, we’re cooking at 98 degrees for 50, 60, 70 years,” says Aaron Cravens, Revel’s founder and CEO.

Cravens and his colleagues tackled N^ε-carboxymethyllysine (CML), the end product of a series of reactions that starts when a sugar finds a solvent-exposed lysine on a protein’s surface. Because no endogenous enzyme can reverse the modification, CML accumulates over time, eventually contributing to tissue stiffening and inflammation.

To develop the glycation-reversing enzyme, the Revel team screened bacterial glycine oxidases for variants that could hydrolyze free-floating carboxymethyllysine. But those starting proteins showed little activity against CML embedded in peptides, so the researchers used directed evolution to develop an enzyme that cleaves the sugar off CML within a peptide.

The enzyme does not react with other amino acids or glycation products; Cravens calls it “a lawnmower that only cuts dandelions.”

Revel’s researchers and collaborators confirmed that their enzyme could break down CML in highly glycated aorta tissue from a 75-year-old human donor, despite the chemical complexity of the tissue.

According to Cravens, the study is an exciting proof of concept that shows that specifically removing advanced glycation end products is possible. But the activity of Revel’s engineered enzyme activity is low. The company wants to obtain better potency before investigating it in an animal model of diabetic retinopathy.