Engineered Enzyme Reverses Age-Related Protein Damage

Scientists have engineered an enzyme that removes an advanced glycation end product (AGE) from proteins. This type of age-related modification, which affects protein function and can trigger inflammation, has previously been considered extremely hard to reverse [1].

Can we rejuvenate proteins?

Many proteins are continually broken down and replaced. Others, especially proteins of the extracellular matrix (ECM) – the collagen-rich material surrounding cells – can remain in the body for years or decades. During that time, they gradually accumulate unwanted chemical changes.

One major category of damage is glycation, which occurs when sugars and sugar-derived reactive molecules become spontaneously attached to amino acids in proteins without the reaction being mediated by an enzyme (nonenzymatically). Over time, some of these early modifications become chemically stable AGEs [2].

AGEs can cause harm by forming crosslinks between proteins, which makes tissues stiffer, or by altering the charge or shape of individual proteins. Some AGEs can also be recognized by immune or cellular receptors, triggering inflammatory signaling. The body has natural defenses that slow the accumulation of this type of damage, but, in general, it cannot be naturally reversed.

Removing these modifications has been one of the hardest problems in geroscience. A new study published in Nature Communications by scientists from Calico, Revel Pharmaceuticals, and the University of Colorado contains what might be the first proof that it can actually be done.

Going enzyme-hunting

The researchers studied a particular AGE, N^ε-carboxymethyl-lysine (CML), which is formed on the amino acid lysine, adding a carboxymethyl group to it. CML can impair the protein’s structure and function, and it can engage RAGE, the pro-inflammatory receptor for AGEs [3].

The authors searched for an enzyme capable of cleaving CML and found that several glycine oxidases could act on free CML, as part of the CML molecule chemically resembles glycine. However, initially none could act when CML was embedded in a peptide, a chain of several amino acids that the researchers used as a model instead of much longer proteins.

A glycine oxidase from the bacterium Bacillus subtilis did indeed convert free CML into normal lysine and two byproducts, but it did so very inefficiently. The researchers then looked at several of its “relatives.” One of them was able to bind free CML much better but still could not process peptide-bound CML, suggesting that the problem was physical access rather than chemistry: when CML is part of a protein, it’s harder for a glycine oxidase to engage it.

The authors found that a particular structural element (α9 helix) likely obstructed the enzyme’s access and searched existing databases such as AlphaFold for similar enzymes without that element. After examining north of 44 thousand candidates, they found one that showed weak but detectable activity on the model peptide.

Evolution, weaponized

To advance from this starting point to an enzyme that actually works, the team harnessed evolution itself, linking the desired enzymatic reaction to bacterial survival. They first created huge libraries of mutated enzyme variants and introduced them into an E. coli strain unable to grow without lysine. The bacteria were then supplied with CML (either free or embedded in a short peptide). Only cells carrying enzyme variants that could convert the modified lysine back toward its normal form were able to obtain enough lysine to form colonies. This allowed the researchers to pick the most promising variants and improve them further.

Across five rounds of mutation and selection, the authors progressively opened the active site, improved CML binding, stabilized the enzyme, and reduced dependence on the surrounding peptide sequence, producing the final variant of the enzyme they called CMLase. In the next experiment, the enzyme removed CML from a complete protein, modified bovine serum albumin, without degrading it. The researchers then confirmed CML-removing activity for several other proteins.

CMLase, however, did not achieve universal results. Some CML sites were more susceptible to “rescue attempts” than others, depending only partly on the protein structure at that particular spot. The exact reasons why some lysines can be restored more easily than others are still unclear.

Moving to naturally glycated proteins

Up to this point in the study, CML damage to proteins was engineered. It was crucial to see whether CMLase can act on naturally aged human proteins. The enzyme indeed reduced endogenous CML in soluble lens proteins from a 64-year-old human donor. Lens crystallins are exceptionally long-lived proteins, making the lens a useful tissue for studying decades of accumulated chemical damage.

CMLase also sharply reduced CML burden in aged human artery (by 70%) and skin (by 55%). Skin CML staining after treatment fell below that observed in 31-year-old skin. Importantly, these experiments were performed in very thin sections and did not demonstrate penetration into a living artery or intact piece of skin, where delivering the enzyme to its target might prove harder.

While a big milestone and an exciting proof of concept, the study had several other important limitations. For instance, the researchers did not perform functional rescue experiments that would determine whether treated proteins or tissues functioned better, and they did not test for immunogenicity; an enzyme of bacterial origin might trigger a human immune response.

CML is also a relatively easy target compared to some other AGEs. The next step might be tackling glucosepane crosslinks: glucosepane attaches itself to two collagen molecules (or two sites on the same molecule), restricting tissue elasticity, which is a major factor in aging.

Literature

[1] Trabosh, N., Smith, J., Hsu, M. Y. H., Panja, S., Nagaraj, R., Olsson, N., … & Cravens, A. (2026). Reversal of protein chemical aging by enzymatic deglycation. Nature Communications, 17(1), 5926.

[2] Chaudhuri, J., Bains, Y., Guha, S., Kahn, A., Hall, D., Bose, N., … & Kapahi, P. (2018). The role of advanced glycation end products in aging and metabolic diseases: bridging association and causality. Cell metabolism, 28(3), 337-352.

[3] Kislinger, T., Fu, C., Huber, B., Qu, W., Taguchi, A., Du Yan, S., … & Schmidt, A. M. (1999). N ε-(carboxymethyl) lysine adducts of proteins are ligands for receptor for advanced glycation end products that activate cell signaling pathways and modulate gene expression. Journal of Biological Chemistry, 274(44), 31740-31749.